yes, i know you meant that for model 3. i did not explain clearly. in order to avoid crossing the 200k usa tax credit boundary, model 3 deliveries will be maximally loaded to canada this quarter. the 10:100 (or 1:10) ratio of reservations will not apply to deliveries in this quarter because they have to dump as much of model 3 production as they can to canada.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Near-future quarterly financial projections
- Thread starter luvb2b
- Start date
yes, i know you meant that for model 3. i did not explain clearly. in order to avoid crossing the 200k usa tax credit boundary, model 3 deliveries will be maximally loaded to canada this quarter. the 10:100 (or 1:10) ratio of reservations will not apply to deliveries in this quarter because they have to dump as much of model 3 production as they can to canada.
I dont think I was right, i think both moved an immaterial amount, so jury is still out.
I think your logic on deliveries is super reasonable, it's not really overstated at all. This is probably the most bullish case i have see using reasonable estimates. I think we can agree that even the smallest of changes in assumptions will have a fairly significant effect on the model since it compounds every quarter in some areas.
Did you build this model yourself? Seems like something you built for work or something first because i can tell it took a lot of time.
Did you build it purely in excel or something else?
1) no if opex scales past a certain % with revenue, then you could have an unprofitable model at even the rosiest of demand assumptions. It doesnt need to be 1% to 1% if it scales anything close to that they are dead in the water.
2) You can financially engineer a lot of stuff within 1 quarter wrt cash, and to a lesser extent, profit. Q3 would be perfect for that for both. Hold back lots and lots of US deliverables (assuming you can produce that many) and dump all of them into Q3 to game the FIT credit. It's perfectly logical. Rec all that revenue in Q3 even tho you will take a huge bath in Q2.
In Q4 you have the last quarter of the FIT credit so I think its POSSIBLE that you get a big push (i do think Mission E/I-Pace/E-tron are going to take a toll on demand here though as it is natural human tendency is to want the 'newest' thing, this is especially true for the coastal elite* demographic Tesla plays to, and the MOdel S/X just arent that anymore). So it comes down to how deep is that high margin Model 3 book. Can you push hard enough and take advantage of just those 2 one time events to try and turn a profit? Maybe by the slimmest of margins.However after that what is left? Only a capital raise whcih just kicks the can down the road if the business model isnt profitable.
Cash flow is easier because you can just fluff cash or pull up a little bit on CAPEX/vendor payments, rollover contract terms, or not pay bonuses or any number of items that you need to make a cash number. I think they are probably out of luck on taking deposits other than model 3 config deposits.
*I consider myself part of this demographic so im not being prejudice here, im very pro environment and green energy
this is 100% my own work. i don't have time for a job, fortunately.
the spreadsheet is all in excel but the data and knowledge in various aspects are deep and sourced from different people who know more about this or that than me.
when i have enough at stake, this is what i do. i buy the products, visit the stores and factories, and go meet the experts (long & short). i met a prominent short a few weeks ago, who could be you by the things he mentioned as primary concerns.
not me, id pay good money for your model though
Also im def not 'prominent' although i have what most would consider a very material amount of money on this and spend a lot of my workday trying to prove myself wrong.
The funny part, is every spot where we REALLY need data to prove our thesis, we don't have it. I think that might be the biggest of the red flags for me.
Production #'s constantly leak, but why don't reserve numbers leak?
The customer deposits number on the balance sheet has more or less been destroyed as an indicator because of the changes in the accounting disclosures on how some of the deferred revenue was reclassed (i dont think that was an accident, i believe there is some flexibility in balance sheet presentation here, Tesla didnt want us seeing that number anymore imo):
(from 10-Q)
Prepayments on contracts that can be cancelled without significant penalties, such as vehicle maintenance plans, have been reclassified from deferred revenue to customer deposits. Refer to the Automotive Revenue and Automotive Leasing Revenue sections below for further discussion of the impact on various categories of vehicle sales.
