Troy Teslike

Q4 2024 Tesla Production & Delivery Estimates, Update #3

3 天前

16 December 2024, Update #3:

Hi everyone. Based on the latest data, my Q4 delivery estimate has increased to 498,000 units, up from 488,000 in the last update, mainly due to stronger-than-expected sales in China.

Tesla is approaching the 500,000 mark and has a real chance of surpassing it. If they do, it will be the first time they deliver 500,000 cars in a single quarter, breaking the current record of 484,507 units set in Q4 last year.

However, even hitting 500,000 wouldn't be enough to avoid a year-over-year decline. Tesla would need 514,925 deliveries this quarter to surpass last year's total, though most Tesla followers would still view 500,000 as a solid result.

Looking Ahead to 2025:

2025 is shaping up to be even more challenging as Trump plans to eliminate the $7,500 federal tax credit for electric vehicles. This change would effectively increase the price of all Model 3 and Model Y cars in the U.S. by $7,500 for most buyers. Additionally, Trump is aiming to end requirements for automakers to produce a certain percentage of Zero Emission Vehicles (ZEVs), which could affect the regulatory credits Tesla currently sells to other automakers.

Tesla Delivery Estimates for Q4 2024:

My Estimate: 498,000
Analyst Consensus: 511,000

Spreadsheet:

Feel free to check out the spreadsheet. I just updated it today with the latest data. Here are the links:

Delivery Table: This tab shows the 2023 and 2024 delivery tables side by side. There are three versions of each table: one grouped by model and then by factory, another grouped by factory and then by model, and a simpler table that doesn’t include the factory.
Delivery Charts: This tab features bar charts for deliveries in each region to provide context for the current quarter.
Production Table: This tab shows the 2023 and 2024 production tables side by side for easy comparison.
Production Charts: This tab includes bar charts for production numbers by model and factory.

Changes Since the Last Update:

My delivery estimate for Q4 has increased by 10,000 units, even though my estimate for U.S. deliveries dropped by more than 8,000 units.

The first table shows my previous estimate from November 24, 2024, the second shows my current estimate, and the third one shows the differences.

USA: I reduced my estimate by 8,398 units due to the latest production and inventory data.

Europe: I increased my estimate by 2,494 units based on November registration numbers from all countries. I’m also monitoring the daily and weekly December registrations from most European countries.

China: I increased my estimate by 14,553 units based on the latest weekly car insurance data.

Analyst Consensus:

The analyst consensus currently stands at 511,000, but this number doesn't accurately predict where Tesla's actual deliveries will end up. Tesla always aims to exceed the consensus, though they don’t always succeed. For instance, here are four quarters where the analyst consensus was notably higher than actual deliveries:

Earlier in the quarter, there was a real chance that deliveries could fall well below the consensus. Fortunately, stronger-than-expected sales in China, thanks to Tesla's recent promotion, have improved the outlook.

China Sales:

On 25 November 2024, Tesla China introduced a ¥10,000 Yuan ($1,374 USD) discount for the Model Y. See the news article here. This came on top of the existing ¥20,000 Yuan "Cash for Clunkers" government incentive for trading in an old gasoline car for a new EV. Both incentives expire at the end of December. Tesla is also offering 5-year, 0% financing with no money down for Model 3 and Model Y purchases in China.

This ¥10,000 incentive was announced one day after my last update on November 24. So my estimate was too low. As a result, I’ve increased my estimates.

The table below shows weekly Tesla sales in China, based on car insurance data. The website here publishes them but they are available on Twitter a few hours earlier every Tuesday morning. The last two columns show what I was expecting in the previous update and what I expect now.

I had estimated 17,300 units for weeks 9 and 10, but the actual numbers came in at 18,730 and 21,930. As a result, I’ve increased my estimates for the remaining weeks from 17,300 to 21,200 units. The next weekly number will be available tomorrow, December 17. 21,200 units would be as expected.

Here’s my monthly table for China. It includes monthly numbers released by the China Passenger Car Association (CPCA). Domestic deliveries were 40,485 in October and 73,490 in November. Production was 76,622 in October and 87,087 in November.

For Q4, I estimate 210,231 domestic deliveries for Model 3 and Model Y combined, which would be a new record. The previous record was 181,883 in Q3 2024. These numbers only include Model 3 and Model Y, not Model S or X.

Including Model S/X, the numbers are expected to be as follows. The first chart shows quarterly and the second one shows yearly deliveries. You can see how strong this quarter is compared to other quarters.

That said, yearly sales growth has slowed. This year, with 672,000 total sales compared to 610,000 last year, Tesla's growth rate is only 10%. That’s a big drop from 39% growth in 2023 and 37% in 2022. Next year could be even tougher with more competition in the EV market.

Tesla Model 3 vs Xiaomi SU7 sales:

Xiaomi is a Chinese manufacturer best known for its mobile phones and home electronics. The brand is also well-known in Europe, and the quality of their mobile phones is pretty good. It's similar to Samsung.

They've decided to enter the electric vehicle market, and their first model, the SU7, has been quite successful. Similar to the Model 3 in many aspects, the SU7 has a sportier exterior and a well-equipped interior. Sales of the SU7 began in April this year, so I'm comparing the sales numbers in China from that point onward.

Xiaomi has announced plans to launch another car, the YU7, which is expected to compete with the Model Y, with deliveries expected to begin in July 2025.

This shows the EV market in China could get tougher for Tesla next year. Sadly, Tesla has been slow at updating their cars or bringing out new ones. For example, the Xiaomi SU7 has Giga Casting, a structural battery pack, a head-up display, and a dashboard, but the Model 3 doesn't have any of these. Tesla has been planning to use Giga Casting for the Model 3 but hasn't done it yet.

U.S. Sales:

Tesla is struggling with demand in the US, especially for the Model Y and Cybertruck.

Production at the Giga Texas Factory:

DMV VIN data shows a 56% drop in Cybertruck production rates after November 10.

I estimate Tesla will produce 13,436 Cybertrucks this quarter, down from 14,774 in Q3. Tesla appears to be scaling back production due to weaker demand, which is disappointing because there was a huge reservation list for the Cybertruck. It seems that only a small fraction of those reservations have converted into actual orders. The $80,000 starting price, less range than announced at the reveal event and the loss of the $7,500 tax credit appear to have reduced demand significantly.

Initially, the Cybertruck was expected to qualify for the $7,500 EV tax credit when Tesla made the regular version available at $79,990 in early October. However, for some unknown reason, it has not been approved. The official list of qualifying Teslas on the government website here was last updated on December 4th, and it still doesn’t include the Cybertruck.

Some reservation holders are still hoping the tax credit will be approved, but I think it's unlikely. More likely is that Trump will eliminate the EV tax credit entirely with an executive order on January 20, 2025. Whether this works is unclear since Congress usually needs to approve such a change, but that might not happen until late March 2025.

Either way, demand for the Cybertruck looks weak, and sales could be less than 40,000 units next year.

Model Y Production at Giga Texas:

Unlike the Cybertruck, there’s no sign of a mid-quarter slowdown in Model Y production. Instead, VIN data suggests a consistently low production rate throughout the quarter.

I estimate Giga Texas will produce 22,500 Model Y in Q4, less than half of what was produced in Q1.

Production at the Fremont Factory:

I don’t see anything unusual when looking at the VIN data for Model 3 and Model Y production at the Fremont factory.

The Model 3 and Model Y numbers seem normal to me. The most interesting part is that Model S/X production keeps declining.

Overall, the US market looks weak this quarter and I expect deliveries to be slightly down compared to Q3.

This is surprising since Q4 is usually Tesla’s strongest quarter for U.S. sales. Typically, there’s a 20% bump in Q4 compared to Q3. But in 2023, that increase fell to just 9%, and this year, we may see an actual decline.

If we look at Tesla's yearly sales in the US, it doesn't look great either because this year will be the first time we see a decline in US sales. The quarter isn’t over yet, and sales could improve a little, but they’re unlikely to recover enough to match 2023.

Outlook for 2025

Next year’s US sales might be even more challenging because Trump is expected to cancel the $7,500 EV tax credit. When that happens, all Model 3/Y versions will effectively become $7,500 more expensive.

Based on DMV data, the average sale price of a Model 3 or Model Y in the US was $47,378 in Q3. Buyers currently get an instant $7,500 tax credit at the point of sale. Not all buyers qualify due to income limits, but around 84% do. On average, this means buyers pay $41,078 (47,378 - 7,500 x 0.84) right now. If the tax credit is canceled, the price goes back up to $47,378.

I asked Grok to estimate how much sales would drop in this scenario, and it projected an 18.4% decline. For context, about 537,000 Model 3/Y vehicles and 25,000 Model X Long Range AWD qualify for the tax credit this year, totaling 562,000 vehicles. A sales drop of 18.4% would mean a loss of about 103,408 units per year (562,000 x 0.184).

However, it’s possible that Congress’s approval would be needed to cancel the tax credit. If that’s the case, the credit might still be available for Q1 2025, and only three of the four quarters would be affected. That would reduce the estimated loss to 77,556 units (103,408 x 0.75).

