@ChrisBloomstran Happy to receive feedback on the way we’ve modeled the insurance business.
I’ll walk through the mechanics of it in a bit and you can let me know how you would amend.
But first I’d like to level set and clear up a few misperceptions.
@ChrisBloomstran 1)
As described in the blog, the “bear case” as we’ve defined it is the 25th percentile outcome.
It is not the *worst* outcome (which, as with all equities, is bankruptcy.)
The case detailed in the blog is one of many possible ways that outcome could occur.
@ChrisBloomstran Because we have tried to distill the model into independent inputs the 25th percentile outcome is not the downside case for each of those inputs.
The bear case of 3 coinflips is not 3 tails, but 1 heads and 2 tails.
@ChrisBloomstran Similarly a bear case will have some things that go better but more things that go worse.
@ChrisBloomstran 2) You don’t like the way we have modeled the insurance business; very clear!
And perhaps you don’t believe that Tesla enjoys any durable competitive advantage along that line.
That’s why we provided the model!
0 it out; there isn’t a material impact on expected value.
@ChrisBloomstran We think it has strategic value to Tesla and will contribute to customer retention and satisfaction, but as we currently have it modeled it is not a material driver.
But since we’ve caught your attention, and I’d love to understand it better, I’ll walk you through our thinking.
@ChrisBloomstran We believe Tesla will enjoy competitive advantage on vertical integration through insurance along 3 vectors.
1) Advantaged customer acquisition cost (due to their direct to consumer model and the digital interface in their vehicle.)
@ChrisBloomstran 2) Integrated telematics and data, inclusive of interior and exterior vehicle cameras, radar, and driver profiles.
(This should yield better underwriting and could be used to positively influence driver behavior.)
@ChrisBloomstran 3) Vehicles that will get safer over time with software upgrades via autopilot and full-self-driving advances.
The company will have a real-time understanding of these improvements and depending upon pricing-strategy may result in underwritten policies that grow more profitable.
@ChrisBloomstran Note, of those 3 advantages, they are only realizing the first at this point. They are not underwriting as you point out, but instead on-selling customers.
We model them as continuing to operate under that model (albeit on an increasing share of incremental new sales) until 2023
@ChrisBloomstran Mechanically they collect ~12% (varies) of commissions through 2023 at which point we assume that they begin to underwrite.
At that point they begin underwriting their own insurance at a ~70% loss ratio.
@ChrisBloomstran We then assume that they hold that per-mile pricing as the vehicle-safety improves by ~5% (varies) per year.
That vehicle safety improvement flows through to the bottom line in a lower realized loss ratio.
@ChrisBloomstran The number of new vehicle sales they tie to their insurance increases (as they open up more geographies), but we don’t assume that they go back and recapture customers that they didn’t sign on initial sale.
@ChrisBloomstran We also assume that they win insurance against all vertically integrated ride-hail (non robotaxi) miles at a per-mile premium to non-ride-hail vehicles; this becomes an important driver of insurance premiums in the non-robotaxi cases.
@ChrisBloomstran One last addition to the structural advantage discussion:
Tesla’s vertical integration through service should also offer them better cost-control on their underwriting.
@ChrisBloomstran Tldr:
not material to our modeled outcome;
Clear to us that they enjoy structural advantage in this space (contingent upon execution);
and we think we have modeled it appropriately.
Curious to hear your thoughts.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Let's dimension the US robotaxi market (since market participants seem unwilling to do so).
People pattern match against structurally ~$3 per mile point to point mobility products and so misunderstand the potential scope of robotaxi as it becomes mass accessible.
The average US adult spends nearly an hour per day driving. The imputed labor cost of all that manual piloting runs in excess of $4 trillion per year.
In addition we pay $1.6 trillion annually for the actual service of driving point to point.
By giving people back time (for which they don't have to pay full freight) and winning spend share, we think the US market could approach $4 trillion annually at saturation.
Given reasonable expectations of supply diffusion and consumer adoption robotaxi service providers could exceed $1.5 trillion in revenue by 2030 with gross profits in excess of $1 trillion.
Let's work through the underlying derivation.