I think that is what I find the most telling. Tesla is notorious about touting GOOD news, they dont hold anything back. So when they intentionally dont say something, they basically tip their hand that bad news is coming.
I honestly feel the CR Report and edmonds review are the nails in the coffin, that was near shut already. People who buy 35-40k cars do TONS of research because its such a material amount of money to them. Those are the 2 main places they look to. People who buy 100k cars tend to do less because 100k is likely not as significant to the demographic that buys 100k cars.
I wish you well either way as I have honestly enjoyed the work you have done here. It's the closest anyone has come to making me question my thesis.
I could be mistaken, but I thought you were saying the percentage drop in SGA vs production was out of line.
And revenue comes from selling things, so until TE picks up, fairly proportional to car production (with factor for ASP of S/X vs 3).
could tesla be profitable as opex scales with revenue? sure it could. it depends on the scale. for example if you assume opex is a linear function of revenue, with opex = fixed expenses + slope * revenue, then the scaling parameter [m] is the key value. if that slope is meaningfully below the gross margin level then you can see nice operating leverage and profit growth as revenues grow.
here is what the shorts see, with opex on the y axis and revenue on the x axis, opex clearly is mostly linear with a high slope parameter. and so tesla is basically a dead man walking.
i think the analysis is confounded by the acquisition of solarcity and spend for energy related things. so i attempt to separate solarcity and tesla spend, which is actually fairly straightforward for most of the post acquisition period (3 quarters have direct disclosures).
with tesla auto-only opex on the y axis and auto-only revenue on the x axis, you can see there is a step function, not a linear function. opex climbs rapidly w/revenue to a new level while they spend ahead of a product ramp. then opex flattens w/revenue as the new product scales up. there is not a single linear function with a single slope parameter, it's a series of line segments with varying slopes, and the slope is much flatter as revenue ramps much higher.
you can see where i have modeled the future in my modeling.
which chart do you want to bet on? that is a primary difference between the longs and the shorts.
there are not accounting tricks i know that could be used to show a profit for 1/2 quarters if the company's financial situation continues to decline. that's because the losses right now are so large that it would take herculean accounting manipulation to swing to a profit from the current level. and it would be too obvious.
define "good money".
we rarely have good data to prove one way or the other when investing.
on reservations, at this stage i don't think disclosures matter much. the order book is so deep and they are so badly supply constrained it's sort of wasted thought. it will matter a lot more as total model 3 deliveries start moving towards 100-150k. one short i talked to thought the take rate on reservations was 7%. i think people like him will be sorely disappointed, although in the longer run i may be disappointed too. but not yet.
btw on the cr report, you should watch this video. notice the facial emotions and body language of the reviewers, esp jake fisher.
compare to a more neutral baseline for jake fisher.
i don't think cr is going to be such a big issue.
This is a fantastic post
Up to which production number are the solarcity costs known on your chart? Approximately
Obviously there is a wide spectrum between opex flatfline from here on out to opex growing in tandem with revenue. Neither one nor the other seems credible. I am personally holding a 5% opex growth for every 10% of revenue growth. That too is obviously a very unsophisticated guess.
fair, but never have we seen a valuation swing, both % and $ the amount that some of these fairly simple data points would swing the stock if they pointed more towards the bull or bear argument. I mean you even commented on take rate as being a super important indicator. 7% is absurd, but it is very telling that Elon dodged this question last EC. I don't know how anyone could think its a good % when he avoids that question.
Also, I am going to find it, but someone logged how many of the (i think it was premium wheels) deliveries occurred on the tesla tracker, the extrapolated it to take rate.
I am going to make up numbers here, but i will find the actual analysis
Someone scraped the premium wheel option for Model S and noted something like 40% of all Model S had a wheel upgrade (scraped from used car data points, i know im being terrible about the data). You could see that the premium wheels option on model 3's had essentially stopped after something like 4,000 were delivered.