However, Tesla could offset some of this loss with a new compact crossover vehicle, a shorter version of the Model Y. In the last Earnings Call, Elon said the following:

Elon: "We are still on track to deliver our affordable models starting in the first half of 2025"

Source: Tesla's Earnings Call on 23 Oct 2024. You can listen to that specific section here.

"The first half of 2025” is a bit vague. It could mean production starts in early April, or it could mean early July. Also, based on the latest rumors, this new vehicle is expected to start at $37,490 without the tax credit.

For comparison, the Model Y currently starts at $44,990, and the Model 3 starts at $42,490 in the US. Neither model has a Standard Range version anymore because those versions used LFP batteries from China that didn’t qualify for the tax credit. So, Tesla discontinued them.

The Model 3 Standard Range was previously priced at $38,990. The new compact vehicle is expected to start at $37,490, meaning that when Elon refers to "more affordable models," he’s talking about a $1,500 price reduction.

I’m not sure how much that’s going to boost sales. But it’s possible there’s a segment of buyers who don’t like the Model 3 or Model Y but might prefer a shorter version of the Model Y. So, let’s hope US sales don’t keep declining next year.

The Potential Cancelation of Regulatory Credits in the US:

In a recent interview published a week ago on YouTube, Donald Trump said he would end the Electric Vehicle mandate on January 20, 2025, with an executive order. He’s talking about Tailpipe Emission Standards, which might affect Tesla's regulatory credits. Tesla benefits from selling Zero Emission Vehicle (ZEV) credits to other automakers who fail to meet clean vehicle sales quotas.

ZEV credits are regulated at the state level. California has a unique position due to its Clean Air Act, which sets stricter emissions standards than the federal government. Other states can choose to follow California’s ZEV program. Currently, 16 states participate in the ZEV program.

Historically, the federal government and California have had an agreement allowing California to regulate its own ZEV program. However, the Trump administration might challenge this arrangement, potentially disrupting the current system of vehicle emissions regulations and the associated credit market.

I looked into how Tesla's earnings might be affected if regulatory credits in the US get canceled. Normally, it’s hard to calculate this without knowing where Tesla earns most of its regulatory credits. I would have assumed they made most of it in the US, but apparently, that’s not the case. In the last earnings call, Tesla’s CFO provided rare insight on this:

Vaibhav Taneja (Chief Financial Officer): We continue to see elevated levels of regulatory credit sales with over $2 billion of revenues so far this year. To expand on this at an industry level, China continues to outperform U.S. and Europe by a factor of three.

A three-to-one ratio means that 75% of the credits come from China, with the remaining 25% coming from the US and Europe combined. Let’s assume Tesla loses 25% of its regulatory credits. If this happens, Tesla’s earnings per share (EPS) would be $2.22 over the last four quarters, down from $2.40, a 7.5% decline.

4680 Cell Costs:

Tesla currently gets a tax credit for producing 4680 cells in the U.S. under the Advanced Manufacturing Tax Credit. The credit provides $45 per kWh for U.S.-made solar and battery products. More details about it can be found here.

The Cybertruck’s battery is estimated to be around 122.4 kWh based on EPA data. This means the $45/kWh tax credit saves Tesla $5,508 per Cybertruck. However, if Trump and Congress eliminate this tax credit, the Cybertruck’s cost would increase by $5,508.

Based on my calculation, the Cybertruck Long Range AWD version, which Tesla currently sells for $80,000, costs them $99,000 to build. If the manufacturing tax credit is canceled, production costs would rise to $104,500.

Europe Sales:

Here is my detailed table for Europe. The December numbers are my estimates but most of these countries report daily or weekly registration numbers. Therefore, we already have a lot of data for December.

Germany is expected to have the biggest drop in sales, down 26,413 units from last year. France is next with a drop of 22,530 units. Meanwhile, the Netherlands is expected to have the biggest increase, with sales up by about 9,353 units.

In terms of quarterly sales, this quarter is expected to be stronger than both Q2 and Q3:

In terms of full-year sales, however, this year is expected to end lower than last year:

The data quality for Europe is very good. Therefore, I'm not too worried about estimating Europe sales accurately. Here’s my error rate for Europe over the past 12 quarters. My errors were bigger in the earlier quarters, but the last 6 quarters have been more accurate.

Sales in the ROW (Rest of the World)

Here’s my current calculation for quarterly and yearly sales in this region:

Here is a detailed table that shows the sales numbers for each country in this region. The Q4 numbers are my estimates, while the numbers for previous quarters are the actual numbers reported by each country.

The Q4 estimates are based on October and November registration numbers and how they compare to past trends. Unfortunately, we get almost no data for December for this region before the month ends, which makes it more difficult to estimate than Europe.

Yearly Production Table:

Here is the latest production table for 2024. You can see this next to the 2023 table on the spreadsheet here. My production estimate for Q4 is now 462,175 units.

My estimate for total production this year is 1,776,173 units, compared to 1,845,985 units produced last year. That's 69,812 units or 3.8% less. Most of this drop is due to lower production at the Fremont factory.

Yearly Delivery Table:

Here is the latest delivery table for this year. My estimate for Q4 is now 498,000 units. If you want to see this table and the 2023 tables side by side, you can open the spreadsheet and scroll down.

Here is a more detailed version of the same delivery table that shows the originating factory for deliveries. For example, Canada was getting a lot of Model 3 and Model Y from Giga Shanghai but they switched to Fremont now.

Here is a detailed look at how deliveries this year in each region compare to last year:

Notes:

The Google Sheet for today's update is available here: Q4 2024 Tesla Production & Delivery Estimates.
If you find these updates helpful, feel free to recommend them to others. I appreciate your support.
This is Patreon-only content. Please don't share anything unless I have tweeted it. Thanks.
There is a public page here with lots of information about my error rate in the past. I'm trying to be as transparent as possible about my error rate.

That's all for now. Thank you!

Troy

2天

Hi Troy, as per this article Tesla sold 16,692 Cybertrucks in Q3 https://www.forbes.com/sites/brookecrothers/2024/10/15/tesla-cybertruck-sales-surge-in-q3-as-used-price-crashes/ but your spreadsheet is showing only 10,750 Cybertruck for Q3. Any idea? Or, am I reading the data incorrectly?

2天

Although, your sum of Model S/X + Cybertruck of 22,915 matches Tesla's Q3 slidedeck of Other models. I know Tesla doesn't give a breakdown between S/X/CT. So, my best guess is your breakdown is too conservative for CT and generous for S/X.

3天

Excellent work Troy!

3天

Troy, Great work. How confident are you with your estimate for ending inventory in the U.S.? How much do you think it could be off, either too high or too low?

创作者

3天

Inventory in the U.S. is dropping fast. There was a big drop in the last 48 hours. See the chart here: https://tesla-info.com/blog/inventory-stats-regional/US?q=2 I could increase US sales by up 5,000 units more assuming all Model 3/Y inventory will be gone. However, the weekly China number released today was 2,748 units lower than I expected, and there are still two more weeks to go. So, the two changes cancel out each other and 498K still looks very good. I plan to post two more updates. One will be on December 31st because we’ll get a new weekly China number that day. Another will be in a week and I’ll also get new VIN data for that update. The big unknown is Cybertruck deliveries in the US.

3天

Hello Troy, I think Europe is a bit too high as many countries just shut down during Christmas And why France is so high after that incentives were cut at the end of November? https://media.roole.fr/transition/voiture-propre/bonus-ecologique-ce-qui-va-changer-en-2024#:~:text=Le%20d%C3%A9cret%20fixant%20les%20nouvelles,000%20euros%20en%20jusque%2Dl%C3%A0.

3天

Hello Troy, thank you for the great analysis! 2 points of criticism! 1. History suggests a sharp drop in deliveries in China from week 49/50... 2023 was an exception! 2. According to my information, the dealine for the China subsidy is January 15, 2025! My guess is that weeks 51-52 are more realistic at 9,000 - 15,000! There is 0 order lead time! Europe: I think that in France the subsidy cap will push small car purchases in 12/24, the Netherlands and Belgium will perhaps increase by 50%! I see more of a drop in the UK, as 11/24 was only the balance to 10/24! If you adjust for the peak in the UK, 30,000 deliveries in Europe cannot be beaten in my opinion!

3天

Uk is already higher than Dec. About the Belgium and the Netherlands was not there that incentives disappeared in December?

3天

Thanks Troy! Tesla China deliveries were just out at 18,452. Does that change your view of China in terms of expecting 21k per week?

创作者

3天

I checked the numbers released today. 496,000 to 499,000 looks most likely. Model Y inventory in the US is going down rapidly. See https://tesla-info.com/blog/inventory-stats-regional/US?q=2 The biggest challenge is estimating Cybertruck deliveries in the US and Canada.

3天

I doubt the new model will spike significant sales. It will probably cannibalize Model 3 / Model Y sales again. Not the prettiest car!

3天

Troy, regarding ZEV credits and the CFO comment, I think it means US is 1, EU is 1 and China is 3 for China being 3/5 or 60% with rounding could be just ovver 50%.

创作者

3天

I asked ChatGPT and Grok, and they both split it as 75% for China and 25% for the rest based on this quote: "We continue to see elevated levels of regulatory credit sales with over $2 billion of revenues so far this year. To expand on this at an industry level, China continues to outperform U.S. and Europe by a factor of three."

