Near-future quarterly financial projections

[damonbrodie]

Great Q posted by Tesla. Congrats to @petit_bateau for nailing the EPS number.

Lot's of good takeaways from the earnings call. One that stands out is the statement about expecting quite a bit of factory downtime in Q3 for upgrades. @Troy has been predicting this for quite a while now - the last public number I saw was 454K deliveries in Q3 against the 466K in Q2. I'm anxious to see the updated Q3 estimates start to flow. I'm assuming the Q3 P&D will probably head lower than that number - given that Elon made it a clear point very early in his remarks.

I'm guessing this is Highland related...

My Auto revenues prediction was a bit high. Actuals were $20.42B against my prediction of $20.89B (ex leasing and ex reg credits). I only started grabbing daily prices about halfway through Q2, so that may have contributed to some of the delta, but I think the big one was not applying the Discounts on the new inventory that Tesla has on their inventory sales pages. I've been collecting it for about a month now, so I can start factoring that in for all of Q3 and see if I can get even closer to the right automotive revenues.

 

[Christine69420]

My numbers show earnings Q2 > Q1.

My model is still too new and fresh to be more precise than that.

Revenue and EPS was in the expected direction.

Margins or rather the costs not so much.

If I am to be honest.

 

[petit_bateau]

Folowing the Q2-2023 results I've updated my longer term model to include a rosier picture for energy margins, and a continued restraint on overheads. Thus - in this scenario - it now has truly outstanding volume performance, and truly outstanding margin performance through to 2030.

It is hard to justify those margins unless one assumes they are software-enabled, i.e. continually improved FSD/ADAS in the vehicles and an equivalent in the storage. By the way this also is such a stunning performance scenario that it assumes Tesla solves solar, which takes a very active imagination indeed.

What is excluded from this is any revenues from RoboTaxi or Optimus. It was 2014 that Tesla first offered pre-purchase of Autopilot and I feel Musk wishes to claim mission-accomplished by the 10-year milestone. Whilst it may by then be a very good ADAS system of some level I think that it will be at least another 10-years before a full-on Robotaxi of any commercial importance might be viable. Similarly so for Optimus. I'd love to be wrong. In any case this is already such a high-end scenario it does not seem necessary to gild the lily any further.

If one then takes the above and runs it through a variety of ways of calculating fair value for share prices this is the result:

Which in graphical terms is this:

It almost seems to me that in the short term the real issue in the share price is what is the correct discount rate to apply. It is hard to justify 2023 shareprices nearing $300 unless one is contemplating superlative business growth and continued improving profitability, and in a near-risk-free manner (i.e. current discount rate of say 5% rather than a risked discount rate of say 10%).

At the moment the EPS-driven valuations tend to underwhelm as 2023 EPS is likely to be below 2022 EPS. Looking forwards the 2025 year is when a real tension will exist between those who value using EPS will be motivated to buy - especially if by then the Cybertruck and Highland are pushing prifitability back towards GM of 30% - versus those who would be somewhat more restrained if using NPV-driven valuations.

Personally I'm not brave enough to trade or use leverage in any way, as the irrationalities in this share price; combined with the outright manipulations and gaming of the market place; plus the information asymmetry that exists all make me very cautious.

Fair value at the moment looks awfully like $210-$220 when one looks at the quarterly data and applies a 10% discount rate made up roughly of a approx 6% equity risk premium, and 4% 10-year bond rate.

It may well be that many other companies are irrationally overvalued in the market by such an approach, but that is a different matter.

Nevertheless am I missing anything ?

 

[Doggydogworld]

Does anyone have a good source for Tesla Shanghai monthly production numbers? Moneyball on Twitter used to report them, but I think he stopped and moreover, Elon made Twitter useless for non-members. Three separate monthly reports come out, one after the other:

1. Wholesale (retail sales plus exports)
2. Retail sales
3. Production

1 is widely reported and 2 isn't that hard to find, but production is tough. Many sites erroneously call the wholesale report "production", ruining search results.

Troy's quarterly production numbers always matched the monthly reports I collected. But he has Q2 production at 84,560 3s and 154,695 Ys for 236,255 total. That means June production was only 24,778 3s and 48,074 Ys for 72,852 total. That conflicts with a random Twitter guy who I can't find now that claimed June production was a bit over 80k. He didn't give an exact number, but his charts and inventory math made me think he was using a real number.