With the assumption that only 20% of model 3s take the premium wheels (cheaper car), and only 30% of the 450k reservations want the LR battery w premium upgrades (that is currently offered) you get
135k orders x 20% prem wheels x take rate = 4,000
take rate = 6.75%
I need to find this analyssi to show you i cant believe i didnt save it. Like even if you assumed really favorable assumptions, like 10% premium wheels and only 20% want the current offering, the drop off in prem wheel deliveries signaled an extremely dismal take rate/
bdy0627
Active Member
This was really helpful. Thank you! Your charts showing the step function vs linear function of opex vs revenue is a great illustration of the two opposing perspectives about what is happening.
bdy0627
Active Member
I very much agree with your perspective on the Consumer Reports review. The video of the interview with Jake Fisher suggested to me that this may actually turn out to be a rather positive development both for Tesla and CR. He mentions his hour long phone call with Elon and it is obvious that he found it to be a very positive conversation. He comes across as quite optimistic about Tesla rapidly improving the Model 3 and wanting CR to continue to evaluate the vehicle as it improves over time. That interview is critical viewing for anyone forming an opinion based upon Consumer Reports' perspective on Tesla and the model 3.
bdy0627
Active Member
I don't see the CR Report as anything like the scathing complaints in the Edmunds Review. Be sure to watch the interview video with Jake Fisher as it's very helpful for getting additional information about CR's perspective on the model 3. The only substantial issue CR really had with the car was the poor emergency braking results from 60 mph. The concerns about the touch screen are not so much an issue as a preference about where to place certain controls. That's going to vary widely by person and isn't likely to affect those buying the model 3 since they almost certainly already know the controls are on the touch screen. They also touched on the ride being firm in the back seat especially. Again, that does not seem likely to deter many buyers of the vehicle. They highly praised the driving performance, which is perhaps the most important aspect of the car for many buyers. Tesla needs to address the brake issue. Once they do that, there isn't much on the CR report to be concerned about affecting demand for the model 3. The Edmunds review is a whole different animal with almost nothing but complaints other than how awesome the car performs. That review stands way out to me as an outlier very negative report compared with lots of positive reports along with the CR report.
so a key assumption you made is that all 135k who want the lr battery w/premium upgrades have been invited. a second key assumption is all premium wheel packages were done in one single batch, vs. multiple batches at different points in time (for example, they may have just run out of the special wheels and stopped delivering that option until they got more).
but i agree: *if* all 135k that want this combination lr + pup have been invited and configured, and assuming all premium wheel packages have been built already *then* the take rate is 6.75%.
but what if the number invited out of the population of reservations who want the lr+pup package is only 50k? take rate would jump to 40%.
you can review various sources and see virtually all non-owner reservations after the 1st day remain uninvited. 180k reservations were made on the first day. many of those by previous owners. so ballpark 250k reservations may yet remain uninvited.
alternatively, consider that out of the 135k of the total population you are using as wanting the package, you are only observing data for the subset which is north american (usa or canada). even if you assumed that 30% of reservations were international and 30% of the population wanted the lr+pup package (assuming no correlation between international and package), then you're math would become:
450k x 30% want lr+pup package x 70% domestic x 20% prem wheels x take rate = 4000.
take rate = 21%. error is a factor of 3+.
and what if you only observed one of two batches of such production b/c they didn't expect so many people to take premium wheels?
450k x 30% want lr+pup package x 70% domestic + 20% prem wheels x 50% x take rate = 4000. take rate = 42%. error is a factor of 6+.
basically there are so many possible grave errors in the methodology presented, i would not want to bet on it.
i went back and checked. 17q1 is a direct disclosure. 17q2-q3 are indirect disclosures but enough to get reasonable. 17q4 and 18q1 are more guesswork as any disclosure is missing. if anything i may have non-auto costs too low b/c research on storage and solar roofs has grown. increasing cost allocation to solar city only makes the chart look better for tesla in the last 3-4 periods.
you bring up an interesting point though, i haven't reviewed this work in a while. so i should go back and double check my assumptions.
Last edited:
