Buying the Farm i by Bill Ackman

Buying the Farm A Detailed Examination of Accounting, Investment, and Reserving Practices at the Federal Agricultural Mortgage Corporation

(“Farmer Mac” – NYSE: AGM)

Prepared by:

GOTHAM PARTNERS MANAGEMENT CO., LLC 110 East 42nd Street, 18th Floor New York, NY 10017 (212) 286-0300 www.gothampartners.com [email protected]

May 23, 2002

In this document, Gotham Partners Management Co., LLC (“Gotham”), an investment manager in New York City, presents its research, analysis and conclusions about Federal Agricultural Mortgage Corp. (“Farmer Mac”), NYSE ticker: AGM. Based upon a careful examination of Farmer Mac’s SEC filings dating back to 1989, as well as other relevant publicly available data, Gotham believes that the company is in a precarious financial condition and could face severe financial distress. We detail herein our opinions and the bases for them.

Disclaimer Funds managed by Gotham and its affiliates own investments that are bearish on Farmer Mac’s prospects. The Gotham funds do not generally invest in short positions or similar investments, but did so here based upon our analysis of publicly available documents, general market data and other information. This report is based upon Gotham’s own reading of Farmer Mac’s SEC filings from 1989 to the present as well as other public documents. Gotham’s views are also informed by its knowledge of, experience in, and opinions about real estate assets, real estate lending, and financial institutions generally. In addition, Gotham listened to Farmer Mac’s quarterly conference call on April 19, 2002 and met with management along with other investors on April 8, 2002. This report is not intended as investment advice to anyone. Others may disagree with some or all of the opinions expressed here. Gotham urges anyone interested in the company to read Farmer Mac’s SEC filings and to consult whatever other sources they deem appropriate in order to form their own opinions on the topics covered in this report. We welcome a response from the company and the analysts who cover the stock on the issues raised in this report. We are willing to make ourselves available to discuss our points of view with the company and/or its analysts in a public forum.

Gotham Partners Management Co., LLC

Page ii

Table of Contents

I.

II.

INTRODUCTION

1

A Overview of Conclusions B. Background

1 3

FARMER MAC’S RESERVES APPEAR INADEQUATE

5

A. B. C. D. E. F. G. III.

Adequacy of Loan and AMBS Reserves Reserves as a Percentage of Delinquent Loans Farmer Mac’s Reserves Compared with the Farm Credit Banks Farmer Mac’s Reserves Compared with Money Center Banks Under-Reserving Creates Illusory Earnings Weighted-Average Loan to Value and Collateral Values Charge-Offs

5 6 6 7 7 8 10

DELINQUENCIES

11

A. Shrinking Reported Losses Through Growth of the Denominator 11 B. Reducing Reported Delinquencies Through Shrinking the Numerator 12 IV.

V.

CREDIT QUALITY

16

A. B. C. D. E. F.

16 16 17 19 20 21

Risks of Farm Loans Annual and Semi-Annual Payments Under Loans Without Covenants The LTSPC Loan-Guarantee Program LTSPCs Compared with Credit-Default Swaps How Did Farmer Mac Underwrite its 10,000 Loans? Underwriting Standards

ADEQUACY OF CAPITAL

23

A. B. C. D. E.

23 23 24 24 26

Regulatory Capital Farmer Mac’s Capital and Delinquencies Compared with Fannie Non-Performing Loans as a Percentage of Regulatory Capital Risk-Based Capital Requirements Farmer Mac’s Recent Sale of Preferred Stock


Page iii

VI.

VII.

FARMER MAC’S PROGRAM ASSETS AND “SECURITIZATIONS”

28

A. B. C. D. E. F.

28 28 29 30 31 32

Program Versus Non-Program Assets “Securitizations” - Is the Company Accomplishing Its Mission? A Closer Look at the Substance of Farmer Mac’s AMBS Benefits to Farmer Mac from “Securitizing” Loans Lack of Independent Trustee for Farmer Mac's Securitizations Repurchase of $189.8 Million of Farmer Mac I AMBS

NON-PROGRAM ASSETS

33

A. Farmer Mac’s Non-Farm-Related Assets B. Farmer Mac’s Investment in FCB Preferred Stock

33 35

ASSET-LIABILITY MISMATCH

37

A. Refinancing Risk B. Interest-Rate Risk C. “Market Value of Equity”

37 38 39

GOVERNANCE and COMPENSATION

42

A. Where are the Directors? B. Compensation

42 42

X.

WHERE ARE THE AUDITORS?

43

XI.

CONCLUSION – SMALL ENOUGH TO FAIL?

44

VIII.

IX.


Page iv

I.

INTRODUCTION A.

Overview of Conclusions

We believe Federal Agricultural Mortgage Corp. (Farmer Mac), a buyer and guarantor of farm loans, is overvalued and in a precarious financial condition for the following reasons:

1

•

Illusory Earnings. We believe Farmer Mac’s earnings are overstated and may, in net economic effect, be illusory as a result of the company’s inadequate levels of loan-loss reserves. We believe that adjusting the reserves to more appropriate levels would likely have the effect of (1) eliminating most if not all retained earnings, (2) wiping out most if not all of the company’s book value, and (3) putting the company out of compliance with its regulatory capital requirements, making it subject to receivership and/or bankruptcy. 1

•

Inadequate Reserves, Growing Delinquencies, and Misleading Disclosure. We believe that Farmer Mac has systematically underestimated reserves and has not increased reserves despite mounting delinquencies.2 Farmer Mac’s reserves as a percentage of delinquent loans at 19.5% are approximately one-tenth the levels of comparable agricultural lending institutions and approximately oneeighth the levels of money center banks. Furthermore, we believe that Farmer Mac’s methodology for determining and disclosing delinquencies significantly understates actual delinquency and non-performance.

•

Funding Risk. Farmer Mac has, over time, increased the proportion of its funding provided by commercial paper to $2.3 billion, or more than 72% of its total debt, and the average term of these discount notes has declined to approximately one week.3 The company’s funding is dependent on its continued ability to refinance these short-term obligations. Adding to this risk is the apparent misperception in the debt market that Farmer Mac’s debt has been rated Aaa by Moody’s, as is the debt of Fannie Mae and Freddie Mac. In fact, Farmer Mac’s debt is not rated by any rating agency.

Because of the nature of long-term lending, the true profitability of a particular loan is not known until the loan is repaid or not repaid. As a result, GAAP accounting for loan losses requires a company making or purchasing the loan to set aside reserves equal to the anticipated losses on a loan or portfolio of loans, projected at the time of the making or purchase of the loan. Lenders must add to these reserves if loss development is greater than anticipated. 2 By using the word “delinquency,” we adopt Farmer Mac’s terminology for a defaulted loan. These loans are not simply in technical default. They are loans that have not paid interest or principal and are more than 90 days past due. 3 Discount notes are the term used for agency commercial paper. Like commercial paper, they are issued at a discount and redeemed at par. Their terms do not exceed twelve months. In this document, we use the term discount note and commercial paper interchangeably.

4

•

Asset-Liability Mismatch. Farmer Mac’s assets and liabilities are substantially mismatched. As mentioned above, 72% of the company’s liabilities are discount notes which have average terms of approximately one week. By contrast, a substantial portion of the company’s assets are fixed-rate, long-term loans and mortgage-backed securities with maturities up to 30 years.

•

Increasing Risk in Program Assets. Farmer Mac has taken increasing risk in its program assets (farm mortgages). Over the past five years, Farmer Mac has innovated new “products” and modified its lending criteria to enable it to grow assets and guarantees. The assumption of these new risks has coincided with a decline in the U.S. agricultural economy that has become increasingly dependent on U.S. government price supports. We believe that one of Farmer Mac’s new products, the Long Term Standby Purchase Commitment (LTSPC), is, in substance, a put option or long-term credit-default swap written by Farmer Mac to other banks. The company’s $2.13 billion of LTSPCs at March 31, 2002 insure the credit quality and performance of designated farm loans and, as a result, suffer from an adverse-selection problem. The LTSPC program has created a potentially enormous unfunded additional liability for the company.

•

Increasing Risk in Non-Program Assets. Farmer Mac has taken increasing risk in its non-program assets. The company modified its 1996 policy of investing its non-farm loan assets only in highly liquid, short-term U.S. Government and agency debt, AAA corporate debt and A1-P1 commercial paper. The company’s changed policy allows it to invest in long-term equity securities and other lowerrated illiquid investments. Farmer Mac recently disclosed that it had purchased $167 million of private preferred stock issued by the banks from whom it buys loans and to whom it has issued LTSPCs.

•

High Leverage. At March 31, 2002, Farmer Mac had a 43 to 1 ratio of on- and off-balance-sheet liabilities to regulatory capital ($5.76 billion4 to $134 million). Over the past six years, the company’s liabilities and guarantee obligations have increased nine times while equity capital has increased less than three times.

•

Inadequate Board Oversight, Conflicts of Interest and Excessive Compensation. Representatives of Farmer Mac’s largest customers, who sell loans to Farmer Mac or purchase loan guarantees from the company, control the board with 10 of 15 seats, creating clear conflicts of interest. We believe that those directors who do not suffer from these conflicts have exercised insufficient independence and oversight over the actions of interested directors and management. Management enjoys an excessive equity compensation program under which they have been granted options and restricted stock totaling 13% of the shares outstanding and will ultimately receive as much as 35% when the balance of the authorized option pool is granted. These options fully vest within two years from each award. We find this option program egregious in its size and

Total Liabilities + LTSPCs + Off-balance-sheet AMBS = $3.28B + $2.13B + $0.361B = $5.76B


Page 2

in its accelerated vesting. Management’s financial self-interest exacerbates its incentive to take on risk and overstate profitability. •

Overvaluation. Even if one were to accept Farmer Mac’s earnings and book value as reported, we believe the company is materially overvalued at approximately 25 times earnings and 3.5 times book value. Similarly, we believe that the company’s debt securities trade at yields which do not adequately compensate investors for the underlying risks.

•

Risk to the U.S. Taxpayer. If the company exhausts its reserves and other resources, Farmer Mac has the ability to borrow up to $1.5 billion from the U.S. Treasury only for the purpose of meeting the company’s mortgage-backed securities guarantee and LTSPC liabilities. In the event of a continued deterioration in Farmer Mac’s loan and guarantee portfolio, the U.S. taxpayer could bear significant losses from this $1.5 billion facility. This financing, however, is not permitted to be used to repay Farmer Mac’s discount and medium-term note liabilities. B.

Background

Farmer Mac is a federally-chartered, public corporation that was created to establish a secondary market for agricultural real estate and rural housing mortgage loans ("Qualified Loans"). Like Fannie Mae and Freddie Mac, Farmer Mac is a GovernmentSponsored Enterprise or GSE. Farmer Mac was created by the Agricultural Credit Act of 1987, which amended the Farm Credit Act of 1971 (collectively, as amended, the "Act"). Farmer Mac participates in the agricultural mortgage market by (1) purchasing newly originated Qualified Loans directly from lenders through its "cash window" and existing, or "seasoned," Qualified Loans from lenders and other third parties in negotiated transactions; (2) exchanging agricultural mortgage-back securities (AMBS) guaranteed by Farmer Mac for newly originated and seasoned Qualified Loans that back those securities through its "swap" program; (3) issuing Long-Term Standby Purchase Commitments for newly originated and seasoned Qualified Loans; and (4) purchasing mortgage-backed bonds secured by Qualified Loans through its "AgVantage" program. Farmer Mac conducts its business through two programs - Farmer Mac I and Farmer Mac II. Under the Farmer Mac I Program, Farmer Mac purchases, or commits to purchase, Qualified Loans, or obligations backed by Qualified Loans, that are not guaranteed by any instrumentality or agency of the United States. Under the Farmer Mac II Program, Farmer Mac purchases the guaranteed portions (the "Guaranteed Portions") of loans guaranteed by the United States Department of Agriculture (the "USDA") pursuant to the Consolidated Farm and Rural Development Act (7 U.S.C. secs. 1921 et seq.; the "ConAct").


Page 3

Pursuant to its statutory authority, Farmer Mac guarantees timely payments of principal and interest on securities backed by Qualified Loans or Guaranteed Portions ("Farmer Mac Guaranteed Securities") and retains those securities in its portfolio or sells them in the capital markets. Farmer Mac is one of the fastest growing financial institutions in the world as measured by asset and guarantee growth. It is also one of the most leveraged. At March 31, 2002, Farmer Mac’s on- and off-balance-sheet liabilities and guarantees were 43 times its regulatory minimum capital.

$ 7.0

(Billions)

$ 6.0 $ 5.0 $ 4.0 $ 3.0 $ 2.0 $ 1.0

Notes Payable & OffBalance-Sheet Guarantees Regulatory Capital

Multiple

19 96 19 97 19 98 19 99 20 00 20 M 01 ar -0 2

$ .0

50 45 40 35 30 25 20 15 10 5 -

Multiple of Liabilities to Capital

Exhibit 1: Leverage Has Increased Substantially

Source: Farmer Mac SEC filings


Page 4

II.

FARMER MAC’S RESERVES APPEAR INADEQUATE A.

