Tesla’s Incremental Approach Paves the Long Road to Autonomy
In September, Tesla released its latest software update Version 10.0. The most notable new feature, Smart Summon, allows Tesla owners who have purchased Full Self-Driving (FSD) or Enhanced Autopilot to request their vehicles to self-navigate a driveway or parking lot and pick them up. The company sells its Full Self-Driving Capability (FSD) software upgrade for $6,000 per vehicle.
By incrementally recognizing revenue on high margin software as they roll out new Autopilot features, Tesla creates a unique system of taking in money to continuously develop and improve the system that consumers have paid upfront for. A second advantage is their proprietary hardware which is capable of powering more advanced features as Tesla inches closer to full autonomy.
The Advantage of Monetizing Incremental Improvements
Smart Summon is a visible step towards full autonomy, made possible in part by the company’s custom self-driving hardware platform. Tesla’s advantage in AV chip design is largely ignored by investors. During the Autonomy Investor Day in April, Tesla announced the industry’s first custom-designed self-driving hardware platform. Instead of continuing to use NVIDIA’s GPUs in its autopilot hardware platform, Tesla reportedly invested $100 million and four years to design its own proprietary chip in partnership with Samsung. Today’s Tesla models S, X, and 3 are all shipping with the Autopilot 3.0 platform. Musk further revealed that Tesla was building a second-generation silicon that’s 3x more efficient at processing sensors and two years away from release. We believe Tesla’s custom-designed hardware is three to four years ahead of any of its competitors.
This hardware is one reason why Tesla is the only company monetizing incremental improvements in autonomous driving. While OEMs that offer advanced driver assistance systems monetize them through upgraded vehicle options, there is no clear path from ADAS to full autonomy.
At the end of Jun-19, the company had booked $883m in deferred revenue related to FSD. Tesla and its auditors determine the timing of FSD revenue recognition (there are 6 components to recognize all of the deferred revenue, of which 4 have been made). In June the company said it expects to recognize $567m in FSD revenue over the next 12 months, with the $316m balance recognized over the 8-year life of the vehicle.
This incremental revenue is a competitive advantage for Tesla since it can reinvest the free cash flow to fund incremental development of autonomous technology. We view this approach as the most appealing model for the years-long investment cycle to build autonomous vehicles. A second, less appealing, but still viable approach is Waymo’s deep pocket model, which resembles a long R&D project funded with the backing of a highly profitable parent company. Traditional auto manufacturers will be challenged, given they lack either incremental revenue or the deep pockets needed to sustain years of investment.
It Will Likely Take a Few Years for Tesla’s Market Cap to Get Credit for Autonomy
Since Tesla’s April autonomy event, the company’s market cap has been flat at about $47B suggesting investors are reluctant to give it credit for its progress on and long-term vision for autonomy. As a point of reference, private, pre-revenue, autonomy only companies are valued in the $7B-$20B range. In May, Cruise raised at a $19B valuation, and Argo.ai was valued at $7B in July, similar to Uber’s ATG which was most recently valued at $7.3B. Waymo is widely thought to be valued at around $20B.
Investors’ dismissal of Tesla’s autonomy efforts is likely tied to Musk’s history of missing near term targets, eroding his creditability related to the company’s current progress and his autonomy vision. That said, we believe history will repeat itself, and Musk will eventually make good on key parts of his vision. . . and investors will handsomely reward him.