TSLA Q2'20 Deliveries and Earnings Forecast: 101k Deliveries & $260M GAAP Profit

Introduction


Until yesterday, I didn't really pay close attention to Q2 deliveries, because I thought it would simply be too hard to estimate with any accuracy this quarter, due to the Fremont shut down, Shanghai ramp, Model Y ramp, and potential demand impacts from COVID-19. I just thought there were too many uncertainties, so I had decided to wait for the official P&D (Production and Deliveries) report from Tesla, before perhaps taking a look at how Q2's earnings could pan out.

However, this post on TMC yesterday, containing global EV delivery numbers for the months April and May, made me want to look into Tesla's Q2 deliveries, because I thought that with this data I might be able to make some semi-accurate predictions. This post presents those predictions, including data and assumptions these predictions are based on, as well as a brief earnings forecast.

P&D

Shanghai

Let's start with production and deliveries from Gigafactory Shanghai, because it's the easiest to accurately predict, and it wasn't affected (much) by shut downs in Q2.

Data

April was a pretty straight forward quarter with no shut downs in China, and we have official delivery numbers and some official information about production. This Teslarati article reports that the CPCA (China Passenger Car Association) reported Tesla's China sales in April as 3,635, and production at over 11,000.

May was a bit more unusual, because according to this CNBC article the Shanghai factory was shut down from May 1st until May 5th as part of an official holiday break, which was extended by Tesla until May 8th. However, for May we also have exact official production and delivery numbers thanks to the CPCA, reported by this InsideEVs article as deliveries of 11,095 and production of 11,501.

The last piece of data comes from Tesla itself. In its Q1'20 Update, Tesla stated:
Our Gigafactory Shanghai ramp is progressing according to plan. Due to better than expected progress, we believe Model 3 will achieve a production rate of ~4,000/week (or ~200k/year extrapolated run rate) by mid-2020.

Predictions

As for production, April was somewhere between 11,000 and 12,000, so I'll assume it was 11,500. We know production in May was exactly 11,501, but this was in only 23 days (May 9th - May 31st), so daily production was 11,501 / 23 = 500/day, or 3,500/week.

Extrapolating that to get June's Shanghai production number would give 500 * 30 = 15,000 produced, but considering Tesla expects to hit ~4,000/week some time in June, I will go with a prediction of 16,000 produced at Shanghai in June.

Therefore, I estimate Shanghai's total production for Q2 to be: 11,500 + 11,501 + 16,000 = ~39,000.

As for deliveries, inventory in China increased by over 8,000 units in April and May, but I suspect that this was due to the price reductions in May, and/or sending cars to more remote locations first. I do expect inventory to increase somewhat in China during Q2 (higher production volume means higher inventory in absolute numbers if inventory stays the same in days of sales/production), but I think an overall inventory increase of ~2,000 units is more likely, so I expect Tesla to deliver a total of ~37,000 vehicles in China during Q2.

Fremont

This is a little trickier due to the shut down. The biggest uncertainty is how quickly Tesla and its suppliers were able to ramp production back up to previous levels, but Model Y production, USA demand, and other uncertainties exist as well. However, I think we're still still able to draw some conclusions from the following data.

Data

https://insideevs.com/news/426128/global-plugin-car-sales-april-2020/

https://insideevs.com/news/426128/global-plugin-car-sales-april-2020/

http://ev-sales.blogspot.com/2020/06/global-top-20-may-2020.html


Furthermore, according to Tesla's Q4'19 ER, Tesla's inventory was 11 days of sales at that time using a TTM average. 2019 deliveries were 367,656, so that was 367,656 / 365 = 1,007 per day, and therefore inventory at the end of Q4'19 was approximately 1,007 * 11 = 11,077 vehicles. In Q1'20, Tesla produced approximately 14,272 more vehicles than it delivered, so inventory at the end of Q1'20 was 11,077 + 14,272 = 25,349.

Fremont reopened on May 12th, leaving pretty much exactly 7 weeks of potential production before the end of the quarter.

Finally, Model Y VINs around 23,000 have been spotted, whereas at the end of March the highest VIN reported was around 3,000.

Predictions

We have a clear picture of how April went. Tesla sold 14,793 vehicles in April, so that means 14,793 - 3,635 = 11,158 vehicles outside of China. This means that, excluding the increase in inventory of vehicles coming out of Shanghai, inventories were reduced from 25,349 to 25,349 - 11,158 = 14,191 by the end of April.

Fremont was open for 7 weeks this quarter (May 12th - June 30th). I'm going to assume Tesla was able to produce at more or less full volume for 6 out of those 7 weeks. This could be too pessimistic if there were no hiccups, or this could be too optimistic if there were some issues for Tesla or one of its suppliers. It's the biggest uncertainty in all of this, but this is what I went with.