72k is also the lowest since February (and a lower average daily rate than even February). And 72k implies a record (by far) 21k inventory drain in China in June. Why slow production so dramatically just before you shut down for retooling? I could understand if they shut down for retooling the final week of June, but I don't recall any reports of that.

Any help appreciated.

 

[Yuha]

Does anyone have a good source for Tesla Shanghai monthly production numbers? Moneyball on Twitter used to report them, but I think he stopped and moreover, Elon made Twitter useless for non-members. Three separate monthly reports come out, one after the other:

1. Wholesale (retail sales plus exports)
2. Retail sales
3. Production

1 is widely reported and 2 isn't that hard to find, but production is tough. Many sites erroneously call the wholesale report "production", ruining search results.

Troy's quarterly production numbers always matched the monthly reports I collected. But he has Q2 production at 84,560 3s and 154,695 Ys for 236,255 total. That means June production was only 24,778 3s and 48,074 Ys for 72,852 total. That conflicts with a random Twitter guy who I can't find now that claimed June production was a bit over 80k. He didn't give an exact number, but his charts and inventory math made me think he was using a real number.

72k is also the lowest since February (and a lower average daily rate than even February). And 72k implies a record (by far) 21k inventory drain in China in June. Why slow production so dramatically just before you shut down for retooling? I could understand if they shut down for retooling the final week of June, but I don't recall any reports of that.

Any help appreciated.

Rob Maurer tracks Shanghai numbers. June production was 27,440 3s and 52,925 Ys for 80,365 total:

 

[petit_bateau]

Does anyone have a good source for Tesla Shanghai monthly production numbers? Moneyball on Twitter used to report them, but I think he stopped and moreover, Elon made Twitter useless for non-members. Three separate monthly reports come out, one after the other:

1. Wholesale (retail sales plus exports)
2. Retail sales
3. Production

1 is widely reported and 2 isn't that hard to find, but production is tough. Many sites erroneously call the wholesale report "production", ruining search results.

Troy's quarterly production numbers always matched the monthly reports I collected. But he has Q2 production at 84,560 3s and 154,695 Ys for 236,255 total. That means June production was only 24,778 3s and 48,074 Ys for 72,852 total. That conflicts with a random Twitter guy who I can't find now that claimed June production was a bit over 80k. He didn't give an exact number, but his charts and inventory math made me think he was using a real number.

72k is also the lowest since February (and a lower average daily rate than even February). And 72k implies a record (by far) 21k inventory drain in China in June. Why slow production so dramatically just before you shut down for retooling? I could understand if they shut down for retooling the final week of June, but I don't recall any reports of that.

Any help appreciated.

Rob Maurer tracks Shanghai numbers. June production was 27,440 3s and 52,925 Ys for 80,365 total:

That looks very plausible @Yuha thank you/

I don't have a better source than @Doggydogworld and have found the same problem with google fu in recent quarters . For my purposes ordinarily quarterly granularity is sufficient. Butfor what @D is likely doing I understand why monthly is necessary.

Somebody recently kindly passed me [1. Wholesale (retail sales plus exports)] but was under the impression it was [3. Production] so there is misunderstanding out there.

Here is how I ended up allocating them, in case that helps. Mind you this may just be me reflecting @Troy .

I wish they'd just publish this stuff. It would lead to better decision making all round and improve the information asymmetry that exists.

 

[Troy]

@Doggydogworld,
I updated the public spreadsheet here with the latest numbers for Q2: Tesla Production & Delivery History It shows production by factory and deliveries by region.

 

[Doggydogworld]

Rob Maurer tracks Shanghai numbers. June production was 27,440 3s and 52,925 Ys for 80,365 total:

Thanks, that's exactly what I needed.
And thanks Troy for keeping all your different numbers so up to date and precise. Don't know how you do it.

 

[petit_bateau]

per @hobbes

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

I look for a reason for today's bump... this is all I could come up with: CyberTruck pre-hype (because people were STILL thinking it wouldn't produce) The 7 day itch (neg/pos) It's been 3 days since this place has been depressing.

teslamotorsclub.com

 

[hobbes]

per @hobbes

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

I look for a reason for today's bump... this is all I could come up with: CyberTruck pre-hype (because people were STILL thinking it wouldn't produce) The 7 day itch (neg/pos) It's been 3 days since this place has been depressing.

teslamotorsclub.com

Something went wrong here, what was quoted is not from me.. what it links to is from me, can you please check and correct?!

