Near-future quarterly financial projections

[luvb2b]

tl;dr structural tailwinds due to accounting and business facts will push a lot of profit down to the bottom line.

lately i have been thinking about some structural tailwinds to profit in the coming quarters. just wanted to throw some of these out and see how others see them.

1. improving gross margin of service is driven by higher utilization of facilities. i see this as structural because as tesla continues to deliver 80-100k vehicles per quarter and some of the 2015 s/x start to roll off warrantee there should be higher utilization of all service centers. last quarter saw a huge quarter over quarter improvement in service margin. another 10% improvement pushes ~$75m/quarter down to the bottom line.

2. positive margins in energy storage for the first time continue to improve slightly. this was a huge surprise last quarter as in my modeling energy storage finally turned to positive gross margin. another several points of improvement can help push another ~$15m/quarter to the bottom line.

3. aggressive strategies to recognize deferred revenue drive revenue realization. it was mentioned on the call, shown in the filings, and apparent in the autopilot/fsd definition changes and release schedules that tesla is aggressively pursuing ways of getting the deferred revenue off the balance sheet and onto the income statement. i estimate this has ~$75m/quarter of potential added profit (vs current levels). they can manage this item a bit by pushing features and recognizing in one quarter or another.

4. recently the db note indicated ir is guiding to an uptick in credit revenue from the fiat chrysler deal starting in january 2020. different people have different numbers but i think saying an additional ~$100m/quarter gets to the bottom line is reasonable for 2020.

the total here is quite meaningful: for 19q3/q4 there's potentially ~$165m/quarter tailwind, or about 90c of gaap eps.
starting in 20q1 the tailwind goes up to ~265m/quarter tailwind, or about $1.40 of gaap eps.

as two of the largest tailwinds seem to be just accounting facts and another a natural byproduct of selling more cars, i feel i should have high confidence in their existence. curious to hear a few other thoughts, ideally contradictory thoughts.

 

[luvb2b]

delivery estimates: 19k s/x, 84k model 3. model 3 could range 80-90k based on level of back-end loading. they've been swearing up and down that they are trying to reduce back-end loading, and if they have (finally) accomplished that goal deliveries should surprise to the downside.

 

[luvb2b]

on the delivery estimates, the best forecasting method i know is to use known/estimated deliveries to date from various sources, then make adjustments based on what is known about back-end loading and one time events. this quarter model 3 estimation is particularly hard because the prior data is only 2-3 data points, and most of it is skewed due to tax credit runoffs and delivery snafus. so the risk as a forecaster is we use the past data, which implies heavy back-end loading, and then realize that in reality it didn't match what was seen in the short history.

as most known delivery points this quarter are tracking above last quarter for model 3, a rational forecast would be to estimate higher units vs last quarter as well. so just emphasizing here due to limited data points there can be meaningful errors in any model 3 forecast. i expect the model s/x forecast to be more accurate as there's much longer history to work with.

 

[AlMc]

I agree on delivery estimates being very hard to make overall. A 10% move either way leads to a great or disappointing quarter.

 

[EVNow]

delivery estimates: 19k s/x, 84k model 3. model 3 could range 80-90k based on level of back-end loading. they've been swearing up and down that they are trying to reduce back-end loading, and if they have (finally) accomplished that goal deliveries should surprise to the downside.

I'm guessing by bac-end loading you are talking about the wave.

I don't see the wave ending anytime soon. If there is any unwinding, it would be because of delivery issues. Tesla has too much pressure on increasing deliveries every quarter to be able to actually do the unwinding which hits their cash flow and deliveries hard in one of the quarters.

My guess is they have decided to wait for GF3, when there will be some unwinding because of China. EU wave will continue.

 

[EVNow]

the total here is quite meaningful: for 19q3/q4 there's potentially ~$165m/quarter tailwind, or about 90c of gaap eps.
starting in 20q1 the tailwind goes up to ~265m/quarter tailwind, or about $1.40 of gaap eps.

I've been tracking deferred revenue because of EAP/FSD. The big quarter will be when Tesla delivery "automatic city driving" or City NOA. Once that happens and they do the HW3 upgrades, they can recognize most of the deferred AP/FSD revenue. Q1 '20 is lining up to be good for that - esp. given likely seasonally low deliveries in Q1, it would be useful. It could be ~200M.