Based on the Model Y VINs reported, I'd guess approximately 15,000 MYs were produced and delivered in Q2, at a production rate of 2.5k / week. It's kind of a guess to be honest, but production and deliveries will definitely be less than the highest VINs reported around 23,000. 15,000 MYs in Q2 on top of the ~1,000 MYs in Q1, puts actual deliveries at around two thirds of the highest VIN reported. It could be a bit higher or lower, but this is what I've gone with.

Tesla reported in their Q1'20 ER that M3+Y production capacity in Fremont at the time was ~8k/week. It's possible that MY production has increased without affecting M3 production, but for now I'm assuming that the total of 8k/week remained steady, and that Tesla produced 5.5k M3 per week in Q2, for a total of 33k M3 produced. This brings total M3+Y production for Q2 to 48k.

Model S+X are the most difficult in my opinion, because I'm unsure what demand will have been like in the wake of the COVID-19 crisis, and these are the only vehicles for which Tesla is demand constrained rather than production constrained. I'll go with MS+X production of 1k/week for 6k total. This brings total Fremont production to 54k for Q2.

Now to get the total Fremont deliveries number in Q2, all we have to do is estimate the change to global inventories. These are some historic data we have:

Q4'19: 11,077 (4 year low in days of sales)
Q1'20: 25,349 (very high due to COVID-19)
April'20: 14,191 (excluding Shanghai)

I don't think inventory will end up as low as Q4'19, because Q4 usually sees lowest inventory, and it was a 4 year low. Exactly how low it'll go will likely largely depend on demand in NA. The country was shut down for the first 1-1.5 months of the quarter, so demand likely suffered during this period, but most of the production that is allocated to NA usually happens towards the end of the quarter because of logistics. This means that production allocated to NA was barely affected, but demand was. This is likely why we saw Tesla pull some demand levers and offer some incentives towards the end of the quarter. I'm going to with an inventory reduction (ex-Shanghai) of 10,000 in Q2, although it could be a few thousand more or less.

In total, this brings Fremont deliveries to 64k for Q2, which would put worldwide deliveries at an astonishing 64k + 37k = 101k, give or take a few thousand.

As for the breakdown between models, I'm going with:

MS+X: 9k
MY: 15k
Fremont M3: 40k
Shanghai M3: 37k

Earnings

Energy


Automotive




Income


Summary


Comments

First and foremost, don't count on these financials being very accurate. I could've not only been off in terms of delivery numbers, Q2's financials are much more difficult to predict than any quarter in recent history, perhaps any quarter ever. He's a non-exhaustive list of things that are particularly hard to predict this quarter:
  • Tesla Energy. TE had a particularly bad Q1, probably due to both seasonality and COVID-19. I'm expecting Q2 to be worse, because it's hard to install energy projects during global shut downs, but 1 or 2 large megapack projects could swing this in the other direction.
  • Fremont shut down's effect on margins. Pretty much a guessing game. There should be a negative impact, but furloughs probably mitigated it to some extent.
  • Low deliveries in Europe's effect on credit sales. Literally a complete guess.
  • FSD deferred revenue recognition. Probably ~100M or so? But hard to estimate precisely.
  • FSD discounts and take-rate's impact on ASPs and Margins. This one is perhaps the easiest to estimate on this list, but still a tough factor to consider.
  • Model Y margins in the middle of the product's ramp-up. Nobody really knows for sure.
  • Shanghai margins while still ramping. There is further uncertainty here.
  • Service & Other revenue and margin during COVID-19 shut downs.
  • Effect of pay cuts, temporary halt of stock-based compensation, and other cost savings' effects on OPEX.
These are on top of the usual hard to predict items such as losses/gains due to changes in FX rates.

Therefore, even after we receive the official P&D it will be very difficult to accurately model this quarter's financial results, but at this point it is simply very difficult to predict, and so a very large margin of error should be attached to these numbers.

With that being said, my model spits out quite a large GAAP Profit of $260M, so I'm fairly optimistic about a Q2 GAAP profit at this point.

Conclusion

It looks like we're headed for a really strong P&D, and a likely GAAP profit that leads to S&P 500 inclusion. However, especially for the GAAP profit a large margin of error should be taken into consideration.

I'll probably update the Earnings Forecast with official P&D numbers, sometime in between now and the actual Q2'20 ER.

Comments

  1. Hi Franck, great blog and research. Any update following Q2 deliveries or are you waiting for the financials ? Thanks !

    ReplyDelete
    Replies
    1. Thank you! I'm glad you're enjoying it.

      I will do a Q2'20 ER forecast updated with official delivery numbers sometime between now and the ER on the 22nd.

      I'm also currently considering writing a blog post about S&P 500 inclusion. If I decide to write it, that would probably be finished within 3-4 weeks, perhaps much sooner. It'll depend on how big of a post it turns into and how motivated I am to finish it.

      Delete

Post a Comment