Run Run Shaw Library
Copyright Warning Use of this thesis/dissertation/project is for the purpose of private study or scholarly research only. Users must comply with the Copyright Ordinance. Anyone who consults this thesis/dissertation/project is understood to recognise that its copyright rests with its author and that no part of it may be reproduced without the author’s prior written consent.
EXCHANGE RATE RATE EXPOSURE, CORPORATE CORPORATE HEDGING AND FIRM MARKET VALUE
CHEN QING
DOCTOR OF PHILOSOPHY CITY UNIVERSITY OF HONG KONG JANUARY 2007
CITY UNIVERSITY OF HONG KONG
香港城市大学 EXCHANGE RAT EXPOSURE, CORPORATE HEDGING AND FIRM MARKET VALUE
外汇风险暴露，套期保值和公司的市场价值 Submitted to Department of Economics and Finance
經濟及金融系 in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
哲學博士學位 by Chen Qing
陳慶 January 2007
二零零七年一月
Abstract This dissertation consists of three essays. The essays examine whether exchange rate movements affect firms’ values and the impact of hedging on exchange rate exposure and the market value of firms.
In the first essay, I empirically examine the firm-specific exchange rate exposure of Japanese multinational corporations (MNC). Exchange rate exposure is defined as the effect of exchange rate changes on the value of a firm. Although financial theory suggests that exchange rate movements should affect the value of a multinational corporation, most previous studies have failed to document significant exchange rate exposure. One major reason for the failure may be the specification of the regression model--in particular, the inappropriate selection of the exchange rate index. I study the problem by using detailed exchange rate data and propose a method to construct a firm-specific exchange rate index for each Japanese MNC bas ed on the number and location of their subsidiaries. In addition, to account for the possibility that exchange-rate crises may have a different impact on a firm from the periods of normal exchange rate fluctuations, I incorporate a crisis indicator in the standard exchange rate exposure regression model proposed by Jorion (1990). Results suggest that using the firm specific exchange rate and allowing for a crisis indicator significantly increases the evidence of exchange rate exposure of the firms in the sample.
In the second essay, I examine the role of hedging in the exchange rate risk management of Japanese multinational corporations. Using a sample of large Japanese multinationals, I examine the impact of hedging on exchange rate exposure. By using foreign currency derivatives as a proxy for financial hedging and geographical dispersion as a proxy for operational hedging, I find that both operational and financial hedging can significantly reduce a firm’s exchange rate exposure. Since the use of the foreign currency derivatives for financial hedging has grown dramatically in the past two decades, I also examine the determinants of the use of foreign currency derivatives. The results suggest that while firm size and the exposure to exchange rate through foreign sales are two important factors
determining the decision to hedge, the exposure to exchange rate through foreign sales is the sole factor affecting the extent of hedging. These results also show that firms mainly use derivatives for hedging and not for speculation.
In the third essay, I examine the use of foreign currency derivatives (FCDs) in a sample of 275 large Japanese nonfinancial firms between 2000 and 2003 and its potential impact on firm market value. Using Tobin’s Q as a proxy for firm value, I find a positive and significant relationship between the use of FCDs and firm market value in the period of Japanese yen appreciation. Specifically I find that, on average, firms that have foreign involvement and use currency derivatives have an 8.6 percent higher market value than firms that do not use currency derivatives. For firms that have no foreign involvement, but may still be exposed to exchange rate movements through export or import competition, I find a small and statistically insignificant hedging premium.
Table of Contents Table of Contents…………………………………………………………………...…i List of Figures…………………… Figures………………………………………………..…………… …………………………..……………………………………v ………………………v
List of Tables………………………………………..... ables……………………………………….....…………………………………...…….v …………………………………...…….viii iii
CHAPTER I. INTRODUCTION………………………………....………… INTRODUCTION………………………………....…………………1 ………1 1.1 What is Foreign Exchange Rate Exposure…………………….…………… Exposure…………………….…………………..……..1 ……..……..1
1.2 Corporate Hedging…………………………………………...……… Hedging…………………………………………...……………………..……6 ……………..……6
1.3 Empirical Questions of This Study………………………………………… Study……………………………………………...….……9 …...….……9
CHAPTER II. FIRM-SPECIFIC EXCHANGE RATE EXPOSURE OF JAPANESE MULTINATIONAL MULTINATIONAL CORPORATIONS
13
2.1 Introduction…………………………… Introduction…………………………………………..……………………………..… ……………..……………………………..……13 …13
2.2 Literature Review…………………………………………..…..………………………...18 Review…………………………………………..…..………………………...18
2.3 Data Description and Summary Statistics…………………………………………..……23 Statistics…………………………………………..……23
2.4 Model………………………………… Model…………………………………………………………………………… ……………………………………………….…25 …….…25
2.4.1 Jorion’s Jorion’s Model………………………………… Model……………………………………..………………………...……25 …..………………………...……25
2.4.2 Ihrig’s Ihrig’s Model………………………………… Model………………………………………..............…………….……...26 ……..............…………….……...26
2.4.3 Adjustments to Ihrig’s Ihrig’s Model…………………………......………………………27
2.4.4 Adding a Crisis Indicator…………………………………………………….…...29
2.5 The Exchange Rate Exposure of Japanese Firms…………………………………..........31
2.5.1 The Results for Jorion’s Model……………………...…………………….....…...31 Model……………………...…………………….....…...31
2.5.2 The Results R esults for Ihrig’s Ihrig’s Model……………………………………………....….…32
2.5.3 The Results of Adjustment to Ihrig’s Model…………………………………......33 Model…………………………………......33
2.5.4 Results of Adding a Crisis Indicator………………..…………….………… Indicator………………..…………….…………...….34 ...….34
2.5.5 Robustness Analyses………………………………………………………..….…36 Analyses………………………………………………………..….…36 2.6 Conclusions……………………
CHAPTER III. HEDGING IN THE RISK MANAGEMENT OF EXCHANGE RATE EXPOSURE OF JAPANESE MULTINATIONAL CORPORATIONS …………….. 51
3.1 Introduction……………………………… Introduction………………………………………………………………...………… ………………………………...…………….51 ….51
3.2 Literature Review………………………… Review………………………………………………………………….….. ……………………………………….….....55 ...55
3.2.1 Empirical Evidences of Hedging and Exchange Rate Exposure.......…………….55
3.2.2 Empirical Evidences of the Determinants of Corporate Hedging………………..58
3.3 Sample Description…………………………………………………....…………………60 Description…………………………………………………....…………………60
3.4 Hedging and Exchange Rate Exposure………………………………..…………………62
3.4.1 Methodology Framework…………………………………..….….………………62
3.4.1.1 Stage One: Estimation of Exchange Rate Exposure……..……………......62
3.4.1.2 Stage Two: Cross-Sectional Estimation………………………………...…63
3.4.2 Regression Results………………………………..…………… Results………………………………..…………………………...….64 ……………...….64 3.4.3 Robustness Check……..…………..............................……………...............……68
3.5 The Determinants of Foreign Currency Hedging………………....……………… Hedging………………....…………………...…69 …...…69
3.5.1 Factors Affecting a Firm’s Firm’s Foreign Currency Hedging…………..…………….....69
3.5.1.1 Corporate Tax Structure………………………………..……………..…...71
3.5.1.2 Transactions Costs of Financial Distress……………………………….....72
3.5.1.3 Underinvestment Costs…………………………….…………………… Costs…………………………….……………………...73 ...73
3.5.1.4 Foreign Currency Exposure………………………………………….……74 Exposure………………………………………….……74
3.5.1.5 Hedging Substitutes……………………….……………...……….… Substitutes……………………….……………...……….………76 ……76
3.5.2 Tests and Results……………………………….……………… Results……………………………….………………………..………..76 ………..………..76
3.5.2.1 Univariate Tests…………………………………………… ests………………………………………………….…...…...77 …….…...…...77
3.5.2.2 Multivariate Tests………………………….……………… ests………………………….…………………………..…..78 …………..…..78 3.6 Conclusions………………………………… Conclusions…………………………………………….……...…………………………79 ………….……...…………………………79 CHAPTER IV. FOREIGN CURRENCY HEDGING AND FIRM MARKET VALUE
92
4.1 Introduction………………………………… Introduction………………………………………………….….……………..................92 ……………….….……………..................92
4.2 Prior Hedging Theories and Empirical Evidences………………....………………… Evidences………………....…………………….97 ….97
4.3 Sample Description and Variable Definitions…………………………....………………99
4.3.1 Sample Description……………………………….………………………...….…99
4.3.2 Firm Value Value and Explanatory Variables………………………………………......102 Variables………………………………………......102 4.4 Empirical Results………………………….....................................……………........…104 4.4.1 Relation between the Use of Derivatives and Firm Value…………………....…104 Value…………………....…104
4.4.1.1 Univariate Tests…………………………………….......…… ests…………………………………….......………….…......106 …….…......106
4.4.1.2 Multivariate Tests……………………….………..……………..… ests……………………….………..……………..……......108 …......108
4.4.1.3 Robustness Tests…….…………………………………… ests…….…………………………………….………...…....112 .………...…....112
4.4.2 Relation between the Amount of Derivative Use of Firm Value.…….….…..…..112 Value.…….….…..…..112 4.5 Conclusions………………………………………… Conclusions……………………………………………………...............……………...1 …………...............……………...114 14 References…..……………………………………… References…..………………………………………………….........…………………………..1 ………….........…………………………..131 31 Appendix 2.1 Crisis Dates in Edison’s (2001) study…..……………..…………………………48 study…..……………..…………………………48
List of Tables Tables Table 2.1A: Histogram of Number of MNCs’ Foreign Subsidiaries………………………........…38
Table 2.1B: Histogram of Countries of MNCs’ Foreign Subsidiaries………….……………........38
Table 2.2: Basic Summary Statistics……………………………….…….…… Statistics……………………………….…….……………....…..……39 ………....…..……39
Table 2.3: Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Trade-Weighted Trade-Weighted Exchange-Rate Index…………………….…..……...……… Index…………………….…..……...……………....…40 ……....…40
Table 2.4: Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Subsidiary-Weighted Subsidiary-Weighted Exchange-Rate Index…………………….…………… Index…………………….…………………........41 ……........41
Table 2.5: Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using an Integrated Firm-Specific Exchange-Rate Index…………………….……… Index…………………….………………………...……42 ………………...……42
Table 2.6: Distribution of Exchange Rate Exposure with Crisis Indicators…….…………….…..43
Table 2.7: Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Trade-Weighted Trade-Weighted Exchange-Rate Index…………………………..…........…… Index…………………………..…........……………...44 ………...44
Table 2.8: Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Subsidiary-Weighted Subsidiary-Weighted Exchange-Rate Index………………………………… Index……………………………………….........45 …….........45
Table 2.9: Distribution of Exposure Coefficients (β ( β2)of 80 Japanese MNCs Using an Integrated Firm-Specific Exchange-Rate Index……………………....….……… Index……………………....….………………………...46 ………………...46
Table 3.1: Descriptive Statistics……………………………………....………........………...........81
Table 3.2: Descriptive Statistics for Foreign Exchange Rate Exposure Coefficients from 1999 to 2001………………………………….…………….............…………………………….82
Table 3.3: Relationship between Trade-Weighted Exchange Rate Exposure and Hedging from 1999 to 2001…..……………………………......………… 2001…..……………………………......……………………………..…...……83 …………………..…...……83
Table 3.4: Descriptive Statistics for Foreign Exchange Rate Exposure Coefficients from 1997 to 2001……………………….........................…………….………………..……………..84
Table 3.5: Descriptive Statistics for Firm-Specific Foreign Exchange Rate Exposure Coefficients from 1999 to 2001……………………….........…………………………......…………..85
Table 3.6: Relationship between Trade-Weighted Exchange Rate Exposure and Hedging from 1997 to 2001…..……………………………......…………… 2001…..……………………………......………………………………….........…86 …………………….........…86
Table 3.7: Relationship between Firm-Specific Exchange Rate Exposure and Hedging from 1999 to 2001……………………………………… 2001………………………………………..…......…………………………...............87 ..…......…………………………...............87
Table3.8: Relationship between Exchange Rate Exposure and Hedging Using a Probit Model………………………………………….……………………...............................88
Table 3.9: Differences Differences between Foreign Currency Hedgers and Nonhedgers…...……………..…89
Table3.10:Probit Regression Estimates of the Likelihood of Using Currency Derivatives…….................................................................................……………...........90
Table3.11:Test for Characteristics and Level of Foreign Currency Derivatives Usage............................................................................................…………….................91
Table 4.3: Summary Statistics…..………………………….....................…………….............…116 Panel A: All firms…………………………… firms……………………………...........................……………...………1 ...........................……………...………116 16 Panel B: Firms with Foreign F oreign Sales>0…………………………………………………117 Sales>0…………………………………………………117 Panel C: Firms with Foreign Sa les=0......................................……………..................117 Table 4.2: Profile of Firms’ Hedging over Time…………………......…………………………..118 Time…………………......…………………………..118
Table 4.3: Comparison of Tobin’s Q…..………………………….....................……………...…119 Q…..………………………….....................……………...…119 Panel A: Difference in Means……………………………................…………………119 Means……………………………................…………………119 Panel B: Difference in Medians.......................................................…………….........120
Table4.4:Foreign Currency Derivative Use and Firm Market Value: The Whole Period…….......................................................................................……………...........121
Table4.5:Foreign Currency Derivative Use and Firm Market Value: The Whole Period…………………………………………...................………………….......……122
Table 4.6:Foreign Currency Derivative Use and Firm Market Value: (Yen Appreciation Period)………………………………………..………….......................................……123
Table 4.7: Foreign Currency Derivative Use and Firm Market Value (Yen Depreciation Period)……………………...............……………….……………………………...…..124
Table 4.8: Foreign Currency Derivative Use and Firm Market Value for the Sample of Firm with Foreign Operations (Fixed-effects)………………………………………… (Fixed-effects)………………………………………….…..……..125 .…..……..125
Table 4.9: Foreign Currency Derivative Use and Firm Market Value for the Sample of Firm without Foreign Operations (Fixed-effects)..………...………………………….……126 (Fixed-effects)..………...………………………….……126
Table 4.10: Factors Explaining Firm Value: Value: The Whole P eriod………….......…………….........127
Table 4.11: Factors Explaining Firm Value: Value: Yen Yen Appreciation Period……….……………….....128
Table 4.12: Factors Explaining Ex plaining Firm Fir m Value: Value: Yen Yen Depreciation Depreciatio n Period…...............…129
List of Figures Figure 2.1: Distribution of Top 500 Companies with Largest Revenue by Country and Top 100 Companies with Largest Foreign F oreign Assets by Country………………………………..………….…47
Figure 4.1: The Japanese Nominal Effective Exchange Rate Index for the Period January 2000 until December 2003 ………………………………………….……………..……....……130
1
CHAPTER I
INTRODUCTION
1.1 What Is Foreign Exchange Rate Exposure
The exchange rate is the price of one country’s currency in terms of another currency and it is one of the most important prices in the open economy. It is a well-known fact that after the breakdown of the Bretton Woods System of fixed exchange rates in 1973, exchange rate volatility has increased dramatically and unpredictably. For example, exchange rates are typically four times as volatile as interest rates and ten times as volatile as inflation rates.
It is widely believed that exchange rate movements should affect the value of a firm. A multinational corporation (MNC) with subsidiaries in foreign locations, an exporter or importer concentrating on international trade or even a firm that has no direct international activities will be affected by the exchange rate fluctuations. The impact of exchange rate changes on firm value is generally determined by a firm’s net position in foreign currency. For example, if firms have a net long position in foreign currency, such as exporters and firms with foreign operations, they will benefit (suffer) from depreciation depreciation (appreciation) (appreciation) of the domestic currency. currency. For firms with a net short position in foreign currency, such as importers and domestic firms that can be affected from foreign competition, they will benefit (suffer) from an appreciation (depreciation) of the domestic currency.
2
The general concept of exchange rate exposure refers to the degree to which a company is affected by exchange rate changes. A number of theoretical papers have investigated the possible sources of exchange rate exposure. For example, Dumas (1978), Adler and Dumas (1980), and Hodder (1982) have defined economic exposure to exchange rate movements as the regression coefficient of the real value of a firm on the exchange rate across states of nature. As Adler and Dumas (1984) point out, from from the viewpoint viewpoint of U.S. U.S. investors, exposure is the regression regression coefficient coefficient or, if there are many currencies, the vector of partial regression coefficients, when an asset’s dollar price is regressed on exchange rates. From the viewpoint of the firm and its managers, exposure is a statistical quantity rather than a (projected) accounting number. The regression coefficient concept of exposure provides a single comprehensive measure that summarizes the sensitivity of the whole firm, to all the various ways in which exchange rate changes can affect it. Once exposure is so measured it can, for purposes of managerial control, be decomposed into distinct components. The decomposition can produce the counterparts of translation and transaction exposure as part of the total exposure. Neither parts should be ignored or hedged alone. Based on the impact of exchange rate changes on a firm, there are mainly two types of exchange rate exposure – accounting exposure and economic exposure.
expo sure, also known as translation exposure, refers to the changes (1) Accounting exposure
in income statement items and the book value of balance sheet assets and liabilities that are caused by an exchange rate change. More specifically, it arises from the need, for the purpose of reporting and consolidation, to convert the financial statements of foreign operations from the local currencies involved to the home
3
currency. If exchange rates have changed since the previous reporting period, this translation, or restatement, of those assets, liabilities, revenues, expenses, gains and losses that are denominated in foreign currencies will result in foreign exchange gains or losses. For example, suppose a U.S. subsidiary of a Japanese company earns a profit of 20 million dollars. This amount of money would be translated into JPY2.56 billion at an exchange rate of JPY128 per US. But if the Japanese yen strengthens to JPY112 per US, the same US dollar profit falls to JPY2.24 billion. Thus the yen representation of the profit depends on the exchange rate. The rules that govern translation are devised by an accounting association such as the Financial Accounting Standards Board. According to the Statement of Financial Accounting Standards No. 52 promulgated in 1982, firms must use the current rate method to translate foreign currency-denominated assets and liabilities into dollars. That is, all foreign currency revenue and expense items on the income statement must be translated at either the exchange rate in effect on the date these items are recognized or at an appropriately appropriately weighted average exchange rate for the period.
(2) Economic exposure emphasizes the market value of the firm and refers to the impact of exchange rate changes on the value of the firm as measured by the present value of its expected future cash flows. Economic exposure can be separated into two components: transaction exposure and operating exposure. Transaction exposure results from transactions that give rise to known, contractually binding future foreign-currency-denominated cash inflows or outflow. Since the transaction will result in a future foreign currency cash inflow or outflow, any change in the exchange rate between the time the transaction is entered into and the time it is settled in cash will lead to a change in the domestic currency amount of cash inflow
4
or outflow. Transaction exposure measures the changes in the value of outstanding foreign currency-denominated contracts that are brought about by an exchange rate change. For example, a U.S. firm that purchases Japanese goods may need Japanese yen to buy goods. While it may know exactly how many yens it will need, it doesn’t know how many dollars will be needed to be exchanged for those yens in a future time of delivery. Also, when considering a Japanese multinational corporation (MNC) that will be receiving a foreign currency, its future receivables are exposed, since it is uncertain of yens that it will obtain when exchanging the foreign currency received. Operating exposure arises because currency fluctuations can alter a company’s future revenues and costs, that is, its operating cash flows. Consequently, measuring a firm’s operating exposure requires a longer-term perspective, viewing the firm as an ongoing concern with operations whose cost and price competitiveness could be affected by exchange rate changes. Therefore, economic exposure is relevant for the short-term as well as the medium- and long-term. It is concluded that the firm faces economic exposure the moment it invests in serving a market subject to foreign competition, the sourcing of inputs, taxes, or the market’s response to prices.
Jorion (1990) is the first paper to address the problem of empirically measuring the exchange rate exposure and the determinants of the exposure at the firm level. He regresses the company’s common stock return on the rate of change in a trade-weighted exchange rate and the rate of return on the CRSP value-weighted market index. The exposure is the exchange rate regression coefficient . He uses a basic sample of 287 U.S. multinationals multinationals and monthly data from 1971 to 1987. The results show that most exposure coefficients are insignificant with a few exceptions.
5
However, the study reports significant cross-sectional differences in the exposure of U.S. multinationals.
While most researchers have focused on empirical tests of firm-level exchange rate exposure, some have also studied industry-level exchange rate exposure. Bodnar and Gentry (1993) examine industry-level exchange rate exposure for Canada, Japan, and the USA. They study a broad spectrum of industries encompassing traded and non-traded, manufacturing and service industries in the three countries. They use an augmented market model to estimate industry-level exchange rate exposure by adding a trade-weighted exchange rate to the market model for industry portfolios in each country. The three countries have similar results: between 20 and 35 per cent of industries have statistically significant exchange rate exposures and exchange rate movements help to determine industry returns at an economy-wide level. The results also suggest that the impact of exchange rate movements on industry returns is larger for Canada and Japan than for the USA. They also examine whether there is a systematic relation among industry exposures and industry activities. They decompose industry exchange rate exposure into a linear relation of a non-traded industry dummy variable, an export ratio, an import penetration ratio, a measure of the reliance on internationally-priced input, and the ratio of foreign assets to total assets. From this specification, the results for all three countries suggest that these characteristics influence an industry’s exchange rate exposure in a manner that is broadly consistent consistent with economic theory. theory.
As the existing literature examining the relationship between international stock prices and exchange rate finds little evidence of systematic systematic exchange rate exposure,
6
some researchers argue the reason may be due to restrictions imposed on the sample of data and the empirical specifications used in previous studies. Instead of using only a single country’s data, they argue to in favor of using a broad sample, including several countries. Dominguez and Tesar (2001) study a broad sample of firms in eight countries over the 1980-99 period. The specific countries in the sample are chosen on the basis of data availability and to include both OECD and developing countries. They use firm-and industry-level stock returns to test for the presence of exchange exchange rate exposure exposure in eight eight countries. They They find a surprising surprising amount of exposure, ranging from 14 percent of the sample of firms in Chile to 31 percent in Japan. At the four-digit industry level, Germany and Japan exhibit significant exposure in greater than 60 percent of the sample, and the remaining countries show between 17 and 46 percent exposure. Their results differ substantially substantially from most previous studies studies in the literature that finds little little evidence evidence of exposure. exposure.
In short, for the past decade, the results of empirical research on exchange rate exposure are mixed.
1.2 Corporate Hedging
Corporate hedging refers to the use of off-balance-sheet instruments such as forwards, futures, swaps and options to reduce the volatility of firm value. Managing foreign exchange exposure is now an important aspect of corporate risk management. Hedging a particular currency exposure means establishing an offsetting currency position such that whatever is lost or gained on the original currency exposure is exactly offset by a corresponding foreign currency gain or loss on the currency
7
hedge. In particular, a firm can hedge the exchange rate-induced volatility by (1) selling a forward contact on the foreign currency, (2) selling foreign exchange futures on the foreign currency, (3) entering into a currency swap in which it receives cash flows in domestic currency and pays cash flows in the foreign currency, (4) buying a put option on the foreign currency, or (5) writing a call option on the foreign
currency.
Alternatively,
the
firm
could
also
hedge
through
an
on-balance-sheet strategy. It can relocate production facilities abroad by setting up subsidiaries in foreign locations or fund itself in the foreign currency by foreign debt.
If financial policy is based on the objective of maximizing shareholders’ wealth, then hedging firm risks makes sense only if this leads to an increase in the value of the firm. Theories of optimal hedging demonstrate that capital market imperfections create incentives for firms to use instrument derivatives. Stulz (1984) assumes that managers decide which position the firm takes in forward contracts or foreign bonds. He examines the role of managerial compensation contracts, of exchange rate dynamics, and of hedging costs in the determinations of optimal hedging policy. The results show that firms follow an active hedging policy, which means one would not expect a firm to take systematically forward positions of opposite sign and equal in value to the promised payment in foreign currency. He also derives optimal hedging policies for risk-averse risk-averse agents agents in the presence presence of uncertainty uncertainty about future commodity commodity prices and of holding costs costs of hedge hedge positions. positions.
Smith and Stulz (1985) also present an analysis of hedging behavior of firms. Rather than assuming that the firm is risk averse, they follow modern finance theory
8
and assume that incentives exist within the contracting process to maximize the market value of the firm. They show that a value-maximizing firm can hedge for three reasons: (1) taxes, (2) costs of financial distress, and (3) managerial risk aversion.
