Barlcays Tesla Full Report

Equity Research 17 April 2018

Global Autos & Auto Parts

INDUSTRY UPDATE

Industrie 4.0 vs. Tesla’s ‘lights out’?

European Autos & Auto Parts

German OEM CEOs don’t need to sleep on the factory floor : In 2017 Tesla won the PR war, even if Chinese OEMs took the electric sales crown. But we expect EU premium carmakers to fight back in 2018/9. And it's not so much about headline-grabbing products (although the Porsche Mission E, Audi E-tron and Jag I-PACE are snapping at the heels of Tesla on range and performance), but about the more mundane war of production efficiency. Despite his vision for full factory automation, Elon Musk is sleeping on the factory floor again during his current Model 3 “production hell” woes, and tweeting about his underuse of human workforce. We think German OEMs are about to launch an onslaught of attractive AND profitable electric products; we expect them to hit targets. The German (and Japanese) model of a harmonious blend of robot and human, perfected over 100 years of 'kaizen', together with the benefit of Industrie 4.0 will ensure a speedy and efficient production ramp. BMW is our new Top Pick in European Autos, replacing VW (OW, PT €212). Underappreciated modular capabilities as well as a supportive product cycle and positive mix through 2018 and 2019 lead us to reiterate our OW on BMW with a PT of €115 (previously €114). We downgrade Daimler from EW to UW on the back of a relatively late and, as a result, more expensive commitment to electrification at a time of weakening earnings momentum and mix at Mercedes. We therefore apply a 15% discount to historical multiples, which takes our SotP-based PT to €73 (previously €82). The German OEMs don't just promise, they deliver: Silicon Valley may be better at PR and building brand, but we believe EU premium OEMs are quietly creeping up on their Space-aspiring counterpart (as are Chinese OEMs). Incumbent OEMs have >100 years of know-how on world-leading processes from Just-in-Time, Lean Manufacturing, to modular design and now Industrie 4.0, utilising AI (artificial intelligence) and IoT (Internet of Things). And the incumbents are generating serious cash and profit, even in the face of declining diesel and tech disruption. We believe BMW's modular production won't be threatened on range by native EV platforms. With the 5th-gen electric about to launch, we believe BMW is well placed to enjoy the returns of profitable BEV production, whilst peer Mercedes, Mercedes, on its first gen, may see margins at greater risk. And don't forget the importance of scale, global reach and product diversity : Whereas Tesla has design innovation to its credit, BMW (Mercedes and Audi too) have more cash to invest, a broader portfolio over which to spread costs and with which to entice customers, customers, global modular manufacturing (and distribution) expertise and long standing supplier relationships. relationships. With the surge in SUVs globally, the slow death of diesel should be more than offset by strengthe strengthening ning mix and modular product breadth. breadth.

Barclays Capital Inc. and/or one of its affiliates does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. This research report has been prepared in whole or in part by equity research analysts based outside the US who are not registered/qualified as research analysts with FINRA. PLEASE SEE ANALYST CERTIFICATION(S) AND IMPORTANT DISCLOSURES BEGINNING ON PAGE 78.

NEUTRAL Unchanged For a full list of our ratings, price target and earnings changes in this report, please see table on page 2. European Autos & Auto Parts Kristina Church +44 (0)20 3134 2199

[email protected] Barclays, UK Dorothee Cresswell +44 (0)20 7773 2192 [email protected] Barclays, UK Adam Rae, CFA +44 (0)20 7116 1579 [email protected] Barclays, UK U.S. Autos & Auto Parts Brian A. Johnson +1 212 526 5627 [email protected]

BCI, US Dan Levy, CFA +1 212 526 8047 [email protected] BCI, US Steven Hempel, CFA +1 312 609 7260 [email protected] BCI, US

Barclays | Global Autos & Auto Parts Summary of our Ratings, Price Targets and Earnings Changes in this Report (all changes are shown in bold) Company

Rating

Price

Price Target

Old New 16-Apr-18 Old

EPS FY1 (E)

EPS FY2 (E)

New %Chg Old New %Chg Old New %Chg

European Autos & Auto Parts

Neu Neu

BMW (BMW GY / BMWG.DE)

OW OW

90.66

114.00 115.00

1

11.88 12.26

3

12.49 12.64

1

Daimler AG (DAI GY / DAIGn.DE)

EW UW

65.31

82.00 73.00

-11

8.99 8.99

-

8.74 8.95

2

Volkswagen AG-PFD Preferred (VOW3 GY / VOWG_p.DE)

OW OW

172.14

195.00 212.00

9

25.43 27.24

7

27.38 30.03

10

Source: Barclays Research. Share prices and target prices are shown in the primary listing currency and EPS estimates are shown in the reporting currency. FY1(E): Current fiscal year estimates by Barclays Research . FY2(E): Next fiscal year estimates by Barclays Research. Stock Rating: OW: Overweight; EW: Equal Weight; UW: Underweight; RS: Rating Suspended Industry View: Pos: Positive; Neu: Neutral; Neg: Negative

Valuation Methodology and Risks European Autos & Auto Parts BMW (BMW GY / BMWG.DE) Valuation Methodology: We continue to use a SotP valuation to reach our PT app lying a 35% EV/Sales multiple, 4.4x EV/EBIT and 11x PE for the Auto business, just ahead of the historical average despite structurally higher margins and cashflow. We continue to value th e FS business at 1x book and China JV at 8x PE. We apply a 5% conglomerate discount to the whole. Risks which May Impede the Achievement of the Barclays Research Valuation and Price Target: The main downside risks to our price target, in our view, are: 1) Europe - shouldEuropean pricing and sales deteriorate due to unexpected macro-economic factors, as well as worries for the leasing and residual value environment intensify, intensify, BMW might struggle to reach our earnings targets; 2) US - should US demand for sedans weaken further and residual value fall to impact the financial services book, BMW might struggle to achieve our forecasts; 3) China risk - should the economic and demographic indicators in China prove misleading and the premium auto market fail to achieve the growth we forecast for the next 5 years, we could see downside risk to our current earnings estimates for BMW. Daimler AG (DAI GY / DAIGn.DE) Valuation Methodology: We derive our price target through a SOTP analysis based on a blend of EV/sales, EV/EBIT and P/E multiples. We apply a blended average of 30% EV/Sales, 3.7x EV/EBIT and 9.4x PE to the MBC business (a 15% discount to those we use for BMW due to weakening model momentum and therefore earnings dynamics at the core Mercedes brand). We value the trucks trucks business at a 15% d iscount to Volvo at market. We apply a 15% conglomerate discount to our overall SotP, which we think applicable given management's focus on many different business lines but improved focus in recent months on making legal structural changes within the overall group. Risks which May Impede the Achievement of the Barclays Research Valuation and Price Target: Risks that may drive the shares to outperform include: 1. if management were to achieve a faster separation of Mercedes and the Truck Division 2. if Mercedes were to show a greater earnings resilience, despite softer model mix, than our c urrent assumptions 3. if profitability at Mercedes were less impacted by the costs of electrification than we currently expect. Volkswagen AG-PFD Preferred (VOW3 GY / VOWG_p.DE) Valuation Methodology: To reach our current valuation for Volkswagen prefs, we employ a sum-of-the-parts valuation methodology using a blend of peer multiples for the different brands/businesses and market values for listed holdings and m inorities. We then apply a 25% conglomerate holding discount to the preference shares, taking into account the risks surrounding recent emissions scandals and corporate governance. Risks which May Impede the Achievement of the Barclays Research Valuation and Price Target: The main risks to our price target, in our view, are:1) The fallout surrounding the NOx emissions scandal and possible regulatory, legal, and recall costs. 2) External risks, such as macroeconomic factors, outside the control of the company, leading to a weaker demand and pricing environment than we currently assume that could make our forecasts difficult to achieve, including rising raw materials and volatile FX rates and 3) risks remain around corporate governance and management's focus on shareholders. Source: Barclays Research.

17 April 2018

2

Barclays | Global Autos & Auto Parts

CONTENTS .................................. ................................... ...................... .... 4 EXECUTIVE SUMMARY: 1) THEMATIC .................

EXECUTIVE SUMMARY: 2) COMPANY IMPLICATIONS ......................... 10 ................................... .... 13 TESLA HAS MANY GLOBAL EV COMPETITORS... ............................... ............ ..... 18 ...AND THE COMPETITION IS ABOUT TO HEAT UP FURTHER .......

GERMAN APPROACH: UNDERPROMISE, OVERDELIVER...ELON MUSK ................................... ................................... ....................... ..... 29 MORE FOCUSED ON AUTOMATION .................. “SMART FACTORIES” OR INDUSTRY 4.0 ARE GERMAN CONCEPT OF PERFECT SYNTHESIS OF ROBOT AND HUMAN AND INCORPORATE ................................... ................................... ................................... .................................... .................................. ................ 37 AI AND IOT................. ................................... .................................... ................................... ..................... 41 GERMAN OUTPUT QUALITY ................. ................................. ............................... .............. 43 GERMAN BEV PRODUCTION STRATEGIES ................

GERMAN OEMS ALREADY HAVE A GLOBAL PRODUCTION FOOTPRINT TO BUILD ON ..................................................................................................... 54 .............................................. ..................... 57 GERMAN OEMS HAVE DEEP POCKETS …. ............................. ................................... ................................... ................................... .................................. ................. 61 COMPANY SECTION ..................

BMW FIRST-MOVER ADVANTAGE AND MODULAR PRODUCTION APPROACH OFFER COMPETITIVE EDGE ON ELECTRIC VEHICLES MOVE TO TOP PICK, PT OF €115 ............................................................... 62 VW – ENJOYING ENJOYING THE BENEFITS OF SCALE IN ELECTRIC VEHICLE ......................... ..... 65 PRODUCTION - REITERATE OW RATING, PT TO €212 .................... DAIMLER - LATE AND THEREFORE MORE EXPENSIVE EV ............................ ..... 68 COMMITMENT - DOWNGRADE TO UW, PT TO € 73 ....................... .................................. ..................... 71 APPENDIX 1: COMPANY-SPECIFIC EV TARGETS ................. ..................................... .......... 73 APPENDIX 2: COUNTRY-SPECIFIC EV TARGETS ...........................

APPENDIX 3: BARCLAYS POWERTRAIN FORECASTS ............................ 74

17 April 2018

3


EXECUTIVE SUMMARY: 1) THEMATIC

The automobile industry is facing fundamental changes. Alongside the electrification of the powertrain, autonomous driving and the development of new markets, it is above all digitalisation that is driving this process of change. This combination of the physical and digital is ’s website) often referred to as "Industrie 4.0". (Daimler (Daimler’s Elon Musk wanted to automate electric vehicle production... The above quote on “Industrie 4.0” is taken from Daimler’s website but could just as much apply to VW and BMW. The old Toyota Production System (TPS), incorporating the notion of continuous improvement (kaizen) and highlighting the indispensability of people to the efficiency of the production line, may now seem anachronistic in a world of AI (artificial intelligence) and the IoT (Internet of Things). But we agree with the arguments put forth by our Macro colleagues (see Robots at the gate: Humans and technology at work , 10 April 2018) that technology and humans should work symbiotically with each other and that soft, not hard, automation should be the first step until technology and AI (artificial intelligence) advance enough to welcome the Era of the Robot. As long as incumbent OEMs remain flexible in their employment requirements, we believe they will be prepared for the shifts.

...but 13 April he tweeted that he had underappreciated the human element Despite Tesla’s CEO, Elon Musk, arguing on multiple analyst earnings calls about the importance of automation and his desire to reach a “lights out” scenario with Model 3 production (ie, full automation with only minimal human intervention), it seems that Tesla’s current production woes have led to a change of heart. In fact on Friday last week, @elonmusk replied replied to a tweet from Wall Street Journal journalist Tim Higgins to say:

Robots won’t kill all automotive jobs imminently By contrast, Tesla’s “lights-out” or fully automated vision may be more fitting with the current zeitgeist, but we worry it is overlooking the importance of the human touch (and could also lead to exploding capital intensity and complexity). We strongly believe traditional carmakers are more focused on the social-economic demands of manufacturing, than their Silicon Valley peer. And as a result, Tesla CEO Elon Musk highlighted at FY17 results that he is living through “production hell” and is now back b ack to sleeping on the factory floor. By contrast, we expect the ramp of BEVs (battery electric vehicles) from traditional OEMs in the coming years to be efficient, profitable and on target.

17 April 2018

4


"Yes, excessive automotion at Tesla was a mistake. To be precise, my mistake. Humans are underrated." @elonmusk (in reply to timkhiggins) 8,54pm, 13 April 2018 ….German Automakers actually know how to make great vehicles (ICE or EV) efficiently We believe there is an Industrial Revolution occurring in the German auto industry but it is about the perfect synthesis of human and robot, not about removing all human elements from the production system The German OEM’s OEM’s may be losing the PR war but that doesn’t mean they won’t win on EV production efficiency (and therefore profitability)

We believe there is an Industrial Revolution occurring in the German auto industry but it is about perfecting the combination of human and robot, not about removing all human elements from the production system. German carmakers are also fully focused on digitalisation within the manufacturing process, but we think valuations do not reflect any belief in the success of new BEV launches or superior production efficiencies to start-up peers.

Traditional OEMs taking the fight to Tesla with a raft of new BEV launches There is no denying that Tesla has been a major disruptive force in the automotive industry over the last few years and garnered reams of press space over the same period. If we were to judge a car brand’s performance by the number of re -tweets or clicks, Tesla would trump the German car industry hands down. And of course, more importantly for investors, the market cap of c.$50bn (or c.$500k/car produced last year) highlights that the PR machine of Tesla has done a better job at convincing the market of the business’ potential future worth, than the incumbent OEMs, with BMW’s market cap per car produced at a mere €23k and GM’s at only $8.3k. But that's before the ramp of new, exciting BEVs (battery electric vehicles) that's coming over the next few years, just as Tesla is struggling to mass produce the Model 3.

There are two key points we plan to explore in more detail in this report: 1) Can Tesla break into mass-market manufacturing and does a higher degree of automation help? 2) Are native/standalone platforms for electric vehicles discernibly better than fully modular production across all powertrains?

1) Does Tesla's highly automated approach make sense? We have questioned for some time now the resilience of Tesla's business model to mass manufacture and believe the "blue pill" pillars may be starting to crack and question the faith of the key ‘purple pillers’ (See Challenging the blue pill pillars of faith, 29 March 2018). Tesla’s production missteps, combined with increased competition, make it harder for the Model 3 to achieve the ‘iPhone moment’ , in our view. We remain bearish on Tesla, as we believe investors have not adequately appreciated the myriad risks involved in Tesla’s aggressive Model 3 ramp. And don't rule out the traditional OEMs, who have been perfecting cost-effective, highly automated processes over 100 years of manufacturing know-how from Lean Manufacturing, to Just-in-Time and TPS (Toyota Production System), to modular toolkits and now Smart Factories/Industrie 4.0, We think they know how to produce EVs that have been designed to cost (DTC). A joint study between McKinsey and A2Mac1 (“Trends in electric vehicle design” design”, October 2017) recently pointed out that second-generation EVs tend to achieve far improved performance and component integration than their predecessor predecessors. s. 17 April 2018

5

Barclays | Global Autos & Auto Parts 1

"One machine can do the work of 50 ordinary men. No machine can do the work of one extraordinary man." Elbert Hubbard, philosopher (1856 – (1856 – 1915) 1915) Soft not hard automation We believe that until machine learning is advanced enough to more fully replicate human intelligence, the importance of the human touch is still necessary. We are currently in a period of “soft” not “hard” automation within automotive manufacturing, but the advent of “Smart Factories” or “Industrie 4.0” (a concept that originated in Germa ny) highlights that incumbent OEMs are very much capable of embracing technological change when needed. For a much fuller discussion on these topics, our Macro colleagues published a series of impact studies last week – for for a summary see Equity Gilt Study 2018, 10 April 2018 and for more in-depth on AI, see Artificial Intelligence: A primer , 10 April 2018.

That is not to say that automotive employment isn’t changing drastically MGI research recently concluded that 45% of activities that people are paid to perform today could be automated in future. Toyota and the incumbent OEMs emphasise the importance of people to ensure quality and “kaizen” in the manufacturing process, but this does not mean that they are not all preparing for a more drastic shift in jobs as the advent of electrification and Artificial Intelligence converge. On a recent roadshow with VW management, the company confirmed that they are approaching automotive employment in a flexible nature as the shift from ICE to electrification and connectivity/autonomy coincides with less appetite from younger workers for blue-collar factor roles. This is enabling VW to push forward with the Future Pact to reduce blue-collar headcount by 23k, whilst also focusing on hiring an additional 9k workers in the field of IT and software development. We think the incumbent OEMs are as aware of the need to actively reallocate resources to face up to technology disruption as Tesla is of the need to retain talent.

