No emotions, minimum speculations, just raw impartial numbers. We will answer once and for all what is the fair value of TSLA.
Chapter 1. Bull thesis (and other lies I tell myself?)
Let's start with the typical bull thesis. The one you have probably encountered many times in the wilderness of reddit or twitter. It goes like this:
- Tesla's car sales will grow 50% annually for foreseeable future. Eventually reaching annual production rate of 20M by 2030. Source: Technoking himself
- Net margins will stay as high or even grow further from the latest 13.1% (Q4 2021). Commonly cited reasons are 4680, Gigacasting… maybe even Alien Dreadnought?
- Tesla is a tech company. They will generate tons of revenue by selling software such as
FraudFull-Self-Driving, and might eventually launch their own marketplace (see AppStore).
- Tesla is… ETF? (cannot add a link to youtube video, but it is from solving the money problem)
- Tesla
sexbot. Enough said.
Let's start with an automotive sales part:
- Say Tesla is such GigachadTM that it reaches 20M sales without reducing the prices or introducing the cheaper model(s). According to Q4 2021 Financial Report, current ASP is $50.7K, derived as auto revenue excl. regulatory credits divided by the number of delivered cars. Assuming 3% average inflation for the next 9 years (incl. current spike) the ASP in 2030 would be $66.2K (50.7 x 1.03^9)
- With net margins of 13% that would result in 66.2K x 20M x 0.13 = $172B net income.
- Eventually growth by 2030 will taper and converge to automotive industry average. As of writing, PEs of auto peers: Toyota – 9.64, Volkswagen – 6.88, Ford – 10.10, GM – 7.10, BMW – 4.75. But Tesla being Tesla, so we award an automotive Tesla PE of 15.
- Tesla market cap by end 2030 is (drum roll…) 172B x 15 = 2.58T. An absolute automotive leader with 20M sales at an average price of 66.2K USD, with outstanding operating margins (~twice the industry average), with PE 15 (approximately twice the industry average) will triple from the current valuation (or double from January 3's)?
- Taking an average of 7% market growth leads to Net Present Value (NPV) of 2.58T / 1.07^9 = 1.4T*, so … Tesla on January 3rd was pretty fairly valued? Although why would anyone invest in a single stock for a 7% growth versus investing into SPY?
- From the other angle, if you invest in TSLA now (market cap of 890B as of writing) it will return you (2.58 / 0.89)^(1/9) = 1.125 or 12.5% annually. Not too shabby, but also anything but impressive in contrast to its growth in the last two years.
But careful observer would remind me that we are talking about Tech company and not an Auto company. But before we go there… let's discuss what is wrong with the bull thesis above.
Chapter 2. Automotive market.
Many analyses that I have read address future volumes only from the perspective of the supply. Analyses argue that the ramp up of the existing factories plus the introduction of new ones can support 50% growth, eventually reaching 20M car sales by 2030. What they often fail to address is the total addressable market (TAM), which is in our case the EV market in 2030. To be clear, below we will include both plug-in hybrids (PHEV) and battery electric vehicles (BEV) as parts of the EV market. The main reasoning is that for a wide target audience PHEV covers 95% of all use-cases (daily trips within a city) with electric power, therefore creates a real alternative to buying BEV (what happened to me and my wife personally).
No doubt the EV market will be enormous by 2030. In particular:
- EU proposes to ban new ICE cars by 2035 (source). Citation: “… if the EU raised its CO2 emission reduction targets to 50% by 2030, it would bring new fossil-fuel car sales across the bloc down to virtually zero by then… Brussels also proposed allowing plug-in hybrids to count as low-emission vehicles up to 2030 …”. From this we can assume EV penetration rate of a 100% in EU by 2030.
- China plans to transition 40% vehicles sales to so-called “New Energy vehicles” (that include plug-in hybrids, fuel cells, and battery electric vehicles) by 2030 (source). So EV penetration rate in China of 40% by 2030.
- and USA target half of all vehicles sold in 2030 in US to be electric (also includes plug-in hybrids, source), i.e. 50% EV penetration rate for the US by end 2030
- The rest of the World mostly do not have any plans for phasing-out Internal Combustion Engine (ICE) cars (source). Anecdotally, when I visited my hometown of 300K population (in former USSR country) last winter I couldn't locate a single EV, whereas they are common in European city where I live now. We will make an assumption of 20% EV penetration rate for the rest of the world.
2019 automotive sales by region as a percentage of the global are as follows (source): China – 26.5%, EU – 25.3%, US – 18.0%, Rest of the World – 30.2%. By taking into account assumptions on regional EV penetrations rate, we obtain: 0.265 x 0.4 + 0.253 x 1.0 + 0.180 x 0.5 + 0.302 x 0.2 = 0.509 or 50.9% global EV penetration rate.
The next step is to evaluate total car sales in 2030. There are various forecasts, however most of them are in the same ballpark. According to ResearchAndMarkets (source) global automotive sales should reach 122.8M units by 2030. Worth noting that global automotive sales did not practically rise since 2016. Yet most of the research firms keep 2030 target by adjusting CAGR, which I personally find as an unlikely scenario. Especially with the recent inflation, chip shortage, supply chain and other issues.
Nevertheless, by multiplying forecasted global automotive sales to global EV penetration rate we obtain 62.5M EV cars (PHEV + BEV) to be sold in 2030. It is important to understand that this is a bullish estimate rather than the base. First of all, we applied a very rude global level calculation. To be more accurate we need to apply analysis on the regional levels. In particular, auto sales for the rest of the World and China are expected to grow much faster than in the EU region. Therefore, lower EV penetration rate of the former (20% and 40%) relative to the latter (100%) would result in the lower global EV sales by 2030 than we estimated. Second, it is clear by commentary of the experts and the press that the aforementioned phase-out plans are ambitious and can be taken as a stretch targets. Elon in 2020 himself believed that the global BEV market would only be 30M by 2030 (source).
Chapter 3. Tesla's market share.
From EV-Volumes.com, we can take the annual global EV sales for the past years. It's easy to estimate Tesla's market share from this graph:
- 2018: 245K / 2082K = 11.8%
- 2019: 367K / 2276K = 16.2%
- 2020: 500K / 3240K = 15.4%
- 2021: 936K / 6750K = 13.9%
Not to raise an alarm, but it looks like Tesla's market share peaked at 16.2% and already started to decay. Two years is a bit short of a timeframe to make conclusions on the trend. But it is difficult to restrain yourself from making a connection between the loss of Tesla's market share and ramp up of Chinese OEMs, VW (id family), and wide range of PHEV from legacy.
For 2030, in my most bullish view Tesla can at most maintain its 13.9% market share. Take into account the combination of increasing aforementioned competition and almost nonexistent roadmap of Tesla. To elaborate, Tesla has in production four models (two original designs from aesthetics perspective – head and tail lamps, bumpers, interior, etc.) – Model S/X and Model 3/Y. Cybertruck is expected to launch soon, however according to Elon himself, the target for CT is a mere 250K annual production.
Model S/X is already a 10-years old design (except for the front facelift and an interior update). Model 3/Y's original design is 5-years old with no major updates yet. Given the 4-5 year median time between announcements and production of Tesla, we should not expect any new mass production model(s) before 2026. Especially given an already long pipeline of unfinished projects (Cybertruck, Roadster – niche product, Semi, etc.). By that time Model 3/Y would be 9 year old design (comparable to the current state of Model S/X).
We have observed firsthand what such aging without any major updates might mean for the sales. Since 2018 combined sales of Model S/X dropped from 101.5K to 24.4K in 2021 (it was going down consistently for all the previous years as well, so do not attribute overall drop just to a model refresh). It is not difficult to understand why. When someone buys a new car for $100K, that person wants to make sure that people around recognize it as a new car for $100K and not say 10-year old used one for the price of $30K.
So in order for Tesla to keep up the market share it needs to step up its game in introducing new models and doing major updates for existing ones. If people will start considering Model 3/Y to be rather outdated, the demand will fall off the cliff as we have seen with Model S/X. The fall of Model S/X can be attributed to the release of Model 3/Y. But unlike in 2017, there are far more alternatives now to the aging Model 3/Y as well.
Despite all that, let's consider Tesla will sustain its 13.9% market share through 2030. Recall our estimates on EV global sales of 62.5M in 2030 and we obtain 8.7M Tesla cars to be sold in 2030. This is whopping 56.5% lower than in the original bull thesis, and will respectively lead to a TSLA valuation of 1.12T USD in 2030. An annual return of 2.5% (below inflation) if you invest at current prices.
Chapter 4. ASP
Perhaps for Teslanaires throwing $50K at a car is no big deal, but for most people said $50K is actually big money. If Tesla wants to sell 8.7M cars it needs to either (or preferably both) reduce the ASP of existing model lines or introduce cheaper ones. Especially given the aforementioned points on increasing competition, poor roadmap and aging line-up.
8.7M correspond to 7.1% of the total projected car sales in 2030. Only two brands (note, not manufacturers) had comparable market shares in 2020, namely Toyota with 8.5% and Volkswagen with 7.8% respectively (source). It is only logical to assume that the price distribution of Tesla cars should follow that of a Toyota or Volkswagen rather than, for example, Mercedez-Benz (3.1%) or BMW (2.7%). Neither Toyota Motor Corporation nor Volkswagen Group do not break down the sales and revenues by brands. We will take Toyota as an example as it only contains 2 major brands (Toyota and Lexus) in contrast to 5 major brands of Volkswagen (Volkswagen, Audi, Skoda, Seat and Porsche).
