•••

(as long as you forget about CO2 credits, •• • cannibalization, EV regulations, price elasticity, EV class demand, income tax credits, ZEV credits, government subsidies, competitive responses, battery advancements, NHTSA and EPA rules ... which you can't)

© �019 Reven�e �l'ilal}'!ics, Inc. 02

TABLE OF CONTENTS

PRICING THE ELECTRIC FUTURE: HOW TO PRICE A TESLA 3

WHAT SHOULD TESLA DO? 3 Model 3 Price Cuts: At First Glance, a Bad Idea 4 Government CO2 and EV Regulations Muddy the Picture 4 Higher Margins Make Price Cuts More Attractive 5 Sales Cannibalization Makes Price Cuts Less Attractive 5 Price Cuts Today Yield Future Benefits 7 Closing Thoughts: The Only Certainty is a Need for Careful Analysis 8

© 2019 Revenue Analytics, Inc. Pricing the Electric Future: 03 How to Price a Tesla ™ Tesla’s pricing seems to be constantly in the news. It makes one wonder: what does it take to make a sustainable profit from electric vehicles? Maintaining prices high makes it hard to grow sales but cutting prices dilutes margins.

WHAT SHOULD TESLA DO?

DISCLAIMER: We're not best buddies with Elon (yet), and Tesla is not a client of ours (yet!). We would never advise a client without leveraging proprietary data including forecasts of costs, sales and margins. Without that data, the answer is pure guesswork...but well-researched and thoughtful guesswork can be fun and informative, so let's give it a go,

shall we? DEMAND INTELLIGENCE

After harmonizing the aforementioned data, one thing Tesla should do (if they haven’t already), is build and apply an analytics capability to understand the total impact of price changes on profit.

Elon Musk hasn’t made this data publicly available, of course - but look at what this might involve using what we do know to run an analysis. Since the entry level Model 3 is Tesla’s top seller, we’ll focus on that vehicle.

Tesla recently announced a $35,000 “standard range” Model 3, but that variant hasn’t sold yet so we’ll focus on the variants sold in 2018.

• The base price was $44,0001 • After options, the average Tesla Model 3 rose to $59,3002 • In Q4 2018, Tesla sold 63,150 of these cars.3

The Model 3’s profit margin is highly confidential. But Tesla did reveal that its gross margin is above 20%, with a target of 25% in 2019.4 Let’s split the difference and use 22.5% as our initial estimate. This translates to around $13,350 per car. This margin will be impacted by Tesla’s recent price cuts of $3,100 (first a cut by $2,000 at the start of 2019, and then another $1,100 price cut5 in February).

Were these price cuts a good move? It depends.

1Electrek, “Where is Tesla’s $35,000 Model 3”, https://electrek.co/2019/01/24/tesla-model-3-base-35000/, January 24, 2019. 2CleanTechnica, “Tesla Model 3 Average Selling Price (ASP) = $59,300, Surveys Find”, https://cleantechnica.com/2018/08/23/tesla-model-3-average-selling- price-asp-59300-surveys-find/, August 23, 2018. 3Tesla Inc. press release, “Tesla Q4 2018 Vehicle Production & Deliveries, Also Announcing $2,000 Price Reduction in US”, https://globenewswire.com/news- release/2019/01/02/1679576/0/en/Tesla-Q4-2018-Vehicle-Production-Deliveries-Also-Announcing-2-000-Price-Reduction-in-US.html, January 2, 2019. 4Tesla Inc., “Tesla Fourth Quarter & Full Year 2018 Update”, http://ir.tesla.com/static-files/0b913415-467d-4c0d-be4c-9225c2cb0ae0, January 30, 2019. 5Autoweek, “Tesla Model 3 price slashed by a further $1,100”, https://autoweek.com/article/green-cars/tesla-model-3-price-slashed-1100#ixzz5glWvTe4b”, February 6, 2019.

© 2019 Revenue Analytics, Inc. 04 Model 3 Price Cuts: At First Glance, A Bad Idea TM

Everything starts with the customer, so how will a typical Tesla customer react?

™ One study found that for every $5,000 increase in the base price of an electric vehicle, demand dropped by 40%.6 So, a $5,000 price increase – 11.4% – would drive a 40% reduction in demand, a 3.5-to-1 ratio. This value, called the Model 3’s price elasticity, is fairly typical of automobiles.

Applying this price elasticity to $3,100 of price cuts, 5.2%, increases demand by 18.2% (i.e. 3.5 multiplied by 5.2%). At its initial margin and volume, the Model 3 brought in about $838 million in gross profit in Q4 2018. With the price decrease, its 18.2% higher sales volume wouldn’t offset the $3,100 lower margin, reducing gross profit to $760 million. We’re just considering gross profit, not net profit, since R&D, capital expenses, etc. are already sunk. So at first glance, the price cut appears to be a bad idea.

Government CO2 And Ev Regulations Muddy The Picture DEMAND INTELLIGENC E But it gets more complicated. Tesla buyers are eligible for electric-vehicle income tax credits. Prior to January 2019, they received $7,500; but this dropped to $3,750 in January 2019 because Tesla’s total sales of electric vehicles passed a 200k-unit threshold.7 Of Tesla’s competitors, none are in the same position, with only GM, maker of the Bolt and Volt, losing half its tax credit starting in April.8 So in the market, Tesla’s price cut really translates into a “perceived” price increase of $650 (i.e. $3,750 less tax benefit minus $3,100 price cut versus a $3,750 “perceived” price increase if Tesla held price. Applying the same math as before to these two scenarios, we calculate that the price cut really prevented a 22% crash in demand to 49,173 units rather than a modest decline of 4%. With this view, the price cut yields a profito f $618M versus $653M. It still doesn’t appear to be a good idea, but the negative impact is less.

We’re not done yet. California and several other states require automotive OEM’s to earn Zero-Emission-Vehicle (ZEV) credits in proportion to their sales in the state. Every time Tesla sells an electric vehicle in these states, it earns credits. Once you reach this credit threshold you can sell the excess to other OEMs who are in short supply and desperately need these credits. Similarly, OEMs can trade other types of non-ZEV emissions credits to comply with NHTSA and EPA rules. In the third quarter of 2018, Tesla reported $189M in sales of credits while it delivered 83,500 vehicles.9

6Green Car Reports, “Tesla sales: how much do rebates matter to buyers in Canada?”, https://www.greencarreports.com/news/1115929_tesla-sales-in-canada- how-much-do-rebates-matter-to-buyers, March 27, 2018. 7Car & Driver, “Tesla Buyers to Lose $7,500 Tax Credit at End of Year”, https://www.caranddriver.com/news/a23743195/tesla-tax-credit-deadline-date/, Decem- ber 26, 2018. 8Electrek, “GM Hits Electric Vehicle Tax Credit Threshold, Phase Out to Start in April”, https://electrek.co/2019/01/03/gm-hit-electric-vehicle-tax-credit-thresh- old-phase-out-april/, January 3, 2019. 9Seeking Alpha, “Tesla’s 10-Q Reveals Q3 May Not Have Been Profitable, After All “, https://seekingalpha.com/article/4217588-teslas-10-q-reveals-q3-may-profit- able, November 2, 2018.

© 2019 Revenue Analytics, Inc. 05

Each vehicle netted over $2,250 from credits alone, increasing its effective ™ profit margin. Factoring in this higher profit margin, the net impact of the price cut is to drive a profit of $755M vs. $764M without the price cut. It’s a much closer call now, but the price cut still doesn’t appear justified.