Adequacy of Loan and AMBS Reserves

At March 31, 2002, Farmer Mac had $17.0 million or 0.45% of reserves against $3.78 billion of loans and guarantees for post-1996 Farmer Mac I. The company does not take reserves against its pre-1996 Farmer Mac I loans because it believes the first-loss risk retained by the seller reduces the company’s risk to zero. Similarly, it takes no reserves for its Farmer Mac II loans which are guaranteed by the USDA. The company’s reserving policy has been remarkably consistent since 1996. This is despite the fact that Farmer Mac’s defaults (defined as a percentage of total loans) have risen from a low of 0.10% of loans and guarantees at June 30, 1997 to 2.32% at March 31, 2002. Exhibit 2: Flat Reserves and Growing Delinquencies. 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% 1996

1997

1998

1999

2000

2001 Mar-02

Reserves as % of Post-96 Act Guarantees % of Post-96 Act Loans that are Delinquent

Source: Farmer Mac SEC filings We believe the risk profile of the company’s loan and AMBS assets has changed meaningfully since 1996 for the following reasons: (1) In July 1996, Farmer Mac began taking first-loss risk. Prior to that time, the originating bank retained the first 10% of any losses on a loan; now Farmer Mac assumes 100% of the risk on the loans it purchases. (2) In 1998, Farmer Mac began a “part-time farmer” program in which it purchases loans in which the home is at least 30% of the value of the property and the property generates at least $5,000 per year in farm income. At an investor meeting on April 8, 2002, Mr. Edelman characterized these loans as more akin to jumbo vacation home loans than true farm loans. (3) In early 1999, the company entered into its first Long Term Standby


Page 5

Purchase Commitment (LTSPC). At March 31, 2002, the outstanding exposure under LTSPCs was $2.13 billion. We discuss LTSPCs below in “IV. Credit Quality.” B.

Reserves as a Percentage of Delinquent Loans

In Farmer Mac’s April 29, 2002 press release in response to The New York Times article,5 the company criticizes The Times for reporting Farmer Mac’s reserves of $15.9 million at year-end 2001 rather than the $17.0 million of reserves that the company announced in its Q1 2002 earnings press release. While Farmer Mac’s reserves were indeed $1.1 million higher at the end of the first quarter, the company did not disclose in the press release or in the 10-Q filed thereafter the dollar amount of loan delinquencies as of March 31, 2002. We calculate March 31, 2002 delinquencies at approximately $88 million, up more than $28 million or nearly 50% from year-end 2001, just three months earlier.6 One common measurement of reserve adequacy used by investors is the comparison between non-performing loans and current levels of reserves. By comparing Farmer Mac’s $17.0 million of reserves with its current level of delinquencies, you can see that Farmer Mac must collect 81%7 of these delinquent loans in order for this reserve level to be adequate for just the $88 million of current delinquencies. Even if the $17.0 million is adequate for the current $88 million, this leaves no reserves available for the company’s other $4.3 billion of total loans and guarantees. Clearly, on this basis, reserves appear wholly inadequate. C.

Farmer Mac’s Reserves Compared with Farm Credit Banks

By comparing Farmer Mac’s reserve levels to banks with similar, if not lower-risk, loan portfolios, one can also see the inadequacy of the company’s reserves. Farm Credit Banks (FCBs) are providers of credit to local agricultural banks. Below, we provide the ratio of reserves to impaired loans at December 31, 2001 at five of the seven FCBs.8 These reserve levels are 9.5 to 17.8 times Farmer Mac’s reserve level of 19.5% of delinquencies.

5

“Big City Paydays at ‘Farmer Mac’,” Alison Leigh Cowan, The New York Times, April 28, 2002, Business, p. 1. 6 We calculated the updated delinquent loan balance by taking the disclosed percentage of delinquencies from the press release of 2.32% and multiplying it by the sum of the columns labeled “Farmer Mac I Loans & AMBS” and “LTSPC’s,” which total $3.781 billion. This same analysis generates a delinquent loan dollar amount that matches the company’s disclosed levels in the December 31, 2001 10-K. 7 ($88 m -$17m)/$88 m = 81% 8 We were not able to obtain the data on Western Farm Credit Bank or Farm Credit Bank of Wichita, the other two FCBs. Gotham Partners Management Co., LLC

Page 6

Farm Credit Bank AgAmerica Agfirst Farm Credit Bank and Associations Agribank, FCB and Associations CoBank Farm Credit Bank of Texas

Ratio of Reserves to Impaired Loans 201% 338% 210% 246% 181%

The irony of the comparison is that these FCBs are the primary customers of Farmer Mac’s LTSPC guarantees. The FCBs are much more conservatively capitalized than Farmer Mac, with an average of 15.8% capital to total assets,9 and more conservatively reserved. Yet, the FCBs look to Farmer Mac to insure them against risky loans in their portfolios. If Farmer Mac were required to increase its reserves to the level of the FCBs listed above, i.e., from 19.5% to a minimum of 181% of delinquencies, its current $17.0 million of reserves would have to be increased by a minimum of $157.8 million. This would reduce Farmer’s equity capital at March 31, 2002 by $140.8 million to negative $6.4 million rendering the company insolvent. Even after adding in Farmer Mac’s recent sale of $35 million of preferred stock, its capital would be only $28.6 million, still well under its minimum, critical, and risk-based regulatory capital requirements. D.

Farmer Mac’s Reserves Compared with Money Center Banks

For another point of comparison, we compare Farmer Mac’s reserves to reserves at large money center banks. These banks arguably have higher quality, more diversified portfolios than Farmer Mac, but have reserves which average 161% of non-performing loans compared with Farmer Mac’s 19.5%.10 Again, on this basis, Farmer Mac’s reserve levels appear materially inadequate. E.

Under-Reserving Creates Illusory Earnings

A lender, like an insurer, that knowingly or unknowingly provides inadequate reserves for the potential losses in its future will, in hindsight, be seen to have manufactured illusory earnings. In the case of Farmer Mac, we believe that the company’s growing risks – the increasing age and default rates on its guarantee obligations, the introduction of higher risk guarantee products, and the post-1996-Act increase in Farmer Mac’s exposure to loss – all suggest that greater reserves are needed.

9

“Performance and Accountability Report – Fiscal Year 2001,” Farm Credit Administration, p. 15. “Is Citigroup’s Umbrella Big Enough? Lower Than Average Loan-Loss Reserves Make Some Investors Uneasy,” Wall Street Journal, May 13, 2002, p. C1.

10


Page 7

In the chart below, we show what increasing the reserves to various levels would do to Farmer Mac’s retained earnings and core capital. We believe that the chart explains why management has not added adequate reserves – even a modest upward adjustment would put the company out of regulatory capital compliance and significantly reduce book value and earnings: Exhibit 3: Impact of Increased Reserves on Capital and Retained Earnings $50 $25 $ (Millions)

($25) ($50) ($75) ($100)

Capital Surplus Eliminated

Retained Earnings Eliminated

March 2002 Capital Reduced to Zero

($125) Capital and New Preferred Eliminated

($150) ($175)

19% 25% 44% 50% 66% 85% 100% 125% 150% 175% 200% 225% Reserves as % of Delinquencies at March 31, 2002 Retained Earnings

Capital Surplus

Source: Farmer Mac SEC filings F.

Weighted-Average Loan to Value and Collateral Values

Farmer Mac argues that its modest level of reserves is adequate because collateral values will protect the company in a foreclosure. This view, it appears, is predicated on the claimed quality of Farmer Mac’s original underwriting of a loan, or on the underwriting at the time a loan is guaranteed under the LTSPC program. Farmer Mac contends that it has an overall low ratio of “weighted-average” loan to value (LTVs) that provides it with adequate collateral protection. According to the 2001 10-K, delinquent loans have a 59% weighted-average loan to value. There are a number of problems with this statistic, however. First, we believe that weighted-average loan to value is a flawed measure for a portfolio of loans that is not cross-collateralized. Farmer Mac cannot use equity in underleveraged properties to make up for the lack of equity in overleveraged properties. This can best be seen by using a somewhat extreme example. Consider two loan portfolios: in one portfolio every loan has a 50% LTV; in the second portfolio, there is one $10,000,000 loan at 10% LTV and one-hundred $100,000 loans at


Page 8

90% LTV. The two portfolios have identical weighted-average LTVs. However, the second portfolio will likely have a much higher delinquency rate with a much higher risk of loss to the lender. The first portfolio is far less likely than the second to experience defaults, and if they occur, the net losses should be far lower. We believe that Farmer Mac’s loan portfolio is more consistent with the second than the first portfolio, and, therefore, its weighted-average statistics understate the true risk of the portfolio. Second, appraisals of farms are notoriously inaccurate. Farms are businesses with unstable and unpredictable cash flows, which makes valuations difficult to perform. We believe that most lenders do not rely heavily on appraisals for this reason. Third, the company’s LTV statistics are based on outdated appraisals for determining value (the denominator of the calculation), while using the current amortized principal amount in the numerator of the calculation. Farmer Mac explains: As of December 31, 2001, the weighted-average loan-to-value ratio for all post-1996 Act loans was 49 percent, and the weighted-average loan-to-value ratio for all post-1996 Act loans that were 90 days or more past due, in foreclosure or in bankruptcy was 59 percent. The following table illustrates the distribution of non-performing loans as of December 31, 2001 by loan-to-value ratio (based on current loan balance and appraised value at the date of initial guarantee by Farmer Mac): [December 31, 2001 10-K; emphasis added.]

The fact that at the time of a purchase or guarantee, a particular loan had a certain LTV is not the critical information a Farmer Mac shareholder needs to know today. In order to determine the likelihood of losses in foreclosure, one needs the collateral values at the time of default, and this information is strikingly absent from the company’s disclosure to investors. This is particularly unfortunate because of the emphasis management and analysts who cover Farmer Mac place on the weighted-average LTV statistic in determining the likelihood of losses. For example, in a press release on April 29, 2002, in response to The New York Times article, the company states: (p) the article ignores the loan-to-value ratios on all of the company's seriously delinquent loans, a weighted average of 59% as set forth in its Form 10-K filed on March 27, 2002, which makes the recovery of 73% of the $59.8 million of delinquent loans referred to in the article a very reasonable expectation to beat;

Even ignoring the problems with “weighted-average” LTV statistics, we find it significant that management did not mention in the press release that the LTV statistics are as of the date of the original loan purchase or guarantee, while the numerator (the current loan balance) is as of a recent date. We believe it is likely that borrowers default because of poor business performance. In other words, the defaulting loan’s underlying collateral is generating insufficient cash flows to cover debt service, meaning there has likely been a reduction in the collateral value compared with earlier LTV estimates. As a farm’s cash flow declines, its value necessarily declines.


Page 9

G.

Charge-Offs

One of the other ways Farmer Mac defends the adequacy of reserves is by referring to the relatively low level of historical charge-offs. Management, however, can control the level of charge-offs by retaining rather than selling foreclosed assets and postponing write-downs. We believe that this may be occurring at Farmer Mac. One might ask, however, why there is no REO listed on the balance sheet. While there is no such category on the company’s balance sheet, we suspect the “Loans” category may contain REO assets. Loans on the balance sheet have increased from $30.2 million at December 31, 2000 to $309.9 million at March 31, 2002. While some of this growth in loans can be explained by the SFAS 140 pronouncement that requires certain post-April 1, 2001 newly issued AMBS to be classified as loans, we believe that some of the growth in the loan category is due to growth in foreclosures and REO. Charge-offs have grown over time from zero as recently as 1998 to a net $2.1 million in 2001 and a projected $3.4 million in 2002. We would expect the rate of increase in charge-offs to be accompanied by a corresponding increase in reserves. Yet, no such corresponding increase has occurred. Moreover, in the December 31, 2001 10-K in which the company projected higher charge-offs for 2002, it listed a number of reasons for increases in charge-offs – all of which are suggestive of the need to increase reserves: In 2002, it is likely that delinquencies will increase and it is expected that quarterly charge-offs will be at somewhat higher levels, similar to the $850,000 in losses recognized during fourth quarter 2001, due to a growing percentage of the guarantee portfolio entering its peak loss years, pressures from growing agricultural inventories, weak markets for agricultural commodities and products worldwide along with low prices and economic uncertainty in the agricultural sector, following the trend of the last several years.

Companies are required by GAAP to predict their loan losses over time by setting aside reserves. We find it unusual, however, that Farmer Mac can predict the absolute dollar amount of charge-offs for the next few quarters with precision. Similarly, it seems strange that charge-offs are expected to be the same approximate dollar amount for the next few quarters. Discretion over the amount of charge offs is entirely in management’s hands. By selectively selling assets to generate charge-offs that don’t exceed $850,000 per quarter, the company can limit charge offs to approximately $850,000 per quarter. Alternatively, management can choose never to write down any of the on-balance-sheet loans or REO.


Page 10

III.

DELINQUENCIES

Farmer Mac has been experiencing increasing delinquencies. We believe, however, the company’s disclosure for and reporting of delinquencies understate the size of Farmer Mac’s actual delinquencies and non-performing assets. A.

Shrinking Reported Losses Through Rapid Growth of the Denominator

Farmer Mac gives little data to help an investor quantify precisely the adequacy of its reserves, particularly because its disclosure focuses on delinquencies as a percentage of total loans including just-originated loans. Because farm loans take at least three to five years to reach their peak default years (according to management and other industry observers), we believe Farmer Mac’s rapidly growing loan and guarantee portfolio has masked loan losses when they are expressed as a percentage of total loans and guarantees. The impact of the growing number of loans and guarantees on the calculation of overall delinquencies is exaggerated further because there can be a meaningful lag between the time when a farm loan in Farmer Mac’s portfolio gets into trouble and the time when it shows up as delinquent in the company’s SEC filings. This has several causes: (1) Farmer Mac loans typically pay interest only once or twice per annum; (2) a loan is not defined as delinquent until it is at least 90 days past due; and (3) there appear to be no interim loan covenants. Thus, most Farmer Mac loans cannot be “delinquent” by Farmer Mac’s definition before a minimum of nine months (the semi-annual payment plus 90 days) or 15 months (the annual payment plus 90 days) from the time the loan is originated or from the last timely payment. Farmer Mac gives clues about its potential default rate in the supplemental footnotes of its filings through a year-of-issuance disclosure that began to appear in public filings in March 31, 1999. By showing the delinquency rates of loans by year of issuance, one can get a better sense of default rates. This table is from the March 31, 2002 10-Q:

By year of origination: Before 1996 1996 1997 1998 1999 2000 2001 2002 Total


Distribution of Post-1996 Act Delinquency Loans Rate(1) ---------------- -------------24% 8% 10% 16% 18% 10% 12% 2% --------------100%

0.82% 7.17% 5.62% 3.24% 1.98% 1.10% 0.16% 0.00% -------------2.32%

Page 11

There is a clear trend in the data: with the passage of time, default rates grow. Note that the loans originated in 1996 have a 7.17% default rate. The 1996 delinquency rate peaked at 8.35% as of March 31, 2001, as the loans approached their fifth year outstanding. As we will show below, however, these numbers understate the true delinquencies overall as well as for each of the cohorts. Even if we accept 8.35% as being the highest level delinquencies reach, we believe that a 45 basis point reserve is inadequate. In order for 45 basis points to be adequate, the company must collect 94.6%11 of its delinquent loans, an unrealistic assumption, particularly in light of the high costs of collection and transaction fees in a foreclosure and sale of a farm. In addition, we think, the risk profile of the company’s loan purchases and LTSPCs has increased since 1996, making the 8.35% rate an underestimate for the future. B.