3天

RoW sales have been running in place since Q2 2023, and have really fallen off in Australia, New Zealand, Taiwan and South Korea while being offset by jumps in other markets. Could Tesla see falling RoW sales in 2025?

3天

With a PE of 125 and a CEO that is doing various side-quests, I'm not so sure that Tesla's current valuation is accurate, but public opinion makes it sound like it is. Thank you Troy for your down to earth analysis! "You're neither right nor wrong because other people agree with you. You're right because your facts are right and your reasoning is right"

Q3 2024

My error rate for deliveries was OK at 2.0%, but my production estimate was off by 3.5% this time, which is the highest error rate I’ve had for production.

Deliveries this quarter were pretty predictable and never went off by more than +-4% at any point during the quarter.

-----------------------------❖-----------------------------

Q2 2024

My error rate was low for production but high for deliveries.

Deliveries were higher than I expected.

The first table shows the actual deliveries in Q2 2024 based on all the data we have so far (the data for Canada hasn't been released yet). The second table shows my estimate as of June 30th. The last one shows the difference:

Deliveries were high in the US because Tesla reduced inventory more than I expected. Deliveries were also high in the Rest-of-the-World region due to strong sales in a few countries.

-----------------------------❖-----------------------------

Q1 2024

This was a difficult quarter because Tesla reported much fewer deliveries than expected. I had the lowest estimate for deliveries but it still turned out to be too high.

Estimates were high early in the quarter because Tesla had just delivered 484,507 cars in Q4 2023 and Q1 2024 was expected to be higher than that because Q1 in 2021, 2022, and 2023 was higher than the previous quarter. Unfortunately, Tesla delivered only 386,810 cars this quarter.

After day 50, it became clear that deliveries were going to be low. So, I started to lower my estimate. Unfortunately, I still ended up 5.7% too high. Analyst consensus on March 31st was 11.5% too high.

Here is a table that shows my error rate versus other estimates. I had the lowest error rate but I wasn't happy with the results.

-----------------------------❖-----------------------------

Q4 2023:

This was another good quarter for me in terms of accuracy. My error rate for my final estimate was 2.2% for production and -1.1% for deliveries. I'm happy with that.

Q4 2023 was a predictable quarter. Tesla's target was 1.8 million deliveries in 2023, requiring 475,926 in Q4. Throughout the quarter, estimates were slightly higher. Towards the end, it seemed Tesla might miss the target, but a strong end-of-quarter push resulted in 484,507 deliveries for the quarter and 1,808,581 for the year.

-----------------------------❖-----------------------------

Q3 2023:

This was a good quarter for me in terms of accuracy. My error rate for my final estimate was -0.5% for production and +1.4% for deliveries. I'm happy with that. Analyst consensus was off by +4.5%

Initially, on day 22 of the quarter, my estimate was too high by +7.3% compared to where the quarter would finish but we didn't know that at the time. Then after day 40, I started to lower my delivery estimate.

I continued to lower it gradually and I finished the quarter only +1.4% higher than the actual number. Most analysts were too slow to lower their estimates and the analyst consensus at the end of the quarter was 454,809 units according to Tesla's company compiled consensus. This ended up being too high by 19,750 units or +4.5%.

When the results were announced, I did get a mention by CNBC on their website here which was nice to see. Actual deliveries in Q3 2023 were 435,059 units.

-----------------------------❖-----------------------------

Q2 2023:

This was not a great quarter for me. My first delivery estimate turned out to be 8% too low. -8% is actually not too bad if we look at some of the other quarters but analysts were only -4.7% off early in the quarter. Therefore, in this case, -8% is not great.

My final estimate was off by -3.9% and I was not happy with that because I'm aiming for a +/-3% error for deliveries. This is outside of that target range.

My production estimate was more accurate with a -1.7% error and I was happy with that because it's usually less than +/-2% and I want to keep it that way.

At the end of the quarter, I finished slightly higher than analysts and there was less up and down fluctuation throughout the quarter but generally speaking, this was not the best quarter for me.

However, it was of course a great quarter for Tesla because deliveries exceeded expectations and that's great.

-----------------------------❖-----------------------------

Q1 2023:

This was a good quarter in terms of my error rate for deliveries. I was happy with the 1.0% error. An interesting development at the end of the quarter was that, CNBC included my final estimate in a news article when the results came out.

This quarter was more predictable than other quarters. Throughout the quarter, my estimate fluctuated between +4% and -3% of the actual number. That's actually a tight range compared to other quarters.

-----------------------------❖-----------------------------

Q4 2022:

After we have the actual numbers, I always post my error rate publicly as well. Here is what happened after Q4 2022. I tweeted the results here.

-----------------------------❖-----------------------------

Q3 2022:

Here is what happened after Q3 2022. I tweeted about it here. I was happy with +1.7% error for production but my delivery estimate was off by +5.6%. This was the largest error I made in recent quarters.

Tesla had some delivery challenges and they couldn't deliver all cars they produced. Therefore there was a large gap between production and deliveries.

Here is my error rate throughout Q3 2022:

-----------------------------❖-----------------------------

Q2 2022:

Here is what happened after Q2 2022. Both my production and delivery estimates were less than 1% off. I was happy with that.

Here is my error rate throughout Q2 2022:

-----------------------------❖-----------------------------

Q1 2022:

After Q1 2022, I was happy with less than 1% error.

Here is my error rate throughout Q1 2022:

-----------------------------❖-----------------------------

Q4 2021:

Q4 2021 was a more difficult quarter to estimate. Analyst estimates were off by -14%. Therefore I was pretty happy with my -3.1% error and my work was featured in a Barrons article here on 3 January 2022. Of course, I was very happy with that too.

Here is a detailed view of Q4 2021:

Here is my error rate throughout Q4 2021:

-----------------------------❖-----------------------------

Data sources I use:

DMV data from a few states in the US. I purchase some of this data. The data includes Vehicle Identification Numbers (VINs). I use that to calculate production. Some of the DMV data includes the average sale price. I use it in my financial calculations. I calculate revenue, gross margin, EPS before each earnings release.
Historical data: I keep track of Tesla sales in all countries where Tesla sells cars and I track production per factory and model using VIN data. I created a public Google sheet for the Tesla community: Tesla Production & Delivery History
Inventory numbers: Tesla publishes global inventory as days of supply and I make sure my global inventory calculation matches Tesla's number but I also calculate inventory for each model produced at each factory.
Shipping data: This includes the number of ships and the capacity of each ship that carries exports.
Production data: I keep track of the following:
[1] The highest quarterly production achieved for each model at each Tesla factory
[2] Stress test scores
[3] Employee counts at each factory
[4] Improvements to production lines;
[5] The number of crews and shifts for each factory and production line
[6] Information released by Tesla's suppliers
[7] Production interruptions reported in the media or spotted through VIN data

-----------------------------❖-----------------------------

DMV VIN Data (Vehicle Identification Numbers):

Here’s what VIN data looks like when visualized. Each dot represents the vehicle identification number or VIN of a Tesla car. In this example, the chart shows Model 3 production at the Fremont Factory in California. The triangle on the left represents production in Q3 2023, while the one on the right shows Q4 2023.

As production continues through the quarter, Tesla assigns higher VINs to the cars they produce, creating a triangular pattern. The height of the triangle indicates the total production for the quarter.

The accuracy of this method is pretty good, and my production estimates are usually more accurate than my delivery estimates.

The accuracy is pretty good and my production estimates are typically more accurate than my delivery estimates.

-----------------------------❖-----------------------------

Here is what the data I collect for Asia-Pacific looks like:

Here is what the data I collect for China looks like:

I calculate inventory originating from each factory for each model. Here is what my inventory tables look like:

Inventory Table 1:

Inventory Table 2:

I regularly update my estimates throughout the quarter but I always post my final estimate on Twitter on the last day of the quarter to be transparent about my error rate. I'm obsessed with Tesla and love creating content about it. Thanks to my supporters on Patreon, I can dedicate lots of time to this, and I appreciate the support!

Thank you,
Troy Teslike, @TroyTeslike
2 October 2024

Q4 2024 Tesla Production & Delivery Estimates, Update #2

11月25日

24 November 2024, Update #2:

Hi everyone! Tesla has been struggling with deliveries this quarter, especially in the US. Based on the latest data, I’ve lowered my delivery estimate by 8,000 units, from 496,000 to 488,000. While my estimate for China has increased by 12,000 units, my estimates for the US and Europe have dropped by 20,000 units in total, resulting in a net decrease of 8,000 compared to my previous update.

The biggest challenge Tesla is facing right now is Cybertruck deliveries in the US. The Cybertruck was expected to qualify for the $7,500 tax credit last month, but it still hasn’t been approved, and as a result, most reservation holders are holding off on taking delivery.

Unfortunately, the data so far for U.S. and European sales doesn't look promising, and the downside risk seems greater than the upside potential.

In the last shareholder letter on 23 October 2024, Tesla said, "We expect to achieve slight growth in vehicle deliveries in 2024" (source: page 10). To achieve this, deliveries this year would need to exceed the 1,808,581 units delivered last year, which means deliveries this quarter would need to exceed 514,925 units.