Higher Margins Make Price Cuts More Attractive As mentioned at the start of the analysis, we assumed an average gross margin of 22.5%, with Tesla targeting 25% in 2019. What if Tesla achieves its 25% gross margin target in early 2019? Its per unit margin before price cuts and credits would then be $14,750. With each incremental sale worth more, repeating our calculations yields gives us a profit of $845M with the price cut versus $837M without. DEMAND INTELLIGENCE This seems to suggest that Tesla may have finely tailored its Model 3 pricing to maximize profit.

But there might be an even more solid case for the price cut. In the new age of mobility, auto OEMs don’t stop profiting once a car is sold; they profit over the entire customer life cycle. For example, with a $7,000 over-the-air software update, essentially zero-cost for Tesla, customers can activate the Model 3’s Enhanced Autopilot functionality, and an even more enhanced “full- self-driving” option is slated in the future.10 Based on take-rate surveys, about 23% of Model 3’s were purchased without either feature.11 If only 10% of owners later upgrade to the enhanced autopilot, the average incremental margin is $700/vehicle, raising the gross profit to $887M with the price cut versus $871M without. Now the price cut is even more attractive.

Sales Cannibalization Makes Price Cuts Less Attractive Unfortunately, we can’t consider the Model 3 in isolation. Every time Tesla sells a Model 3, the sale has to come from somewhere, as shown in the figure on the next page (6). Either the buyer’s second choice was to not buy a car, to buy a competitor’s vehicle, or to buy another Tesla. The first two situations are great for Tesla; they either grew the market or stole a sale from a competitor. But in the last case, Tesla steals a sale from its more expensive and more profitable vehicles – Model S or Model X. At an estimated 25% margin, these vehicles, with an average selling price of over $121,000, yield an average gross profit of $30,250

10Tesla Model 3 Configurator, https://3.tesla.com/model3/design#autopilot, February 27, 2019. 11Teslike Model 3 Survey #1, https://docs.google.com/spreadsheets/d/1YeLtMxFt9Lh8mndZhjOqrzFfWQ_r5T9LS-l3gQzPXCk/edit#gid=756674994, February 27, 2019.

© 2019 Revenue Analytics, Inc. 06

Competitor Pure incremental profit & competitor loss BEST TM

™ Not Buying Pure incremental profit BETTER Model 3 Sale Base Tesla Lower incremental profit GOOD Model 3

Tesla Model X BAD or Model S Negative incremental profit

Such “sales cannibalization” seems to be confirmed by delivery figures, which show that Tesla delivered fewer Model S’s in Q2 2018, after the widespread introduction of the Model 3, than in Q2 2017.12 Data on cannibalization is difficult 13

DEMAND INTELLIGENC E to come by, but Tesla notes that 4% of Model 3 buyers trade in a Model S. Since few Model S’s have been on the road long enough to be traded in, it’s reasonable to assume the cannibalization rate is considerably higher, say 10%.

Let’s consider what this 10% cannibalization rate means. Out of every ten Model 3 buyers, on average nine wouldn’t have bought another Tesla had they not bought the Model 3. The profit from these sales is totally incremental, but the tenth customer diverted away from a Model S or Model X. While Tesla gained the roughly $15,000 to $16,000 gross profit from the Model 3 sale, it lost the $30,250 gross profit from the lost sale of a Model S or Model X. So, on average, over the ten customers, Tesla loses and incremental $3,025 per sale from cannibalization. This lower incremental “diversion-adjusted” gross profit per sale means the business case for the Tesla price cut is again negative, yielding $704M in incremental portfolio profit versus $722M.

What about Tesla’s recently announced $35,000 Standard-Range Model 3 and massive price reductions to the Model S and Model X? What effect will they have on sales cannibalization? Since the Model 3 is now closer to the higher-end Teslas, it will likely lose sales to them initially. But conversely, it’s more likely to get those sales back if its price drops, raising the cannibalization rate.

12InsideEVs, “Are Sales Dropping Due to Model 3?”, https://insideevs.com/tesla-model-s-sales-dropping/, July 7, 2018. 13AutoNews, “Musk: After Q4 Profit, Tesla to Be in the Black Every Quarter in ‘19”, https://www.autonews.com/manufacturing/musk-after-q4-profit-tesla-be- black-every-quarter-19, January 30, 2019.

© 2019 Revenue Analytics, Inc. 07 Similarly, the standard-range Model 3 will also eventually steal sales from the other Model 3’s. If Tesla constrains production to only use excess Model 3 capacity to build its base model, it might be able to mitigate those diversions. ™ Given Tesla’s massive order backlog, it’s hard to tell how many Model 3 buyers purchased a more basic Model 3 just because they were impatient to get a Model 3, versus really wanting the additional options. Also, despite having the same name, we must keep in mind that a $35,000 car and a $59,000 car don’t really compete for the same customers. Substantially longer range, quicker acceleration and a premium interior differentiate the Model 3 variants.

For the purposes of our analysis, the standard-range Model 3 should have little impact. It’s expected to not quite break even in terms of profit, so diverting one of its sales is slightly beneficial from a profit standpoint and should have little or no impact on Tesla’s incremental margin.14 Our analysis is based on percent change in price vs. percent change in sales, so even if the DEMAND INTELLIGENCE baseline sales volume and profit change, any increased or decreased profit from a price change could change in magnitude but not in direction.

A good pricing decision will remain a good pricing decision.

Price Cuts Today Yield Future Benefits Consider Tesla’s competitive position. For the first six months of 2019, the federal government subsidizes each unit it sells by $3,750. This subsidy drops to $1,875 for the last six months of 2019. As 2020 begins, Tesla will face a horde of competitors benefitingf rom the $7,500 tax advantage. To compete, it needs to get its prices down. To do that profitably, it needs to reduce its costs by gaining further experience and economies of scale.

As of January 2019, Tesla had produced about 160k Model 3’s.15 After the price cut, it’s on track to produce about 243k more (non-standard-range) Model 3’s to meet demand in 2019 versus about 196k without it. Based on one study, Tesla’s battery pack cost for the Model S dropped by 20% each time it doubled its cumulative production.16 Assuming a similar learning curve for the Model 3 and applying it just to the cost of the battery, about $100/kWh for a 75kWh battery at the end of 2018,17 Tesla would reduce its costs by about $200 more per vehicle in 2020 thanks to the added 47k volume generated by the price cut. (Similar analysis might underly Tesla’s decision to sell the standard-range Model 3.)

14Benzinga, “A Week In Tesla: SEC Alarms, Low-Cost Model 3 And Dealership Closures”, https://www.benzinga.com/analyst-ratings/analyst- color/19/03/13279569/a-week-in-tesla-sec-alarms-low-cost-model-3-and-dealers, March 1, 2019. 15InsideEVs, “Cumulative Tesla Model 3 Production Estimate Exceeds 160,000”, https://insideevs.com/cumulative-tesla-model-3-production-160000/, January 10, 2019. 16Forbes, “Tesla’s Innovations Are Transforming The Auto Industry”, https://www.forbes.com/sites/innovatorsdna/2016/08/24/teslas-innovations-are-transform- ing-the-auto-industry/#5204cc6f19f7, August 24, 2016. 17Electrek, “Tesla’s Gigafactory 1 Battery Cells Have a 20% Cost Advantage Over LG, New Report Says”, https://electrek.co/2018/11/20/tesla-gigafactory-battery- cells-made-cost-advantage-panasonic-lg-report/, November 20, 2018.

© 2019 Revenue Analytics, Inc. 08 Since Tesla’s upcoming Model Y promises to share 76% of its parts with the 18 ™ Model 3, this $200 per unit savings from increased production would likely translate to that model as well. Conservatively, if Tesla sells only 350k Model 3’s and Model Y’s in 2020 (half its predicted total global production of 700,000 vehicles)19, the $200/unit cost savings adds up to $70M. When this is taken into consideration, the net result of the price cut is a profit of $773M with the price cut vs. $722M without.

Now the price cut again looks like a great idea.

