Battery Day: The Reality Vs. Expectations
Personally, my expectations going into battery day were blown away by Tesla’s (NASDAQ:TSLA) execution, and that was already after I set my expectations fairly high. The Street, however, seemed underwhelmed post Battery Day, with the stock selling off.
Let’s dive into what Tesla actually did deliver.
The Battery Day event was an additional event added onto the normal proceedings of the Tesla annual shareholder meeting. Before going into Tesla’s new battery technology, Elon Musk made a few key announcements. A link to the majority of source material for this note can be found here. The first announcement Musk made was that he anticipates the Giga Shanghai plant can reach one million units per year at some point in the future. This is a big deal, because capacity seemed locked at ~250K units between the Model 3 and Model Y combined. With Phase 2 of Giga Shanghai underway, continued capacity unlocks will likely boost deliveries over the long term in the Chinese market in particular.
Elon Musk then goes on to say that despite the challenges that Tesla has faced this year, the team at Tesla is pulling together well, and they anticipate 30-40% growth in automotive deliveries for the full year. This means that Tesla anticipates 477-514K units in the year of the pandemic. So, in spite of the ~month long shutdown of the Fremont factory, Tesla still anticipates to come roughly in line with their guidance provided for 2020 at the end of 2019. This is, to say the least, very impressive and is a positive update in the 2H delivery push.
And finally, at the end of the shareholder presentation (not the Battery Day itself), Tesla gave us two key updates on the autonomy front. First of all, it sounds like the fundamental autopilot software rewrite has been completed, and automated video labelling is now being implemented. This likely greatly speeds up software development and neural net development time. Essentially, this is a big boost for Tesla in terms of ramping to full autonomy. And second of all, Elon Musk said we can anticipate a full self-driving beta in roughly a month or so. This is huge, and accelerates the prior timeline I had in my valuation model for Tesla’s commercialization of autonomy.
Finally, we have the Battery Day presentation.
The Meat & Potatoes: Battery Day
The bulk of the Battery Day was focused primarily on Tesla’s multiple-year, multi-stage plan to halve the cost of their batteries. The battery is one of the largest costing components of the overall vehicle. A drastic reduction in the battery cost will likely lead to an improved margin profile on the vehicles. Bear in mind, third-party analysis suggests that Tesla is spending $158.27/kWh at the pack level at the end of last year. The first part of the presentation involves an innovation in tablets architecture in their battery cells.
Anybody with a remote level of understanding of battery cell architecture understands that batteries feature a component called the tab. The tab has been very useful, particularly when it comes to thermals, and improving the flow of heat throughout the cell. The key argument made against a tabless cell is that it will come at the expense of the thermal design of the overall battery. But by removing the tab, the overall electrical path length reduces by 5x, improving the amount of time it takes for energy to get to the battery drastically. This will likely lead to further improvements in overall supercharging. Removing the tab simplifies the manufacturing process, and removes parts. Removing the tab also enables Tesla to increase the dimensions of the cell itself, hence the 46 mm x 80 mm architecture. This leads to a 16% improvement to range itself. Production has already begun at Tesla’s pilot-facility in Fremont.
The second development is Tesla’s new cell factory. One of the most prevalent innovations from this factory is their dry-electrode manufacturing technology. This technology was largely proof of concept when Tesla bought its owner, Maxwell Technologies. However, Tesla is on revision number four of this technology, and the overall yield has been low, but improving. This dry-electrode technology will lead to a 10x reduction in footprint and energy consumption. Tesla is estimating another three more manufacturing revisions (which take 3-4 months). Thus, we are likely 9-12 months away from volume production. In addition, the manufacturing process itself at this new cell factory is likely going to be a major innovation for Tesla. Tesla claims that one assembly line can produce ~20 gWh, which is ~7x the normal assembly line. How will Tesla do this? Two ways: a continuously moving assembly line and vertical integration of manufacturing teams. Firstly, a continuously moving assembly line reduces overall idle time in the factory, improving output. In addition, the acquisition of Grohmann Automation leads to greater vertical integration in the supply chain and manufacturing process. Elon Musk tells us that Tesla’s biggest advantage over traditional auto OEMs will be efficient auto manufacturing, not necessarily the EVs themselves, or autonomy software. It’s manufacturing. This efficiency will likely lead to a 1 tWh scale at the Texas Terafactory compared to the 150 gWh at Giga Nevada, on a smaller footprint. This is just a great show of Tesla’s improved manufacturing efficiency. As the cell factory scales up, more volume potential is unlocked. Tesla anticipates 100 gWh of internal cell production by 2022 (~1 million units, assuming 100 kWh/vehicle). Tesla anticipates this scales to 3 tWh in production by 2030. All of this leads to another ~18% reduction in $/kWh at the battery pack level.
The third major development is Tesla’s use of new materials in their batteries. They are moving to the use of silicon instead of graphite. Silicon is in much more abundance relative to graphite, and more importantly, it stores 9x more lithium than graphite. Shifting from graphite to silicon is expected to improve range by an additional ~20%. It also leads ~5% $/kWh reduction at the pack level.
The next upgrade in Tesla’s battery technology is improving the underlying cathode material. For those that don’t know, the cathode can be thought of as a storage material for lithium in the battery. The material Tesla has used in the cathode since essentially day one has been cobalt. This has led to worries of battery inefficiency as well as human rights issues, as obtaining cobalt can often involve the use of child labor in African cobalt mines. Diversifying cathode materials away from cobalt will likely improve cost dynamics. Long life cycle products would likely use iron, long range products would use nickel and manganese, and mass sensitive (Cybertruck and Semi) use high nickel. Elon Musk goes on to note that 35% of the cost of the cathode comes solely from manufacturing and processing. Tesla plans to address this by building a cathode plant, drastically simplifying the cell supply chain. All of these advancements lead to another 12% reduction in $/kWh.
The final upgrade Tesla has made is improving cell-to-vehicle integration. Tesla’s new casting process for the Model Y has led to a reduction of 79 parts from the final vehicle. This reduces manufacturing complexity and cost. In addition to this, Tesla batteries are becoming duel use as both an energy device and part of the structure of the car. The new “body+battery” design allows for 170 less parts, a 10% mass reduction, 14% more range, and another 7% $/kWh reduction at the pack level.
All of these fundamental breakthroughs have led to a 56% cost improvement, a 54% range improvement, and a 69% improvement in investment per gWh. The long term is to build 20 million vehicles per year, which is supported by 3 tWh of internal capacity by 2030.
In addition to these announcements, Tesla had some very interesting tangible product announcements. The first announcement was well covered by the media: Tesla plans on building a $25,000 autonomous car in roughly three years. They also announced the highly anticipated Model S Plaid, with a 0-60 <2 seconds (competitive with the Roadster), a top speed of 200 mph, and 520+ miles of electric range. This model is expected to begin deliveries towards the end of next year.
Taking This All Into Consideration: Is Tesla Really A Bubble?
I often hear commentary surrounding Tesla that goes as follows: Tesla is evidence that we operate in a bubble market. And while I agree that the market has in large part been artificially inflated, Tesla is not the poster child for this market whatsoever. Let’s delve into my valuation framework for Tesla to explain why this stock really isn’t the bubble bears would like you to believe it is.
Starting on the EV side:
|2025E||Model S/X||Model 3||Model Y||Cybertruck||Semi||Roadster||Total|
|Auto Operating Profit||$18.153B|
|2025 Auto Valuation||$453.832B|
|2021YE Auto Value per Share||$348.27/share|
I am anticipating that despite the opportunity to expand unit sales of the Model S and Model X with plaid versions, units will remain roughly flat from last year’s 66K deliveries. I also have a $90K ASP for these vehicles, as they are high-end premium luxury vehicles. I estimate that before the battery day announcements, Tesla’s high end gross margins were ~25% for the S/X. Considering that (as I mentioned earlier), third-party research indicated that Tesla battery packs costed $158.27/kWh at the end of last year, and that Tesla expects a 56% cost decline, how will this affect margins? Assuming this 56% cost decline comes to fruition by 2025, the battery pack cost should be ~$69.63/kWh. On a 100 kWh battery pack per vehicle, the savings and margin upside total ~$8,864/vehicle, an additional ~9.8% of margin. This brings me to a gross margin profile of 34.8% on the S/X units.