According to the latest Toyota Financial report (Q1-Q3 combined) ASP of Toyota car is 3.8M yen or 33K USD, estimated by dividing automotive revenue of 23.3T yen by car deliveries of 6.1M. In reality these 23.3T yen also included financial services, and 6.1M deliveries also include Lexus, but it's a good enough approximation. Under the assumption that Tesla can dictate $5K premium for the same market share, Tesla's 2030 ASP is $49.5K (38K x 1.03^9) or 25% lower than the original bull thesis assumption of $66K.
Deducting these extra 25% results in TSLA valuation by end 2030 of $840B, or -0.7% annual return if you invest today. See the discrepancy between these numbers and 3-10T valuations TSLAnalysts target for as soon as 2025? And they often claim that nothing other than auto sales are included in their models.
Margins.
One topic I will not touch in this post is net margins, as it deserves its own DD. For now we assumed the same margins in all of the cases. In fact, lower ASP (e.g. cheaper models), increasing number of service centers (to keep up with production), etc. would definitely put a pressure on margins. On the other hand Tesla investments in Gigacasting and structural batteries might (or might not) help to increase margins. Drawbacks of the latter two is lower (to none) repairability that would lead to higher warranties costs. As I said, the topic deserves its own DD.
Chapter 5. Share dilution or Twitter polls
When we discuss the share price we should also touch such concept as share dilution. Even if Elon personally says enough and stops diluting shareholders via his out-of-this-universe bonus plans. Note that for the last 5 years alone number of outstanding shares increased from 0.8B to 1.12B (source), and to my understanding that might not yet include non-executed options of Elon (experts please weigh in).
Due to the expected high-growth, i.e. ramp ups of existing factories Gigafactories and introduction of new models, Tesla is unlikely to offer stock buybacks until 2030. And even if we assume that Tesla will not raise any more funds either, share dilution will still take place via employee stock compensations alone.
A good comparison would be Amazon, unlike Microsoft or Apple who offer a lot of buybacks. For the last 7 years Amazon experienced an average share dilution of 1.1% (source). Needless to say this is a bullish target for a company in a more infancy stage such as Tesla. Applying average of 1.1% over the course of 9 years (end of 2030) brings total share dilution to 10.3% (1.011^9).
On top of that, Elon demonstrated that not only he loves to bonus compensations, he is open to sell them, i.e. increase the float. Which is in short to mid term is even more important for a stock price than outstanding shares as it increases the supply on the open market. But in shouldn't play a role in theory for the long term (again, in theory).
The results:
If I would want to invest in Tesla now, such that it returns me in average annually 10% (vs 7% average of SPY) and we apply:
- our estimated target for market cap of 840B USD,
- and take into account bullish 10.3% share dilution,
Tesla should not be valued more than: 840 / 1.1^9 / (1.103) = 323B USD today
Or with the current number of outstanding shares: 323B / 1.123B = 287 USD per share today
For Tesla bulls: before you say it's outrageous, note how this model still results in $TSLA current market cap equivalent of Toyota and way bigger than VW group. And all that due to the high expectations of growth alone. However, expectations of high growth over the long timeframe involves a lot of risks, that we didn't even account for.
Chapter other product lines of Tesla:
As for the other product lines, it's difficult to judge them now as they are in their infancy. Solar installation seems to be dropping since the days of SolarCity (source). Since 2018 solar installations seems to be recovering and the energy storage seems to be increasing (source: latest quarterly report). However, it is clear from the financial statements that both of these businesses lose money already on the gross margin level. In particular, Tesla reported Automotive Gross margins of 29.3% and Total Gross margins of 25.3%.
How a company exactly calculates gross expenses might differ, but losing money on the gross margin is rarely a good sign. It often means that the costs of goods sold already exceeds the selling price. Think of it as Tesla spending $100 to buy solar tiles, another $50 for shipping, and $200 for labor to install it, whereas only sells it for $250 to a customer. On top of that there are operational expenses that include general management and accounting, engineers, marketing, their bonuses, office expenses, etc. that affect Operating margins.
The TAM of storage and solar by 2030 is debatable. It is clear however, that the biggest solar companies in the world (source) have valuations of just few billions. So adding 100s of billions to Tesla's valuation based on Solar business is unreasonable. I bet the same holds for energy installation business.