2) Are native EV platforms always better than powertrain modularity? Additionally, we question the perceived wisdom (see McKinsey 'Trends in electric-vehicle design' October 2017) that only native/standalone EV platforms will succeed for manufacturing a consumer-acceptable battery electric vehicle. We believe that native platforms are definitely the ultimate process over the long term and, for those businesses with scale over which to spread the additional expense (eg, VW), also make sense in the nearer term, but for premium manufacturers the return does n’t fit with the cost outlay. We acknowledge that a native BEV (ie, vehicle developed on a standalone platform designed designed just for BEVs) has the ability to optimise the battery packaging and position and achieve optimal energy capacity. Native EVs also achieve a larger interior space (sometimes by as much as 10%) for the same wheelbase. In theory this means native EVs can obtain best range too. We therefore think it makes sense for a company with the scale of the VW Group to pursue this strategy. It may be higher risk, given the greater costs associated, but given the learnings of MQB (which stands for Modularer Querbaukasten, translating from German to "Modular Transversal Toolkit"), we think it will prove a successful and cost-effective solution. But we also note the nearer term premium cars (Audi E-tron and Porsche Mission E), will be launched on a derivative of the existing MLB-platform i.e. a modular platform

1

17 April 2018

https://www.economist.com/blogs/sch https://www.econom ist.com/blogs/schumpeter/2012/10/z-business-qu umpeter/2012/10/z-business-quotations-2 otations-2

6

Barclays | Global Autos & Auto Parts similar to BMW's strategy, rather than the new performance car standalone BEV platform (PPE) which will come in the early 2020s. FIGURE 1 Native electric vehicles may be the ultimate solution for interior space but non-native vehicles can reach a high range at a more cost-effective cost-effective price point Vehicle range versus price 90 80 70 60

Price1

Sales thousands (EUR)

50 40

Non-native

30

Native

20 10 0 0

10 0

200

30 0

400

50 0

600

Rangekm

Source: Barclays Research, US Department of Energy

We think for premium OEMs, a modular strategy is more fitting shorter term as BEV penetration penetrat ion ramps. We believe the advances in modularity made by premium OEMs such as BMW are being underappreciated by the market for their ability to generate superior range (the iX3 is expected to have >300 miles), but at a margin-prote margin-protective ctive fixed cost base. We also believe that given the multiple approaches to EV powertrain and battery design (eg cylindrical, pouch and prismatic) in the market currently, a company with a fully flexible manufacturing approach, approach, like BMW, provides itself with the best ability to adjust design if at a later date the industry converges on an optimal technology solution. And even if no one solution prevails, we believe BMW will be able to offer its customers a fully modular choice of battery pack, vehicle styling and range. Given the advances already made on ICE (internal combustion engine) vehicles with powertrains designed to cost (DTC), BMW will leverage on these cost advantages via i ts modular approach, while peers pursuing standalone standalone/native /native platforms will have a much steeper learning curve to overcome.

17 April 2018

7


FIGURE 2 Design approaches to managing EV powertrain and battery thermal management still vary widely among OEMs, which is why we believe keeping a flexible approach is key to manufacturing until a convergence on a single technology solution solution Pow ertrain

Battery

Charge

DC-DC

AC-DC

Electric

Module

converter

inverter

Motor

Gearbox

Cooling

Liquid

Resistive

Heating

heating

N issan LEA F

N one

VW e-Golf

N one

While plugged in or on battery

N one

BMW i3

While plugged in

T esla S 60

While plugged in

Legend Water cooling

Active

Passive

cooling

cooling

Interconnections

Source: Barclays Research, A2 Mac1, McKinsey Centre for Future Mobility

Other areas explored in today’s report include: Incumbent OEM EV launches are set to threaten Tesla’s early lead With BMW poised to launch its fifth-generation electric electric drivetrain and GM far advanced now via the Chevy Bolt, we believe these vehicles will achieve considerably higher performance and cost efficiency than peers that are only at first-generation. We look at how different electric vehicle specs are converging with those of Tesla o n range, performance and, in our view, with much more ability to ramp to scale production smoothly and in a cost-efficient manner. manner.

German OEMs are underestimated for their advances i n “Smart Factories” Industrie 4.0 should not be overlooked, in our view. It is the German concept of ultimate symbiosis of man and machine that sets the German engineering capabilities of the incumbent automakers automakers apart from their Silicon Valley peer. We believe this understanding is hard to replicate quickly and also offers protection versus the ramp of new BEV offerings coming out of China.

We challenge the market assumption that >100 years of automotive engineering know-how needs to be thrown out the window with the advent of BEVs. We are not questioning Elon Musk’s visionary prowess nor his advances in increasing consumer acceptance of electrification, thus forcing automotive incumbents to face up to disruption, but we highlight why we think the market is underestimating the German OEMs (and GM too). We also challenge the view that the Chinese have as much opportunity to deliver on EVs as their European counterparts. We agree that they have government support, and access to cheap battery cell supply is in their favour, but we think this underestim underestimates ates the importance of the deep learning that the traditional OEMs have enjoyed in the field of efficient 17 April 2018

8

Barclays | Global Autos & Auto Parts manufacturing and the importance now of “smart factories”. We believe the Germa n OEMs in particular are leading the way in this field.

TPS, Lean-manufacturing and kaizen have been perfected by traditional automakers over many years... We think Elon Musk knows how to make a great car, but BMW and GM (and not to rule out the traditional Asian automakers too) know how to make great AND profitable BEVs at scale. From the Toyota Production System (TPS), to VW’s MQB (modular platform), the incumbents have been applying “kaizen” over many years of manufacturing processes to produce “Just -in-Time” products at the highest quality levels.

We believe scale may aid incumbent OEMs versus Tesla... We look at the global production footprints of traditional OEMs and their ability to spread costs over a wider base, as well as the benefit of long-established supplier and dealer relationships. The market may be more excited about potential break-up opportunities amongst the German conglomerates, but we think currently the benefits of scale can have its advantages for incumbents versus start-up disruptors.

...and traditional OEMs have deep pockets too Finally, we explore the superior spending capabilities of traditional OEMs and the years of R&D and capex know-how that has enabled them to build high quality vehicles. In our view, OEMs have been refining exactly the right levels of modularity, flexibility (fixed vs variable costs; automation vs human expertise), working capital requirements and vertical integration, as well as long-standing supplier relationships. With EV launches ramping in 2018/19 we believe the Porsche Mission E and BMW iX3 will be able to perform very strongly versus Tesla’s product of fering as well as leveraging the groups’ global breadth. We think it would be unwise to underestimate the importance of mass market BEVs like GM’s Bolt and the Hyundai Ioniq and their ability to steal some of the Model 3’s thunder. And the investment isn't just into vehicle production but also consumer understanding. understanding. The industry needs to persuade customers to want to buy BEVs as much as it has to work out how to produce BEVs profitably. We b elieve the OEMs’ superior cash balances compared to Tesla, as well as years of customer understanding via Captive Finance offerings, will help persuade customers to buy OEMs’ BEVs en masse.

17 April 2018

9


EXECUTIVE SUMMARY: 2) COMPANY IMPLICATIONS FIGURE 3 Three phases of electrification – electrification – who who is best placed?

Near term (2018-2019) – Funding Electrification BMW (OW, Top Pick)

VW (OW)

Daimler (UW)

Renewal of entire SUV line-up, drives mix

A flurry of new product launches

Waning new product momentum at Mercedes

Mid term (2020-2021) – Ramping Up EV Production BMW

VW

Modular production limits costs, maximises flexibility

Cost of dedicated BEV platform supported by scale

Capex and R&D expenditure begin to normalise normalise

Capex and R&D expenditure begin to normalise normalise

Daimler Dedicated BEV platform platform and associated cost Capex and R&D spending remain remain elevated

Long term (2022+) – Winning the EV race Ultimate winners/losers winners/losers as yet unclear

Source: Barclays Research

FIGURE 4 VW and BMW have an improving SUV mix whilst Daimler has peaked % SUV Models 35% 30%

% SUV Models 38%

% SUV Models 34%

36%

32%

34%

30%

32% 25% 20%

30%

28%

28%

26%

26%

24%

24%

15%

22%

22% 10% 2 01 01 5

2 01 01 6

20 17 17

20 18 18 VW

20 19 19

2 02 020 2 02 021 OEM average

2 02 022

20% 2 01 01 5

2 01 01 6

20 17 17 20 18 18 BMW

2 01 019 2 02 020 2 02 021 OEM average

2 02 02 2

20% 2 01 01 5

2 01 01 6

2 01 01 7

2 01 01 8 D AI AI

20 19 19 20 20 20 2 02 021 O EM EM a ve ve ra ra ge ge

2 02 022

Source: IHS, Barclays Research

17 April 2018

10


FIGURE 5 Tailwinds from volume/price/mix volume/price/mix versus headwinds for technology transition transition for Mercedes and BMW 1690

740 490 290

190 10

-247 -566

-130

-570

-813 -1010 2018E

2019E

2020E

MBC tailwind fr from price/volume

BMW tailwinds fr from price/volume

MBC headwinds from tech transition

BMW headwinds from tech transition

Source: Barclays Research forecasts

FIGURE 6 BMW has made large advances as it moves from 1st to 5th gen electric drivetrain

Source: BMW company reports, Barclays research

BMW moves to Top Pick in EU Autos We believe BMW has used the years since the i-brand launch launch to test the benefits of in-house full vertical integration versus outsourcing of electric drivetrain components. We expect more outsourcing on the iX3 and electric Mini of commoditised parts, which should keep costs within a reasonable range. We think BMW’s advancements in electrification are currently misunderstood by the market, which clearly favours the use of native BEV platforms. But with a positive mix of product (including a strong SUV renewal) and also less risk from residuals than the market currently credits, BMW is our favoured stock in the EU (previously ly €114). Autos space and we select it as our Top Pick with a price target of €115 (previous

17 April 2018

11


VW remains an OW stock with high optionality We also like VW's approach to electrification - to leverage on its knowledge of modularity via MQB to build a standalone/native modular EV platform called MEB. We don't think this strategy works outside of the mass market, given the additional costs in terms of retooling, etc, but with Dieselgate propelling VW along, we think its advanced outlook on EVs (electric vehicles) will stand it in good stead to gain profitable market share with a highly flexible EV product offering. With the appointment of Dr Herbert Diess as new group CEO and cultural and structural changes ongoing, we still see strong optionality within VW’s valuation and continue to rate the shares OW. Our price target increase s from €195 to €212 as we raise our 2018 and 2019 profitability forecasts for the VW and Audi brands to take greater account of new product momentum and the potential for additional synergies under the group’s new corporate structure .

DAI downgrade to UW and €73 PT for short-term earnings pressure at Mercedes We believe the Daimler Group is very well placed on future mobility trends and will by no means be shown up as a ‘dinosaur’ in the industry. However, it is as yet unclear exactly where Mercedes' approach to electrification fits between its German peers. We suspect if Mercedes pursues only native EVs, their margin dilution will be substantial. In today's note we downgrade the DAI stock to UW, not because we don't think the Group will survive in a world of electrification and autonomy, but because for the next few years we believe that the step-up in spending relative to peers and lower mix improvements from new product will leave Mercede Mercedess’ margins at greater risk. We apply a 15% discount to historical premium auto multiples in our SotP analysis to take into account waning earnings momentum at Mercedes and therefore reduce our price target from €82 to €73.

Tesla – Model Tesla – Model 3 highly attractive, but we remain UW the stock on the production ramp ramp Meanwhile for Tesla, we give credit for the leadership Tesla has displayed in accelerating global EV uptake, and indeed we acknowledge that at this time the Model 3 is clearly the most attractive EV option at its price point. However, we continue to remain bearish on Tesla, with an UW rating and $210 price target, as we believe investors have not adequately appreciated the myriad risks involved in Tesla’s aggressive ramp. Margins are likely to remain compressed, while cash burn remains an issue. And while there may be some merit to the simplicity of the Model 3 structure and to Tesla’s lead in battery cost, any such advantages over legacy OEMs are likely more than offset by the legacy OEM s’ cost advantages in scale and other efficiencies.

17 April 2018

12


TESLA HAS MANY GLOBAL EV COMPETITORS... Incumbent OEMs selling a wider range of EVs than Tesla To date volumes from traditional car makers of BEVs (battery electric vehicles) have been relatively low, with many preferring to focus on PHEVs (plug-in hybrid vehicles) until infrastructure keeps up with consumer demand for product. But we see 2019 as the tipping point to full battery electric cars when costs have come down sufficiently for manufacturers to offer products at attractive price points and where the build out of charging infrastructuree (including the Ionity ultra-fast shared charging network funded by BMW, DAI, infrastructur Ford and VW which will see 400 stations with a capacity of up to 350kW across Europe by 2020) make the offerings more attractive to consumers, particularly as range extensions occur. To date, there have been few BEV vehicles to truly threaten Tesla’s lead (other than the largely unheard of EC-Series from BAIC, which is not only China’s top selling electric vehicle but also delivered >100k NEVs or new energy vehicles globally in 2017).We see the competitive competiti ve landscape changing drastically as we go through 2018 and 2019.

We expect BEV penetration to ramp on an S-curve We have been highlighting for many months, see Tesla's competition hots up as the ICE melts, 12 July 2017 and Future Powertrain 2: Are OEMs disruptors? , 15 August 2016, that we expect a sharp and steep S-curved ramp in BEV penetration, once battery costs come down, regulations force OEMs to comply with tougher CO2 targets and infrastructure builds. We have consistently viewed the hybrid period as an interim technology and argued that car makers are assured of their ability to make money on BEVs once they can break even on battery costs - we calculate this to be at $100/kWh, see below: FIGURE 7 We estimate the tipping point to EVs is $75-100 kWh Battery cost / kWh Component

$75

$100

$150

$200

$300

$350

$26, 000

$26, 000

$26, 000

$26, 000

$26, 000

$26, 000

Addi t i ona l HE P cost

$1, 600

$1, 600

$1, 600

$1, 600

$1, 600

$1, 600

Addi t i ona l 48V

$1, 100

$1, 100

$1, 100

$1, 100

$1, 100

$1, 100

$28, 700

$28,700

$28,700

$28, 700

$28,700

$28,700

-$5, 500

-$5, 500

-$5, 500

-$5, 500

-$5, 500

-$5, 500

E l ec ect ri c dri ve vet ra i n / pow er er el ect roni cs cs

$3, 25 250

$3, 250

$3, 25 250

$3, 25 250

$3, 250

$3, 250

50 k Wh ba t t ery (c. 240-290k m ra nge)

$3, 750

$5, 000

$7, 500

$10, 000

$15, 000

$17, 500

$27, 500

$28,750

$31,250

$33,750

$38, 750

$41,250

Mi d-si z e ICE vehi cl e

Mid-siz e ICE v e hic le with HEP & 48V less E ngi ne / t ra nsm i ssi on plus

T otal

Source: ICCT, University California Berkeley, UC Davis, Barclays Research

But car makers are largely in the camp that while hybrids (ranging from 48v mild hybrids to plug-in full hybrids) are a helpful interim technology to gain consumer trust in range and understanding of electric vehicles capabilities, as well as a helpful aid to achieve 2020/1 emissions targets in the face of decline diesel decline, they can never be produced profitably. We therefore assume a much steeper and sustainable ramp for BEVs than for hybrid vehicles (see Appendix 3 for our detailed powertrain powertrain forecasts by region). We are not alone in assuming that battery costs can come down significantly – it it is outside the scope of this report to focus on battery technologies, but we highlight two charts from P3, a 17 April 2018

13

Barclays | Global Autos & Auto Parts third party engineering and technology consultancy to the Automotive industry, highlighting their view of the economic case for BEVs vs ICEs (internal combustion engine vehicles): FIGURE 8 Battery costs are coming down and chemistry improvements evolving EUO / kWh 180

Between 2020 and 2021 100 EUR/kWh on battery cell level will be reached ～

160 140 120 100 80 60 2014

2016 NMC111

2018 NMC622

2020

2022

2024

NMC811

Source: PS3, Barclays Research. Note: NMC is L ithium Nickel Manganese Cobalt Oxide (the numbers represent ra tios)

FIGURE 9 On a TCO basis, BEVs are quickly reaching parity with ICE competitors New business models & approaches in emobility

BMW X4 xDrive35i

Jaguar I-PACE S / Audi Audi ETron

Tesla Model X

€61,250

€75,000

€91,250

8.3 l kWh/100km

18.0 kWh/100km

20,8 kWh/100km

Power:

225 kW

294 kW

245 kW

Battery:

-

90kWh

75kWh

Range:

-

500km

417km

Vehicle

875

987

1,003

Taxes / insurance

122

125

150

Consumption

164

78

84

Service

96

46

137

1,257

1,236

1,374

1.01 €/km

0.99 €/km

1.10 €/km

Model: List Price: Consumption:

Monthly costs (€/month)

Total cost of ownership Cost/km Source: PS3, Barclays Research

Tesla takes the sales crown in the US In the US, Tesla has clearly outsold all other brands for BEVs (battery electric vehicles) in 2017 and YTD 18 with the Model S, X and now 3. But the GM Bolt is hot on its heels and we think the market will look very different as we go beyond 2018 and into 2019. Not only is the Bolt a significant competitor to the Model 3 but Hyundai ’s Kona SUV also looks to be a realistic threat to Tesla’s attempt to break into mass market EV manufacture. And tha t’s before the premium OEMs start to launch their offerings.

17 April 2018

14


FIGURE 10 Tesla leading the charge in US BEV sales by model in 2017

FIGURE 11 Renault still leader for BEV sales in Europe in 2017

35,000

35,000

30,000

30,000

25,000

25,000

20,000

20,000

15,000 10,000 5,000 0

Source: IHS, LMC and Clean Technica

15,000 10,000 5,000 0


In Europe Renault-Nissan is currently the leading BEV manufacturer And the US isn’t the key market for EVs , as we expect penetration rates to explode fastest in China and then Europe, driven by government regulations and incentives. In Europe, the Renault Zoe remains the top selling car, followed by BMW’s i3. The Renault-Nissan group was early to the game on EVs with the Renault Zoe launched in Dec 2012 and ranked among world’s top sell electric cars 2015 and 2016. However, the jury is out on the success of the push into electric and whether the group was too early, and struggled to sell enough vehicles to reach profitability. The revolutionary concept of allowing consumers to lease the battery in the vehicle, helped to get around residual value fears but range anxiety and cost issues restrained sales.

Equally, the i3, which we will argue later in the report was revolutionary in its production ethos and set BMW as one of the earlier manufacturers to build a deep understanding of standalone electric architectures, has also suffered a mixed reception – with consumer worries more focused focused on the styling of the vehicle. However, in 2017, the i3 continued to surpass Tesla’s Model S in European sales volumes.

China is a much more competitive but fast-growing EV market Of course, with restrictions increasing in China for the sale of combustion engines (for further detail see pg 6f of European Integrated Oil & Refining: Postcard from China, 19 Jan 2018), European and US automakers are readying themselves for a speedy uptake of BEVs . We forecast BEV growth in China of 52% 2015-2020E. VW plans to deliver 400k NEVs by 2020 and 1.5mn by 2025 and GM 150k NEVs by 2020 and 500k by 2025. BYD is targeting 200k by 2018 (see Appendix 1 & 3). But it is a hugely fragmented market with multiple domestic players already making inroads into the BEV market, as can be seen from Figure 10 below. Currently 93% of EV sales belong to domestic brands, although we expect this to change in the coming years as European and US implants seek to compete in this lucrative market. While we will not focus on China in detail in today’s note or the competitive offering from the Chinese manufacturers, who are largely starting from a blank slate on BEVs, it is worth noting that there are numerous Chinese EV start-ups unveiling concept cars (we saw the Byton offering at CES in Jan, as well as XPeng’s G3). So we mustn’t forget that the Chinese may well lead the global push in EV penetration, given the substantial government support and also access to inexpensive Chinese-sourced battery cells.

17 April 2018

15


FIGURE 12 The Chinese BEV market is much more diverse and currently dominated dominated by domestic players 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0


Globally a Chinese manufacturer takes the lead on BEVs In 2017, BAIC’s EC-Series BEV product took the global sales lead with >100k vehicles, and Tesla ’s Model S came in second place. And it was also a Chinese manufacturer (this time BYD) which took the global lead on EVs (ie, including plug-in hybrids). But we think the landscape will change dramatically over the next 2 years. Until now, we believe European premium manufacturers manufacturers have been holding back their firepower on BEVs as they refine their manufacturing capabilities and wait for the optimum moment when consumer acceptance starts to coincide with regulatory demands.

FIGURE 13 Consolidated Consolid ated US, European and Chinese BEV sales 2017 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0


Global OEMs have aggressive plans for electrification, even if Tesla has been earliest to the game In Appendix 1 at the back of this report, we highlight the aggressive electrification plans of numerous global manufacturers, all looking to take on Tesla’s lead on electric. Are the German OEMs going to be too late to the PR war (if Renault’s experience in Europe is 17 April 2018

16

Barclays | Global Autos & Auto Parts anything to go by, you can definitely be too early to a new concept too). We think the brand engineering that Tesla has built can be surprisingly quickly eroded, especially when the impressive stats of the upcoming premium models are reviewed. Much has been written about Tesla’s lead on battery technology, providing a superior range to competitors, but in today’s note we focus more on the less -discussed lead that we expect to come from European production efficiencies. But first, let’s take a look at some of the upcoming, supposed “Tesla-killer” models:

17 April 2018

17


...AND THE COMPETITION IS ABOUT TO HEAT UP FURTHER What BEV products are hitting the market imminently? As we have noted previously, the competition is heating up for TSLA and we see that competition competiti on is accelerating in the coming years (see: “Tesla competition hots up as the ICE melts,,” 7/12/17). Importantly, we are now seeing some of the concept cars of the past melts coming to showrooms. While we concede that the Model 3 doesn’t face competition yet in the ~$50k fully loaded segment (which we note is smaller than the $35k segment), competitors are coming in both at the Model S/X ~$80-100k price point, and also at the more affordable affordable EV price point. Just in the US and Europe (never mind mind the profusion of of Chinese EVs) key scheduled launches include: Upper end 

Jaguar iPace (on (on sale 2Q18 in Europe, Europe, 3Q18 in US)



Audi eTron (on sale 2H18 in Europe then US)



Porsche Mission E (2Q19)



Mercedess EQC (late 2019) Mercede



BMW iX3 (early 2020)

Middle/lower end 

Chevy Bolt (on sale now)



Nissan Leaf (on sale now)



Hyundai IONQ EV (2018)



Hyundai Kona EV SUV (summer 2018 in Europe, 4Q18 in US)



GM Buick variant of Bolt



VW ID Neo (2020)

While there’s not enough room in this report to focus on all th e exciting new pure electric products, we go into a little more depth on a few of the launches which we think will be critical to Tesla’s competitive landscape:

Jaguar’s i -PACE -PACE The launch of Jaguar’s first all-electric offering this month, the I-Pace, is a product pitched as a direct competitor to the Tesla Model X, given its SUV styling. Its credentials look very attractive, with 0-60mph in 4.5 seconds, a range of around 500km and an approximate price point of €75-80k, ie, cheaper than an entry range Model X. Two electric motors are powered by a 90kwH lithium ion battery and it takes 2hours for a full charge or 45 minutes for 80% using super fast charge. Like the Porsche BEV, it uses the more expensive (but better performance) PSM motor and can reach 400PS and 696 Nm of torque. The first vehicles are due to be delivered this summer. Like Tesla, Jaguar is utilising the idea of OTA (over the air) updates to ensure the latest technology, 3 screens and ensuring best customer experience, with live charging updates to deal with range-anxiety. The company have given no detail on their volume targets but early reviews have been very favourable.

17 April 2018

18

Barclays | Global Autos & Auto Parts FIGURE 14 Jaguar I-Pace – I-Pace – the the first battery electric SUV launched to appeal to similar customers to Tesla ’s Model X

Source: Jaguar website

Audi ’s ’s e-tron e-tron Audi will supply the next big EV release in August of this year. We are looking forward to the hotly-anticipated hotly-anticipat ed launch event of the Audi e-tron in Brussels over the summer, but from the specifications already in the market, it looks to compete with Tesla on both range and performance. The e-tron CUV will be built using the MLB-Evo architecture, a derivative of VW’s MLB for longitudinally mounted combustion engines. However, to save costs in future years, Audi and Porsche are looking at a shared architecture for electric cars, called PPE (Premium Platform Electromobility) from c2022, and will be used for higher utility vehicles an Weckbach, Porsche’s head of BEVs, said he plans to take the best than the Mission E. Stef an modules from both the J1 and MLB-Evo platforms for the PPE development, which began 18 months ago (for a deeper discussion on VW’s platforms plans, see the following section of this report). But with 0-60mph in 4.5 seconds and a range of >310 miles and a 95 kWh battery, we expect the car to be a strong performer. Range-anxiety will also be aided by fast-charge capabilities within 30 mins.

17 April 2018

19

Barclays | Global Autos & Auto Parts FIGURE 15 Audi e-tron – e-tron – the the VW group’ group’s first electric SUV and launched using the MLB-evo architecture

Source: Audi website

Porsche “Mission E” One of the most exciting new BEV launches, we think, will be Porsche’s offering from next May (the concept is named Mission E). It will be a sport sedan which will seat 5 people, with an 800 volt battery, a range of c300 miles/500km (under NEDC so more like >250 miles under WLTP) and 0-60 mph in 3.5 secs. It will be capable of “Turbo Charging” , ie charging to 80% in 15 mins from ultra-fast DC at 350kW and 800 volt. All Ionity chargers will support 800 volt as will Electrify America, the VW built EV charging network in the US.

The Mission E will be produced at Zuffenhausen from 2019 with factory capacity of c20k/annum, although with some flexibility to flex higher if necessary. We will discuss platform strategies in the later section of this report, but while the parent group VW is focused on modularity, Porsche is aiming to retain its exclusivity via a standalone architecture for the Mission E, called J1. The Mission E Cross Turismo BEV crossover concept unveiled at Geneva this year will be based on the same platform.

FIGURE 16 Porsche Mission E – E – will will be the first electric 5- door sports-car product to fully compete against Tesla ’s Model S

Source: Porsche website

17 April 2018

20

Barclays | Global Autos & Auto Parts What is different about the Porsche BEV versus competitor products is that it will utilise a Permanent Magnet Synchronous Motor (PSM) versus Asynchronous Motors (ASM) used by other vehicles. We believe this will be a key differentiating factor versus Tesla’s Model S, as it will enable the car to maintain high speeds for a sustained period of time, eg on the German autobahn. ASM has high peak performance but lower continuous performance. From recent discussions with Porsche AG executives, it seems that Porsche management are comfortable that the lower complexity of BEVs will ultimately bring opportunities in terms of profitability (definitely in comparison to PHEVs, which require a much more complicated design due to the need for both a battery and combustion engine). We imagine BEVs will still come in below the profitability of ICE (internal combustion engine) for now, but management assure us that they can be profitable at 20k/annum. The project budget for the Mission E was c€700mn but employee employeess shouldered some of the burden for t his cost via paycuts, which they will hope to recoup at a later date. And don’t forget that Porsche also sells PHEVs, with the company to offer two variants in future including the high-performance “Turbo S E -Hybrid”. In fact, almost 60% of Panamera sedans Porsche now sells in Europe are the plug-in E-Hybrid version. And in addition, in the U.S. 20% of its Cayenne SUV sales are the E-Hybrid, and that's before the new generation of Cayenne E-Hybrid goes on sale. We believe this breadth of offering, along with the highly successful and profitable combustion engine vehicles, should enable the Porsche brand to compete head on with Tesla.

Wait until late 2019/20 for Mercedes and BMW’s new BEV launches Consumers have to wait for late 2019/early 2020 for the competitive offerings from Mercedes and BMW. Both companies will launch new BEV products pitched against the Model X given their SUV body styles.

Mercedes’ EQC is the first in the EQ range So far Daimler’s pure electric vehicle sales have been limited to the electric Smart (ForTwo and ForFour) first produced in 2009 and the electric Mercedes B Class launched in 2013. Daimler has sold a total of 28k electric smarts and just under 14k electric B class to date. Production of the electric B class was halted in Q3 2017 on the back of soft demand levels. We note the B Class Electric Dri ve’s electric powertrain was co developed and sourced from Tesla, which is in direct contrast to BMW’s i-brand capabilities, which ensured total inhouse competency and sourcing.

17 April 2018

21

Barclays | Global Autos & Auto Parts FIGURE 17 Mercedes EQC is the first of the EQ family products for Mercedes and shows the consumer preference for SUV styling

Source: Mercedes website

In November 2016, Mercedes-Benz announced a more serious EV effort. Under Daimler’s €10bn “emission free driving plan” the company will launch more than 10 pure electric vehicles by 2022, with BEVs contributing 15-25% of total volumes sold by 2025. As a first step Daimler will bring the EQC SUV to market in 2019, with two further variants to follow soon thereafter. The vehicle is expected to have a range of roughly 500km and can accelerate to 100km/h in under 5 seconds. We understand the battery pack with a capacity of over 70kWh is supplied by Deutsche Accumotive with cells from Korean manufacturer SK Innovation. Daimler intends to manufacture electric vehicles in 6 global locations across Europe (Bremen, Rastatt, Sindelfingen), China and the US (Tuscaloosa). We expect the product, like BMW’s iX3 to offer a range of different energy storage options and price levels, to ensure maximum customer appeal. However, consumers must wait another year to get their hands on the product. We also argue later in the report that we believe Mercedes are behind their Munich-peer in terms of in-house capabilities on electric and therefore have a steeper ramp up in spending over the coming years. Whilst BMW is preparing to launch the fifth-generation electric drivetrain, Mercedes’ EQ brand is still very much in its infancy. We think Mercedes’ recent delivery on product styling should enable a very successful reception for the EQ family at launch, but we believe the phasing of spending will hold back Mercedes ’ margins for the – hence hence today’s downgrade to UW. next two years and restrain share performance –

BMW came early to the EV game with i-brand/i-NEXT In Feb 2011, BMW launched the sub-brand i-brand , as a follow on from Efficient Dynamics. BMW were very early to the electric offerings with the fully electric i3 launched in 2013 and also the sportier plug-in-hybrid i8 in the same year. What we think the market often forgets about the ibrand is that it was not only about these two cars but focused on wholly renewable sourcing, even down to using responsibly sourced wool and leather and eucalyptus wood from sustainably managed forests. And the use of Carbon Fibre Reinforced Plastics (CFRP) in the vehicles, as well as light-weight aluminium, helped maintain the vehicles lighter footprint. The i3 had many critics on its styling but it still won two World two World Car of the Year Awards: World Awards: World Green Car of the Year and also World Car Design of the Year. The i3 was also given an iF an iF Product Design Gold Award. And Award. And at the 2017 the 2017 New York International Auto Show the newly launched 94Ah version of the i3 was named "World 17 April 2018

22

Barclays | Global Autos & Auto Parts Urban Car of the Year". BMW were also the first to offer the concept of fully flexible leasing packages on the i3 to overcome early adopters “range-anxiety” on BEVs (eg providing an i3 customer with an X5 for longer journeys for a specified number of days of the year). FIGURE 18 After the iX3, BMW returns to the sedan shape for the BMW iVision – iVision –

FIGURE 19 But the new electric Mini will keep more traditional styling

Source: BMW presentation

Source: BMW presentation

Currently BMW offers nine electrified models and management have stated their aim to launch 25 new electrified models, including 12 BEV models by 2025. In addition to the existing i3 and i8, other future models include the ix3 coming in 2020, iNEXT crossover, and an i4 sedan (previously called the i Vision Dynamics), which are due in 2021. BMW recently patented a list of nine names spanning from iX1 to iX9 in addition to the i1 to i9 names patented back in 2010. We believe the company has plans to fully electrify its fleet, but with minimal disruption to profitability by ensuring complete flexibility across powertrains powertrains (as we discuss in the following sections of the report). We believe the benefits of being about to launch the fifth generation of eDrive will provide BMW with superior electric technological know-how than compet competitors. itors.

Mini BEV coming in 2019 An all-electric version of BMW’s Mini is due in 2019 and will draw heavily on the electric drivetrain of the i3. Indeed the original Mini E concept from 2009 went on to spawn the i3 project. The new BEV Mini will be built in Oxford, although the batteries and electric motors motors required for the new Mini will be built in BMW’s Dingolfing and Landshut plants in Bavaria. Interestingly, BMW management have said that the existing Mini factory at Cowley will require minimal alterations before production of the new EV could begin – we we think largely because of the modularity modularity of the electric concept across the the BMW group. It is possible the new BMW will use an aluminium and carbon-fibre mix to keep down weight and increase the car’s range and efficiency, with the concept vehicle capable of 190 miles. BMW has also signed a letter of intent with Great Wall to develop and produce the electric Mini in China. We understand the drivetrain and the battery technology will be sourced locally, but the launch date and exact production location are yet to be announced.

iX3 due in early 2020 In early 2020, BMW will offer the iX3 to rival Mercedes’ EQC. Little is known as yet of the iX3’s spec but it is expected to have a range of >300 miles. It’s the first of BMW’s electric cars that uses its normal range of cars as a base, unlike the i3 and i8 and will come in two forms: a plug-in hybrid version and the fully electric model as part of BMW's fifth-generation EV platform (which will be used in 11 further electric BMWs due to arrive by 2025). The iX3 will have twin electric motors (one at the front and one on the rear axle) to maximise torque 17 April 2018

23

Barclays | Global Autos & Auto Parts and traction. We expect this product to be based on the fifth generation eDrive and to be a very credible BEV product, appealing to SUV-orientated consumers.

VW has aggressive electric plans post-Dieselgate The VW Group’s push into electrification may have been spurred into action by Dieselgate , but we believe that this was one of the few positive outcomes of the affair and could now position VW at the forefront of electrification. We explain in the following section how VW are taking a different approach to electrical manufacturing than BMW (with a bottom-up approach to building a stand-alone electric platform called the MEB). We can see why such a strategy makes sense for a mass market, global automaker that can build enough scale on its electric platform. And VW clearly has one of the most aggressive approaches to electric penetration penetrat ion with a stated aim of having 20-25% 20 -25% of its sales comprised of BEVs by 2025 and offering 80 electrified products products by the same year, of which 50 will be BEVs. The group is also investing €70bn by 2030 to bring electrification across its entire range of 300 models.

