Near-future quarterly financial projections

[bhzmark]

saying some random person/people hated amazon has nothing to do with the conversation

Wha?
Another odd data point on the quality of analysis.

 

[Reality]

Guys,


Full Self Driving is not a variable attribute. Either it is Level 5 or it is not.


"FSD features" is meaningless garble stated to get headlines. IM shocked some of you are as smart as you are and do not realize this.


My microwave has "Deep Learning features" but it' not going to cause the Singularity anytime soon.

 

[luvb2b]

they wouldn't be able to recognize all fsd revenue for partial feature set enablement. it's approximately percentage of features enabled as an added multiplier. so 73.6m x 15% for example.

Take rate on EAP for model 3 is 67.7% and FSD is 9.7% according to Model 3 Invites Spreadsheet Published Web Version.

By end of Q3, assuming the S/X Fleet has similar take rates, with roughly ~170k AP2 S/X vehicles delivered, and 50k+23k+10k (8182 rounded up to include 2017 deliveries), there will be roughly 253k vehicles delivered with the ability to upgrade to FSD. The revenue recognition from a Q3 delivery of some basic FSD would give us roughly

253,000*.097*$3000 = 73.623 million added to revenue with 100% gross margin.

I'm not sure what percentage of EAP users would upgrade to FSD, but I'm assuming almost all of them would at this point.

 

[Reality]

they wouldn't be able to recognize all fsd revenue for partial feature set enablement. it's approximately percentage of features enabled as an added multiplier. so 73.6m x 15% for example.

disagree, unless PwC has the F team on this there is no way that FSD would ever be able to qualify for % of completion treatment.

 

[kbM3]

i agree, but my thesis is that tesla is not thinking about the long term, so i think RD gets hit hard

Good point! Elon never looks far into the future. He’s always so concerned with the next quarter!

 

[gluu]

they wouldn't be able to recognize all fsd revenue for partial feature set enablement. it's approximately percentage of features enabled as an added multiplier. so 73.6m x 15% for example.

I see. How did you determine this? All I could find from https://www.sec.gov/Archives/edgar/data/1318605/000156459018011086/tsla-10q_20180331.htm was this statement

Revenue related to Autopilot and full self-driving features is recognized when functionality is delivered to the customer.

 

[gluu]

Guys,


Full Self Driving is not a variable attribute. Either it is Level 5 or it is not.


"FSD features" is meaningless garble stated to get headlines. IM shocked some of you are as smart as you are and do not realize this.


My microwave has "Deep Learning features" but it' not going to cause the Singularity anytime soon.

Aren't they already doing this with EAP (in regards to percentage of completion)?

 

[schonelucht]

Isn't it because of the obvious close call liquidity-wise that Tesla finds itself in right now? 6-9 months ago, they still didn't seem to believe, or plan for, the possibility of the ramp progressing poorly, despite optimizing financially for a rapid ramp. They finally got to a point a few months ago where the financial ramifications of the poor ramp relative to the plan became clear. This is one of the consequences. This is the most obvious answer to me. Does this not seem a reasonable explanation?

Had these reductions been made let's say in Q4 last year, you'd be right. But now, not really. Because they actually don't help. Remember Tesla willll pay almost full salary AND severance packages in Q2 for those workers. So Q2 is going to be worse than if they had not done this. Q3 is the first quarter that it could make a positive contribution. But that chapter was already projected to be profitable and cash flow positive regardless!

Here's my guess : production isn't as efficient as hoped for and they had to higher more production associates than originally planned. To compensate for lower margins on M3 they are pruning costs elsewhere.

 

[Reality]

Aren't they already doing this with EAP (in regards to percentage of completion)?

Not to my knowledge, I am pretty knowledgeable about rev rec rules.


The only thing I can think of that would delay recognition is if it was tied to a subscription and the updates were considered part of the sale, or necessary for effective use of the software. Even that would just mean that it would get the 'over time' treatment, not % completion.

The fact they deliver the software via a car and the buyer takes ownership of the software would automatically make it recognized at the point of sale though.

Other than leases I dont think there is any AP revenue amortizing out of the balance sheet. I could be wrong though, i guess.

 

[Reality]

Here's my guess : production isn't as efficient as hoped for and they had to higher more production associates than originally planned. To compensate for lower margins on M3 they are pruning costs elsewhere.

I agree that it is this or they thought they would be able to raise capital by now and can't for some reason, im not really sure which i would prefer

 

[luvb2b]

i had checked with ir at one point and eap was not fully recognized.

Not to my knowledge, I am pretty knowledgeable about rev rec rules.

The only thing I can think of that would delay recognition is if it was tied to a subscription and the updates were considered part of the sale, or necessary for effective use of the software. Even that would just mean that it would get the 'over time' treatment, not % completion.

The fact they deliver the software via a car and the buyer takes ownership of the software would automatically make it recognized at the point of sale though.

Other than leases I dont think there is any AP revenue amortizing out of the balance sheet. I could be wrong though, i guess.

 

[Reality]

i had checked with ir at one point and eap was not fully recognized.

Can you send me what they said if you still have it? That doesnt make sense to me on anything except leases maybe.

It would seem like in order for that to be true you would have to have deferred expenses related to the development of the software too. Maybe I am not thinking about it right.

 

[luvb2b]

refer to 17q1 shareholder letter:

Enhanced Autopilot revenue recognized in Q1 on cars delivered in Q4 contributed $35 million to gross profits. This was partially offset by $26 million in warranty reserve for the Takata airbag recall and equipment impairment charges. We have yet to recognize a significant amount of deferred revenue from future functionalities related to Enhanced Autopilot and Full Self-Driving Capability options. ​

does it bother you that you've been wrong on so many different issues? not sure how concentrated you are... i talked to a short and he was like 2% short. that's small enough that he could afford to not pay much attention. i don't get the sense you are at 2% though.