Chapter Hype: Fraud Self Driving
This one is the closest to my heart. Disclaimer, I work for the top automotive semiconductor company and contribute to automotive sensors for high-level autonomy. And by proxy, I also have some understanding of the post-processing side of things, what Silicon Valley folks refer to as Machine Learning, Sensor Fusion, Behavioral Planning, etc. So I could probably write the whole DD just related to this topic, but instead I will try to keep this chapter simple. No discussions on the strategy, sensor suits, architectures. We will only talk about simple concept – disengagements.
Since Tesla doesn't share any statistics on disengagements of FSD, we can only rely on the videos coming from the OG Tesla shills beta-testers. If you explore the prairies of Youtube you will encounter hundreds, if not thousands, of FSD videos. At first, you would be even impressed. But we fellow investors should not mix emotions with raw numbers.
After your careful research you would realize that (anecdotally) average disengagement rate is about 1 disengagement per 1-5 miles. Elon's statements on Tesla being on the path of marching nines is heavily misleading. If you think emotionally, a car driving all by itself for 1 to 5 mile is an impressive feat. And maybe it is, which is not an achievement of Tesla per se, but the whole industry since the days of Darpa's challenges and even before.
But if we think practically, we realize that 1-5 miles is too short of a distance. In average US driver drove 14000miles in 2019 (source). For the sake of the argument, let's say that not all FSD disengagements would have led to lethal accidents if not taken. Be it 10%… f**k it, say 1%. That is still 1 lethal accident per 100-500 miles. Or 28 – 140 lethal accidents per year. Would you trust a system to drive you or your loved one home, if you know that the system will try to (or successfully) kill you every second week or even day.
If Tesla reduces disengagement rate from there by 100, You still end up with 0.28 – 1.4 dangerous disengagement per year. That's where the big problem starts to appear. Since a car is NOT trying to kill you for 364 days in a year, you start to become complacent and that's where the first accidents will happen. After few lethal accidents people perhaps will become very cautious again.
Fast forward, Tesla reduces disengagement rate by another factor of 100. Now it's one lethal accident in 100-300 years! Tesla so far produced 2.5M cars with FSD take rate of 10%, i.e. 250K wild FSDs out there. And that results still in 830 to 2500 lethal accidents per year due to FSD.
And that is how marching nines looks like. When Tesla will fight against statistics as people will get more and more complacent. But we are long way from this.
Chapter Hype: To be continued…
I could also rant about 4680, Gigacasting, vertical integration. Especially on the last topic I have something to say from semiconductor perspectives (given Tesla's ambitions with FSD chip and DoJo). But all of these topics I might include in some other DD later on.
Chapter History.
A bit of a detour into a history of stock market. I like to compare Tesla to Cisco. Just like Tesla, Cisco was the stonk in 2000. Cisco actually was the World's biggest company by market cap with a valuation of 500B, adjusted to inflation – 800B. But that number makes no justice to what Cisco was. In 2000 the World GDP was about 34B vs 84B now (source: statista), SPY was around 150 vs 470 now. So, Cisco price was equivalent to 1.25 to 1.5T of today's dollars.
And yet, market analysts did claim that Cisco still had a lot of room to grow. For instance, this bloomberg article claims Cisco was the safest Net play back then. And another nice fella from Credit Suisse believed Cisco will be valued at 1T in just a few years! 1T of 2000 dollars no less. Does such claims sound familiar? At the time of the article, 37 investment banks rated it buy or strong buy, and NONE sell or even hold! By the way, article was released on 19 March 2000. See how they almost perfectly timed the top?
By looking at CSCO all-time chart you can see how the story ended. In 20 years the price haven't recovered to it's ATH. Add to that how much market has grown, inflation, and you will realize that the real returns are much worse than -28%. Nowadays Cisco is the real solid company with a current valuation of 230B and PE ratio of 20. The problem is it was just too overvalued and too overhyped around 2000. Was Cisco a part of the future back in 2000? Absolutely. But sometimes you need to ask yourself how much that future is worth.
It doesn't really matter whether Tesla is 1-5-10 years ahead of competition. What matters is how much that lead actually worth?
Conclusion
My conclusion results that the bullish target for TSLA is 287USD. I am not a financial advisor so only you yourself are responsible for you financial decisions.
P.S. Fun fact, $TSLA is valued at approx. $890B / 2.5M = $356K per every car Tesla ever sold (it was $480K per car as late as January 3). When Hertz “announced” 100K order from Tesla, $TSLA jumped around $400K per every car. This creates an interesting philosophical question: didn't we just discover perpetuum mobile? You can buy a Tesla car from a wealth generated by $TSLA which in fact would increase the value of former even more. Could it be that all Tesla buyers are former or current $TSLA holders? khm….
Leave a Reply