Froot and Stein (1993) develop a model to show how a firm’s optimal hedging strategy, in term of both the amount of hedging and the instruments used, depends on the nature of its investment and financial opportunities. They show that hedging mitigates the underinvestment problem that would result when cash flow is volatile and access to external financing is costly.
Empirical examination of hedging theories has been difficult due to the general unavailability of data on hedging activities. It is only recently that corporations have been required to disclose in footnotes of their annual reports, reports, the notional amount of derivatives they are using. In the absence of reported information on derivatives, most of the earlier studies used survey data to examine why firms use derivatives. For example, Nancy, Smith and Smithson (1993) use survey data to compare characteristics of firms that actively hedge with those that do not. Their results show that high R&D firms are more likely to hedge; firms with more leveraged capital structures hedge more and high dividend-paying firms are more likely to hedge.
More recent empirical studies have employed a binary variable indicating whether a firm uses derivatives or not. Specifically, Geczy, Minton, and Schrand (1997) examine currency hedging activities for a sample of Fortune 500 firms and find that
9
the use of currency derivatives is directly related to the amount of research and development (R&D) expenditures, which is consistent with the use of hedging to reduce underinvestment problems. Tufano (1996) examines hedging activities in the gold-mining industry and finds that the use of commodity derivatives is negatively related to the number of options and positively related to the value of stock held by managers and directors. This evidence is consistent with theories of managerial risk aversion. Haushalter (2000) examines the commodity hedging activities of firms in the oil and gas industry and finds evidence consistent with theories of transaction costs of financial distress.
While there are an increasing number of studies providing evidence that firms that hedge fit the profile of one theory or another, no study had addressed the question of whether there is a direct relation between hedging and firm value until using Tobin’s Q as a proxy for firm market value, they find a positive relation between firm value and the use of foreign currency derivatives. In particular, the hedging premium is statistically and economically significant for firms with exposure to exchange rates and is on average 4.87% of firm value.
1.3 Empirical Questions of This Study
In this thesis, I study several empirical topics about the foreign exchange rate exposure and hedging activities of Japanese firms by examining the relation between changes in exchange rates and changes in firm value; the determinants of exchange rate exposure and the relationship between hedging and firm market value. In particular, particular, the research research questions that that I examine examine are:
10
(1)
Do exchange rate movements affect firm value? Is exchange rate exposure
significant for the Japanese multinational corporations?
(2)
Can financial and operational hedging reduce the exchange rate exposure
of Japanese multinational corporations and what determines the use of foreign currency derivatives by corporations?
(3)
Does hedging affect firms’ market value?
The specific objectives of each hypothesis are elaborated below: Despite corporate views and portfolio managers’ beliefs that exchange rate movements affect firms’ values, earlier empirical work was unable to document significant exchange rate exposure for the vast majority of U.S. multinationals (Jorion (1990)), or largest U.S. exporters (Amihud (1993)). In the next chapter, I argue that the lack of significant evidence of exchange rate exposure documented in the previous literature is due to the reason that the selection of exchange rate index does not appropriately capture the firm’s sensitivity to exchange rate changes. For example, most studies use trade-weighted exchange rate indices with weights deriving from national trade figures with major trading partners by assuming that changes in the trade-weighted exchange rate affect multinational corporations uniformly. However, the international linkage of multinational corporations, which have operations in different countries, may not uniformly relate to these national trade figures. Therefore, I construct a firm-specific exchange rate for each individual firm based on both the number and location of its subsidiaries as well as the general
11
trade figures of Japan. In addition, as small movements in the exchange rate may not be very significant to a firm’s balance sheet while large movements movements in the exchange rate, e.g. crisis period, may be costly to a firm, I add an early warning system crisis indicator to capture the impact of crisis on exchange rate exposure. By doing this, I find more firms have significant exposures than previous studies.
In the third chapter, I examine the impact of financial and operational hedging on exchange rate exposure. Because of the unavailability of derivatives data, many previous papers have tried to uncover the impact of financial derivatives use on exchange rate exposure by using a series of variables that existing studies find to be important in explaining firms hedging policy to proxy for financial hedging. Although most studies have examined firms’ financial hedging, few studies have examined firms’ operational hedging. Since diversifying operations in different countries is also a very important way to eliminate exchange rate risk, I incorporate proxies for financial and operational operational hedging as determinants determinants of exchange rate exposure with control of firms’ foreign operations in a model. I find that both financial and operational hedging can significantly reduce firms’ exchange rate exposure. In addition, I also examine the determinants of the use of derivatives. Using a two-stage Cragg framework, I find that firms with larger size and exposure to exchange rates through foreign sales are most likely to use currency derivatives. I also identify the ratio of foreign sales to total sales as the sole determinant of the extent of hedging. This result also suggests that firm mainly use derivatives for hedging rather than speculation.
12
Finally, I examine the effectiveness of derivatives use on firm market value. Theories of optimal hedging suggest that hedging can increase firm value by reducing capital market imperfections. If firms are exposed to the exchange rate changes and use currency derivatives to reduce this exposure, are these hedging activities effective in reducing their exchange rate exposure and consequently, increasing firm market value? I first examine whether firms that use foreign currency derivatives have a higher market value than those that do not use foreign currency derivatives. I find that there exists a positive and statistically significant relationship between firm’s firm’s market market value and the use of of foreign currency currency derivatives derivatives in in the period period of Japanese yen appreciation. I also test whether the extent of hedging using foreign currency derivatives is related to firm market value. I find that the level of hedging of firms is marginally related to their market values, which suggests that the firms are not hedged optimally. It also implies that firms use different hedge ratios according to their firm-specific characteristics.
13
CHAPTER II
FIRM-SPECIFIC EXCHANGE RATE EXPOSURE OF JAPANESE MULTINATIONAL CORPORATIONS
2.1 Introduction
It is widely believed that exchange rate changes should affect the cash flows of a firm and hence a firm’s value. A multinational corporation (MNC) is a company engaged in producing and selling goods or services in more than one country with price denominated denominated with respective local currencies. currencies. Since the breakdown of the Bretton Woods system of the fixed exchange rate in 1973, exchange rate volatility has increased dramatically. Exchange rate movements affect both the price and quantities of the inputs and outputs of multinational corporations, which lead to the change of their competitive advantage or disadvantage in the global market, therefore, their cash flows change, and so do their firm values. While finance theory strongly supports that firm value is sensitive to exchange rate movements, existing empirical support is fragile. Most empirical studies have failed to find a strong relationship between exchange rate changes and a firm’s stock market return, which is a proxy for the change in firm value. The early paper of Jorion (1990) and later studies of Amihud (1994) and Bartov and Bodnar (1994) have so far documented a weak link between contemporaneous exchange rate fluctuations and stock returns of U.S. multinational corporations. Similar international studies that investigate several countries, such as Bodnar and Gentry (1993), Dominguez and Tesar (2001) and Doidge, Griffin and Williamson (2002), also find few firms with significant
14
exchange rate exposure.
There are several potential reasons for the weak evidence of a relationship between firm value and changes in exchange rate. One possible reason is that prior research has mainly focused on studying the exchange rate exposure of U.S. multinationals and it is not clear whether these empirical results are merely a spurious correlation that may not be confirmed in other countries.
The second major reason previous literature often mentioned is that the selection of the exchange rate index does not appropriately capture the firm’s sensitivity to exchange rate changes. For example, most existing studies use trade-weighted exchange rate indices with weights deriving from national trade figures with major trading partners. By doing this, they generally assume that changes in the trade-weighted value of the domestic currency affect multinational corporations uniformly (Shin and Soenen, 1999). It is easily understood that the international linkage of individual firms may not uniformly relate to these national trade figures, especially in the case of large multinational corporations, which have operations in different countries. The exchange rates that have great impacts on their revenues should be the relative exchange rate that affects their foreign revenues, e.g. bilateral exchange rate between home currency and the currencies of the countries their subsidiaries operate in. Therefore, the trade-weighted exchange rate indices may not appropriately appropriately capture the exchange rate exposure faced by individual firm.
Finally, the third reason is that firms may shield themselves against exchange rate risk through the use of various hedging instruments. Corporate on- and off-balance
15
sheet hedging activities can reduce the firm’s exposure. Empirical studies of the incentives to hedge for multinational firms also support this conjecture. The magnitude of the exposure can be explained by variables that proxy for a firm’s hedging incentives (Chow and Chen, 1998; He and Ng, 1998).
In this chapter, I address the first two issues mentioned above. The third issue will be discussed in the next chapter. chapter. In the first place, I examine the exchange rate exposure of Japanese MNCs for the period from 1995 to 2001. In 1996, Fortune Magazine reported that most of the multinational corporations, by revenues, are United States and Japanese. From Fig. 2.1, we can see that 162 of the top 500 multinationals with largest revenues from the United States, and 126 from Japan in 1996. When measured by foreign assets, the distribution of largest companies looks very much the same. Most of the top 100 companies with largest foreign assets are from U.S. and Japan. Thus, in order to test the robustness of prior findings on exchange rate exposure outside the US environment, Japan is a good option. Also, many Japanese manufacturing sectors are heavily globally oriented and are more susceptible to unanticipated fluctuations in foreign exchange rates.
The second contribution of this chapter is the use of firm-specific exchange rates and crisis indicators to capture the impact of crises. Ideally, we would like to have detailed information on the relevant exchange rates for each firm. Ihrig (2001) proposes a method method to construct construct a firm-speci firm-specific fic exchange rate for each individual firm on the base of the number and locations of a multinational’s subsidiaries and uses these exchange rates in the simple Jorion model for a sample of 226 US multinationals. She finds that the number of US multinationals with significant
16
exposure rises from 10 percent by using trade-weighted exchange rates to 16 percent by using firm-specific firm-specific exchange rates. Based on her results, she argues that by using firm-specific exchange rates, exchange rate exposure is much more prevalent than the previous studies suggest.
Based on a sample of Japanese MNCs, I apply Ihrig’s method to create a firm-specific exchange rate for each Japanese MNC. In contrast to Ihrig’s results, I find that fewer firms are significantly exposed to the firm-specific exchange rates than to the trade-weighted exchange rate. The reason for such results may be that the method Ihrig proposed mainly captures the impact of exchange rate changes on the foreign subsidiaries of the MNCs. As a matter of fact, the MNCs are exposed to exchange-rate risk through two major channels. On the one hand, the domestic subsidiaries of a MNC are exposed to exchange-rate risk because the value of foreign input and output, in terms of the domestic currency, changes when the exchange rate changes. On the other hand, a MNC has a large portion of assets and liabilities abroad. The foreign subsidiaries of a MNC with productions and sales in foreign countries can also increase or decrease a firm’s exposure. In order to combine the two channels through which the exchange rates have effects on the value of the MNCs, I adjust Ihrig’s method to construct a firm-specific exchange rate based on both the number and location of their foreign subsidiaries subsidiaries as well as the general trade figures of Japan. By using the new firm-specific exchange rate index, I find that the number of Japanese MNCs with significant exposure rises to 31.3 percent, compared compared to 30 percent percent using the Ihrig’ Ihrig’ss method, which which is the same as using using the trade-weighted exchange rate index. Since the three indices provide similar results, I test the correlation between every pair of exchange rate indices. The
17
correlation coefficients are all above 97%, which means that these exchange rate indices are highly correlated. Therefore, although these exchange rate indices conceptually capture different aspects of firms’ exposure, the estimation of exposures provides similar results.
In addition, Chen and So (2002) show that increases in exchange rate variability during the crisis are associated with statistically significant increases in stock return volatility for the multinationals. Ihrig (2001) suggests that small movements in the exchange rate may not be very significant to a firm’s balance sheet while large movements in the exchange rate may be costly to a firm. Thus the firms’ returns may be affected from the crisis period of drastic changes in the exchange rate differently differently from normal period of exchange rate movements. The financial crises that erupted in East Asia in the second half of 1997 after the devaluation of the Thai baht have a widespread and devastating impact of the economy of most Asian countries. As most of Japanese MNCs’ subsidiaries are located in Asia, the impact of crises may be an important factor in estimating exposure. Therefore, I follow Ihrig (2001) and add an early warning system crisis indicator to capture the impact of crises on exposure. By allowing stock returns to be affected by exchange rate changes differently in crisis period than in normal period, I find the number of significant significant firms increases increases substantially. My results show that more than a half of my sample has significant exchange rate exposure.
The rest of this chapter is organized as follows. Section 2 gives a brief review of the theoretical and empirical literature of exchange rate exposure. Section 3 describes the data and summary statistics. Section 4 describes the research
18
methodology to estimate exchange rate exposure, including Jorion’s (1990) model, Ihrig’s (2001) model and the adjustments to Ihrig’s model. Section 5 presents the empirical results and compares the different results by using different methodology. Section 6 concludes this chapter.
2.2 Literature Review
A number of theoretical papers have investigated the exchange rate exposure. Dumas (1978), Adler and Dumas (1980), and Hodder (1982) have defined economic exposure to exchange rate movement as the regression coefficient of the real value of the firm on the exchange rate across states of nature. As Adler and Dumas (1984) point out, from from the viewpoint viewpoint of U.S. investors, investors, exposure exposure is the regression coefficient coefficient or, if there are many currencies, the vector of partial regression coefficients, when an asset’s dollar price is regressed on exchange rates.
None of the theoretical theoretical studies mentioned mentioned above have provided empirical estimates of their models until Jorion (1990), which first addressed the problem of empirically measuring the exchange rate exposure and its determinants at firm level. Jorion (1990) regresses the company’s common stock return on the rate of change in a trade-weighted exchange rate and the rate of return on the CRSP value-weighted market index. He finds very weak support for exchange rate exposure. Using a basic sample of 287 U.S. multinationals and monthly data for a period from 1971 to 1987, he only finds 15 out 287 multinational corporations show significant exposure at 5% significant level. His study also reports significant cross-sectional differences in the exposure of U.S. multinationals.
19
Following Jorion (1990), Amihud (1994) tests the exchange rate exposure of a portfolio of 32 large U.S. exporting firms between 1982 and 1988. He finds that the exposure coefficient is not significantly different from zero, which suggests that there is no contemporaneous impact of exchange rate movements on the value of exporters. However, using quarterly data, he finds a lagged effect of exchange rate changes on firm value. Bartov and Bodnar (1994) also examine contemporaneous and lagged effects for a sample of 208 US firms that report foreign currency adjustments in their past annual financial statements and have exposures of the same sign. Like Jorion (1990) and Amihud (1994), they find no significant contemporaneous exchange rate exposure. However, they find significant exchange rate exposure when regressing contemporaneous stock returns against a lagged change in the value of the dollar.
Choi and Prasad (1995) examine the exchange risk sensitivity of 409 U.S. multinational firms during the 1978-89 period. In contrast to previous studies, they find that exchange rate fluctuations do affect firm value. More specifically, they find 61 firms demonstrate significant exposure at the 10%-level and approximately sixty percent of firms with significant exchange rate exposure exposure gain from a depreciation depreciation of the dollar. They also find that cross-sectional differences in exchange risk sensitivity are linked to key firm-specific operational variables, such as sales and assets. When they aggregate the data into 20 SIC-based industry groups, they observe some cross-sectional and inter-temporal variation in the exchange risk coefficients. Their subperiod analysis also reveals that more firms with significant exchange rate exposure during the dollar depreciation period as compared to the appreciation
20
period.
While most researchers have focused on empirical tests of firm-level exchange rate exposure, some have also studied industry-level exchange rate exposure. Bodnar and Gentry (1993) examine industry-level exchange rate exposures for Canada, Japan, and the USA. They study a broad spectrum of industries encompassing traded and non-trade, manufacturing and service industries in the three countries. They use an augmented market model to estimate industry-level exchange rate exposure by adding in a trade-weighted exchange rate to the market model for industry portfolios in each country. The three countries have similar results: between 20 and 35 per cent of industries have statistically significant exchange rate exposures and exchange rate movements help to determine industry returns at an economy-wide level. Their results also suggest that the impact of exchange rate movements on industry returns is larger for Canada and Japan than for the USA. They also examine whether there is a systematic relation among industry exposures and industry activities. They decomposed industry exchange rate exposure into a linear relation of a non-traded industry dummy variable, an export ratio, an import penetration ratio, a measure of the reliance on internationally-priced input, and the ratio of foreign assets to total assets. From this specification, the results for all three countries suggest that these characteristics influence an industry’s exchange rate exposure in a manner that is broadly consistent consistent with economic theory. theory.
The aforementioned studies have primarily focus on U.S. firms and found low level of significant exchange rate exposure. It seems reasonable to look at other countries. Several studies investigate firms in small and open economies and find
21
they are more sensitive to exchange rate movements than firms in larger and less open economies. Glaum, Brunner and Himmel (2000) examine the economic exposure of German corporations to changes in the DM/US-dollar exchange rate. They find that German firms are significantly exposed to changes in the DM/US-dollar rate, 31 percent of 71 firms are significantly exposed in full period from 1974 to 1997. Nydahl (2001) investigates the exchange rate exposure for a sample of Swedish firms. Using weekly data, he finds that about 26 percent of the 47 firms in the sample are significantly exposed to exchange rate changes over the period 1990-1997. Jong, Ligterink and Macrae (2002) examine the relationship relationship between exchange rate changes and stock returns for a sample of Dutch firms over 1994-1998. More than 50% of the firms in their sample are significantly exposed to exchange rate risk. Moreover, they also find that all firms with significant exchange rate exposure benefit from a depreciation of the Dutch guider relative to a trade-weighted exchange rate index.
Instead of using only a single country’s data, some studies argue to use a broader sample comprising several countries. Dominguez and Tesar (2001) study a sample of eight industrialized and developing countries over the 1980-1999 period. They find a statistically significant level of exposure in the pooled eight-country sample: 23 percent of firms and 40 percent of industries industries are exposed to at least one of the trade-weighted exchange rate, the US dollar and the currency of the country’s major trading partner. Doidge, Griffin and Williamson (2002) construct a comprehensive international database of over 17,000 non-financial firms from 18 countries between 1975 and 1999. Consistent with previous studies, they find that more firms are exposed to exchange rate movements than can be attributed to chance. Moreover,
22
they find that exchange rates cannot explain a large portion of the variation in the firm value.
A few studies have examined the exchange rate exposure of Japanese firms in the recent years. He and Ng (1998) examine a sample of 171 multinational firms that have export ratios of at least 10 percent in the sample period from 1979 to 1993. They find that about 25 percent of the 171 firms have significant positive exposure coefficients and about 2 percent have negative coefficients, which suggest that most of the firms in the sample benefited from the depreciation of the yen. Chow and Chen (1998) use 1,110 Japanese firms included in the PACAP database for a period from 1975 to 1992. In contrast to the findings of the He and Ng (1998), they find that more than 80% of Japanese firms are negatively exposed, indicating that yen depreciation hurts firms. Moreover, there is also some indication that the number of firms with significant exposure increases with time horizon.
This chapter also focuses on Japanese firms for the period between 1995 and 2001. Unlike previous studies on Japan, this study examines large multinationals during a period which includes includes the Asian financial financial crisis. crisis. The exchange rates rates of several Asian Asian countries have fluctuated dramatically after the 1997 financial crisis. The change from highly stable exchange rate regimes to floating regimes was associated with a sharp increase in exchange rate variability. Using a firm-specific exchange rate and also a crisis indicator, I aim to shed some light on the measurement and significance of the exchange rate exposure of Japanese multinational corporations.
23
2.3 Data Description and Summary Statistics Statistics
I focus on a sample of large Japanese multinational corporations listed in The Directory of Multinationals, which are the world’s largest corporations with consolidated sales of over US$1.5 billion and overseas sales in excess of US$750 million during 1999/2000. The directory excludes firms predominantly or exclusively in banking, insurance and commodity broking, which means that my sample only includes nonfinancial firms. The locations and number of principal subsidiaries for each MNC are also obtained from this source. Only those with full information on the locations of principal subsidiaries are chosen. The number and locations of subsidiaries is very dispersed (see Table 2.1). 9 MNCs of the sample have more than one hundred subsidiaries, while 6 MNCs of the sample have less than five subsidiaries. Regarding the number of countries in which the MNC subsidiaries are located, only 1 MNC has subsidiaries located in more than forty countries and 18 MNCs have subsidiaries located in less than five countries. The list of MNCs obtained from The Directory of Multinationals is then matched with stock price data for seven years between January 1995 and December 2001. In total, 80 MNCs meet this requirement.
End-of-month stock price data are retrieved from the Datastream database. I first use a trade-weighted Japanese yen exchange rate index as a reference point. I construct the exchange rate index by taking the weighted average of nine normalized bilateral nominal nominal exchange rates 1 between Japan and its major trading partners (base period December 1994=100), defined as Japanese yen per unit of foreign currencies for the United States, Canada, Australia, France, Germany, Italy, Netherlands, Switzerland and the United Kingdom. The weights are calculated based on each
24
country’s proportion of the nine countries’ total trade with Japan. This information is obtained from the Direction of Trade Statistics Yearbook . In order to construct a firm-specific exchange rate, I collect all relevant end-of-month bilateral nominal exchange rates between Japan and the countries the subsidiaries of the MNCs operate in from the IMF’s International Financial Statistics. For the market rate of return, I use the Nikkei 225 index obtained from the Datastream database.
The sample period selection procedure is an important aspect in the analysis of exchange rate exposure as it can strongly influence the results. In this study, my sample period is from 1995 to 2001. I limit my time horizon to 1995-2001 for the following reasons. First, I choose to focus on a crisis period, as in 1997, the Asian Financial Crisis occurred. A period from 1995 to 2001 includes both the pre-crisis period as well as the post-crisis post-crisis period. Second, earlier studies (e.g., Jorion, 1990) have shown that exchange rate exposure can be time varying. Measuring exposure over longer time horizons may lead to biased results. Therefore, I restrict my sample period to a seven-year horizon. horizon.
The sample contains monthly data on 80 MNCs that I use in the estimation of exchange rate exposure over a period from 1995 to 2001, so that there are 6720 firm-year observations in the sample. Table 2.2 provides summary statistics of all the variables I use in the analysis.
1
I use nominal exchange rate here is because the nominal exchange rates are much more volatile than prices at frequencies less than one year. In this regards, regards, the correlation between nominal exchange rate and real exchange rate is extremely high. Therefore, it is very likely that the results I report using nominal exchange rate would extend to real r eal exchange rates.
25
2.4 Model
2.4.1 Jorion’ Jorio n’ss Model
Jorion’s (1990) model is the benchmark model to estimate exchange rate exposure. For each MNC, exchange rate exposure is estimated by regressing the MNC’s stock return on the market portfolio return and exchange rate change. Jorion’s model is given by
R itit = β0i +β1i R mt mt +β2i et +εt,
(2.1)
where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the trade-weighted exchange rate in month t. β2i measures exchange rate exposure of firm i after taking into account the overall market’s exposure to currency fluctuations. If we reject the hypothesis that β2i is zero, we find evidence of exchange rate exposure. If we do indeed find evidence of exchange rate exposure, this indicates the existence of some form of market inefficiency. A rejection of no exposure suggests either that investors are not fully diversifying their portfolios—so that exchange rate risk remains—or that firms themselves are not fully hedging their exchange risks. The sign of the exchange rate coefficient can either be positive or negative depending on the net exposed asset and liability positions of the firm. For example, firms that use their foreign subsidiaries principally to import finished goods and sell them in domestic countries will benefit from an appreciation of its domestic currencies. This benefit arises due to a reduction in the domestic value of foreign costs. In contrast, firms that
26
incur most of their production in the domestic countries and sell in foreign markets have exposed foreign sales revenue. They will find that their products become less competitive in overseas markets and their foreign sales revenues decline with any appreciation of the domestic currency. Similarly, MNCs with net exposed assets abroad will lose with a strengthening domestic currency, while firms with net exposed liabilities gain.
2.4.2 Ihrig’ Ih rig’ss Model
More generally, we should expect variation in individual firm to various exchange rates. Any test that restricts the measurement of exposure to one exchange rate (whether it be a trade-weighted rate or a bilateral rate) is likely to be biased. One possible research strategy to mitigate this problem is to create a firm specific exchange rate. The difficulty with this approach is that it is not clear on what basis these exchange rates should be chosen. Firms may hedge exposure to the more obvious currencies, for example, currencies of the countries where they export or import good, but remain exposed to currencies of countries with whom their goods compete on world markets (with whom they do no direct business, while have indirect impact of their exchange rates change). Following the standard Jorion (1990) model of exchange rate exposure, Ihrig’s model uses a firm-specific exchange rate. Based on the number and locations of the subsidiaries of each MNC, Ihrig’s model proposes an approach approach to create a firm-specific firm-specific exchange exchange rate. With With the available data data of locations and numbers of MNCs, for each MNC, the exchange rate is a weighted average of the home currency bilateral exchange rates for where its subsidiaries are
27
located. The weight given to a country’s exchange rate is associated with the number of subsidiaries located in that country relative to the total number of foreign subsidiaries of the MNC. Following Ihirg’s (2001) model, for each month, the MNC-specific subsidiary-weighted exchange rate, sub-ER is defined as:
N
sub − ERi = ∑ j =1
# subsidiaries j
ER j , Total # foreignsubsi
where N is the total number of countries that MNC i operates in, ER is the bilateral exchange rate between the MNC’s home currency and the currency of country j. In this study, ER is the normalized bilateral exchange rate between the Japanese yen and the currency of country j (base period December 1994=100). The denominator of the above formula is the number of the total foreign subsidiaries of an MNC. If the MNC operates in only one country other than the home country, then sub-ER is just a bilateral bilateral exchange exchange rate.
