-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
The coming Tesla cash cow and the short burn of the century
- Thread starter DaveT
- Start date
Thanks for the answer, very thorough.
You are right that I think the warranty costs are understated but overall I think your analysis if pretty reasonable and well thought out.
i don't think your interpretation is wrong.
Our long-term gross margin target of 25% for Model 3 has not changed. In the medium term, we expect to achieve slightly lower margin due to higher labor content in certain areas of manufacturing where we have temporarily dialed back automation, as well as higher material costs from recently imposed tariffs, commodity price increases and a weaker US dollar. On the other hand, our average selling price is significantly higher than prior projections, so we expect to achieve higher gross profit per vehicle than we previously estimated.
the combination of this paragraph plus the 20% gross margin comment on the conference call lead me to conclude this is a 5% gross margin reduction for model 3 production in q3 and q4. i found this disturbing because q3 and q4 should have very high option mix vehicles. maybe i can comment more later.
"imitation is the sincerest form of flattery" so thanks!
I'm curious what the gross margin is on a true $35k base Model 3. And if it's even positive.
How do you guys reconcile that Q3/Q4 will likely have a burst of production from built up demand for the AWD/LR/EAP M3 that will likely be the BY FAR highest margin M3, but after that it will be much cheaper versions yet you assume that "25%" will be achieved and then just maintained?
I think 25% is POSSIBLE on the extremely high end M3's, but at full scale without backlog that will likely be a very tiny % of the M3 sales if you assume an ASP in the low 40s (which i think is probably reasonable).
I assume the general assumption is the economies of scale will offset the declining ASP/high margin items in future periods? I am not saying that is wrong but just asking if that is the general assumption?
I think 25% is POSSIBLE on the extremely high end M3's, but at full scale without backlog that will likely be a very tiny % of the M3 sales if you assume an ASP in the low 40s (which i think is probably reasonable).
I assume the general assumption is the economies of scale will offset the declining ASP/high margin items in future periods? I am not saying that is wrong but just asking if that is the general assumption?
Mario Kadastik
Active Member
I would assume 10% GM on the base model when the line is operating at nominal pace and ca 40+% for the high option mix high ASP Model 3. Total averaging around 20-25%
From the SpaceX guy interview, Model S marginal cost was 30k years ago (S60), Deepak mentioned 2k of depreciation per car. Definitely positive. If the 3 line is more efficient and the BOM is lower cost, it could be 25k to build a 35k sales price 3 or 29% GM. That's without EAP which tacks on 12.5% GM at the 40k price.
i can understand your concern. they guided down gross margin 5% for q3 and q4. so we're looking at 20% on the highest asp cars. as lower priced options get mixed in it creates a gross margin headwind.
i don't think the overall asp will ever be low 40s.
to get to 25% you have to assume additional operating efficiencies being realized in 2019. i think it's quite possible from what i have seen on the s/x, but the timeline has pushed out several (2?) quarters.
sundaymorning
Active Member
But Tesla is still adding robots and additional lines such as the semi automatic. Would that contribute to the decline in margins as well? Next year they’re still guiding for 25% on less optioned M3s.
I dont feel like its reasonable to assume
1) ASP of over 45k
AND
2) Demand of over 100k a year, let alone 200k
You have to pick one or the other. 50k+ is basically mid size luxury sedan, which the TOP selling car in 2017 sold just over 55k in the US
Mid-size luxury cars - U.S. sales by model 2017 | Statistic
Obviously that is just the US, and you can average in some lower spec'd M3's that will capture some of the entry level luxury car demographic, and maybe somehow the market reverses trend and starts growing (doubtful), but from just an eye test, it seems like to get to 200k/yr demand you have to have an ASP in the low 40s, and my personal opinion is the M3 won't even have that much demand @ 40k without the FIT credit. I don't feel like many bull cases really capture what the total market is for 50k cars. The whole "best selling mid size luxury car already" or whatever was in the update letter is incredibly misleading since the M3 has 2 years of backlog to work through.
correct. but they were also guiding that 25% long term gross margin before. and i was thinking 25% gross margin in q4. i thought we'd see gross margin 25% in q4 with high option vehicles, and then efficiencies offset the several % margin drag from mixing in lower option vehicles.
but now he's saying 20% gross margin in q4, and efficiencies have to offset *another* 5% extra beyond just the headwind of lower option vehicles. that seems a longer bridge to cross to me.
I agree with this, also the fact that EM notoriously overstates and now you almost need his claims to be understated to make the projections work.
