One of the things that amazes me currently in this market is that, it feels like we are trying to justify every new coming hot tech/growth company as if they are the next AMZN in their field. I think one of the delusions people are having lately is that they look at what people projected for AMZN 5-10-15 years ago and compare that to what actually happened.
While it is true that AMZN exceeded any sort of expectation by a huge margin (and did this consistently, year over year) it feels to me that we are taking perhaps one of the most unique success stories ever and then extrapolate that to others, thinking that as if those sort of growth and expectations are in fact more common and easy to achieve than not, regardless of the industry the subject company is operating in.
I could discuss this for a number of stocks, I wanted to open discussion for TSLA. As we all know, TSLA stock went up almost 8 fold in the last year. Today, it is common occurrence in the market that analysts are justifying these price levels and state why TSLA is just more than a car company. This argument obviously has its own merits, but aren’t we little ahead ourselves and already attribute from today, all those things that could potentially be achieved and bake it into the price?
If we went back in in time, we could probably say AMZN was more than just a book company, but could we really say that they would have almost monopolistic power in most of the areas they operated in? Could we really say back in 2008 that AMZN would emerge the market leader in something like for instance Cloud? Or, did these things happen and then we realized how great AMZN became, after the fact. It feels like this was much easier to do in hindsight. But what if AMZN is truly that one outlier, anomaly, in the history of stock markets in terms of sustained growth and market power?
Isn’t this what we are trying to do with TSLA here? I do think Elon is one of the greatest minds of our time and can achieve far beyond any other has achieved so I am not comparing Bezos or any other with Elon here. But what I am trying to say is that, just thinking TSLA as a company, aren’t we already saying today that they already have monopolistic power in EV cars, autonomous driving, solar-energy trade, underground tunnels, robotaxis, etc. – you name it, as if they are already there and waiting to be switched on? It’s as if we have already attributed everything we could imagine could happen and already announced TSLA as the winner.
Now this is the theory side of things. How about financial side of things?
I don’t want to approach this from a car sales point of view, but I will make certain comments.
But what if we applied AMZN’s growth trajectory into TSLA, without focusing too much on car sales, and just see what growth and valuation that would lead us? If TSLA is just more than a car company, what other example could serve best in terms of growth, thinking from an array of products and technological advantage perspective?
So I did a quick and dirty DCF calculation. In there, my starting point is the first year when AMZN hit $24 billion in revenue, which is 2009. TSLA achieved it’s $24 billion in 2019, so from there onwards I am applying AMZN’s growth trajectory for the following 10-11 years. Then, growth is gradually reduced in going into the terminal year. What this means is TSLA will be roughly $900 billion revenue generating company by year 20 (2040). From a car sales perspective, I can’t really project how many cars it could sell in 20 years, but lets say it is 10 million cars. (I think GS said 15 million). Toyota in 2019 sold around this number I believe so with all the competition that will happen in this space, I think this is by no means and under-estimation. In 2019, TSLA sold roughly 367,500 cars and for a total revenue of $19.2 billion (excluding regulatory credits). On average this equates to $52k per car. In 20 years, car prices will be higher than what they are today obviously, but there will be competition and economies of scale to bring the price down. This one is hard to assume but let’s say it grows 1.0% – 1.5% per year with long term inflation, again not necessarily higher because of these offsetting factors. If they do indeed sell 10+ million cars, this is roughly $600-650 billion from cars.
This leaves you with another $200-250 billion or so for other things, whatever you want to attribute this for – AI, robotaxies, insurance, and so on. Now this figure is a future value, so from today’s perspective (using a discount rate of 8% used in my DCF below) that’s roughly $60 billion (revenue). This is again a huge number, as it is almost another Facebook, 2x Coca-Cola etc., or half of GM/Toyota type business from revenue generating capability, again, from today’s perspective – none of which has already happened.
Once again I want to point out that my revenue projections are not driven by bottoms-up car sales, rather, I am using simply AMZN trajectory in terms of growth, then try to attribute the resulting final revenue figure into parts, obviously majority being from car sales, and the rest from whatever you can think of. (Simply because we don’t know what AI/Robos/Insurance/Transportation businesses can generate from revenue/profitability/market share perspective and more importantly what the competition in those areas can be). If you think TSLA can grow even more than AMZN, then I just want to remind you that you won’t easily justify 40-50% revenue growth rates once you’re in the 50+ billion zone, which is the point of the argument. AMZN, still achieved those growth rates when it had 100+ or 200+ billion, which is what we are already assuming here.