Reducing Reported Delinquencies Through Shrinking the Numerator

Farmer Mac appears to have reduced reported delinquencies by shrinking the numerator in its calculation of delinquent loan percentage. The company does so through two methods. First, it excludes delinquent loans that are “in the process of collection”: Non-performing (or "impaired") loans are loans for which it is probable that Farmer Mac will not receive all amounts contractually due and include all loans 90 days or more past due unless the loan is in the process of collection. [December 31, 2000 10-K; emphasis added.]

It is not clear to us which delinquent loans are not “in the process of collection.” What lender does not begin and continue the process of collection of a loan until the loan is either repaid or foreclosed upon? We are quite puzzled by this carve out, but believe that it could reduce the numerator of delinquencies quite substantially. The other numerator-reducing device is the removal from the calculation of Real Estate Owned or REO. This appears to have occurred in the second quarter of 1997. The March 31, 1997 10-Q stated: At March 31, 1997, loans that were 90 days or more past due, loans that were in foreclosure or bankruptcy and loans that had been foreclosed upon and the related mortgaged property not yet liquidated ("REO Property") represented 0.4% of the principal amount of all loans underlying Farmer Mac Guaranteed Securities. Management believes that no losses will be incurred by Farmer Mac as a result of the loans in foreclosure or the REO Property because of the existence of the 10% subordinated interests with respect to the related securities. [Emphasis added.]

Note that in the same section of the subsequent 10-Q, Farmer Mac has excluded REO from the calculation:

11

(8.35%-0.45%)/8.35% = 94.6%


Page 12

At June 30, 1997, loans that were 90 days or more past due and loans that were in foreclosure or bankruptcy represented 0.1% of the principal amount of all loans underlying Farmer Mac Guaranteed Securities. [Emphasis added]

While it is impossible to determine the precise impact of the numerator changes on the reported delinquencies, we suspect the impact could be significant. This can be shown by comparing the June 30, 2000 distribution of delinquencies by year of origination with the following two quarters’ data. The tables have been cut and pasted directly from the relevant SEC filings so you can see exactly how they are presented:

[Balance of page intentionally blank to present the following chart.]


Page 13

From the June 30, 2000 10-Q Distribution of Post-1996 Loans --------------------

By year of origination: Before 1996 1996 1997 1998 1999 2000

Delinquency Rate ----------------

38% 9% 11% 19% 19% 4% ---------------100%

Total

0.41% 6.28% 2.22% 0.99% 0.00% 0.00% 1.16%

From September 30, 2000 10-Q

By year of origination: Before 1996 1996 1997 1998 1999 2000 Total

Distribution of Post-1996 Loans --------------------

Delinquency Rate ----------------

23% 8% 28% 9% 11% 20% ---------------100%

0.12% 0.00% 0.45% 6.43% 3.23% 3.51% 1.80%

From the 2000 10-K Post-1996 Act

By year of origination: Prior to 1996 1996 1997 1998 1999 2000

Delinquency Rates ----------------0.89% 4.40% 2.20% 1.70% 0.22% 0.22%

By comparing the three tables, you will note large changes in quarterly delinquencies and in distributions by year of origination. How 1996 delinquencies can go from 6.28% to 0.00% and then back to 4.40% is difficult to ponder. Perhaps the movement of a loan from “in the process of collection” to not “in the process of collection” or from “loan” to “REO” accounts for some of the movements. Alternatively, perhaps there is an error in


Page 14

the table; however, the large quarterly changes in distribution and in delinquency rates by cohort are repeated in future quarters. The year of origination changes are also dramatic. Note that the distribution of 1997 loans go from 11% on June 30, 2000 to 28% on September 30, 2000. This is the first clue that the cohorts of loans by year of origin are changing in unanticipated ways. Intuitively, we initially thought loans by year of origin should decline as loans are repaid and as loans are purchased in the current year. We now believe something else is at work. The second clue is the “Before 1996” year-of-origination category for “Post-1996 Act Loans.” This category doesn’t seem to make sense because Farmer Mac obviously did not begin purchasing newly originated Post-1996 Act loans until July 1996. We believe that the explanation for the “Before 1996” category and the changing distributions is that the company appears to be including newly issued LTSPCs in the data. Most LSTPC transactions involve seasoned loans that have performed for more than five years. As a result, when Farmer Mac enters into new LTSPCs on seasoned loans, it adds these new performing loans by year of origination to the table. For example, assume for a moment that at the beginning of the March 31, 2002 quarter, the 1996 origination year had $100 million of loans, 10% of which, or $10 million, were delinquent. Next, assume that later in the quarter, the company issues an LTSPC on $100 million of 1996 originated loans. By definition, these loans under the new LTSPC are performing on the date of the guarantee because LTSPCs are issued only on performing loans. As a result, by the end of the quarter, the company’s 1996 delinquency rate, assuming no change in delinquencies since the beginning of the quarter, would drop to 5% ($10 million out of $200 million). The introduction of performing loans into an earlier origination year necessarily obfuscates the delinquency data by increasing the principal amount of performing loans in the denominator of the delinquency calculation. As a result, the company can effectively go back in time and “average down” the delinquency data for any distribution year by adding performing loans under LTSPC. We find the delinquency data by year of distribution materially misleading if this is indeed what is going on. The whole idea behind loss development is to show how a fixed set of loans behaves over time. Adding performing loans to a prior-period portfolio distorts loss-development data. If the company provided two tables -- one for the performance of loans by year of acquisition and another for the performance of LTSPCs by year of initial guarantee -- shareholders would have a much better sense of the true delinquency statistics.


Page 15

IV.

CREDIT QUALITY A.

Risks of Farm Loans

Farm loans are among the riskiest real estate and business loans because of the highly volatile nature of farm income. Below, we quote a risk factor disclosed in an information statement filed by Farmer Mac on November 21, 1997: Repayment of agricultural loans is typically dependent upon the success of the related farming operation, which is, in turn, dependent upon many variables and factors over which farmers may have little or no control, such as weather conditions, economic conditions (both domestic and international) and even political conditions. If the cash flow from a farming operation is diminished (for example, adverse weather conditions destroy a crop or prevent the planting or harvesting of a crop), the farmer's ability to repay the loan may be impaired. Significant loan payment defaults by farmers are likely to necessitate payments under Farmer Mac's guarantees and could have a material adverse effect on Farmer Mac's financial condition and results of operations.

Other risk factors include changing food preferences, pestilence, and mad cow disease. In addition, a particular farm’s income is dependent on the business experience of the borrower and members of his or her family and other employees. Generational changes as well as other more traditional business factors can affect the cash flow of a farm. We believe that as a result, most lenders do not think of farm loans as pure real estate loans, like home mortgages or apartment house lending, but rather as business loans. Because it is extremely difficult for a long-distance lender to assess borrower and property quality, farm lending is a specialized and local lending business. Unlike singlefamily mortgage lending, for which a high degree of standardization, many decades of credit performance information and other lending statistics exist, farm lending is not readily subject to standardization because of the unique nature of the land, equipment, crop-type, and borrower. As a result, most farm loans are made by local banks or local branches of the Farm Credit System, an extremely well-capitalized group of lending institutions that also benefit from GSE status. The history of our country is replete with periods of farm failure and bankruptcy as recently as the early 1980s. In fact, the early 1980’s farm credit crisis was the catalyst behind the formation of Farmer Mac. Congress hoped that Farmer Mac could create liquidity in the farm credit markets through the establishment of a secondary market in agricultural loans, similar to the secondary market for residential loans created by Fannie Mae and Freddie Mac. Unfortunately, this has not occurred. B.

Annual and Semi-Annual Payments Under Loans Without Covenants

Unlike nearly all other business or commercial real estate loans with which we are familiar, Farmer Mac loans apparently do not have typical post-closing commercial loan covenants, such as periodic Debt-Service Coverage Ratio (DSCR) maintenance tests, borrower net worth tests, annual loan-to-value (LTV) standards, appraisal requirements,


Page 16

and the like. About the only way for a Farmer Mac loan to go into default is for the borrower to be more than 90 days late in making a payment. Payments on most of the company's loans are due only once or twice a year. This may partly explain why Farmer Mac loans typically, according to the company, default more in the first and third quarters. Beginning in March 31, 1998 when delinquencies began to rise, Farmer Mac stated in that 10-Q: The increase in delinquencies is due to the aging of the portfolio and the fact that a greater proportion of the loans backing Farmer Mac I AMBS have payments due on January 1 than any other date during the year as a result of the number of semi-annual and annual pay loans in the portfolio. [March 31, 1998 10-Q; emphasis added.]

One quarter later in the June 1998 10-Q, the company blamed growing defaults on additional factors: The increase in delinquencies is due to the growing number of loans held or securitized by Farmer Mac that are approaching their anticipated peak default years. Farmer Mac anticipates that the level of delinquencies will fluctuate from quarter to quarter with higher delinquency rates reported in the first and third quarters of each year due to the semiannual payment characteristics of most Farmer Mac loans. On average, Farmer Mac anticipates that delinquency levels during 1998 will be higher than those experienced during 1997 due to the aging of loans held or securitized by Farmer Mac and adverse conditions affecting certain sectors of the agricultural economy. [June 30, 1998 10-Q; emphasis added.]

In effect, the company is saying that borrowers can only default when interest payments are due. While this is standard practice for monthly-pay single-family loans, it is almost never the case for commercial loans that by design have numerous covenants which give a lender rights in the event its collateral or its borrower's finances begin to deteriorate. Because a borrower's finances or collateral can deteriorate meaningfully in the six to twelve months from the time of a borrower’s last interest payment, Farmer Mac’s loans, have inherently higher risk than conventional monthly-pay business and real estate loans. C.

The LTSPC Loan-Guarantee Program

Farmer Mac has been criticized for its growth in new “products” out of concern as to whether these products are within the mission chartered for the company by Congress. Representative Richard H. Baker, Chairman of the Congressional Subcommittee on Capital Markets, Securities and Government Sponsored Enterprises stated: It seems increasingly evident that the new products and services aimed at increasing Farmer Mac’s market share must be reevaluated. A determination of the net benefits generated by Farmer Mac and the extent to which its mission activities compete with or complement other GSEs must be reconsidered. [Press Release by the Congressional Subcommittee on Capital Markets, Securities and Government Sponsored Enterprises, June 9, 1999.]


Page 17

The product that we believe deserves the most scrutiny is the Long-Term Standby Purchase Commitment or LTSPC which was first issued by Farmer Mac a few months before Representative Baker’s statement. Since its inception three years ago, the LTSPC program has grown from zero to $2.13 billion at March 31, 2002. In the past five quarters, from January 1, 2001 to March 31, 2002, LTSPCs increased by $1.26 billion. LTSPCs now constitute more than half of Farmer Mac’s exposure to loan losses. Farmer Mac’s LTSPC was described at its introduction in 1998: In 1998, Farmer Mac introduced a variation on swap transactions for Sellers seeking to obtain the benefits of a Farmer Mac guarantee on Qualified Loans retained in their portfolios. Under a "long-term standby purchase commitment," Farmer Mac commits to purchase any Qualified Loan in a segregated pool of loans subject to the commitment, if: (a) the Qualified Loan becomes four months delinquent; (b) the Qualified Loan meets Farmer Mac's cash window requirements at the time the Seller requests that Farmer Mac purchase the loan; or (c) the Seller requests that Farmer Mac purchase all of the identified Qualified Loans. In the case of a delinquent Qualified Loan, Farmer Mac will pay the Seller a predetermined price for the loan - generally, principal plus accrued interest; in the case of a Qualified Loan under clause (b) or (c), the price for the Qualified Loan(s) would be negotiated at the time of purchase. This structure permits the Seller to retain the segregated loans in its portfolio while reducing its credit and concentration exposures and, consequently, its regulatory capital requirements. In consideration for Farmer Mac's assumption of the credit risk on the segregated loans, the recipient of the commitment pays fees to Farmer Mac on the outstanding balance of the loans approximating what would have been Farmer Mac's guarantee fee had the loan been exchanged with Farmer Mac in a swap transaction. [December 31, 1998 10-K; emphasis added.]

In essence, the LTSPC is a guarantee or reinsurance policy. For an annual fee of 40 to 50 basis points of potential liability, Farmer Mac assumes the risk of a borrower’s default. In addition, the lender can force Farmer Mac to purchase LTSPC guaranteed loans at any time at its option. In the event Farmer Mac’s creditworthiness deteriorated, lenders with LTSPCs would probably exercise the put option and force the company to acquire the underlying loans even if the loans were not delinquent. With more than $2 billion of LTSPCs outstanding, we believe that Farmer Mac does not have sufficient liquidity to fulfill these and its other guarantee obligations in the event of a run on the bank. The company acknowledges that its contingent liabilities exceed its resources: Furthermore, Farmer Mac anticipates that its future contingent liabilities in respect of guarantees of outstanding securities backed by agricultural mortgage loans will greatly exceed its resources, including its limited ability to borrow from the United States Treasury. [ “$29,241,646 Guaranteed AMBS,” Farmer Mac Mortgage Securities Corp. April 30, 2002.]