BTW, what do you think will be the impact of GF3 coming online in Q1 '20 ? How will the depreciation (fixed), margins, ASP etc get affected ? I think Q1 for all these reasons will be wild.

 

[Doggydogworld]

1. improving gross margin of service is driven by higher utilization of facilities.

I don't see utilization as the issue. Most SCs are well-utilized, perhaps overly so. A few under-utilized ones in the hinterlands doesn't move the needle. Warranty work is charged against warranty reserve, so it helps utilization (unless they charge most of it to "goodwill", as the death cult claims).

Tesla says the Services category is driven by used cars, and the numbers seem to scale almost linearly with deliveries (using a one month lag). If utilization is the problem, Musk's talk of opening more SCs to attract buyers who live too far from existing ones does not bode well.

2. positive margins in energy storage for the first time continue to improve slightly. this was a huge surprise last quarter as in my modeling energy storage finally turned to positive gross margin. another several points of improvement can help push another ~$15m/quarter to the bottom line.

Storage margin should improve, but IMHO Q2's margin jump was seasonal. The solar lease fleet produces a lot more kWh revenue in Q2 and Q3, which boosts margins.

The Wal-Mart fiasco is a pretty big drag on solar margins. Fixing that would help a lot.

I mostly agree deferred revenue and emissions credits, though it's hard to nail down amounts and timing. I expected more of the FCA deal to count in Q2 and Q3. Also, if Trump attacks the Federal GHG system we could see credit buyers play wait-and-see.

BTW, what do you think will be the impact of GF3 coming online in Q1 '20 ? How will the depreciation (fixed), margins, ASP etc get affected ? I think Q1 for all these reasons will be wild.

My theory is GF3 is owned by a government entity and leased to Tesla for $1/month or whatever. They'll have to count it as an asset, as with Buffalo and the Panasonic equipment in GF1, but they'll amortize an offsetting liability to match the faux depreciation. The bigger issue is labor, a repeat of Model 3 where tons of workers produced very little for almost a year would obviously hit margins.

 

[Reinhard]

the total here is quite meaningful: for 19q3/q4 there's potentially ~$165m/quarter tailwind, or about 90c of gaap eps.
starting in 20q1 the tailwind goes up to ~265m/quarter tailwind, or about $1.40 of gaap eps.

as two of the largest tailwinds seem to be just accounting facts and another a natural byproduct of selling more cars, i feel i should have high confidence in their existence. curious to hear a few other thoughts, ideally contradictory thoughts.

Model Y cannibalizes Model 3 sales without providing a significant boost. Lack of resources to give an aging Model S a full interior and exterior refresh causes sales to further decline to a rate similar to other luxury car designs that are >5 years old. The massive boost in European sales because of the Netherlands and the UK ends with Netherlands incentive ending Jan 1st and novelty factor ending in the UK. China goes into a deep recession because of the trade war and Chinese customers can no longer purchase anywhere near 3,000 Model 3s/week. Large fixed costs of Gigafactory 3 cannot be recouped as a result.

 

[mrdoubleb]

Model Y cannibalizes Model 3 sales without providing a significant boost. Lack of resources to give an aging Model S a full interior and exterior refresh causes sales to further decline to a rate similar to other luxury car designs that are >5 years old. The massive boost in European sales because of the Netherlands and the UK ends with Netherlands incentive ending Jan 1st and novelty factor ending in the UK. China goes into a deep recession because of the trade war and Chinese customers can no longer purchase anywhere near 3,000 Model 3s/week. Large fixed costs of Gigafactory 3 cannot be recouped as a result.

You forgot about the bit where the aliens landed, but we couldn't fight them as we were too busy running away from the zombies.

 

[EVNow]

You forgot about the bit where the aliens landed, but we couldn't fight them as we were too busy running away from the zombies.

I thought we were running away from the Alien Zombies.

Model Y cannibalizes Model 3 sales without providing a significant boost.

I think it is super important that Tesla is able to shift the product mix of 3/Y - like they can apparently do with S/X.

We certainly don't know how much of the 3 demand will shift to Y.
​

 

[trentbridge]

What I don't see mentioned is any discussion about Tesla accessories as an income source. Wall connectors and floor mats etc. I imagine these are high profit margin items and with the rapid deployment of the Model 3 globally, I'd expect this to be growing rapidly.
Right now the Gloss Black Wall Connector selling at $500 is sold out - demand is real.

