Why I’m betting on Tesla
People are often puzzled when I say I invest in Tesla. Even people who never invested themselves will quickly dismiss Tesla stock as risky, or even just plain stupid. Like many other bullish investors, I spent quite some time following this company. In this article I’ll explain why “the machine that builds the machine”, software-driven autonomous cars and the energy revolution drove me to invest in this company.
1. Tesla is not a car company
The confusion normally starts when I say that Tesla is not a car company. Sounds strange right? How can they sell cars and not be a car company?
According to their website “Tesla’s mission is to accelerate the world’s transition to sustainable energy”. Their current product line-up does exactly that:
- Model S, X, 3 have much lower carbon dioxide emission than their gas counterparts. Yes, even if you factor in the production of the car.
- Solarpanels and solarroof provide people with a way to generate renewable energy.
- Powerpack and megapack provide a way to store renewable energy for those moments when it can’t be generated.
Looking at the broad picture, Tesla aims to make the entire process of energy generation, storage and usage (transport) sustainable.
It’s true that they currently mostly make cars. However, CEO Elon Musk recently indicated that in time the energy generation and storage business might be just as big for Tesla as their car business.
2. Competition is coming
Ok, so let’s dive into the electric car market. Some skeptics proclaim that Tesla will get crushed by competition from traditional OEM’s. This reminds me of the following by Wired (don’t forget to look at the date).
Yes that’s right, this article was written in 2009. That’s a full 10 years ago. In fact, this article was written three years before the initial Model S was released.
People have been proclaiming the upcoming competition for a long time, and so far it’s been quite underwhelming. For those wondering, the e-Tron was released in 2018 (nine years after the above article). While it is a great addition to Audi’s line-up, it has been far from a “Tesla killer”.
Obviously, Tesla will not end up with 100% of the electric car market. However, so far other electric cars seem to replace gas cars rather than Tesla’s. Apart from a seasonally weak Q1, Tesla car sales growth remains strong.
Note that the sales are lumpy. Most growth occurs when a new model is introduced or when a new production facility comes online. This is exactly what’s about to happen in the next 6–9 months with the Model Y, and Gigafactory 3 in Shanghai.
3. Tesla’s are not other cars
Ok, but why are Tesla’s cars popular? I could boast about the longer range, faster charging, higher safety and supercharger network, but that would fail to paint the bigger picture.
The iPhone from 2007 was not just a feature phone with a larger screen or a better camera. It was a different paradigm altogether. Phones could suddenly be updated with new software and apps.
In the same way, a Tesla is not just an electric car. You can buy it online and the price is fixed. The user-interface is pretty similar to an iPad, and the smartphone app is a meaningful part of the whole experience.
Most important of all, Tesla provides free over the air software updates. This is absolutely revolutionary in the auto industry. Your car actually gets better over time. It is updated regularly just like your smartphone, without the need of leaving your garage.
Things an early Model 3 has now that it didn’t ship with initially:
- Smart-summon
- Sentry mode
- Dashcam
- Dog mode
- Netflix, YouTube
- Several games
- Longer range and more power
4. “The machine that builds the machine”
Tesla CEO Elon Musk has referred to the Gigafactory as the “machine that builds the machine”. Ultimately a car company is not just about designing a car, but also about designing a machine (factory) that can make that car.
Tesla’s Gigafactory is a new approach to this. Ultimately, these huge factories will be supplied with mostly raw materials and produce everything from the individual battery cell to the entire car.
This highly vertically-integrated approach takes time to perfect and Tesla definitely overstepped with trying to automate too much from the outset. However, they are continuously improving and control a huge part of the entire manufacturing process. Tesla even makes parts such as seats themselves, which most other car companies buy from suppliers.
I believe that ultimately this level of control allow them to make better cars at higher margins. Optimising the whole process is generally more efficient than optimising a sub-process. An excellent example of that is Tesla’s superbottle.
5. The future
Ok, so Tesla doesn’t just make electric cars and even its electric cars are not like other cars. What about autonomous driving?
By lunch Tesla has already collected far more data on driving than you as a human ever will in your entire life!
Every time you drive your Tesla car, it will collect driving data and automatically send it back to Tesla. This data is then used to improve the self-driving Autopilot algorithm. Worldwide, there are over half a million Tesla’s on the road that do this.
Let’s say you live to be a hundred years, and you drive for your whole life. Over your entire life, you would collect 100 x 365 = 36.500 driving-days worth of data. That is a mere 1/15th of 500.000 driving-days that Tesla collects every single day. By lunch Tesla has already collected far more data on driving than you as a human ever will in your entire life!
This data is highly diverse, from driving in the U.S. to driving in Thailand. From calm country roads to driving at night on mountain roads with heavy snow. This diversity is essential for making an algorithm that functions in any conditions.
What about the others?
Google’s Waymo, Cruise Automation (now part of GM) and some other companies are also developing self-driving cars. Instead of directly producing self-driving cars, they retrofit existing cars with their own sensors, which is generally more expensive.
While Tesla automatically collects driving data for free, Waymo and Cruise pay expensive engineers to drive around. This process is obviously much more expensive and less scalable. All these competitors combined collect data at about 1% of the rate of Tesla by some estimates, giving Tesla a huge advantage.
Self-driving approach
Waymo and GM rely on a combination of Lidar, camera’s and HD maps. Lidar shoots out lasers to detect the distances to objects in any direction. This data is then compared to an HD map. A HD map is like Google Maps, except that it contains 3D information of anything in the surroundings such as buildings. The vehicle can then determine it’s exact location, such as what lane it is in, and use this information to guide it.
