Diary of a Tesla Short Seller (Sept 15, 2023)

Morgan Stanley pumps Tesla AI Tech that can't be "validated" yet as Tesla's fundamentals weaken

The Tesla Pump of the Year Just Happened this Week

Tesla is up 11% so far this week after Morgan Stanley put out a 66-page report last Sunday on how Tesla’s Dojo AI chip for vision processing could add $500 billion to Tesla’s enterprise value. In short, the 7-member team that produced the report, says that Dojo will not only enable advanced autonomous driving, but also a lead to revenue streams from robotics, aviation/air mobility, railways and even healthcare.

This should lead to $218 billion of EBITDA, which the team estimates will account for 62% of Tesla’s “group EBITDA” of $352 billion—by 2040 (!). I’ve always found that when you have to project earnings 17 years out from now to justify your investment thesis, you’re probably grasping at straws. Also, there is so much “ass covering” in this pie-in-the-sky report that it’s worthy of display (see Figures 1 and 2).

Figure-1: Caveat About Possible Vaporware

Figure-2: Section of the Report Designed by MS’s Legal Team

It should be noted that there is absolutely zero mention of “Dojo” in any of Tesla’s financial statements so far. It was first officially mentioned at AI Day in August 2021 and then again on Tesla’s second AI Day in September 2022 (strangely, there was no mention of it at Investor Day in March).

On the Q2 earnings call, Musk said that Dojo would need around $1 billion of capex/R&D in 2024, which is quite chunky for a “science project” being conducted while demand for your old cars declines.

The Morgan Stanley report essentially engraved the Dojo story into the psyche of the Tesla fan base. It will now be something to dream about as earnings deteriorate on lower prices and likely lower volumes from 2024 onward. It will be an excuse to buy the dip whenever Tesla sells off on increasingly bad numbers and keep Tesla trading at perverted valuations.

And the biggest question is: why now? A Tesla pump of this magnitude reeks of trying to lay up the share price ahead of an equity offering or Musk trying to offload more Tesla shares to fund Twitter or SpaceX (Twitter’s next $325 million of quarterly interest expense is due next month).

Finally, Dojo might actually be useless if the US transportation authorities order a recall of Tesla’s Autopilot/FSD (there are currently four government agencies, including the Justice Department, investigating FSD). If Autopilot/FSD is indeed recalled—the investigation is now 2 months overdue—Morgan Stanley will have officially beclowned themselves again (their estimates include large licensing revenues to third parties).

Meanwhile, Tesla’s Fundamentals are Weak

• Weak China Sales: Weekly insured vehicles in China were bad, at only 11,729 units (-23% YoY, despite 32% higher capacity than this time last year). This was mainly due to no more Model 3 supply as Tesla prepares to launched its refreshed Model 3 “Highland” next month.

• Output at Giga Texas has slowed down dramatically: Production in recent weeks has dropped to levels around 30% below weekly output rates in Q2. Musk said that production would be down in Q3 due to “global factory upgrades”, but at this rate, deliveries may also be down versus Q2 (Tesla IR was telling European investors as much last week). This is alarming, as Tesla was sitting on 100,000 units of inventory entering Q3. If deliveries are once again lower than output in Q3 (doubtful, but can’t be ruled out), we might see negative free cash flow for the first time since Q1 2020.

• Cheapest Model Y variant was removed from Tesla’s US website: This happened yesterday. It was being produced at Giga Texas and outfitted with Tesla’s in-house 4680 battery cells. Tesla began mentioning the 4680s as a cost headwind since Q3 2022 and continues to do so each quarter since then. The fans are saying the lower-end variant was removed because Tesla is preparing to launch the Cybertruck, which needs the 4680 cells for its battery pack. But this begs the question of what happens to the low-end Model Y at Giga Texas—which is only operating at 37% of capacity now—once the Cybertruck actually does go on sale. By my estimates, Tesla has enough in-house 4680 capacity to make 6,000 Cybertrucks per week (assuming a 160 kWh battery pack). The question is whether Tesla sacrifices the low-end Model Y in order to make the Cybertruck, which appears to be the situation. Given that the cheaper Model Y saw greater demand and is much more profitable, allocating its 4680 cells to the Cybertruck—which will be loss-making in its first year due to start-up costs—sounds like a big headwind for Q4 earnings, if this is the situation.

Outlook Still Negative

I see Q3 non-GAAP EPS at $0.64, which is 21% below Street estimates of $0.81. My estimates are largely based on the following 3 assumptions:

1. Q3 deliveries of 446,000 versus the Street’s 463,062

2. Average price estimate is $44,869, which is down 2.2% QoQ and in line with Street estimates. While the weak Chinese yuan (-3% QoQ so far) should lower revenues booked in China, the higher mix of Model Ys versus Model 3s in Q3 should lead to roughly flat pricing in US-dollar terms.

3. COGS/unit of $37,318, which is down 2.0% QoQ. This is a generous assumption in case Tesla was buying lithium in the spot market, which has dropped another 20% since Q2 to new lows not seen since late 2021 in China. Tesla is mostly locked into long-term contracts, but did admit to having benefited from lower prices in Q2, when COGS/unit was -1.1% QoQ.

For the full year, my non-GAAP EPS estimate is $3.02, which is only 10% below the Street’s estimates of $3.36, as there should be a slight bump from sales of the refreshed Model 3. These estimates are based on deliveries of 1,777,929, which is 3.5% below consensus estimates of 1,841,680.

While the Street sees a 43% recovery in non-GAAP EPS to $4.79 in 2024, I currently forecast a 20% decline to $2.41, which feels a bit too high given the continued price cuts at Tesla and softening auto demand globally.

As always, none of the above is investment advice. It’s simply an excerpt from my diary.