However, my current estimate for this quarter is 488,000 deliveries, which is 27,000 short of the target. That's not a small gap to close in the remaining five weeks. Moreover, my estimate includes several optimistic assumptions that might not happen.

Tesla Delivery Estimates for Q4 2024:

My Estimate: 488,000
Analyst Consensus: 509,000

Analyst Consensus

The analyst consensus for Q4 deliveries is now 509,000 units, up slightly from 508,000 units a month ago. This is 21,000 units higher than my estimate of 488,000, which is a bit unusual—usually, the two estimates are much closer.

That said, we've seen something similar before, back in Q1 2024. That quarter turned out to be a tough one for Tesla, as deliveries came in much lower than expected.

If you look at this chart, you'll see that by day 59 of Q1 2024, I started to pull away from the analyst consensus, as my delivery estimate dropped significantly. The analysts were slower to adjust their estimates—by day 82, their forecast was still 82,000 units higher than what actually happened.

If we look at the final estimates for Q1 2024, these were the numbers at the end of March 2024:

My final Q1 estimate ended up being 22,190 units too high, but the analyst consensus was even further off. I’m bringing this up again because I see some similarities between Q1 and Q4, though on a smaller scale.

For example, let’s say Tesla finishes this quarter with 472,000 deliveries. In this case, let’s assume my final estimate is 478,000 units and the analyst consensus is 490,000 units. That would mean my estimate was off by 6,000 units, and the analyst consensus would be 18,000 units too high. I wouldn’t be surprised if we end up seeing something like that.

Full-Year Estimates:

The analyst consensus for the full year is now 1,803,000 deliveries. My estimate is 1,782,000 units. Both numbers are below the 1,808,581 units Tesla delivered last year.

Spreadsheet:

Feel free to check out the spreadsheet. I just updated it today with the latest data. Here are the links:

Delivery Table: This tab shows the 2023 and 2024 delivery tables side by side. There are three versions of each table. One is grouped first by model, then by the factory. The other is grouped first by factory, then by model and there is a simpler table that doesn't show the factory.
Delivery Charts: This tab shows bar charts for deliveries in each region to provide context for the current quarter.
Production Table: This tab shows the 2023 and 2024 production tables side by side for easy comparison.
Production Charts: This tab shows bar charts for production numbers for each model at each factory.

Possible Scenarios for Tesla Q4 Deliveries:

This chart shows four different scenarios for Q4 deliveries:

Currently, Scenario 2 looks most likely. In other words, deliveries this quarter are expected to be between 484,507 and 500,000 units. The 484,507 number is important because that's how many vehicles Tesla delivered in Q4 last year and it also happens to be the current record quarter.

Therefore, if Tesla exceeds that number, it will set a new quarterly delivery record. However, Scenario 2 would still be disappointing since it falls short of the 514,925 units needed to avoid a decline in annual deliveries. It would also miss the symbolic milestone of 500,000 units.

I added the 470,000 figure to the chart. In the previous version, Scenario 1 was simply for deliveries less than 484,507. Now, Scenario 1 is between 470,000 and 484,507 deliveries. The reason is that I have looked at the worst-case scenario and if everything goes wrong, we could see 470,000 deliveries.

That said, I don’t expect this to happen. While some might view my 488,000 estimate as too conservative, it’s actually based on a set of optimistic assumptions. In fact, I consider it a relatively optimistic estimate.

Optimistic Assumptions included in my 488,000 Estimate:

The first table shows a hypothetical worst-case scenario of 470,000 deliveries, which could occur if things continue as they are and no improvements are made. The second table shows my current 488,000 estimate and the third one shows the difference.

USA: I'm assuming 9,729 more sales than what the data currently suggests.

Registration data for October shows very low Cybertruck deliveries in the US compared to July. It's always best to compare the first month of the current quarter to the first month of other quarters. Therefore, we are comparing October 2024 to July 2024.

This is most likely because many Cybertruck buyers were waiting for the $7,500 tax credit to be approved in October which never happened. Some of these buyers will likely stop waiting and take delivery anyway in November. Therefore, I expect an improvement in Cybertruck deliveries in November, and I will have some data for November later in the quarter.

For the Model 3 and Model Y, I’m assuming that Tesla will be able to clear out all inventory in the US by the end of the quarter. This means I expect the inventory levels for both models, as shown on this chart to drop to around 1,000 units each by the end of December. The chart reflects the number of new inventory listings, but based on my calculations, the actual inventory is roughly four times higher. This is because Tesla doesn’t create separate listings for identical configurations at the same delivery center.

Europe: October registrations in Europe were lower than in April or July, but my current estimate for Q4 is higher than Q3. To support that estimate, we would need to see 21,800 registrations in November. We should have the November numbers by December 7th. You can follow the purple table here:

The purple table is part of a Tesla community project aimed at making Tesla’s European registration numbers easier to access. While registration data is publicly available in Europe, it’s published separately by each country, and someone has to collect and organize the Tesla numbers into a single table. A group of volunteers does this every month, and I’ve been part of this group since 2015.

In 2015, I was browsing Tesla forums and came across a group of Tesla enthusiasts sharing European registration numbers in forum posts. Each time there was more data from another country, somebody would post a forum message but it was too messy because the data changes all the time and you can't find all the numbers for the past months in one place. So, I offered to create a Google Sheet and that's how the purple table was born in 2015. Nine years later, we still use it every month. It’s a small side project I enjoy being a part of.

China: We get weekly car insurance registration numbers for 12 of the 13 weeks in the quarter, but we don’t get data for the final week in time. Unfortunately, the final week happens to be unpredictable. It's sometimes lower and sometimes higher than the previous weeks.

For example, in Q3 2024, week 13 was a whopping 8,800 units higher than week 12 but in Q2 it was 3,350 units lower. For this quarter, I assume it will be around 2,900 units higher but it's difficult to estimate.

My estimate for the next weekly China number that will be released Tuesday morning is 17,300 units. That would be as expected.

Cybertruck Delivery Challenges

Based on the latest data, Tesla had 8,986 unsold Cybertrucks in inventory at the end of Q3 2024. This shouldn't be a surprise because at the time Tesla was selling only the Foundation Series of the Cybertruck which had a starting price of $99,990 for the non-performance Long-Range AWD version and $119,990 for the Performance version.

Because there were not enough buyers, on 4 October 2024 Tesla opened up orders for the regular version with a $20,000 lower price.

In the US, the non-performance Long-Range AWD version of the Cybertruck is priced at $79,990, just under the $80,000 threshold required to qualify for the $7,500 federal tax credit. As a result, there was an expectation that this version would qualify for the credit.

However, despite this, the Cybertruck has yet to be approved for the tax credit. The reason for the delay is unclear, but the official list of qualifying Teslas on the government website (last updated on November 1st) still does not include the Cybertruck.

In addition to the 8,986 Cybertrucks carried over from Q3 to Q4, production is also higher this quarter. Based on VIN data, I estimate that 19,736 units will be produced this quarter, up from 14,774 in Q3. This means Tesla could theoretically have 28,722 Cybertrucks available for delivery in Q4.

However, most of the Cybertrucks produced in the final days of Q4 won't make it to delivery in time, and there are always some buyers who don't take delivery due to issues like cosmetic problems or payment complications. Taking these factors into account, Tesla could deliver about 27,000 Cybertrucks this quarter compared to 11,948 units in Q3, if they had enough buyers.

Unfortunately, Tesla is facing a shortage of buyers. In fact, the Cybertruck is now demand-limited. You can see this for yourself by visiting Tesla's order page here, where the Cybertruck is listed for immediate delivery in California and several other states. If you're in the US, you can scroll down and enter your zip code on that page to check delivery timing for your area.

We don’t know the exact size of the reservation backlog, but it may have been as high as 400,000 at one point. Now, it's down to zero, with Tesla having delivered only 25,000 Cybertrucks since production began.

Some have argued that the $7,500 tax credit issue shouldn’t impact Cybertruck or Tesla sales. However, that's unrealistic—how could it not affect sales if the vehicle is $7,500 more expensive?

While some buyers don’t qualify for the tax credit due to income limits, most people do. The income limits are as follows:

$300,000 for married couples filing jointly
$225,000 for heads of households
$150,000 for all other filers

Only 8% of households in California earn over $300,000 a year, according to this website:

Let’s assume Tesla buyers are more selective than the general population, so let's say 16% of potential buyers fall above the income limit and don't qualify for the $7,500 tax credit. This still leaves 84% of potential buyers eligible for the credit, making it clear that the $7,500 federal tax credit plays a significant role in purchase decisions.

While we don’t know exactly how many Cybertruck reservation holders Tesla had in early October, we do know that Tesla reached out to all of them when orders for the regular version opened on October 4, 2024. Despite these efforts, Tesla has run out of Cybertruck buyers.

Tesla's Production in the U.S. Based on VIN Data:

I use DMV VIN data (Vehicle Identification Numbers) to estimate Tesla's production in the U.S. Each dot you see on the chart represents a VIN.

The VINs form a triangular pattern because Tesla assigns higher VINs to new vehicles as production continues. Therefore, I extend the triangle until the end of the quarter to estimate where total production will end up if the same production rate continues. Of course, the production rate could change which is why I regularly get new VIN data to update my calculation.

