Cathie Wood vs. Value Investor: a look at Tesla´s valuation

Everyone has a professor fondly remembered who made us love a field of study we would not have otherwise. Over the past year, I have experienced it again, but exclusively through YouTube! It is professor Aswath Damodaran from NY Stern University, and the field he teaches about is corporate valuation. Corporate valuation can be pretty sterile stuff to study, but he makes it fun. He teaches that corporate valuation is about storytelling and creativity as much as it is about number crunching.

And it seems there are very few things I enjoy more than these 2 things: creativity and data analysis. So I will share more of that newfound passion with you, starting with a thrilling practical exercise: valuing Tesla.

What happens when the valuation dean values Tesla - Is Tesla a value stock?

Aswath recently made its own detailed valuation of Tesla, providing all the source formulas and numbers he used. According to him, the fair value of Tesla would have been 571 USD back in Nov 2021, when the risk-free rate was around 1.5%. It is now at 3%, making Tesla’s fair value smaller at 481 USD. I would argue that he was forgiving when inputting Tesla’s cost of capital. He did not consider the stock’s huge Beta, which means greater volatility and should mean greater return expectations for the investors taking such a risk. According to Aswath’s own lecture on Beta’s importance for calculating the cost of capital, I used a much higher cost of capital at 15%, which means Aswath’s valuation story on Tesla would put a fair price of around 306 USD per share, less than half of today’s price.

What story does the professor tell about Tesla?

Enough with the boring part; what stories does the professor tells about Tesla?

With the wind behind its back, Tesla has consolidated its hold on the electric car market and will continue to grow that market at the expense of conventional car makers. As the crisis handicaps its more indebted, slower-moving competitors. Tesla will consolidate its hold on the electric car market and push its revenues to roughly $400 billion (about ten million cars, augmented by revenues from ancillary businesses) in 2032.

Here you have the first part of the story to input the future source of growth. According to industry forecasts, that would represent about 9% of worldwide car sales. The story makes sense if you consider almost every car sold will be electric by then. The biggest carmaker, Toyota, has a 10% market share today. However, Aswath does not tell a story where Tesla will make more than 10% of its revenues from autonomous ride-hailing services and other ancillaries.

It will also deliver higher margins than traditional auto companies in a steady state. It will remain an electric car company at its core. Its superior manufacturing efficiency and current excess capacity will boost its reinvestment efficiency to new records for the next 5 years.

This is the second part of the story, which focuses on profitability. The professor translates it to an operating margin of 16%, in the range of the best manufacturing firms (but far from the Software firms) and a sales-to-capital ratio of 4 to then fall back to its current level of 2.7. This would be close to the bear case in my valuation story as you will see.

Finally, on risk, Aswath used the typical US company’s cost of capital. The stock is almost exclusively financed through equity and has very low debt and a decent amount of cash. I instead added a risk premium to the cost of capital because of the historic high stock volatility compared to the S&P500.

What happens when Cathy Wood’s team values Tesla - Is Tesla a meme stock?

First, we need to recognize ARK analysis has some real substance: they are probably the only investment fund open sourcing their entire valuation model. I applaud such a disruptive approach. I went through their model in detail to understand what part of the story makes their valuation so crazy far away from the professor’s. How far away? ARK’s “base case” puts a $4.600 2026 target price, which means a 57% CAGR from Aswath’s price for the next 5 years, quite a stretch.

ARK’s story about Tesla software business

The first big difference in their story is that Tesla will be a software company at its core instead of an auto company. More than half of its revenues will come from auto sales, but 2/3 of the profits will come from new revenue streams (mostly autonomous and human ride-hailing). Their model uses Monte Carlo simulation, which means the final result of their price target is a statistically weighted average between thousands of scenarios that combines different input variables. Their assumptions are not one-sided, but rather a range of options they think are plausible, with a wider standard deviation if they are less sure about a specific variable. For example, their input on when their autonomous ride-hailing service will launch is a range between 2022 and 2030. But they force the upper and lower bound of the standard deviation to be between 2023 and 2026. Their mean revenue estimate for that business is 216 Billion USD by 2026. They forecast approx. 10% of the cumulative Tesla sold fleet would be added to the robotaxi network within a year of the launch, each traveling 110.000 miles on average, with a platform cut of 50%. These are a lot of assumptions, which is why the Monte Carlo simulation makes sense. Their bear case would actually not see any revenue from this robotaxi business, so why is their bear case valuation still so high?