With other variables defined as before, Ihrig model may be written as i R itit = β0i + β1i R mt mt + β2i e t + εt,
t=1,· · ·T
(2.2)
where eit is the percentage change in the MNC-specific exchange rate sub-ER in month t. The other variables in model (2.2) are defined the same as the model (2.1).
2.4.3 Adjustments to Ihrig’s Ihrig’s model
As far as an MNC is concerned, it has both domestic subsidiaries and foreign
28
subsidiaries. The domestic and foreign subsidiaries of an MNC are affected by exchange rates through possible two different channels. The domestic subsidiaries of an MNC are exposed to exchange rate risk because the value of foreign input and output, in term of the domestic currency, changes when the exchange rate changes. Therefore, their international trade pattern may follow the national trade pattern. A trade-weighted exchange rate may be more proper to capture the effect of exchange rate change on firm value. On the other hand, MNCs have a large portion of assets and liabilities abroad, which often produce and sell products in various counties with prices denominated denominated in corresponding corresponding local currency. currency. Therefore, Therefore, an MNC’s MNC’s foreign revenue will be mostly affected by the bilateral exchange rate between the home currency and the currency of the locations of its subsidiaries. Using Ihrig’s approach to construct a subsidiary-weighted exchange rate may only capture one aspect of the exchange rate exposure, which is the foreign activity of each MNC. In order to account for these two major effects, I combine the trade-weighted exchange rate and subsidiary-weighted exchange rate together to create an integrated exchange rate index for each MNC as follows 2:
int − ER
i
=
Total # foreignsubsi Total # subsi
N
# subsidiaries j
∑ Total # foreignsubs ER j + j =1
i
Total # domesticsubsi Total # subsi
.
2
I want to test the overall exposure to exchange rate, so I integrated the two exchange rate indices into one. I also ran the regression by separating the two exchange rate indices and the results worsen significantly, only 4 percent of the sample have significant exposures. I then test the correlation between the trade-weighted exchange rate index and the subsidiary-weighted exchange rate index, they are highly correlated at above 97%. Therefore, separating of them might cause the problem of multicollinearity that makes explanatory variables insignificant.
ERtw
29
The first part of the int-ER is the sub-ER defined earlier weighted by the proportion proportion of foreign subsidiaries to total subsidiaries. subsidiaries. The second second part of the int-ER is as the trade-weighted exchange rate weighted by the proportion of domestic subsidiaries to total subsidiaries.
Therefore the adjusted Ihrig-model is as follows: R itit = β0i + β1i R mt mt + β2i *Δint-ER it it+ εt ,
t=1,· · ·T
(2.3)
where Δint-ER itit is percentage change in the integrated MNC-specific exchange rate in month t. The other variables in model (2.3) are defined the same as defined earlier.
2.4.4 Adding a Crisis Indicator
It is intuitively quite plausible that a crisis period differs from a period of normal exchange rate fluctuations in many aspects. As a result, it is important to allow for a crisis factor in estimating exchange rate exposure. The early warning system literature suggests various measures of exchange rate pressure. Frankel and Rose (1996), Kaminsky, Liondo, and Reinhart (1998), and Kamin, Schindler and Samuel (2001) create monthly country indicators that take on values of zero or one. A value of zero indicates that there is no exchange rate pressure in the country for that date. A value of one indicates that there was abnormal exchange rate pressure (i.e. a crisis) based on the author’s criteria criteria for crises. Edison (2001) provides a good overview of the abovementioned studies on the early warning systems and extends the former crisis indicators to 1990s. In order to capture the effect of financial crises, I take the
30
aforesaid crisis indicator and add it to the adjusted Ihrig-model. That is, I allow returns to be affected by exchange rate fluctuations during crisis periods differently from months of normal exchange rate fluctuations.
The crisis indicator added into the adjusted Jorion-model is based on the Edison (2001) indicator. Appendix 1 shows the crisis dates from Edison’s (2001) study. I code a country’s crisis indicator as one for that date and zero otherwise for the period from 1995 to 1999. I take these indicators and create a MNC-specific indicator using the same method as for sub-ER. The subsidiary weighted crisis indicator, sub-I, is defined as:
N
# subsidiaries j
j =1
Total# subsi
sub − I i = ∑
I j ,
where N is the total number of foreign countries that MNC i operates in, and I is the crisis indicator of country j. Sub-I can take on a value between zero and one. If none (all) of the countries where the MNC has operations in has a crisis, then sub-I is zero (one).
Introducing the crisis indicator to capture the effect of a crisis on exposure in the adjusted Ihrig model, the model is adjusted to:
i R itit = β0i + β1i R mt mt + (β2i + β3i I t)*Δint-ER i+ εt
t=1,· · ·T
(2.4)
where Ii is an exchange rate crisis indictor function sub-I. When none of an MNC’s
31
subsidiaries is located in crisis countries, the indicator function is zero and model (2.4) reduces to model (2.3). The indicator function used in the analysis takes on a value between zero and one, where the value indicates the proportion of its subsidiaries in crisis countries. Now firm i’s exposure at date t is given by β3i + β4iIit, exposure varies through time as I i fluctuates between zero and one. The other variables are the defined the same as in model (2.3).
2.5 The Exchange Rate Exposure of Japanese firms
In this section I discuss the results of my estimation of different models of the exchange rate exposure of Japanese MNCs. In accordance with previous researches in this area, I first examine the standard Jorion model.
2.5.1
The Results for Jorion’s Jorion’s Model
Equation (2.1) is estimated using a trade-weighted exchange rate over a period from January 1995 to December 2001. Table 2.3 presents the results of regressing the monthly stock returns on the value-weighted market portfolio index return and the change of trade-weighted Japanese Yen.
Table 2.3 shows that 25(31.3%) of the 80 Japanese MNCs are significantly exposed to exchange rate risk at the 10% significant level and 18 (22.5%) firms are significantly exposed to exchange rate risk at the 5% significant level. Compared with previous studies of US MNCs, I find that Japanese firms are more frequently exposed to exchange rate change. For example, Jorion (1990) finds only 5% of the
32
US firms to be exposed. My results are even more substantial than the findings of He and Ng(1998), which also find about 25 percent of a sample of 171 Japanese multinationals with significant exchange rate exposure. The reason I find more significant firms may be that my sample period includes the Asian Financial Crisis, during which the exchange rate fluctuated dramatically. The firms may be more vulnerable to exchange rate change in such a time period.
The median (mean) exposure coefficient is 0.13(0.17), indicating that the average Japanese MNC in my sample gains 0.13% (0.17%) in value in case the Japanese yen depreciates 1%. Furthermore, of the total 25 MNCs which are significantly exposed, 19 MNCs have a positive exposure coefficient. In line with He and Ng (43 positive and 2 negative), my results indicate that on average, Japanese MNCs benefit from a depreciation of the home currency.
2.5.2
The Results for Ihrig’s Ihrig’s Model
An important reason that most previous literature cannot find significant exchange rate exposure is the use of a broad exchange rate index. Ihrig (2001) proposes a method to create a firm-specific exchange rate based on the number and locations of subsidiaries of firms. For each Japanese MNC listed in The Directory of Multinationals, I follow Ihrig’s approach to create a firm-specific exchange rate which is a weighted average of the Japanese Yen bilateral exchange rates for the foreign countries that its subsidiaries are located in. Equation (2.2) is estimated using a firm-specific exchange rate for each MNC. The results are presented in Table 2.4.
33
Table 2.4 shows that when using a firm-specific exchange rate, 24 firms out of 80 (30%) are significantly exposed to exchange rate risk at 10%-level and 20(25%) firms are significantly exposed to exchange rate risk at 5% significant level. Different from Ihrig(2001)’s results, which find that the number of U.S. MNC with significant exposure rises from 10 percent in the standard Jorion estimate to 16 percent in her model, my results show that the number of Japanese MNCs’ with significant exposure in the Ihrig model (30%) is less than that in Jorion model (31.3%).
As far as the magnitude of the exposure is concerned, the median (mean) exposure coefficient is 0.10(0.15). Both the mean and the median are smaller than the trade-weighted exposure, which are 0.13 and 0.17 respectively. Furthermore, out of the 24 MNCs with significant exposures, 18 have a positive exposure coefficient. By proportion, proportion, 75 percent of total significant significant MNCs MNCs benefit from from a depreciation depreciation of home home currency, slightly less than that using the Jorion model, which finds 76 percent of total MNCs which are significantly exposed benefit from a depreciation of Japanese Yen.
2.5.3
The Results of Adjustment to Ihrig’s Ihrig’s Model
As mentioned above, in order to measure the exchange rate exposure in broader framework, I propose a way to combine the two approaches together to construct a firm-specific exchange rate for each MNC. The results are reported in Table 2.5.
Table 2.5 shows that 25(31.3%) MNCs have significant exposures. The magnitude
34
of the exposure is similar to the results estimated by Jorion’s model. The median (mean) is 0.12(0.17).
Jorion’s model and Ihrig’s model conceptually capture different aspects of exchange rate exposure of a firm. In particular, the trade-weighted exchange rate index focuses on the foreign activities of domestic subsidiaries of a firm, while the subsidiary-weighted exchange rate index focuses on the foreign subsidiaries’ activities of a firm. The integrated exchange rate index combines these two effects to capture the overall exchange rate exposure of a firm. However, the empirical results show that these three methods provide similar results. I also test the correlation between the trade-weighted trade-weighted index and the subsidiary-weighted subsidiary-weighted index as well as the correlation between the trade-weighted index and the integrated index. They are all highly correlated with correlation coefficient above 97%. Therefore, although these methods conceptually capture different aspects of exchange rate exposure, they provide similar similar results because of the high correlations correlations in the particular sample used in this study.
2.5.4
Results of Adding a Crisis Indicator
As Ihrig (2001) suggests, small movements and large movements in exchange rate may have different impact on a firm’s balance sheet, thus the exchange rate exposure. Lee (2003) also finds that there are structural changes in the relationship between the exchange rate return and the stock return before and after the economic crisis. Therefore the crisis indicator adjusted Ihrig model with a crisis indicator is estimated. The results are presented in the Table 2.6.
35
Table 2.6 reports the regression coefficients of the equation (2.4) on the left panels and the combined exposure including the effect of a crisis on the right panels. Due to the lack of data of crisis indicators, my sample period changes to 1995-1999. The number of firms also drops from 80 to 69, as 11 MNCs did not have any one of their subsidiaries located in a crisis country in the 1995-1999 period.
According to column 1 table 2.6, β2 is the standard exposure estimate as my previous estimation. The median value is 0.13, which is approximately the same as that of the adjusted Ihrig model. About 34.7 percent of the sample is significantly exposed to exchange rate fluctuation in normal period. Column 2 reports β3, which captures the impact of financial crisis on a MNC’s exchange rate exposure. The median value of β3 is –9.67, the negative coefficient means that a crisis has a negative effect on a firm’s stock return and it also suggests that a depreciation of the Japanese yen benefit firms less during a crisis period than it does in normal period. In order to calculate the additional effect of exchange rate crises on firm i’s return, I multiply β3i by Ii. Column 3 reports the combined exposure distribution across the 69 firms. Exposure for firm i at time t is calculated as β2i+β3i Iit. Exposure is significant if either β2 and/or β3 is significant. Thus, 52.2 percent of the sample is significantly exposed to exchange rate movements, which is considerably higher than 31.3 percent of the sample that has significant exposure estimated by the previous model without a crisis indicator. From the analysis above, one might expect that the exchange rate movements have different impact on firms’ stock return during normal period and crisis period. Adding a crisis indicator is better to capture the effect of a crisis on the exchange rate exposure of firms.
36
2.5.5
Robustness Analyses
Chow and Chen (1998) point out that exchange rate exposure increases with horizons. Studies to date have looked at many different return horizons. Although most of the literature uses monthly data, some studies also use shorter horizons, such as weekly or bi-weekly and try to capture the short-term effect of exchange rate impact. In this section, I test the sensitivity of my analysis using weekly data. Table_2.7, Table_2.8 and Table_2.9 represent the results of weekly exchange rate exposure using the Jorion’s model, the Ihrig’s model and the adjustment to Ihrig’s model respectively. The results confirm Chow and Chen’s (1998) findings that exchange rate exposure increases with horizons. More firms are significantly exposed to exchange rate movements using weekly data than those using monthly data in all the three models. More specifically, when estimating Jorion’s (1990) model, 42.5% of firms have significant exchange rate exposure when using weekly data, while only 31.3% of firms have significant exchange rate exposure when using monthly data. The results are similar when testing the other two models. As for the magnitude of the exchange rate exposure, the weekly exchange rate exposure tends to have smaller magnitude, the mean exposure is 0.15, ranging from -0.48 to 0.97. The mean monthly exchange rate exposure is 0.17, ranging from -1.06 to 1.47.
2.6 Conclusions
I examine the relationship between exchange rate changes and stock returns for a sample of Japanese MNCs for a period between 1995 and 2001. In addition to estimating the standard Jorion-model, I also apply Ihrig’s method to create a
37
firm-specific exchange rate for each Japanese MNC. I then combine the two methods to construct an integrated firm-specific exchange rate index, considering both the domestic and the foreign subsidiaries of the MNCs. Although the trade-weighted exchange rate index and the firm-specific exchange rate index conceptually capture different aspects of the exchange rate exposure, the empirical results are similar due to the reason that these exchange rate indices are highly correlated in my sample.
Furthermore, some studies suggest that exchange rate movements during normal verses crisis periods may have a differential impact on a firm’s exchange rate exposure. To test this hypothesis, I add a crisis indicator into the adjusted Ihrig-model. The number of significant MNCs rises dramatically from 31.3 percent to 52.2 percent. These results suggest that emphasize the importance of using firm-specific information in measuring the exchange rate exposure of Japanese MNCs, we can better capture the exchange rate exposure.
38
Table 2.1A-Histogram of Number of MNCs’ Foreign Subsidiaries
# of subsidiary subsidiar y
Frequency
0-5
6
6-10
13
11-15
3
16-20
3
21-25
4
26-50
26
51-75
12
76-100
3
>100
9
Table 2.1B-Histogram of Countries of MNCs’ Foreign Subsidiaries
# of countries
Frequency
0-5
18
6-10
8
11-15
16
16-20
13
21-25
9
26-30
9
31-35
4
36-40
2
>40
1
39
Table 2.2 Basic Summary Statistics
N R i 6720 R m 84 Δtw-ER 84 Δsub-ER 6720 Δint-ER 6720 sub-I 4140
Mean 0.0005 -0.006 0.002 0.001
Median 0.003 0.003 0.005 0.002
Min -0.570 -0.139 -0.136 -0.149 -0.14 9
Max 1.153 0.149 0.105 0.134
Std Dev 0.122 0.060 0.035 0.035
0.002 0.004
0.003 0
-0.138 -0.13 8 0
0.121 0.4
0.034 0.022
R i is the monthly return on firm’s stock R m is monthly return on Nikkei 225 index Δtw-ER is the percentage change in the trade-weighted exchange rate Δsub-ER is the percentage change in the subsidiary-weighted subsidiary-weighted exchange rate Δint-ER is the percentage change in the integrated exchange rate sub-I is the subsidiary-weighted crisis indicator
40
Table 2.3 Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Trade-Weighted Exchange-Rate Index
The table reports the quartiles for β2 from the following model, R itit = β0i +β1i R mt mt +β2i et +εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the trade-weighted exchangerate index in month t, β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-1.06
First quartile
-0.08
Median
0.13
Third quartile
0.45
Maximum
1.47
Firms with significant exposure at the 5%-level 10%-level
18 25
Firms with positive significant exposure at 10%-level
19
Firms with negative significant exposure at 10%-level
6
41
Table 2.4 Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Subsidiary-Weighted Exchange-Rate Index
The table reports the quartiles for β2 from the following model, i R itit = β0i + β1i R mt mt + β2i e t + εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e it is percentage change in the MNC-specific exchange rate sub-ER in month t. β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-1.02
First quartile
-0.07
Median
0.11
Third quartile
0.43
Maximum
1.46
Firms with significant exposure at the 5%-level 10%-level
20 24
Firms with positive significant exposure at 10%-level
18
Firms with negative significant exposure at 10%-level
6
42
Table 2.5 Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using an Integrated Firm-Specific Exchange-Rate Index
The table reports the quartiles for β2 from the following model, R itit = β0i + β1i R mt mt + β2i *Δint-ER it it+ εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, Δ int-ER itit is percentage change in the integrated MNC-specific exchange rate in month t. β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-1.06
First quartile
-0.08
Median
0.12
Third quartile
0.44
Maximum
1.45
Firms with significant exposure at the 5%-level 10%-level
18 25
Firms with positive significant exposure at 10%-level
18
Firms with negative significant exposure at 10%-level
7
43
Table 2.6 Distribution of Exchange Rate Exposure with Crisis Indicators
The table reports the estimated results from the following model, i R itit = β0i + β1i R mt mt + (β2i + β3i I t)* Δint-ER i+ εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, Δ int-ER itit is percentage change in the integrated MNC-specific exchange rate in month t. I i is an exchange rate crisis indictor function sub-I. The data represent observations of 69 firms from 1995 to 1999.
β3
Exposure β2 + β3 I (3) All
Coefficients (1)
(2)
β2
Minimum
-1.37
-2558.71
-5.23
First Quartile
-0.19
-29.28
-0.25
Median
0.13
-9.67
0.02
Third Quartile
0.50
17.50
0.52
Max
1.66
308.01
1.79
No. of firms firms Significant at 10% Median Exposure
24
19
36
n/a
n/a
0.02
Mean Exposure
n/a
n/a
0.04
No. of MNCs
69
69
69
44
Table 2.7 Distribution of Exposure Coefficients (β ( β2) of 80 Japanese MNCs Using a Trade-Weighted Exchange-Rate Index
The table reports the quartiles for β2 from the following model, R itit = β0i +β1i R mt mt +β2i et +εt Where R itit is the weekly return on the firm i’s stock at time t, R mt is the weekly return on market portfolio, e t is percentage change in the trade-weighted exchange- rate index in week t, β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-0.47
First quartile
-0.06
Median
0.10
Third quartile
0.28
Maximum
0.95
Firms with significant exposure at the 5%-level 10%-level
23 34
Firms with positive significant exposure at 10%-level
26
Firms with negative significant exposure at 10%-level
8
45
Table 2.8 Distribution of Exposure Coefficients ( β2) of 80 Japanese MNCs Using a Subsidiary-Weighted Exchange-Rate Index
The table reports the quartiles for β2 from the following model, i R itit = β0i + β1i R mt mt + β2i e t + εt Where R itit is the weekly return on the firm i’s stock at time t, R mt is the weekly return on market portfolio, e it is percentage change in the MNC-specific exchange rate sub-ER in week t. β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-0.45
First quartile
-0.07
Median
0.13
Third quartile
0.27
Maximum
0.97
Firms with significant exposure at the 5%-level 10%-level
23 31
Firms with positive significant exposure at 10%-level
22
Firms with negative significant exposure at 10%-level
9
46
Table 2.9 Distribution of Exposure Coefficients ( β2) of 80 Japanese MNCs Using an Integrated Firm-Specific Exchange-Rate Index
The table reports the quartiles for β2 from the following model, R itit = β0i + β1i R mt mt + β2i *Δint-ER it it+ εt Where R itit is the weekly return on the firm i’s stock at time t, R mt is the weekly return on market portfolio, Δint-ER itit is percentage change in the integrated MNC-specific exchange rate in week t. β2i measures exchange rate exposure of firm i. The data represent observations of 80 firms from 1995 to 2001.
β2i
Minimum
-0.48
First quartile
-0.07
Median
0.12
Third quartile
0.28
Maximum
0.97
Firms with significant exposure at the 5%-level 10%-level
23 34
Firms with positive significant exposure at 10%-level
23
Firms with negative significant exposure at 10%-level
11
47
Fig 2.1 Distribution of Top 500 Companies with Largest Revenue by Country and Top 100 Companies with Largest Foreign Assets by Country
180 162 160 140
126
120
United Stated Japan
100
France 80
Germany Britain
60 42
41
40
34
20 0 United Stated
Japan
France
Germany
Britain
35 30 30 25
USA 18
20
Japan United Kingdom
15
11
11
10
9
5 0 U SA
Japan
United Kingdom
France
Germany
France Germany
48
Appendix 2.1:
Crisis dates in Edison’s (2001) study
Crisis Date Exchange Rate Pressure Index Country
Original KLR Dates
Replication of KLR Crisis Date 1970-1995
Replication of KLR Crisis Date 1970-1999
1. Argentine
Jun-70 Jun-75 Feb-81 Jul-82 Sep-86 Apr-89 Feb-90 Nov-82 Nov-83 Sep-85 Feb-83 Nov-86 Jul-89 Nov-90 Oct-91
Mar-75 Jul-82 Apr-89 Dec-89
Mar-75 Jul-82 Apr-89 Dec-89
Nov-82 Nov-83 Sep-85 Sep-82 Mar-90 Nov-90 Oct-91
Nov-82 Nov-83 Sep-85
2. Bolivia
3. Brazil
4. Colombia
Mar-83 Feb-85
Jan-85
5. Denmark
May-71 Jun-73 Nov-79 Sep-92 Jun-73 Oct-92 Nov-91 Sep-92 Nov-78 Apr-83 Sep-86
Jun-73 Apr-77 Oct-82 Sep-92 Jun-73 Apr-77 Oct-82 Sep-92 Nov-78 Apr-83 Sep-86
Nov-74 Nov-77 Oct-83 Jul-84
Nov-74 Nov-77 Oct-83 Jul-84 Jun-71 Dec-74 Jan-80
6. Finland
7. Indonesia
8. Israel
9. Korea
10.Malaysia
Jul-75
Sep-82 Mar-90 Nov-90 Sep-91 Jan-99 Jan-85 Sep-97 Sep-98 Jul-99 Jun-73 Nov-79 Aug-93 Jun-73 Apr-77 Oct-82 Sep-92 Nov-78 Apr-83 Sep-86 Dec-97 Nov-74 Nov-77 Oct-85 Jul-84
Jan-80 Nov-97
Nov-73 Nov-78 Mar-80 Feb-85 Dec-92 Jul-97
49
Exchange Rate Pressure Index
Table 3 continued Country
Original KLR Dates
Replication of KLR Crisis Date 1970-1995
Replication of KLR Crisis Date 1970-1999
11. Mexico
Sep-76 Feb-82 Dec-82 Dec-94 Jun-76
Sep-76 Feb-82 Dec-82 Dec-94 Jun-76
Sep-76 Feb-82 Dec-82 Dec-94 Jun-76
Oct-87 Feb-70 Oct-83 Jun-84
Oct-87 Oct-83 Jun-84 Feb-86
Oct-87 Oct-83 Jun-84 Feb-86 Dec-97 Mar-76 Feb-77 Jun-82 Jun-83 Sep-75 Jul-84 Aug-85 May-86 Jun-98 Jun-73 Feb-76 Jul-77 Mar-83 Sep-92 Mar-95 Aug-77 Sep-81 Oct-82 Nov-92
12. Peru
13.Philippines
14. Portugal
Mar-76 Feb-77 Jun-82 Jun-83 Sep-75 Jul-84 Aug-85 May-86
15. South Africa
16. Spain
Feb-76 Jul-77 Dec-82 Sep-92 May-93
17. Sweden
Aug-77 Sep-81 Oct-82 Nov-92 Nov-78 Jul-81 Nov-84
18. Thailand
19. Turkey
20. Uruguay 21. Venezuela
22. Chile
Aug-70 Jan-80 Mar-94 Dec-71 Oct-82 Feb-84 Dec-86 Mar-89 May-94 Dec-71 Aug-72 Oct-73 Dec-74 Jan-76 Aug-82 Sep-84
Jun-73 Feb-76 Jul-77 Mar-83 Sep-92 Mar-95 Aug-77 Sep-81 Oct-82 Nov-92 Nov-78 Jul-81 Nov-84 Aug-70 Jan-80 Mar-94 Mar-72 Nov-82 Mar-84 Dec-86 Mar-89 Jul-71 Sep-72 May-73 Dec-74 Jul-85
Jul-97 Aug-70 Jan-80 Mar-94 Mar-72 Nov-82 Mar-84 Dec-86 Mar-89 Dec-95 Jul-71 Sep-72 May-73 Dec-74 Jul-85
50
Exchange Rate Pressure Index
Table 3 continued Country
Original KLR Dates
Replication of KLR Crisis Date 1970-1995
Replication of KLR Crisis Date 1970-1999
23. Norway
Jun-73 Feb-78 May-86 Dec-92
Jun-73 Feb-78 May-86 Nov-92 Jan-83 Oct-85
Jul-91 Mar-93 May
Apr-91 Mar-93 May-72 Oct-82 Nov-77 Dec-70 Jul-75
Jun-73 Feb-78 May-86 Nov-92 Jan-83 Sep-84 Oct-85 Jul-91
24. Greece
25. India 26. Pakistan 27. Sri Lanka 28. Singapore
Dec-70 Jul-75 Nov-78 Mar-80
May-72 Oct-82 Nov-77 Nov-77 No data
Mar-80 Dec-97
Notes: The KLR crises dates are derived from calculating an index of weighted average of exchange rate changes and reserve losses. The reading of this index that was 2.5 standard deviations or more above the mean was cataloged as a crisis.