Compared to Farmer Mac’s other guarantee products, we believe LTSPCs carry the greatest risk with the least adequate compensation.


Page 18

(Billions)

Exhibit 4: LTSPC’s Have Increased Rapidly and Account for 56% of Total Guarantees $ 2.5

60%

$ 2.0

50%

Loans & AMBS

40%

$ 1.5

30% $ 1.0

20%

$ .5

10%

$ .0

LTSPC LTSPC as % of Post96 Act Guarantees

0% Dec- Dec- Dec- Dec- Dec- Mar97 98 99 00 01 02

Source: Farmer Mac SEC filings By entering into a LTSPC, a lender shifts 100% of the credit risk of the selected loans to Farmer Mac and frees up regulatory capital (the regulatory capital requirement for guaranteed loans is 80% lower than for standard loans). Farmer Mac is similarly advantaged by entering into LTSPCs; its regulatory capital requirement is only 0.75% of exposure versus 2.75% of principal amount for on-balance-sheet assets. Because of the informational advantage a lender has about its own seasoned loans versus the knowledge that Farmer Mac can obtain in its limited due diligence review, there is little to prevent a lender from offloading its riskiest loans into the LTSPC program. D.

LTSPCs Compared with Credit-Default Swaps

We find it useful to compare the LTSPC to an analogous Wall Street product called a credit-default swap. Alternatively, one can think of the LTSPC product as a long-term reinsurance policy and the annual fee as the premium.12 When viewed in this context, we conclude that the LTSPCs are materially underpriced given the level of risk assumed by Farmer Mac. A key difference between Farmer Mac’s LTSPCs and credit-default swaps is in the guarantor’s or insurer’s length of exposure to the borrower’s possible default. LTSPCs are usually very long term because they extend for the term of the underlying loans, which can be 10 to 15 years and as much as 30 years. Credit-default swaps, by contrast, are almost always for five years or less; 10-year protection is available only for the highest-grade credits and only if the buyer of insurance pays significantly higher annual premiums. Another important difference between LTSPCs and credit-default swaps is that the farm lender can cull certain loans from its portfolio rather than being required to 12

The UBS analyst who covers Farmer Mac appears to agree with this characterization for he refers to the outstanding balance of LTSPC guarantees as “insurance in force” in his research reports. Gotham Partners Management Co., LLC

Page 19

insure the overall credit quality of an entire portfolio. These two differences, we believe, make the LTSPC significantly riskier than credit-default swaps. We compared the pricing of Farmer Mac’s LTSPCs (40-50 basis points per year times the amount insured) to the pricing of Wall Street’s credit-default swaps. Below we have listed recent credit-default swap pricing provided by Deutsche Bank for some blue chip corporations.13 All pricing is for five-year credit default swaps. Company J.P. Morgan GE Capital American Express Merrill Lynch Disney Daimler Chrysler Ford Motor Credit Hilton Hotels Toys R Us Goodyear Delta Air Lines ATT Gap

Annual Fee in Basis Points 70 72 75 95 95 135 200 210 330 340 500 575 610

In order to compare the relative risks and rewards of the LTSPC program with creditdefault swaps, we compare the credit quality of the above issuers to farm loans selected for insurance by the original underwriter of the loans. Next, we compare pricing: 70-610 basis points per year for J.P. Morgan et al. versus 40-50 basis points per year for LTSPCs. Finally, we compare the length of exposure: for the above credit-default swaps the exposure is limited to five years, while the LTSPC exposure can be 10 to 30 years. To us, this demonstrates that the 40-50 basis point cost of credit-default insurance on farm loans is extraordinarily cheap. It is therefore not surprising that Farmer Mac’s customers are entering into large dollar amounts of LTSPCs. Despite the increasing risk and, we believe, extreme mispricing of Farmer Mac’s current high-growth product, the company has held reserves as a percentage of outstanding loans and guarantees constant for the last six years at approximately 45 basis points of post1996 Farmer Mac I loans. We find troubling the combination of high risk, insufficient reward, and minimal reserves. E.

How Did Farmer Mac Underwrite its 10,000 Loans as well as New Acquisitions and Guarantees?

With only 21 employees at year-end 1996, growing to 30 employees at year-end 2001, Farmer Mac, we believe, has never had the capacity to underwrite what we estimate to be

13

All pricing as of May 3, 2002.


Page 20

its approximately 10,000 loans.14 Instead, according to management, it oversees underwriting done by others. According to Mr. Edelman, the majority of the company’s underwriting is done by Zions Bancorp and AgFirst, which are major shareholders of the company, appoint directors to the company's board (Zions’ director sits on the compensation committee, AgFirst's sits on the audit committee), and are major sellers of loans to the company. Last year, for example, Zions Bancorp sold the company 34% of the loans purchased by Farmer Mac and, according to Mr. Edelman, has sold Farmer Mac more than $1 billion of the total loans in its guarantee portfolio. In 2001, AgFirst purchased $213 million of LTSPCs.15 We believe that underwriting is a core competency of any lender or purchaser of loans and that contracting it out is antithetical to good business practice. Contracting underwriting to primary sellers of loans seems risky and fraught with potential for conflicts. F.

Underwriting Standards

Farmer Mac’s underwriting criteria have not changed, as far as we can tell, since the inception of the post-1996 Act program.i To summarize: for a newly issued Qualified Loan, the criteria are as follows: (1) No greater than 70% loan to value (other than a part-time farm loan and a loan on an agricultural facility with a related integrator contract) (2) a pro forma, post-closing debt-to-total-asset ratio of 50 percent or less; (3) a pro forma DSCR on the mortgaged property of not less than 1:1; (4) a pro forma total debt service coverage ratio, including farm and non-farm income, of not less than 1.25:1; and (5) a pro forma ratio of current assets to current liabilities of not less than 1:1. Of the above five criteria, the one that we believe lenders primarily rely on in determining the borrower’s ability to pay interest and principal on the loan is (3), the property Debt Service Coverage Ratio (DSCR), particularly because there are apparently no debt covenants that would prohibit the borrower from obtaining additional leverage. Even in the most bullish real estate lending environments of which we are aware, first mortgage lenders’ DSCRs have not declined below 1.1 to 1, or 110%, coverage of the 14

This assumes an average loan size of approximately $450,000 as supported by a Robertson Stephens analyst report. 15 Also, see annual proxy for detailed disclosure of affiliate relationships. Gotham Partners Management Co., LLC

Page 21

required debt service payment except in extraordinary circumstances. Even for the least risky real estate asset classes like apartments, DSCR of 1.2 to 1, or 120%, are typically the minimum requirements. In addition, the standards for business loans, rather than real estate loans, are significantly higher, even if they are secured by real estate assets, because of the greater volatility of business income when compared to multi-family properties or multi-tenant office buildings. In light of the high volatility of farm income because of the farm-lending risk factors mentioned previously, a 1.0 to 1 DSCR, with zero cushion against a decline in income, is inadequate. Even if one were to use the 1.25 to 1 borrower cash flow requirement, there is little, if any, cash flow cushion in a Farmer Mac loan that just meets this criterion. Bearing in mind that there are apparently no covenants to prevent the borrower from taking on additional indebtedness on other assets of the borrower or on an unsecured basis, relying on the pro forma borrower coverage is fraught with risk. Farmer Mac’s underwriting standards are not rigid. The company has the right to waive them: The Underwriting Standards provide that Farmer Mac may, on a loan-by-loan basis, accept loans that do not conform to one or more of the Underwriting Standards when: (1) those loans exceed one or more of the Underwriting Standards to a degree that compensates for noncompliance with one or more other Standards ("compensating strengths"); and (2) those loans are made to producers of particular agricultural commodities in a segment of agriculture in which such compensating strengths are typical of the financial condition of sound borrowers. Farmer Mac's acceptance of loans that do not conform to one or more of the Underwriting Standards is not intended to provide a basis for waiving or lessening in any way the requirement that loans be of consistently high quality in order to be eligible Qualified Loans. [December 31, 2001 10-K]

We cannot tell from the company's public filings when, how often, and to what extent it has used or continues to use this discretion to waive underwriting standards nor, where it does so, how it ensures "consistently high quality" loans. In addition, in most cases when a loan is “seasoned,” has performed for five years without a payment more than 30 days late, and meets the company’s LTV test, it will qualify under Farmer Mac’s LTSPC standards. Unfortunately, however, past borrower performance is not necessarily indicative of future performance nor does seasoning protect Farmer Mac from issuing LSTPCs on loans with deteriorating credit or market trends. Further, when Farmer Mac is buying a seasoned loan, we believe the lender selling the loan or obtaining a guarantee from Farmer Mac has a much better understanding than does Farmer Mac of which seasoned loans are more likely to default in the future.


Page 22

V.

ADEQUACY OF CAPITAL A.

Regulatory Capital

Farmer Mac’s regulatory capital requirements were established by the 1996 Act as amended. There are three requirements that must be met which include minimum, critical, and risk-based capital. Since the risk-based capital standards don’t take effect until May 23, 2002 (the date of this report),16 the minimum capital and critical requirements (50% of minimum) are the standards the company has had to meet. In light of the fact that off-balance-sheet guarantees have, at a minimum,17 the same credit risks that on-balance-sheet assets have, it strikes us as odd that the capital requirement for off-balance-sheet guarantees (0.75% of exposure) is only one-fourth of the capital requirement for on-balance-sheet assets (2.75% of principal). These vastly disparate capital requirements have encouraged the company to grow its off-balancesheet risks – its LTSPC program, for example, grew from zero four years ago to $2.13 billion as of March 31, 2002. While there is no direct interest-rate risk for an offbalance-sheet guarantee, credit risks under farm loans vastly outweigh interest-rate risk. Unlike home loans, many of Farmer Mac’s farm loans have prepayment penalties to help mitigate interest rate risk. B.

Farmer Mac’s Capital and Delinquencies Compared with Fannie Mae’s

The regulatory capital requirements of Farmer Mac are surprisingly similar to those of Fannie Mae. Fannie Mae’s are 2.50% for on-balance-sheet assets and 0.45% for offbalance-sheet guarantees versus 2.75% and 0.75% respectively for Farmer Mac. We believe, however, that the regulatory authorities who composed the capital requirements for Farmer Mac may not have adequately considered the credit differences between Fannie Mae's one-to-four family conforming home loans versus the low debt-servicecoverage farm loans in Farmer Mac’s portfolio. Even if one were to use Farmer Mac’s delinquencies as a percentage of total loans and guarantees outstanding compared with Fannie’s delinquency rates calculated on its “average book of business,” the comparison is striking. Farmer Mac’s delinquencies were 2.32% versus 0.06% or six basis points for Fannie Mae.18 Farmer Mac’s delinquencies as expressed in their most favorable light are nearly 39 times Fannie Mae’s as a percentage of the portfolio. We do not understand how only 0.25% of extra capital for on-balance-sheet assets and 0.30% for off-balance-sheet guarantees can be adequate in light of the substantially higher delinquencies and credit risks at Farmer Mac compared to Fannie Mae. 16

Originally they were supposed to take effect three years after the Act, but efforts by Farmer Mac to delay the implementation as well as the slow Federal bureaucracy have been successful in creating delay. 17 Not considering the greater risks of LTSPCs. 18 Source: Fannie Mae 2001 Annual Report Gotham Partners Management Co., LLC

Page 23

C.

Non-Performing Loans as a Percentage of Farmer Mac’s Regulatory Capital

The chart below shows Farmer Mac’s delinquent loans as a percentage of its regulatory capital. This calculation stresses the importance of management’s discretion in setting reserve levels. Increasing reserves from the current 0.45% level by fewer than six basis points to just under 0.51% would put the company out of capital compliance at March 31, 2002. As we have discussed previously, if the company were to increase reserves as a percentage of non-performing loans to the standards that are typical for large commercial banks or for the Farm Credit Banks (resulting in an increase in reserves from 19% of nonperforming assets to 161% or more), the company would be below its minimum and critical capital requirements, causing dire regulatory consequences including forebearance from further growth, receivership, and potentially liquidation or winding up. Exhibit 5: Non-Performing Loans as a Percentage of Regulatory Capital 70% 60% 50% 40% 30% 20% 10% -Q1

-Q4

2002

-Q3

2001

-Q2

2001

-Q1

2001

-Q4

2001

-Q3

2000

-Q2

2000

-Q1

2000

-Q4

2000

-Q3

1999

-Q2

1999

-Q1

1999

-Q4

1999

-Q3

1998

-Q2

1998

-Q1

1998

-Q4

1998

-Q3

1997

-Q2

1997

-Q1

1997

1997

1996

-Q4

0%

Source: Farmer Mac SEC filings D.