While this technique can get to decent results fast, it has some huge downsides:
- Lidar sensors are very expensive. Though, in 2017, Waymo CEO John Krafcik stated that they could get the industry-standard of $75.000 down by 90%, that’s still $7.500 for a single sensor. A fully sensor-suite will easily set you back tens of thousands of dollars.
- In order to use HD maps, you first need to map every single road where you want to drive. While Google has shown this capability with Google Maps, it still requires huge effort to keep these maps up-to-date.
- When the maps are not up-to-date (e.g. when construction is takings place), this method is far less reliable.
Tesla on the other hand uses a technique much more similar to how we humans drive. It relies mostly on vision from camera’s, and is assisted by ultrasonic sensors and radar.
The downside of this approach is that is requires much more data for the algorithm to work well. As we discussed earlier, this is exactly what Tesla has.
However, despite relying on a much cheaper sensor-suite, this approach is way more scalable down the road. Since it doesn’t require special maps, it will work on any road and in any conditions.
The point here is that while Tesla’s approach might take longer, it is ultimately the better option for autonomy.
But Tesla has only level 2 autonomy
In the five levels of self-driving, Tesla is indeed at level 2 (partial automation) while some others are at level 4 (high automation but human still responsible). However, these levels couldn’t be more useless. I believe there are only three true levels:
- Non self-driving
- Capable of self-driving but not legally allowed to
- Self-driving
As long as a human needs to oversee the car, we can’t have the full benefits of self-driving. While assisted driving might be more pleasant and safe, it does not give us autonomous taxi’s. In this regard, the only thing that matters is how fast a company can go from non self-driving to fully capable of self-driving (on any road, in any condition).
Why autonomy is important
In short, Tesla seems a likely candidate to get to real full self-driving first. Autonomous driving would not only enable your car to work as a taxi while you’re gone, but it also enables you to do something different while driving.
It’s hard to assign an exact monetary value to all these benefits. It’s clear however that an autonomous car with identical hardware could sell for a much higher price than it’s non-autonomous counterpart. A fair portion of this higher price would trickle down to Tesla’s already high 22.8% automotive gross profit margins (as of Q3 2019).
Besides, Tesla plans on having a RoboTaxi fleet. Any Tesla owner can add their car to the fleet when not in use, and receive money for the rides it makes. Tesla would get a cut (~30%) of this revenue.
5. Solar and energy storage
Tesla has recently launched their solarglass roof (roof tiles with integrated solar panels) and Megapack (grid-scale battery). These products are only months old, so it’s hard to estimate their full potential.
However, electric power makes up 5% of the U.S. GDP. With Tesla’s track record of innovation and efficiency, I suspect they can capture a decent share of this huge market.
6. If everyone sees it…
So far we’ve seen that Tesla‘s car business has huge potential and that Tesla does far more than just make cars, but what about investing in it? If the growth potential is huge, then isn’t this already priced in?
Well, most Wall St. analysts covering Tesla are car company analysts. They value Tesla as a regular car company, and to them it must be absolutely ludicrous that Tesla is valued higher than the much larger GM.
Price-sales ratio
The price sales ratio can be calculated by dividing the market cap by the full year revenue. This can be a useful metric for determining how “expensive” a growth company is, since these companies often don’t have significant profits yet.
At $350 a share, Tesla’s price-sales ratio is about 2.5. This is much higher than your average car companies’ 0.3, but much lower than the 5.0–10.0 that is common in tech stocks. This further illustrates that Wall St. is uncertain how to price Tesla.
However, I imagine that software-driven autonomous cars will ultimately have tech-company-like profit margins. If this is the case, then Tesla’s current valuation seems pretty low.
Tesla stock hasn’t gone up in years
Exactly. While the company has been growing tremendously, the stock is still at its mid 2017 level.
Meanwhile, Tesla has been profitable in two of the past four quarters. Their current production run-rate is about 400k cars/year, much lower than the ~10 million than VW group or Toyota produces. We can only expect the profit to go up once Tesla hits the true economies of scale that the rest of the industry enjoys.
7. The jockey and the horse
No one could summarise this better than Chamath Palihapitiya. When you invest in Tesla you invest in Elon Musk. In short, this man is an incredible engineer that has (co-)founded multiple billion dollar companies (PayPal, SpaceX, Tesla).
While this is a reason to invest all by itself, it also attracts great talent. Tesla is positioning itself as a young and energetic company that excites engineering talent and drives innovation.
8. Short shorts
Shorting a company is this weird phenomenon where you sell a stock you don’t own, only to buy it back at a (hopefully) lower price in the future. Shorting is betting on the failure of a company , and Tesla is one of the most shorted stocks in history.
These shorts are a large driver of Tesla FUD (Fear, uncertainty, doubt). After all, a lot of people will make a lot of money if Tesla fails.
Shorts obviously don’t get a vote in company decisions. Besides spreading negative news, there is nothing they can do really. As long as a company can still raise money when needed (which Tesla demonstrated last May), there is little reason to worry about short sellers.
9. Summary
We have looked at Tesla’s potential in the EV market, the autonomous vehicle market, the energy storage market and the solar market. While Wall Street is still unsure on whether to price Tesla as a car company or tech company, I believe we’re looking at huge upside potential from the current $60 billion valuation.
DISCLAIMER: This is purely my opinion and not financial advice. I hold Tesla shares.
Further reads