Cybertruck Production at Giga Texas:

Based on the latest data, I estimate 19,736 Cybertruck production in Q4, up from 17,976 in my previous calculation.

The 19,736 units in Q4 would represent an increase of 4,962 units compared to 14,774 in Q3. What's most interesting about this increase is that to achieve it, Tesla seems to be reallocating employees from the Model Y production line at Giga Texas to the Cybertruck line, as the data shows a decrease in Model Y production.

Model Y Production at Giga Texas:

The data shows that production began at VIN 336,300 and is expected to reach VIN 360,800 by the end of December, which would represent 24,500 units. However, based on historical data from previous quarters, actual production tends to be lower than what the VIN data suggests. Therefore, I'm estimating a production total of 22,500 units in Q4.

Here are the production numbers for Giga Texas: Cybertruck production is up, while Model Y production is down. Overall, however, total production remains relatively stable, averaging just over 40,000 units in most quarters.

Model 3 Production at the Fremont Factory:

I'm calculating 52,800 production in Q4, which is less than Q3.

Cybertruck's Questionable Financial Future

I want to explain how we can determine how much it costs Tesla to produce the Cybertruck. We have DMV data that shows the average sale price for each model and trim level. This was the average sale price:

Additionally, Tesla said the Cybertruck was gross margin positive for the first time in Q3 2024 which means it had a negative margin in Q2 2024. Also, we know the unit sales for each trim level. The numbers were as follows:

Additionally, we know Cybertruck was barely profitable in Q3 2024 because Tesla pushed performance version deliveries from other quarters into Q3 2024 to achieve profitability. Also, looking at the hardware differences between the dual motor Long Range AWD version and the performance version with three motors, I estimate that the performance version costs Tesla $8,000 more to produce (they sell it for $20,000 more).

Based on these numbers, I have come up with the following Cost of Goods Sold (COGS) numbers for each trim level that makes the Cybertruck gross margin negative in Q2 but barely positive in Q3.

Based on my calculations, the COGS for the Foundation Series Long Range AWD Cybertruck was $106,400 in Q3, and it was not profitable, while the COGS for the performance version was $114,400, and it was profitable.

Tesla now sells the regular version in the U.S., and the Long Range AWD version should cost Tesla less than $106,400 to produce because the Foundation Series included a few extras for free, whereas the regular version doesn't have these.

I estimate the retail price for the options included with the Foundation Series was around $13,415 but the cost to Tesla was only $3,710 because FSD doesn't add extra hardware costs since the required hardware is already included in all Teslas, regardless of whether you buy FSD. However, I'm adding a $500 cost for FSD to account for the salaries of programmers working on FSD.

So, without the Foundation Series options, the cost of the regular version drops to 106,400 -3,710 = $102,690. However, there will be additional cost reductions due to higher volume and other improvements.

Therefore, in Q4, it’s likely that the Long Range AWD will cost Tesla approximately $99,000 to produce, with an average sale price of around $83,000. This means Tesla is expected to lose about $16,000 per Long Range AWD Cybertruck sold in the U.S.

The Performance version should cost Tesla around $107,000 to produce and the ASP should be around $103,000. So, I expect only a $4,000 loss for that version.

However, Tesla is selling only the Foundation Series in Canada in Q4. So, that will help with the financials, but the most important factor for Q4 financials I think will be Bitcoin sales. I believe Tesla has already sold some of its Bitcoins this quarter to offset the negative margins on the Cybertruck and the low margins on the Model 3/Y, which have been caused by 0% promotions and discounts.

Canada Sales:

We finally have the sales numbers for Canada for September (and October as well). In the last update, we didn't have September numbers. The Q3 numbers were lower than I expected which means, I increased my numbers for US sales in Q3.

The best way to calculate US sales for past quarters is by subtracting all other countries from Tesla's global numbers. We get the numbers for Europe and China a week after the quarter is over. A few countries in Asia Pacific take longer but Canada numbers always take the longest to be released.

Here is the updated table showing Tesla sales for the first three quarters, comparing last year and this year:

After three quarters, U.S. sales are down by 33,000 units. If it were not for the 22,000 Cybertruck deliveries, the U.S. would be down by 54,000 units.
Europe is down by 29,000 units.
China is up by 23,000, Canada by 3,000, and the Rest of the World by 6,000.
Global sales are down by 30,000 units.

Changes Since the Last Update:

My estimate for Q4 deliveries is down by 8,000 units even though my estimate for China deliveries is up by 12,577 units. Here is a closer look:

The first table shows my previous estimate on 31 October 2024, the second table shows my current estimate and the final table shows the difference.

USA: I reduced my estimate by 16,863 units. 8,551 of that is due to lower Cybertruck deliveries. My estimates for Model 3 and Model Y are down too and it's because of lower production and high inventory.

Canada: I increased my estimate for Cybertruck deliveries which compensated for the drop in Model 3 and Model Y. In other words, Cybertruck deliveries are expected to be higher than previously thought but Model 3 and Model Y are expected to be lower.

Europe: I reduced my estimate by 3,271 units because Tesla's sales performance in Europe has not been great so far in Q4. A few countries in Europe report daily registration numbers but they are not a good representation of the entire Europe. Sales in daily reporting countries increased in October but total sales in Europe were down. We'll have a much clearer picture once we receive the November registration numbers around December 7th.

I will calculate an estimate for each country in Europe once we have the November numbers.

China: I increased my estimate by 12,577 units based on the weekly can insurance numbers. Unfortunately, this increase was not enough to compensate for the drop in other regions.

Yearly Production Table:

Here is the latest production table for 2024. You can see this next to the 2023 table on the spreadsheet here. My production estimate for Q4 is now 466,505 units, down by about 10K compared to the last update.

My estimate for total production this year is 1,780,503 units, compared to 1,845,985 units produced last year, a decrease of 65,482 units.

Here is how production is expected to change compared to last year. The first table shows last year's total production, the second one shows this year and the final table shows the difference.

Fremont Factory: This year, production is expected to be down by 74,582 units compared to last year.

Giga Shanghai: I estimate that production will decrease by 36,367 units this year.

Giga Texas: Production is expected to be up by 38,964 units thanks to around 50,121 Cybertrucks produced this year. I hope Tesla confirms the 50,000th Cybertruck production milestone around December 30, 2024. It would help with calculations.

Giga Berlin: I estimate a minor increase of 6,503 units this year.

Global Production: I estimate a 65,482 units decline compared to last year.

Yearly Delivery Table:

Here is the latest delivery table for this year. My estimate for Q4 is 488,000 units, which is 21,000 units less than the 509,000 analyst consensus and 27,000 units less than the 515,000 units Tesla needs to avoid a decline in total yearly deliveries. If you want to see this table and the 2023 tables side by side, you can open the spreadsheet and scroll down.

Here is a more detailed version of the same delivery table. It shows the originating factory for deliveries. For example, you can see that in Q4, Canada is getting most of its Model 3s from the Fremont factory and not Giga Shanghai.

Here is how total deliveries this year are expected to change compared to last year. The first table shows last year's deliveries, the second one shows this year and the final table shows the difference.

USA: Deliveries are expected to be down by 39,000 units. If it were not for Cybertruck deliveries, deliveries would be down by 75,000 units which is probably a fairer comparison because there were no Cybertruck deliveries last year and sales of other models dropped sharply. However, this 75,000-unit drop will seem relatively small compared to the likely impact of Trump canceling the $7,500 federal tax credit. I expect a 25% decline in sales or 155,000 units.

Europe: Sales this year are expected to be down by 41,000 units, mostly due to poor Model Y sales but also a complete collapse of Model S/X sales from 9,000 units last year to 3,800 units this year, a 58% decline.

China: Sales are expected to be up by 47,000 units. Model S/X sales in China are also down by a lot from 7,000 units last year to 1,800 units this year, a 74% decline. The reason is unknown.

ROW (Rest of the World): Sales are expected to be up by 5,800 units, despite Model S/X sales dropping by 51% or 2,600 units.

What to Expect in 2025:

Tesla delivered 1.8 million vehicles last year, and this year’s total is expected to be slightly lower but still close to that number. Next year is still unclear because it all depends on what Tesla means by "more affordable models" expected to launch in the first half of 2025. These new vehicles will be built on the same production lines as the Model 3 and Model Y.

Elon: "We are still on track to deliver our affordable models starting in the first half of 2025"

Source: Tesla's Earnings Call on 23 Oct 2024. You can listen to that specific section here.

The best-case scenario would be a VW Golf-style hatchback version of the Model 3 or Model Y, available in Europe and China. It would be very popular in those markets.

The worst-case scenario is that "more affordable" might just mean upgrading the Model 3 with dual castings, a structural battery pack, and switching to 4680 cells. It would still look the same on the outside but would be cheaper to produce. If that's the case, we could be looking at around 1.7 million deliveries next year.

This estimate also assumes a 25% drop in US sales if Trump eliminates the $7,500 federal tax credit, which could result in Tesla losing about 155,000 US sales next year. That would reduce Tesla’s global deliveries from 1,782,000 to about 1,627,000 units. However, I assume China sales will continue to increase, So, 1.7 million is a realistic estimate for deliveries next year in this scenario.

