Thread by @wintonARK

It looks like our 2024 EV unit sales expectations are going to land @ 37mm units.

(Scooping @skorusARK and @TashaARK)

@skorusARK @TashaARK We believe that 200+ mile EVs will be sticker-price cheaper than the median vehicle before 2022

(supported by Wright's Law work analyzing both lithium ion battery cost declines and the Model III's production cost learning rate which align on that conclusion)

@skorusARK @TashaARK Imagine walking into an auto-dealership, on your left is a Toyota Camry, on your right is an EV version of the same that costs less money out of pocket *and* saves you money over time *and* is faster off the line.

Why would you keep buying the Camry?

@skorusARK @TashaARK You may keep buying the Camry because the EV version is "available" but not immediately so--the biggest risk to our forecast is the degree to which traditional OEMs have dragged their feet on committing capital to the new drivetrain-paradigm.

@skorusARK @TashaARK (And keep in mind that by 2024 we believe that something of Model III quality/range could be substantially less expensive than the median ICE car sold, yet we only forecast EVs making up ~40% of sales--there is a lot of friction already designed in!)

@skorusARK @TashaARK It is possible that there is so much institutional friction (/restructing/consolidations and bankruptcies) within the traditional OEM space that they don't sufficiently invest--this would likely allow Tesla's EV business to be more cashflow-generative than it rightfully should.

@skorusARK @TashaARK It is also possible that those OEMs playing from furthest behind, buy drivetrain assemblies (plus sensors and compute assemblies?) from Tesla, allowing Tesla to more capital-efficiently scale production.

@skorusARK @TashaARK Barring Tesla's borrowing against other OEM balance sheets in order to scale (which is how you should think of the Skate sales opportunity), we think that Tesla's maximum possible 2024 production is roughly 7mm units.

Where could the remaining 30mm come from?

@skorusARK @TashaARK China could provide a substantial chunk. The Chinese government has indicated that they anticipate 35mm auto sales by 2025 (we're not quite so bullish on that front tho 😬), and there is clear strategic intent to subsidize their battery suppliers as they scale production.

@skorusARK @TashaARK If China + Tesla (ex China) EV sales in 2024 = ~20mil
Then 17m units have to come from other places.

Some quick math.
At OEM capital efficiencies producing 17m units would require ~$150 billion in Capex. Seems a lot.

@skorusARK @TashaARK BUT the top 10 OEMs ex-China had a collective Capex budget of ~$140 billion in 2018.

It's not that they can't do it, they just have to spend in the right places (and they have to start soon if they are actually going to get on a glide-path to expected 2024 EV production.)

@skorusARK @TashaARK Much of that Capex is devoted to keeping existing factories running; net of depreciation, capital expenditure budgets summed to $40 to $50billion.

3 years of that net Capex would be sufficient to get the OEMs on a manufacturing ramp to hit our 2024 forecast.

@skorusARK @TashaARK Alternatively the OEMs are paying out ~30 billion in dividends.

If they were really taking this transition seriously they could cut those dividends keep existing ICE-related Capex going and hedge their bets against the (what we see as inevitable) EV future.

@skorusARK @TashaARK All of this sounds nice.

So far as we can tell, however, the OEMs don't believe it.
Those that have developed EV product shave had trouble selling them, regardless of whether or not they seemed to spec competitive (Bolt) or not (I-Pace, e-tron).

@skorusARK @TashaARK Where else could the additional 17mm units come from?

It depends upon how you define an electric vehicle.

Right now city-cars/neighborhood electric vehicles are a niche category because owning a separate vehicle just for non-highway trips doesn't make sense for most households.

@skorusARK @TashaARK With robo-taxi capability, however, those sorts of vehicles will make a *lot* of sense and deploy appropriate to the journey that the consumer wants to make.

@skorusARK @TashaARK Production scaling for those vehicles can occur substantially more capital efficiently and could easily fill the gap (left by OEM underinvestment)

@skorusARK @TashaARK In our previous work we have demonstrated that robo-taxis, by winning share of miles onto EV platforms from legacy ICEs could massively shift forward peak oil demand (this forecast from 2016)

ark-invest.com/research/oil-d…

@skorusARK @TashaARK This most recent forecast has brought to light--for me--the degree to which the EV build cycle (not just share of miles) will be sensitive to the delivery of robo-taxi capability.

/fin?

😅