Can we all agree there’s no more short squeezes? Elon predicting another short squeeze is about as believable as FSD capability.
with all the short interest out there, a short squeeze is still one of my main concerns
Main concerns as short seller:
1) Some sort of terrible govt bailout, likely in the form of some weird SpaceX reverse merger or something
2) Some really dumb chinese company buying Tesla for the brand at a really terrible price
3) Short Squeezes
Im not really scared of anything else
adiggs
Well-Known Member
A few reasons why I think a high 40's / 50k ASP for Model 3 over the long haul is reasonable:
- EAP is practically a requirement for every car and I believe is priced at $4k. Therefore the $35k base model is really $39k (whether Tesla collects the incremental $4k at original purchase, or after the second owner activates it).
- When it comes along, FSD won't be quite as much of a requirement, but the take rate is going to be >>50%. That's another $5k I believe, and now the base car is more like $44k.
- The history of Tesla car models is that they attract people to a more expensive car than they otherwise would purchase. Plenty of anecdotes - it's an important component of how well S/X are doing in the luxury car market - part of it is that they are bringing purchasers that wouldn't otherwise be in those markets. (I'm 2 of them - first a Roadster, then a Model X; ICE car competition is I've never otherwise spent $20k for a car, and won't).
- The history of Tesla car models is that base models are passed over. The original Model S 40KWh that everybody expected to be the volume car after the initial expensive cars were shipped, instead had so few orders that the company canceled the car without ever shipping an actual 40KWh battery pack. They filled the orders using software limited 60KWh packs to avoid the developments costs for that smaller pack (and crash tests and stuff).
- Seems like the most common S/X is a non-performance, AWD, 100 KWh car.
- There are enough high volume car drivers that the lower cost of ownership (fuel, maintenance) can be used to swap a cheaper car for a more expensive car, and still come out ahead (for these drivers). This will bring people in the $20-35k range, up into the Model 3 range as they shift fuel/maintenance expense into a car payment; some of them will keep going to and beyond $50k. It won't take Model 3 to 1M units/year in the US, but I have no trouble believing 100k units/year in the US on a sustained basis. (And this won't come about for a couple of years anyway).
My expectation is that a rear wheel drive, small battery, no options Model 3 will be one of the rarest of unicorns in the market of Model 3's. Once both that combination and the AWD performance version is shipping, both readily available, there will be significantly more of the AWD performance being bought, than there will be of the no options Model 3.
I realize that's backwards from any other car model in like the history of ever, but its consistent with Tesla's history. I expect the single biggest volume configuration to be the AWD big battery, not performance, with EAP. And that's going to be really close to or higher than $50k.
Remember that the original business plan for Model S was a sustained 20k/year demand / production, and thus the importance of bringing Model X to market quickly on the heels of Model S. Instead Model S is more like 50-70k/year on a sustained basis, with Model X approaching balance with Model S, and an overall expectation of the 2 cars to be 100k/year indefinitely (or even for demand to go unfilled at that level).l
Thanks for the thorough answer--your post contains a lot of substantive explanations, but are you sure there is no Supercharger expense included in SG&A?
In 2014 Tesla told the SEC:
"As of December 31, 2013, we allocated 40% of our Supercharger network costs to cost of automotive sales and the remaining 60% to selling, general and administrative. . .
At least annually, we review the utilization of our Supercharger network and adjust our cost allocation, as appropriate, based upon changes in utilization trends."
Letter to the SECAt least annually, we review the utilization of our Supercharger network and adjust our cost allocation, as appropriate, based upon changes in utilization trends."
The most recent 10Q stated:
"Cost of automotive sales revenue includes direct parts, material and labor costs, manufacturing overhead, including depreciation costs of tooling and machinery, shipping and logistic costs, vehicle connectivity costs, allocations of electricity and infrastructure costs related to our Supercharger network, and reserves for estimated warranty expenses.
The current percentage allocation to SG&A may no longer be 60%, but wouldn't Tesla delete "allocations" in the description of Auto COGS if 100% of Supercharger expense were charged there?
sundaymorning
Active Member
One of the things that really stood out for me in the earnings report was this:
For months, the battery module line was our main production bottlenecks. After deploying multiple semi-automated lines and improving our our original lines, we have largely overcome this bottleneck. Consequently l, we now expect to reach a module production rate of 5,000 car sets per week even before we install the new automated line designed and built by Tesla in Germany.
What this this tells me is that the new semi automated lines that are being installed were not part of the original plan, which means the extra spending will likely offset some margins. Once the automated lines are working as intended, then workers on the semi automated can be directed elsewhere. This is the same for the current situation at Fremont, where more hands are being hired to compensate for the robotics that are being pulled offline for the “hackathon”, then once those robotics are hacked, Tesla will reinsert them and draw those workers elsewhere to increase production even further. Here’s the excerpt from the letter:
We have temporarily dialed back automation and introduced semi-automated or manual processes while we work to eventually have full automation take back over. This flexibility has enabled us to continue the Model 3 to new levels.
