Near-future quarterly financial projections

[FrankSG]

Q2 Estimate: $69m GAAP Profit
View attachment 566202

Assumptions: Q2 vs Q1
View attachment 566203

Free Cash Flow Estimate: $975m
View attachment 566204

It looks like we are quite close on almost everything, except for one thing: FSD take rate.

It looks like you only counted an additional $5M in total for an uptake in FSD sales? At $7,000 a piece, that'd be less than a thousand extra FSD packages sold in Q2 compared to Q1?

If I recall correctly, @Troy estimated FSD take rate of 50% this quarter compared to the usual ~20%?

 

[The Accountant]

It looks like we are quite close on almost everything, except for one thing: FSD take rate.

It looks like you only counted an additional $5M in total for an uptake in FSD sales? At $7,000 a piece, that'd be less than a thousand extra FSD packages sold in Q2 compared to Q1?

If I recall correctly, @Troy estimated FSD take rate of 50% this quarter compared to the usual ~20%?

yes - there are 2 upsides I did not take in Q2.
1. - The FSD uptake you mention.
2. MiC margins. I improved my Shanghai margins to 23.5% (from 17%) but there are claims from some on twitter that margins are in the 35% - 40% range. I want to see the margin improvement in Q2 before putting that into my model.

 

[UCF3]

Q2 Estimate: $69m GAAP Profit
View attachment 566202

Assumptions: Q2 vs Q1
View attachment 566203

Free Cash Flow Estimate: $975m
View attachment 566204

Cost of Revenues increases at Fremont when the plant was shut for 40 days? I would expect cost of revenues there to decrease.

 

[Bullish_Bear]

yes - there are 2 upsides I did not take in Q2.
1. - The FSD uptake you mention.
2. MiC margins. I improved my Shanghai margins to 23.5% (from 17%) but there are claims from some on twitter that margins are in the 35% - 40% range. I want to see the margin improvement in Q2 before putting that into my model.

For the MIC margin, the 35-40% estimate was before MIC SR price reduction on 5/2 (17.5% reduction) as long as I remember, so it should be too optimistic now FYI.

 

[The Accountant]

Cost of Revenues increases at Fremont when the plant was shut for 40 days? I would expect cost of revenues there to decrease.

There is special accounting for idle capacity. When the factory was shut down for 40 days, there were still costs being incurred (fixed costs). These would include depreciation, property maintenance, Insurance, security, etc. These costs are not applied to vehicles/inventory produced (as no vehicles were produced in the 40 days). These fixed costs during the 40 days go straight to Cost of Goods and not to Inventory.
A complex topic but there is a downside when a plant is idled.

EDIT: Tesla may take a different accounting approach to the idle capacity. The accounting I outlined above is the most conservative.

 

[kengchang]

It looks like we are quite close on almost everything, except for one thing: FSD take rate.

It looks like you only counted an additional $5M in total for an uptake in FSD sales? At $7,000 a piece, that'd be less than a thousand extra FSD packages sold in Q2 compared to Q1?

If I recall correctly, @Troy estimated FSD take rate of 50% this quarter compared to the usual ~20%?

Why would you have 600 Model Y under lease for Q2? Leasing on the Y is not available until today

 

[smorgasbord]

Q2 Estimate: $69m GAAP Profit
View attachment 566202

Assumptions: Q2 vs Q1
View attachment 566203

Free Cash Flow Estimate: $975m
View attachment 566204

I understand why Model Y increase ASP over Model 3, but doesn't the loss of 1600 Models S&X hurt?

It's also a little scary to see that $60M of the $69M profit is from FSD deferred revenue. Do we know for sure they're going to do that, and at this level?

As always, thanks for all this effort, much appreciated.

 

[The Accountant]

I understand why Model Y increase ASP over Model 3, but doesn't the loss of 1600 Models S&X hurt?

It's also a little scary to see that $60M of the $69M profit is from FSD deferred revenue. Do we know for sure they're going to do that, and at this level?

As always, thanks for all this effort, much appreciated.

You're correct, the reduction in Models S&X sales vs Q1 did reduce the overall Average Selling Price in my model. I did not highlight it in my notes....but it is accounted for.
On the $60m FSD recognition, we do know that Tesla recognized $30m in Q3 2019 for Smart Summon. I am estimating that the Traffic Light and Stop Sign Control will release twice that. But it is only an estiamte. I am sure they will release an amount to revenue....developing a good model for how much they will release is difficult. $60m is based on my discussions with FSD owners on how they valued Smart Summon vs Traffic Light.

 

[JohnnyEnglish]

It seems from this post Future of FSD after "traffic light & stop sign response" release that traffic lights and stop signs have been released in Europe on 20.28.10. If this is confirmed that is some more deferred revenue that can be released in Q3.

Of course we should expect there to be further new FSD features released in Q3 (and hence even more deferred revenue) if Tesla are to hit feature complete by the end of THIS year

 

[kbM3]

It looks like we are quite close on almost everything, except for one thing: FSD take rate.

It looks like you only counted an additional $5M in total for an uptake in FSD sales? At $7,000 a piece, that'd be less than a thousand extra FSD packages sold in Q2 compared to Q1?

If I recall correctly, @Troy estimated FSD take rate of 50% this quarter compared to the usual ~20%?

I don’t understand. Tesla can’t recognize the full $7k for FSD, right?

