The trouble with making Tesla’s economics work at any cost

Drew Dickson is the founder and chief investment officer of Albert Bridge Capital. He is currently short Tesla in his personal account.

Teslas had been getting cheaper long before Elon Musk took to the stage for the automaker’s 2023 investor day in March.

To some, these price cuts were a sign that competitor offerings from traditional manufacturers like Mercedes, Volkswagen or Ford were becoming more viable, or that Tesla products were not moving as quickly as they hoped, or that perhaps consumers’ appetites for the brand’s current unrefreshed models was waning. To others, this was a nimble move that put pressure on potential challengers in the EV space.

At that presentation, Musk told investors that the market for Teslas was extremely elastic. In his words, “even small changes in the price have a very big effect on demand.” This gave his followers hope. Moreover, many of these Tesla investors believed that when coupled with falling costs of materials such as lithium, the increased scale that resulted from higher production would result in no degradation to gross margins.

I disagreed, particularly with the naïve notions espoused by the Muskeratti that Tesla can sell as many cars as it makes, or that lowering prices was part of some grand master plan.

To me, that Tesla chose to avoid so many important value drivers during that investor day was striking. Management instead chose to devote the whole of a four-hour presentation to the planned streamlining of production and reducing of production costs. There was no mention of when to expect a new mass market vehicle (temporarily dubbed the Model 2), nor the Cybertruck (announced in 2019; now delayed until 2024). In fact, Tesla made it clear from the beginning of the presentation that there would be no mention of the impact of the aforementioned price cuts, nor of any delivery metrics, even in the Q&A.

From my perspective, that wasn’t good news. If Tesla were seeing a positive response to the price cuts, why would Musk and his underlings choose not to give the market some production or delivery metrics to support the previous claims? 

A month later, on April 3, we learned why. The elasticity that Musk had claimed to be seeing was either not real, or something happened in March that prevented it from developing. It turned out that Tesla couldn’t sell every car it made.

Despite sequential price cuts of approximately 10-15 per cent, deliveries only grew by 6 per cent sequentially. In fact, Tesla actually overproduced in the quarter, growing inventories. An implied price elasticity of much less than 1.0 was evidence that price changes, whether big or small, had very little effect on demand.

As smarter heads prevailed, some folks started focusing more on gross margin pressure. During its full-year earnings presentation on January 25, Tesla CFO Zach Kirkhorn had stated that clean automotive gross margins would would fall after these price cuts, but stay above 20 per cent. Most of the $TSLA faithful took this guidance as gospel. Some wiser Tesla bulls, however, such as Emmett Peppers and Gary Black, felt that 20 per cent was an aggressive expectation but retained belief in the long-term story.

And sure enough, it wasn’t a pretty picture for gross margins, or even for revenues.

We should remember that even before the discounts were announced, Tesla shareholders expected tremendous unit growth in 2023. The sell-side consensus last November had been for 37 per cent first-quarter group revenue growth. What was delivered instead was 24 per cent growth year on year, which equated to a $2bn revenue miss. Price cuts of approximately 12-15 per cent led automotive revenues to fall 7 per cent.

For investors anticipating 50 per cent annual growth in perpetuity, this was not the kind of figure they wanted to process.

Likewise the gross margin miss. Tesla failed to reduce costs through scale effects or lower input costs so first-quarter automotive gross margins (ex regulatory credits and leases) were 18.3 per cent, down from the 29.7 per cent generated a year before. A reminder that management had guided for “at least 20 per cent” just a month earlier.

Moreover, Tesla received Inflation Reduction Act (IRA) tax credits during the quarter. In a filing with the SEC, Tesla stated that a higher cost per unit during the first quarter was “partially offset by manufacturing credits earned as part of the IRA”. Management didn’t disclose a precise number but if, hypothetically, 17.5 per cent of costs was the “partial” offset then Tesla would receive a $269mn subsidy against production expenses. In this (theoretical but plausible) example, this means automotive gross margins ex all the funnies didn’t merely drop to 18.3 per cent, they may have plummeted to around 16.9 per cent.

Tesla did not reveal anything about the IRA credits optically supporting margins on the conference call.

Decomposing the numbers, I estimate that the profit per vehicle likely dropped from circa $15,000 last year to just above $7,500. In other words, here was a price cut of approximately 12.5 per cent and profits almost halved.