VW’s ID Range After VW’s e-Golf and e-UP! City car, VW will launch its I.D. range of BEVs beginning with the A-segment sized ID Neo in 2019, the first VW-badged MEB model. The ID Neo will be followed by the ID Crozz, a small CUV that also will be sold in the U.S. beginning in 2020. Next is a battery-electric sedan (likely around 2021), followed by a CUV one size bigger than the Crozz. The I.D. Buzz, a modern interpretation of the VW Bus/Campervan, is due in 2022, followed by additional derivatives and yet another MPV/CUV prior to 2025. Commercial vehicles based on the MEB platform may also be on the cards. All the cars will have >250 mile range. The final car in the group was launched at this year’s Geneva Motorshow, the I.D. Vizzion, and will come by 2022 with >400 miles of range and 111kWh battery, which is greater than Tesla’s current offering.

FIGURE 20 VW ID Crossover concept showcases the popularity of mini SUVs, or CUVs (crossover utility vehicles)

Source: VW website

The ID Neo, is claimed to boast a range of 249-373 miles, easily eclipsing the 186-mile range of the face-lifted version of the e-Golf . And it will have its electric motor at the rear, freeing up space within the front section of the ID, which means that VW can exploit the packaging advantages inherent in pure electric drivetrains to provide a roomy four-seat interior offering accommodation similar to today’s larger Passa t. VW describes the battery used by the ID as being scalable and hints at differing capacities in each of its upcoming electric models in much the same way that it offers differing po wer outputs in today’s 17 April 2018

24

Barclays | Global Autos & Auto Parts combustion engine models – ie, ie, a fully flexible offering due to its modular manufacturing concept (more on this later). Little is being said about the charging system for the ID, although VW management claim that its battery can be recharged to 80% within 30 minutes, so it is likely to be enabled for an 800V system similar to that employed by Porsche on the Mission E.

And VW believes believes battery costs will become become competitive competitive Earlier in March, VW held its annual analyst day and revealed that via its contract with battery suppliers, it expects to get to battery cell costs below €100 per kWh by 2020 (or ~$120/kWh), and to continue to reduce costs beyond that. This is similar to GM’s guidance, which based on its contracts with supplier LG assumes $100/kWh on a cell basis by 2022. Assuming pack costs add another ~40%, it could imply $140/kWh for GM and VW in 2022 – well – well within the range of Tesla’s goal of $100/kWh. Or put another way, on a 50kWh battery system, the legacy OEMs would be at a $2,000/vehicle cost disadvantage. But with ~$15,000 of materials and assembly costs in the rest of the car, a 10% scale advantage for the legacy players could close all but $500 of the cost gap.

Chevy Bolt – Bolt – another another step of GM’s aggressive push into electrification The Chevy Bolt marks GM’s increasingly aggressive push into the arena of electrification.

We acknowledge that at a starting price of ~$36k (pre EV tax credits), the Bolt is a fairly less compelling offering than the Tesla Model 3, which offers much more of a luxury appeal, and which will comp more to a vehicle like the BMW 3-Series. Rather, the Bolt, while offering a nice option package, will still comp c omp against other mass-market offerings. offerings. That said, we believe the introduction of the Bolt marked a very crucial step for GM’s push into electrification. In particular, GM will be able to leverage the Bolt as it expands in electrification. GM is expanding its current EV platform with two new crossover entries in 2020. And perhaps more importantly, the learnings of the Bolt can be leveraged as GM launches an all-new modular EV platform in 2021, which will encompass multiple brands and segments. The platform will support vehicles in the US and China, and will have a structurally integrated new battery system – benefiting from the ability to manufacture batteries at scale, while modular cell assemblies enable flexibility.

Moreover, while the Bolt currently sells at a loss (we’ve heard estimates up to $9k loss/vehicle), loss/vehicle ), the Bolt is a crucial step toward making GM’s EVs profitable. GM is t argeting a 30% cost reduction per unit, driven by lower battery costs (cell cost to decline to sub$100/kwhr vs. the current level of $145/kwhr), as well scale and manufacturing improvement. And with its EVs to ultimately be priced comparably to current ICE prices, it likely implies that GM expects its EV powertrain cost to be at parity or better by the time the new platform emerges. Finally, beyond serving as GM’s first legitimate purely electric offering, the Bolt is also significant, as it is at the center of GM’s autonomous ridesharing platform, which GM highlights offers simpler integration of technologies, while also is optimal for urban environments.

17 April 2018

25


FIGURE 21 Latest generation of GM’ GM’s autonomous Chevy Bolt, highlights GM’ GM ’s advances in autonomy as well as electrificatio electrification n

Source: media.gm.com, used with permission

Hyundai’s Kona Electric SUV, another Model 3 competitor Hyundai’s all-electric Kona coming later this summer is another SUV-based electric car and will likely be pitched at a similar price point to the Chevy Bolt. It will be front wheel drive. There’s a short -range edition with a 39.2kWh battery that can travel up to 186 miles on a single charge. The longer long-range 64kWh version can manage 292 miles on one charge. The list of upcoming BEVs could go on, but suffice it it to to say there’s lots of competition coming in the battery electric vehicle market and much of the competition comes from global auto manufacturers that know how to make cars efficiently and a t scale.

17 April 2018

26


FIGURE 22 Comparison of upcoming BEV product offerings Nissan

Hyundai

VW

Renault

BMW

GM

Tesla

BAIC

Tesla

Jaguar

BMW

Mercedes

Audi

Porsche

Tesla

Principal model

Leaf 2

IONIQ EV

ID Neo

Zoe

i3 94A

Bolt

Model 3

EC180 EV

Model X 100D

I-PACE

iX3

EQC

e-tron

Mission E

Model S 100D

Body style

Small hatch

Small car

compact

Small hatch

Mid-sized sedan

Small hatch

SUV

SUV

SUV

SUV

SUV

Sports

Sports

Release date

2018

2018

early 2020

Sep-16

Jul-16

Dec-16

Jul-17

Dec-16

2015

Mar-18

early 2020

Late 2019

Aug-18

May-19

Jun-12

0-60 mph

7.4 secs

8.5 secs

>8 secs

8.2 sec

6.4 secs

NA

4.7 secs

4.5 secs

5 secs

>5 secs

4.5 secs

3.5 secs

4.1 secs

Battery partner

Nissan

LG Chem

LG Chem, Samsung, CATL

Cylindrical, Prismatic, or Pouch?

Pouch

Pouch

Range (miles)

150

125

200-300

Full recharge time*

16 hours (60% in 30 mins via fastcharging)

4 hours

30 hours, 80% in 30 mins using fast-charging

Energy storage

compact Small hatch

5.1 secs / 5.6 6.3 secs sec (ex range / std)

LG Chem Samsung

LG Chem

Panasonic / Insourced

Panasonic

LG Chem

LG Chem Samsung CATL + SK (specifically SDI and Samsung Innovation Mission E LG Chem designed)

Panasonic

Cylindrical

Pouch

Prismatic Cylindrical Prismatic Prismatic

Cylindrica l

295

310

60 hours on slow charge. Supercharge can do 80% in 40 mins

10 hours, 45 mins to 80% via fast charging

Pouch

Prismatic

Pouch

Cylindrica l

250

114 (180 rangeextended)

238

220/310 (standard / long-range)

16.530.5h

125

9.5 26 hours. hours, Supercharge 1hr to can do 60% 7 hours 67% via in 30 mins for fast standard charging battery

8.5hr

40 kWh

28kWh

90 kWh

41 kWh

33kWh

60kWh

100 kWh (options also 50/75 kWh 21kWh 90kWh exist for 75 kWh)

1475

1358-1438

NA

1468

1343

1,746

1610/1730

Weight (Kg)

1085

Estimated battery $220/kW Not Not $190/kW $300/kWh $300kWh < $190/kWh cost h disclosed*** disclosed h Vehicle price (exGrant) in €

25,000

c28k

comparable to well-spec'd >23,600* diesel Golf * (c35k)

c36k

c30k

c28k

2,459

>300

>250

tbc

tbc

range of options

range of options

2133

>310

300

9hr45 9hr45, 30 45 hours on hours, 15 mins to slow charge. mins to 70% via Supercharge 80% via fastcan do 80% in "turbocharge 40 mins charging"

95kWh

90 kWh

100 kWh (options also exist for 75 kWh)

2199

2199

2,148

Not Not Not Not Not < $190/kWh disclosed disclosed disclosed disclosed disclosed

c20k

c78k

>70k

c70k

>45k

335

>70k

< $190/kWh

>80k

c80k

Source: Company data, Barclays Research *2.3kW (10A, 230V) domestic ** battery leased *** VW targeted <$100kWH for cell cost by 2020

17 April 2018

27

Barclays | Global Autos & Auto Parts FIGURE 23 There is a proliferation of BEV launches coming from traditional automakers automakers over the next two years OEM

SOP

Production region

Comment

Range (mi)

Brand

Model

Type

GM

Chevrolet

Bolt

Car

Sep-16 North America

On sale in US winter 2017, Europe as Opel Ampera

238

Volkswagen

Volkswagen

Golf Golf

Car

Apr-17 Europe

EV version of Golf, 125m range, US and Europe

125

Hyundai

Hyundai

KONA

SUV

Jun-17 Japan/Korea

On sale in Europe and NA in 2018

210

Tesla

Tesla

Model 3

Car

Jul-17

North America

In slow ramp in NA, Europe sales later in 2018

310

Renault/Nissan

Nissan

Leaf

Car

Dec-17

Europe & NA & Asia

On sale in Europe and NA in January 2018; 151m range

151

Tata

Jaguar

I-PACE

SUV

Feb-18 Europe

On sale in Europe spring and NA summer

298

e.GO

e.GO

Life

Car

Apr-18 Europe

Expected pricing ~$16k, 130km range

78

Chery

Chery

Tiggo2

SUV

Sep-18 Europe

Electric variant of Chinese SUV

-

2017 Launches

2018 Launches

Barclays | Global Autos & Auto Parts FIGURE 23 There is a proliferation of BEV launches coming from traditional automakers automakers over the next two years OEM

SOP

Production region

Comment

Range (mi)

Brand

Model

Type

GM

Chevrolet

Bolt

Car

Sep-16 North America

On sale in US winter 2017, Europe as Opel Ampera

238

Volkswagen

Volkswagen

Golf Golf

Car

Apr-17 Europe

EV version of Golf, 125m range, US and Europe

125

Hyundai

Hyundai

KONA

SUV

Jun-17 Japan/Korea

On sale in Europe and NA in 2018

210

Tesla

Tesla

Model 3

Car

Jul-17

North America

In slow ramp in NA, Europe sales later in 2018

310

Renault/Nissan

Nissan

Leaf

Car

Dec-17

Europe & NA & Asia

On sale in Europe and NA in January 2018; 151m range

151

Tata

Jaguar

I-PACE

SUV

Feb-18 Europe

On sale in Europe spring and NA summer

298

e.GO

e.GO

Life

Car

Apr-18 Europe

Expected pricing ~$16k, 130km range

78

Chery

Chery

Tiggo2

SUV

Sep-18 Europe

Electric variant of Chinese SUV

-

McLaren

McLaren

Ultimate Series

SPORT Sep-18 Europe

Pure electric hypercar likely > €1mn ???

-

Volkswagen

Audi

e-tron quattro

SUV

Aug-18 Europe

Dimensions like Q5, 300m range, on s ale in fall

275

Hyundai

Kia

Soul

SUV

Oct-18 Japan/Korea

Redesigned for 2018 with 111m range

111

Hyundai

Kia

Telluride

SUV

Jan-19 North America

EV version of three-row SUV

PSA

Peugeot

208

Car

Feb-19 Europe

EV variant of 208 Car, 50kWh bat

~250

PSA

Peugeot

DS3 Crossback

SUV

Apr-19 Europe

EV variant of new small SUV

280

Volkswagen

Porsche

Mission E

Car

May-19 Europe

0-60 in 3.5s; capable of long-range at high speed

>310

Daimler

Mercedes-Benz

EQC

SUV

Jun-19 Europe

0-60 in <5s;

~300

General Motors

Buick

B-CUV EV

SUV

Oct-19 North America

Bolt based Buick crossover

~230

Volkswagen

Volkswagen

ID

Car

Nov-19 Europe

Based on Vizzion concept, 413m NEDC range

~350

Honda

Honda

B-Hatch EV

Car

Nov-19 Japan/Korea

Urban EV concept, targets 2019 Europe launch

155

Tesla

Tesla

Model Y

SUV

Nov-19 North America

May be more like 2020 or 2021, no factory yet

~300

Ford

Ford

B-CUV

SUV

Dec-19 North America

Small SUV EV, for US, Europe and Asia by 2020

~300

Volkswagen

Audi

e-tron Sportback Car

Dec-19 Europe

Concept introduced at Shangahi show in 2017

275

BMW

Mini

Mini E

Car

Dec-19 Europe

Urban EV concept

Tata

Jaguar

XJ replacement

Car

May-19 Europe

Expected spring 2019 as an "indulgent, super-luxury" car

300

Geely

Volvo

XC40

SUV

Dec-19 Europe

Based on 40.2 concept

310

BMW

BMW

iX3

SUV

During Europe 2020

All electric version of popular SUV

300

Renault

Renault

TBC

TBC

During Europe 2022

Nothing confirmed for 18/19 but 8 by 2022

2017 Launches

2018 Launches

2019 Launches -

>190

Source: Barclays research, company data, Autocar, Electrek, Autoexpress, carwow Note: ranged stepped down from NEDC to EPA

17 April 2018

28


GERMAN APPROACH: UNDERPROMISE, OVERDELIVER...ELON MUSK MORE FOCUSED ON AUTOMATION In our experience most German management teams are extremely hesitant to promise anything they might possibly not be able to deliver. The cautious approach to targetsetting we tend to see from the management teams at VW, Daimler and BMW when it comes to volumes, profitability and cash flow is likely more than frustrating for bullish investors. Yet on the flip side, this modest attitude when it comes to forecasting and the typically quiet confidence in communication also makes us feel comfortable that the EV targets set by the German OEMs over recent months will in all likelihood be met. We believe these management teams know how they are going to deliver on their volume guidance, they know how many cars they can produce and sell and they know how much they will cost. For example, BMW m anagement announced they’d sell 100k electrified units in 2017, and they did just that (in fact the company ended up selling slightly more at 103k). Their approach stands in stark contrast to Elon Musk’s track record of broken promises. Tesla may have won many devotees on its vehicle design and tech features, but it has become clear that Tesla’s execution on mechanical issues is far from legacy auto industry standards — slow slow production ramps with quality issues, especially in the early days of each model. And the Model 3 has been no exception – in in February, 2017 Tesla forecast that it would achieve an exit rate of 5000/week for the model 3 in 4Q17, a forecast Elon Musk repeated repeated in a tweet in July 2017. In reality, production ended 1Q18 at a rate of ~2000/7-day week, a pace that appeared to continue into April.