ARK’s story about the Tesla Car business.

In terms of cars sold, their input is a segment penetration range for each model by 2026, and the range spread is vast. Their mean is 16 million units sold with a std deviation of around 12 million. It means they are concentrating their simulation in a range between 4 and 28 million units sold. The professor models 5 million units sold in 2026, closest to ARK’s worst scenario. The thing is, with the Monte-Carlo method, the average bear case either has some Autonomous ride-hailing business running or a higher volume of cars sold than 4 million units. This is why the average bear case for ARK has a 49% gross margin, much higher than today’s 29%, and similar to Ferrari’s gross margin. I wouldn’t call that a “pessimistic scenario”.

ARK ultimately ends up using multiples to value Tesla

There is a lot of research that backs ARK’s model to forecast TSLA fundamentals. But when translating it to a share price, they use multiples. Tough it is not the source of their overwhelming optimism about the stock, as they use a pretty conservative multiple for their car business (13 for the car business and 19 for the robotaxi business). Maybe I would be slightly more conservative on the robotaxi multiple, but it’s no-nonsense. Instead, the professor uses a discounted cash flow model using an estimated cost of capital based on the risk-free market rate and the Equity market risk premium. Anyway, both methods are approximative.

Conclusion

Valuation is fun but a very personal exercise. The more you go deep into others’ valuation models, the more you realize why you should do your own valuation before trusting others. Professor Aswath’s template is an excellent start to play with your own story. Here is the bear and bull story I built for myself.

Tesla bear case

I gave Tesla a 10% global market share by 2032, close to the Professor´s estimate. In this story, the business is still centered on car sales, not Sofware. I maintained its operating margin similar to its already high level compared to the industry (17%). But I also penalized its cost of capital for its high volatility relative to the S&P500, making equity investors demand more investment yield. With the professor discounted cash flow model, I get to an intrinsic value of $460. Link to the model

Tesla bull case

In this story, Tesla makes 1/3 of its revenues from a hypothetical autonomous ride-hailing service 10 years from now, not far from Cathy Wood’s forecast, but further in time. It represents a $320 billion-a-year business. I calculated this assuming that

• 20% of the most recent 7-year sold fleet would be offered in the robotaxi service,
• The average cost per mile would be $0.6,
• Each car travels an average of 20 thousand miles a year in it.
• Tesla’s platform cut is 35%, with an operating margin of 28%, thus accounting for about half of Tesla´s profits.
• The cost of capital for Tesla is average compared to the market, at 8.5%.

Instead of running a Monte Carlo simulation and using EV/EBITDA multiples to translate it into a stock price, I used the simpler DCF model used by the professor. It gets me to an intrinsic value of $1.200, close to ARK´s bear case 2026 target price brought to 2021 present value with an 8.5% discount rate. Link to the model

Is Tesla a good investment right now in 2022?

There is a massive spread between my bull and bear case, and the average would put Tesla's fair price at $870. It means Tesla would be about 25% undervalued, as I am writing this. 25% is a decent safety margin when discussing more traditional, less volatile businesses than Tesla. But TSLA is not that kind of business and has a beta of over 2. I would expect at least twice the usual 20% safety margin. I know this would mean penalizing Tesla twice for its high volatility (once in the cost of capital of the bear case, once in the expected safety margin), so I am adjusting the cost of capital to 8.7% in the bear case to get an intrinsic value of $720. The new average case would then be $1.000: we have our 40% safety margin! So I do think Tesla looks good for high-risk-tolerant investors at $700. Keep in mind that even the bear case story is all about future growth, every misstep means huge volatility in the stock price. Ultimately, it also depends on how comfortable you are with Elon Musk´s style. In my case, I'm becoming increasingly uncomfortable with him since Twitter´s takeover news. I am for now staying on the sideline as I have too much volatility already built in my portfolio. Next on my valuation list: $NVDA (NVIDIA Corporation), stay tuned!