To The Moon

Tesla’s ($Tesla Motors(TSLA)$) Q1 earnings release came with few surprises, given that sales of the company’s products were already flagged as having been in decline. Relative to the same period in the previous year, Tesla’s revenue fell 9% with automotive revenue in particular down 13%. Pass-through line items were far more affected.

While operating expenses and capital expenditures rose 37% and 34% respectively, adjusted EBITDA dropped 21% while earnings per share dropped 55%. The company’s damaging price war with other EV makers has been particularly damaging, as evidenced by its gross margin falling 18%.

Given that EVs are no longer niche and Tesla’s current catalogue is facing stiff competition and “brand fatigue”, the doldrums seen are both entirely expected and likely to be sustained past this past quarter. What wasn’t expected was the price surge seen in the stock. At least a part of this was on account of what the company promises to be and do in the future.

New Directions

In its 2006 Master Plan, the company envisioned that a $25,000 dollar car – referred to as the Model 2 – would be made possible by the revenues generated by its higher-end products. Recent reports had already indicated that the company has abandoned this project, which would have been an entirely different platform. Instead, the company announced in its latest earnings its commitment to launching “affordable” versions of its existing catalogue built around the same platforms and existing production lines. However, it is entirely unclear if this is possible as per the company’s own admission in the Q4 2023/FY 2023 earnings call in January, wherein CEO Elon Musk stated that the company was approaching the "natural limit" of cost reductions on its existing lineup. It is entirely possible that feature-reduced (or “stripped down”) versions of the current catalogue might be made available. The “basic versions” of the Model 3 and Model Y – both of which are built on the same platform and retail at the $40,000 range – might be marginally lower in cost but unlikely to bring it close to the “$25,000” goal.

Also mentioned during the earnings call was that setting up additional Gigafactories was not being explored. The sales decline and associated cost trends would have been visible to the company for months now, which brings to question all the hype generated over the course of several months regarding the scouring of locations for new plants in Mexico and India.

The recent earnings call, however, indicate that this limitation isn’t universal. A plant for the Tesla Semi – an electric trick estimated to retail in the $150-180,000 range – continues to be under development in Nevada. While sales of the Cybertruck – which retails at the $81-100,000 range – have been slow, the company continues to ramp up production facilities in Texas. Both the Semi and the Cybertruck attract high prices and can be expected to have high margins. In light of the declining prospects for the revival of the Model 2, it is clear that Tesla is working on becoming a company for the classes and not the masses.

The matter on what kind of company Tesla is also came into a form of reckoning during the call. For almost 18 years now, Tesla was considered a carmaker by most analysts. However, at many moments during the call, Elon Musk stated that Tesla should be considered a company that works in “AI Robotics”, a keyword combination that is of great interest to analysts and investors everywhere but in which the company had so far shown no substantial evidence of leadership in.

The company’s Robotaxi endeavour – heralded to be unveiled (yet again) in August – is seemingly a mark of the company’s expertise in AI at the very least by projecting the ability of a Tesla product becoming a “fleet operator’s choice”. Central to this is advancing the company’s Full Self-Driving (FSD) suite of software solutions to the point wherein it can be feasibly used with substantial autonomy and minimal error.  Like all algorithms, this requires a massive amount of training data and networked computation capacity. Over the past one year, the number of miles driven under FSD has seen a manifold increase, thus accelerating the gathering of training data.

With greater neural network capability integrated into FSD V12, the accumulated billion-plus miles of training data coupled with nearly 40,000 equivalents of Nvidia H100 GPUs should in theory be bringing autonomous usage closer to reality, although nearly every expert projects that truly autonomous driving without significant potential for mishap is still 3-5 years away.

In Conclusion

During the earnings release, the company shared a “preview” of a “Cyber Cab” app, envisioning it as a combination of Airbnb and Uber, wherein some of the vehicles available for use will be robotaxis owned and operated by Tesla whilst others will be owned by end users. Given that regulatory approval – let alone the prospect of operating vehicles fully autonomously – will be hard to achieve and likely to take years, this preview can logically be earmarked as a matter of scant interest. The company also mentioned that the Optimus humanoid robot, which is stated as being used in small numbers and for simple tasks on the plant floor, will likely be made available sometime in 2025. However, how these robots will find currency any time soon – especially given that the problem set required to be solved to make this a generic industrial labour solution is likely several times more computationally complex than autonomous driving – is a massive question mark. Try as the company might in rebranding itself as an AI/robotics solutions provider, the simple fact remains that there are many more companies that have delivered far more substantive solutions in this space in the years that Tesla has been a carmaker. Such a tag cannot be claimed in such a matter-of-fact manner without a sustainable solution. Until then, the company will (and must) continue to be regarded as a carmaker.

Subsequent to the earnings, it has been reported that the Tesla Roadster, estimated to be priced somewhere around the $200,000 price level, would be made available sometime later in 2024. The higher the price, the better the margin. Positive developments in the development of more models and high-quality premium-priced products would give the company’s stock additional tailwinds. The stated acceleration of product development is a matter of interest but until they’re made available, the company faces being in the grip of a long-term market decline – “AI Robotics” or not.

*****************

For broader articles that deep-dives into business and culture in Asia, visit asianomics.substack.com. Recent articles feature a comparison between Indian and Chinese market trends that formed the core of my commentary to Bloomberg, commentary on EV markets hat were featured in CNN and Investing.com, and a summary of all of my talking points that featured in media outlets regarding Nvidia’s earnings and the ongoing US sanctions regime regarding tech exports to China.