Risk-Based Capital Requirements

In the recent first quarter earnings press release, Farmer Mac states that it believes that the new risk-based capital requirements would have been $32.3 million as of March 31, 2002 compared with the existing minimum capital requirements of $112.2 million. We find it astonishing that, according to Farmer Mac, the risk-based capital standards are below the current minimum capital requirements and below even the critical capital requirements of $56.1 million. At $32.3 million, Farmer Mac’s leverage including off-


Page 24

balance guarantees (but not derivatives) would be at least 137 to 1. The risk-based capital level was supposed to be calculated by determining: The minimum level of risk-based regulatory capital necessary for Farmer Mac to maintain positive capital during a 10-year period in which the most stressful credit and interest rate conditions occur. [Federal Register, Vol. 66, No. 71/April 12, 2001]

We do not find it credible that $32.3 million of capital is sufficient to support $4.4 billion of farm loans and guarantees particularly in light of the company’s current $88 million of non-performing loans. Therefore, we believe that either Farmer Mac is incorrect in its calculations or the risk-based requirements have not achieved their mandated goal. Despite the apparently modest requirements of the new risk-based requirements, Farmer Mac’s expressed concerns about the new requirements, in its filings and on the recent quarterly conference call, seem inconsistent with the minimal risk-based capital dollar amount disclosed in the press release. On the recent conference call, the UBS analyst asked the following questions and received the following responses from management. UBS Analyst: “Henry, you talked a lot about the regulatory capital standards and what you perceived as a distortion of the rules for some time so long. Well, the stress test capital [risk-based] requirements at thirty-two million is so far below your leverage test [i.e., minimum capital requirements] and I am struggling to figure out why that would limit your behavior even if they [the regulators] in their minds mispriced something, you know, you have so much wiggle room, why would it make a difference to your business plan?” Mr. Edelman: “It would only be if we had very rapid growth in newly originated large loans guaranteed off the balance sheet. We do not anticipate that right now that that constraint will be significant. As I said, we don’t see a [problem] on the three-year horizon.” UBS Analyst: “Are you implying that the regulatory capital – that the stress test requirement should be growing a lot faster than your leverage test [minimum capital].” Mr. Edelman: “No. I think what I am saying is that we see technical flaws in the riskbased capital rules; legal flaws and functional flaws that are inconsistent with the statute but that we are very able to live with it and, as you know, if the calculation of risk-based capital comes at a level lower than statutory minimum, we are still obligated to comply with statutory minimum.”

Management statements and the company’s public filings are inconsistent on the issue of risk-based capital. We suspect that the requirements may, in fact, impact the company’s ability to grow. If off-balance-sheet growth is constrained by the new requirements, it will interfere with Farmer Mac’s ability to grow its LTSPC product. On-balance-sheet growth, in addition to being more capital intensive, requires Farmer Mac to actually purchase loans from other institutions and finance them. The need for financing to buy loans on balance sheet will require the company to access more discount and mediumterm notes in the marketplace. At this juncture, we do not know what the appetite for additional discount and medium-term notes is for Farmer Mac in light of the relatively limited growth of on-balance-sheet assets in recent years.


Page 25

Obviously, the direct impact of any regulatory intervention would likely be dwarfed by the effect such a capital shortfall would have on Farmer Mac’s continued ability to raise funds to meet its obligations. E.

Farmer Mac’s Recent Sale of Preferred Stock

On May 6, 2002, Farmer Mac announced the placement of $35 million of 6.40% perpetual preferred equity. This issuance increased Farmer Mac’s equity capital to $169 million, but left its leverage at 34 to 1 of on- and off-balance-sheet liabilities and guarantees to total equity, and increased slightly the leverage to the common equity. This announcement raises a number of questions. Farmer Mac has repeatedly told shareholders in public statements, at investment meetings, and in its public filings that it has more than sufficient equity capital to grow its business and therefore has no need to access new capital for a number of years. According to the company’s December 31, 2001 10-K, it has the ability to grow its guarantee portfolio by more than $14 billion without raising additional equity capital: Based on the current minimum capital requirements established in the 1996 Act, Farmer Mac's current capital surplus of $15.5 million would support additional guarantee growth in amounts ranging from $562 million of on-balance sheet guarantees to more than $2.0 billion of off-balance sheet guarantees. Furthermore, should Farmer Mac deem it appropriate, on-balance sheet non-program assets (cash and cash equivalents and investment securities) of $1.4 billion could be replaced with on- and off-balance sheet program guarantees, resulting in the ability to carry additional guarantees ranging from $1.4 billion of on-balance sheet guarantees to over $5.1 billion of off-balance sheet guarantees. Ultimately, Farmer Mac could sell on-balance sheet program assets of $1.9 billion in order to support further increases of on- and off-balance sheet program guarantees, resulting in the ability to carry an additional cumulative $14.1 billion of offbalance sheet guarantees. Any of these transactions would, of course, be evaluated to optimize Farmer Mac's return on equity and capital flexibility. Accordingly, in the opinion of management, Farmer Mac has sufficient capital and liquidity for the next twelve months. [Emphasis added.]

We were surprised that the company was able to identify an investor19 to purchase $35 million of unrated, non-traded perpetual preferred stock at a 6.40% coupon. This interest rate is only a modest premium over the yield on the 30-year U.S. Treasury. The sale is a particularly remarkable accomplishment when you remind yourself that this is private perpetual preferred stock of an unrated company, not fixed-term debt of the U.S. government. One question that occurs to us is whether the purchaser of Farmer Mac's new preferred stock is an affiliate or customer of the company. Recently, in management’s press release on April 29, 2002, the company disclosed that the $167 million of preferred stock owned by Farmer Mac was issued by members of the Farm Credit System. Based on our 19

The private placement agent, Bear Stearns, has told us that the preferred stock was purchased by one institution that Bear would not identify. Gotham Partners Management Co., LLC

Page 26

review of the public filings, it appears that there are only two Farm Credit Banks that have preferred stock outstanding which is held by non-Farm Credit System members (Farmer Mac is not a member of the Farm Credit System). 20 The first preferred issue sold to non-System members was a $225 million term preferred completed in May 2001. In June 2001, another Farm Credit Bank issued $300 million. Both issues were sold to non-FCS members. If indeed the $167 million of preferred held by Farmer Mac is from these two system institutions, then its purchase represented 31% of the total preferred issued, making Farmer Mac critical to the success of these offerings. Recently, the SEC has focused on “round-trip” transactions between two corporations which attempt to better the two companies’ income statements and/or balance sheets, but, in substance, have no economic effect. We wonder whether one or more of the FCBs which previously issued preferred to Farmer Mac returned the favor by buying the $35 million of Farmer Mac preferred. Alternatively, perhaps Farmer Mac contemporaneously purchased a mirror-image block of $35 million in preferred stock issued by the same party that purchased Farmer Mac's $35 million of new preferred stock. At a minimum, substantially more disclosure on who purchased the preferred as well as details as to its terms would be helpful to shareholders in their understanding of these issues. So far, the company has not released a prospectus or certificate of designations for the offering, nor identified the purchaser.

20

We have not been able to confirm whether there are other Farm Credit Banks that have issued preferred stock to outside investors.


Page 27

VI.

FARMER MAC’S PROGRAM ASSETS AND “SECURITIZATIONS” A.

Program Versus Non-Program Assets

Farmer Mac’s assets are comprised of two broad categories: program and non-program assets. The program assets are largely agricultural mortgage-backed securities (AMBS) that are guaranteed as to payment of interest and principal by Farmer Mac. Farmer Mac has the ability to borrow up to $1.5 billion from the U.S. Treasury to meet its obligations under this guarantee, but cannot use these funds to repay any of its own discount or medium-term notes or other obligations.21 Non-program assets are those that are not invested in farm loans or related AMBS. Instead, these assets were purchased using the proceeds of excess borrowings by Farmer Mac. We begin our discussion with AMBS and securitizations. B.

Farmer Mac’s AMBS “Securitizations” - Is the Company Accomplishing Its Mission?

Farmer Mac was formed for the purpose of initiating a secondary market for securitizations of farm loans, which are otherwise illiquid once they are originated on a lender's books. Despite the company’s statements about creating an active secondary market for its AMBS, we have not been able to identify a single MBS trading desk on Wall Street that has executed a trade in recent years in Farmer Mac AMBS. Farmer Mac owns approximately 83% of the AMBS it has issued, with the balance held by institutions who received the AMBS in swap transactions and a few who purchased seasoned AMBS outright for a purchase price of face value in 2000 ($159.9 million) and 2001 ($65.9 million).22 Since the third quarter of 1998, Farmer Mac has not been able to sell its AMBS at a premium at issuance, which was the company’s business strategy and an important component of its profitability in the first two years of its post-1996-Act securitization program. We find the company’s disclosure of its failure to sell AMBS striking:

21

“Farmer Mac may, in extreme circumstances, issue obligations to the U.S. Treasury in a cumulative amount not to exceed $1.5 billion. The proceeds of such obligations may be used solely for the purpose of fulfilling Farmer Mac's guarantee commitments under the Farmer Mac I and Farmer Mac II Programs…

The United States government does not guarantee payments due on Farmer Mac Guaranteed Securities, funds invested in the stock or indebtedness of Farmer Mac, any dividend payments on shares of Farmer Mac stock or the profitability of Farmer Mac. [December 31, 2001 10-K]” 22

There is no disclosure on who purchased the $225.8 million of AMBS held by third-party investors. In light of the lack of interest on Wall Street for these securities, we wonder whether the buyers are in any way affiliated with Farmer Mac. We would like the company to disclose to shareholders whether any of the FCBs or other banks that sell loans or obtain guarantees from Farmer Mac purchased any of the company’s AMBS.


Page 28

Market volatility in the latter part of the third quarter [of 1998] resulted in lower rates on Treasury securities, but wider spreads on Farmer Mac debt securities and even wider spreads on AMBS. These conditions diminished the economic attractiveness of capital market sales of AMBS, due to lower potential gains on issuance, but facilitated Farmer Mac's retention of the AMBS in its portfolio at favorable spreads. Retaining the AMBS is expected to generate net interest income over the long term with a present value in excess of the foregone up-front gain on issuance. [September 30, 1998 10-Q; emphasis added.]23

Note how the deterioration in the value of Farmer’s AMBS somehow “facilitated” retention of these securities. Our translation of the company’s statement is as follows: “After the failure of Long Term Capital Management, investors’ tolerance for risk declined and there was a general flight to quality. As a result, newly issued Farmer Mac AMBS could no longer be sold at a profit and probably was only salable at a loss. Because of its inability to sell securities at a profit, the company elected to retain them.” While this retained AMBS strategy was initially described as “temporary” in the company’s filings [See December 31, 1998 10-K], it has effectively become a permanent strategy. At March 31, 2002, of the total Farmer Mac I and II securities issued by Farmer Mac, 83% were held in Farmer Mac’s portfolio, up from 30% at December 31, 1998. As a result, we do not believe that Farmer Mac is, in fact, accomplishing its stated mission to create liquidity in the agricultural capital markets because there are few, if any, real buyers for its AMBS at issuance. Farmer Mac has only been able to sell seasoned AMBS in recent years for prices precisely at par. In light of the decline in interest rates since these seasoned AMBS were originally created, we find it significant that Farmer Mac could not sell these higher interest-rate securities at a premium.24 C.

A Closer Look at the Substance of Farmer Mac’s “Securitizations”

We believe that a closer examination of Farmer Mac’s securitizations explains the reasons for this phenomenon. Farmer Mac's most recently filed prospectus supplement dated April 30, 2002 for “$29,341,646 Guaranteed AMBS” is representative, we think, of the “securitizations” they have done in recent years. You can obtain a copy of the prospectus from EDGAR although it is filed under a different entity from the one listed under the ticker symbol for Farmer Mac, namely Farmer Mac Mortgage Securities Corporation.25 23

Management’s statement that “Retaining the AMBS is expected to generate net interest income over the long term with a present value in excess of the foregone up-front gain on issuance,” is disingenuous if there was indeed no “upfront gain” available in the market. Management had previously sold AMBS securities with gains averaging less than 1% of principal amount, and, we believe, would probably have sold AMBS if even a small profit could have been generated because of the reduced capital requirements for offbalance-sheet guarantee assets. 24 If the securities could only be sold for par after interest rates declined, then it is highly likely that they could only be sold at a loss on the date of their original issuance. 25 While you are looking at the filings for Farmer Mac Mortgage Securities, you might want to examine the 8-Ks that are filed which purport to provide updates for “investors” who own Farmer Mac’s AMBS. Farmer Mac does not give delinquency or special servicing data in these reports as is customary for commercial mortgage-backed securitizations. As result, the data are largely irrelevant. Gotham Partners Management Co., LLC

Page 29

The first unusual fact about the securitization document is that there is no investment banker or dealer listed on the cover as the underwriter for the issue. Clearly, Farmer Mac with 30 total employees does not have its own sales force to sell these issues leading one to the conclusion that the company does not need a banker to sell securities to itself. Next, note that there are six “pools” of securities offered in the issue, making each pool extremely small compared to typical securitizations. As a result of the small size of these pools, we believe there is no possibility for a liquid trading market to develop in them. Five of the six pools in the prospectus are characterized by a near total lack of diversification. For example, pool JM1025 includes “two mortgage loans that constituted approximately 47.92% and 42.56% (by principal balance) of the aggregate principal balance of that pool.” Pool LM1023 includes “one mortgage loan that constituted approximately 92.29% (by principal balance) of the aggregate principal balance of that pool.” Pool QM1024 includes “two mortgage loans that constituted approximately 38.49% and 29.51% (by principal balance) of the aggregate principal balance of that pool." We do not believe that Farmer Mac’s securitizations are securitizations in reality, i.e., securities that are suitable for purchase by sophisticated third-party investors and that bring new sources of capital to the agricultural mortgage market. Therefore, we do not believe that anything of economic substance takes place when Farmer Mac purchases loans, puts them in a REMIC, and takes back “newly issued” AMBS kept on its balance sheet. D.

