Tesla Q1 Results: A Shady Set of Numbers

Tesla booked deferred FSD revenues but didn't disclose how much. Stripping out the estimated amount leaves real GAAP EPS at $0.12 or 65% below reported EPS

Deferred FSD Revenue Recognition May Have Boosted Q1 Profits Significantly

Tesla’s Q1 revenues missed by 2.3% and non-GAAP EPS was 8.2% below estimates (see Figure 3). This was not as bad as had been feared and the stock ripped upwards by 13.3% in after-hours trading.

Despite continued price cuts, a one-month shutdown of production at their German factory, and 20% lower deliveries than in Q4 2023, Tesla somehow managed to eke out roughly flat Automotive gross margins of 18.5% in Q1 versus 18.9% in Q4 2023. Imagine having flat gross margins despite 20% lower volumes and roughly 5% lower prices.

This sounded alarm bells in my head and it turns out to be the biggest takeaway from the Q1 results: deferred FSD revenue recognition must have been huge. While Tesla noted in their Shareholder Deck that they recognized deferred FSD revenues in Q1, they did not provide a number and none of the analysts on the earnings call asked about it.

A few facts about Tesla’s FSD deferred revenue recognition:

• Deferred FSD revenue recognition is only disclosed in Tesla’s 10-Qs and they are booked within Automotive Revenues. This means we won’t know the amount booked in Q1 until next week when the 10-Q is released.

• Contribution to profits is substantial as these revenues have nearly 100% profit margins. They also boost average revenue/unit calculations unless stripped out.

• Only the full-year numbers are audited.

• Deferred revenues are a non-cash item so do not impact cash flows.

• Tesla clearly uses these deferred revenues to prop up profits during weak quarters. Figure 1 shows that the highest bookings (highlighted in red) were made during the weakest quarters of the past 3 years.

Figure 1: Tesla Taps Deferred FSD Revenues During Weak Quarters

Source: Company data

The clearest sign that Tesla booked a large amount of deferred FSD revenues in Q1 is the 1.5% QoQ rise in average unit prices versus Q4 2023. Tesla cut prices in China by around 3% and rampant inventory discounts in the US and the EU likely led to a 5% decline in those regions.

Assuming that Tesla should’ve seen average prices drop by 3% QoQ in Q1 (excluding deferred FSD revenue bookings), I can extrapolate $700 million of recognition in my model. While this is highly speculative at the moment, it is worth noting as there is no way Tesla’s car prices rose by 1.5% QoQ. Also, note that deferred revenues on the balance sheet within current assets were slightly up QoQ to $3.0 billion rather than declining. Non-GAAP EPS would’ve been $0.27 or 40% below Tesla’s reported $0.34. And this would’ve undershot consensus by 67% rather than the 8% miss of reported numbers.

Figure 1: Tesla’s Q1 Results Minus an Estimated $700m of Deferred FSD Revenue Recognition

Source: Company data & MH estimates

Tesla Scraps the Model 2 & Pumps Unidentified “New Models”

Tesla all but admitted to scrapping the Model 2—a compact EV that would be priced at $25,000 with 50% lower production costs—as Reuters reported on April 5th. That day, Musk tweeted, “Reuters lies (again)” without saying how or why. This was the key reason behind the 17% slide in Tesla’s stock since the Reuters article came out through last Friday’s close. It was the most important topic of today’s earnings call given how Tesla’s old model lineup has led to price cuts and volume declines that belie its “growth stock” valuations. Nevertheless, Musk gave very few details about its new model strategy. These are the main points about Tesla’s extremely ambiguous guidance on new products:

• Tesla says that it updated its future vehicle lineup “to accelerate the launch of new models ahead of our previously communicated start of production in the 2H of 2025”. This is extremely vague because Musk originally said in January that the Model 2 would be out in the 2H of 2025. This statement says otherwise.

• The new products that Tesla is referring to are also a mystery. Tesla merely stated, “These new vehicles, including more affordable models, will utilize aspects of the next generation platform as well as aspects of our current platforms, and will be able to be produced on the same manufacturing lines as our current vehicle line-up.” This sounds like a “work in progress” mandated by Musk in recent weeks. Expect Tesla to miss the deadline as usual.

• When the analyst from Bernstein asked Musk for some color about these new models—as it sounded like Tesla would simply be tweaking the Models 3 & Y instead of launching new products—Musk curtly replied, “We have said all we want to say on that”. Musk had in fact said nothing about the new models during the earnings call.

• Tesla did admit that their next-generation platform—which was planned to have a revolutionary “unboxed production system” that would lead to 50% cost reductions—will not have the targeted amount of cost cuts (see text below from Tesla’s Shareholder Deck). Given that the Model 2 was intended to sell at a $25,000 price tag, this line below is the closest thing we have to an admission by Tesla that the Model 2 was scrapped. And this makes sense given the fact that (a) even Toyota struggled for years to cut costs by 20% with their TNGA platform; and (b) there is simply too much competition from Chinese EV makers who are mass-producing EVs priced as low as $10,000.

Inventory Climbs to a Record High—Q2 Price Cuts Could Be Deep

Tesla’s Q1 inventory rose by $2.4 billion (+18% QoQ) versus Q4 2023 and was likely the biggest factor behind free cash flow (FCF) burn of $2.5 billion during the quarter.