On the other hand, if the "more affordable" models turn out to be a VW Golf-style hatchback, I believe Tesla could reach 1.9 million deliveries next year. Here’s why:

The highest production at Giga Shanghai was 90,802 units in November 2023. If they maintain full production next year, the total from Giga Shanghai would be 90,802 × 12 = 1,089,624 units.
Giga Berlin, operating at full capacity, can produce 375,000 units per year.

Together, these two factories could produce 1,464,624 vehicles in 2025, up from 1,109,513 this year, an increase of 355,000 units over four quarters.

However, since Elon mentioned "the first half of 2025," it’s unlikely the new vehicles will roll out in early January. Assuming production starts in Q2 2025, we’d need to adjust for only three-quarters of production next year. Taking 75% of the 355,000-unit increase gives us an improvement of 266,000 units for the year. So, that's 1,782,000 -155,000 +266,000 =1,893,000.

Summary

Without a VW Golf-style hatchback: I estimate 1.7 million deliveries next year due to the potential loss of the US tax credit.
With a VW Golf-style hatchback: I estimate 1.9 million deliveries, including 266,000 units of the new, more affordable vehicle.

Notes:

The Google Sheet for today's update is available here: Q4 2024 Tesla Production & Delivery Estimates.
If you find these updates helpful, feel free to recommend them to others. I appreciate your support.
This is Patreon-only content. Please don't share anything unless I have tweeted it. Thanks.
There is a public page here with lots of information about my error rate in the past. I'm trying to be as transparent as possible about my error rate.

That's all for now. Thank you!

Troy

Q4 2024 Tesla Production & Delivery Estimates, Update #1

11月1日

31 October 2024, Update #1:

Hey everyone! This is the first update for Q4 2024. Here is what to expect:

Tesla's Delivery Goals for Q4 2024:

In a recent shareholder letter, Tesla said, "We expect to achieve slight growth in vehicle deliveries in 2024" (source: page 10). To achieve growth, they must deliver over 514,925 vehicles this quarter. Before the earnings call, the analyst consensus for Q4 deliveries was 494,000. Now, it’s 508,000 units, but some analysts haven’t updated their estimates yet, so it’ll likely continue rising toward 515,000 units.

However, 515,000 deliveries this quarter is not guaranteed to happen. Even if Tesla sells out all its production and inventory of the Model 3, Model Y, and Cybertruck in the U.S., my numbers still only add up to 496,000 units.

To hit 515,000 deliveries, in addition to maximizing U.S. sales, Tesla would need around 202,000 sales in China in Q4. The current record is 182,206 sales in Q3, and unfortunately, we're trending below that so far. My current estimate assumes that China sales will pick up and match Q3 levels later in the quarter. We will have a better idea as we get more data.

Tesla Delivery Estimates for Q4 2024:

My Estimate: 496,000
Analyst Consensus: 508,000

Spreadsheet:

Feel free to check out the spreadsheet. I just updated it today with the latest data. Here are the links:

Delivery Table: This tab shows the 2023 and 2024 delivery tables side by side. There are three versions of each table. One is grouped first by model, then by the factory. The other is grouped first by factory, then by model and there is a simpler table that doesn't show the factory.
Delivery Charts: This tab shows bar charts for deliveries in each region to provide context for the current quarter.
Production Table: This tab shows the 2023 and 2024 production tables side by side for easy comparison.
Production Charts: This tab shows bar charts for production numbers for each model at each factory.

Four Scenarios for Tesla's Q4 Deliveries:

Tesla's Q4 delivery performance depends on which of these targets they can hit.

Scenario 1: Less than 484,507 deliveries

This would be very disappointing since Tesla delivered 484,507 vehicles in Q4 last year. Falling short of that leaves little room for positive discussion, and Tesla would miss all three criteria shown in the chart.

Scenario 2: 484,507 to 500,000 deliveries

This scenario would meet just one of the three criteria: it would set a new record for the quarter. However, it falls short in two areas: it’s below the 500,000 milestone, and total deliveries for the year would be lower than last year. Overall, this would still be disappointing.

Scenario 3: 500,000 to 514,925 deliveries

This scenario would fall short of expectations, as Tesla’s statement in the shareholder letter has raised the bar to at least 514,925 deliveries in Q4.

Tesla: "Despite ongoing macroeconomic conditions, we expect to achieve slight growth in vehicle deliveries in 2024". (Source: Page 10 of Q3 2024 Update)

That said, more than 500,000 deliveries isn’t too bad, since it would still set a new record (anything above 484,507 counts as a new record). It would also mark the first time Tesla crosses the 500,000 milestone. So, many Tesla fans might overlook a decline in yearly deliveries if Tesla manages to hit that milestone. However, if they end up below 500,000 deliveries, I think most people would see that as disappointing.

Scenario 4: More than 514,925 deliveries

This would be great. It’s the new expectation set by Tesla’s statement in the shareholder letter. While this scenario seemed unlikely until recently, Tesla has now launched a 0% financing promotion in the U.S. that doesn’t require purchasing the $8,000 FSD (Full Self Driving) option. As a result, U.S. sales are expected to rise. Plus, there’s an incentive deadline in China that should boost Q4 sales as well. So, this target is now looking achievable, though getting there will still be a challenge

Since Tesla deliveries were only 462,890 units in Q3 2024, you might be wondering why Q4 deliveries should be much higher than that. Fortunately, there are a few new factors in Q4 that should help drive sales. Let’s go over them.

Factors Expected to Positively Impact Q4 Sales Compared to Q3 2024:

1. Seasonality:

We don't see this effect in China, Canada, or Asia-Pacific but we see it clearly in the U.S. and Europe where Q4 sales are typically the highest of the year.

This chart shows sales in the U.S. The red columns show Q4. Over the past four years, Q4 was the strongest quarter each year except in 2023, when sales flattened.

Q4 sales in the U.S. increased over Q3 by:

17% in 2020,
20% in 2021,
24% in 2022,
5% in 2023,

For this year, excluding the Cybertruck, I estimate a 9% increase in Model S/X/3/Y sales in the U.S. over Q3. It's not higher than 9% because I’m already estimating Tesla will sell out all production and inventory, so I can’t increase my estimate unless production increases and I haven’t seen that reflected in the VIN data yet.

By the way, here is a chart that shows my delivery estimate for North America in Q4 2024 compared to past quarters. There are two interesting details: Cybertruck deliveries are expected to nearly double from Q3 and Model Y deliveries are expected to increase compared to Q3 which was unusually low.

We see a similar situation in Europe. In the last four years, Q4 was the strongest quarter of the year except for 2023 when sales flattened. However, even in 2023, Q4 was 9% higher than Q3.

My assumption for this year is an 8% increase but fortunately, the data quality for Europe is excellent and we won't need to rely on assumptions as we get more vehicle registration data later in the quarter.

2. Deadline for China's Cash-for-Clunkers EV Incentive:

In April 2024, China introduced a 10,000 Yuan ($1,408) incentive for replacing old gas cars with new EVs. Then, in July, they doubled the incentive to 20,000 Yuan ($2,816), which significantly boosted Tesla’s sales in Q3.

Now, this incentive is set to end. Buyers have until December 31, 2024, to purchase a new EV and must register it by January 10, 2025, to qualify. More details are available on the government website here.

With the incentive ending, overall EV sales in China, including Tesla’s, will likely see a boost in Q4 as buyers rush to take advantage. However, current weekly car insurance data shows Tesla’s Q4 sales in China at 33,700 units after the first four weeks, compared to 41,800 units in Q3. So, while the incentive deadline may help, we can’t automatically assume Q4 sales will exceed Q3 when the weekly data doesn’t support that. Therefore, for now, I estimate that Q4 sales will be similar to Q3.

Here’s my detailed table for China, with actual numbers up to September 2024 from the China Passenger Car Association (CPCA) and my estimates for October, November, and December 2024.

Here is a chart that shows my estimates for China deliveries in Q4 compared to past quarters. I expect Q4 to be pretty much the same as Q3. We will have a better idea as we get more data.

3. Additional Import Fees for Model 3 in Europe:

The European Commission is implementing a new import fee for EVs imported from China. This will be an additional fee on top of the 10% import fee that applies to all imported cars. For Tesla, the additional fee will be 7.8% but for other automakers like BYD, it will be much higher.

The Reuters article here says the new tariff is expected to be imposed from next month and is expected to be effective for five years. Therefore, it's likely that Tesla will increase its prices in Europe soon but I don't expect any price increase in Q4 because they have plenty of Model 3 inventory to sell.

In Q3 2024, Tesla sold 23,000 Model 3s in Europe and I'm calculating 43,000 Model 3s exported from China but not delivered anywhere yet at the end of Q3. In October, they will probably export another 22,000 Model 3s before the new tariff kicks in. So, there is plenty of inventory for this quarter and the next one.

Though the new tariff may soon impact costs, Tesla could use it to drive Q4 sales by pre-announcing a price increase and encouraging buyers to take delivery before it takes effect. Tesla hasn't done that yet but we might see it next month.

Here is a chart for deliveries in Europe. I don't see anything unusual in Q4. It looks like a normal quarter, slightly higher than Q3.

4. Over 7,000 unsold Cybertrucks to be delivered in Q4:

Tesla doesn’t release production and delivery numbers by model. Instead, they group Model 3/Y and Model S/X/Cybertruck. This is all we got for Q3. See Tesla's page here.

However, based on various data sources, including Tesla registration numbers for California and a few other states, I estimate that Tesla finished Q3 with around 8,225 unsold Cybertrucks, as they ran out of buyers for the Foundation Series which starts at $100,000.

Now, Tesla has begun delivering the regular Cybertruck version, starting at $80,000. By the end of Q4, some Cybertrucks will still be in transit. Therefore the inventory gain won’t be 8,225 units, but it should be around 7,000 units. In other words, I estimate that Tesla will deliver approximately 7,000 more Cybertrucks this quarter than they produce.

The table on the left shows my calculation for Model S/X and the one on the right shows the Cybertruck:

Based on VIN data, I estimate 17,976 Cybertruck production in Q4, up from 14,774 in Q3. The data shows an increase in the production rate after September 7, 2024. With 13 weeks in a quarter, this suggests an average production rate of 1,383 units per week in Q4.

I assume production will continue at this rate through Q4 without further increases but things could change. Therefore, I plan to get new VIN data regularly throughout the quarter.

Because of these four factors, deliveries in Q4 are expected to exceed Q3. Let's have a closer look. The first table shows what happened in Q3 based on all available vehicle registration data. The second table shows my estimate for Q4 and the third one shows the difference:

In Q4, I estimate 33K more deliveries worldwide compared to Q3. 24K of those are in the U.S. I estimate a big increase in Cybertruck and Model Y deliveries in the U.S. but almost no increase in Model 3 because the Model 3 is expected to be production-limited this quarter.

Tesla used to ship Model 3s from China to Canada but Canada imposed new import fees for cars from China. Therefore, they are not getting any Model 3s from China this quarter. Instead, Tesla will ship them from the U.S., which limits the available supply for U.S. deliveries.

Cybertruck Achieved Profitability in Q3 2024:

Tesla had initially forecasted that the Cybertruck would reach profitability by the end of this year, but surprisingly they achieved this in Q3 2024, one quarter sooner than planned. This was well timed because, in early October, Tesla announced that they are now delivering the regular Cybertruck version instead of the Foundation Series.

This change means the Cybertruck likely won’t be profitable in Q4, with its average sale price dropping by $20,000 compared to the Foundation Series. Achieving profitability in Q3 was therefore well-timed, and Tesla used a clever trick to make it happen. They increased deliveries of the higher-priced performance version. Registration data shows that 41% of Cybertrucks delivered in Q3 were performance models.

Interestingly, if we look at the total deliveries in the last four quarters, 31.7% were performance models. This suggests Tesla may have pushed around 1,700 performance deliveries from Q2 to Q3 to hit profitability in Q3.

I'm calculating the following numbers for the Cybertruck:

For the regular Long Range version, the Q4 average sale price should be approximately $82,000, though production costs are around $106,400 per unit. The Performance version’s average sale price should be around $102,000, with an estimated production cost of roughly $114,400.

However, if Tesla prioritizes delivering Performance models with Full Self-Driving (FSD), the average sale price could reach around $110,000. Combined with further cost reductions, this might bring them closer to break-even.

Analyst Consensus:

The consensus estimate is important because missing it often leads to negative headlines for Tesla. Currently, the Q4 consensus is at 508,000 deliveries, up from 494,000 before the earnings call.

This number will likely continue climbing toward 515,000 as more analysts update their estimates. This means the stakes will be high at the end of December, as Tesla needs 514,925 deliveries this quarter to match last year’s total annual deliveries.

When the consensus is within reach, Tesla often rolls out new promotions to hit that target. In the quarters I’ve highlighted below, deliveries ended up just slightly above the consensus, which is probably no coincidence. We frequently see new promotions and inventory discounts pop up, even in the last two weeks of a quarter. This suggests that Tesla closely monitors delivery numbers at the end of each quarter and introduces new incentives if necessary.

Anyone following Tesla is likely familiar with the end-of-quarter delivery push, and I expect the same this quarter. Tesla will probably do everything possible to maximize deliveries, and we might even see delivery tents set up outside the Fremont Factory and Giga Texas on December 31 to hand off cars as soon as they roll off the production line. It’s also possible that the Fremont factory will keep production going even on Christmas Day, just like they’ve done a few times in the past.

Full-Year Estimates:

The analyst consensus for the full year is now 1,802,000 deliveries. My estimate is 1,790,000 units. Both numbers are below the 1,808,581 units Tesla delivered last year.

What Happened in Q3 2024?

My final delivery estimate for Q3 2024 was 472,000 units, but actual deliveries came in at 462,890 units. That means I was off by 9,110 units, or 2.0%. I’m okay with an error margin of less than 3%, but in this case, the analyst consensus was more accurate. Therefore, for the sake of transparency, here’s a chart that shows my error rate compared to the analyst consensus over the last 12 quarters:

My average error was 2.8% compared to 4.2% for the Wall Street consensus. Also, there were two quarters, when analysts were more than 10% off. I'm not always more accurate, but I usually avoid the big mistakes.

Here is a closer look at what happened in Q3. The first table shows my final estimate on the last day of the quarter. The second table shows what actually happened in Q3 based on all the data we have now. The final table shows the difference.

U.S. Sales: My estimate was too high by 12,914 units or 8.2%. The U.S. is the most difficult region to estimate because of limited data and because the numbers change significantly from one quarter to another.
Canada: We still don't have the September numbers. So let's wait for that to have a better idea.
Europe: My estimate was too low by 1,672 units or 2.1%.
China: I was too low by 6,768 units or 3.7%. The data quality for China is pretty good because we get the weekly car insurance numbers for 12 of the 13 weeks in a quarter. We don't get the number for the final week and in this instance the final week happened to be much higher than expected.
ROW (Rest of the World): My estimate for this region was too high by 4,245 units or 13.5%.

Yearly Production Table:

Here is the latest production table for 2024. You can see this next to the 2023 table on the spreadsheet here. My production estimate for Q4 is 475,976 units, the highest this year but not a record quarter because Q2 2023 and Q4 2023 were higher than that.

Yearly Delivery Table:

Here is the latest delivery table for this year. My estimate for Q4 is 496,000 units, which is 12K units less than the 508,000 analyst consensus and 19,000 units less than the 515,000 units Tesla is aiming for. If you want to see this table and the 2023 tables side by side, you can open the spreadsheet and scroll down.

Year-To-Date Deliveries After the First 3 Quarters in 2024 vs 2023

Here is an interesting comparison. The first table shows total deliveries in the first three quarters of 2023. The second table shows 2024 and the last one shows the difference.

After three quarters, U.S. sales are down by 36K units. If it were not for the 22K Cybertruck deliveries, the U.S. would be down by 58K units.
Europe is down by 29K units.
China is up by 23K, Canada by 6K, and the Rest of the World by 6K.
Global sales are down by 30K units.
Model S/X sales collapsed in Europe, China, and the Rest of the World. The reason is unknown. In the U.S., Model S/X sales were OK.

Notes:

The Google Sheet for today's update is available here: Q4 2024 Tesla Production & Delivery Estimates.
If you find these updates helpful, feel free to recommend them to others. I appreciate your support.
This is Patreon-only content. Please don't share anything unless I have tweeted it. Thanks.
There is a public page here with lots of information about my error rate in the past. I'm trying to be as transparent as possible about my error rate.

That's all for now. Thank you!

Troy

Q3 2024 Tesla Earnings Per Share Estimate (Changed from $0.51 to $0.55)

10月22日

21 October 2024, Update 3:

Hi everyone. After reviewing the data again, I’ve adjusted my non-GAAP EPS estimate from $0.51 to $0.55. I’m trying to be as accurate as possible while we still have time, so I went over everything again and noticed some improvements, particularly in how the 1.99% financing offer affected U.S. revenues. I also updated the Q3 currency exchange rates, which impacted revenue in China and Europe. Here are the updated numbers:

Non-GAAP EPS:

Analyst Consensus: $0.60
My Estimate: $0.55

Automotive gross margin excluding regulatory credits::

Analyst Consensus: 14.9%
My Estimate: 14.0%

Overall Gross Margin::

Analyst Consensus: 16.9%
My Estimate: 16.5%

Here is the updated table. I will tweet this new table to avoid any confusion. Also, after the earnings release, I will tweet an updated version of this table with the actual numbers next to my estimates.

I’ve also updated the spreadsheet. You can find my estimates for each category in Column O here.

By the way, within the last two hours, Tesla announced 0% financing for Model 3 and Model Y in the U.S., without purchasing FSD. This will increase sales. I think 500K deliveries this quarter are now likely. Tesla needs 514,925 deliveries in Q4 to match the 2023 total this year. Even though 500K is now likely, it's not clear yet if they can reach 515K.

However, the 0% financing offer will impact margins for Q4. In Q2, Tesla offered 0.99% financing in the U.S., but only for the Model Y, and it lasted less than a month. Later in Q3, they offered 1.99% financing for both the Model 3 and Model Y throughout the quarter. Now in Q4, they’re offering 0% financing for both models.