Closing Thoughts: The Only Certainty is a Need for Careful Analysis Is this our final conclusion? By now you’ve probably guessed that, of course, it

DEMAND INTELLIGENC E isn’t… Tesla even raised its prices just after this eBook was written! We’ve come a long way from our simple analysis, but undoubtedly additional refinement remains which would require Tesla’s involvement and proprietary knowledge. The figure below illustrates how varied our profit estimates were in our analysis – and underscore how critical it is to fully understand the wide- ranging implications of pricing decisions have on not just sales, but also regulatory compliance and future manufacturing efficiency.

18CleanTechnica, “ To Share ~76% Of Parts With Model 3, Be Built At Gigafactories”, https://cleantechnica.com/2019/01/31/tesla-model-y-to-share- 76-of-parts-with-model-3-built-at-gigafactories/, January 31, 2019. 19Business Insider, “ Backs Off Tesla’s Goal of Making 1 Million Vehicles by 2020”, https://www.businessinsider.com/elon-musk-hedges-on-tesla- making-1-million-cars-by-2020-2018-8, August 2, 2018.

© 2019 Revenue Analytics, Inc. 09 We made a few simplifying assumptions. We assumed that all of Tesla’s sales

were in the U.S., which is relatively accurate for the Model 3 – European and ™ Chinese sales have just started,20,21 and the Canadian market is small, only about 2k per quarter, and closely linked to the U.S. market.22 Most importantly, we assumed that the Model 3’s Q4 2018 sales represented unconstrained demand representative of the future.

What we don’t know is how many of the hundreds of thousands of Model 3 orders remain backlogged on Tesla’s books, how soon production can catch up, and whether some sales were pulled ahead by the rush to gain the full $7,500 tax credit in 2018. Nor do we know how the market will respond to the many new electric vehicles being offered by competitors.

In the future, Tesla’s growing model lineup will reduce the potential benefits

of isolated price changes to individual models or trim levels. The less the DEMAND INTELLIGENCE differentiation between products, the more likely customers will be to just move around within the portfolio in response to a single price change. If a price goes up, they’ll buy a different Tesla – maybe a more profitable one, maybe not. If a price goes down, more sales growth will come at the expense of other Teslas. To really profit from this situation requires a simultaneous set of price changes across the portfolio, factoring in the resulting diversions. The complexity of pricing decisions will continue to rise rapidly.

This highlights what we believe to be the most important takeaway from this our analysis of Tesla’s price cut:

In a dynamic market with ever growing portfolios and competition, the complexity of pricing decisions is growing exponentially.

We briefly touched on simple portfolio implications on Tesla’s relatively small model lineup. Consider what full-range OEMs have to consider with simultaneous price changes across dozens of models in their portfolio, each with multiple trim levels and powertrains and each competing for shared production capacity.

We briefly discussed the impacts of the sales of ZEV and other credits, but it’s important to note that the value of these credits depends on the credits earned by OEMs and on their expected fuel efficiency relative to regulations in future years. Throw in multiple-year product development cycles and impending major changes to fuel economy and GHG regulations across the globe and the need for scenario analysis becomes critical.

20InsideEVs, “Tesla Model 3 Oficially Approved for Sale in Europe”, https://insideevs.com/tesla-model-3-approved-sale-europe/, January 19, 2019. uary 22, 2019. 22InsideEVs, “Tesla Model 3 Outsells All Other EVs in Canada”, https://insideevs.com/tesla-model-3-outsells-evs-q3-canada/, November 11, 2018.

© 2019 Revenue Analytics, Inc. 10 TM

With all these interdependencies, even evaluating the impact of a pricing

™ strategy is exceedingly challenging, let alone optimizing a pricing strategy.

Automotive pricing isn’t easy, and it’s getting tougher all the time. DEMAND INTELLIGENC E If you enjoyed this eBook, please visit our Auto OEM Resources Section to find other helpful Revenue Management content.