In the footnote where Tesla shows days’ inventory, Tesla has an arcane formula for counting inventory: “Days of supply is calculated by dividing new vehicle ending inventory by the relevant quarter’s deliveries and using 75 trading days.” Based on this, Tesla’s Q1 inventory rose from 15 days’ supply in Q4 2023 to 28 days in Q1.

This implies Q1 vehicle inventory of 144,409 units (see Figure 2 below), which amounts to a whopping 33% of total Q1 production. And while Q2 is a seasonally higher quarter for sales, having this much inventory on hand will likely lead to deep price cuts, the first of which came worldwide over the weekend.

Figure 2: Tesla’s Q1 Inventory Was 33% of Q1 Output

Source: Company data

Other notable items from Q1 results include the following:

• Tesla’s tax rate in Q1 soared to 26.3% versus 9.3% in Q1 2023. This is due to an increase in the tax rate in China from 15% last year to 25% this year per the Grant Contract with China, as well as paying normal taxes in the US. Based on these two factors, I had estimated a 24% blended tax rate, but Q1 seems to be much higher for some reason. More details in the 10-Q.

• Tesla continued to finance for a third quarter in a row during Q1, although proceeds were only $216 million versus $952 million in Q4 2023. While there aren’t any details so far, Tesla did finance roughly $3.9 billion in the 2H of 2023 via asset-backed notes linked to leased vehicles and finance receivables.

• SG&A rose by 6% QoQ, which is surprising because I had expected Tesla to have sold its Bitcoin holdings (10,000 coins) given the tough quarter and the need for cash. Because Tesla books its Bitcoin gains or losses in SG&A, I had expected a $235 million gain due to Bitcoin having risen to the $60,000 handle in March. This may not have been the case, but we’ll know when the 10-Q drops.

• Q1 capex of $2.5 billion included $1 billion of investment in H100 GPUs. Operating cash flow barely broke even in Q1, so it remains to be seen how much more Tesla can spend to bolster its computing power.

A Meticulously Orchestrated Stock Pump

There was a lot of sloppiness in this earnings report (mistaken EPS numbers in the Shareholder Deck, new stock-based compensation disclosures, etc.), as well as a new product story that made little sense and felt like it was made up on the fly. Musk even postponed a trip to India, including a planned meeting with Prime Minister Narendra Modi, due to “very heavy obligations” at the Tesla.

However, what was impressive was the concentrated effort to tick all the right boxes in order to alleviate fears surrounding Tesla’s new product strategy and pump the stock. Below are some of the key “pumps” from the earnings call to give you an idea of how shameless Musk and his team are. And it reeks of an incoming equity raise after the 10-Q is released:

• Musk says deliveries will grow YoY in 2024: Musk said, “I think we’ll have higher sales this year than last year.” This is an astounding proclamation given that April sales in China are down over 50% YoY and the EU is down 20% YoY. Having cut headcount by nearly 20% last week, it is clear that Musk knows that demand has fallen off a cliff. Note that Musk did something similar on the Q4 2022 earnings call in January 2023 with regard to Tesla’s unprecedented 20% price cuts a few weeks before: in his prepared statement, Musk said, “We currently are seeing orders at almost twice the rate of production”. This was either a lie or the peak of orders before they fizzled out, as deliveries only grew 4% QoQ despite the 20% price cuts. I lean towards the latter having heard a similar tone today.

• Optimus on sale from late 2025: Musk said that Tesla’s humanoid bot, Optimus, is doing simple factory tasks now and will be working in the factory by year’s end. Sales may start by the end of next year.

• FCF to return to positive in Q2: The CFO sounded like he was reading off of a script the entire evening. This was especially so when he said that FCF would be positive again in Q2 after having a “supply-demand mismatch” in Q1. This points to production cuts in Q2 which would massively depress profits and operating cash flow.

• Underlying costs/unit are down: The CFO also said that “profit per unit was slightly better QoQ if the Cybertruck start-up cots & Model 3 Highland production snags were stripped out.

• “The future is extremely bright”, says the CFO near the end of the call.

Final Thoughts: Relief Rally Could be Strong but Short-Lived

There was an uptick in volume today ahead of the earnings report which implies that investors were betting that the “worst is over” or “all the bad news is baked in”. It’s strange to see volumes so elevated on the day of Tesla’s scheduled earnings report.

There will undoubtedly be call-buying to amplify the gamma squeeze today which will provide the “liquidity exit” sought after by large institutions. Sales in China and Europe are down over 50% and 20%, respectively, while Tesla has 144,409 vehicles in inventory from the start of Q2 (this is equal to 33% of Q2 2023 deliveries).

I have no doubt that Q2 earnings will be worse than Q1 (currently, I’m forecasting non-GAAP EPS of $0.30 versus $0.47 in Q1) and that FCF should continue to be negative (my estimate is $2 billion of FCF burn).

While Musk is desperately trying to spin Tesla as an “AI/robotics stock”, the market is becoming increasingly suspicious. How can Musk build his robotaxis and robots at Tesla’s Austin headquarters if the car business burns all of Tesla’s cash to keep the lights on?

Figures 4 and 5 show just how overvalued Tesla is as both a carmaker and as an “AI/robotics” stock.

Figure 3: Tesla’s Q1 Results vs Consensus & MH Estimates

Note: Q1 2024 consensus was compiled by Tesla IR

Figure 4: Tesla is the Most Overvalued Auto Stock

Figure 3: Tesla is the Most Overvalued “AI” Stock

Note: None of the above is investment advice.