 

[petit_bateau]

Something went wrong here, what was quoted is not from me.. what it links to is from me, can you please check and correct?!

There is an error in how TMC presents internal links. The problem is that the TMC preview is typically of the first post in a thread, not the actual post being linked. Only of you click the link do you then go to the specific post that is intended. In this case I had linked to your post #424,466 which in turn contained the stationary market info.

Six Chinese companies were among top 10 global energy storage system suppliers in 2023H1 | Shanghai Non ferrous Metals

According to SMM statistics, the global energy storage system shipments in 2023H1 reached 72.4 Gwh. China’s shipments were 47Gwh, accounting for 65%; overseas shipments were 25.4Gwh, accounting for 35%; global energy storage system shipments were still dominated by Chinese integrators. Tesla’s...

news.metal.com

Wie bei Elektroautos: Tesla führt Weltmarkt für stationäre Akkus an, großer Rest aus China

Laut einer Studie ist Tesla bei stationären Akkus weltweiter Marktführer – wie bei Elektroautos direkt vor BYD und weiterer China-Konkurrenz.

teslamag-de.translate.goog

(Also it is not easy to give source info attribution, hence my noting @hobbes )

 

[petit_bateau]

There is an error in how TMC presents internal links. The problem is that the TMC preview is typically of the first post in a thread, not the actual post being linked. Only of you click the link do you then go to the specific post that is intended. In this case I had linked to your post #424,466 which in turn contained the stationary market info.

Six Chinese companies were among top 10 global energy storage system suppliers in 2023H1 | Shanghai Non ferrous Metals

According to SMM statistics, the global energy storage system shipments in 2023H1 reached 72.4 Gwh. China’s shipments were 47Gwh, accounting for 65%; overseas shipments were 25.4Gwh, accounting for 35%; global energy storage system shipments were still dominated by Chinese integrators. Tesla’s...

news.metal.com

Wie bei Elektroautos: Tesla führt Weltmarkt für stationäre Akkus an, großer Rest aus China

Laut einer Studie ist Tesla bei stationären Akkus weltweiter Marktführer – wie bei Elektroautos direkt vor BYD und weiterer China-Konkurrenz.

teslamag-de.translate.goog

(Also it is not easy to give source info attribution, hence my noting @hobbes )

Six Chinese companies were among top 10 global energy storage system suppliers in 2023H1 | Shanghai Non ferrous Metals

According to SMM statistics, the global energy storage system shipments in 2023H1 reached 72.4 Gwh. China’s shipments were 47Gwh, accounting for 65%; overseas shipments were 25.4Gwh, accounting for 35%; global energy storage system shipments were still dominated by Chinese integrators. Tesla’s...

news.metal.com

Turning to the substance of the news item,

"According to SMM statistics, the global energy storage system shipments in 2023H1 reached 72.4 Gwh. China’s shipments were 47Gwh, accounting for 65%; overseas shipments were 25.4Gwh, accounting for 35%; global energy storage system shipments were still dominated by Chinese integrators. Tesla’s shipments in the first half of the year exceeded 7Gwh, ranking first in the world."

This is an area that is desperately short of good data in either the public domain or in proprietary reports. Over the years I have put together a table of what I think the shipments were, plus my own forecast, but I have always been concerned about the possibility of significant under-reporting of actual shipments for all sorts of reasons. The news item is most useful and shows how industry/media are taking the topic more seriously, and can usefully be compared with my previous historical & forecasts.

The most obvious thing is that they (SMM) are suggesting 2023-H1 shipments were 72 GWh, which implies it will likely be approx 150 GWh for the full year. In comparison my forecast was only 47 GWh, so I now tend to think I have been undercounting in the past as it is not credible that volumes leap from 30 GWh to 150 GWh in one year. Also the SMM data includes 7 GWh attributed to Tesla which corresponds well with Tesla's Q1+Q2 public reports of 3.9 + 3.6 = 7.5 GWh and in turn suggests that the SMM data is reasonably solid. Looking at the list of companies and the corresponding volumes also suggests to me that the data is reasonably reliable.