51
CHAPTER III
HEDGING IN THE RISK MANAGEMENT OF EXCHANGE RATE EXPOSURE OF JAPANESE MULTINATIONAL CORPORATIONS
3.1 Introduction Introduction
Exchange rate exposure is becoming an increasingly important issue in international financial management. Exchange rate fluctuations are a major source of risk for multinational corporations. To mitigate exchange rate uncertainty, it has been claimed that hedging, not only financial hedging, but also operational hedging can protect companies from unexpected movements of exchange rates (Allayannis, Ihrig, and Weston (2000)). Corporate hedging refers to the use of off-balance-sheet instruments and on-balance-sheet strategy (Nance (1993)). Firms that are exposed to exchange rate risk can actively reduce currency risk by two ways. On the one hand, firms can hedge exchange rate risk by financial hedging, entering into derivatives contracts, such as forwards, futures or options. The hedging with these off-balance-sheet contracts has a negative effect on exchange rate exposure. On the other hand, firms can reduce their exposure through on-balance-sheet hedging. Firms can move factories to its sales markets and as a result revenues and expenses are incurred in the same currency. This is also called operational hedging, which has a negative effect on exchange rate exposure.
52
According to the classic Modigliani and Miller theory, risk management is irrelevant to the firm. Investors can hedge on their own by holding well-diversified portfolios. portfolios. Optimal hedging hedging theories, however, however, suggest that a firm’s firm’s hedging activities activities affect the extent to which the firm is exposed to exchange rate movements. In a world of market imperfections, firms have incentives to employ derivative instruments to hedge against exchange rate risk. Extensive use of derivatives as hedging instruments should diminish a firm’s exposure. Because of the unavailability of derivative data, many previous papers have tried to uncover the impact of financial derivatives’ use on exchange rate exposure by using a series of variables that existing studies find to be important in explaining a firm’s hedging policy to proxy for financial hedging. For example, He and Ng (1998) relate the exchange rate exposure to the level of its export ratio and variables that are proxies for its hedging needs. Using a sample of 171 Japanese multinationals, they find that highly leveraged firms, or firms with low liquidity, tend to eliminate exchange rate exposure. SFAS105 requires firms to report information on financial instruments with off-balance sheet risk (e.g., futures, forwards, options and swaps) for fiscal year ending after June 15, 1990. In particular, firm mush report the face, contract or notional amount of the financial instrument, and information on credit and market risk of those instruments, the cash requirements, and the related accounting policy. With the availability of derivative data, Allayannis and Ofek (2001) examine the effect of the use of derivatives on foreign exchange rate exposure of a sample of 378 US non-financial firms in 1993. Their results show that the use of derivatives significantly reduces the exposure of the sample firms to exchange rate risk.
53
For the past few decades, the effectiveness of currency hedging has been studies by a number of papers providing empirical results of both US and European countries. In addition to Allayannis and Ofek’s study indicating that the use of foreign currency derivatives can significantly reduce the exchange rate exposure of US firms, Jong, Ligterink and Macrae (2002) take a look at Dutch firms. They find that financial hedging by using derivatives has no significant effects on the exchange rate exposure, which is in contrast to their hypothesis. In line with theory, they find that exchange rate exposure is significantly reduced through operational hedging by producing in factories abroad. Nydahl (2001) has also studied the Swedish firms and found that the foreign exchange exposure is negatively related to the use of currency derivatives. These abovementioned results are only applied to the US and European countries, few studies thus far have examined the hedging effects of firms in Asian countries. As in the first chapter, we can see that the Japan has extraordinary high portion of MNCs that are significantly exposed to exchange rate risk, does hedging can effectively reduce these exchange rate exposures? In this chapter, we will examine the hedging effectiveness on Japanese MNCs.
While most studies have examined firms’ financial hedging, few studies have examined firms’ operational hedging. Diversifying operations in different countries is a very important way to eliminate exchange rate risk. However, it is difficult to measure operational hedging. Jong, Ligterink and Macrae (2002) gather the operational hedging information through a tailor-made questionnaire by asking the firms directly to indicate what percentage of total exposure the firm hedges through its operation. They find that operational hedging has significant negative effect on exchange rate exposure for a sample of Dutch firms. Allayannis, Ihrig, and Weston
54
(2000) develop several measures of geographic dispersion that proxy for the use of operational hedging strategies, they find that geographic dispersion through the location of subsidiaries across multiple countries or regions does not reduce exchange rate exposure.
This chapter investigates both financial and operational hedging in the risk management strategies of multinational corporations. Using a sample of large Japanese multinationals, I examine the effectiveness of financial hedging and operational hedging, how each strategy contributes to the risk management of eliminating exchange rate exposure and improving firm value. I use a model that incorporates proxies for financial hedging and operational hedging as determinants of exchange rate exposure with control of firms’ foreign operations. I find that both financial hedging and operational hedging can significantly reduce firms’ exchange rate exposure. Moreover, as the use of the foreign currency derivatives for financial hedging has grown dramatically in the past two decades, I take a further step to examine the determinants of the use of derivatives. I use a two-stage Cragg framework to examine what determines corporations’ level of derivative use. This two-stage process allows to examine separately a firm’s decision to hedge from its decision of how much to hedge. I find that firms with larger size and exposure to exchange rates through foreign sales are more likely to use currency derivatives. While these tests reveal the factors that prompt corporations to hedge, they do not answer the question of what determines the extent of hedging. In the second stage of the estimation, I use the fair value of currency derivatives for those firms that chose to hedge to identify the factors that are important determinants of the amount of hedging, given that a firm has decided to hedge. The results suggest that the exposure factor
55
(foreign sales/total sales) is the sole determinant of the level of hedging. Therefore, the results suggest that a firm’s decision to hedge and the extent of hedging is primarily primarily affected by its exposure to foreign currency movements movements through foreign sales. This also implies
that firms mainly use derivatives for hedging and not for
speculation.
The remainder of the chapter is organized as follows: Section 2 discusses the existing empirical research on hedging strategies in risk management of exchange rate exposure and the use of derivatives. Section 3 describes the sample and data. Section 4 examines the impact of financial and operational hedging on exchange rate exposure. Section 5 presents the tests on the use of derivatives; and Section 6 concludes.
3.2 Literature Review
3.2.1 Empirical Evidence on Hedging and Exchange Rate Exposure
Although financial theory strongly supports that exchange rate movement should affect the value of a multinational corporation, most previous studies have failed to document significant exchange rate exposure. One major reason suggested by Bartov et al. (1996) is that the limited success of empirical studies in documenting significant exchange rate exposure is partly because firms have effectively managed their exposures. However, until the beginning of 1990s, data on hedging were not available, so researchers could not directly examine the effect of hedging on exchange rate exposure, they could only test the hedging activities by using survey data or variables that are regarded as for hedging.
56
He and Ng (1998) investigate the determinants of exposure for Japanese firms, but include no direct measure of hedging activity. Instead, the authors define an indirect measure by choosing variables that are hypothesized to explain hedging. Their results for dividends, the quick ratio, book-to-market and leverage confirm expectations. Unfortunately, the indirect variables may partly represent direct determinants of exposure, which makes the findings difficult to interpret.
Nydahl (2001) investigates investigates the determinants determinants of exposure for Swedish firms and includes survey data for a single year on derivative usage (futures, swaps and options) and foreign debt. Unfortunately, only the currency positions in monetary values are reported. For example, off-balance and on-balance instruments are included on an equal basis, while the sensitivity to currency changes of off-balance instruments is expected to be much higher. For on-balance hedging on the asset side, the fraction of foreign wages is included. The variable is strongly biased towards service-oriented firms and against manufacturers. The results show that hedging significantly reduces the exposure.
Allayannis and Ofek (2001) use detailed derivative data to examine whether firms use foreign currency derivatives for hedging or for speculative purposes. Using a sample of S&P 500 nonfinancial firms for 1993, they test the relationship between derivative use and exchange rate exposure and find evidence that firms use currency derivatives as hedging to significantly reduce the exchange rate exposure. The results indicate a strong negative relation between derivatives and exposure.
Makar and Huffman (2001) examine the lagged exchange rate exposure in relation
57
to the use of short-term foreign exchange derivatives (FXD) for US multinationals. They provide evidence that the lagged firm value effects of changes in exchange rates are particular to low FXD users, and that the magnitude of such currency exposure deceases monotonically across FXD user groups. Their results indicate that cross-sectional differences in the magnitude of lagged currency exposure are inversely related to FXD use.
While most studies use samples of US firms, Nguyen and Faff (2003) investigate the role of foreign currency derivatives (FCD) in alleviating foreign exchange rate exposure for a sample of Australian firms. Their results support the view that FCDs are used to hedge existing exchange rate exposures and that exchange rate exposure of shorter horizons appears to be negatively related to a firm’s price earning ratio (proxying for growth opportunities) while the exposure of longer horizons is positively positively related to a firm’s firm’s liquidity, liquidity, supporting the view that that liquidity is a substitute substitute for hedging.
The above studies all examine the relationship between financial hedging and exchange rate exposure, few studies examine the impact of operational hedging on exchange rate exposure. Allayannis, Ihrig and Weston (2000) investigate both financial and operational exchange rate risk management strategies of multinational firms. Using a sample of US multinational nonfinancial firms between 1996-1998, they find that operational hedging is not an effective substitute for financial risk management. However, they find that the more geographically disperse a firm is, the more likely it is to use financial hedges. Their end result is that operational hedging strategies benefit shareholders only when used in combination with financial hedging
58
strategies.
3.2.2 Empirical Evidence on the Determinants of Corporate Hedging
There are several theories of optimal hedging, most of which rely on the introduction of some friction to the classic Modigliani and Miller paradigm. For example, in Stulz(1984), corporate hedging arises as a result of managerial risk aversion. In Smith and Stulz (1985), the structure of the tax code (i.e., if taxes are progressive) progressive) or the transaction transaction costs of financial distress could prompt firms to undertake hedging activities. In Froot et al. (1993), hedging mitigates the underinvestment problem that would result when cash flow is volatile and access to external financing is costly. Finally, in Demarzo and Duffie (1995), corporate hedging is optimal when managers have private information on the firm’s expected payoff despite shareholders’ ability to hedge by themselves.
Empirical examination of hedging theories has been difficult due to the general unavailability of data on hedging activities. Until the beginning of 1990s, a firm’s exact position in derivatives was privately held information, and was considered a very important of strategic competitiveness. It is only recently that corporations have been required to disclose in footnotes in their annual reports, the fair value of derivatives they are using.
In the absence of reported information on derivatives, most of the earlier studies used survey data to examine why firms use derivatives. For example, Nance et al.(1993) uses survey data on Fortune 500 firms’ use of forwards, futures, swaps, and
59
options and finds that firms that hedged firms face more convex tax functions, have less coverage of fixed claims, are larger, and have more growth options in their investment opportunity set.
Recent, studies have focused on the type of hedging (commodity, interest rate, or currency), recognizing that different factors can be important for each type of hedging. In particular, Geczy et al. (1997) examines currency hedging activities for a sample of Fortune 500 firms. He finds that firms’ use of currency derivatives is positively related to the amount of R&D expenditure, which is consistent with the use of hedging to reduce underinvestment; size, which is consistent with fixed-costs of hedging explanations; and exposure factors (foreign income and trade). Tufano (1996) examines commodity hedging activities in the gold mining industry. He finds that firms’ use of commodity derivatives is negatively related to the number of options their managers and directors hold, and positively related to the value of their stock holdings, evidence consistent with theories of managerial risk aversion (e.g., Stulz, 1984). Haushalter (2000) examines the hedging activities of oil and gas producers. He finds that total debt ratio is positively related to the percentage of production hedge, which is consistent with theories of transaction costs of financial distress. Visvanathan (1998) examines the use of interest rate swaps by S&P 500 nonfinancial firms, and also finds evidence supporting theories of transaction costs of financial distress (e.g. Smith and Stulz, 1985). Finally, Mian (1996) investigates all three types of hedging activities for a sample of 3022 firms and finds mixed evidence for theories of managerial risk aversion and taxes and evidence that strongly supports the hypothesis that hedging activities exhibit economies of scales (i.e., that larger firms hedge more). All of the studies cited above have examined which factors could be associated with
60
the probability that a firm hedges. With the exception of Thufano (1996) and Haushalter (2000), who also examine the level of hedging in a particular industry (gold, and oil and gas respectively), no other study in the general derivatives area has looked for the factors that are associated with the extent of hedging.
3.3. Sample Description
I construct my sample of large Japanese multinational corporations from The Directory of Multinationals, which are the world’s largest corporations with consolidated sales of over US$1.5 billion and overseas sales in excess of US$750 million during 1999/2000. The directory excludes firms predominantly or exclusively in banking, insurance and commodity broking, which means that my sample only includes nonfinancial firms. The locations and number of principal subsidiaries for each MNC are also obtained from this source. Only those with full information on the locations of principal subsidiaries are chosen. On the basis of the assumption that firms with foreign exchange exposure, I further restrict my sample with firms that have foreign sales to total sales ratio reported in the Worldscope database. This approach produces 75 firms in the final samples.
SFAS 133 requires firms to report information on financial instruments with off-balance-sheet risk. In particular, firms mush report the face, contract, or notional amount of the financial instrument, and information on the credit and market risk of those instruments, the cash requirements, and the related accounting policy.
I obtain
data on year-end notional value of forward contracts reported in the footnotes of the annual reports of all the multinationals in the sample in 2000. My sample’s notional
61
values of foreign currency derivatives also include foreign currency options. However, these values do not include foreign currency swaps, because currency swaps are mainly used in conjunction with foreign debt by translating foreign debt into domestic liability.
To examine the impact of financial and operational hedging that might have on exchange rate exposure, I first estimate exposure for the period from January 1999 to December 2001. Consistent with Allayannis and Ofek (2001), I believe that a three-year return period surrounding the year in which derivative data are disclosed and operational hedging proxies are constructed is appropriate to measure the contemporaneous impact of financial hedging and operational hedging on a firm’s exchange rate exposure. However, I will also use a five-year return period as a robustness check.
To examine the determinants of derivative use, I employ a series of financial valuables which are obtained from Worldscope database. In case of missing data I use data from annual reports.
Table 3.1 presents summary statistics of the main variables that I use in this chapter. My total sample consists of 75 firms with a mean value of total sales of Japanese yen 2560 billion. Out of the 75 firms in our sample, 59 firms (78.7%) report the use of foreign currency derivatives. Data regarding individual firms’ stock prices, market index as measured by Nikkei 225 Index are obtained from Datastream. Exchange rate and trade volume data to construct trade-weighted exchange rate are obtained from IFS.
62
3.4 Hedging and Exchange Rate Exposure
In this section, I examine the impact of the use of financial and operational hedging on the exchange rate exposure. I hypothesize that both financial hedging and operational hedging have a negative impact on exchange rate exposure. I assume firms use foreign derivatives for financial hedging -- that is, we should expect that the use of derivatives would reduce exchange rate exposure. The operational hedging proxy is constructed constructed by a measure of geographic dispersion, dispersion, which is proposed by Allayannis (2002). A priori, we expect that the dispersion of a firm’s subsidiaries across countries is likely to reduce exchange rate exposure. I first present my methodology, and then the tests and results.
3.4.1 Methodology Framework
3.4.1.1 Stage one: Estimation of Exchange Rate Exposure
Jorion’s (1990) model is the benchmark model to estimate exchange rate exposure. For each MNC, exchange rate exposure is estimated by regressing the MNC’s stock return on the market portfolio return and exchange rate change. Consistent with previous literature, I estimate estimate the foreign exchange rate exposure ( β2i) for the period from January 1999 to December 2001 using the following model:
R itit = β0i +β1i R mt mt +β2i et +εt
(3.1)
63
where R itit is the monthly return on the firm i’s stock at time t; R mt is the monthly return on Nikkei 225 market index; e t is percentage change in the trade-weighted exchange rate in month t, measured in Japanese yen per one unit of foreign currency, β2i measures exchange rate exposure of firm i.
Since the trade-weighted exchange rate index is extensively used in estimation of exchange rate exposure in literature, I also use it to estimate the exchange rate in this stage. In addition to the trade-weighted exchange rate, I also use a firm specific exchange to estimate the exchange rate exposure.
I use model (3.1) to estimate exchange rate exposure in my sample for the three years surrounding 2000 (1999-2001). A firm is defined as having positive (negative) exchange rate exposure if its stock return increases (decreases) as the trade-weighted exchange rate index depreciates (appreciates). My first stage estimation results show that out of the 75 sample firms, 28 firms have negative exposure while 47 firms have positive exposure, which means more firms in my sample benefit from the depreciation of Japanese yen.
3.4.1.2 Stage Two: Cross-Sectional Estimation
According to He and Ng (1998), a firm’s exchange rate exposure is determined by its level of foreign operations and hedging policy. There are two kinds of hedging activities firms may employ to mitigate exchange rate exposure. One is financial hedging, which is often proxied by the use of foreign currency derivatives; the other is operational hedging. I use a dispersion factor to proxy for operational hedging. To
64
construct the geographic dispersion factor, I follow the measure that is proposed by Allayannis, Ihrig and Weston (2000). Measure of geographic dispersion for firm i is calculated as:
K
Dispersioni = 1 − ∑ [ j =1
# subs j Total # subsi
]2
(3.2)
This measure is constructed with the Hishman-Herfindahl concentration index over all the countries that a firm operates in. Where K is the total number of countries that firm i operate in. When a firm has subsidiaries in many countries, the second part of the measure is close to zero, which makes the dispersion factor close to one. In the meanwhile, if a firm has subsidiaries only in one country, the value of the dispersion factor will be zero.
Once the exchange rate exposure is estimated, the relationship between hedging and exposure is tested using the cross-sectional regression framework by Allayannis and Ofek (2001). However, in Allayannis and Ofek’s framework, they only include variable proxied for financial hedging in the regression. In my estimation, I also include an operational hedging variable into the regression to examine both the impact of financial hedging and operation hedging on the exchange rate exposure with control of the foreign operations. The equation is shown below:
β2i=α1i +α2i (FS/TS)i +α3i(FCD/TA) i +α4iDispersion i+ηi
i=1,…N,
where β2i is a firm’s exchange rate exposure estimated in model (1)
(3.3)
65
(FS/TS)i (FCD/TA) i
is a firm’s ratio of foreign sales to total sales, is a firm’s ratio of
foreign currency derivatives to total asset,
Dispersion i is defined in equation (2).
This equation incorporates a firm’s financial hedging and operational hedging as well as its foreign operations. These three factors that are believed to simultaneously determine the level of exchange rate exposure. If firms employ financial hedging (as measured by the use of FCD) to hedge exchange rate exposure and assuming FCDs are effective in hedging the exposure, it is expected that the more FCD a firm uses, the smaller exposure it has. On the other hand, firms may employ operational hedging strategy by locating subsidiaries in many different countries. By this way, the more disperse a firm’s subsidiaries’ locations, the less it is exposed to exchange rate risk. Therefore, both hedging measures are expected to reduce exchange rate exposure. In particular, particular, the use of hedging strategies strategies should decrease exposure for firms with positive exposure and increase (decrease in absolute value) exposure for firms with negative exposure. The ratio of foreign sales to total sales, as evidences from previous studies, I expect foreign sales to increase a firm’s exposure, because foreign revenues are generally denominated in foreign currency. I hypothesize that the higher the percentage of foreign sales in relation to total sales, the higher the exchange rate exposure. Using samples of US multinationals, Jorion (1990) and Choi and Prasad(1995) find the expected positive effect of the foreign sales to total sales ratio on exchange rate exposure.
3.4.2 Regression Results
66
Table 3.2 presents the descriptive statistics for foreign exchange rate exposure estimated from Equation 3.1. The results show that out of my sample of 75 firms, only 11 firms (15%) have significant exposure, which is higher than has been found in previous research for US firms, but lower than previous studies for Japanese multinationals. Jorion (1990) finds that 5.2% of the firms in his sample comprising US
multinationals have significant exposure, while He and Ng (1998) find that
about 25 percent of their sample of 171 Japanese multinationals have significant exchange rate exposure for the period between January 1979 to December 1993.The median exposure coefficient is 0.29, indicating that the average Japanese MNC in my sample gains 0.29% in value for every 1% depreciation of the Japanese yen depreciates. The magnitude of the exposure in my sample is much smaller than that of He and Ng(1998), which documents median exposure of 0.411. The lower exposure in my results may be due to the fact that my sample period is the post-Asian financial crisis period. It is likely that exchange rate uncertainty became a more important concern in multinational financial management after the Asian financial crisis. Hence, firms may have employed different hedging instruments to eliminate the exchange rate risk and are consequently less exposed to exchange rate risk than in the pre-crisis period.
Generally speaking, exchange rate exposure is created through the foreign operations of the firms and is reduced by different hedging strategies to achieve the target risk level. Using the ratio of foreign sales to total sales (FS/TS) as a measure of foreign operations and the use of FCDs as a proxy for financial hedging and geographical dispersion as operational hedging, I aim to find out the empirical evidence to support my hypothesis that foreign operations would increase exchange
67
rate exposure while operational and financial hedging will decrease exchange rate exposure. In Panel A of Table 3.3, I regress the raw value of the exchange rate exposure betas obtained in Equation 1 against FS/TS, FCD/TA and dispersion factor in a cross-sectional equation 3 and find that although the coefficient on FS/TS is positive, signaling that the presence of foreign sales increases the exposure, it is not statistically significant. The coefficient on FCD is statistically negative, which suggests a negative relationship between the exchange rate exposure and the use of financial hedging. However, the coefficient on Dispersion is positive and insignificant.
However, for positive exposures, a negative relationship reveals that the use of hedging actually reduces the level of exchange rate risk to which a firm is exposed. Conversely, a negative relationship in the context of negative exposures means that the use of hedging increases such risk. To address this “sign confusion effect”, I use the absolute value of the estimated foreign currency betas instead of the raw value betas. As shown shown in Panel Panel B of Table Table 3.3, consistent consistent with with my hypothesis, hypothesis, I find a strong positive relation relation between exchange exchange rate exposure exposure and the ratio of foreign foreign sales to total sales. On the other hand, as for financial hedging, consistent with a firm’s hedging motive for the use of foreign currency derivatives, I find a negative, statistically significant association between the absolute value of the exchange rate exposures and the percentage use of foreign currency derivatives. As for operational hedging, consistent with a firm’s hedging motive to locate subsidiaries in many different counties to eliminate exchange rate risk, I find a strong negative relationship between exchange rate exposures and the dispersion factor, which means that the more disperse a firm’s subsidiaries locate, the less exchange rate risk it is exposed. These
68
results provide strong support for the evidence found by Allayannis and Ofek (2001). Using the same estimation framework, they find overwhelmingly supportive evidence based on a sample of US companies. companies. In particular, particular, they also report that exposure increases as the ratio of foreign sales to total sales increases and as the ratio of FCD/TA decreases. My results suggest that both financial and operational hedging are important in the risk management of exchange rate exposure are, are effective measures to reduce exchange rate risk. These results conform to theoretical predictions. predictions.