So in the stock price we have baked in, from a revenue perspective, a car business that will sell at least 10+ million cars (this number can obviously be higher but even 10 million will be big market share), and other businesses that are worth, today, more than most companies you can think of.
Now this is the growth side of things. How about margins?
We know TSLA generated a last twelve months EBITDA margin of 14% as of Q3 2020. While margins from car manufacturing is certainly improving, most of this margin is actually attributed to sales of regulatory credits, which are by definition almost 100% profit. And so in reality, the margins are much lower, maybe around low single digits between 1-5%. Now this is important because even with all this revenue growth (and if you disagree with me and think it will even be higher), you still need to achieve big margins on your business. I mean really really good margins, we are talking about 15-25% each year. Interestingly, AMZN, while growing its top line crazily over the years and becoming a behemoth as it is today, always lacked in terms of their margins. That goes without saying they are a tech company that typically exhibit great returns on capital. So the point is they had to sacrifice their margins to be able to achieve this growth and this is an important point.
In my DCF, I am extremely generous. I kept TSLA’s EBITDA margins at 15% for the next 3 years, bumped to 17.5%, 20.0%, then to finally 25.0% every 3 years in going forward. Once again these regulatory credits will not be there forever so those margins are extremely optimistic at this point. They might indeed achieve 25% margins in 10 years or so but it won’t be this smooth.
With these two major factors, other items in the DCF are not super important. The discount rate is 8.0%, which I believe is reasonable, NWC and Capex assumptions are in line with what they are today and I don’t think they would necessarily go down (if not go up). I just want to mention that you may completely disagree with some of the assumptions I outlined above. The point however is no matter what sort of DCF you come up with, you will end up either with lower or slight higher value for TSLA, using very, very aggressive assumptions for 10+ years. So the point and question is around the margin of safety here.
It seems I can’t directly post pictures so here is the link for the DCF:
You might wonder, well your price is even higher than what it is today so why did you write this.
Once again my whole point in the argument was that the margin of error in TSLA’s stock price right now is non-existent, because we’ve already treated it the next AMZN, gave the best margins possible, attributed big bucks on its other stuff. I think my DCF is still in the unreasonable side, but I wanted to point this out that even a slight change in trajectory would mean a totally different path and stock price for TSLA.
So the question is, are we ahead of ourselves with TSLA? This could also apply to others as well, this so called growth companies. It is as if we are pricing everything today and think that they have already achieved incredible market share in everything they touched and it’s only a matter of time that these revenue/profits are realized. Are we taking one anomaly and applying this rationale for every other great company under the name of “it’s tech”? Elon is great and TSLA will be great. But the business model may not just allow them to be the next AMZN, regardless of technology or improvement.
One final comment I will make is related to autonomous driving and data TSLA is collecting. While it is for certain that this data gives TSLA an incredible power in developing this technology, I see couple of issues with this as it relates to future of this thing.
Companies such as Facebook are big not because of their software, but because of the network effects they exhibit. And because of those network effects they become larger and larger and are able to collect more data, which creates an incredibly profitable cycle for them as they can monetize that data right away. Data for TSLA here only matters in developing that technology. Monetization however, is another story.
What I am not entirely sure with all this AI story is that if it needs to continuously get updated for use, then how do regulators/companies determine the older version was safe to use to begin with. What I understand is that AI is something that needs to get constantly updated based on data, challenge here is that you are carrying people’s lives here (as opposed to targeting ads) so there needs to be an infliction point where the full autonomous driving AI software should be at a point where it is certainly 100% safe to use. Because if it is not, then the technology fails until it reaches to that point. If it does reach out to that point, then how difficult is it going to be for others to get to that point or use similar technology if what you need is to get from point A to B safely.
Besides there is the price factor in as well. Let’s say TSLA is the only one who can do this. If autonomous driving is too expensive, then how much will people pay for it? Say if the car is 50k and you have to pay extra 50k on top of that – would you do this? Would this have a cannibalization effect? Or if it is monthly subscription fees, would you go on to pay this forever? How significant will the cost be in 20 years from now when others presumably also start providing some sort of service.
On the other hand, if it is not expensive and only a fraction of the car price, then how are you going to achieve these growth levels with that data and technology to begin with. It is one thing to have a great technology, another to monetize it. They might license it for sure, but then you lose all that power right away. So the question then becomes, will it be a niche product or commodity at the end of the day – either of which will not allow you to have incredible returns.
What are your thoughts?
Disclaimer: This information is only for educational purposes. Do not make any investment decisions based on the information in this article. Do you own due diligence or consult your financial professional before making any investment decision.