Benefits to Farmer Mac from “Securitizing” Loans Even if There is No Functioning Secondary Market for Agricultural Mortgages

We believe that Farmer Mac continues to "securitize" its loans even without any hope of selling them to third parties because there are other significant benefits to the company as a result of its securitizations. Retaining AMBS on balance sheet means these securities can be classified as “held-tomaturity” or “available-for-sale” at the company’s election. Held-to-maturity securities are carried at cost allowing the company to avoid writing down the value of these securities. Available-for-sale securities must be carried at fair value according to GAAP. If a company’s available-for-sale securities are not actively traded, as with the AMBS, the holder must calculate fair value. Farmer Mac retains a high degree of management discretion in valuing its AMBS: 11. FAIR VALUE DISCLOSURES Significant estimates, assumptions and present value calculations are used for the following disclosure, resulting in a high degree of subjectivity in the indicated fair values. Accordingly, these estimated fair values are not necessarily indicative of what Farmer Mac would realize in an actual sale or purchase. [December 31, 2001 10-K]


Page 30

In addition, by “securitizing” its loans, Farmer Mac may gain access to the U.S. government’s $1.5 billion credit line, if needed, to meet its AMBS guarantee obligations, although it is hard to understand how Farmer Mac could draw on the Treasury where Farmer Mac itself is both guarantor and beneficiary of the guarantee. Until the adoption of new accounting rules on April 1, 2001, transforming loans into AMBS also allowed Farmer Mac to book as income a 50-basis-point guarantee fee “paid” by the REMICs26 which “issued” the securities. While this guarantee fee reduces the net interest income generated by the mortgages underlying the securities, the changed character of this income allows Farmer Mac to promote the substantial growth of what it calls "annuity-like" guarantee fees. In the last few years, guarantee fee income paid by Farmer Mac REMICs to Farmer Mac has averaged in excess of 20% of pre-tax income. The company omitted this break-out of guarantee fees for the first time from its 2001 10K. The importance of guarantee fee income to Farmer Mac is represented by the manner in which management promotes this income stream to Wall Street. In a 2001 third quarter press release, CEO Edelman stated: The steady pace of Farmer Mac loan purchases through the Farmer Mac I and Farmer Mac II programs, as well as long-term standby commitments, took our outstanding guarantee portfolio past the $4.0 billion mark for the first time. That represents a net increase of over $1.0 billion compared to the year-earlier level. Revenue growth from the annuity-like income stream from outstanding guarantees, combined with prudent interest rate risk management, again delivered a strong bottom line. [Emphasis added.]

As management presumably knows, Wall Street values “annuity-like” income streams at much higher multiples than net interest margins. E.

Lack of Independent Trustee for Farmer Mac's Securitizations

Farmer Mac recently replaced the Trustee for a number of its securitizations. The REMIC trustee's role is to report to investors and to insure that the issuer of the securitization meets its legal obligations to investors. Since the inception of the securitization program, Farmer Mac had always used an independent, third-party trustee. In the December 31, 1999 10-K, however, Farmer Mac made the following statement: U.S. Bank Trust National Association, a national banking association in Minneapolis, Minnesota, serves as trustee for each trust underlying Farmer Mac I Securities, although Farmer Mac may assume some or all of the trustee function and, thus, eliminate some of the cost associated with a third party trustee as soon as practicable. [Emphasis added.]

In a June 27, 2001 prospectus supplement, Farmer Mac was listed as the Guarantor and the Trustee. We find it extremely surprising that Farmer Mac, with 30 total employees in 26

REMIC is an acronym for Real Estate Mortgage Investment Conduit. A REMIC is a trust with flowthrough tax treatment that is used to hold mortgage securities.


Page 31

the entire company, has the resources to take on the additional role of trustee. The REMIC trustee business is a notoriously low-margin business that requires enormous scale, expensive computer and other systems as well as appropriate staffing to be done effectively, resources which we believe are not present at Farmer Mac. The lack of a third-party trustee is certainly a negative for the AMBS purchaser, unless of course, Farmer Mac is both the purchaser and the issuer. F.

Repurchase of $189.8 Million of Farmer Mac I AMBS

In the Q2 1999, Farmer Mac purchased $189.8 million of Farmer Mac I AMBS and $69.4 million of Farmer Mac II AMBS for a total of $259.2 million of AMBS from “capital market investors." [June 30, 1999 10-Q] The sheer size of this repurchase, representing more than 36% percent of outstanding AMBS not owned by Farmer Mac at the time, strikes us as unusual. When questioned about the purchase at our April 8, 2002 meeting, the CFO, Nancy Corsiglia, stated that “the AMBS were sold at a premium and purchased at a premium.” When asked from whom the purchase was made, management stated that the AMBS were purchased from one fund that went out of business in 1999, but would not disclose the name of the fund. What puzzles us is why Farmer Mac had to pay a premium to buy back such a large percentage of AMBS from one investor. In light of the highly illiquid nature of Farmer Mac’s AMBS, one would normally expect such a large repurchase to be achieved at a discount because of the lack of other buyers in the marketplace. Also, one might reasonably ask why Farmer Mac would pay a premium in repurchasing its own AMBS when, since the third quarter of 1998, it has not been able to sell its newly issued AMBS at a premium. We believe the reason why Farmer Mac purchased these AMBS securities at a premium may have something to do with SFAS 115 that requires a company to mark its availablefor-sale securities at fair value. Were a large block of AMBS to have traded at a discount to Farmer Mac’s carrying value for similar securities, its auditor would likely have required the company to mark its securities on-balance sheet at similar values. This would likely have had the effect of reducing the company’s capital, potentially below its regulatory minimum requirements.


Page 32

VII.

NON-PROGRAM ASSETS A.

Farmer Mac’s Non-Farm-Related Assets

Beginning in early 1997, Farmer Mac decided to borrow more funds in the short-term debt market than were needed to purchase agricultural loans. These excess funds became what are called “non-program assets,” that is, assets not invested in farm loans. As of December 31, 2001, non-program assets (other than cash) on Farmer Mac's balance sheet totaled $1.0 billion. Farmer Mac’s charter, however, restricts the company’s ability to borrow to purchase non-program assets.ii The practice of overborrowing and investing the proceeds in non-program assets is called GSE arbitrage and has been widely criticized.27 Farmer Mac explained its initial decision to overborrow as follows: During the first quarter of 1997, Farmer Mac also undertook a strategy to increase its presence in the capital markets, particularly the debt markets, in order to attract more investors to its debt and mortgage-backed securities and thereby improve the liquidity of its securities and reduce its borrowing and securitization costs. The Board and management believed that increasing Farmer Mac's presence in the capital markets would improve the pricing of its AMBS, and thereby enhance the attractiveness of the loan products offered through its programs for the benefit of agricultural lenders and borrowers. Since the implementation of the debt strategy, the Corporation has experienced a tightening of its AMBS spreads relative to other comparable agency securities and anticipates continued improvements in pricing as liquidity and investor recognition increase through the expanded debt issuances. The Corporation's eventual objective for the proceeds of its increased debt issuances is investment of those proceeds in Qualified Loans purchased under the Farmer Mac programs. During the phase-in of that objective, Farmer Mac will be investing a portion of those proceeds in high quality interest-earning assets, which have generated, and should continue to generate, increased interest income. [March 31, 1997 10-Q]

Up until the implementation of GSE arbitrage, Farmer Mac either lost money or was only modestly profitable. The significant incremental income from GSE arbitrage has contributed meaningfully to the company’s profitability. We do not consider GSE arbitrage to be related to the company’s mission despite its initial argument that overborrowing will reduce spreads on its AMBS. The fact that AMBS spreads have widened beginning in Q3 1998 to a level which no longer allows Farmer Mac to book a profit on the sale of newly issued AMBS, further demonstrates the fallacy of the company’s initial argument for increasing “its presence in the capital markets.” The new strategy has had a dramatic effect on Farmer Mac’s income statement with 36% of interest income coming from non-program assets in 2001, up from 19% in 1996. Initially, the company pursued a low-risk strategy with respect to the proceeds of its borrowings, which were invested in short-term, high quality, highly liquid assets: Pending use of the proceeds of Notes issued for the above-described purposes, such proceeds will be invested in accordance with the policies established by the Board in the following investments: (1) securities and obligations issued or guaranteed by the United 27

“Baker Raises Concern about Farmer Mac Viability,” Press Release, June 9, 1999.


Page 33

States Treasury ("Treasury Securities") or by agencies and instrumentalities of the United States government ("Agency Securities"), including those on which interest is payable at maturity ("zero coupon" or "strip" securities); (2) repurchase agreements for Treasury and Agency Securities; (3) commercial paper rated Al by Standard & Poor's Corporation and Pl by Moody's Investors Service, Inc.; (4) guaranteed investment contracts issued by banks or insurance companies that are rated AAA by Standard & Poor's Corporation and Moody's Investors Service, Inc.; (5) short-term borrowings between banks, known as "federal funds," provided that such banks have a rating of C or better from Thomson's BankWatch; (6) negotiable certificates of deposit and bankers acceptances issued by commercial banks and thrift institutions rated at least "B/C" by Thomson's BankWatch; and (7) corporate money market funds, such as the Merrill Lynch Institutional Fund and comparable funds, that invest in short-term diversified, high quality money market securities; and (8) asset-backed securities that are rated AAA. [December 31, 1995 10-K]

Eighteen months later, this policy apparently changed: The increase in the average balance of interest-earning assets was primarily due to an increase in the average balance of investments and cash equivalents resulting from the implementation of Farmer Mac's expanded debt issuance strategy. The increase in net interest yield was due to a shift in the composition of the investment portfolio from short-term, highly liquid investments to longer-term floating-rate investments which generally have higher spreads. The shift toward longer-term investments was primarily attributable to growth in floating-rate agency securities. [June 30, 1997 10-Q; emphasis added.]

Over time, the company’s policy toward risk in its investment portfolio has become more liberal. It now allows investments in preferred equity and large exposures to individual credits. Note the relatively high degree of exposure of Farmer Mac to a large number of different issuers in its investment portfolio. As of December 31, 2001, Farmer Mac had investments in commercial paper, corporate debt securities, asset-backed securities and preferred stock issued by 52 entities totaling $750 million, of which 26 exceeded 10 percent of Farmer Mac's stockholders' equity (the cumulative balance of investments in such entities totaled $624.5 million), and 9 entities with a total balance of $331.0 million exceeded 15 percent of stockholders' equity. In addition, as of December 31, 2001, Farmer Mac held $373.2 million of securities issued by GSEs or agencies of the U.S. government and $298.6 million in money market investment accounts, with the maximum amount held in any one money market investment fund at any time during 2001 being approximately $350 million. The shortterm nature of the investment portfolio also limits Farmer Mac's credit risk. As of December 31, 2001, 41.6 percent of the investment portfolio, excluding GSE and agency investments, consisted of short-term highly liquid investments. [December 31, 2001 10K; emphasis added.]

Federal Reserve regulations for bank holding companies limit a bank’s investments in debt securities of any one firm to 10% of the bank’s total equity. It is surprising to us that Farmer Mac’s regulators do not require similar diversification. At December 31, 2001, investments in each of 26 issuers “exceeded 10%” of Farmer Mac’s stockholder’s equity and investments in each of nine issuers securities “exceeded 15% of stockholder’s equity.”iii The failure of any one of these issuers could put Farmer Mac near or even below its minimum capital requirements. The 2001 10-K does not


Page 34

disclose how much above 15% of Farmer Mac's equity any one of the nine represents. We can presume that it falls under the “concentration limit” of 25% that Farmer Mac has set to “mitigate” the credit risk in the non-GSE, non-governments portion of its investment portfolio, but we find troubling such a high degree of exposure to various undisclosed credits in its portfolio.iv In light of the complex and concentrated nature of the investment portfolio, we believe that it is critical that the manager of the portfolio has sufficient acumen and experience. During our meeting with the company on April 8, 2002, we asked Mr. Edelman for the name of the employee who manages the investment portfolio and what his or her background and experience was. Mr. Edelman refused to answer the question saying that there was insufficient time left in the meeting to respond. He proceeded to take questions from other investors for the next thirty minutes, but would not tell us nor the other investors in the room who manages the portfolio. We believe that investors are entitled to know who the Chief Investment Officer is, and should have the opportunity to meet with the CIO and ask pertinent questions about the company’s investments, risk parameters, and portfolio strategy. B.

Farmer Mac’s Investment in Preferred Stock Issued by Its Customers

As of December 31, 2001, $750 million of the company’s $1.0 billion of non-program assets were invested in commercial paper, corporate debt securities, asset-backed securities and preferred stock. Of the $750 million, preferred stock represents $166.7 million. The company’s investment in preferred stock represents 124% of the company’s $134.4 million of book value at year-end 2001. Even a 10% decline in the value of this preferred stock could have put the company below its minimum capital requirements at December 31, 2001. In Farmer Mac’s press release on April 29, 2002 in response to The New York Times article of April 28, 2002, the company stated: While it is true that Farmer Mac owns approximately $166 million in preferred stock, the article omits to state that these are mission-related assets, issued by Farm Credit institutions (also Government Sponsored Enterprises) to facilitate their lending activities, and supported by their agricultural mortgage assets, pp. 1-2.

The April 29th press release is the only public disclosure we know of that states that the preferred stock was issued by members of the Farm Credit System. In light of the material nature of this disclosure, we find it surprising that it appears only in a press release response to a critical article in the media and in no other public filings. We beg to differ with the company that preferred stock in its Farm Credit Bank customers is a mission-related asset. Farmer’s Mac’s mission was to create liquidity in agricultural mortgages by guaranteeing and securitizing portfolios of qualified loans. Its mission, as we understand it, is not to make investments in preferred stock of members of the Farm Credit System. The riskiness of preferred stock in an institution as leveraged as


Page 35

a bank is self-evident. The illiquidity of private preferred could also pose a significant problem. Finally, we believe the potential conflict in making investments in its customers is similarly evident. The December 31, 2001 10-K also highlights that the portfolio included 41.6% of highly liquid, short-term investments, with the remaining 58.4% presumably invested in illiquid, long-term investments. If you trace the development of Farmer Mac’s investment policies for its non-program assets over time, the risk profile of the company’s non-program assets has increased materially while their liquidity has declined. When we asked Mr. Edelman the reason for the increasing investment risk, he stated that the company wanted to generate more income. He did not, however, seem concerned about the additional potential risks.