Constructive criticism welcomed.
The richest income earners spend the most time manually driving, and can command $50 per hour after tax.
Higher earners are willing to pay a higher share of after tax wages to win time back.
Our research suggests that highest income earners would turn down something less than the equivalent of overtime pay in order to win time back. For other cohorts they buy back time at a discount to what they could otherwise take home.
This is a fairly sensitive input to overall market size.
That millenials are so obviously willing to trade time for money by hiring doordash drivers rather than schlepping to the takeout counter themselves provides decent anecdata that there is some truth to this curve.
When a consumer decides to take a robotaxi they are not just trading time for money, they are also avoiding the cost of running their own vehicle.
Top decile earners spend $.76 per mile, inclusive of the cost of purchasing vehicles, on getting from place to place (excluding air travel).
Pretty consistently, by income decile, the marginal cost of mobility runs at ~$.17 per mile.
This model assumes that people that already own vehicles are only willing to pay that $.17 at first, plus the value of their time. Over the typical vehicle life-cycle we assume that consumers avoid new vehicle purchases as they grow increasingly reliant on robotaxi.
2 car households become 1 car households and more of the transportation budget shifts into robotaxi.
(note that the fixed ownership bumpiness across income decile is almost certainly just an artifact of extracting this from a single year's CEX data crossed with a line item--vehicle purchases--that is infrequent but large across households; I clearly should smooth that but it's not particularly material to conclusions.)
The AI-driven collapse in the cost to write software should simultaneously
-cause software demand to explode and
-radically change value chain structure
👇
How we think about demand:
AI promises radical productivity improvements for knowledge workers
At 4.5x productivity improvement against $30 trillion in 2030 knowledge work wages with 10% value capture into the tech stack suggests $13 trillion in 2030 AI software demand
Seems like that should be very good for software vendors.
If SaaS were to maintain its position in the industry it would.
Though CoGs in SaaS flows down into other layers of the software stack, SaaS revenues predominate.
On current industry structure we would expect $6.5 trillion in revenues, which suggests nearly $40 trillion in enterprise value on current multiples.
people ask why google isn't shipping
isn't it obvious?
they are caught in the thorns of the innovator's dilemma
how does google make money?
they charge a toll to deliver a user from the text prompt to another website.
language models take the information in *all* other websites, compress it, and deliver it directly to the user.
No transit required.
And no toll paid.
A tangible example.
I want to buy a vacuum cleaner; what do I do a year ago?
I google: "which is the best vacuum cleaner"
(I'm not satisfied with just any old vacuum--I need the best)
Google serves me up a bunch of cost per click ads which I ignore (though perhaps they get some CPM revenue off the results--and some users click thru: kaCHING)
No, I am seeking to be informed. I scroll down to the reddit page where people vociferously debate the pros/cons of different vacuums. After some skimming, reading and considering, I have selected my vacuum.
Back to google.
A new search. Probably a cost per click to actually consummate the transaction.
What do I do now. Same query, but not in google's interface. In chatGPT. It has scraped all of reddit; it has synthesized the vacuum cleaner debate, not just across that site but across the wirecutter and cnet and probably amazon customer reviews as well.
And within that interface I can refine the things I want to know based upon how it answers. What if I want a vacuum that works dynamite on pet hair? What if I need cordless, or something that won't make too much noise?
GPT-4 will answer all of that, more precisely, completely and quickly than I can manage scanning across a few websites.
And it's pretty easy to say you're way clear to the OpenAI platform facilitating the commercial transaction at the end of the interaction as well.
But cynics wonder: 1) What if robotaxi never happens? 2) What if production scaling is arbitrarily constrained? 3) What if Tesla has to keep pricing aggressively to move units?
🧵
The value of an open source model is that you can reasonably disagree with our assumptions and generate your own expected value.
People ask what happens to our investment case if robotaxi capability never deploys.
The short answer: expected value of $900 per share.
As you can see, conditional upon robotaxi not working, we believe that Tesla will shift its business towards vertical integration through human driven ride hail.
What happens if Tesla doesn’t pivot to that opportunity either?