17 April 2018

29


FIGURE 24 Model 3 production forecasts

Source: Company reports, @elonmusk on Twitter

Incumbent OE Production is based on >100 years of expertise. Tesla aims to throw out the rule book Tesla CEO Elon Musk bounded into the auto industry and challenged it with a bold vision vi sion of accelerating the pace of production through automation. A clean slate may have aided his vision for vehicle design but we fear it may present an obstacle, rather than an opportunity, to ensuring quality, reliability and efficiency in mass production. Tesla suggests that by throwing out the rule-book, starting starting with a blank page for engineering and by increasing the levels of vertical integration and automation, the company can attain a level of production efficiency that the rest of the industry can only dream of. And yet the Model 3 production ramp is still not going to plan and we keep coming back to our conference call questioning of Mr Musk and his leadership team on the differences between the Toyota Production System and what Tesla is trying to do in the Gigafactory and the M3 line in Fremont. From their answers, answers, and from subsequent subsequent discussions with Tesla IR, it appears appears that they are seeking seeking to get to a nearly ‘lights ‘lights out’ factory. That is, assembly lines with little human labour beyond some workers monitoring computer displays of production production flows. As Mr. Musk said:

17 April 2018

30


Yes. Imagine like if the Model S was -- or the way you design a Model S, design your factory like it's a car. You still have a lot of workers. You still have a lot of people. I mean, just like with Model S, say, we have a large service organization. There's scheduled maintenance. maintenance. There are things that break. There are crashes that need to be repaired. There are technology upgrades. But you don't actually ship people with the Model S. That would be weird. That's not like hanging people in the car. So you have -we expect that the Tesla factory has people -- a lot of people around the factory but very few people in it. (Source: Tesla 4Q17 call) The details around involving humans in the process by building a front line culture of quality cannot be ignored, in our view. But as even Tesla executive Doug Doug Field (ex Apple) noted, even if this were possible for an iPad – which which is still highly manual production at Foxconn – a vehicle is far more complex. Yes, the model at Foxconn was very different where very quick product ramps and very high scale was achieved through manual processing of also what is fundamentally a product whose simplicity is orders of magnitude below ours. And iPad is less complicated than our centre screen in many ways. So it's a very different order of magnitude in terms of the kind of product you're building, and it's extremely manual because that is the way that you have to ramp very quickly and then end the life of our product and bring up a new one. (Source: Tesla 4Q17 call)

The question at this point, in our view, is whether Tesla went for too much automation, and is having difficulty – not just at the Gigafactory, where they acknowledge production equipment was not operating well and the assembly line had to be redesigned – but at Fremont as well. This is what has led to Mr Musk’s description of “production hell”. And what was his solution out of the hell? Bringing in German automation companies offering “plug-and play” systems. Mr Musk further explained: “The most fundamental difference is thinking about the factory really as a product, as a quite vertically integrated product.” He envisions the most automated vehicle plants in the world, where material delivery, manufacture, and assembly will be done without human intervention, and where his production lines will be far faster than the conventional manual assembly lines that he mocks: “Grandma with a walker can exceed the speed of the fastest production line.” We disagree with this view. We think the traditional automotive production process has been evolving over many years to ensure minimal waste, optimum quality and the optimal interfaces between human and machine interaction (which is by no means zero human intervention). interventio n). Automation works well in a stable environment, but in the auto industry there is a turbulent environment of rapid product and technological change – people are much more capable of adapting to these changes. This is something we think the traditional OEMs have learnt over many years, starting with the early paradigm of lean manufacturing. We think it makes sense to take a step back and remind ourselves just how far the traditional OEMs have evolved in i n their methods of auto production: Traditional OEMs cannot risk their brand reputation with poor execution on early models (hence months of beta testing in deserts and tundra) and have therefore been forced to take a less disruptive route to electrification, but one, we believe, that is more assured of success.

17 April 2018

31


Refresher #1: Prior state of the art – Toyota Toyota Production System Toyota was the first in the auto industry to develop the idea of organising its manufacturing and logistics into a defined system known as the Toyota Production System (TPS), focused on “just-in-time” production, which was a precursor of both “lean manufacturing” and modularity. There has been copious academic research 2 focusing on both the Toyota Way and also Lean Manufacturing, but what we wish to highlight is that the automotive industry industry is still to this day held up as the innovators in the field of production systems as a result of the advances started by Toyota. The company even shared its processes with suppliers and donated the system to non-profit organisations to help improve efficiencies. The system remains in practice today across multiple global production locations. Three main pillars run through the TPS, all of which remain relevant today: 1. Just-in-time, ie, making what is needed, only when it is needed and only in the amount that is needed in order to eliminate as much waste as possible 2. Jidoka, ie, automation with a human touch, with a focus on stopping to fix problems immediately in order to ensure the highest quality right from the start. Toyota were adamant that the human touch was necessary to ensure this quality and also focus on future development development of the business via organisational learning 3. Kaizen – continual continual improvement and always striving for innovation and evolution Toyota’s now well-known manufacturing system was first popularised in a book by three MIT academics called The Machine That Changed the World, which became a best-selling in the 1990s and sparked many other manufacturers to emulate the strategy. The principle of the indispensability of people and their importance to the process of “kaizen” were heavily emphasised by Toyota.

These concepts have been copied by many other automakers (as well as non-automotive companies) over the years to ensure minimal production glitches, downtimes, etc. The efficient production concept has also ensured that the car industry has better managed its cost base over the cycle to ensure a tight balance between fixed and variable costs, employing a high level or temporary labour to ensure maximum flexibility but also best quality. Toyota learnt that given the high fixed costs in the industry, building to demand was the best way to ensure as flexible a cost base as possible. However, this requires the very opposite of Mr Musk’s vision, ie, a reduction in automation, to lighten the fixed cost load and ensure “intelligent automation”.

The social part of the equation shouldn’t be forgotten – traditional traditional OEMs have strong relationships with their employees,, suppliers and employees a nd distributors – distributors – a a culture built over many years which is hard to replicate, particularly by a company like Tesla, which we think is so intent on ignoring the sociological aspect.

Refresher #2 - VW’s Modular Production System The next big innovation in manufacturing came with the advent of modularity. The best known proponent of the modular system is VW. VW announced a Modular Toolkit Strategy back in 2007, which from 2012 became known as MQB (modular transverse matrix) and has since been benchmarked by most of the world’s top automakers.

2 For

a more popular read on the topic, see The Machine That Changed the World, 1990 Womac, Jones and Roos and Lean Thinking

17 April 2018

32


IHS said of VW’s MQB strategy: “it could be the single most important automotive initiative of the past 25 years...it really changes the game” Michael Robinet, MD of IHS (a third party automotive consultancy), consultancy), Reuters 10 Feb 2013

The modular design philosophy is to manufacture and sell as many models as possible while using the same parts as much as possible. U nder the MQB approach, VW’s platforms were intended to feature a greater degree of plug-and-play modularity, flexibility and parts commonality, with products tailored for specific markets/segments but at reduced cost. The idea was to protect the different brands’ characters and individuality but significantly enhance profits. The ultimate targets of modular design are so called Lego blocks 3 . A modular design strategy depends on design technology, and due to the high technological hurdle, the first vehicle maker to achieve this goal was leading truck maker Scania4, now part of the wider VW Group: Figure 25: 25: Scania’s high profit margins stemmed partly from modular design technologies Scania: Modular design concept

Source: Barclays Research, based on company data , used with permission

Need to guarantee quality based on combination of individual parts

Modular design has a high technological hurdle because it requires at the design stage for each part that products can be made by assembling all of the various parts. In other words, quality assurance is required not only at the individual parts level, but also at the stage of assembling these parts. parts. We do not think Tesla has been able to replicate this idea, partly due

3

see Satoshi Hino’s practical modular design. reasons for Scania ’s high margin include:1) better product mix due to its focus on heavy dut y, and 2) higher service business ratio. 4 Other

17 April 2018

33

Barclays | Global Autos & Auto Parts to lacking scale and partly due to Mr Musk’s manufacturing vision which relies on hiring engineers (and now engineering systems of automation from Germany) to make the system work, rather than developing talent from within. Also the speed of product design was such that Tesla did not focus on a design that could be built with quality and adaptability in mind. In particular, we think strict safety standards in the automobile industry serve to increase the modular design hurdle. Should a specific part be defective, the number of models subject to recall increases in proportion to the number of models that use that part. As a result, the development and design of individual parts requires greater caution than under non modular systems (and would therefore not suit the “production hell” scenario of Tesla’s current Model 3 production woes). Greater cost benefits if achieved

However, the achievement of modular design (the ability to make different products using the same parts) results in significant cost benefits. Reducing the number of parts facilitates: 1) lower manufacturing fixed costs (dies and machine tools) for parts production, and 2) economies of scale. Tesla may currently have the cost-leadership on batteries but we think the gap is closing here and yet they are unable to close the gap on production cost capabilities with traditional OEMs.

Another advantage of modular design is an inability to imitate

Another advantage of modular design is the inability to imitate, as indicated by other global commercial vehicle makers historically being unable keep pace with Scania despite seeing its high profit margins. Modular design expertise is in the design technology, and competitors are unable to learn this expertise through the reverse engineering of finished products. Not only do we believe this sets the Western OEMs apart from Tesla, but also from their Chinese competitors in terms of manufacturing – it may be easy to copy a vehicle’s styling but less easy to replicate a global modular engineering strategy, even if the companies enjoy a lower overall fixed cost base. We also believe the success of modularity relies on a degree of scale which Tesla so far has b een unable to attain. When VW launched MQB in 2012 (modularer querbaukasten; modular transverse matrix) the company said it would allow it to: 1) reduce capex, development costs, and manufacturing costs; 2) benefit from economies of scale; 3) increase production flexibility; and 4) shorten development times. We believe VW spent up to $60bn on the development of MQB and many critics point to the lack of payback in terms of margin progression. However, we believe the scale of the group required such an intense investment to shift all production facilities onto the new system and that some of the benefits are yet to be achieved, as more vehicles ramp onto MQB. Today there are over 20 different models across the group already sitting on MQB with more still to come (see Figure below).

17 April 2018

34


FIGURE 26 MQB share in overall production is rising steeply

80% 60% 40% 20% 2015

2017

2018

2020

Source: VW Shaping the Transformation Together Presentation

One of the negatives of the system may have been to allow VW to let differentiation of product spiral out of control, but we believe this is something they are now working hard to bring back into line. FIGURE 27: Parts-sharing synergies cross over vehicle sizes and segments based on modular toolkit strategy VW: Technological concept of modular design strategy Platform strategy

Modular strategy

Hat Platform

Modular Toolkit strategy

Module

Module

Synergies limited to same class

Synergies extend across two or more classes

Synergies extend to all classes

C s s B a l C e A l c i h e A0 V

▷

◁

C

▷

◁

B

▷

◁

A

▷

◁

A0

A00

▷

◁

A00

Body design

Body

Body design

Vehicle-specific

Body design

Vehicle-specific

Modules Platform

e l % i c 0 h 0 e 1 V

Source: Barclays Research, based on company data.

17 April 2018

35

Barclays | Global Autos & Auto Parts The Germans are not alone in deploying modular design systems, with Hyundai (NR) in Korea starting to pursue a platform integration strategy from 2009 and Renault- Nissan’s CMF concept operating along similar lines to VW’s modular architecture. Toyota also moved on from TPS to launch Toyota New Global Architecture (TNGA) to enhance product strength and reduce costs. In other words, production processes in the auto industry have been rapidly evolving over the years via constant “kaizen”, “kaizen” , and we think it is naive to assume that a disruptor such as Tesla can announce new ideas for automation of processes and wipe out the entire learning of the incumbent industry over the last 100 years.

17 April 2018

36


“SMART FACTORIES” OR INDUSTRY 4.0 ARE GERMAN CONCEPT OF PERFECT SYNTHESIS OF ROBOT AND HUMAN AND INCORPORATE AI AND IOT While in the previous section we argued that Tesla’s approach to production has been to over-automate (and therefore increase complexity and expense), we are arguing that the German OEs are using technology to their advantage within their factories. In fact, we believe while the Japanese led the way with lean manufacturing and the TPS, German modularity and now “smart factories” show the industry is at the cutting edge of manufacturing progress. progress. It may be less headline grabbing to be leaders in Industrie 4.0 than Tesla’s space-ambitions, but given the increased use of AI (Artificial Intelligence) and IoT (Internet of Things) within traditional OEM factories, we believe the industry is still leading the way in its knowledge of how to make profitable and cash generative vehicles, whether they be internal combustion powered powered or electric.

BMW CEO Krueger believes: “Silicon Valley isn’t the only place where rapid progress is being made in technology” Analyst FY17 results presentation, 21March 2018 The latest developments in both factory automation and robotics, particularly in Germany, continue to reduce the required workforce, relieve workers of unpleasant or unsafe jobs and help to improve reliability and product quality. We believe this enables German OEMs to have the most efficient manufacturing processes but also to ensure the highest quality levels for their products products,, whilst equally protecting profitability.

Key benefits of “smart factories” include: 









Greater flexibility – flexibility – the the ability to respond faster and more flexible to customer demands Greater efficiency – enabling an efficient use of resources and improved inventory control Greater speed – speed – this can enable shorter innovation cycles and faster product innovation and hence improve time-to-market Logistic management – helps with “just-in-time” delivery but also enables greater flexibility in vehicle configuration More attractive working environment environment – – a a better synthesis of man and machine, doesn’t take away the importance of human intervention but can improve the ergonomics and provide more attractive work space for employees, especially given demographic changes and differing working demands (eg, blue collar work holds less appeal in general to the younger generation) g eneration)

Industrie 4.0 originated in Germany We remind investors that t he term “Industrie 4.0” originated in Germany to promote the computerisation of manufacturing. Moreover, we think the German auto industry has been at the forefront of this manufacturing revolution. The auto industry has been developing increasingly efficient manufacturing processes by utilising Industry 4.0 to create “smart factories” focused not only on modularity but also on increased usage of Artificial Intelligence (AI) and the Internet of things (IoT). In combination these systems ensure the optimal real-time communication between cyber-physical systems and humans. On a continuous basis the insights gained from int elligent data analysis in Germany’s auto plants improve quality, enhance the production processes, reduce throughput time and, as a result, cuts costs. German automakers have worked for many years improving the manufacturing process to reach what we believe is an optimal symbiosis between man and machine to ensure the highest quality and flexibility of output. 17 April 2018

37


FIGURE 28 Germany leading the push on automatio automation n and manufacturing efficiency “The full potential of the industrial Internet will be felt when the three primary digital elements – intelligent devices, intelligent systems and intelligent automation – fully fully merge with the physical machines, facilities, fleets and network networks. s. When this occurs, the benefits of enhanced productivity, lower costs and reduced waste will propagate through the entire industrial economy.” Source: Peter Evans and Macro Annunziata, GE, Industrial Internet: Pushing the Boundaries of Minds and Machines (2012)

In particular, we believe the BMW production lines are far more advanced than their West Coast competitor and that the market is underestimating the benefits of BMW’s “smart factories”. Tesla may be emulating the concept, and potentially why the purchase of Grohman was felt necessary, but we believe Mr Musk has misunderstood the concept of “smart factories” and has gone too far in his automated vision. We discuss each OEM ’s strategies on EV production in more detail below.

BMW is not the only German manufacturer focused on Industrie 4.0 "At Mercedes-Benz, Mercedes-Benz, we use the term 'Industrie 4.0' to describe the digitalisation of the entire value chain, from design and development to production, where the term has its origin, and finally to sales and service," says Markus Schäfer, Member of the Divisional Board Mercedes-Benz Mercede s-Benz Cars, Manufacturing and Supply Chain Management Management,, Daimler AG. "For us at Daimler, there is no question that the digital revolution will fundamentally change our industry. This applies to the methods by which we develop, plan and produce our vehicles. It applies to the way we make contact with our customers. And not least, it can be experienced through our products themselves."

"If man, machine and industrial processes are intelligently networked, individual products of high quality can be created more rapidly, and production and manufacturing costs can be made competitive." (Daimler website) Given the increasingly diverse requirements of customers around the globe, it is important for automotive manufacturers to ensure the most efficient manufacturing processes and the maximum diversity for minimal cost outlay. Customers are increasingly demanding shorter innovation cycles and the benefits of Industrie 4.0 are enabling the German OEMs to provide this and deal with the increasing demands of electrification and digitalisation consecutively.

Toyota’s TPS has now given way to TNGA Toyota is in the process of rolling out an updated version of its vaunted Toyota Production System (TPS), called the Toyota New Global Architecture (TNGA). In a world where automation and AI are at the fore, it might seem that lean and the human touch are taking a back seat as slightly antiquated. And yet Toyota’s new production processes are more fitting with the below quote: 17 April 2018

38


"One machine can do the work of 50 ordinary men. No machine can do the work of one extraordinary man." Elbert Hubbard, philosopher (1856 – (1856 – 1915) 1915) 5

Do robots kill jobs or create them? The auto industry may be responsible for over half the commercial robot purchases purchases in North America, but Wil James, president of Toyota Motors Manufacturing in Kentucky, recently stated that “our automation ratio today is no higher than it was 15 years ago” with robots responsible for less than 8% of the work on Toyota’s assembly lines. Mr James continues that “Machines are good for repetitive things but they can’t improve their own efficiency or the quality of their work. Only people can.” Toyota are heavily focused on ensuring the human touch is not lost via automation, but that automation works alongside humans to ease the burden of difficult work, ensure greater efficiency but always ensuring that humans have the ultimate connection to the production line to ensure quality. This does not mean that the Japanese system stymies innovation. Toyota is extremely proud of allowing their employees to use their ingenuity to design new automated processes and ensure more ergonomic comfort and more efficient jobs. And the system also fosters a process of employee satisfaction and self-regard. Indeed in 2015 Toyota promoted a 52year veteran of the firm, who started on the factory floor, to head up global manufacturing, just showing showing how much much importance importance is placed placed on on factory factory experience experience.. The TNGA TNGA strategy strategy aims to reduce costs not primarily through labour cuts but via smarter use of materials (lightweighting), improved modularity and trimming platform variants. Robots have a place but not as replacements for people but rather as basic tools to enhance factory performance. This approach may be at odds with the perceived wisdom as evidenced in many academic studies that automation trumps all. Tesla’s CEO Elon Musk clearly follows this aspiration with his ‘lights-out’ concept at his Model 3 plant, and Kia (NR) in Korea is also claiming productivity improvements of nearly 200% as a result of automation (some Kia plants have >1,000 robots but <1,000 humans on the assembly line). We have seen analysis suggesting over the next two decades some 47% of American jobs will be lost to automation, and in China and India the numbers could be as high as 77% and 69% respectively. However, Toyota (and many other incumbent automakers) believes in the importance of focusing on the human ability to innovate and improve while even the most advanced AI systems still lack full cognition skills. When the next level of AI is available, the robot era will likely arrive but we’re not there yet.

Robots at the Gate – Gate – Humans Humans and technology at work Our Macro team last week published a series of reports focused on the interaction of technology and people (for an overview of the multiple reports see Equity Gilt Study 2018, 10 April 2018), in which they argued similarly to Toyota and the incumbent OEMs that it takes time for a technology to mature before full automation can be achieved and that “soft automation, where only parts of a job are automated, is more dominant than hard automation, where technology fully substitutes labour”. As we will discuss in the “Tesla Production System’ section, we worry that Tesla is pushing too fast for hard automation either before the technology has fully matured or at a detriment to capital intensity and complexity.

5

17 April 2018

https://www.economist.com/blogs/sch https://www.econom ist.com/blogs/schumpeter/2012/10/z-business-qu umpeter/2012/10/z-business-quotations-2 otations-2

39


Soft not hard automation encouraged by incumbent OEMs In our recent management meetings with VW, the company commented that the advent of technological advances is working in tandem with a decline in popularity of blue-collar jobs with the younger workforc e. While VW are as keen as Toyota to ensure the ‘human element’ remains prevalent in their factories, the company is also aware that electrification brings also simplification (a BEV has c40% fewer components versus ICE) and therefore a shift in employment away from traditional ICE manufacturing roles towards software development and other technology-related jobs is no bad thing. Hence while VW is aiming for 24k job losses over the coming years, some of these will be replaced by 9k additional jobs in IT and software. software.

17 April 2018

40


GERMAN OUTPUT QUALITY Not only do we believe that the German OEMs know how to manufacture EVs reliably, we also believe that their production systems give them a clear competitive edge when it comes to the cost efficiency and quality-focus of the product. product. We think the past 100 years of manufacturing know-how have led to continuous advances in technology and process. We wrote at length in our feedback pieces from Israel about traditional OEMs leadership in technology and how investors are missing the Auto Tech advances sitting within GM and the European premium OEMs. See: Global Autos: How Elon spurred the Israeli Auto Tech scene, and gave legacy auto an edge, 12 March 2018, and Global Autos: Who's leading the Auto Tech Arms Arms Race? Findings from Israel Israel , 12 March 2018. And we also think the market is missing the ability of the German premium OEMs to make the highest quality electric cars. In their quieter PR approach to both electric and autonomous, the traditional industry has been viewed as being the “dinosaur” to Tesla’s “disruptor” credentials, credentials, but we believe the “slower but surer” ap proach will lead to the most lasting products with the greatest consumer acceptance. Have you ever sat in a German premium car and touched the leather interior and compared it to the quality of the Model S? (although Tesla fans would point out the software experience is a more important brand attribute than plush interiors – which which we agree with for the lower volume S/X but may be an issue as Tesla moves into the mid-luxury market)

German premiums have consistently performed well in JD Powers quality reviews It is extremely hard to show concrete evidence of the higher quality of products. The JD Powers dependability surveys have for years been used to show benchmark quality in the industry (and it is clear from a look at the historical surveys that the 3 German premium manufacturers have consistently recorded above average results, even in weaker product years). But Tesla sells too few cars currently to be included in the survey for comparison purposes: FIGURE 29 German and Japanese brands score highly in the JD Powers’ Powers ’ US dependability surveys over the past few years (consistently above the industry average 2015

Name plate

2016

Problems per 100 vehicle

Lexus

89

Buick

Name plate

2017


Name plate

2018


Name plate


Lexus

95

Lexus

110

Lexus

110

Porsche

97

Porsche

110

Porsche

100

Toyota

111

Buick

106

Toyota

123

Buick

116

Cadillac

114

Toyota

113

Buick

126

Infiniti

120

Honda

116

GMC

120

Mercedes

131

Kia

122

Porsche

116

Chevrolet

125

Hyundai

133

Chevrolet

124

Lincoln

118

Honda

126

BMW

139

Hyundai

124

Mercedes

119

Acura

129

Chevrolet

142

BMW

127

Scion

121

Ram

129

Honda

143

Toyota

127

Chevrolet

123

Lincoln

132

Jaguar

144

Lincoln

133

BMW

146

Mercedes

135

Mercedes

147

BMW

142

Industry Average

152

Industry Average

142

Industry Average

147

Industry Average

156

99

Source: JD Power

17 April 2018

41


Tesla has had more difficulty with initial quality in early production models From blog reports it is clear that Tesla customers have had many grumbles in the early days since the launch of each new product. For example, the S and particular the X had numerous early quality issues (such as the issue with the falcon doors), which did not dent the love for the vehicle amongst the early adopters 6. And legions of bloggers have noted numerous issues with the Model 3, including panel gaps and touch screen malfunctions.7 German OEMs do not have the luxury of being able to support such a multitude of complaints, which are more accepted from a “start-up”. BMW’s i3 may have had its critics as regards the styling of the first generatio generation n vehicle, but the money invested into the electric and light-weighting technology clearly paid off in terms of the quality of the engineering.

Consumer Reports feedback has improved for Tesla – Tesla – Model Model S now top ranked From 2014, owners reported many powertrain issues with Tesla’s Model S sedan but in 2015 US magazine Car and Driver named the Model S the “car of the century”. However, However, in October 2015, two months after naming the Tesla 'the best car ever tested,' Consumer Reports (a US independent consumer product review magazine) declined to give the Tesla Model S a "recommended" designation, citing too many complaints from owners. Complaints ranged from minor issues, such as misaligned doors and squeaky body, to severe ones – e.g. e.g. total drive train replacement and inoperable door handles. In their 2016 Annual Auto Reliability Survey, Consumer Reports improved the Model S rating to average reliability, while reporting that the Model X has had significant malfunction issues. The magazine also raised "serious concerns about how some automakers, including Tesla, have designed, deployed, and marketed semi-autonomous technol ogy”. By 2017, Consumer Reports finally marked the car as “above average” for the first time – 3 3 years after launch. And for 2018, the Model S i s Consumer Reports’ highest highest rated electric car with a score of 94 (the Chevy Bolt comes in next at 77 and the BMW i3 scores 66). But the Model X is much more poorly scored at just 53.

But Model X is still struggling to get a “recommended” rating (Model rating (Model 3 yet to be tested) When Tesla’s Model X was first launched, Consumer Reports wrote that the all-wheel drive Model X 90D “largely disappoints, as rear doors are prone to pausing and stopping, the second-row seats that cannot be folded, and the cargo capacity is too limited. Even its panoramic, helicopter-like windshield was disapproved of as it is cranky-sounding and it is not tinted enough to offset the brightness of a sunny day.” The report continued that overall "the ride is too firm and choppy for an $110,000 car ”. That said, Tesla addressed these issues with several software updates, and no known issues remained after the 8.0 firmware was released. But in June 2016, Tesla had to settle a lawsuit over usability concerns and accept the Model X was rushed to production before being ready. It may therefore be no surprise that the model model still fails fail s to score highly in the 2018 review.

Traditional OEMs cannot risk a poor quality review We believe the money invested into capex and R&D by the traditional OEMs to ensure the smooth production of their vehicles is not only to ensure the best possible cost base but also to ensure the highest quality products. They have too high a reputation at stake to risk with a “rushed out” EV that cannot maintain their brand reputation. It therefore comes as no surprise that C onsumer Reports says of BMW “t hat single mission focus contributes to making the i3 the most energy-efficient car we've ever tested....combining the superefficient powertrain with lightweight construction creates an extreme extremely ly energy-efficient car that gets the equivalent of 139 mpg when running on electricity -- the highest we've ever seen in our testing.” We suspect, when tested, the Model 3 may fail to achieve high accolades in the early years.

6 See

http://www.jdpower.com/press-releases/tesla-beyond-hype https://www.greencarreports.com/news/1115659_tesla-model-3-quality-is-terrible-but-does-it-matter-tobuyers 7 See

17 April 2018

42


GERMAN BEV PRODUCTION STRATEGIES Key BEV strategies: 









VW standalone/native EV platforms (MEB and PPE), although first Audi and Porsche launches will be built on adapted ICE platforms (MLB) BMW fully modular manufacturing across all powertrains and body styles, building on ibrand know-how and launching 5th gen electric drivetrain (likely more outsourced outsource d than original i3) Mercedes- strategy unclear. Marketed as standalone/native platform for EQ but more likely adaptation of existing modular platform. Likely behind on spending with EQ in its nascent generation generation Tesla: standalone BEV platform but struggling to ramp to scale and also failing to offer consumer flexibility of non-BEV powertr powertrains ains GM: flexible, modular platform similar to BMW but leveraging scale to bring down purchasing costs, particularly of battery cells, and utilising non-vertical integration to buy-in commoditised components

The VW Production System MQB benefits continuing just as MEB system is launching As discussed in the earlier section, VW was the first proponent of a modular system of production. Over the years since its inception in 2012, we have written at length on VW’s modular system called MQB (starting with an in-depth note comparing MQB to Japanese production systems: Toyota, Nissan step forward: Target global mass market with lower costs but better product appeal , 27 August 2012) and many academic and press articles have highlighted how revolutionary the modular system was at conception. MQB was developed for transverse-engine vehicles, alongside Audi ’s MLB for longitudinalengine vehicles and NSF (new small family) for small vehicles. It involved a great expense and due to the scale of the VW Group and its multiple brands and production locations, a complex procedure procedure to adjust all production to the new modular concept. And just when the benefits of MQB might fully be being enjoyed within the VW group, the decision was made, post Dieselgate, to develop a new modular system, the Modular Electrification Toolkit (MEB), a modular system for manufacturing electric vehicles. MEB has been undergoing development since 2015, to further optimize the strengths of the MQB, while making it suitable for electro-mobility. The so-called Early Phase of development will come to a conclusion in the spring of 2017, and series production of the first e-vehicle under the I.D. banner is due to start in 2019.

MEB the next step beyond MQB but is more evolution than revolution Some in the market are disappointed that the Lego-blocks basis of MQB was not sufficient to provide the flexibility needed for electrification, but VW is adamant that an entirely new system is necessary to optimise electric vehicle design. The e-vehicles currently available from Volkswagen are being made using the MQB, which shows it is possible to build EVs on an ICE platform, but newer EV products from the ID range and beyond will launch on the MEB. VW management’s belief is that they need an entir ely standalone production system for BEVs to ensure optimum positioning of the battery and optimum flexibility and cost base. This differs from BMW’s fully modular concept of production across all powertrains, 17 April 2018

43

Barclays | Global Autos & Auto Parts but we believe is possible due to the scale of cars produced by VW (with a target of 2-3mn or 20-25% Group sales of BEVs alone by 2025). VW has bullish market prediction for BEV BEV sales by 2025 – 2025 – 10% 10% of the global market and 35% of global premium sales

VW predicts industry industry BEV sales to account for about 10% of the global market in 2025, with VW outpacing that share slightly. BEVs will control a much larger 35% of the global premium segment by then, the automaker says. Pricing will be among the keys to market growth, and VW is promising to position its MEB vehicles competitively based on each model’s interior volume. Because the MEB’s purpose -built design doesn’t have to accommodate an engine and transmission, it allows for a flat floor and a roomier-thannormal cabin. The upcoming I.D. Crozz CUV, for example, has a Tiguan footprint but the interior package of the bigger Atlas model. VW says it will be priced similarly to an Atlas. As we discussed in the ‘upcoming BEV products’ section earlier, VW has ensured a cost effective supply of battery cells, which helps with their plans to sell BEVs at rates competitive with comparable diesel engines. Management have stated that they are confident of securing battery cells at below €100 per kWh by 2020 (or ~$120/kWh), and to continue to reduce costs beyond that. Similarly to GM, we believe this is due to the Group’s scale and its ambitious BEV targets, but when combined with the cost advantages of MEB it should enable VW to have significantly greater flexibility and over time a reduced cost base than its Silicon Valley competitor, Tesla. Combined with a greater breadth of product offering across all powertrains and bodystyles and brands, this should be a significant competitive competiti ve advantage by the early 2020s.

VW is pursuing a dedicated/standalone approach to BEV production For VW, therefore, dedicated production lines make the most sense. Matthias Erb, executive vice president and chief engineering officer for the North American region, says of the MEB platform that the benefits to product design from having a dedicated electric-vehicle platform are worth the gamble of having standalone plants and outweigh the cost of any manufacturing inefficiency from mixing production with ICE vehicle. VW is starting by converting its Zwickau plant in Germany, currently home to Golf production, to 100% MEB output. The company also plans production in China – VW VW wants to build 1.5 million newenergy vehicles, mostly BEVs, annually in China by 2025. US MEB production is planned too, although no firm commitment has been made. “We (are) more convinced than ever that volume-wise you can build real (full-scale) factories for electric vehicles,” says VW’s Erb

VW isn’t specifying volume plans for the MEB dedicated -BEV platform, but says it expects demand in the U.S. for all-electric vehicles to reach 10% of the market by 2025 and that the Volkswagen brand should exceed that level. There will be no mix of internal-combustionengine cars and BEVs in production on a single assembly (as opposed to the strategies of GM and BMW which we discuss below) , though it’s possible for the vehicles to share a paint shop and portions of a body shop.

Audi/Porsche to share share platforms While Audi will utilise MEB for compact cars over time, unlike the VW brand's ID products, the first Audi and Porsche BEVs will come on adaptations of the existing ICE-based MLB platform - C-BEV for Audi and J1 for Porsche. But by 2021 a jointly developed standalone platform called Premium Platform Electric (PPE) will be launched. Close to 1000 employees are currently working on the project and given the performance credentials of the upcoming Porsche Mission E, we expect it to be equally capable of 800-volt charging and likely with an even higher range. Porsche CEO Blume said in a recent statement, "if we had to tackle the challenge ahead on our own, the costs would be c30% higher". And we believe the benefits of the Group’s scale really give VW an advantage versus the competition on EVs, as long as the company learns from its mistakes with MQB and ensures there is more discipline in model derivatives.

17 April 2018

44


The BMW Production System Flexibility via modularity is key; standalone EV platforms may come later BMW have also spent many years perfecting and refining the production processes and we do not believe the CEO Krueg er is wrong to claim that the company now has “the most flexible production network in the world”. However, we do think that the market may have missed the importance of this and views BMW’s approach to electrification (ie modular ity that can be flexed to any powertrain penetration rates) as less exciting than peers with a more dedicated (ie separate platform) approach. VW’s standalone MEB strategy makes sense given its global scale, but we believe a more niche, premium OEM needs to ensure greater flexibility to maintain profitability.

BMW started investing in electrification early... BMW has the money to invest in electrification. This year alone €200mn will be invested in one factory in Leipzig which will be fully modular BMW started early with EfficientDynamics and then moved on to the ibrand and now iNEXT

That is not to say that BMW has scrimped on its spending on electrification. In fact the company was one of the earliest to the electrification race, firstly with its focus on Efficient Dynamics and then via the launch of the i-brand . BMW invested $100mn into its plant (joint with SGL Carbon) in Moses Lake, Washington for carbon-fibre reinforced plastic bodywork and its Leipzig plant was inaugurated in 2010 with an investment of €400mn. And as we will argue in the following section, BMW still have plenty more money to invest. €200mn will be invested this year in its new plant at Leipzig (which will finish by 2020) and this plant will be fully modular with electric capabilities.

...and focused on the entire lifecycle of the EV via sustainable production BMW has been focused on modularity since at least 2014 but has also been keen to highlight its “well to wheel” sustainability focus, ie, looking at the entire production process. The carbon footprint of the original i3 was around a third smaller than that of the BMW 118d (which was already voted World Green Car of the Year 2008) and around 50% smaller if the car is running on power generated from sustainable sources. The use of CFRP on the scale used in BMW’s i models is the highest in the automotive industry worldwide. It’s still unclear whether the lightweight material will ever be cost-effective but it has clearly helped to enhance the group’s environmental credentials. The production process for the i-brand models consumes around 50% less energy and 70% less water vs current average production at the BMW group, which is already best-in-class. All the electricity used to produce i-models at the Leipzig plant is wind-generated and 100% sustainable. Likewise, all of the energy used in carbon fibre production at Moses Lake is entirely derived from renewable, locally generated hydroelectric power and is completely carbon-free.

And BMW also pioneered pioneered engine modularity: BMW’s modular approach to product ion involves the engine as well as the vehicle design. In 2014 BMW started migrating toward a modular engine family for all gasoline and diesel engines or the new “B” family engines to replace the outgoing “N” family. This approach enabled >60% of all gasoline engines to share components, and 40% sharing between gasoline and diesel.