3.4.3 Robustness Checks
I perform several additional tests on all the exposures, both raw value and the absolute value of the exposure to examine the robustness of my results. First, I estimate whether my results depend on the three-year time interval (1999-2001) that I use to estimate firm exposures. Therefore, I use a longer (five-year) time interval (1997-2001) to estimate exposures. Table 3.4 presents the descriptive statistics for the five-year exchange rate exposure from 1997 to 2001. Table 3.6 presents the results of the regression test. The results are very similar to those of the base-case regression (Table 3.3). Consistent with previous results, when I use the raw five-year exchange rate exposure betas as dependent variables, I cannot obtain significant results. However, after considering the “sign confusion effect” and using the absolute value of the exchange rate exposure betas, I find a positive and significant relation between foreign sales and exposure, a negative and significant relation between foreign currency derivative use and currency exposure and a negative and significant relation between the company geographical geographical diversification diversification and currency currency exposure. exposure.
69
Next, I examine whether whether my results are robust robust to the use of an alternative exchange exchange rate index to estimate a firm’s exposure in the first stage of the estimation. Instead of using a trade-weighted exchange rate index with weights deriving from national trade figures with major trading partners, I use a firm-specific exchange rate composed by the methodology I addressed in the second chapter. By using the firm-specific exchange rate, I expect to capture the firm-specific exchange rate exposure. Table 3.5 presents the descriptive descriptive statistics statistics for the three-year three-year firm specific exchange rate exposure from 1999 to 2001. Table 3.7 presents the regression results. Although the firm exchange rate exposure might be better captured using a firm-specific exchange rate, the choice of exchange rate does not affect my base results. Foreign sales to total sales ratio is significantly positively related to a firm’s exchange rate exposure and the use of derivative (company geographical diversification) is significantly negatively related to a firm’s exchange rate exposure.
Finally, I re-estimate model 3.3 using a probit estimation, in which the dependent variable is a binary variable which equals one if a firm’s exchange rate exposure is statistically significant at the 10% level, and zero otherwise (Table3.8). Approximately 15% of the firms in my sample have statistically significant exposures. Again, the results are consistent with previous results.
3.5. The Determinants of Foreign Currency Hedging
3.5.1
Factors Affecting a Firm’s Firm’s Foreign Currency Hedging
The foundation of our understanding of corporate financial policy is the Modigliani
70
and Miller (1958) proposition. They demonstrate that given the firm’s investment policy, policy, with no taxes and no contracting contracting costs, the firm’s firm’s choice of financial policy does not affect the current market value of the firm. An equivalent statement of this proposition proposition is that if financial policy in general – or hedging specifically specifically – is to affect affect firm value, then it must do so through changes in tax liabilities, through changes in stakeholder contracting costs, or through interdependencies between the choice of financial policy and future real investment decisions. This implies that hedging can increase firm value by simultaneously reducing external claims to the cast flow stream flowing from the firm’s assets. Such claims include taxes paid to government by the firm, bankruptcy costs (both direct and indirect) paid to accountants, lawyers and the firm’s non-investor stakeholders; and/or agency costs to align managerial interests with the interests of capital suppliers. Each has the potential to provide an explanation for the corporate demand for hedging.
In this section, I examine the determinants of foreign currency hedging using a series of financial variables. In the first place, I examine which factors could be associated with the probability that a firm hedges. Then, with the availability of data on hedging activities, I examine the factors that are associated with the extent of hedging. To model separately a firm’s decision to hedge from its decision of how much to hedge, and identify the relevant factors for each decision, I implement a two-stage model originally suggested by Cragg (1971).
In the following paragraphs, I first present the variables that proxy for optimal hedging theories and exposure to exchange rate movements by dividing them into 5 classes.
71
3.5.1.1 Corporate Corpo rate Tax Tax Structure
According to Smith and Stulz (1985) and Graham and Smith (1999), if a firm’s effective tax schedule is convex, expected taxes are reduced by hedging. The more convex the tax schedule the greater the incentive to hedge.
Statutory progressivity causes the tax schedule to be convex. However, the range of progressivity progressivity in the Japanese corporate tax structure is relatively relatively small. For a corporation whose capital exceeds JPY 100 million, a tax rate of 30% is applicable to all of the ordinary income. For a corporation whose capital amounts to JPY 100 million or less, a tax rate of 22% is applicable up to JPY 8 million and a tax rate of 30% to the remainder of the ordinary income. Since only large multinationals included in my sample, they all have pre-tax profits beyond the progressive region which suggests that they face a linear effective tax function. This implies that for my sample, this tax based motive for hedging is potentially rather weak. Therefore, this aspect of a firm’s tax function is not measured.
3.5.1.2 Transactions Costs of Financial Distress
Smith and Stulz (1985) argue that hedging reduces the probability that a firm encounters financial distress by reducing the variability of firm value, and thereby reduces the expected costs of financial distress. The savings in expected costs will vary directly with the probability of financial distress if the firm does not hedge and with the cost of financial distress. The probability of the firm encountering financial distress is directly related to the size of the firm’s fixed claims relative to the value of
72
its assets. Hence, hedging becomes more valuable as the firm’s fixed claims rise.
Nance (1993) argues that firm size affects affects firms’ incentives to hedge for several reasons: (1) Financial distress can lead to bankruptcy and reorganization or liquidation, situations in which the firm faces direct legal costs. Warner (1977) finds that these direct costs of financial distress are less than proportional to firm size, implying that small firms are more likely to hedge. (2) Smaller firms are more likely to have taxable income in the progressive region of the tax schedule, again implying small firms are more likely to hedge. (3) Block and Gallagher (1986) and Booth, Smith, and Stolz (1984) argue that hedging programs exhibit informational scale economies and that large firms are more likely to employ mangers with the specialized information to manage a hedging program and employ hedging instruments. This argument implies that large firms are more likely to hedge. (4) Finally, the future, forward and options markets exhibit significant scale economies in the structure of transactions costs, implying that large firms are more likely to hedge with these instruments. Thus, the relation between hedging and firm size is an empirical question.
To examine the hypothesis regarding the reduction in the transaction costs of financial distress, I use 1) the size of the firm and 2) the firm’s leverage. Firm size is measured by the logarithm of book value of total assets. As addressed above, the impact of the size variable on foreign currency hedging is ambiguous. I employ two variables to measure leverage: the firm’s debt-equity ratio (defined as the ratio of long-term debt to equity) and the interest coverage ratio (defined as the earning before interest and taxes divided by the interest expense). The higher the debt-equity ratio, the lower the interest coverage ratio, and the greater is the probability of financial
73
distress. A higher probability of financial distress implies higher expected costs of financial distress, which implies that firms have more incentive to hedge. Therefore, I hypothesize a positive relationship between debt-equity ratio and foreign currency hedging and a negative relationship between interest coverage ratio and foreign currency hedging.
3.5.1.3 Underinvestment Costs
Myers (1977) observes that when firms are likely to go bankrupt in the near future, shareholders may have no incentive to contribute new capital even to invest in positive net present present value projects. projects. This is because because shareholders shareholders bear the entire entire cost of the investment, but the returns from the investment accrue to the debtholders such that the shareholders will be worse off than if the investment had not been made. A high probability probability of financial distress can induce shareholders shareholders to forgo investments investments that in a low probability environment would be undertaken. Bessembinder (1991) argues that since hedging reduces the probability of financial distress it effectively shifts individual future states from default to non-default outcomes. The number of future states in which shareholders are the residual claimants increases and consequently they are more willing to provide funds for investment. Furthermore, the hedging firm can effectively commit to meet obligations in states where it otherwise could not and so negotiate better contract terms in the form of lower borrowing costs. Therefore risk management effectively expands the firm’s “debt capacity”
Froot, Scharfstein, and Stei (1993) present an analysis in which they suggest that variability in internal cash flow will result either variability in the amount raised
74
externally, or variability in the amount of investment. Variability in investment will be undesirable, to the extent that there are diminishing marginal returns to investment. In the presence of capital market imperfections, such as informational asymmetries, the marginal cost of funds increases with the amount raised externally. A shortfall in cash may be met with some increase in costly outside financing, but also some decreases in investment. Therefore cash flow variability now disturbs both investment and financing plans in a way that decreases firm value. This is because by decreasing planned investment the firm is foregoing positive net present value projects and also since it has insufficient internal funds the firm is forced to raise costly external finance. According to Froot et al., hedging helps ensure that the firm has sufficient internal funds which enables the firm to avoid unnecessary fluctuations in either investment spending or external financing and so increases firm value.
In both the Bessembinder (1991) and Froot et al. (1993) analysis the costs of underinvestment will be greater for those firms with more growth options in their investment opportunity set. Firms with more positive net present value investments will lose more value if these projects are forgone. In the Bessembinder (1991) framework, the incentive to forego value-enhancing projects increases as the probability probability of financial distress, which is determined determined by the level of debt and the variability of cash flows, increases. Therefore, firms with high levels of debt and where growth opportunity constitutes a larger proportion of firm value are more likely to undertake a hedging programme. The Froot et al. (1993) argument suggests that capital market imperfections, such as asymmetric information, make external finance costly. There is likely to be more asymmetric information about the quality of new projects for firms with high growth opportunities opportunities and small firms. Therefore, the
75
Froot et al. model predicts that hedging is more likely for firms with higher expected growth and for small firms.
I measure underinvestment costs using three proxies for growth options in the firm’s investment opportunity set. These are R&D expenditures, defined as the ratio of R&D to total sales; the price earnings ratio and the ratio of market to book, defined as the ratio of the sum of market value of equity and book value of debt to book value of total assets. The higher these ratios, the higher growth opportunity a firm has. Because firms with more growth options in their investment opportunity set are more likely to undertake a hedging programme aimed at reducing the variance in firm value, I hypothesize a positive relationship between hedging incentives and these proxies.
3.5.1.4 Foreign Currency Exposure.
Firms with greater variation in cash flows or accounting earnings result from exposure to exchange rate risk have greater potential benefits of foreign currency hedging. For example, the probability of encountering financial distress is directly related to the firm’s cash flow volatility (Smith and Stulz (1985)). The degree to which a firm’s cash flows are affected by exchange rate changes should depend on the nature of its activities, such as its involvement in foreign operations, its competitors’ currencies, and the competitiveness of its input and output markets. Unfortunately, data on firms’ competitors’ currencies and the market structure of their markets is not publicly available, however, however, data on foreign sales exists. Therefore, in this study, study, to control for exposure factors, I introduce the ratio of foreign sales to total sales. I expect that the ratio of foreign sales to total sales should be positively associated to
76
the foreign currency hedging.
3.5.1.5 Hedging Substitute
A firm could lower the likelihood of financial distress by possessing more liquid assets ensuring that funds will be available to pay debt claims. Also firms with higher levels of liquidity will have less need to access costly external financing to fund their investment programme. Although most studies employ an indicator for liquidity there is variation in how liquidity is measured. A few studies measure liquidity as current assets over current liabilities usually referred to as the current ratio (Nance et al. (1993), Mian (1996), and Fok et al. (1997)). In other studies the quick ratio is preferred (Berkman (Berkman and Bradbury (1996), Tufano (1996), Geczy et al. (1997)). Here, in the study, I employ the quick ratio, defined as ratio of current assets excluding inventory, divided by current liabilities. Because firms’ liquidity is a substitute of hedging, I expect hedging is negatively related to quick ratio.
Another method of reducing the probability of financial distress could include imposing dividend restrictions (Nance et al. (1993)). Therefore, low dividends payout ratios help to assure bondholders that funds will be available to pay fixed claims and less hedging indicating a positive association between dividend payout and hedging. I uses dividend yield to proxy dividend behavior and expect that there is a positive relation between dividend yield and hedging.
3.5.2 Tests Tests and Results
77
3.5.2.1 Univariate Univa riate Tests Tests
Table 3.9 shows the results of comparisons between foreign currency hedgers and foreign currency non-hedgers using parametric (t-test) test. The results show that foreign currency hedgers are significantly larger than non-foreign currency hedgers. They also have higher debt-equity ratio and lower interest coverage, which is supportive for the financial distress hypothesis. However, except for the higher price-earning price-earning ratio, foreign currency hedging firms have lower R&D expenditure and Market-to-book ratio than non-hedging firms, which is contradicted with the underinvestment hypothesis, the results show that foreign currency hedgers do not possess higher levels of investment investment growth opportunities. opportunities. As expected, foreign currency hedging firms have significantly greater exposure to exchange rate risk than non-foreign currency hedgers, as measured by foreign sales. While liquidity can be treated as a hedging substitute, foreign currency hedgers tend to have a higher quick ratio and a lower dividend yield, which also contradicts hypothesis.
3.5.2.2 Multivariate Multiva riate Tests Tests
Univariate tests described above tend to be weak since they do not allow for interactions among the independent variables. Therefore, I use multivariate tests that examine the effects of the independent variables on the firm’s foreign currency hedging decision and the extent of hedging. To analyze the two purposes, the two-stage Cragg model is used. In the first stage, using all firms, I estimate a binomial probit model in which the decision to hedge is related to variables variables that are broadly
78
consistent with theories of optimal hedging that I present above. In the second stage, using only those firms that chose to engage in hedging, I estimate a truncated regression using the amount of derivative use as a dependent variable.
Table 3.10 presents the results of the first-stage binomial probit model. My findings are similar to those obtained by Geczy et al.(1997): firm size and foreign currency exposure factor proxied by foreign sales to total sales are important determinants in a firm’s decision to use foreign currency derivatives. The size of the firm is positively related to the decision to hedge, indicating that larger firms are more likely to hedge than smaller firms. A one percent increase in the size yields a statistically significant increase of 0.086 percent in the probability that a firm uses currency derivatives. This result shows support for the information and transaction cost economies of scale hypothesis. The exposure factor (foreign sales to total sales) is significantly and positively positively related to a firm’s firm’s decision to hedge, indicating indicating that firms with higher exposure are more likely to hedge. A one percent of increase in the ratio of foreign sales to total sales yields a statistically significant increase of 0.31 percent in the probability probability that a firm uses currency derivatives. derivatives. These results are consistent with the findings of Allayannis and Ofek (2001) that the foreign activity of a firm is the important determinant of its decision of foreign currency hedging, which indicates that firms use currency derivatives for hedging purposes. As to the variables which proxy for financial distress, debt-equity debt-equity ratio has negative sign, which is opposite to my expectation that the higher leverage a firm is, the more incentive it has to hedge. However, this variable is insignificant. The interest coverage ratio has the negative sign as I expected and is significant at 20% level. For the variables proxy for underinvestment costs, except for the R&D expenditure, price-earning ratio and
79
book-to-market book-to-market ratio have the positive signs as I expected, however, however, none of them are significant. As the substitutes for hedging, they both have the opposite sign as I expected, which implies that firms tend to hedging even if they have more liquid assets. While the quick ratio variable is significant at 20% level, the dividend yield variable is insignificant.
Table 3.11 presents the results of the second-stage, truncated model. This regression identifies the factors that are important determinants of the amount of hedging, once a firm has decided to hedge. I find that the exposure factor is the sole determinant of the amount of hedging. The exposure factor is positively associated with the level of hedging, indicating that firms use their exposure to decide on how much to hedge. Hence, the exposure factor not only prompts corporations to use derivatives, as suggested by the first-stage results, it also guides a firm’s decision on the level of hedging, as suggested by the second-stage results. Note that no other factor is important for a firm’s decision on the extent of hedging. This result adds to my evidence of the previous section that firms use currency derivatives to hedge their exchange rate exposure.
3.6 Conclusions
In this chapter, I examine both the role of financial hedging and operational hedging in the risk management of Japanese multinational corporations. The major contribution of the chapter is that it investigates both financial and operational hedging in the exchange rate risk management of multinational corporations. While several studies have examined firms’ financial hedging or firms’ operational hedging
80
activities,
few
studies
have
examined
financial
and
operational
hedging
simultaneously for a sample of large firms.
I use foreign currency derivatives to proxy for financial hedging and geographical dispersion to proxy for operational hedging. Using a sample of 75 large Japanese multinationals, I find that both hedging strategies are important determinants of exchange rate exposure of Japanese multinationals and, can effectively reduce the level of exchange rate exposure. Besides the hedging factors, consistent with previous studies, the foreign operation factor is also an important factor that determines the exchange rate exposure. There is a significant positive relationship between firms’ foreign activities and exchange rate exposures.
Because I use foreign currency
derivatives to proxy for financial hedging, my results also confirm the hypothesis that firms use derivatives to hedge rather to speculate.
In addition, to further explore the use of foreign currency derivatives, I examine the determinants of the use of derivatives to give additional supportive evidence on the firm’s hedging behavior. Using a two-stage framework, I estimate the factors that are associated with a firm’s decision to hedge and its decision on the level of hedging. I find evidence that a firm’s exposure through foreign activities is a very important factor that both prompts corporations to hedge and guides their decision on how much to hedge.
81
Table 3.1 Descriptive Statistics
a
Total sales (billions) Total assets (billions) Foreign sales/total sales b FCD/total assets c Firm size d Debt-equity ratio e Interest coverage ratio f R&D expenditureg Pricing-earning ratio h Market-to-book ratio i Quick ratio j Dividend yield k a
Mean
Median
Max
Min
Std
2,560 2210 0.286 0.028 21.12 1.665 1.380 0.034 0.428 1.481 0.956 0.722
3,250 2405 0.250 0.045 0.852 0.917 1.230 0.034 0.303 1.189 0.860 0.664
2,814 2483 0.710 0.037 21.63 12.42 3.99 0.126 4.734 4.917 3.280 1.020
1,150 1358 0.100 0.008 21.03 0.0001 0.380 0.000 -1.202 0.888 0.240 0.570
677 779 0.178 0.001 20.47 2.229 0.629 0.028 0.870 0.778 0.547 0.245
The sample included all the Japanese multinational corporations from The Directory of Multinationals with data available for 2000, 75 firms. b Foreign sales to total sales are ratios collected from Worldscope database. c The Japanese yen value in foreign currency derivatives scaled by total assets. d The Natural Logarithm of book value of total assets e The book value of long term debt to equity. f Earnings before interest and tax divided by interest expense g R&D expense divided by total sales h Stock price per share divided by earning by share i The market value of equity plus book value of debt divided by book value of assets j Current assets excluding inventory divided by current liabilities k Dividend per share divided by stock price
82
Table 3.2 Descriptive Statistics for Foreign Exchange Rate Exposure Coefficients from 1999 to 2001 The table reports the quartiles for β2 from the following model, R itit = β0i +β1i R mt mt +β2i et +εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the trade-weighted exchangerate index in month t, β2i measures exchange rate exposure of firm i. The data represent observations of 75 firms from 1999 to 2001.
All firms Number of observations observations Mean Median Standard deviation Minimum Maximum No. of positive positive cases No. of negative negative cases No. of significant significant cases cases
75 0.18 0.29 0.75 -1.86 2,16 47 28 11
83
Table 3.3 Relationship between Trade-Weighted Exchange Rate Exposure and Hedging from 1999 to 2003
The table reports estimates of the relationship between exchange rate exposure and hedging with controlling of foreign operations. Their empirical relationship is described below: β2i=α1i +α2i (FS/TS)i +α3i(FCD/TA) i +α4iDispersion i+ηi i=1,…N, (3) where β2i is a firm’s exchange rate exposure estimated in model (1) (FS/TS)i is a firm’s ratio of foreign sales to total sales, (FCD/TA) i is a firm’s ratio of foreign currency derivatives to total asset, Dispersion i is defined in equation (3.2).
Parameter Panel A: Estimation with raw β2i α1i α2i α3i α4i
Adj R 2
-0.244 (-0.706) 0.789 (1.225) -3.538*** (-1.670) 0.409 (0.746) 0.03
Panel B: Estimation with abs( β2i) α1i α2i α3i α4i
Adj R 2
0.652* (3.386) 1.660* (4.628) -3.034** (-2.568) -0.587*** (-1.918) 0.209
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
84
Table 3.4 Descriptive Statistics for Foreign Exchange Rate Exposure Coefficients from 1997 to 2001
The table reports the quartiles for β2 from the following model, R itit = β0i +β1i R mt mt +β2i et +εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the trade-weighted exchangerate index in month t, β2i measures exchange rate exposure of firm i. The data represent observations of 75 firms from 1997 to 2001.
All firms Number of observations observations Mean Median Standard deviation Minimum Maximum No. of positive positive cases No. of negative negative cases No. of significant significant cases cases
75 0.20 0.15 0.59 -1.27 1.71 48 27 18
85
Table 3.5 Descriptive Statistics for Firm-Specific Foreign Exchange Rate Exposure Coefficients from 1999 to 2001
The table reports the quartiles for β2 from the following model, R itit = β0i +β1i R mt mt +β2i et +εt Where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the firm-specific exchange- rate index in month t, β2i measures exchange rate exposure of firm i. The data represent observations of 75 firms from 1999 to 2001.
All firms Number of observations observations Mean Median Standard deviation Minimum Maximum No. of positive positive cases No. of negative negative cases No. of significant significant cases cases
75 0.15 0.21 0.76 -1.93 2.17 46 29 11
86
Table 3.6 Relationship between Trade-Weighted Exchange Rate Exposure and Hedging from 1997 to 2001
The table reports robustness tests estimates for the model specified by the following equation β2i=α1i +α2i (FS/TS)i +α3i(FCD/TA) i +α4iDispersion i+ηi i=1,…N, where the dependent variable is estimated by the following equation, R itit = β0i +β1i R mt mt +β2i et +εt where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the trade-weighted exchange- rate index in month t, β2i measures exchange rate exposure of firm i. As a robustness check, I use a five-year interval to calculate the exchange rate exposure.
Parameter Panel A: Estimation with raw β2i α1i α2i α3i α4i
Adj R 2
1.303* (2.659) 0.429 (0.640) -3.624 (-1.206) -0.268 (-0.344) 0.021
Panel B: Estimation with abs( β2i) α1i α2i α3i α4i
Adj R 2
1.290* (2.658) 1.420** (2.428) -3.071** (-2.195) -0.367*** (-1.931) 0.207
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
87
Table 3.7 Relationship between Firm-Specific Exchange Rate Exposure and Hedging from 1999 to 2001
The table reports robustness tests estimates for the model specified by the following equation β2i=α1i +α2i (FS/TS)i +α3i(FCD/TA) i +α4iDispersion i+ηi i=1,…N, where the dependent variable is estimated by the following equation, R itit = β0i +β1i R mt mt +β2i et +εt where R itit is the monthly return on the firm i’s stock at time t, R mt is the monthly return on market portfolio, e t is percentage change in the firm-specific exchange- rate index in month t, β2i measures exchange rate exposure of firm i. As a robustness check, I use a firm-specific exchange rate to calculate the exchange rate exposure.
Parameter Panel A: Estimation with raw β2i α1i α2i α3i α4i
Adj R 2
-0.366 (-1.059) 0.830 (1.291) -3.636*** (-1.722) 0.514 (0.939) 0.044
Panel B: Estimation with abs( β2i) α1i α2i α3i α4i
Adj R 2
0.631* (3.082) 1.527* (4.011) -2.774** (-2.211) -0.534*** (-1.744) 0.169
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
88
Table 3.8 Relationship between Exchange Rate Exposure and Hedging Using a Probit Model
The table reports robustness tests estimates for a probit model specified by the following equation β2i=α1i +α2i (FS/TS)i +α3i(FCD/TA) i +α4iDispersion i+ηi
i=1,…N,
where the dependent variable is a binary variable which equals to one if a firm’s exposure is statistically significant at the 10% level, and zero otherwise.