 

[JohnnyEnglish]

I don’t understand. Tesla can’t recognize the full $7k for FSD, right?

Correct. Tesla will only be able to recognise 100% of FSD revenue when FSD is feature complete.

 

[FrankSG]

Why would you have 600 Model Y under lease for Q2? Leasing on the Y is not available until today

Tesla reported 4% of Model 3/Y as leases. I guess I was lazy and put it like this in my model, because it barely changes anything. I guess for perfection, I would've had to fill in 0% leases for Model Y, and then change the formula for M3 lease # to take 4% of M3 + MY sales.

 

[FrankSG]

I don’t understand. Tesla can’t recognize the full $7k for FSD, right?

Correct. Tesla will only be able to recognise 100% of FSD revenue when FSD is feature complete.

This is a very good point. I calculated it as if they can recognize the full $7k, because that's the way I had to calculate it last time, when I made an FSD software model going out to 2030, at which point they should be able to recognize the full $7k.

Does anyone know what percentage of the $7k they are able to recognize at this point? @The Accountant ?

if they're only able to recognize half, that'd negatively affect my numbers by ~$60M.

 

[The Accountant]

This is a very good point. I calculated it as if they can recognize the full $7k, because that's the way I had to calculate it last time, when I made an FSD software model going out to 2030, at which point they should be able to recognize the full $7k.

Does anyone know what percentage of the $7k they are able to recognize at this point? @The Accountant ?

if they're only able to recognize half, that'd negatively affect my numbers by ~$60M.

I have read that about half gets recognized immediately and the remainder is deferred. As FSD capability has advanced (with Smart Summon and Traffic Light Control) maybe it is more 60% at sale and 40% deferred.

 

[XHabjab]

I have read that about half gets recognized immediately and the remainder is deferred. As FSD capability has advanced (with Smart Summon and Traffic Light Control) maybe it is more 60% at sale and 40% deferred.

... and they'll recognize some small amount from my Model X from 2018 too, correct? I bought mine originally without FSD but bought the $3000 upgrade when they had an end of quarter sale sometime in 2019.

As time goes on, and there are more and more FSD sales in quarters past, I would think that this FSD recognition would get incrementally larger over time.

My point being that @FrankSG 's estimate of full recognition for 'this' quarter's sales might not be far off but would perhaps be pessimistic going forward as there are more and more FSD sales from past quarters brought forward (assuming an ongoing supply of more FSD features that are enabled 'this' quarter). [Whew! Runon sentence but I hope you can follow along...]

 

[The Accountant]

... and they'll recognize some small amount from my Model X from 2018 too, correct? I bought mine originally without FSD but bought the $3000 upgrade when they had an end of quarter sale sometime in 2019.

As time goes on, and there are more and more FSD sales in quarters past, I would think that this FSD recognition would get incrementally larger over time.

My point being that @FrankSG 's estimate of full recognition for 'this' quarter's sales might not be far off but would perhaps be pessimistic going forward as there are more and more FSD sales from past quarters brought forward (assuming an ongoing supply of more FSD features that are enabled 'this' quarter). [Whew! Runon sentence but I hope you can follow along...]

Yes - that's true. However, from what I have observed, the portion of FSD revenue that gets deferred gets recognized to income in chunks over time. For example, in Q3 2019, Tesla took in $30m for Smart Summon. That $30m included some of your Model X 2018 FSD payment I am estimating that in this quarter, they will take in $60m for Traffic Light and Stop Sign Control (again - some of the $60m includes a portion of your payment on your Model X FSD in 2018).

 

[Doggydogworld]

... and they'll recognize some small amount from my Model X from 2018 too, correct? I bought mine originally without FSD but bought the $3000 upgrade when they had an end of quarter sale sometime in 2019.

As time goes on, and there are more and more FSD sales in quarters past, I would think that this FSD recognition would get incrementally larger over time.

My point being that @FrankSG 's estimate of full recognition for 'this' quarter's sales might not be far off but would perhaps be pessimistic going forward as there are more and more FSD sales from past quarters brought forward (assuming an ongoing supply of more FSD features that are enabled 'this' quarter). [Whew! Runon sentence but I hope you can follow along...]

They can't recognize "old FSD" as aggressively because it was defined by the ability to summon from across town (or across the country) instead of by a list of "features" that only need to work 95% of the time.

 

[The Accountant]

They can't recognize "old FSD" as aggressively because it was defined by the ability to summon from across town (or across the country) instead of by a list of "features" that only need to work 95% of the time.

Interesting - I had not considered that wrinkle.

 

[The Accountant]

Deferred Revenue
Since there have been a few posts discussing this, I wanted to provide some numbers from the Q1 10Q as an FYI.

- Deferred Revenue at March 31, 2020 is $1.5B
- This amount relates to payments received but not yet recognized to income for access to the Supercharger network, internet connectivity and FSD features
- In the 10Q, Tesla states that they expect to recognize $0.8B of the $1.5B to revenue in the next 12 months (I assume this is due to FSD being feature complete by year-end).

 

[thx1139]

I still having a hard time thinking that during this quarter with higher deliveries and smaller inventory at the end of the Q that Tesla wouldnt be profitable. That is compared to previous quarter. I would think somewhere in finances that shutdown saved on some expenses. At the bare minimum electricity and things like that. I would think the majority of costs to actually shutdown would have been incurred Q1 since thats when shutdown started. I know that prices were reduced, by I am assuming that costs also reduced.