17 April 2018

45


FIGURE 30 High commonality between between gasoline and diesel in the new EfficientDynamics family Cylinders (inline)

-3

Gasoline

-4

-5

>60%

% 0 4 0 3

Diesel

Common parts shared within gasoline or diesel engines is c.60%

Between the gasoline and diesel engines there is 40% sharing

>60%

Source: BMW, used with permission

Engine modularity has given BMW greater flexibility than peers This has enabled major flexibility for BMW giving all engines a common interface to the car, thus enabling hybrid and now battery electric vehicles to join the modular process too. We believe the market is underestimating the degree of this flexibility – which which has become ever more necessary given the uncertainties of penetration rates, as evidenced by the sharp decline of diesel across Western Europe. The efficiency of BMW’s engine plant in Stey r in Austria (which has seen over 20mn engines leave the three assembly lines since its inception in 1979, helped by running 3 shifts daily and being able to run 7 days/week when necessary), with an engine rolling off the line every 14 seconds and reaching its vehicle manufacturing plant within 8 hours, we believe is something that Tesla, or other start-up companies might find hard to replicate. Although Tesla need only focus on electric drivetrains, we believe BMW’s ability to have full flexibility across all powertrains should enable the group to offer the best products to customers, whether they require petrol, diesel, hybrid or BEV. Over the last 40 years BMW has invested €6.4bn in the plant at Steyr, including in emissions testing.

The ibrand wasn’t just just about electrification but also about sustainable manufacturing We think the market is forgetting quite how game- changing the production of BMW’s i brand vehicles was when compared to traditional methods of manufacturing. The company has been combining innovative materials, smart data analytics and renewable energy sourcing for more than ten years now. BMW uses automated bonding (no need for screws, rivets and welding), combined with the innovative use of Carbon Fibre Reinforced Plastics (CFRP). The production lines are significantly shorter (110m versus traditional lines of several kilometres) and smart robots proliferate. But this isn’t the fully automated production of Elon Musk’s vision, but a smart working environment where robots coexist with human engineers to increase flexibility, shorten production times but not at the risk of quality. For further details on the production differences of the original i3 versus Tesla’s i – a a different concept vs Tesla, 22 May 2014. model S, see BMW i –

What are BMW doing today on electric? We believe BMW’s culture of quiet confidence means that they tend to hold back from any major announcements around technology technology until they have 100% confidence in delivery. We remain convinced that the company’s strategy to keep production facilities flexible for any powertrain (be it ICE, PHEV or BEV) is the least risky strategy given future penetration rates 17 April 2018

46

Barclays | Global Autos & Auto Parts are largely outside the control of the industry. And yet, we understand why the market views this strategy as ‘boring’ when GM (OW) across the pond is promising to take on Tesla with an accelerated autonomous timeline and VW is promising aggressive BEV penetration and a wide range of products produced on a standalone platform.

BMW are not behind on electrification The market seems to think that BMW took an early lead on electric with the i-brand but but has since seen competitors competitors leapfrog on EV strategy. strategy. We don't think this is a fair perception. The technology workshops in Munich in December showcased how advanced BMW's modular strategy truly is. It is not a straightforward toolkit but a complex system to ensure that all production facilities can produce any drivetrain from combustion engine (ICE) to BEV. And we have been impressed with the advances in electric drivetrain engineering since the first generation i3. The 5th gen electric vehicles starting with the iX3 BEV at end 2019 and Mini E in the same year and continuing with the i-NEXT in 2021 should be a significant step on from the earlier generations and will utilize the latest battery technologies and drivetrain engineering. BMW has the capabilities inhouse to understand the entire battery supply chain and development process and are investing heavily in the new Battery Competence Centre in Munich

Currently BMW are producing all electric motors in-house and also the battery packaging, as management view this as an area where they have a technological edge. The benefit of the Gen 5 electric drivetrain is that it will be fully scalable to different power and range requirements, from entry level to performance level, by the addition or subtraction of cell modules. BMW also believe they have a cost advantage with their electric motor, with no loss of functionality, by using no rare earth metals, but are still extremely competitive on efficiency and energy density. BMW also has the capabilities in-house to understand the entire battery supply chain and development process, which they expect to advance further via the investment in the battery competence centre in Munich due to open in 2019 – management have said they will invest €200mn at the site over the next four years. However, BMW do not intend to manufacture batteries in-house but do currently see advantages of keeping the pack engineering capabilities in-house. The engineering team sees many improvements in battery technology coming in the next few years, although solid state batteries remain only in the development stages. Production facilities are fully scalable globally to ensure BMW will have pack production capabilities worldwide as penetration rates rise. But BMW aren't just focused on battery production and raw material procurement; they are also aiming to look along the whole lifecycle of the battery and factor in 2nd-life usage and recycling to their business plan.

BMW is about to move to its 5 th gen electric engine when other OEMs are just starting development of the 1 st gen We believe that BMW’s progression to the 5 th generation of electric platform, and the billions invested into modular engineering across the business’s global product range, will ensure BEV production production which in our view is not only of a higher quality than Tesla’s but also with a much more profitable cost base. So all in in all, we do not believe that BMW is at all behind on EVs, despite the market’s current perception. What we believe is that they will have highly credible EV options across the spectrum of power and performance options and will be ready to deliver as penetration rates ramp, and we clearly think this is in contrast to Mercedes, which appears not only to be attempting the less cost-efficient standalone platform strategy for EVs, but also to be further behind in terms of generation development of electric drivetrains (as we will discuss below, Daimler’s current drivetrains drivetrains were bought from Tesla).

BMW a pioneer of Industrie 4.0 and Smart Factories In addition to its advances in modularity, BMW is also an innovator in terms of utilising “Smart Factories”. In many academic studies on Industrie 4.0, BMW’s plants across Germany and Austria are used as case studies. It may not garner sensational headlines, but 17 April 2018

47

Barclays | Global Autos & Auto Parts BMW’s knowledge of data analytics to ensure the smoothest functioning of all parts and machines, cuts the error rate and minimises time. Indeed the production time of a vehicle in the body shop has been halved (from 40 hours to 20 hours in a smart factory). Additionally, the production requires 50% less power and 70% less water versus the old BMW production average. average. And of course the i-brand factory at Leipzig is all hydro-pow hydro-powered. ered. BMW knows how to revolutionise manufacturing manufacturing processes, in our view.

BMW on Industrie 4.0 "We leverage the opportunities of digitalisation in many areas. Complex processes processes can be made even more efficient by using IT-supported technologies in production" Oliver Zipse, BMW Board Member for Production As CEO Krueger stated last month: “The BMW iNEXT is far more than just a car. By that, we mean it is a future-proof, scalable modular system. It will enable the entire company and all our brands in terms of technology, design and new approaches. The iNEXT combines Autonomous Driving, Connectivity, Electrification and Services.” FIGURE 31 BMW have been highlighting their “smart factories” factories” for many years, including this chart from a 2015 production presentation presentatio n which emphasizes not only the flexibility and quality but also the importance of human-robot collaboration collaboration FLEXIBILITY MANAGEMENT

BODY SHOP

• Upto 9 derivatives per line

PAIN T SHOP 1

SUSTAINABILITY

• Continuous innovation to meet sustainable produc tion tion claim (e.g. IPP2) • Use of high-tech, sustainable finish

INNOVATION

CAPITAL PRODUCTIVITY

AS A SSEMB LY

LOG ISTICS

• Smart automation of automation of line production • Production and quality increase th increase th rough human-robotcollaboration

• Just-in-time Just-in-time logistics logistics for efficiency and productivity

LEAN PRODUCTION

• Continuous improvement

PRESS SHOP

• “Buy” decision to ensure best capital efficiency 1. Including M derivatives, 2. Integrated paint process Source: BMW, Barclays Research

The Daimler Production System Mercedes marketing its EV production as ‘standalone’ In contrast to BMW, and perhaps in reflection of Daimler’s relatively limited production expertise when it comes to manufacturing Mercedes electric vehicles (the electric B class powertrain was outsourced to Tesla), the group is launching its EQ range (SUVs, saloons, coupes and cabriolets) on an entirely separate platform. Daimler argues that today’s battery technology requires purpose built electric vehicles as the physical size of the battery takes up a much larger space than a traditional ICE. Only a unique vehicle design is able accommodate this. Management concedes, however, that longer term improvements in energy density may make it possible to produce both vehicle types on the same platforms as eventually batteries will require no more space than a typical combustion engine.

17 April 2018

48


...but in reality it may be b e more similar to BMW’s modular approach However, we note that the incremental electric vehicle production capacity Daimler has announced is to be added to existing plants rather than at Greenfield sites. Given the advantages in terms of cost, flexibility and greater potential for additional model variants / derivatives we would expect Daimler to ultimately end up with the same shared electric and ICE platform set up as is the case at BMW.

We fear Mercedes have greater electric production costs to swallow We understand that the electric vehicles planned near term will mainly run on a separate production line. Only during the final part of the production process is it possible for different models to share the same assembly line as ICE cars. We suspect that given the lack of detail regarding the full EQ brand production processes, as well as the rushed announcement of Mercedes’ BEV plans at the Paris Motorshow in 2016, the brand may be more in its electric manufacturing infancy than its Munich-peer. Mercedes are clearly not approaching the 5th generation of electric drivetrain and we suspect have more to invest in plant capex too. In the longer term, Daimler’s innovative cubeTEC production production concept will allow electric and ICE vehicles to share production lines from the body shop onwards, and we believe this too explains the location location of the new EV capacity. We do not see Mercedes Mercedes as behind in the arena of “Smart Factories” or use of A.I. to improve productivity. We just worry that the company is pursuing a more expensive strategy for converting platforms platforms to electric. All the parts for various models/derivatives including EVs will be provided in the so-called “Supermarket”. Driverless transport systems deliver these to GE O stations. A fully flexible GEO station uses the jigless process: several robots on an adjustable NC axis work together holding and turning the relevant parts into the optimal welding positions. In this way parts of varying size can be combined automatically in the body shop. In other words, DAI fully understands modularity and robotics and may well end up pursuing more of a similar, modular strategy to BMW if they discover native electric platforms require greater scale.

The Tesla Production System – or or “A Robot Too Far” After the spirited discussion on the 4Q call where we probed Tesla management around the differences between Toyota ’s Production System and the approach Tesla is taking to manufacturing, we have further looked at what we think Tesla might be trying to achieve. Let's call it, to give it the benefit of the doubt, the Tesla Production System. Based on Elon Musk’s conference call comments, as well as interactions with Tesla IR, and comments made by Mr Musk during a Fremont plant tour on CBS Morning News TV last week 8 and in a tweet posted last week, we see several key elements to the Tesla Production System: 1.

Simpler design of a vehicle v ehicle to make it easier to manufacture

2.

Battery manufacturing automatio automation n

3.

Automation of the initial stages of the vehicle manufacturing (stamping, welding, paint)

4.

Automation of a portion of final assembly – which attempted to push automation too far and now, by Mr. Musk’s own tweeted admission 9, needs to be dialled back

Simpler design – design – electric electric and few options To its credit, Tesla has attempted to design the Model 3 as a vehicle with a relatively simple and easy manufacturing process. While the idea of design for manufacturability is by no 8 https://www.cbsnews.com/news/elon-musk-tesla-model-3-problems-interview-today-2018-04-13/ 9

See Gibbs, “Elon Musk drafts in humans after robots slow down Tesla Model 3 production,” The Guardian, April 16, 2018

17 April 2018

49

Barclays | Global Autos & Auto Parts means new10 or unique to Tesla (i.e. many of the legacy OEMs, such as Toyota, have led in this way), the Model 3 appears a ppears relatively simpler in three ways: 





As an electric vehicle not only is its powertrain simpler with fewer moving parts than an ICE and transmission, but the body and frame of the car can be optimized to sit on top of the battery/drivetrain ‘sled,’ as opposed to having to wrap around an internal combustion engine, transmission, and emission system. The design of the vehicle interior itself is rather sparse, without buttons, dials, movable vents, etc. that add to manual labour assembly. Moreover, Moreover, the lack of buttons and dials around the car, combined with a multi-domain controller which consolidates vehicle controls (and note, luxury vehicles currently have around ~150 ECUs, or electronic control units), implies a more compact electrical architecture architecture and wiring harness. There are relatively few trim option variations on the Model 3. Powertrain options are for different battery sizes and future front wheel drive – which Tesla appears to have delayed. Software options f or autopilot and ‘full self driving’ of course require no manual labour. The only interior option is the premium package, which adds premium heated seating and cabin materials, power adjustable front seats, steering columns and side mirrors, a premium audio system, tinted glass roof, auto dimming mirrors, and a center console with covered storage and docking for smartphones. smartphones.

We would note while the product simplicity and lack of choice is fine for the dedicated early adopters, it may not be enough variety for the mass market buyer. Tesla would need to expand the option set to hit 500k+ annual sales for the Model 3. After all, Henry Ford’s Model T strategy was ultimately displaced by Alfred Sloan’s GM, who was able to build car varieties ‘for every purse and pocket.’

Gigafactory automation for battery cell and pack We've seen the Fremont battery pack assembly area several times, and toured the Gigafactory in its early days. We are frankly somewhat surprised that there had been manufacturing bottlenecks around battery pack assembly at the Gigafactory (which Tesla called out as the primary Model 3 bottleneck in 4Q17). Indeed, one would assume that this is a core Tesla skill, especially relative to other manufacturers, and has benefited from continuous improvement at the Fremont Model S/X pack assembly over the years (see: Kaizen comes to California – California – 12/5/14). 12/5/14). Note, however, that as opposed to Tesla’s cylindrical battery cell format, most other OEMs have chosen a large pouch battery format, which makes the assembly of the packs themselves less complex than Tesla's efforts to assemble several thousand small cell formats. We assume Tesla’s production of the motor and drivetrain to be automated, although other automakers told us that some hand-assembly is needed on electrical motors to piece together delicate parts.

Automated stamping, stamping, welding and paint – paint – like like the legacy industry In the Fremont plant, the stamping, paint, and welding frame operations for Models S, X and 3 appear very similar to those of legacy OEMs, with a high degree of automation - And the use of robotics robotics is by no means new for legacy OEMs. OEMs. For example, in 1961 GM installed the Unimate #001 prototype robot in a Trenton, NJ diecasting factory, and used the Unimate for welding when it revamped its Lordstown plant in 196911.

10

For example, we found mention of the concept as far back as 1988, see Whitney, “Manufacturing by Design,” Harvard Business Review , July 1988 11 See Unimate/The First Industrial Robot at https://www.robotics.org/joseph-engelberger/unimate.cfm

17 April 2018

50

Barclays | Global Autos & Auto Parts However, in our past visits to Tesla’s Fremont factory, we observed on severa l occasions manual polishing was needed on the stamped aluminium panels, and in the CBS Morning News workers could be observed hand sanding rear panels and using a hand-held power grinder to smooth out a door frame.

Final assembly – assembly – attempting attempting more, but not yet full automation, and likely the source of some bottlenecks Tesla appears to have attempted a path of significantly greater automation than traditional OEMs in its vehicle assembly. While on several conference calls Elon Musk has outlined a vision of what would be a lights-out factory (his so-called “Alien Dreadnought”), with humans just monit monitoring oring the the process process as as opposed opposed to actually actually touch touching ing the prod product, uct, itit appears appears at least least for for the Model 3 this is more of a long-term aspiration than an achievement. We have been assured by Tesla management that Model 3 still uses hand labour for delicate tasks involving soft parts, such as threading the wiring harness into the vehicle, applying the headliner, and other steps. However, Tesla has also indicated that tasks that are performed by humans in the traditional assembly line, including securing the dashboard and applying the door panels to the hinges, are done by robots. In legacy OEMs plants we have toured, these are performed by workers using tools such as torque wrenches with built-in pressure sensors to make sure the nut is neither over- or under-tightened. Indeed, as discussed in a recent piece highlighting how Toyota is accentuating human craftsmanship amid the push toward automation:

Robots can’t do everything Fundamentally, Toyota’s production principles were keyed to the notion that people are indispensable, the eyes, ears, and hands on the assembly line – identifying identifying problems, recommending recomme nding creative fixes, and offering new solutions for enhancing the product or process. See: Rothfeder, Jeff, “At Toyota, The Automation Is Human -Powered,” Fast Company, 9/5/17

So without having seen the Model 3 assembly (except at a distance during our most recent Fremont visit), we make several observations observations:: 

Elon Musk himself has now admitted (via Twitter) that Tesla – under under his direction – has has pushed automotion too far: Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated. -- Elon Musk (@elonmusk) April 13, 2018.





17 April 2018

It’s unclear how long it will take to get the automated parts of the line working consistently at a 5000/week rate – or or if it will even get there without significantly more human labour.

With people involved now in the soft assembly and, given Mr. Musk’s tweet , likely to be more involved in other processes, the importance of involving front-line workers in quality and productivity improvement becomes paramount. Of course, the Toyota Production System, famously stresses the involvement of front-line workers in improving operations operations and quality quali ty (and indeed, a prime example of the respect for frontline workers at Toyota is evident with Mitsuru Kawai, a 52-year veteran of Toyota, at the head of global manufacturing – the the highest position ever held by a former blue-collar worker at Toyota).