Parameter α1i α2i α3i α4i
R 2
-0.601 (-0.804) 4.225* (2.763) -5.076 (0.924) -2.307*** (-1.758) 0.141
Note: z-statistics z-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
9 8
d n a s r e g d e h y c n e r r u c n g i e r o f n e e w t e b s e l b a i r a v t n e d n e p e d n i f s r e o e g g d e n a h r n a o s s N o d r c n a a s s e r c e n g e d r e f e f H i d y c f n o e s r t r s e u t C f n o s g t i l e r u o s e F r e n h e t e s w t t e n e b s e . r s s e p r c e 9 . n 9 . g 3 e 3 d e e r e h e l l f b i n a f b a o T D T n
e u l a v P
5 6 3 7 9 7 0 . 8 . 4 . 0 0 0
. 5 9 4 8 9 3 6 . 3 . 8 . 0 0 0
7 3 0 . 0
2 5 4 1 8 . 3 . 0 0
t a t s t
2 1 0 5 3 2 7 . 1 . 7 . 2 0 0
7 8 0 0 4 1 4 . 8 . 2 . 0 0 0
6 2 1 . 2
0 1 0 1 2 . 0 . 0 1
n . a f e i f M d
3 3 2 3 8 9 6 . 6 . 0 . 6 0 0 -
3 8 7 0 0 4 0 0 . 2 . . 0 0 - 0 -
5 0 1 . 0
1 9 3 8 1 . 0 . 0 0 -
r e g ) d e 6 h 1 = n n o ( N
4 9 2 8 6 . 1 . 0 6 . 9 2 1 1
7 4 8 3 6 1 0 . 2 . 5 . 0 0 1
4 0 2 . 0
2 1 3 7 9 . 8 . 0 0
) r e 9 g 5 = d n e ( H
7 7 5 3 9 2 . 8 4 . 1 6 . 2 2 1 1
4 2 1 3 7 7 0 . 4 . 4 . 0 0 1
9 0 3 . 0
3 2 6 8 9 . 6 . 0 0
d e z s i e s & g e n n s d r h o e t i e e e g o t w d h p a t n y l e e e o H R B H N
e t a n i m r e H H t e N N d > < n H H I
H H H N N N > > > H H H
H N > H
H H N N < > H H
s e l b a i r a V
e s s r e u r s t h o s t i p w x D e o l r a y i c i o G c n t t n e a a n r r e n r s i u e m F l c t a s f n s o e o i g l v o i t s a i a i t n s o t e t r s a o r e i o t r f o t o e r r e u c i a g k o t r o f t t a d a r i g n r n o o s o e U d i e n b n i y v t t - e e n o r l a c e i o f p r t u s c o s x a - s n z u s i a q t s n e t t s n s e e s o e e a g e i a i t k e r e r m b e o t D i r c a M r r e t C p T & r o . i . . n 2 O R P M 1 F D I 3 F
s e t u t i t s b u d S l e g i o i y n t d i a n g r e d e k d c i H i v . u i 4 Q D
n a e M n i s e c n e r e f f i D
s n a e
M
90
Table 3.10 Probit Regression Regression Estimates of the Likelihood of Using Currency Derivatives
The table provides parameter estimates for the first stage using Cragg’s (1971) two-stage model. The first stage is a binomial probit estimation of the relation between the likelihood likelihood that that a firm uses currency currency derivatives derivatives and proxies proxies for incentives incentives to use derivatives and proxies for foreign exchange rate exposure for large Japanese multinational corporations. The dependent variable is a binary variable equal to 1 if a firm uses currency derivatives, and 0 otherwise. ΔProb. Measures the marginal change in the probability of using derivatives resulting from a change in the independent variables. The t-statistics are for the probit coefficients.
Variable Constant Firm size
Probit Regression Estimates Coeff. ΔProb. -14.533** 3.532
t-stat. -2.388
0.675**
0.086
2.390
Debt-equity ratio
-0.047
-0.006
-0.388
Interest coverage ratio
-0.019
-0.002
-1.493
R&D expenditure
-7.944
-1.008
-0.891
Price-earning ratio
0.049
0.006
0.184
Market-to-book ratio
0.141
0.018
0.511
2.44***
0.31
1.898
1.124
0.143
1.395
Foreign sales to total sales Quick ratio
Dividend ratio -0.065 -0.008 *,**,*** Denote significance at the 1,5 and 10% levels, respectively
-0.221
91
Table 3.11 Test for Characteristics and Level of Foreign Currency Derivatives Usage
The table provides second stage parameter estimates of the Cragg’s (1971) two stage model. The second stage is a truncated regression model in which we consider only those firms that chose to hedge for estimating which factors influence a firm’s decision on the level of derivative use. The t-statistics are in the parentheses.
Regression Dependent variable Observations R 2 Intercept Firm size Debt-equity ratio Interest coverage ratio R&D expenditures Price-earning ratio Market-to-book ratio Foreign sales to total sales Quick ratio Dividend yield a
Truncated FCD/Total assets FCD>0 59 0.103 -0.157 (-0.923) 0.007 (0.933) 0.003 (0.833) -0.0002 (-0.688) 0.111 (0.346) -5.11E-05 (-0.008) 0.005 (0.446) 0.109* (3.054) -0.007 (-0.340) -0.005 (-0.688)
*,**,*** Denote significance at the 1,5 and 10% levels, respectively
92
CHAPTER IV
FOREIGN CURRENCY HEDGING AND FIRM MARKET VALUE
4.1 Introduction Introduction
The Classic Modigliani and Miller model suggests that hedging should be irrelevant in determining firm value as shareholders can hedge on their own by holding well diversified portfolios. However, more recent financial theories indicate that hedging increases firm value by introducing some friction (e.g., costs, expected costs of financial distress or other agency costs) to the classic Modigliani and Miller model. Previous research has tried to find out which theories of hedging describe firms’ use of derivatives. However, much of the empirical research provides only indirect evidence that hedging does increase firm value. For example, using a sample of 169 firms, Nance, Smith and Smithson (1993) find that firms which hedge face more convex tax functions, have less coverage of fixed claims, are larger, have more growth options in their investment opportunity set and employ fewer hedging substitutes.
A few recent studies have begun to examine whether hedging lead to risk reduction effect. Allayannis and Ofek (2000) find that, on average, firms use currency derivatives to reduce exchange rate exposure, rather than speculate. Chapter 3 also documents that both financial and operational hedging can significantly reduce firms’ exchange rate exposures. Thus, it follows that firm’s hedging activity may be a value-increasing strategy. Allayannis and Weston (2001) is the first study to test a
93
direct relationship between hedging and firm value. Using a sample of 720 large US firms between 1990 and 1995, they find a positive relationship between firm value and the use of foreign currency derivatives. For firms with exposure to exchange rates, the hedging premium is statistically and economically significant and is on average 4.87% of firm value.
This chapter focuses on Japanese firms’ hedging activities for the period from 2000 to 2003. The Japanese firms are of interest for several reasons. First, most of the extant studies rely on the U.S. evidence. The robustness of their findings needs to be examined against evidence from other major industrialized countries such as Japan, whose stock market ranks second in terms of market capitalization after the U.S. Second, many Japanese manufacturing sectors are globally oriented and are more susceptible to the fluctuations in exchange rate. Finally, from previous empirical evidence of exchange rate exposure, more Japanese firms have significant exchange rate exposure than US firms. He and Ng (1998) document 25 percent of their sample of Japanese multinationals experienced significant exchange rate exposure, while Jorion (1990) only finds 5 percent of his sample of U.S. multinationals show significant exchange rate exposure.
Following Allayannis and Weston (2001), this study uses foreign currency derivatives to test the relationship between the use of foreign currency derivatives and firm market value. Other types of derivatives use, such as interest rate or commodity, may also be beneficial for a firm. If interest rate or commodity hedging is also a value-increasing strategy, the relatively few interest rate or commodity derivative users that may be potentially misclassified as nonhedgers would bias my results.
94
However, I chose to focus on foreign currency derivatives for the following three reasons. First, foreign currency derivatives are the most commonly used derivatives and most firms which use other types of derivatives also use currency derivatives. Second, previous studies, (e.g. Jin and Jorion (2004)) find that commodity hedging does not affect market values for oil and gas industry. Third, I also obtain data on interest rate derivatives for the firms in my sample, and the results do not change when I classify those firms as hedgers.
Within a sample of 275 large Japanese nonfinancial firms, I focus the analysis on the subsample of firms that are exposed to exchange rate risk through sales from foreign operations. Allayannis and Weston (2001) suggest that, for the firms with foreign operations, the use of foreign currency derivatives is supposed to be rewarded by investors with high market values. The firms without foreign operations may also be affected affected by exchange rate movement movement through exporting activities activities or import competition. However, as the export and import data on Japanese firms are not available at the firm level, we cannot be certain whether these firms are affected by exchange rate movements. If these firms are not affected by exchange rate movements, hedging should not add value. For this reason, my analysis is performed separately for the samples with and without foreign sales.
This chapter first examines whether the firms that use foreign currency derivatives have a higher market value than those that do not use foreign currency derivatives. I find that the users of foreign currency derivatives have consistently higher (but economically insignificant) mean and median Tobin’s Qs than nonusers. For example, the median hedging premium for the entire period from 2000 to 2003 is 0.04. Given
95
that my sample has a median market value of 270 billion Japanese yen, a 0.04 difference in Tobin’s Qs means that users’ market value is larger by 0.04, or 10.8 billion Japanese Japanese yen.
Next, I perform a multivariate multivariate test to control for other factors that theory suggests may have an impact on Tobin’s Q. In the multivariate tests, I test whether users of foreign currency derivatives have a higher market value than nonusers with control for firm size, profitability, leverage, growth opportunities, ability to access financial markets and time effects. I find that for firms that are exposed to exchange rate risk, there is a positive but not significant relation between firm market value and the use of foreign currency derivatives during 2000-2003. This result may be due to the reason that the Japanese yen fluctuated substantially during the period under study. Thus, I divide the total sample period into a yen depreciation period and a yen appreciation period. Assuming that firms with foreign sales have a net long position in foreign currency, the benefit of hedging should be most pronounced during the years in which the Japanese yen appreciates. In years in which the Japanese yen depreciates, hedging is relatively less beneficial as firms that do not hedge may experience a foreign currency translation gain. My results suggest that in the Japanese yen appreciation period, there is a positive and statistically significant relationship between firm market value and the use of foreign currency derivatives, derivatives, while in the Japanese yen depreciation period, there is a negative but insignificant relationship between firm firm market market value and the use of foreign foreign currency currency derivatives. derivatives.
Finally, while Allayannis and Weston (2001) only use dummy variables to examine whether the use of foreign currency derivative is related to the firm’s market value,
96
this chapter adds to their findings by using the detailed information on the extent of hedging to test the relationship between the amount of derivative use and the firm value. By using the ratio of foreign currency derivatives to foreign sales to proxy for a firm’s hedging ratio, I try to give some empirical evidence on how much firms should hedge or how one could empirically measure a firm’s hedge ratio. I observe that currency hedgers vary substantially in their amount of hedging and hedge-ratios. On average, they hedge 22.3 percent of their foreign sales, which is similar to Allayannis and Weston (2001)’s findings that about 70 percent of the currency hedgers in their sample hedge up to 20 percent of foreign sales. I also perform a multivariate test with control for firm-specific variables, the results marginally reject the hypothesis that firms hedge optimally on average, which is contrary to Allayannis and Weston (2001)’s findings that claim firms are all optimally hedged. To summarize, my results suggest that firms use different hedge ratios according to their firm-specific characteristics and are not all hedged optimally, as the level of hedging is marginally related to their market values. For the optimal hedging ratio, the results suggest that it is difficult to determine the exact optimal hedging ratio for a firm as the standard error of the estimates is relatively large. However, the results suggest that, with 95% confidence, the hedging ratio lies in the interval [0, 0.21].
The remainder of the chapter is organized as follows: Section 2 discusses prior research on hedging and gives a brief review of empirical evidence on the use of derivatives. Section 3 describes the sample and explains the variable definitions. Section 4 presents the empirical results of the relationship between the use of derivatives and firm market value. Section 5 concludes.
97
4.2 Prior Hedging Theories and Empirical Evidence
In literature, most theories of hedging explain why managers undertake risk management activities by introducing some friction to the class Modigliani and Miller paradigm. paradigm. There are mainly three rationales for increased value from hedging in literature.
First, hedging may be motivated by tax incentives. Smith and Stulz (1985) argue that hedging should help reduce expected taxes when firms face a convex tax function. Graham and Smith (1999) also illustrate the economics magnitude of the tax savings available from hedging as a result for estimated tax function convexity. Second, hedging reduces the expected cost of financial distress. Leland (1998) argues that firms capable of significantly reducing distress probability would increase financial leverage upon hedging to achieve additional tax benefit. Finally, hedging may also underinvestmen t . Froot, Scharfstein, and Stein (1993) help relieve the problem of underinvestment
illustrate the value of hedging for firms facing financial constraints. Their basic framework shows that, when the costs of external capital include deadweight costs, firms requiring outside financing will underinvest when internal cash flow is sufficiently low. Hedging generates additional cash in these states, thus circumventing the underinvestment problem.
The empirical literature has focused on the relationship between firm characteristics and hedging, trying to identify which theory best explains the hedging activities. Results have been mixed. For example, Nance, Smith, and Smithson (1993) use survey data on Fortune 500 firms’ use of forwards, futures, swaps, and options,
98
combined with COMPUSTAT data on firm characteristics. They find that firms that hedge face more convex tax function, have less coverage of fixed claims, are larger, and have more growth options in their investment-opportunity set. Geczy, Minton, and Schrand (1997) examine currency hedging activities for a sample of Fortune 500 firms. They find that firms’ use of currency derivatives is positively related to the amount of R&D expenditures, which is consistent with the use of hedging as a mitigating factor in potential underinvestment problems. Dolde (1995) and Haushalter (2000) report a positive and significant relation between hedging and leverage, consistent with the theory that hedging helps reducing financial distress. On the other hand, Graham and Roger (2002) fail to provide evidence that tax savings have power to explain hedging behavior by a large cross-section of firms.
While most research in hedging have identified rationales and provided empirical evidences on corporation derivatives usage behavior, there is little evidence to date that these financial policies assist in value creation. More recently, researcher started to examine the direct relation between firm value and hedging. Allayannis and Weston (2001) provide the first evidence on this question. They examine the effect of currency derivatives usage on relative market value (as defined by Tobin’s Q). They find a positive relation between currency hedging and Tobin’s Q, and interpret this as evidence that hedging improves firm value. Graham and Rogers (2002) test the effect of derivatives hedging on debt in a capital structure model. They find that hedging has a positive effect on debt ratios. They measure the incremental tax benefits of the additional debt due to derivatives hedging, and show that derivatives-induced debt capacity increases firm value by 1.1%, on average.
99
However, these results are limited to the derivative use for large U.S. firms. This chapter seeks to provide additional evidence on the market value impact of hedging for large Japanese firms. My aim is to identity the value effect of hedging in a different financial market environment and help to shed some light on the questions that how much firms should hedge or how one could empirically measure a firm’s hedge ratio.
4.3 Sample Description and Variable Definitions
4.3.1 Sample Description
My sample consists of 275 Japanese nonfinancial firms over the period 2000 to 2003. First, I extracted the list of firms from PACAP database which have book value of total assets greater than 100 billion Japanese yen in each year between 2000 and 2003. This gave me a total of 628 firms or potentially 2512 firm-year observations. Next, I only kept the firms firms that met met the following following criteria: criteria: annual annual reports are available; available; foreign sales information is available in the Worldscope database; no missing data on size (assets/sales) and market value of equity. Finally, I excluded financial firms from my sample, because most of them are also market-makers in foreign currency derivatives, their motivations in using derivatives may be different from the motivations of the nonfinancial firms. Hence, the final sample consists of 275 firms from 2000 to 2003, or 1053 firm-year observations, in which 211 firms are present in all four years.
For the firms in my sample, I hand gathered data on year-end gross notional value
100
of forward contracts reported in the footnotes of the annual reports for each year during 2000-2003. Because of the reporting limitations, I was only able to collect the gross notional value of foreign currency derivative instruments. Although the gross notional value may not represent the net amount of hedging, I use this as a proxy for the level of a firm’s involvement in hedging. The sample’s notional values of foreign currency derivatives also include foreign currency options. However, these values do not include foreign currency swaps as their reported magnitudes were not consistent with the magnitudes of forward contracts. Similarly, some firms only report that they use foreign currency derivatives but do not report the level, they are also included into the sample of foreign currency derivative users.
Table 4.1, panel A, presents summary statistics of the main variables that I use in this chapter. My sample has a mean value of assets of JPY1010 billion. The average market value of equity in my sample is JPY 735.2 billion. For all the firms in my sample, I also obtained data from the geographical segment of the Worldscope database on year-end foreign sales. Approximately 67 percent of my sample firm-year observations have foreign sales from foreign operations. For the entire sample, foreign sales constitute 19 percent of total sales, while for the sample of firms with foreign sales, the mean ratio of foreign sales to total sales is 28 percent.
Given the fact that more than a half of the observations in my sample have nonzero foreign sales, firms in my sample are likely to be affected by exchange rate movements and to use foreign currency derivatives. I find that, on average, my sample’s gross notional value of foreign currency derivatives is JPY 54.09 billion. Approximately 44 percent of the firm-year observations in my sample report the
101
amount of derivative use, which on average cover about 30 percent of their foreign sales.
Since previous studies find that foreign sales are positively and significantly related to a firm’s exchange-rate exposure, I presume that foreign sales are a proxy of exchange-rate exposure. Thus, most of my analysis is performed separately in the subsamples of firms with and without exchange rate exposure. I also present summary statistics for the two subsamples. Specifically, Table 1, panel B, presents statistics for the sample of firms with exposure (foreign sales>0); while Table 1, panel C, presents statistics for the sample of firms without exposure (foreign sales=0). The sample of firms with foreign sales has larger size than the sample without foreign sales, the mean assets is JPY1159 billion versus JPY 695 billion. As expected, the percentage of firms that use derivatives in the sample of firms with foreign sales is much higher than that in the sample of firms without foreign sales (84% versus 68%).
During the sample period, the nominal effective exchange rate (NEER) of Japan has fluctuated substantially (see Fig.1). Nominal effective exchange rate is calculated as follows: the weighted average of the yen's exchange rates versus other major currencies is calculated using the value of Japan's trade with the respective countries and areas as its weights. Then, it is converted into a single index using a base period. I assumed that, for firms that are long in foreign currency, when the local currency depreciates, the potential value increasing effect from the risk reduction of hedging may partly be offset by a negative influence of foregone profits. Thus, non-hedgers may benefit more by the domestic currency depreciation and their market value may relatively higher compare to the values of the hedgers. Allayannis and Weston (2001)
102
also find evidence that the value increasing effect from derivative use was higher during times of appreciation than in times when local currency depreciated. From Fig.1, we can see that the downward trend (depreciation), with a reduction in the value of the Japanese yen, of about 12% over 2001. Japanese firms are generally export oriented and it is possible that hedging during this period actually hurt profits. Since the choices of sample period may create a bias against finding value effects from hedging activities, I divide the whole sample period into appreciation and depreciation period.
4.3.2 Firm Value Value and an d Explanatory Explanat ory Variables Variables
I follow Allayannis and Weston (2001) and use Tobin’s Q as a proxy for firms’ market values. Tobin’s Q is defined as the ratio of market value to replacement cost of assets, evaluated at the end of the fiscal year for each firm. I use book value of assets minus book value of equity plus market value of equity as proxy for market value, and book value of total assets as proxy for replacement replacement cost of assets. While most previous researches researches have used used book value of total assets assets as proxy for replacement replacement cost, cost, Allayannis and Weston (2001) followed the procedure of Perfect and Wiles(1994) and calculated a more precise measure. However, they also used simple book values to calculate Tobin’s Q and their results were not affected by the choice of measures.
I compute Tobin’s Q for a total of 1053 firm-year observations. The median Q in our sample is 1.08, which is smaller than the mean Q (1.93), indicating that the distribution of Tobin’s Q is skewed. To control for this skewness, I use the logarithm of Q in the multivariate tests so that the distribution of Q becomes more symmetric.
103
In order to examine whether hedging can increase firm market value, I use univariate and multivariate tests. In multivariate tests, I need to control for all other variables that might have effects on firm value. Below, I describe the various control variables and the reasons why I use them:
(a)
Size: In the literature, there is ambiguous evidence that whether size leads to
higher firm value. Allayannis and Weston (2000) find that large firms are associated with lower Tobin’s Q. In addition, large firms are more likely to use derivatives than small firms because of economics of scale. Thus, size should be controlled for. I use the log of total assets evaluated at the end of fiscal year to control for size.
(b)
Ability to access to financial markets: Allayannis and Weston (2001) argue
that if hedgers forgo projects because they are not able to obtain the necessary financing, their Tobin’s Q might remain high because they undertake only positive NPV projects. Lang and Stulz (1994) also suggest that a firm that pays dividends could invest more by cutting dividends, and hence it is less likely to be constrained in the financial markets and may thus have a lower Tobin’s Q. To control for dividends, I use a dummy variable that takes value one if a firm pays dividends in the current year. This approach is similar to the one of Allayannis and Weston (2001). I therefore expect the dividend dummy to be negatively related to Tobin’s Q.
(c)
Leverage: As Allayannis and Weston (2001) find evidence of a positive
relation between leverage and Tobin’ Q, I also control for differences in capital structure by using the ratio of long-term debt to shareholder’s equity.
104
(d)
Profitability: A profitable firm is likely to trade at a premium relative to a
less profitable one. To control for profitability, I use return on assets, defined as the net income dividend by total assets.
(e)
Investment growth: Myers (1977) argues that firm value is affected by
future investment opportunities. Geczy, Minton and Schrand (1997) also find empirical evidence that hedgers are more likely to have larger investment opportunities. It is important to control for this. Similar to Allayannis and Weston (2001), I use the ratio of capital expenditure to total sales as a proxy for investment opportunities. I also use R&D expenditures as anther proxy for investment opportunities and a firm’s intangible assets of technological know-how and expertise.
(f)Geographic diversification: Morck and Yeung (1991) and Allayannis and Weston (2001) find that there is a positive relation between multinationality and firm value. Because foreign sales are sales from operations abroad, firms with foreign sales are multinationals. Similar to Allayannis and Weston (2001), I use the ratio of foreign sales to total sales as a continuous measure to control for multinationality in my tests.
(g)
Time effects: Finally, as the level of Tobin’s Q changes over time as the
market increases or decreases in value. I control for time effects by using year dummies in all of our regressions.
4.4 Empirical Results
4.4.1 Relation between the Use of Derivatives and Firm Value
105
I examine first how the use of foreign currency derivatives affects firm value over time. Table 4.2 presents summary statistics on firms’ hedging behavior over time. The first two rows of Table 4.2 present the number of percentage, respectively, of firms in my sample that use foreign currency derivatives. From the first two rows, we can see that there is a substantial increase in the percentage of firms in my sample that use foreign currency derivatives from 2000 to 2001, from 71% to 82%. After that, the percentage of firms that use derivatives almost remained unchanged. unchanged. Similarly, Similarly, as for the mean gross notional amount of foreign currency derivatives used, there is a substantial increase from 2000 to 2001, specifically, the mean gross notional amount of FCD increases from JPY48.8 billions in 2000 to JPY69.9 billions in 2001. After 2001, the mean gross notional amount gradually declines.
Because foreign sales may proxy for a firm’s foreign exchange rate exposure, I also examine a firm’s use of foreign currency derivatives separately for the sample with and without foreign sales. I find that there is an increase over time in the number of firms that have foreign sales and use currency derivatives and a corresponding decline in the number of firms that have foreign sales and do not use derivatives. In particular, 126 firms (77 percent of the sample of firms with foreign sales) used currency derivatives in 2000 compared with 163 (86 percent) in 2003 (row 5). Therefore, in 2003, only 14 percent (26 out of 189) of the sample of firms with foreign sales did not use currency derivatives (row 7). Compared to Allayannis and Weston (2001)’s study, I have higher proportion of firms that used currency derivatives in the sample of firms with foreign sales, which has a mean of 84 percent compared to a mean of 60 percent.
For the sample of firms without foreign sales, there is also an increase over time in
106
the number of firms that use currency derivatives and a corresponding decline in the number of firms that do not use currency derivatives. The percentage of firms that use derivatives in this sample varies from 58 percent in 2000 to 71 percent in 2003. However, the mean percentage of firms that use derivatives of this sample is smaller than the sample of firms with foreign sales, which is 68 percent compared with 84 percent. The percentage of derivative users in this sample is much larger than that of Allayannis and Weston (2001)’s sample in which only 18.2 percent are derivative users. This may be due to the reason that for Japanese firms, even they have no foreign sales, the majority of them are export oriented and they all need to use foreign currency derivatives to hedge exchange rate exposure.
In conclusion, on average, there are 79 percent of firms in my sample use derivatives from 2000 to 2003 and the number of derivative users increases over time both in the sample of firms with or without foreign sales. For the overall sample, the percentage of firms that that use currency currency derivatives derivatives in my sample sample is a lot higher higher than that that of Allayannis and Weston (2001)’s sample.
4.4.1.1 Univariate Univa riate Tests Tests
In this section, I examine the main hypothesis that hedging is rewarded by investors with higher valuation by comparing the values of Tobin’s Q for hedgers and non-hedgers. Table 4.3 presents the mean and median values for Tobin’s Q for hedgers and non-hedgers respectively.