I think that once automation takes over, then margins will increase, and the 5k number won’t stop there, because Tesla will continue to push to 10k, which means the current workers will be deployed elsewhere once automation becomes more efficient. I think the headwind in lower margined cars will be offset by more efficiencies of automatic lines as well as a higher ramp to 6k, 7k-10k.
Those equipments Tesla is bringing in right now is not cheap, the learning curve on how to use them, make them functional, and then rectifying non working parts evidently costs a lot of initial investments as well as time. But once the learning curve and experimental phase is done, it’s prety much replicating the same line. To me, it’s much harder and more expensive to ramp from zero -1000 Model 3s than it is to ramp from 2,000-5,000 Model 3s, simply due to the learning curve. From the letter:
Once we hit the 5,000 per week milestone, we intend to incorporate our learnings to continue to increase output on our existing manufacturing lines beyond 5,000 units per week, and then in a capital efficient manner to add incremental capacity to ultimately get to a 10,000 unit weekly rate.
Simply put, Tesla is not capital efficient right now because there is still a learning curve, once we get over that curve, then efficiencies will make up the lost revenues/time.
The evidence? Just look at the SX ramp.
Singer3000
Member
Hello Brian, thanks for that. An interesting snippet on the opex/gross profit split for Suoerchargers. I wasn’t very clear but my point was more that Superchargers are not considered a Service and Other item, as some many at first assume but included within Auto.
I disagree with the above. For the production to improve and the gm on the 3 to go down from -10%, they'd need to see a ridiculous increase in depreciation, or a insane growth in workforce, or some combination of both.
I am not in front of a computer, but it would be instructive to come up with a split of total $ depreciation, per unit depreciation, direct material and direct labor, for the 3 production q1. Repeat the same for q2, and see if you can make the math work out and give reasonable numbers for increase in capex or direct labor. I'll try to post mine at some point.
The under utilization appears to be because of the larger sizes of the new centers rather than the increase in the absolute number of service locations. In the table below, the ratio of 5 mobile = 1 permanent is included in the Locations count; Fleet =Cumulative Vehicles Delivered and Ratio = Fleet/Locations
| QUARTER | LOCATIONS | FLEET | RATIO |
| 2Q16 | 233 | 136,655 | 586 |
| 3Q16 | 250 | 161,476 | 645 |
| 4Q16 | 265 | 183,728 | 693 |
| 1Q17 | 271 | 205,779 | 759 |
| 2Q17 | 300 | 227,805 | 759 |
| 3Q17 | 350 | 254,164 | 726 |
| 4Q17 | 376 | 285,354 | 759 |
| 1Q18 | 399 | 321,972 | 806 |
If the ratio correlates to "utilization," there has been a steadily improving trend which dipped slightly with the initial addition of mobile units but resumed in the most recent quarters. (To this layman, it still looks low)
Aren't many of the Sales & Service locations operating leases rather than outright purchases?
Are the Service Centers doing much body repairing other than for collisions? IIRC, Jon Mcneill was the CEO of a nationwide network of collision repair facilities. Elon assumed his responsibilities when he resigned in February. Elon doesn't seem to have similar collision repair experience and could be stretched too thin to oversee properly the start-up of a new segment in the service business. Again, as a layman, outside of specific high market penetration areas such as California, HK, and Norway, there doesn't seem to be enough volume to warrant the acquisition of specialized equipment and skills. Perhaps, this is a business segment that could be temporarily curtailed (or re-thought entirely) during the current cash preservation regime as M3 ramps to profitability.
It's opaque. Services & Other Gross Profit decreased between 3Q17 and 4Q17 by $25.4 million (from a loss of $63.1 million to a loss of $88.6 million). The Services & Other Gross Profit decrease between 4Q17 and 1Q18 was $29.0 million (from a loss of $88.6 million to a loss of $117 million). Revenue declined in both quarterly periods, while COGS increased sequentially. In 4Q17 the "Net changes in liability for pre-existing warranties" was -$3.5 million vs + $0.5 million in 1Q18 (There was an additional $37.1 million in accrued in 1Q18 because, with the new revenue standard, warranty repairs on Resale/Residual Guaranty Cars reduce the accrual (amount reserved) rather than being expensed as incurred.)
I may have it backasswards (and am not an accountant), but was under the impression that a reduction in amounts previously accrued (as shown for 4Q17) would benefit the corresponding COGS account to which it relates. (My impression was reinforced years ago when Chanos whined about $10.1 million reduction reported for 4Q13.)
Back at you. There is definitely some hair here, both because S&O is a hodge-podge and because of the deteriorating trend, but the losses may reverse if Tesla can execute profitably with the M3.
Similar threads
- Article