The language (Chinglish) is slightly ambiguous but looking at the company list and the volumes I am fairly sure that there is no doublecounting of vehicle mobility product going on. These volumes appear to be purely stationary storage, for a combination of consumer-scale and utility-scale usage. Confusingly this also gets termed domestic storage so one needs to be careful about terminology.

The implication is that stationary market is aready taking approx one quarter of the volume versus the vehicle market, i.e. FY 2023 will be approx 150 GWh vs 626 GWh, approx. This is great news in terms of overall stationary storage adoption/scaling/deployment and further suggests to me that we are significantly underestimating the speed of the transition. I did not think that the (historical data + my projections) stationary market would reach that relative fraction until approx 2030.

Again this is the previous forecast I had, which I now think I will have to very substantially revise.

Does anyone have access to the SMM historical data to compare with my own scrapings of the various datapoints I have amassed out there, which now seem to be confirmed as a significant under-count ?

Oh, and by the way, this is another new-world industry where Chinese dominance is fully-locked in. It also happens to be one that is of vital strategic importance for the Chinese in their internal use as well as for Chinese exporting.

(I'll repost this on the main board as well)

(once again thanks @hobbes for snagging the original news item)

 

[MP3Mike]

There is an error in how TMC presents internal links. The problem is that the TMC preview is typically of the first post in a thread, not the actual post being linked. Only of you click the link do you then go to the specific post that is intended. In this case I had linked to your post #424,466 which in turn contained the stationary market info.

It you just paste a link to a TMC post, it will "unfurl" it and show a preview of the first post of the page that contains what you linked to. They expect you to use the quote function to link to content on the site. (But if you really only want to paste a link, and not quote the post, then you can edit the post after you save it to set the "unfurl" option to off so it just shows the URL, or you can use the link function of the editor to supply the text you want displayed for the link.)

 

[petit_bateau]

credit @Singuy and RobMaurer - that is remarkably smooth operating, I wonder if Shanghai will manage to reach 1m total in 2023.

 

[hitchhiker]

probably not. 80k/month doesn't get you there, plus they have a bunch of shorter months in the first half of the year to make up for. But its pretty probable for '24

 

[DarkandStormy]

https://x.com/TroyTeslike/status/1702085589460242472

Six weeks ago this was 467k.

 

[petit_bateau]

credit @Zaddy Daddy also re @GhostSkater and ref mine

 

[Doggydogworld]

Noodling around with Tesla's 4 and 5 millionth car announcements, alongside quarterly P&D reports, I came up with a few estimates for production rate per calendar day. These estimates have error bands since they said "this week" for 5m vs giving an actual day (I assumed Wednesday, 9/13/22) and their early year P&D reporting was a bit haphazard.

5.6k/day - March production
5.3k/day - Q2 production
4.7k/day - Q3 production through 9/13

The numbers point toward 435k Q3 production, with a 95% confidence interval (haha) of 420-450k. Note that while 4.7k/day looks low, it's higher than the 4.6k/day averaged in Jan/Feb of this year. Lunar New Year and other slowdowns meant Shanghai alone averaged 0.5k/day less than max the first two months of the year, plus some start of year down days at the other factories.

 

[petit_bateau]

cortesy @Buckminster and of course @Troy from main thread #427,653

 

[DarkandStormy]

That looks very plausible @Yuha thank you/

I don't have a better source than @Doggydogworld and have found the same problem with google fu in recent quarters . For my purposes ordinarily quarterly granularity is sufficient. Butfor what @D is likely doing I understand why monthly is necessary.

Somebody recently kindly passed me [1. Wholesale (retail sales plus exports)] but was under the impression it was [3. Production] so there is misunderstanding out there.

Here is how I ended up allocating them, in case that helps. Mind you this may just be me reflecting @Troy .

I wish they'd just publish this stuff. It would lead to better decision making all round and improve the information asymmetry that exists.

View attachment 958443

@Troy has said Q3 production from Austin is going to be ~21k. And due to the Highland refresh, I think the MIC Model 3 number there for Q3 will also be lower.

We'll get an idea (global only since Tesla doesn't break anything out by region like literally every other OEM) in a few days.