So assume $250 average purchase x 100,000 deliveries = $25 million per quarter.

 

[luvb2b]

just noticing today's news about end of quarter incentive changes.
Tesla changes end-of-quarter incentives after some 'abuse' - Electrek

i find it interesting we have this end of quarter push/scramble sort of thing happening again. and the incentive structure makes me wonder if tesla made the same forecasting error that many analysts would make. they wouldn't have a way to know until some of september had passed.

on the delivery estimates, the best forecasting method i know is to use known/estimated deliveries to date from various sources, then make adjustments based on what is known about back-end loading and one time events. this quarter model 3 estimation is particularly hard because the prior data is only 2-3 data points, and most of it is skewed due to tax credit runoffs and delivery snafus. so the risk as a forecaster is we use the past data, which implies heavy back-end loading, and then realize that in reality it didn't match what was seen in the short history.

as most known delivery points this quarter are tracking above last quarter for model 3, a rational forecast would be to estimate higher units vs last quarter as well. so just emphasizing here due to limited data points there can be meaningful errors in any model 3 forecast

@Reinhard sounds completely loony at first, but it's an interesting alternative take. china is currently doing (my estimate 2-3k model 3's per month). would there be demand for 4x that amount somewhat lower prices?

Model Y cannibalizes Model 3 sales without providing a significant boost. Lack of resources to give an aging Model S a full interior and exterior refresh causes sales to further decline to a rate similar to other luxury car designs that are >5 years old. The massive boost in European sales because of the Netherlands and the UK ends with Netherlands incentive ending Jan 1st and novelty factor ending in the UK. China goes into a deep recession because of the trade war and Chinese customers can no longer purchase anywhere near 3,000 Model 3s/week. Large fixed costs of Gigafactory 3 cannot be recouped as a result.

 

[gluu]

There's also the SolarCity aspect of the business. They're doing 'rentals' vs leases now and I'm not clear on how dangerous or positive this is. It seems like they limited the states and they're somehow trying to improve cash flow using the ITC. Don't want to go down the SCTY rabbit hole again.

 

[trentbridge]

Very similar to my numbers, with slightly higher deliveries and slightly lower ASP. @trentbridge, what is your margin assumption?

20% Model S/X - I think the production costs of the Raven are cheaper because the motors are now made in Nevada - and there's no inventory hang of Model S/X 75s to be discounted.
19.6% Model 3 - slightly higher than Q2.
14% Energy Storage - think the supply of batteries has increased and is more stable allowing greater production of Powerwalls etc.
-20% services - I think the repair/service shop business is growing, CPO is under better inventory control, and that Tesla is selling a boatload of accessories..

 

[ReflexFunds]

just noticing today's news about end of quarter incentive changes.
Tesla changes end-of-quarter incentives after some 'abuse' - Electrek

i find it interesting we have this end of quarter push/scramble sort of thing happening again. and the incentive structure makes me wonder if tesla made the same forecasting error that many analysts would make. they wouldn't have a way to know until some of september had passed.

I posted the below in the main thread, - I wouldn't read much into end of Q discounts until Tesla unwinds the delivery wave/ends the end of Q inventory reduction push. If anything Tesla reversing on discounts over 10 days before end of Q seems positive.

There are two parts to the end of quarter discounts.
1) Every quarter Tesla predicts demand for each option at each region and then ships/trucks the cars.
This cannot be done perfectly so towards the end of the quarter there is a mismatch at certain service centres/regions between supply and demand for certain options.
In each specific region there are some unsold cars with a certain set of options, while there is also an order backlog of cars with different options which have not yet arrived.
At the end of each quarter Tesla tries to balance this out and clear as much inventory as possible in each region. They do this by offering discounts on the remaining cars and sometimes use discounts to persuade customers to purchase a car from currently available stock rather than wait for their preferred option to arrive.
2) Many customers now know there is a chance of getting a small discount or some free options at the end of each quarter, so often customers now wait until towards end of Q to order their cars. To meet sales targets Tesla introduces some discounts again for the final few weeks of the quarter to push these customers into making their order with enough time remaining to receive delivery before end of Q.

sounds completely loony at first, but it's an interesting alternative take. china is currently doing (my estimate 2-3k model 3's per month). would there be demand for 4x that amount somewhat lower prices?