Can you send me what they said if you still have it? That doesnt make sense to me on anything except leases maybe.

It would seem like in order for that to be true you would have to have deferred expenses related to the development of the software too. Maybe I am not thinking about it right.

 

[bdy0627]

Had these reductions been made let's say in Q4 last year, you'd be right. But now, not really. Because they actually don't help. Remember Tesla willll pay almost full salary AND severance packages in Q2 for those workers. So Q2 is going to be worse than if they had not done this. Q3 is the first quarter that it could make a positive contribution. But that chapter was already projected to be profitable and cash flow positive regardless!

Here's my guess : production isn't as efficient as hoped for and they had to higher more production associates than originally planned. To compensate for lower margins on M3 they are pruning costs elsewhere.

That's plausible. I get that this would have made a lot of sense to do in Q417 or Q18. Wouldn't it still make sense from a cash standpoint to target Q3 for improvement since that is when cash is projected to be lowest, thus higher risk for Tesla?

 

[Waiting4M3]

Had these reductions been made let's say in Q4 last year, you'd be right. But now, not really. Because they actually don't help. Remember Tesla willll pay almost full salary AND severance packages in Q2 for those workers. So Q2 is going to be worse than if they had not done this. Q3 is the first quarter that it could make a positive contribution. But that chapter was already projected to be profitable and cash flow positive regardless!

Here's my guess : production isn't as efficient as hoped for and they had to higher more production associates than originally planned. To compensate for lower margins on M3 they are pruning costs elsewhere.

It sounds to me that you're assuming that the RIF decision came after they saw Q3 profitability. I don't think that's so clear cut. The RIF was announced yesterday but who knows how long it has been in the works? and whether it factored into the math that lead Elon to tweet Q3/4 profitability.

I don't disagree with you that production efficiency of M3 is not as high as it could be in Q2, but I disagree that the RIF is directly a result of that. Cost saving is cost saving. Any CEO if he/she sees the chance to do it, why wouldn't they? Even if M3 margin is already good, saving more cost would only give them more money to invest into future products, why wouldn't they do it.

 

[dakh]

Guys,


Full Self Driving is not a variable attribute. Either it is Level 5 or it is not.


"FSD features" is meaningless garble stated to get headlines. IM shocked some of you are as smart as you are and do not realize this.

You would do better without jabs on collective IQ level here. Read up on iterative development. "FSD features" means exactly what one would think, stuff that can drive the car, this time in a limited set of conditions. Expanding (iterating over) the set of conditions computer can keep driving while validating "FSD features" already shipped to production is going to eventually get to L5. Unlike any other player in this space Tesla will have the confidence of millions of miles driven for the features that are already shipped, data to tweak implementation etc.

 

[bdy0627]

Some analysis on the job cuts from this article:
Elon Musk Finally Admits the Hard Truth About Tesla

"According to Fortune, Tesla’s employee headcount had grown 15 percent in the past six months; cutting that number by 9 percent could save about $80 million a quarter in operating expenses in the future, after taking an extra financial hit this quarter (Tesla is “providing significant salary and stock vesting” to people who are being laid off, Musk wrote). “In the context of the company’s high cash-burn rate, $80 million per quarter may not sound like enough to have an impact,” venture capitalist Gene Munster said, according to Bloomberg. “But as the next several months may decide the fate of the company, every dollar counts.”" [my emphasis]

 

[schonelucht]

Cost saving is cost saving. Any CEO if he/she sees the chance to do it, why wouldn't they? Even if M3 margin is already good, saving more cost would only give them more money to invest into future products, why wouldn't they do it.

I agree for residential solar. They can’t seem to turn that sector around. Better to scrap it completely. But not for their car business. That’s set to double for two more years at least. They’ll need those people.

 

[Reality]

refer to 17q1 shareholder letter:

Enhanced Autopilot revenue recognized in Q1 on cars delivered in Q4 contributed $35 million to gross profits. This was partially offset by $26 million in warranty reserve for the Takata airbag recall and equipment impairment charges. We have yet to recognize a significant amount of deferred revenue from future functionalities related to Enhanced Autopilot and Full Self-Driving Capability options. ​

does it bother you that you've been wrong on so many different issues? not sure how concentrated you are... i talked to a short and he was like 2% short. that's small enough that he could afford to not pay much attention. i don't get the sense you are at 2% though.

Honestly, things like that have nothing to do with my thesis so I dont pay much attention. I was mostly speaking to my understanding of GAAP.


Also, a 2017 Q1 SH letter stating there is def revenue (bundled with FSD) that relates to EAP and not specifying leases does not exactly dispute anything I stated.

I would find it very odd for any part of the EAP sales price on a sold car to not be fully recognized. And even if it's not, it is meaningless to the overall thesis.

 

[Reality]

You would do better without jabs on collective IQ level here. Read up on iterative development. "FSD features" means exactly what one would think, stuff that can drive the car, this time in a limited set of conditions. Expanding (iterating over) the set of conditions computer can keep driving while validating "FSD features" already shipped to production is going to eventually get to L5. Unlike any other player in this space Tesla will have the confidence of millions of miles driven for the features that are already shipped, data to tweak implementation etc.

So we agree it is nonsense then? Lane assist has been around for awhile, it would be considered a FSD feature, rolling it out is not newsworthy.