Page 36

VIII. ASSET-LIABILITY MISMATCH COULD CAUSE A LIQUIDITY CRISIS Prudently managed financial institutions match fund their assets and liabilities so that they can repay their liabilities when due from the proceeds of assets that have reached their term. By match funding, a financial institution is not speculating on short-term interest rate changes and can therefore better understand the true profitability of its lending. More than 72% of Farmer Mac’s liabilities - approximately $2.3 billion - come due approximately every seven days. In the event that Farmer Mac could not refinance its short-term debt, it would not have adequate liquidity available from its assets (most of which are long-term and of limited liquidity) to provide the cash to meet these obligations. Farmer Mac, unlike nearly every corporate borrower of which we are aware, does not have bank lines to back up its commercial paper program. We consider this to be a paradigm of an asset-liability mismatch despite management’s statements to the contrary.28 Farmer Mac’s balance sheet is exposed to both interest rate and refinancing risk as a result of the mismatch: (1) The interest rates on Farmer Mac's seven-day debt can fluctuate rapidly, potentially turning against the company just as rapidly and dramatically as recent interest rate declines have benefited the company; (2) A general disruption in the commercial paper markets, or in the specific market for Farmer Mac’s debt, could lead to a potentially fatal liquidity crisis for the company. A.

Refinancing Risk

Farmer Mac relies heavily on its GSE status in adopting a funding policy that we believe is extremely high risk. We do not believe that investors in Farmer Mac debt securities have done adequate due diligence on the underlying credit of the company because they believe that Farmer Mac is a AAA credit or alternatively, that the U.S. government will assume its obligations in a default. 1. Farmer’s Mac’s Debt is Widely Perceived as AAA, but is in Fact Unrated. Investors may be under the mistaken impression that Farmer Mac’s debt is rated Aaa by Moody's, which it is not. Until recently, Bloomberg mistakenly listed Farmer Mac as Aaa rated. The error on Bloomberg has now been corrected, and all Farmer Mac securities are now described as NR or Not Rated.v Particularly in the currently skittish debt market, the realization that Farmer Mac debt is unrated and riskier than previously thought could cause buyers of its discount notes to balk at providing further financing.

28

According to perhaps the preeminent text on money market instruments, Marcia Stigum’s, The Money Market, 3rd Edition, under her glossary definition for “match fund”: “A bank is said to match fund a loan or other asset when it does so by buying (taking) a deposit of the same maturity.” p. 1222. Gotham Partners Management Co., LLC

Page 37

2. Farmer Mac’s Debt is Not Guaranteed by the Government. Some investors and analysts have argued – and the behavior of the credit markets seems to indicate that this belief is widespread – that a funding mismatch is a non-issue for Farmer Mac because GSE debt carries the “implicit guarantee” of the U.S. Treasury and that, therefore, there will always be willing buyers of the discount notes. By charter, however, Farmer Mac must explicitly state that its securities are not guaranteed by the U.S. government: Each obligation shall clearly indicate that the obligation is not an obligation of, and is not guaranteed as to principal and interest by, the Farm Credit Administration, the United States, or any other agency or instrumentality of the United States (other than the corporation). [Farmer Mac Charter from Title VIII of the Farm Credit Act of 1971, as amended 12 U.S.C. 2279 ss et seq.]

Moreover, the only access Farmer Mac has to the U.S. Treasury is its ability to borrow up to $1.5 billion (on unspecified terms) to meet Farmer Mac’s guarantee obligations for its farm-mortgage guarantee obligations. The company cannot access these funds to repay any of its own discount or medium-term notes or other obligations.29 The savings & loan crisis of the 1980’s demonstrated the financial hazards of government guarantees which encouraged investors to ignore risk in their desire for return. The hazards are greater in the case of Farmer Mac because there is no government guarantee for discount or medium-term note holders, just a hope of one. For reasons we will discuss in greater detail in our conclusion to this report, we do not believe the U.S. government will step in to guarantee Farmer Mac’s debts. While the U.S. government protected depositors in the S&L crisis as it expressly promised to do, bondholders of insolvent S&Ls and their shareholders bore substantial, if not total, losses. B.

Interest-Rate Risk

Farmer Mac has been a significant beneficiary of the recent decline in short-term rates and the currently steep yield curve because of the mismatch between its short-term liabilities and long-term assets.30 This has led to a widening of the company’s historically narrow interest margins, greater prepayment penalty gains, and short-term profits for the company. These earnings, however, have not come without risk. Were short-term rates to rise substantially, the company would quickly become unprofitable because of its narrow net interest margin. 29

“Farmer Mac may, in extreme circumstances, issue obligations to the U.S. Treasury in a cumulative amount not to exceed $1.5 billion. The proceeds of such obligations may be used solely for the purpose of fulfilling Farmer Mac's guarantee commitments under the Farmer Mac I and Farmer Mac II Programs…

The United States government does not guarantee payments due on Farmer Mac Guaranteed Securities, funds invested in the stock or indebtedness of Farmer Mac, any dividend payments on shares of Farmer Mac stock or the profitability of Farmer Mac. [December 31, 2001 10-K]” 30 The company’s stock price performance neatly coincides with Greenspan’s aggressive short-term rate reduction over the last two years. Gotham Partners Management Co., LLC

Page 38

The company claims that it is match funding, but we find little evidence that it is in fact doing so. On the most recent earnings conference call, the Chief Executive Officer, Mr. Henry D. Edelman, was asked by the Robertson Stephens analyst why the company does not take advantage of the current low level of long-term interest rates to replace its shortterm debt with long-term debt. Mr. Edelman answered: We try to the best of our ability to match fund – so for us to go out on the [yield] curve although it is intuitively attractive, it would be inconsistent with our policy [1st Quarter conference call, April 19, 2002]

We have not been able to decipher what Mr. Edelman meant by this statement. The company seems to suggest that match funding is taking place, despite the fact that a substantial amount of its assets are fixed rate, because many of its assets, like its liabilities, are floating rate: Farmer Mac's primary strategy for managing interest-rate risk related to Farmer Mac I and II Securities and other assets held for investment is to fund them with liabilities that have similar durations, or average cash flow patterns over time, and provide flexibility to accommodate changing prepayment rates in changing interest rate environments. To achieve the desired funding objective, Farmer Mac uses a mix of short-term Discount Notes and callable and non-callable Medium-Term Notes. See Note 5 to the Consolidated Financial Statements. By using a mix of liabilities that includes callable debt, the duration of the liabilities will tend to increase or decrease as interest rates change in a manner similar to changes in the duration of the assets (the rate of change in the duration of an asset or liability to a change in interest rates is referred to as convexity). [December 31, 2000 10-K; emphasis added]

We are uncertain what the company means by “average cash flow patterns over time.” In any event, it is unlikely that the adjustment frequency of the company’s long-term floating rate assets match those of its short-term, seven-day liabilities. This mismatch benefits the company in declining interest rate environments, but can render it unprofitable and potentially insolvent in rising rate environments where its short-dated liabilities reprice faster than its longer-term assets. C.

“Market Value of Equity”

The company defends its management of interest rate risk by utilizing a “Market Value of Equity” (MVE) test. In an April 29, 2002 press release in response to a recent article on the company in The New York Times, the company states: While the article suggest that the company’s assets and liabilities are not matched, they are in fact closely matched, as evidenced by the relative insensitivity of “market value of equity” to interest rate fluctuations, as disclosed in the previously mentioned form 10-K, pp. 3-4.


Page 39

Despite the company’s and some analysts’ apparent comfort with this measure, one of the several disclosed qualifications to MVE makes it, we believe, a largely useless indicator. Below, the company describes the methodology for the MVE test: The simulation of MVE involves generating multiple paths for future interest rates starting from a "base" yield curve and then discounting the estimated cash flows under those rate paths to arrive at the estimated fair value of Farmer Mac's assets, liabilities and off-balance sheet items. Farmer Mac uses a commercially developed model to perform the MVE analyses. The analysis, which is based on Farmer Mac's existing assets, liabilities and off-balance sheet financial instruments, does not assume any new business and measures the change in MVE under seven interest rate scenarios. The interest rate scenarios include a "base case" in which the "base" yield curve is equal to the current yield curve, and six parallel and instantaneous shocks to the "base" yield curve (plus and minus 100, 200 and 300 basis points). Inherent in the MVE sensitivity analysis presented is the assumption that interest rate changes occur as instantaneous parallel shifts in the yield curve; in reality, such shifts are rarely instantaneous or parallel. In addition, actual future market conditions may differ materially from those assumed in the analysis. For example, actual loan prepayments and Farmer Mac AMBS and debt spreads may differ significantly from those assumed in the analysis. Accordingly, the results of the MVE sensitivity analysis should not be viewed as a projection of future results. [December 31, 2001 10-K; emphasis added]

We find it unsupportable to assume, as does the MVE model, that the shape of the yield curve will remain the same over time, i.e., assuming that, for every interest-rate period from seven days to thirty years, that rates will move up or down in lockstep. Farmer Mac currently benefits from the current upward-sloping yield curve. The resulting wide spread between the highest-grade commercial paper rates and the 10-year treasury is near a five-year high. The following chart illustrates the volatility in the spread between 7-day commercial paper and the 10-year Treasury. Exhibit 6: Spreads are Volatile and Near a 5-Year Peak

M ay Au -97 g N -97 ov Fe -97 b M -98 ay Au -98 g N -98 ov Fe -98 b M -99 ay Au -99 g N -99 ov Fe -99 b M -00 ay Au -00 g N -00 ov Fe -00 b M -01 ay Au -01 g N -01 ov Fe -01 b M -02 ay -0 2

3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% -0.5% -1.0% -1.5%

Spreads Between A1/P1 Rated 7-Day CP and 10-Year Treasuries

Source: Bloomberg


Page 40

For the company to rely on an MVE model which makes such an unfounded, artificial assumption strikes us as imprudent, particularly when there are much more sophisticated, real-world, interest-rate sensitivity models available in the marketplace.


Page 41

IX.

GOVERNANCE and COMPENSATION A.

Where are the Directors?

Farmer Mac’s board is comprised of 15 members. Five are appointed by the President of the United States. The 10 others are appointed five each by the Class A and Class B shareholders. The Class A stock is permitted to be held only by banks and other financial institutions and trades in limited volume on the NYSE. The Class B stock is privately held by institutions of the Farm Credit System. The non-voting Class C stock trades on the NYSE and is held by non-affiliates of Farmer Mac which are primarily institutional and individual investors. All institutions that wish to sell loans or obtain guarantees from Farmer Mac must purchase Class A or Class B shares, depending on whether or not they are members of the Farm Credit System. These transactions are disclosed at length in the related-partytransaction section of the proxy that is filed annually. The most recent disclosure is a good example of the extent of the affiliate transactions.vi The company’s only independent directors are the five political appointees (average age of 70) who by charter cannot have any affiliations with financial institutions, but two must have farming experience.vii These five directors are responsible for vetting fairly sophisticated loan purchase and guarantee decisions including those with affiliated shareholders, yet, remarkably, they are not permitted to have financial institution experience: which members shall not be, or have been, officers or directors of any financial institutions or entities [Farmer Mac Charter]

One wonders how directors with no financial institution experience can adequately represent shareholders in these matters. B.

Compensation

Farmer Mac’s option plan is large and extremely favorable to management. Currently, approximately 13% of outstanding shares have been issued to management in the form of options and restricted stock. The authorized option pool is more than 20 percentage points larger at 35% of the outstanding shares. Needless to say these are relatively and absolutely large numbers. When one considers that there are only 30 employees at Farmer Mac, these option and restricted stock grants seem disproportionately larger. In addition, the company’s option plan has an accelerated vesting schedule which provides that one third of the options granted vest immediately, an additional one third in 12 months, and the balance 12 months thereafter. By comparison, most company option plans vest ratably over four years, with the first grant vesting 12 months after the options are issued.


Page 42

X.

WHERE ARE THE AUDITORS?

Farmer Mac was audited by KPMG from its inception in 1988 to 1999. The company dismissed KPMG in early 1999, replacing it with Arthur Andersen. When we asked management why the company switched auditors, Mr. Edelman said that the company was not getting adequate attention from KPMG because it also audited Fannie Mae. This answer might make sense except for the fact that Andersen was Freddie Mac’s auditor until recently. While Freddie and Farmer shared the same auditor, they also shared a practice of booking guarantee fees on their own securities held in portfolio. This is not a material source of income for Freddie Mac, but, as we have shown above, it is a material source of income for Farmer Mac. Fannie Mae and its auditor KPMG do not use this same aggressive practice. Farmer Mac’s audit costs historically have been extremely low. We find it difficult to believe that Andersen could possibly audit the validity of the company’s reserves on approximately 10,000 loans for a total 2001 audit fee of $94,700. Farmer Mac recently dismissed Andersen replacing it with Deloitte & Touche. The company will not release a D&T audited financial statement until the filing of the December 31, 2002 10-K. We question whether D&T will allow Farmer Mac to continue accounting practices that we believe are aggressive and disclosure that we believe is inadequate and misleading.


Page 43

XI.

CONCLUSION – SMALL ENOUGH TO FAIL?

Investors have been willing to purchase Farmer Mac discount and medium-term notes at an immaterial incremental yield to Fannie Mae and Freddie Mac securities. This is despite the fact that Fannie and Freddie notes offer substantially greater liquidity, receive the highest AAA rating by multiple rating agencies, and have underlying assets of much higher quality, diversification, and liquidity. We believe that the reason why Farmer Mac’s securities trade at comparable prices to Fannie Mae and Freddie Mac securities is that investors have done little research on Farmer Mac’s credit and rely on its so-called GSE “implicit guarantee.” As previously shown, however, by charter Farmer Mac must explicitly state that its securities are not in fact guaranteed by the U.S. government. Based on our research on the issue, the notion of an “implicit guarantee” began in the 1970s. Nearly every mention of “implicit guarantee” in the media and in Congress is coupled with a discussion of how Fannie Mae and Freddie Mac are “too big to fail,” and how they have made a significant contribution to homeownership through their dominance of the home mortgage market. In a letter written by Alan Greenspan to Rep. Richard H. Baker (R-LA) on May 19, 2000, Mr.Greenspan states: The lower borrowing costs of these institutions, of course, reflect the belief of purchasers of their debt that the government is unlikely to let a GSE fail. In part this perception results from these housing-related GSEs initially being established as federal government entities to carry out specific government policies, and that, despite their subsequent privatization, these institutions continue to have government missions, which confer upon them a special status in the eyes of many investors.

The three criteria that appear to be responsible for the perception that Fannie Mae and Freddie Mac have an implicit guarantee are: (1) Fannie and Freddie were originally established as federal government entities that were subsequently privatized; (2) they have accomplished and continue to accomplish their stated missions of creating liquidity in the U.S. housing market; and (3) they are too big to fail. We do not believe Farmer Mac meets any of these criteria. In particular, Farmer Mac is not too big to fail. We believe this to be true because Farmer Mac has largely failed in its chartered mission to create an active secondary market in farm loans, and its assets are de minimis compared to Fannie and Freddie. In addition, with its guarantee products covering less than 3% of the agricultural loan market, the company is largely irrelevant to the health of the agricultural lending market. Unlike the home mortgage market, which is very dependent on Fannie’s and Freddie’s participation, the farm lending market is not dependent on Farmer Mac. The Farm Credit System has more than ample capital to provide farm loans on attractive terms to farmers. Similarly, commercial banks and insurance companies continue to be active participants in farm lending. In addition, we do not see any significant repercussions to the U.S. capital markets from a failure of an entity of the relatively small size of Farmer Mac. Enron was allowed to fail


Page 44

despite its enormous size and its financial interconnectedness with numerous counterparties in the capital markets. Treasury Secretary O’Neill’s commentary on Enron’s failure is suggestive of his limited desire to rescue failing businesses with taxpayer funds. In a statement on January 11, 2002, he stated: Companies come and go. It's part of the genius of capitalism. People get to make good decisions or bad decisions and they get to pay the consequences or enjoy the fruits of those decisions.

We believe that the Treasury department would be extremely averse to stepping in to save Farmer Mac because of the greater implications for Fannie and Freddie. In effect, a bailout of Farmer Mac would be tantamount to an acknowledgement by the Federal government that it guaranteed Fannie Mae’s and Freddie Mac’s debts. In any case, the Treasury could protect investors who are relying on Farmer Mac’s AMBS guarantee as well as banks that are counterparties to the LTSPC guarantees without helping the equity or debt holders of Farmer Mac. The moral hazards of saving these risk-taking investors are obvious and significant. In bank failures, the U.S. government has traditionally protected depositors who had an explicit guarantee. Stock and bond holders of these institutions, however, did not have this safety net and therefore suffered significant if not total losses. In the event of a financial crisis at Farmer Mac, we expect the same government policies to apply. Caveat Emptor!


Page 45

Endnotes

i

From the 2001 10-K: The Underwriting Standards require, among other things, that the loan-to-value ratio for any Qualified Loan (other than a part-time farm loan and a loan on an agricultural facility with a related integrator contract) not exceed 70 percent. In the case of newly originated Agricultural Real Estate Qualified Loans that are not part-time farm loans, borrowers must also meet certain credit ratios, including: (1) a pro forma (after closing the new loan) debt-to-asset ratio of 50 percent or less; (2) a pro forma cash flow debt service coverage ratio on the mortgaged property of not less than 1:1; (3) a pro forma total debt service coverage ratio, including farm and non-farm income, of not less than 1.25:1; and (4) a pro forma ratio of current assets to current liabilities of not less than 1:1. In early 1998, Farmer Mac introduced a premium loan program for loans to highly creditworthy borrowers. Under that program, Qualified Loans meeting certain more stringent Underwriting Standards than the foregoing loan-to-value and credit ratios would qualify for guarantee at a lower fee than those applicable to loans not meeting the higher standards. In 1999, Farmer Mac introduced a loan product for borrowers with high credit scores and whose security properties have low loan-to-value ratios. For these borrowers, loan processing has been simplified and documentation of the credit ratios described above is not necessary. In the case of a seasoned loan (a loan that has been outstanding for five or more years), Farmer Mac considers sustained performance to be a reliable alternative indicator of a borrower's ability to pay the loan according to its terms. A seasoned loan generally will be deemed an eligible Qualified Loan if it has been outstanding for at least five years and has a loan-to-value ratio (based on an updated estimate of value) of 60 percent or less, and there have been no payments more than 30 days past due during the previous three years and no material restructurings or modifications for credit reasons during the previous five years. Existing loans that have been outstanding for fewer than five years must comply with the Underwriting Standards for newly originated loans when the loan was originated. In the case of Rural Housing Qualified Loans and Qualified Loans under the part-time farm program, up to 85 percent of the appraised value of the property may be financed if the amount above 80 percent is covered by private mortgage insurance. For newly originated Qualified Loans on part-time farm properties, the borrower must generate sufficient income from all sources to repay all creditors. A borrower's capacity to repay debt obligations is determined by two tests: (1) the borrower's monthly mortgage payment-toincome ratio should be 28 percent or less and (2) the borrower's monthly debt payment-to-income ratio should be 36 percent or less. ii

The Corporation (and affiliates) may issue debt obligations solely for the purpose of obtaining amounts for the purchase of any securities under paragraph (1) [securities that “represent interests, or obligations backed by, pools of qualified loans.”], for the purchase of qualified loans (as defined in section 8.0(9)(B)), and for maintaining reasonable amounts for business operations (including adequate liquidity) relating to activities under this subsection. [Title VIII of the Farm Credit Act of 1971as amended 12 U.S.C. 2279aa et seq. [Emphasis added.] iii

It is interesting to compare the 2001 10-K disclosure with the 2000 10-K which states:

As of December 31, 2000, Farmer Mac had investments comprised of commercial paper, corporate debt securities and asset-backed securities outstanding with 78 entities totaling $637 million, of which 20 exceeded 10 percent of Farmer Mac's stockholders' equity (the cumulative balance of investments in such entities totaled $315 million), but none of which exceeded 15 percent of stockholders' equity. In addition, at December 31, 2000, Farmer Mac held $470.8 million of securities issued by GSEs or agencies of the U.S. government and $230.1 million in money market investment accounts, with the maximum amount held in any one money market investment fund at any time during 2000 being approximately $141 million. [December 31, 2000 10-K10-K, emphasis added.]


Page 46

iv

The policy is outlined in the December 31, 2001 10-K:

The credit risk inherent in other investments held by Farmer Mac is mitigated by Farmer Mac's policy of establishing concentration limits and investing in highly-rated instruments, which reduce exposure to any one counterparty. Farmer Mac's policy limits the Corporation's total credit exposure to a single entity by limiting the dollar amount of investments with one entity, excluding GSEs and agencies of the U.S. government, to the greater of 25 percent of Farmer Mac's regulatory core capital or $25 million. The policy also requires the entity to be rated in one of the three highest rating categories of at least one nationally recognized statistical rating organization for investments with terms greater than 270 days and in one of the two highest rating categories for investments with terms of 270 days or less. v

When we first began our research on Farmer Mac, we examined the trading prices of Farmer Mac on Bloomberg by typing the following commands: FAMCA Corp Go, and a list of the company’s securities appeared. In addition to coupon and maturity, the table of securities includes a column labeled “RTNG” which stands for credit rating. In this column up until February 20, 2002, all of Farmer Mac’s securities were listed as AAA. Typing DES to obtain a more detailed description on any of the securities on the list gave further detail indicating that Farmer Mac was rated Aaa by Moody’s. One of the brokers with whom we made a bearish investment on the company called Moody’s to get a copy of the Farmer Mac rating report. The representative at Moody’s said that they had never heard of the company, and certainly did not rate it. This prompted the broker to call Bloomberg to ask why the securities were listed as AAA Moody’s-rated on Bloomberg. The Bloomberg ratings representative said that the “issuer told us that it was rated Aaa.” The broker informed Bloomberg that Moody’s did not, in fact, rate the securities of Farmer Mac, and shortly thereafter, in the RTNG column on Bloomberg Farmer Mac’s securities are described as “NR” or not rated as it continues to specify today. While we did not speak to this representative at Bloomberg directly, the broker that had the conversation is someone whom we trust implicitly. He works for one of the most highly regarded, bulge-bracket investment firms on Wall Street. We do not doubt the veracity of his statements. Despite the fact that Farmer Mac’s securities are unrated, they trade as if they were. The spread between Fannie, Freddie and Farmer is typically fewer than 10 basis points on even longer-term paper. For example, Farmer Mac’s 5.4% of October 2011 closed on May 8, 2002 at a yield to maturity of 5.87% versus Fannie Mae’s 5 3/8 of 11/11 which closed on the same day at a yield to maturity of 5.79%, an eight basis point spread. vi

“Certain Relationships and Related Transactions

John Dan Raines is a member of the Board of Directors of AgFirst Farm Credit Bank ("AgFirst"), a Farm Credit System institution with which Farmer Mac and Fannie Mae have entered into a joint arrangement for the pooling of Rural Housing Qualified Loans. Under the arrangement, AgFirst purchases eligible Rural Housing Qualified Loans for pooling through the Farmer Mac I program and Farmer Mac guaranteed securities issued in connection therewith are to be purchased by Fannie Mae with a guarantee fee payable by AgFirst to Farmer Mac and Fannie Mae. During 2001, Farmer Mac guaranteed securities having an aggregate principal amount of approximately $223.9 million were issued under the arrangement among AgFirst, Fannie Mae and Farmer Mac. AgFirst also acts as a central servicer and contract underwriter for Farmer Mac in the Farmer Mac I program. During 2001, AgFirst received approximately $99,700 in fees as a central servicer and $17,000 in fees as a contract underwriter. Kenneth E. Graff is the President of Farm Credit West, ACA, which is the successor to Central Coast Farm Credit. Central Coast Farm Credit acted as a central servicer in the Farmer Mac I program. During 2001, Central Coast Farm Credit received approximately $221,000 in servicing fees as a central servicer. Peter T. Paul is a director of GreenPoint Financial Corp., an affiliate of GreenPoint Mortgage, which acts as


Page 47

a central servicer in the Farmer Mac I program. During 2001, GreenPoint Mortgage received approximately $30,750 in servicing fees as central servicer. From time to time, Farmer Mac purchases, or commits to purchase, Qualified Loans under the Farmer Mac I program and Guaranteed Portions under the Farmer Mac II program from institutions that own five percent or more of a class of Voting Common Stock or that have an officer or director who is a director on the Farmer Mac Board. These transactions are conducted in the ordinary course of business, with terms and conditions comparable to those applicable to lenders unaffiliated with Farmer Mac. In 2001, Farmer Mac purchased: (a) 206 Qualified Loans having an aggregate principal amount of approximately $90.8 million from Zions (Zions is the holder of approximately 32% of Farmer Mac's Class A Voting Common Stock and W. David Hemingway, a Class A director, is Executive Vice President of Zions and its holding Company); and (b) one Qualified Loan having a principal amount of approximately $185,000 from AgFirst (John Dan Raines, a Class B director, is a member of the Board of Directors of AgFirst). In 2001, Farmer Mac guaranteed: (a) through a long-term standby purchase commitment with AgFirst, 1,404 Qualified Loans having an aggregate principal amount of approximately $213.3 million; (b) through a long-term standby purchase commitment with AgStar Farm Credit Services, two Qualified Loans having an aggregate principal amount of approximately $94,700 (Paul A. DeBriyn, a Class B director, is President and Chief Executive Officer of AgStar Farm Credit Services); and (c) through a long-term standby purchase commitment with Central Coast Farm Credit, 446 Qualified Loans having an aggregate principal amount of approximately $349.4 million (Kenneth E. Graff, a Class B director, is the President of Farm Credit West, the successor to Central Coast Farm Credit). The principal amount of Guaranteed Portions that Farmer Mac purchased under the Farmer Mac II program from director-affiliated institutions or five percent or greater stockholders was approximately 10.6% of that program's volume in 2001. During 2001, Farmer Mac (a) entered into Farmer Mac II transactions with Zions involving Farmer Mac's purchase of Guaranteed Portions or the issuance of Farmer Mac II guaranteed securities backed by Guaranteed Portions in an aggregate principal amount of approximately $16.2 million (8.2% of the program's total); and (b) purchased 26 Guaranteed Portions having an aggregate principal amount of approximately $4.7 million (2.4% of the program's total) from Bath State Bank (Dennis L. Brack, a Class A director, is the President and Chief Executive Officer of Bath State Bank). In addition to its participation as a seller of loans in the Farmer Mac programs, Zions also acts as a dealer in Farmer Mac's discount and medium-term note programs; is a counterparty to Farmer Mac on certain interest rate swap transactions; and acts as a central servicer and contract underwriter for Farmer Mac in the Farmer Mac I program. See "Compensation Committee Interlocks and Insider Participation" for quantitative information concerning Zions' contractual relationships with Farmer Mac.” vii

“5 members shall be appointed by the President, by and with the advice and consent of the Senate -

(i) which members shall not be, or have been, officers or directors of any financial institutions or entities; (ii) which members shall be representatives of the general public; (iii) of which members not more than 3 shall be members of the same political party; and (iv) of which members at least 2 shall be experienced in farming or ranching.” [Farmer Mac Charter.]


Page 48

Buying the Farm i by Bill Ackman

Recommend Documents