If you didn't catch ARK's #BIS2022 summit yesterday, here's some highlights from my latest autonomous technology research:
In our original work, we estimated that an autonomous ride-hail service could be priced as low as $0.25 per mile for consumers, while still leaving room for operator profits.
Now, we estimate that there should be substantial demand at higher price points. Note that today's ride-hail services average $2/mile in western markets, higher than consumer's perceived value of time AND higher than the marginal cost to drive your own car.
Takeaway: the low cost and convenience of autonomous ride-hail should win share of miles, even from today's car owners. Consumers already pay up for convenience. And at its lowest price points, autonomous travel is even cheaper than the marginal cost to drive a personal car.
All in, the addressable market for autonomous ride-hail could be $11 trillion.
If I've bored you so far, here's why you should pay attention to autonomous cars: we think this could be the most impactful innovation in history. Autonomous ride-hail could add roughly $26 trillion to global GDP by 2030.
This accounts for losses from gas powered car sales, fuel, maintenance revenues (electric AVs need less of it), insurance (we estimate AVs over 80% safer), and property repair and medical revenue from the accidents prevented from autonomous.
Gains include service revenues from ride-hail, a productivity uplift from freed up time as a passenger (whether you're emailing or watching TV), a gain from saving some of the 1M+ people that die every year in accidents so they can live happy productive lives, and AV sales.
That's all to say that this is a HUGE opportunity. AVs should save money for consumers, benefit public health, boost economic growth, and create a massive investable market - autonomous platform providers could be worth over $11T in enterprise value in 2030.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Our updated Tesla model is now out! Autonomous driving likely will be one of the most life-changing AI unlocks in history, and quite meaningful for Tesla shareholders: ark-invest.com/articles/valua…
We conservatively assumed that Tesla does not sell Optimus externally in our model, and that Optimus manufacturing savings modestly impact Tesla’s costs in single digit percentages over the next five years.
We still believe that Tesla could leverage its learnings from robotaxis to become a leader in embodied AI, although the financial benefits may extend past our five-year forecast.
1/ Our latest Tesla price target is out! $2,000 per share in 2027. Check it out! ark-invest.com/articles/arks-… Similar to last year, we believe autonomy will represent a significant portion of Tesla’s future enterprise value.
2/ The more Teslas on the road that are autonomy-capable, the better, because each can become a recurring revenue generating robotaxi, a business which we think will generate much more cash flow than electric vehicle sales alone.
3/ As you can see @DMaguireARK and @skorusARK analysis shows that a robotaxi offers the highest return for batteries at Tesla.
1/ Drone delivery is having a moment. In 2015, I estimated that Amazon could deliver a package in less than 30 minutes for <$1. ark-invest.com/articles/analy…
2/ Since then, I've updated that work and estimated the market potential of both meal and parcel delivery. All in all, I think this could be a $1-2T market.
3/ In reality, drones should have many many use cases. I chose these two markets because they might have the largest potential in terms of volume and dollars, but drones could be used for imaging, inspection, and delivery of medical supplies much like @zipline does now.
At Tesla's Investor Day, for the first time we saw the accident rate per million miles driven in FSD. Previously the Tesla Safety Report included only Autopilot statistics. 🧵1/10
1 accident per 3.2M FSD miles. Looks better than the national average, right? However, I'd argue the comparison is not 🍎to🍎 as FSD worked only on surface streets until v11.3. Also, how does this compare to a manually driven Tesla? Let's compare across surface streets only. 2/10
Surface streets have more frequent accident rates than highways. Adjusting the national average and the "Non Autopilot" rate given in Tesla's Safety Report, a Tesla in FSD looks ~5X safer than a Tesla in manual mode 👀‼️3/10
1/ Waymo is on the cusp of making history with a fully commercial robotaxi launch. It must be VERY confident in fully driverless rides to make this push. blog.waymo.com/2020/10/waymo-…
2/ In the past 5 years we've seen a handful of automakers promise a 2020 launch and fail to come through. Today, many are not even close. group1nissan.co.za/blog/nissan-ne…
3/ Waymo even pushed back its commercial timeline, which was originally supposed to be last year. A lot of doubt has been cast on the pace of progress.. nytimes.com/2020/05/12/tec…
1/ Here's why we think Tesla and every other automaker should pursue autonomous driving - the market likely will be worth much more than the auto market today. We estimate autonomous ridehailing will be worth $9T, or more than today's energy sector, in the next ten years.
2/ Why? Because we think autonomous rides will be much cheaper than today's taxis and personal cars. At scale, they could price at $0.25/mile. This is a big deal. The price of personal transport hasn't changed in over 100 years. These economics can drive widespread adoption.
3/ AVs likely will offer more safety and convenience than ridehailing today in what could be a winner takes most market in each geography. Fees could exceed Uber and Lyft's 20-30% cuts and come in at the higher end of the spectrum below: