It took a little more time than I expected to pull this together into a post, but below are the basics of my modeling of the core auto business for Tesla I mentioned in the google hangout.
Before I get into the numbers, here’s a little context on why I’ve taken this approach to valuing Tesla. I am quite confident that even confining a forecast to a conservative take on only the parts of Tesla’s business I think we can have considerable confidence projecting 8-10 years out strongly indicates that Tesla is substantially undervalued based on fundamentals. In other words, the value of other businesses in development at Tesla (such as the Tesla Network, and Tesla’s as yet unannounced strategies with both the semi truck and the smaller urban bus currently under development) which have tremendous potential but could have wildly varying outcomes are not needed(1) at current prices to justify a bullish position. Of course, if the stock price rises above the valuation of the core auto business before the other opportunities become more realistic to forecast, this valuation approach will be at the limits of its useful application.
I think many underestimate how possible it is to do a reasonable valuation of Tesla’s core auto business 8-10 years out. That the ease and utility of such a forecast is misunderstood is not surprising, as 1) it is very unusual for such a long range forecast to be meaningful for a high growth business, 2) for several years there appears to have been something of an ongoing campaign to convince the public that Tesla is not and cannot be rationally valued despite the fact that the analysts covering the stock do indeed use the tools of fundamental analysis to build models to do precisely this. The ongoing seeming campaign to misinform I mentioned is found in the repetition of the “can’t value the darn thing”, “story stock”, “completely detached from fundamentals”, “cult stock” mantras we’ve seen for years. Hopefully this post helps some realize that Tesla’s basic auto business can be rationally valued, and it’s not so difficult to jump in and make your own rough forecasts with whatever assumptions you find most compelling. Fwiw, I recommend creating multiple scenarios to determine a range of valuations.
On to the model...
I’m going to list the key inputs one needs for this basic model, and put in italics my assumption here for this example of a forecast of Tesla c.2025-26 (in projected 2025-26 dollars):
Auto Business Volume, Average Sale Price, Operating Margin (2) by vehicle: (2.510 million total)
Model 3: 1 million, $45K, 11%
Model Y: 1 million, $47K, 11%
Tesla Pickup Truck: 350K, $50K, 12%
Model S/X: 150K, $110K, 13%
Roadster: 10K, $140K, 15%
Operating Profit from core Vehicle Business: $14.575 Billion
Tesla Energy, Conservative Scenario
Volume: 50 gWh (50 million kWh)
ASP: $200 kWh
Operating Margin: 15%
Operating Profit from TE Business: $1.5 billion
Total Operating Profit: $16.075 billion
Effective Tax Rate: 25%
Earnings: $12.056 billion
Increase in Share Count(3): 75 million shares
2025/26 Share Count: 235 million shares
EPS: $51.30
Expected Annual Earnings 5-Year Growth Rate beyond 2026(3): 15% to 25%
PEG(4) : 1.5
PE (PEG*Growth Rate): 22.5-37.5
Projected 2025/26 stock price: $1155-$1925 (market cap $265 billion to $442 billion)
If you like, from here you can use a discount rate of your own choosing to estimate the present fair value of the stock.
Notes:
(1) I think there’s a reasonable case to be made that to merit the PE of the high end of the valuation range in this example, Tesla in 2025/26 will either need to look like it will reach Toyota volumes in another 5 to 6 years and continue from there to grow roughly 50% faster than the auto industry as a whole (which I do think is possible), or, the market will need to value the potential that other markets like those I’ve not included in the earnings model will be growing earnings for Tesla five years further out (by about 2030). Put another way, to justify a PE near 40, Tesla in 2026 may need to look like it’s headed to be an automaker with market share beyond the share the world’s largest automaker has in 2017, or, a business that is “more than an automaker.”
(2) My memory is of Elon using the phrase “net profits” when describing expectations of reaching low to mid teen profitability. I made a brief search for Elon’s comments to clarify this, but did not find them. Given that the media descriptions of profitability I’ve seen on other automakers focus on gross profits or operating profits, and Porsche, whom Tesla aspires to approach in profitability, has a 15% operating margin, I assume that the low to mid teen profit rate target Elon has stated numerous times refers to operating margin rather than net margin. As operating profit takes into account the depreciation of plants and equipment, this model in effect incorporates Tesla’s expectations for the cost of additional plants and equipment. While the expense of the coming capex for these new plants is taken care of in the operating margin, in point (3) below I discuss modelling the growth in share count that also needs to be taken into account if the money for the capex spending will in part come from future secondary offerings.
(3) Estimating Tesla’s future capex needs, employee stock compensation and the increase to the share count to pay for both could be a separate model of its own. While I think this is worth diving into more deeply, in fact, I hope some of us take this on, I think the rough estimate here of increased shares is conservative enough to allow people to play with the model as is (of course, feel free to tweak the assumed amount of additional shares needed as you like). As far as this model taking a conservative approach on share count, even if Tesla were to need to issue another 30 million shares beyond the additional shares I’ve forecasted here, this model’s projection of Tesla’s 2025/26 share price would still be $1,025 to $1705.
I assume Tesla will add about 75 million shares, or about 47% over their current count to cover shares granted to employees and offered in secondaries over the next 8-10 years for any of the funding of future factories for which retained earnings not adequate. Of course, Tesla may substantially shift to using loans rather than secondaries, in which case, they would have interest expense impacting their earnings, but a smaller increase in the share count.
I’ve used the 75 million rough estimate of the increase in the share count, which comes to about a 15% larger increase in share count from today through 2025/26 than the roughly 65 million shares Tesla has added since it went public in 2010. While Tesla will have more employees to grant shares to in the next 8 to 9 years than the past 7, 1) the shares are already a 10X more powerful unit of currency than they were pre-Spring 2013, 2) 10-15 million of the past share increase was from the Solar City acquisition, and 3) Tesla will have dramatically more money from their sales of existing products to help pay for growth than they did in the past. I estimate that between now and the end of 2024, ongoing operations of the core auto business will provide $20-25 billion of cash which Tesla can invest in new plants and equipment. What’s more, there is the possibility that one or more of the other businesses under development could generate cash on a similar scale in this time period.
It would be helpful to research some finer detail on Tesla’s stock compensation policy to project future share distributions. I’m not totally clear how much of stock compensation already committed to is reflected in the current 160 million share count, and such research would also be helpful in modeling shares needed for compensation plans yet to be added that will impact the share count in 2025/26.
Finally, I think the amount of additional capex needed to reach the volumes assumed in this model could vary very significantly from roughly $20 billion to $45 billion, due to factors such as, a) how successful Tesla is in re-engineering the machine that builds the machine (i.e., potential for more efficient use of capital with “alien dreadnoughts”), b) the range of incentives Tesla might receive from governments competing to entice Tesla to build a GF and/or auto factory in their jurisdiction, c) Panasonic’s TBD future role in the financing of GFs, d) increasing cell energy density from improving cell chemistry leading to more output for each dollar spent on the equipment to build the cells, e) inflation’s impact on the cost of building and equipping these factories).
(4) I assume a PEG of 1.5, considerably more conservative than the S&P 500’s current PEG of 2.1 (using a current PE of 24.84 and State Street Global Advisors reporting of an 11.94% earnings 3-5 year growth rate for the S&P 500 (they describe their methodology for the growth rate projection as using earnings estimates for the index’s underlying holdings from FactSet, First Call, I/B/E/S Consensus and Reuters)).
As I mentioned on the hangout, I really recommend learning more about PEG, popularized by Peter Lynch, if you’re not too familiar with it. I didn’t mention it then, but among other things, it’s a very helpful tool in walking through a rational reconciliation of Tesla’s valuation to other automakers.
Additional Notes, Discussion Points:
(5) I referred to my assumptions here as a conservative take on Tesla’s core auto business. Is 2.5 million vehicles conservative? If you think either a) long range EVs will be a niche part of the auto market in 8-10 years, or b) the other automakers will switch to producing tens of millions of such EVs within that timeframe, my assumptions may look aggressive to you. If you think there will be a gap in the 2020s between the amount of long range EVs available and the demand for them, these assumptions are conservative. I created a thread a couple of months ago making a pretty detailed case for the latter.
(6) I have more confidence in being able to make an estimate for the core auto business than TE. I put TE in anyway because 1) there already are sales, 2) even if my volumes turn out not to be conservative some of this will be made up for if Tesla continues selling solar panels and roofs for which I have not put in any earnings (I also did not put in earnings from several other smaller core automotive sources such as the CPO businesses, a simple base semi truck business of selling the semis, and battery upgrades to older vehicles), and 3) its contribution to the valuation is only about 10%.
(7) Tesla Service and the SuperChargers are not included in this model as they are assumed to run at break even. Tesla has publically stated this if their policy for the service centers, but the SuperChargers do have the potential to be a source of profits.
(8) I’ve chosen to model 2.5 million in sales. I actually think Tesla will have 3 sets of GFs and Auto plants up and running by then, likely producing about 3 million vehicles per year. One of these locations will most likely be in China, where it is very probable Tesla would be sharing costs and profits with a partner. I think modelling for 2.5 million sales is a simple means of hedging for Tesla sharing the profits from their sales in China with a partner.
(9) I’ve not put any interest expense or repayment of existing debt into the model. I base this on the assumption that Solar City’s currently contracted business is sufficient to pay for the debt and interest due on Tesla’s debt, most of it from Solar City. I’ve not modelled here additional earnings from future solar panel or roof sales.
(10) Please feel free to point out anything you think is substantial that this simplified model of the core auto business has missed. At this point I’m aware of some possible auto business cash generators I’ve omitted (semi trucks, the CPO business, potential SuperCharger profits, and profits from replacement battery upgrades on older vehicles), but not expenses. Please let me know if it looks like I missed something, particularly an oversight that leads to an overestimate of the company’s value.