Definitely there is a question mark on exactly what demand level Tesla can reach in China in such a weak auto market, but it is not somewhat lower prices - it is a gigantic reduction in prices.
Made in China cars will benefit from 40% tariff removal (or 15% until December before tariffs are raised) and 6.4% lower VAT (2.4% until December), together with new access to up to $6k local and national EV subsidies and also the 10% purchase tax exemption which all Tesla's have recently been given. The initial prices of Made in China cars are a bit higher to pay for the production ramp (as they did in US), but these will go down as Tesla get to volume production.
I think it is highly likely Tesla adds to this price demand driver with its first proper advertising. In China people don't have access to Elon's Twitter feed so it makes sense for this to be the first market Tesla experiments with paid advertising.

 

[Doggydogworld]

@Reinhard sounds completely loony at first, but it's an interesting alternative take. china is currently doing (my estimate 2-3k model 3's per month). would there be demand for 4x that amount somewhat lower prices?

It's more than just lower prices, it's lower prices that everyone knows are coming. Just as an expiring tax credit tends to pull demand forward, a looming price cut should push demand back. How much demand? Unfortunately I have no idea.

There's also the issue of Made in China vs. imported. Tesla bears claim wealthy Chinese love to flaunt their expensive imported goods and look down their noses at the hoi polloi who buy domestic. Bulls claim the opposite - that Chinese in the age of Trump are all burning their imported products and lining up to support local producers. Nobody has data, they're just talking their book.

 

[EVNow]

It's more than just lower prices, it's lower prices that everyone knows are coming. Just as an expiring tax credit tends to pull demand forward, a looming price cut should push demand back. How much demand? Unfortunately I have no idea.

There's also the issue of Made in China vs. imported. Tesla bears claim wealthy Chinese love to flaunt their expensive imported goods and look down their noses at the hoi polloi who buy domestic. Bulls claim the opposite - that Chinese in the age of Trump are all burning their imported products and lining up to support local producers. Nobody has data, they're just talking their book.

First, Tesla will still be importing higher end trims. So, people who want to flaunt still have the opportunity.

Second, even if made locally Tesla is still "foreign". The same way BMW is foreign. Nobody is going to confused Model 3 with a BYD car.

ps : Even now in US BMW, Toyota, Nissan etc are still "imports" and only GM/Ford/Chrysler are "domestics".

pps : We can figure out the demand in China by looking at the volume of similarly priced models. How many BMW 3s are sold ?

 

[owlafaye]

Model Y cannibalizes Model 3 sales without providing a significant boost. Lack of resources to give an aging Model S a full interior and exterior refresh causes sales to further decline to a rate similar to other luxury car designs that are >5 years old. The massive boost in European sales because of the Netherlands and the UK ends with Netherlands incentive ending Jan 1st and novelty factor ending in the UK. China goes into a deep recession because of the trade war and Chinese customers can no longer purchase anywhere near 3,000 Model 3s/week. Large fixed costs of Gigafactory 3 cannot be recouped as a result.

The US-China trade war is just a game of chess. There will be an amiable accord and then there will be other tariff complaints...business as usual. I think people are reacting a little much to all this trade "war" talk. The Chinese will buy every Model 3 offered for the next few years. It will be quite the parade.

 

[30seconds]

I thought we were running away from the Alien Zombies.

I think it is super important that Tesla is able to shift the product mix of 3/Y - like they can apparently do with S/X.

We certainly don't know how much of the 3 demand will shift to Y.
​

GF3 will also allow to shift a bit more on a global basis. Although I expect that EU & US sales will be limited by 3 & Y output at GF1 until EU GF is running.

 

[EVNow]

BMW 3-Series L China auto sales figures

BMW sells about 10k per month of 3 series in China. If we assume similar figures for Model 3, we are looking at ~2.5k/week.

Largest plugin in China sells about 8k per month. So, 10k per month is not that far fetched.

EV Sales: China July 2019

I expect Tesla to only produce ~1k per week in the beginning in Q1 '20, though. Gradually increasing to meet demand. They should be able to dial back production in case of lower demand.