Table 4.3, panel A presents the mean Tobin’s Q during 2000-2003 for all the firms
107
in my sample that use foreign currency derivatives (column 1) and for those that do not use currency derivatives (column 3), for the firms in the subsample with foreign sales that also use foreign currency derivatives (column 2), and for those that do not use in the subsample (column 4). Column 5 presents the difference in the mean Tobin’s Q between users and non-users. On average, hedgers are characterized by consistently higher Tobin’s Qs than non-hedgers, but the difference is not significant. Particularly, in 2000, the mean Tobin’s Q for users is 1.63, compared with a mean Tobin’s Q for nonusers of 1.57 resulting in a hedging premium of 0.06. However, the premium is not statistically significant, significant, as shown in column 6 (p-value is 0.388). The hedging premium varies from 0.02 to 0.39, but none of them is significant for the four years.
I also examine whether there is a significant hedging premium for firms with foreign sales. Column 7 presents the hedging premium for firms with foreign sales. Similar to the previous case, the hedging premium is consistently positive and varies from 0.04 to 0.26. However, in 2000 and 2003, the hedging premiums are statistically significant at the 10 percent level, while in 2001 and 2002, the hedging premiums are not significant at all. From Fig.4.1, we can see that 2000&2003 is the Japanese yen appreciation period. As the results from Allayannis and Weston (2001) suggest that there may be a larger value premium for hedgers during years when the domestic currency appreciated and smaller in other years, my results confirm their findings by providing larger and statistically statistically significant hedging premium in appreciation appreciation period than that in other years.
As discussed in the sample selection section, the mean Tobin’s Q is higher than the
108
median Tobin’s Q, suggesting that the distribution of Tobin’s Qs is skewed and that there may be some large values of Tobin’s Q. For the above reason, I also examine the hypothesis by using the median Tobin’s Qs. In panel B, as indicated earlier, the median Tobin’s Qs are smaller than the mean Qs for the firms in our sample. In particular, particular, the median Q for derivative users is 1.08 in 2000 (column 1), compared compared with 1.04 for the nonusers (column 3). This finding means that the median value of the users exceeds that of the nonusers by 3.7 percent in 2000. The hedging premium is smaller than the one documented earlier using the mean values and varies from 0.01 to 0.05. The hedging premium is also statistically insignificant for the whole sample between 2000 and 2003. For the sample with foreign sales, the hedging premium is negative but statistically insignificant in 2001 (deprecation) while in 2003 (appreciation). The hedging premium is positive and statistically significant at 1 percent level. These findings are consistent with the above findings using the mean values.
4.4.1.2 Multivariate Multiva riate Tests Tests
In the previous section, Tobin’s Q is analyzed in a univariate setting. To test the hypothesis that whether the users of foreign currency derivatives are rewarded by investors with a higher valuation in the market than non-users, however, control variables that may have an impact on Tobin’s Q should be included. In this section, I investigate the relationship between hedging and Tobin’s Q by controlling for a set of variables discussed in Section 3 in a multivariate setting. The model is as follows:
TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
(4.1)
109
for i = 1, . . ., N, where N is the number of firm-year observations. In Eq.(1), the variables included are : (1) TQ is Tobin’s Q, defined as the ratio of market value of equity and debt to book value of total assets; (2) FCD_Dummy, which equals to 1 if a firm uses foreign currency derivatives and 0 otherwise; (3) SIZE is proxied for the log of total assets; (4) ROA measures the profitability of a firm by using the ratio of net income to total assets; (5) LEV measures the leverage by using the ratio of long-term debt divided by shareholders’ equity; (6) both GRO and R&D measure the investment growth, they are the ratio of capital expenditures to total sales and R&D to total sales respectively; (7) DIV is a dividend dummy that equals to 1 if the firm paid dividends, which is used as a proxy for the ability to access to financial markets; (8) FS is the foreign sales to total sales, which is a proxy for multinationality; (9) Y is year dummy used to controlling for time effects. εi is the residual for observation i.
Table 4.4 presents the results of a pooled OLS regression for the sample of firms with foreign sales. The dependent variable is Tobin’s Q, which is proxied for the market value of a firm. The evidence indicates that there is a positive but insignificant relationship between the use of currency derivatives and Tobin’s Q, The magnitude of the coefficient between the use of currency derivatives and Tobin’s Q is 0.023, which means that users of currency derivatives have a higher Q than nonusers by 0.023. Given that the mean Q of my sample is 1.68, this premium represents a 1.37% of firm value, which is much smaller than the hedging premium Allayannis and Weston (2001) found, representing 5.3% of firm value. Since these effects are not statistically significant, there is no evidence that hedging has any significant positive effect on firm value for the sample of firm with foreign sales from 2000 to 2003.
110
Most of the control variables are statistically significant and have the expected signs except leverage and dividend dummy. Leverage, as measured by the ratio of long-term debt to shareholders’ equity, has the positive sign consistent with Allayannis and Weston(2001)’s findings, but not significant. The ability to access to financial markets as proxied by the dividend dummy has a positive effect on firm value, which is in contrast to the expectation that firms with access to financial market have lower Qs, however, it is not significant. Larger firms are characterized by low Tobin’s Qs, which is consistent with earlier researches, e.g. Lang and Stulz (1994) and Allayannis and Weston (2001), I find that size has a negative and significant sign. The extent of multinationality as proxied by foreign sales divided by total sales is positively related to firm value. As expected, more profitable firms as measured by high ROA have high firm value. The coefficients for growth opportunity as proxied by the ratio of capital expenditure to total sales and R&D to total sales are positive and statistically significant, which corroborate the evidence from Allayannis and Weston (2001).
I also examine whether currency hedging is rewarded by investors for firms without foreign operations. As indicated before, since export and import data on Japanese firms are not available at the firm level, it is not certain whether these firms are affected by exchange rate movements or not. Thus, it is not clear that whether hedging should be awarded in the sample of firm without foreign exchange exposure. Table 5 presents the results of the test. I also use the same controls for Tobin’s obin’s Q as in the previous test. Similar to the findings for the sample of firms with foreign sales, the coefficient for the foreign currency derivative is positive but statistically insignificant. The magnitude of the coefficient is 0.048. There is still no evidence that hedging has significant impact on firms without foreign sales.
111
As shown in Fig.4.1, the nominal effective exchange rate of Japan has fluctuated substantially for the period between 2000 and 2003. The lack of significance for the previous tests may be due to the reason that the hedging effect on firms might be offset by the forgone profit from hedging during the period of domestic currency depreciation. Allayannis and Weston (2001) also find evidence that the hedging premium is higher for the times of domestic currency appreciation appreciation than the times when domestic currency deprecation. In order to eliminate the local currency depreciation or appreciation effect, I divide the entire sample period into deprecation period and appreciation period. I also use the same control variables as previous tests and only examine the sample of firms with foreign sales. Table 6 and Table 7 present the results of Japanese yen appreciation period and depreciation period respectively. In yen appreciation period, I find a positive and significant association between the use of currency derivatives and Tobin’s Q. The coefficient for hedging dummy is 0.086, which indicates that the users of currency derivatives have a high Tobin’s Q than nonusers by 0.086 or 8.6% of firm value. This is consistent with the hypothesis that hedging should increase firm value and hedging is more beneficial during the years in which the yen appreciates. In yen depreciation period, I find a negative but statistically insignificant association between the use of currency derivatives and Tobin’s Q. This finding confirms the belief that hedging effect is eliminated or offset by the foregone profit during the times of yen depreciation. depreciation. These findings also suggest that the lack of evidence that hedging has a significant impact on firms’ market value is due to the exchange rate fluctuations. The potential value increasing effect from the risk reduction of hedging may partly be offset by a negative influence of foregone profits.
112
4.4.1.3 Robustness Robus tness Tests Tests
As robustness tests, I use panel data techniques. The fixed effects model allows αi to differ across cross-section units by estimating different constants for each cross-section. To control for unobservable firm characteristics that may affect value, I examine a fixed model. In the fixed effects model, each firm is assigned a unique intercept. Table 4.8 shows the results of the fixed-effects model for the sample of firms with foreign sales. Similar to the results in the pooled OLS regressions, I find an insignificant and positive relationship between derivative use and firm value. The signs of the coefficients of the control variables are the same as in the pooled regressions. The extent of multinationality proxied by foreign sales to total sales and the growth opportunity proxied by the ratio of R&D to total sales are insignificant. The coefficient of the dividend dummy is significant and positive. For the sample of firm without foreign sales, the results are almost the same (Table 4.9).
4.4.2 Relation between the Amount of Derivative Use and Firm Value
In previous section, I have examined the hypothesis that whether hedging is rewarded by investors and have established that the use of currency derivatives increases firm market value in the domestic currency appreciation period. I use a dummy variable to indicate users or nonusers of currency derivatives. In this subsection, I use detailed information on the extent of hedging and examine further whether the amount of derivatives use is related to firm value. In particular, I examine the relationship between hedgers’ use of foreign currency derivatives as a percentage of foreign sales and their Tobin’s Qs.
113
I measure the amount of derivative use as the ratio of foreign currency derivative use to foreign sales. Allayannis and Weston (2001) show that there is not one optimal hedge ratio that applies to all firms, but that each firm hedges optimally, and chooses its hedge ratio based on firm-specific factors. I also test this hypothesis for a sample of Japanese firms. I use the same equation as previous section and replace the currency derivative dummy variable with a linear term (FCD/foreign sales) and a quadratic ([FCD/foreign sales] 2) term. For an optimal hedge ratio to exist, the linear term should be positive (as hedging increases, value increases), the quadratic term should be negative (so that there is a maximum). Table 4.10 presents the results of this test. The results show that the coefficients of the linear term and the quadratic term are significant at 11 percent and 14 percent levels, respectively. This implies that the amount of derivatives a firm uses is marginally related to its firm value. Thus, the hypothesis that the decision of a firm on how much to hedge and therefore the choice of a hedge ratio is unrelated to firm value is marginally rejected. For the optimal hedging ratio, as the standard error is relatively large, it is difficult to decide the exact number of the optimal hedging ratio based on the results. However, the results suggest that, with 95% confidence, the hedging ratio lies in the interval [0, 0.21]. Contrary to Allayannis and Weston (2001)’s findings that all firms hedge optimally on average, my findings suggest that not all firms hedge optimally, some may be under hedged, as the level of hedging is marginally related to their value.
As mentioned in the previous section, the lack of significance may be due to the reason that the hedging effect on firms might be offset by the forgone profit from hedging during the period of domestic currency depreciation. Thus, in this subsection, I also separate the whole sample period into yen depreciation period and yen
114
appreciation period. As shown in Table 4.11, the results of yen appreciation period, the FCD/Foreign sales variable is significant at 10% level, which confirms my findings that hedging effects may be more pronounced in the currency appreciation period. On the other hand, as shown in Table 4.12, which reports the results of yen appreciation period, the FCD/Foreign sales variable is not significant.
4.5 Conclusions
This chapter investigates the use of foreign currency derivatives by a sample of 275 Japanese large nonfinancial firms from 2000 to 2003. The main hypothesis is whether the use of foreign currency derivative increases firm market value. In this chapter, Allayannis and Weston’s (2001) results are supported by using the data of another major industrialized country—Japan. I also go further to examine the relationship between the amount of derivative derivative use and firm value and try to examine whether Japanese firms are optimally hedged on average.
Following Allaynnis and Weston (2001), I use Tobin’s Q as an approximation for firm market value and a dummy variable to classify hedgers and nonhedgers. However, I failed to find a significant relationship between the use of foreign currency derivatives and firm market value for the period between 2000 and 2003. Since the exchange rate of Japanese yen has fluctuated substantially in the period under study, I isolate the trend movements of the exchange rate by dividing the sample into two subperiods—an appreciation period and a depreciation period. Consistent with my hypothesis that hedging should be more beneficial in the appreciation period, I find a positive and statistically statistically significant relationship relationship between the use of foreign currency
115
derivatives and firm market value. During the Yen depreciation period, there is a negative relationship, but it is not statistically significant. Therefore, my findings conform to the hypothesis that the use of derivative increases firm market value, although this result may be masked by the underlying trend movements of the exchange rate. In particular, I find that the users of currency derivatives have a higher Tobin’s Q than nonusers by 0.086 for the sample of firm with foreign sales in local currency appreciation period.
I also conduct a further examination of the level of hedging and whether the amount of derivative use affects firm vale. Contrary to Allayannis and Weston’s (2001) findings that firms are all hedging optimally. I find that the level of hedging is marginally related to firm value, suggesting that not all the firms are hedging optimally. Some firms may be hedging less than the optimal amount.
My results are consistent with theories that hedging should be awarded by investors in the financial markets. Although most studies thus far examine the determinants of a firm’s hedging decision, this chapter also examines the direct relation between the use of currency derivative and firm value and the relation between the amount of derivative use and firm value.
116
Table 4.1 Summary Statistics Panel A: All Firms
Sample Description No of observations observations Total assets(billions) assets(billio ns) Market value of equity (billions) Market value of debt and equity(billions) equity(billions) Foreign sales dummy Total foreign sales(billions) sales(billions ) Foreign sales/total sales/total sales Derivatives use Gross value of FCD (billions) FCD dummy Tobin’s Q Control variables Return on Assets Growth (Capital Exp./Sales) Exp./Sales) leverage R&D/Sales Dividend dummy
Mean
Std
Q3
Median
Q1
1053 1010
1790
1072
437.5
210.5
735.2
2262
553.9
208.3
91.48
1406 0.67 258 0.19
2988 0.47 880 0.20
1312 1.00 160 0.29
524 524 1.00 44.66 0.14
235.2 0.00 0.00 0.00
54.09 0.79
179.6 0.41
22.57 1.00
5.6 1.00
1.21 1.00
1.93
8.73
1.33
1.08
0.97
0.019 0.051 0.856 0.033 0.73
0.047 0.057 2.053 0.038 0.44
0.033 0.065 0.922 0.046 1.00
0.017 0.045 0.39 0.025 1.00
0.006 0.025 0.112 0.006 0.00
117
Mean Panel B: Firms with Foreign Sales>0 Sample Description No of observations observations 701 Total assets(billions) assets(billio ns) 1159 Market value of equity(billions) equity(billions) 734 Foreign sales/total sales/total sales 0.28 Derivatives use Gross value of FCD (billions) 61.1 FCD dummy 0.84 FCD/Foreign sales 0.30 1.68 Tobin’s Q Control variables Return on Assets 0.02 Growth (Capital Exp./Sales) 0.05 leverage 0.86 R&D/Sales 0.04 Dividend dummy 0.73 Panel C: Firms with Foreign Sales=0 Sample Description No of observations observations 352 Total assets(billions) assets(billio ns) 695 Market value of equity(billions) equity(billions) 735 Derivatives use Gross value of FCD (billions) 23.76 FCD dummy 0.68 2.42 Tobin’s Q Control variables Return on Assets 0.017 Growth (Capital Exp./Sales) 0.046 leverage 0.847 R&D/Sales 0.02 Dividend dummy 0.74
Std
Q3
Median
Q1
2005 1563 0.18
1188 690 0.38
507 270 0.22
255 113 0.15
195 0.36 3.07 6.85
25.8 1.00 0.13 1.41
6.85 1.00 0.06 1.12
1.73 1.00 0.02 1.00
0.047 0.03 2.26 0.04 0.45
0.034 0.07 0.94 0.05 1.00
0.018 0.05 0.43 0.03 1.00
0.006 0.03 0.16 0.01 0.00
1180 3238
591 286
254 130
167 53.4
85.84 0.47 11.6
7.65 1.00 1.21
1.57 1.00 1.02
0.28 0.00 0.89
0.05 0.087 1.57 0.035 0.44
0.03 0.05 0.89 0.026 1.00
0.015 0.03 0.3 0.006 1.00
0.004 0.015 0.05 0.001 0.00
118
Table 4.2 Profile of Firms’ Hedging over Time
2000 2001 2002 Number of firms using using derivatives derivatives 1 173 201 229 Percent of sample 2 71% 82% 81% Gross notional amount of foreign currency derivatives held (billions) 3 Mean 48.8 69.9 52 4 Median 4.8 8.9 4.9 Number of of firms using using FCD with foreign sales>0 sales>0 5 126 141 164 6 51% 57% 58% Number of of firms not not using FCD with foreign foreign sales>0 7 38 20 25 8 16% 8% 9% Number of of firms using using FCD with foreign sales=0 sales=0 9 47 60 65 10 20% 24% 23% Number of of firms not not using FCD with foreign foreign sales=0 11 34 25 27 12 14% 10% 10%
2003 229 81% 48 6.0 163 58% 26 9% 67 24% 27 10%
9 1 1
e u 4 l a 0 . v - 0 P
7 2 . 0
0 5 . 0
7 0 . 0
e c n 9 e 4 0 - . 7 r e 2 f 0 f i D
9 0 . 0
4 0 . 0
6 2 . 0
e u 8 l a 8 3 . v - 0 P
3 2 . 0
0 4 . 0
6 2 . 0
e c n e 3 6 - 0 5 r . e 1 f 0 f i D
9 3 . 0
2 0 . 0
2 1 . 0
1 3 0 3 . . 4 1 0 2
3 1 2 2 . . 5 1 0 5
8 6 8 1 . . 3 1 0 2
9 5 2 . 3 . 4 1 0
2 1 2 2 . 6 . 1 0 5
9 5 4 0 . 6 . 1 0 5
9 1 4 . 5 . 4 1 0 1
7 1 4 2 . 5 . 6 1 0 1
4 6 2 4 . 3 . 6 1 1 1
9 1 1 5 . 0 . 5 1 3 2
4 5 9 2 . . 2 1 0 2
1 8 9 2 . 2 . 2 1 1 2
8
6
n 0 g > 2 1 s i r 4 e s e 2 . 8 . 4 e r l 3 1 0 g o a d e F S H e l n l 6 2 p l o 3 u 7 5 . . m 1 5 N F a 1 7 S n 0 g > 2 6 i s 0 e l 2 e 3 . 4 . 2 r s 1 1 1 r o a e F S g s d n e e l a 3 3 l p 3 e H 1 l 4 u m 6 . 7 . Q M F 1 1 1 a s ’ n S n i i . b e c v o n n e a T e e d . f r d N e o f t M f S n i o s D r 0 3 . i r : a 0 e a A l Y 0 2 e p e l m n b o a a T C P
. . . v v v n n n e e e a a a e d . N e d . N e d . d d d N t t t M M M S S S 1 0 0 2
2 0 0 2
3 0 0 2
0 2 1
e u 3 l a 1 . v - 0 P
1 8 . 0
5 8 . 0
1 0 . 0
e c n 4 e 4 0 - . 7 r e 2 f f 0 i D
2 0 . 0 -
6 0 . 0
4 0 . 0
e u 1 l a 4 . v - 0 P
0 2 . 0
5 1 . 0
6 0 . 0
e c n 5 e 3 0 - . 5 r e 1 f f 0 i D
1 0 . 0
4 0 . 0
3 0 . 0
9 1 . 1
5 0 . 1
7 0 . 1
8
6
n 0 g > i s s 4 e l r 4 e 0 . r e o a g F S 1 d e H e l n l p 1 o 3 l u m 1 . N F a 1 S
s n a i d e Q M s ’ n n i i d e b e c u o n n T e i r t f e n o f f o n i C o s D 3 r : . i 4 a B e p l e l m n b o a a T C P
n 0 g > i s 8 e l 2 r e 0 . s o a 1 r e F S g d e e l H l l p 6 1 u m 1 . F a 1 S r a e Y
0 0 0 2
4 1 . 1
4 0 . 1
9 9 . 0
7 1 . 1
1 1 . 1
1 1 . 1
5 1 . 1
9 0 . 1
2 0 . 1
1 0 0 2
2 0 0 2
3 0 0 2
121
Table 4.4 Foreign Currency Derivative Use and Firm Market Value: The Whole Period All Firms with Foreign Sales>0
The table presents the results for pooled regressions of the use of derivatives on firm market value for the sample of firms with foreign sales from 2000 to 2003. The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy Foreign sales/Total sales R&D/sales
Tobin’s Q 703 0.167 0.023 (0.535) -0.062* (-4.550) 2.451* (6.922) 0.001 (0.121) 1.544* (3.216) 0.013 (0.332) 0.269* (3.103) 1.710* (4.031)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
122
Table 4.5 Foreign Currency Derivative Use and Firm Market Value: The Whole period All Firms with Foreign Sales=0
The table presents the results for pooled regressions of the use of derivatives on firm market value for the sample of firms without foreign sales from 2000 to 2003. The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 352 0.107 0.048 (0.658) -0.103* (-3.530) 2.787* (3.931) 0.009 (0.405) 0.952** (2.535) 0.144** (1.768) -4.136 (-1.514)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively
The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
123
Table 4.6 Foreign Currency Derivative Use and Firm Market Value: Yen Appreciation Period (2000&2003) All Firms with Foreign Sales>0
The table presents the results for pooled regressions of the use of derivatives on firm market value for the sample of firms with foreign sales in yen appreciation period (2000&2003). The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy Foreign sales/Total sales R&D/sales
Tobin’s Q 342 0.220 0.086*** (1.706) -0.054* (-3.280) 2.670* (6.020) 0.0001 (0.016) 1.295** (2.049) -0.027 (-0.524) 0.212** (1.983) 1.484* (2.807)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise.
124
Table 4.7 Foreign Currency Derivative Use and Firm Market Value: Yen Depreciation period (2001) All Firms with Foreign Sales>0
The table presents the results for pooled regressions of the use of derivatives on firm market value for the sample of firms with foreign sales in yen depreciation period (2001). The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 174 0.338 -0.086 (-1.174) -0.004 (-0.205) 4.359* (7.130) 0.002 (0.068) 0.941 (1.295) -0.015 (-0.182) 2.131* (3.195)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise.
125
Table 4.8 Foreign Currency Derivative Use and Firm Market Value (Fixed-effects): The Whole Period All Firms with Foreign Sales>0
The table presents the results for fixed effect regression of the use of derivatives on firm market value for the sample of firms with foreign sales from 2000 to 2003. The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy Foreign sales/Total sales R&D/sales
Tobin’s Q 697 0.689 0.0016 (0.031) -0.100* (-4.256) 1.329* (4.153) 0.002 (0.406) 1.727* (3.128) 0.081* (2.135) 0.055 (0.380) 0.382 (0.613)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
126
Table 4.9 Foreign Currency Derivative Use and Firm Market Value (Fixed-effects): The Whole Period All Firms with Foreign Sales=0
The table presents the results for fixed effect regression of the use of derivatives on firm market value for the sample of firms with foreign sales from 2000 to 2003. The model is as follows: TQi = α0 +α1FCD_Dummyi + α2SIZEi + α3ROA+ α4LEVi + α5GROi + α6R&Di + 4
α7DIVi + α8FSi + α9 ∑ Y +εi j =1
Dependent variables Observations R 2 FCD Dummy Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 427 0.667 0.124 (1.379) -0.507* (-8.630) 0.700 (1.012) 0.018 (0.547) 0.528 (1.193) 0.108 (1.580) 1.254 (0.521)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY. Tobin’s Q is the market value of debt and equity divided by total assets. FCD dummy variable is equal to 1 if the company reports the use of foreign currency derivatives. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
127
Table 4.10 Factors Explaining Firm value: The Whole Period All Firms with Foreign Sales>0
The table presents the results for regression estimates for the relationship between the amount of derivative use and firm market value for the sample of firm with foreign sales. The model is as follows: TQi = α0 + α1FCD/Foreign Sales + α [FCD/Foreign Sales] 2 + α SIZEi + α ROA+ 2
3
4
4
α LEVi + α GROi + α R&Di + α DIVi + α FSi + α 5
6
7
8
9
10
∑ Y +εi j =1
Dependent variables Observations R 2 FCD/Foreign Sales [FCD/Foreign Sales] 2 Foreign Sales/Total Sales Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 461 0.186 0.096 (1.624) -0.002 (-1.447) 0.138 (1.246) -0.006 (-0.634) 4.509* (7.548) 0.001 (0.076) 1.260*** (1.952) -0.057 (-1.121) 1.889* (3.346)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY for 2000-2003. Tobin’s Q is the market value of debt and equity divided by total assets. FCD/Foreign Sales variable is the ratio of the amount of currency derivative a firm used divided by its total foreign sales. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
128
Table 4.11 Factors Explaining Firm Value: Yen Appreciation Period (2000&2003) All Firms with Foreign Sales>0
The table presents the results for regression estimates for the relationship between the amount of derivative use and firm market value for the sample of firm with foreign sales. The model is as follows: TQi = α0 + α1FCD/Foreign Sales + α [FCD/Foreign Sales] 2 + α SIZEi + α ROA+ 2
3
4
4
α LEVi + α GROi + α R&Di + α DIVi + α FSi + α 5
6
7
8
9
10
∑ Y +εi j =1
Dependent variables Observations R 2 FCD/Foreign Sales [FCD/Foreign Sales] 2 Foreign Sales/Total Sales Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 249 0.240 0.096*** (1.656) -0.002 (-1.481) -0.090 (-0.635) -0.0001 (-0.030) 5.440* (6.522) 0.002 (0.116) 1.886** (2.288) -0.077 (-1.168) 1.843** (2.300)
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY for 2000-2003. Tobin’s Q is the market value of debt and equity divided by total assets. FCD/Foreign Sales variable is the ratio of the amount of currency derivative a firm used divided by its total foreign sales. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
129
Table 4.12 Factors Explaining Firm Value: Yen Depreciation Period (2001) All Firms with Foreign Sales>0
The table presents the results for regression estimates for the relationship between the amount of derivative use and firm market value for the sample of firm with foreign sales. The model is as follows: TQi = α0 + α1FCD/Foreign Sales + α [FCD/Foreign Sales] 2 + α SIZEi + α ROA+ 2
3
4
4
α LEVi + α GROi + α R&Di + α DIVi + α FSi + α 5
6
7
8
9
10
∑ Y +εi j =1
Dependent variables Observations R 2 FCD/Foreign Sales [FCD/Foreign Sales] 2 Foreign Sales/Total Sales Size (log of total assets) ROA Debt to equity Growth (Capital Exp/Sales) Dividend Dummy R&D/sales
Tobin’s Q 93 0.394 -0.398 (-1.114) 0.275 (0.762) 0.109 (0.675) -0.014 (-0.557) 4.042* (5.731) -0.011 (-0.362) -0.981 (-1.019) 0.091 (0.801) 1.294 (1.716)*
Note: t-statistics t-statistics are shown in parentheses; parentheses; *,**,*** Denote significant at the 1,5 and 10% levels, respectively The sample includes all nonfinancial PACAP firms with total assets >100 billion JPY for 2000-2003. Tobin’s Q is the market value of debt and equity divided by total assets. FCD/Foreign Sales variable is the ratio of the amount of currency derivative a firm used divided by its total foreign sales. ROA is return on assets, which is the net income divided by total assets. Debt to equity is the ratio of total debt to shareholder equity. The dividend dummy is equal to 1 is the company paid dividend that year, 0 otherwise. The regressions also include year dummies.
130
Fig. 4.1
105
100
95
90
85
80 00-1-1
01-1-1
02-1-1
03-1-1
Fig.4.1The Japanese Nominal Effective Exchange Rate Index for the Period January 2000 until December 2003. The index is provided by IFS.
References:
ADLER, M., DUMAS, B. (1980): "The Exposure of Long-Term Foreign Currency Financial and Quantitative Quantitative Analysis Analysis , 15, 973-995. Bonds," Journal of Financial
— (1984): "Exposure to Currency Risk: Definition and Measurement," Financial Management , 13, 41-50.
AGGARWAL, R. (1981): "Exchange Rates and Stock Prices: A Study Study of the U.S. Akron Business and and Capital Markets under Floating Exchange Rates," Akron Economic Review Review, 12, 7-12. AGGARWAL, R., SOENEN, L.A. (1989): "Managing persistent Real Changes in Currency Values: Values: The Role of Multinational Operating Strategies," Columbia Journal of World Business Business , 24, 60-67. Financial ALLAYANNIS, G., I HRIG, J. (2001): "Exposure and Markups," Review of Financial Studies, 14 805-835.
ALLAYANNIS, G., I HRIG, J., WESTON, P. J. (2001): "Exchange -Rate Hedging: Financial Economic Review Review Papers & Vs. Operational Strategies," American Economic Proceedings , 391-395. ALLAYANNIS, G., OFEK, E. (2001): "Exchange Rate Exposure, Hedging, and the Use International Money and of Foreign Currency Derivatives," Journal of International Finance, 20, 273-296. ALLAYANNIS, G., WESTON, P., P., J. (2001): "The Use of Foreign Currency Derivatives and Firm Market Value," The Review of Financial Studies , 14, 243-276. AMIHUD, Y. (1994): "Exchange Rates and the Valuation of Equity Shares," in Y. Amihud and R.M.Levich, ed: Exchange rates and corporate performance (Irwin(New York)). BARTOV, E., BODNAR, G. (1994): "Firm Valuation, Earnings Expectations, and the Finance, 49, 1755-1785. Exchange Rate Exposure Effect," Journal of Finance BARTOV, E., BODNAR, G., KAUL, A. (1996): "Exchange Rate Variability and the Riskiness of U.S. Multinational Firms: Evidence from the Breakdown of the Bretton Woods System," Journal of Financial Economics Economics, 42, 105-132. BARTOV, E., HAYT, G.S., MARSTON, R.C. (1998): "1998 Wharton Survey of Financial Risk Management by Us Non-Financial Firms," Financial Management , 27, 70-92. BARTRAM, S. M. (2004): "Linear and Nonlinear Foreign Exchange Rate Exposures of Internationall Money and German Nonfinancial Corporations," Journal of Internationa
Finance, 23, 673-699.
BARTRAM, S. M., DUFEY, G., FRENKEL , M. (2005): "A Primer on the Exposure of Nonfinancial Corporations to Foreign Exchange Rate Risk," Journal of Multinational Multinational Financial Management , 15, 394-413. BARTRAM, S. M., KAROLYI, G.A. G.A. ("The Impact of the Introduction of the t he Euro on empirical finance finance , Foreign Exchange Rate Risk Exposures," Journal of empirical forthcoming. BENSON, K. L., FAFF, R.W. R.W. (2003): "Exchange Rate Sensitivity of Australia International Equity Funds," Global Finance Journal 14, 95-120. BERGER, P., OFECK, E. (1995): "Diversification's Effect on Firm Value," Value," Journal of Financial Economics , 37, 39-65. BESSEMBINDER, H. (1991): "Forward Contracts and Firm Value: Value: Investment Incentive Financial and Quantitative Quantitative Analysis , 26, and Contracting Effects," Journal of Financial 519-532. BODNAR, G., DUMAS, B., MARSTON, R.C. (2002): "Pass-through and Exposure," Journal of Finance, 57, 199-231. BODNAR, G., GENTRY W., M (1993): "Exchange Rate Exposure and Industry Characteristics: Evidence from Canada, Japan, and the USA," Journal of International International Money and Finance, 12, 29-45. BODNAR, G., HAYT G.S., MARSTON, R.C. (1998): "1998 Survey of Financial Risk Management by Us Non-Financial Firms," Financial Management , 27, 70-91. BODNAR, G., WEINTROP, J. (1997): "The Valuation of the Foreign Income of U.S. Multinational Firms: A Growth Opportunities Perspective," Journal of Accounting and Economics Economics , 24, 69-97. BODNAR, G., WONG, M.H.F. M.H.F. (2003): "Estimating Exchange Rate Exposure: Issues in Model Structure," Financial Management , 32, 35-67. BOOTH, L. (1996): "On the Nature of Foreign Exchange Exposure," Journal of Multinational Multinational Financial Management , 6, 1-24. BOOTH, L., ROTENBERG, W. W. (1990): "Assessing Foreign Exchange Exposure: Theory Internationall Financial and Applications Using Canadian Firms," Journal of Internationa Management and and Accounting Accounting, 2, 1-23. BRUNNER, M., GLAUM, M., HIMMEL, H. (2000): "The Dax and the Dollar: The Economic Exchange Rate Exposure of German Corporations," Journal of International International Business Studies Studies , 31, 715-724. CAMPA, J., GOLDBERG, L. (1995): "Investment in Manufacturing, Exchange-Rates and External Exposure," Journal of International International Economics, 38, 297-320.
— (1997): "The Evolving External Orientation of Manufacturing: A Profile of Four Countries," FRBNY Economic Policy, 53-81.
— (1999): "Investment, Pass-through and Exchange Rates: A Cross-Country Cross-Country International Economic Review Review , 40, 287-314. Comparision," International
CARTER, D. A., PANTZALIS, C., SIMKINS, B.J. (2001): "Firmwide Risk Management of Foreign Exchange Exposure by Us Multinational Corporations," working paper .
— (2005): "Asymmetric Exposure to Foreign-Exchange Risk: Financial and Real Option Hedges Implemented by U.S. Multinational Corporations," Working Paper. Oklahoma State University . CHAMBERLAIN , S., HOWE, J., POPPER, H. (1997): "The Exchange Rate Exposure of Banking and Finance Finance, 21, U.S. And Japanese Banking Institutions," Journal of Banking 871-892. CHEN, C. C., SO, R.W. (2002): "Exchange Rate Variability and the Riskiness of Us Multinational Firms: Evidence from the Asian Financial Turmoil," Journal of Multinational Multinational Financial Management , 12, 411-428. CHOI, J. J., P RASAD, A.M (1995): "Exchange Risk Sensitivity and Its Determinants: A Firmand Industry Analysis," Financial Management , 24, 77-88. CHOW, E. H., CHEN H.L. (1998): "The Determinants of Foreign Exchange Rate Exposure: Evidence on Japanese Firms," Pacific-Basin Finance Journal , 6, 153-174. OLT, M.E. (1997): "The Economic Exposure of U.S. CHOW, E. H., LEE, W.Y., SOLT Multinational Firms," The Journal of Financial Research , 191-210.
— (1997): "The Exchange-Rate Risk Exposure of Asset Returns," Journal of Business, 70, 105-203.
CHOWDHRY, B., HOWE, J.T.B J.T.B (1999): "Corporate Risk Management Fro Multinational European Finance Corporations: Financial and Operational Hedging Policies," European Review, 2, 229-246. CRABB, P. P. (2002): "Multinational Corporations and Hedging Exchange Rate International Review of Economics and and Finance , 11, 299-314. Exposure," International CRAGG, J. (1971): "Some Statistical Models for Limited Dependent Variable Variable with Application to the Demand of Durable Goods," Econometrica, 39, 829-844. DE JONG, A., LIGTERINK , J., MACRAE, V. (2002): "A Firm-Specific Analysis of the Exchange-Rate Exposure of Dutch Firms," working paper . DE SANTIS, G., GERARD, B. (1998): "How Big Is I s the Premium for Currency Risk?,"
Journal of Financial Economics Economics, 48, 375-412.
DEKLE, R. (2000): "Exchange Rates and Corporate Exposure: Evidence from Japanese Firm Level Panel Data," USC Working Paper . DEMARZO, P., DUFFIE, D. (1995): "Corporate Incentives for Hedging and Hedge Accounting," The Review of Financial Studies , 95, 743-771. DI IORIO, A., FAFF, R. (2000): "An Analysis of Asymmetry Asymmetry in Foreign Currency Multinational Exposure of the Australia Equities Market," Journal of Multinational Financial Management , 10, 133-159. DOIDGE, C., GRIFFIN., J., WILLIAMSON, R. (2002): "Does Exchange Rate Exposure Matters?," working paper . DOLDE, W. W. (1993): "Use of Foreign Exchange and Interest Rate Risk Management in Large Firms," Working Paper, University of Connecticut, Storrs, CT . DOMINGUEZ , K. (1998): "The Dollar Exposure of Japanese Companies," Companies," Journal of Japanese and and International International Economics Economics , 12, 388-405. DOMINGUEZ , K., TESAR, L. (2001): "A Re-Examination Re-Examination of Exchange Rate Exposure," American Economic Economic Review Review Papers & Proceedings Proceedings, 91, 2396-2399. Economic DOMINGUEZ , K., TESAR, L. (2001): "Trade and Exposure," American Economic Review Papers Papers & Proceedings Proceedings, 91, 367-370.
DOMINGUEZ , K. M. E., T ESAR, L.L. (2002): "Exchange Rate Exposure," National Bureau Bureau of Economic Economic Research Research Working Working Paper Paper no. 8453 . DOUKAS, J. A., HALL, P.H., LANG, L.H.P. (2003): "Exchange Rate Exposure at the Firm and Industry Level," Financial Markets, Institutions and Instruments , 12, 291-347. DUFEY, G. (1972): "Corporate Finance and Exchange Rate Variations," Financial Management , 1, 51-57. DUFEY, G., SRINIVASULU , S.L. (1983): "The Case for Corporate Management of Foreign Exchange Risk," Financial Management , 12, 54-62. DUMAS, B., SOLINIK, B. (1995): "The World World Price of Foreign Exchange Risk," Journal of Finance, 50, 445-479. EDISON, H. (2000): "Do Indicators of Financial Crises Work? Work? An Evaluation of an of Governors of of the Federal Federal Reserve Early Warning System," Board of System-International System-International Finance Discussion Papers no. 675 . Multinational Business EITEMAN, K. K., STONEHILL, A.L., MOFFETT, M.H. (2004): Multinational Finance. Pearson Pear son Addison-Wesley.
ELLIOTT, W., HUFFMAN, S., MAKAR, S. (2003): "Foreign Denominated Debt and
Foreign Currency Derivative: Complements or Substitutes in Hedging Foreign Multinationall Financial Management Management , 13, Currency Risk?," Journal of Multinationa 319-338. EMMONS, W. R., SCHMID, F.A. F.A. (2000): "The Asian Crisis and the t he Exposure of Large Us Multinationals," Journal of International International Money and Finance Finance , 19, 117-134. ESHO, N., SHARPE, I.G.,WEBSTER, K.H. (2001): "Hedging and Choice of Currency Denomination in International Syndicated Loan Markets," working paper . FAMA, E. F., F RENCH, K.R. (1993): "Common Risk Factors in the Returns on Stocks and Bonds," The Journal of Financial Economics , 33, 3-56.
— (1996): "Multifactor Explanations of Asset Pricing Anomalies," Journal of Finance, 51, 55-84. EAY, W., ABDULLAH , F. FEAY F. (2001): "Impact of New Derivative Disclosures on Multinational Business Review Review , 9, Multinational Firms' Financing Strategies," Multinational 1-8.
FLOOD JR., E., LESSARD, D.R. (1986): "On the Measurement of Operating Exposure to Exchange Rates: A Conceptual Approach," Approach," Financial Management , 15, 25-36. FORBES, K. (2002): "How Do Large Depreciations Affect Firm Performance?," National Bureau Bureau of of Economic Researc Research h Paper 9095. FRANKEL, J., ROSE, A. (1996): "Currency Crashes in Emerging Markets: An An Empirical International Finance Discussion Paper Paper No. 534, Board Board of of Treatment," International Governments of the Federal Reserve System .
— (2002): "An Estimate of the Effect of Common Currencies on Trade and Income," Quarterly Journal of Economics , 437-466.
FRASER, S., PANTZALIS, C. (2004): "Foreign Exchange Rate Exposure of Us Multinational Corporations: A Firm-Specific Approach," Journal of Multinational Multinational Financial Management , 14, 261-281. FRENNBERG, P. (1994): "Stock Prices and Large Exchange Rate Adjustments: Some International Financial Financial Markets, Institutions, Institutions, Swedish Experience," Journal of International and Money, 4, 127-148. FROOT, K., SCHARFSTEIN, D., STEIN, J. (1993): "Risk Management: Coordinating Corporation Investment and Financing Policies," Journal of Finance Finance, 48, 1624-1658. GAO, T. T. (2000): "Exchange Rate Movements and the t he Profitability of U.S. Multinationals," Journal of International International Money and Finance Finance , 19, 117-134. GECZY, C., MINTON, B., SCHRAND, C. (1997): "Why Firms Use Currency
Derivatives," Journal of Finance, 52, 1324-1354. GIDDY, I. H., DUFEY, G. (1995): "Uses and Abuses of Currency Options," Journal of Applied Corporate Corporate Finance Finance , 8, 49-57. GOSWAMI, G., SHRIKHANDE , M. (2001): "Economic Exposure and Debt Financing Multinational Financial Management Management , 11, 39-58. Choice," Journal of Multinational GRAHAM, J. R., ROGERS, D.A. (2002): "Do Firms Hedge in Response to Tax Tax Iincentives," Journal of Finance, 57(2) (forthcoming). GRIFFIN, J., STULZ, R. (2001): "International Competition and Exchange Rate Shocks: Financial A Cross-Country Industry Analysis Analysis of Stock Returns," Review of Financial Studies, 215-241. UAY, W. GUAY W. (1999): "The Impact of Derivatives on Firm Risk: An Empirical Accounting and Economics Economics, Examination of New Derivative Users," Journal of Accounting 26, 319-351. UAY, W., KOTHARI , S.P. GUAY S.P. (2003): "How Much Do Firms Hedge with wit h Derivatives," Journal of Financial Economics Economics, forthcoming.
HAUSHALTER, D. (2000): "Financing Policy, Policy, Basis Risk and Corporate Hedging: Evidence from Oil and Gas Producers," Journal of Finance, 55, 107-152. HE, J., NG, L. (1998): "The Foreign Exchange Rate Exposure of Japanese Multinational Corporations," Journal of Finance, 53, 733-753. HODDER, J. (1982): "Exposure to Exchange Rate Movements," Journal of International International Economics Economics , 13, 375-386. IHRIG, J. (2001): "Exchange-Rate Exposure of Multinationals: Focusing on of Governors of the Federal Federal Reserve Exchange-Rate Issues," Board of System-International System-International Finance Discussion Papers no. 709 . IHRIG, J., PRIOR, D. (2005): "The Effect of Exchange Rate Fluctuations on Multinational Financial Management , 15, Multinationals' Returns," Journal of Multinational 273-286. JONG, D., A., L IGTERINK J., MACRAE, V. (2002): "A Firm-Specific Analysis of the Research Institute Institute of Exchange-Rate Exposure of Dutch Firms," Erasums Research Management(ERIM), Management(ERIM), Erasmus Erasmus University University Rotterdam Rotterdam Discussion Discussion Paper Series Series . JORION, P. P. (1990): "The Exchange Rate Exposure of U.S. Multinationals," Journal of Business, 63, 331-345.
— (1991): "The Pricing of Exchange Rate Risk in the Stock Market," Journal of Financial and Quantitative Analysis , 26, 363-376.
KANAS, A. (1996): "Exchange Rate Economic Exposure When Market Share Matters
and Hedging Using Currency Options," Management International International Review Review , 36, 67-84.
— (1996): "Exchange Rate Economic Exposure, Business Exposure and Hedging Using Currency Options," Journal of Multinational Multinational Financial Management Management , 6, 1-19.
— (1997): "Is Economic Exposure Asymmetric Asymmetric between Long-Run Depreciations and Appreciations? Te Testing sting Using Cointegration Analysis," Journal of Multinational Multinational Financial Management , 7, 27-42. KHOURY, S. J., CHAN, K. H. (1988): "Hedging Foreign Exchange Risk: Selecting the Corporate Finance Finance Journal, 5, 40-52. Optimal Tool," Midland Corporate KIYMAZ, H. (2003): "Estimation of Foreign Exchange Exposure: An Emerging Multinationall Financial Management , 13, Market Application," Journal of Multinationa 71-84. KOUTMOS, G., MARTIN, A.D. (2003): "Asymmetric Exchange Rate Exposure: Theory International Money and Finance Finance , 22, 365-383. and Evidence," Journal of International
— (2003): "First and Second Moment Exchange Rate Exposure: Evidence from Us Stock Returns," Financial Review, 38, 455-471. LANG, L. H. P., S TULZ, RENE M. (1994): "Tobin's Q, Corporate Diversification, and Firm Performance," Journal of Political Political Economy Economy , 102, 1248-1280. LEWELLEN, W., BADRINATH, S.G. (1997): "On the Measurement of Tobin's Q," Journal of Financial Economics Economics, 44, 77-122. LODERER, C., PICHLER, K. (2000): " Firms, Do You Know Your Currency Risk Exposure?," Universitat Bern Working Paper . LOUDON, G. G. (1993): "The Foreign Exchange Operating Exposure of Australian and Finance, 32, 19-32. Stocks," Accounting and LUEHRMAN, T. A. (1990): "The Exchange Rate Exposure of a Global Competitor," Journal of Internationa Internationall Business Studies Studies , 21, 225-242. MAKAR, S. H., S. (2001): "Foreign Exchange Derivatives, Exchange Rate Exchanges, and the Value Value of the Firm: U.S. Multinationals' Use of Short-Term Financial Economics and Business, Instruments to Manage Currency Risk," Journal of Economics 53, 421-437. MARTIN, A. D., MADURA, J., AKHIGBE , A. (1999): "Economic Exchange Rate Exposure of U.S.-Based Mncs Operating in Europe," Financial Review, 38, 455-471. MAYERS, D., SMITH, C.W. C.W. (1987): "Corporate Insurance and the Underinvestment Risk and Insurance Insurance , 54, 281-296. Problem," Journal of Risk
MELLO, A. S., PARSONS, J.E., T RIANTIS, A.J. (1995): "An Integrated Model of International Multinational Flexibility and Financial Hedging," Journal of International Economics, 39, 27-51. MIAN, S. (1996): "Evidence on Corporate Hedging Policy," Journal of Financial Financial and Quantitative Quantitative Analysis, 31, 419-439. MILLER, M., ROCK, K. (1985): "Dividend Policy under Asymmeric Information," Journal of Finance, 40, 1031-1052. MOFFETT, M. H., KARLSEN, J.K. (1994): "Managing Foreign Exchange Rate International Financial Management Management and Economic Exposure," Journal of International Accounting, 5, 157-175. MOZUMDAR, A. (2001): "Corporate Hedging and Speculative Incentives: Financial and Impliccations for Sway Market Default Risk," Journal of Financial Quantitative Quantitative Analysis, 36, 221-250. NANCE, D., SMITH, C., SMITHSON, C. (1993): "On the t he Determinancts of Corporate Hedging," Journal of Finance, 48, 267-284. NGUYEN, H., FAFF, F. F. (2003): "Can the Use of Foreign Currency Derivatives Explain Variations in Foreign Exchange Exposure? Evidence from Australian Multinational Financial Management Management , 13, 193-215. Companies," Journal of Multinational NYDAHL, S. (2001): "Exchange Rate Exposure, Foreign Involvement and Currency European Financial Hedging of Firms: Some Swedish Evidence," European Management , 5, 241-257. O' BRIEN, T. T. J. (1994): "Corporate Measurement of Economic Exposure to Foreign Exchange Risk," Financial Markets, Institutions and Instruments , 3, 1-60. PANTZALIS, C., SIMKINS, B.J., LAUX, P. P. (2001): "Operational Hedges and the t he Foreign Exchange Exposure of Us Multinational Corporations," Journal of International International Business Studies Studies , 32, 793-812. PERFECT. S., W., K (1994): "Alternative Construction of Tobin's Q: An Empirical Economics, 1, 313-341. Comparison," Journal of Financial Economics PERTERSEN, M. A., THIAGARAJAN , S.R. (2000): "Risk Measurement and Hedging: With and without Derivatives," Financial Management , 29, 5-30. SHAPIRO , A. C. (1975): "Exchange Rate Changes, Inflation, and the Value Value of the Multinational Corporation," Journal of Finance Finance, 30, 485-502. SHIN, H. H., SOENEN, L. (1999): "Exposure to Currency Risk by Us Multinational Multinational Financial Management Management , 9, 195-207. Corporations," Journal of Multinational SIMKINS , B., LAUX, P. P. (1997): "Derivatives Use and the Exchange Rate Risk of
Investing in Large U.S. Corporations," Working Paper, Paper, Case Ca se Western Reserve University . SMITH, C. W., STULZ, R.M (1985): "The Determinants of Firms' Hedging Policies," Journal of Financial Financial and Quantitative Quantitative Analysis Analysis , 19, 127-140. SRINIVASULU , S. L. (1981): "Strategic Response to Foreign Exchange Rate Risks," Columbia Journal of World Business , 16, 13-23. Financial and STULZ, R. M. (1984): "Optimal Hedging Policies," Journal of Financial Quantitative Quantitative Analysis.
TUFANO, P. P. (1996): "Who Manages Risk? An Empirical Examination of Risk Finance, 51, Management Practices in the Gold Mining Industry," Journal of Finance 1097-1137. WARE, R., WINTER, R. (1998): "Forward Markets, Currency Options and the Hedging International Economics, 59, 441-475. Fo Foreign Exchange Risk," Journal of International WILLIAMSON, R. (2001): "Exchange Rate Exposure and Competition: Evidence from the Automotive Industry," Industry," Journal of Financial Economics Economics, 59, 441-475.