Due to certain accounting rules, these promotions don’t increase the Cost of Goods Sold but instead reduce revenue. Here’s what I’m calculating:

1.99% financing (instead of the regular 5.75%) reduces revenue by 5.94%.
0.99% financing reduces revenue by 7.69%.
0% financing reduces revenue by 9.14%.

Therefore, the new promotion won't be great for Q4 earnings.

That's all for now. Thanks,

Troy

-----------------------------❖-----------------------------

21 October 2024, Update 2:

Hi everyone. It turns out the company-compiled consensus was just released and it shows $0.60 Non-GAAP EPS instead of the $0.58 shown on Yahoo Finance. In other words, the consensus is now even harder to beat. Here are the updated numbers. My estimate remains the same:

Analyst Consensus: $0.60
My Estimate: $0.51

Here is Tesla's Company-Compiled Consensus. It has two sections. This one is for Q3 2024:

This section is for the full year:

-----------------------------❖-----------------------------

21 October 2024, Update 1:

Hi everyone! Here are Tesla's Non-GAAP Earnings Per Share estimates for Q3 2024:

Analyst Consensus: $0.58
My Estimate: $0.51

In other words, I’m estimating a miss of $0.07. This wouldn’t be surprising since Tesla has missed the consensus estimate for the last four quarters. The last time they beat the consensus was in Q2 2023. Since then, they've missed it four times in a row.

We don’t have the company-compiled consensus yet. At the end of each quarter, Tesla collects estimates from analysts and emails the results back to the analysts who participate. They do this for both deliveries and earnings per share. We might get the company-compiled consensus tomorrow, so there’s a chance the $0.58 consensus could change. I’ll update this page with the company-compiled consensus if I get it tomorrow, but there won’t be another email notification for that.

The actual results will be published on Tesla's webpage here on October 23, 2024, at 5:30 p.m. Eastern Time.

Because analysts estimate $0.58 and I estimate $0.51, a reasonable question to ask is who was more accurate in the past. I have an excellent chart for that. Let me explain what it shows.

For example, in Q2 2024, the Wall Street Consensus was $0.62, and my estimate was $0.61. The actual Non-GAAP EPS came in at $0.52, meaning I was off by $0.09, and the consensus was off by $0.10. When you add up these errors for the last four quarters, I was off by $0.18 in total, while the consensus was off by $0.22.

Of course, we're adding the absolute values here. So if someone was $0.10 too high in one quarter and $0.10 too low in the next, those errors don't cancel each other out. The total error in that case would be $0.20. That’s what the chart shows.

If you look at the chart, you’ll see that James Cat has been more accurate than me, and he’s estimating $0.58 for Q3, the same as the analyst consensus. You can check out his tweet here.

My estimate is based on a detailed calculation for each model and trim level sold in each region. For example, here’s a section from my spreadsheet that shows the Average Sale Price. The U.S. numbers are based on DMV data. I don’t have Average Sale Price data for Canada, so I use the U.S. data and increase the prices a bit to account for higher transport costs.

In addition to that, I also calculate unit sales by trim level. For instance, the data I have shows that 41% of Cybertrucks delivered in Q3 2024 were the Performance version, which is quite high. In comparison, only 24% of Model S/X sales were for the Performance version.

The spreadsheet is pretty large and difficult to read but here is the link.

In addition to Earnings Per Share, the most important number analysts focus on is Automotive Gross Margin Excluding Regulatory Credits. Unfortunately for Tesla, this number has been on a constant decline for a long time, and I expect the same trend to continue. My estimate for Q3 2024 is 13.2%.

Here is a detailed table that shows my estimates for all categories:

I have another version of this table that mimics Tesla's shareholder letter for easy comparison. This table is also available on the spreadsheet here.

Here are charts for Average Sale Price in case you find this interesting. This one includes all models.

This one shows the blended average sale price for Model 3 and Model Y sales in each region.

I haven’t posted any updates for Q4 deliveries yet, but I just received new VIN data today and will be looking into it soon. Right now, the analyst consensus for Q4 deliveries is 494,000, and I think that’s a realistic estimate.

That's all for now. Thanks.

Troy

Tesla's Market Share, includes Q1 2022 [last update: 19 Jun 2022]

2022年3月4日

23 Oct 2023: Because this is an old post, it is now public.

-----------------------------❖-----------------------------

19 June 2022: Hi everybody. The data needed to update this file was complete in early May but I never got around to updating it until now. So, here is the Q1 2022 data.

The link to the Google Sheet is still the same: Tesla's Market Share

There are three sections:

Tesla's Market Share in Battery Electric Vehicles (BEV)
Tesla's Market Share in Passenger Vehicles
BEV Adoption Rate in Passenger Vehicles

'Battery Electric Vehicles' means it excludes hybrids and fuel cells. Let's go over each section:

Tesla's Market Share in (BEV) Battery Electric Vehicles

In the US, Tesla's market share in BEVs in Q1 2022 was 72.1% which is of course very impressive but in line with what the rate was in 2021.

In Europe, Tesla's market share in BEVs increased to 18.1% in Q1. It is expected to drop in Q2 because imports from China were low.

Tesla's Market Share in Passenger Vehicles:

I'm using the term 'Passenger Vehicles' instead of 'Passenger Cars' because the US number includes pickup trucks which are mostly used for personal transport in the US instead of commercial use.

There were two interesting details in Q1 2022:

In the US, Tesla maintained its 3.5% market share in passenger vehicles.
In China, Tesla's market share increased to 2% of passenger cars for the first time.

BEV Adoption Rate in Passenger Vehicles

This section is not related to Tesla. It shows all BEVs vs all passenger vehicles.

The BEV adoption rate is increasing fast in China. It reached 17.2% in Q1 2022. In other words, 17 out of every 100 passenger cars sold in China in Q1 2022 was a battery electric vehicle.

Another interesting detail was the drop in Europe. BEV adoption dropped from 16% in Q4 2021 to 11.8% in Q1 2022.

That's all for now. Thanks.

Troy

-----------------------------❖-----------------------------

3 Mar 2022: Hi everybody. This was the most requested topic in the last poll here, so here it is. I will keep updating the data after each quarter but it takes time until the data is published. It's not immediately published after the quarter is over.

Here is the Google Sheet for this report: Tesla's Market Share in the US, in Europe & in China

There are three sections:

1. Tesla's Market Share in (BEV) Battery Electric Vehicles
2. Tesla's Market Share in Passenger Vehicles
3. BEV Adoption Rate in Passenger Vehicles

Let's go over each section:

1. Tesla's Market Share in (BEV) Battery Electric Vehicles:

Tesla's share of BEVs is much higher in the US than in China and Europe. There are two reasons for that:

Tesla's share in passenger vehicles in the US is higher than in Europe and China. See the chart in section 2.
BEV sales in the US are low. See the chart in section 3.

Tesla's 77.8% market share of BEVs in the US in Q4 2021 is pretty amazing. It's also interesting that Tesla was able to maintain its market share in the US better than in China and Europe.

EV vs BEV: The term electric vehicle (EV) includes battery electric vehicles (BEV) like Teslas and plug-in hybrid vehicles (PHEV). However, in this report, I have only used BEV numbers and not EV numbers.

Passenger Vehicle vs Passenger Car: In the US, light trucks are commonly used as passenger vehicles but the numbers published for passenger cars don't include light trucks. Therefore instead of passenger cars, I'm using the term passenger vehicles which includes light trucks but only for the US.

2. Tesla's Market Share in Passenger Vehicles:

Tesla's combined market share in passenger vehicles in China + Europe + USA was 2.3% in Q4 2021. This percentage is expected to continue to increase over time because global passenger vehicle sales are stable but Tesla sales are increasing.

An interesting detail here is that Tesla's market share in passenger vehicles in the US is higher than in China and Europe. This suggests more people in the US who don't necessarily care about EVs still buy Teslas.

3. BEV Adoption Rate in Passenger Vehicles:

The most interesting detail here is that BEV adoption has increased much faster in China and Europe in the last two years than in the US. The US is lagging behind China by 5 quarters and behind Europe by 6 quarters.

The links to the data sources are available on the Data tab in the spreadsheet.

Again, here is the link to the Google Sheet for this project: Tesla's Market Share in the US, in Europe & in China

Behind the Scenes:

This screenshot is from a spreadsheet that we use to collect monthly Tesla registration numbers for Europe. Tesla only releases global production and delivery numbers and each country in Europe releases its own registrations numbers. So, we collect these numbers and put it together, and make it public for the community. You can see it here. I created this spreadsheet in 2015. We have been collecting this data each month since then.

Notes:

I'm working on a new update for 2023, 2024, 2025 estimates.
I have requested DMV data. It typically takes 7 to 12 days. There will be a new update when I have this data. I use it to calculate US production including Model S/X. Then at the end of March, there will be new DMV data. I will be able to see Texas production as well.
China February domestic sales and export numbers are expected to be released by CPCA on 8 March. Domestic sales are not important because what's not delivered in February will be delivered in March anyway. However, exports are important. Ideally, it would be good to see 22,500 exports in February. China numbers in February are expected to be low because of the Chinese New Year Holiday between 31 Jan and 6 February and because February is 28 days. However, this was already included in my calculation. My 324K Q1 delivery estimate here on 27 February still looks good.

That's all for now. Thanks.

Troy