51

Barclays | Global Autos & Auto Parts Tesla has posted job listings for rework at the end of its assembly line, which would seem to make sense, as in a more automated line there aren't employees to check the fit and finish as the product goes down the line. Moreover, the Toyota front-line culture has also been adopted by the Detroit automakers, who spent painful decades working with the UAW to adopt the process. For example, in the mid 2000s Ford F-150 was ramping up during a tour of the Dearborn plant, we saw the assembly line stopped on several occasions while a frontline blue collar worker explained to engineers with clipboards what she found wrong in that process. This process breeds a culture that respects both the front-line as well as the low-level manufacturing employees, who might be wondering whether a culture of the CEO sleeping at the assembly line really helps (ie, Elon Musk).

GM progressing on “Smart Manufacturin Manufacturing, g,” leading in manufacturing of autonomous vehicles There are those who will criticize GM as one of the “dinosaurs” – an an automaker which will inevitably be disrupted by the likes of Tesla, Waymo, etc. – a a point with which we disagree, as GM’s initiatives to lead in the disruption of the auto industry are underappreciated. However, if there is one thing GM excels at which the “dinosaur” critics can’t ignore, it’s GM’s leadership and proficiency at vehicle manufacturing. Simply, while Tesla has struggled with the task of taking a vehicle to concept and having flawless production (unsurprising, (unsurprising, as Tesla is still relatively young as an automaker), GM on the other hand excels at taking a vehicle concept and bringing it to fruition via flawless volume production. GM’s Global Manufacturing System has five key elements: 1. People Involvement, 2. Standardization, 3. Built-in Quality, 4. Short lead time, 5. Continuous Improvement. Moreover, GM has developed GMA as a global manufacturing process, which allows it to quickly scale up new platforms in new geographies. As an example, consider GM’s large pickup/SUV platform (K2XX), a platform with ~1.2mn units annually, which likely accounts for well over half of GM’s total company profit. Over the course of ~2 years, GM is expected to undergo a highly orderly changeover of the platform to the next-gen version – similar similar to the way Ford went through a highly orderly and largely flawless changeover of its F-Series several years ago to an aluminium body. Such manufacturing expertise cannot be underappreciated.

But beyond its traditional lean manufacturing expertise, GM is also engaging in next-gen “smart” manufacturing. While areas such as 3D printing and drone technology are being employed, we’d highlight the Auto Big Data component of GM’s manufacturing, which is being used for active monitoring and predictive mainten ance. Indeed, alongside data efforts by GM’s OnStar business, the data initiatives in manufacturing provide another indication of GM’s leadership in the realm of Auto Big Data (see: The coming ocean of Auto Big Data , 4/26/17). And while technology is transforming the manufacturing process, at the same time GM is keeping the humans at the center of the manufacturing. Through the likes of “collaborative robots” and technology which assists the front-line worker, arguably GM has an opportunityy to see enhanced manufacturing efficiency. opportunit Finally, GM deserves credit for its leadership in manufacturing of autonomous vehicles. A bit over a year ago GM became the first automaker to assemble self-driving test vehicles (its autonomous Chevy Bolt) in a mass- production facility. Indeed, GM’s autonomous Chevy Bolt, a vehicle without a steering wheel or pedals, will be assembled at its Orion Township facility, alongside the non-autonomous Chevy Bolts. This is an important development, as 17 April 2018

52

Barclays | Global Autos & Auto Parts we believe that success in the autonomous world will require aviation-grade safety and deep automotive engineering rigor to balance out the quick development of autonomous functionality – a balance which we believe GM offers (see: AVs – from app-grade to aviation-grade aviation-gra de safety , 3/26/18).

17 April 2018

53


GERMAN OEMs ALREADY HAVE A GLOBAL PRODUCTION FOOTPRINT TO BUILD ON In addition to considering the German OEMs ’ substantial expertise in auto manufacturing – their ability to reliably produce high quality cars – we feel it is important to bear in mind quite how global their production and distribution footprint already is. We believe this presents a significant advantage as they begin to roll out the manufacture and sale of electric vehicles across geographies. FIGURE 32 German OEMs’ OEMs’ global unit sales 2017 9% 3%

BMW

VW

Daimler

17%

4%

16%

Greter China

42%

45%

9%

46%

10%

25%

Middle East / Africa Asia (ex-China)

2%

2% 40%

Europe

2%

1%

2%

24%

South America North America

Source: IHS Markit

All three German OEMs have announced electric vehicle capacity not just for Europe but also for the US and China. We note that VW, Daimler and BMW have been making cars in the US since 2011, 2005 and 2003 and in China since 1984, 1997 and 1994 respectively. By contrast, Tesla has struggled to gain a foothold in China and sales in Europe have been more lacklustre than in the US (outside of heavily incentivised countries such as Norway and Netherlands). Just over half of Tesla’s unit sales in recent years ha ve been in the US. While the growing exports of Model 3 to global markets, as well as the potential creation of international manufacturing capabilities, may change that mix, we think investors may be under appreciating risks in expanding internationally. German OEMs are also at an advantage advantage when it comes to distribution. distribution. VW sells its vehicles in 153 countries. Daimler runs 8500 sales sales centres globally. globally. And BMW is represented represented in more than 150 countries with 3400 BMW, 1580 Mini and 140 Rolls Royce dealerships. It seems hard to deny that these networks present a solid launch pad for their electric vehicle roll out.

17 April 2018

54

Barclays | Global Autos & Auto Parts FIGURE 33 Global passenger car production locations VW, BMW, Daimler – Daimler – harder harder for Tesla to replicate a global manufacturing and distribution footprint

Source: Barclays Research, Company Reports

17 April 2018

55

Barclays | Global Autos & Auto Parts FIGURE 34 Global EV production plans VW

B MW

Daimler

>EUR72bn investment in EV capacity and projects by 2030

Global EV Plans

> Investment of EUR10bn

>16 production facilities

>25 Electrified Models by

by 2025 for the expansion

for EVs by 2022 (currently

2025, including 12 fully

of the Mercedes electric

3)

electric cars

fleet

>80 new electric models by

> Electrified models to

> Launch 10 pure EVs by

2025, including 50 BEVs

contribute 20-25% of sales

2022

(currently (currently 8 BEV and PHEV

by 2025

>EVs t o contribute 15-25%

models)

of sales by 2025

>Production >Production of 3m EVs/annum

European EV Plans

>EUR1bn investment in Zwickau for EV production ramp up

>Electric Mini to be

>EQ models to be made In

manufactured at Cowley

Bremen from 2019 and

plant UK from 2019

later Rastatt, and

>€200 investment investment i n

Sindelfingen

Barclays | Global Autos & Auto Parts FIGURE 34 Global EV production plans VW

B MW

Daimler

>EUR72bn investment in EV capacity and projects by 2030

Global EV Plans

> Investment of EUR10bn

>16 production facilities

>25 Electrified Models by

by 2025 for the expansion

for EVs by 2022 (currently

2025, including 12 fully

of the Mercedes electric

3)

electric cars

fleet

>80 new electric models by

> Electrified models to

> Launch 10 pure EVs by

2025, including 50 BEVs

contribute 20-25% of sales

2022

(currently (currently 8 BEV and PHEV

by 2025

>EVs t o contribute 15-25%

models)

of sales by 2025

>Production >Production of 3m EVs/annum

European EV Plans

>EUR1bn investment in Zwickau for EV production ramp up

> No formal North American EV Plans

announcemen announcement, t, but E V production likely at

>Electric Mini to be

>EQ models to be made In

manufactured at Cowley

Bremen from 2019 and

plant UK from 2019

later Rastatt, and

>€200 investment investment i n

Sindelfingen

'Battery Cell Competence Competence

>EUR500m investment in

Centre' in Munich

lithium ion battery pack

> Production of electric X3 to start at Spartanburg in 2020

Chattanooga Chattanooga facility

>Investment of $1bn to expand Alabama plant to build EQ SUV from 2020

>Electric X3 to be manufactured in Da Dong >Joint VW JAC investment of EUR10bn to develop and China EV Plans

produce 40 electric vehicles for the Chinese market by 2025

from 2020 >Letter of Intent with Great Wall Motor for production production of electric MINI > High-Voltage Battery Centre opened in Shenyang

>EQ SUV to be made in China from 2019 >Joint Daimler BAIC investment of €655m for the production of BEVs and battery packs in China

in Oct-17 (JV with Brilliance) Rest of Asia EV Plans

> BMW to expand PHEV battery capacity in Thai plant

>Daimler and TAAP to invest €100m (in part) for a new battery assembly facility

Source: Barclays Research, Company reports, Reuters, Bloomberg, FT

Of course EV infrastructure will be necessary too, to entice consumers to buy the vehicles, but the Ionity venture across Europe should help considerably and the Chinese government’s plans for 500,000 public charging points to encourage EV sales of 2m by 2020 provide significant support. In any case, the advantage of having a fully global distribution network, supplier relationships and globalised modular production facilities already in place stand in stark contrast to Tesla.

17 April 2018

56


GERMAN OEMs HAVE DEEP POCKETS… German OEMs are investing heavily in the future

The German OEMs have had to ramp up their spending on electric mobility significantly over recent years. All three have seen their expenditure for capex and capitalized R&D drift up not just in absolute terms but also as a percentage of sales. First and foremost this increase reflects their efforts to meet emission reduction targets primarily through vehicle electrification. More recently spending has been further inflated by the costs associated with autonomy and mobility services. We expect this high level of expenditure to continue in the coming years. VW has announced it intends to spend €72bn on electric vehicle capacity and related projects by 2030. Daimler has committed €10bn by 2025 to the expansion of the Mercedes electric fleet. BMW has never released an overall spending commitment for electrification, perhaps reflecting the fact that it committed to the EV game much earlier than its peers. The group does, however, drip feed regular investment announcements into the market, most recently publishing a €200m plan for a Battery Cell Competence Centre Centre in Munich.

FIGURE 35 Spending on capex and R&D continues to drift up EUR bn

VW Passenger Cars Capex and Capitalised R&D

20

15

10

% of sales EUR bn 10%

BMW Automotive Capex and Capitalised R&D

9%

8%

8%

7% 15 6%

7%

5%

5%

2%

10

EUR Mn

% of sales

20

2%

7% 6% 5%

10

4% 3%

5

2% 1%

1% 0

0%

EUR Mn

% of sales

10% 8%

15

3% 5

% of sales

9%

4%

1% 0%

Mercedes Capex and Capitalised R&D

6%

3%

0

10%

9% 20

4%

5

% of sales EUR bn

0

0%

EUR Mn

% of sales

Source: Company reports

With capex and capitalized R&D of $3.4bn (c.€2.8bn) in 2017, Tesla’s spending is around 50% at the level of BMW and Daimler. And yet as a percentage of sales Tesla outspends its German premium competitors competitors by a multiple of 4.

17 April 2018

57


FIGURE 36 Tesla outspends competitors competitors by 4x as a % of revenue… revenue …

FIGURE 37 …but has significantly less net cash than German OEMs

EUR Mn

% of sales

18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0

35.0% 30.0% 25.0% 20.0% 15.0%

EUR Mn 25,000 20,000 15,000 10,000

10.0% 5.0% 0.0% VW BMW Merc Me rced edes es Tesl Te sla a passenger automotive cars 2017 Capex and Capitalised R&D

5,000 0 -5,000

VW Industrial

BMW Industrial

Daimler Industrial

Tesla

2017 Net industrial cash

% of sales (rhs) Source: Barclays Research, company reports

Source: Barclays Research, company reports

But their pockets are deep…

Moreover, while Tesla ended FY2017 Moreover, FY201 7 with a net debt position of $4.4bn, Daimler, BMW and VW are sitting on comfortable net cash positions of between €16 and €22 bn.

…and cash generation is set to continue

Even taking into account VW, BMW and Daimler’s significant expenditure plans on electrification we expect to see solid free cash generation in the coming years. Daimler, BMW and VW are all poised to benefit from favourable mix developments as the global shift towards higher margin SUVs continues. Combined with moderate volume growth, an ongoing focus on cost discipline and, in the case of BMW, new product momentum, they are well placed to meet the challenges of the future, not just the electrification of their line up but also the transition to autonomous driving. By contrast we forecast a continued cash burn at Tesla until at least 2023.

FIGURE 38 Free cash flow estimates 2017-2020 EUR Mn 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 -2,000

VW industrial

BMW automotive

Daimler industrial

Tesla

-4,000 -6,000 2017A

2018E

2019E

2020E

Source: Barclays Research, company report. Note: Barclays definition of FCF = Operating Cash Flow – (Capex (Capex + Capitalised R&D). VW FCF is pre diesel provisions

17 April 2018

58


Tesla ‘core vs. context’ – Tesla Tesla Software vs. Tesla Auto Auto – – reinforces reinforces the cash risks of manufacturing woes We highlighted Tesla’s dilemma of ‘core vs. context’ in a piece in 2015, yet it still remains highly relevant (see: “Core vs. context – the the dilemma of Tesla Software vs. Tesla Auto ” 11/13/15).

Simply, the challenge Tesla faces is that while it has an advantage over its competition in the crucial area of software (which will be the defining factor for vehicles moving forward), this advantage is at risk of being overshad overshadowed owed by Tesla’s challenges in the more mundane, yet crucial, task of manufacturing efficiency. We think the ‘core vs. context’ framework is an appropriate one to use for Tesla. Broadly, if it creates differentiation, it’s core; if not, it’s context. Both core and context are necessary for success, but context does not provide a competitive advantage. Ideally, you invest more resources in core and less in context. Applying the ‘core vs. context’ model to Tesla, “core” is software and “context” is its auto manufacturing ops. While Tesla is currently spending the right amount on software, we are concerned that it has been inefficient on the “context” side. While this model may have worked for Models S/X, it may not work as Tesla ramps on Model 3. Simply put, if Tesla doesn’t significantly improve its capital efficiency on the context products/processes, then it will likely be unable to invest enough on the core products which will make Tesla truly differentiated. Thus far, the capital efficiency track record for Tesla hasn’t been good, and the capital calls are only going to be more intensive moving forward. We think the market likely underappreciates the capital intensity that comes with vertical integration (supporting elevated machining and engineering costs – tooling, tooling, forging, casting, product design, etc). Moreover, with Tesla expanding its focus into other areas (i.e. energy storage, trucks), risk exists that capital efficiency could get worse.

We estimate all of this adds up to nearly $27bn of capex for Tesla through 2023, with a FCF burn of $7.5bn during that period, as profit per vehicle remains low. Of course, the saving grace to all this is if the capital markets remain an open well for Tesla – if if investors remain willing to fund their losses and cash requirements, then none of this is an issue. FIGURE 39 The flow of innovation – innovation – core core vs. context

The Flow of Innovation Invest more resources here

Core

Context

Extract some resources here

Mission Critical 2. Deploy

3. Manage Manage

1. Develo Develop p

4. Offload Offload

Enabling Invest some resources here

Extract more resources here

Source: TCG Advisors

17 April 2018

59


FIGURE 40 Tesla will need to spend $27bn of capex through 2023… 2023 … $27bn of capex between between 2018 and2023

Capex, $mn

FIGURE 41 …driving a $7.5bn cash burn Free Cash Flow, $mn

$7.5bn cash burn between between 2017 and 2023

$1,000

$6,000 5,064 $5,000

5,235

$0

4,577 4,100

4,209

290

$500

4,312

-6

-$500

$4,000

-$1,000

3,415

-635

-853

-1,024

-$1,500

$3,000

-219

-505

-1,591

-$2,000

-1,954

-$2,500 $2,000

1,635

-$3,000 1,281

239

-2,964

-$3,500

970

$1,000

-807

-$4,000

264

-3,932

-$4,500

$0 2012

2013

2014

2015

2016

2017

2018E

2019E

2020E

2021E

2022E

2012

2023E

2013

2014

2015

2016

2017 2018E 2019E 2020E 2021E 2022E 2023E

Source: Company reports, Barclays Research estimates

Source: Company reports, Barclays Research estimates FCF defined as cash from operations less capex, in clusive of collateralized lease borrowing

FIGURE 42 Capex per unit, capex as % of sales – sales – both both will remain elevated through Tesla’ Tesla’s ramp… ramp…

FIGURE 43 …while operating profit per unit is likely to remain low given significant investment in R&D and SG&A

Capex per unit

Capex as % sales

Adj. EBIT per unit

$90,000

70%

$6,000

$75,000

60%

$4,000

4,671 1,656

$2,000 50%

$60,000

40% $45,000 30% 20%

$15,000

10%

$0

0% 2 01 01 2 2 01 01 3 2 01 01 4 2 01 01 5 2 01 01 6 2 01 01 7 2 01 01 8E 8E 2 01 01 9E 9E 2 02 02 0E 0E 2 02 02 1E 1E 2 02 02 2E 2E 2 02 02 3E 3E Ca pe pex p er er unit ( LH LH S) S)

Ca pe pex as % sa le les ( RH RHS )


17 April 2018

2,354

2,945

195

$0 -$2,000

-1,195

-$4,000 -4,085 -4,160

-$6,000

$30,000

1,501

-6,113

-$8,000 -$10,000 -$12,000

-11,294

-$14,000 2013

2014

2015

2016

2017 2018E 2019E 2020E 2021E 2022E 2023E


60

Barlcays Tesla Full Report

Recommend Documents