Selling cars is a hard business. Toyota, VW and others will not gleefully cede market share. Tesla has to keep refreshing and expanding its product line just to stand still; and economic realities get in the way of all product roadmaps for all manufacturers, even Tesla.

But Elon — and this was either an act of brilliant diversion or complete absurdity — had this to say to everyone during the first quarter results:

Tesla is in a uniquely strong strategic position because we’re the only ones making cars that technically we could sell for zero profit for now, and then yield actually tremendous economics in the future through autonomy.

This was a shocker of a comment. And while Waymo, Baidu and others might be asking Elon to hold their beers, one wonders if the Tesla CEO is implying that a “full service” full-self-driving vehicle is really, finally, just around the corner. (He has some form in that regard.)

Let’s imagine for a moment that FSD is real this time. Tesla fans may be enthusiastic to be guinea pigs for Level 5 autonomy, but do they really believe the rest of us will want to ride in our own vehicles as passengers, or trust a computer to take our kids to a soccer practice?

 In my view, most of us will be driving our own cars for a very long time. For robo-chauffeurs to truly take over, it will likely require federal government mandates. Regulation is the only way to reduce the novel cases to a point where L5 machines could actually function with acceptable error rates.

But, again, that is my view. And for what it’s worth, in June of 2022 Elon said something that $TSLAQ bears loved to hear, but that the $TSLA congregation claimed was merely a bit of characteristic hyperbole. Some said he really didn’t mean what he said. (Again, he has some form in that regard.)

(Our) overwhelming focus is solving Full Self Driving. So, um, yeah, that’s essential. That’s really the difference between Tesla being worth a lot of money, and being worth basically zero.

 So maybe his more recent comments about selling cars for no profit weren’t flippant? Maybe profiting from FSD down the road is his real plan? Maybe it’s the only way he can imagine — after Tesla shares fell over 60 per cent in 18 months — that the stock doesn’t continue to give back the tremendous gains between 2019 to 2021 that made him the world’s richest person?

My opinion is that fully autonomous roads remain decades away, and that current demand even for “free” autonomous software is lower than Tesla fans believe (let alone the demand for a monthly or annual subscription costing thousands of dollars). I also side with Ars Technica, ABi Research, and Guidehouse Insights on the likelihood that Tesla is not even in the technological lead for autonomous.

And yes, I can be dead wrong, but what I am more certain about is this:

Tesla is a niche automaker that has made impressive strides selling EVs over the last few years. Its success spurred other manufacturers into action. They are starting to produce extremely impressive EVs as well.

EV demand today is limited. About 85-90 per cent of all cars and light truck sales this year will still be ICE, but it is possible that this number falls to 50 per cent by 2030 and even 25 per cent by 2035. Sure, government edicts might be required to force people to buy EVs (or heavily incentivise them, as already happens in Norway), but these market share figures probably won’t be far off.

Moreover, it may be no more than a temporary blip that Tesla’s revenues, gross profits, and EPS are all decelerating or in decline. Maybe interest will somehow be rekindled in their products. Maybe the price cuts will start working. Maybe something else happens that I just haven’t thought about.

And where could I be wrong? Well, I’ll be dead wrong if Tesla discovers Level 5 FSD in the next year and brings it to market first, then governments mandate its adoption, and other vendors are forced to license the software. In that scenario, it is hard to predict what those economics would be worth, and even harder to predict what Mr Market would think they were worth.

Or there is the “Tesla wins automaking” scenario. Logic says selling 10mn Teslas with just a handful of unrefreshed models is extremely unlikely. The 20mn units the company has mentioned (and some Tesla shareholders strongly believe) implies a market share figure that is just about impossible. But logic could be proved wrong, and so might I.

But I know exactly what would change my mind. Global demand for 2.5mn Model Ys; a resurrection of demand to 750,000 each of the Model X, S, and 3; at least 750,000 more Roadsters (which are basically a top-level trim of the 3) and 1.5mn Cybertrucks. Tesla also needs to make over 3mn of the currently theoretical Model 2 (or whatever it ends up calling a mass-market Tesla). That would get the company to 10mn in unit production. It is half what the $TSLA faithful are anticipating, but would be more than enough for me to become hopeful.

The Tesla investment story hangs on high expectations and low-probability outcomes. Until the bears are proven wrong, we should remember that exactly four years ago, Tesla traded with a market cap of $46bn. Today, it is over 12 times higher at $570bn. So if you think about it, one really has to wonder what annual production or autonomous software success the share price is already pricing in?

Further reading:
FT.com/Tesla