For the Model 3, I expect general growth in unit sales from here to 2025. This will be driven by international demand for the car, primarily in China right now, and eventually in Europe. I’m projecting that US demand stabilizes at ~150K/year, with an additional ~150K in China, ~100K in Europe. This additional demand will come online as capacity ramps up in these geographies. While the Model 3 starts at $35,000, the majority of buyers purchase some form of upgrade. Whether it be a $1,500 paint option, improved tires, or even the full self driving package, the price is more than often materially higher than the $35,000 starting price. For this reason, I’m using an ASP attach of $45,000. Prior to the battery day event, I estimate that the Model 3 cost ~$35K per unit to assemble. With a 56% cost reduction per kWh on a 75kWh battery pack, I estimate an additional ~$6,648/unit in gross margin. This brings the assembly cost down to $28,352/unit. Assuming the same ASPs of $45,000 per unit, we can assume a new gross margin profile of 37% on the Model 3.
With regards to the Model Y, I expect this to be Tesla’s biggest product (excluding mass market $25K model). Domestic volume should be ~300K (i.e. double Model 3) because of the sheer size of the addressable market in the US, the crossover market. The Model Y has the right combination of range, size, and price point among other features to make it a compelling vehicle for the masses. In China, I anticipate volumes of ~100K, less than the Model 3. And in Europe, I anticipate ~400K units, the bulk of the units to come from Europe. Tesla anticipates ~500K units of annual capacity from Giga Berlin. I believe the Model Y will likely be a much bigger hit than the 3, just because of the size of the European crossover market relative to the mid-priced sedan market. I estimate ASPs stabilize ~10% above Model 3 ASPs at $50K, because the product is ~10% larger. Elon Musk has said in the past that when the Model Y hits sufficient scale, production costs will be roughly equivalent to the Model 3. This is backed up by a third-party teardown analysis of the Model Y by automotive expert Sandy Munro. The Model Y should cost roughly the same to produce as the Model 3. So, assuming that previous cost per unit of the Model 3 of $28,352/unit, Model Y margins can reasonably be expected to be at ~43.3% by 2025.
Moving on to the Cybertruck. This is a vehicle that remains difficult to project for a few reasons. First of all, we don’t know how big this market is. Why? Right now it seems like Tesla’s plan is to deliver units to the US before needing to create a smaller version of the product for European/Chinese markets. If Tesla only intends on building the Cybertruck for the US market, I believe that could seriously limit the potential scale of Tesla’s target market. That being said, Tesla is attacking two markets simultaneously.
The first market is the traditional trucking market. This market is made up of millions of units, with the F-150 being the best selling vehicle in the whole US. If Tesla’s specs are better than competing ICE pickups, then small businesses and normal truck buyers will have a reason to make the switch. The second, and likely larger market is the enthusiast market. People that just think the Cybertruck is a good product, rather than looking at it as a utility for their business. This is likely the market Tesla will go after internationally, as trucking is only biggest in the US (and China to a lesser extent).
I estimate ~150K domestic sales for truckers by 2025. These are individuals who buy the Cybertruck for its utility value. I estimate an additional ~150K international sales for enthusiasts. Keep in mind, this estimate could prove to be pessimistic considering reports indicate reservations of 650K units for the Cybertruck. My ASP estimate is $55K, as I believe the majority of buyers will skew towards the high-end, high-range variants rather than the $40K base model. Margin-wise, it’s still a tough call. I’m going to be conservative and estimate a 10% gross margin, as we still don’t have a third-party teardown or some even details on the size of the battery pack. Going off of third-party estimates, we’re looking at a vehicle with a battery pack of 250 kWh. That’s a massive number, and assuming the same $/kWh by 2025, that means the battery alone could cost Tesla ~$17.4K. Forget about the exoskeleton, tires, interior, computational hardware, and everything else in the car for a second and focus on the battery. Excluding the battery, there is ~$37,600 left for profits and the assembly of the rest of the vehicle. That’s not a whole lot considering the size of the vehicle. So, my margin estimate is 10%.
Finally, we have the Semi and the Tesla Roadster. Starting with the Semi, third-party market analysis pegs the domestic electric Semi market at 54,000 units by 2025. Assuming Tesla, the likely technological leader, scales to a good percentage of the market, I believe that 25,000 units in 2025 (~46% market share) is feasible if not conservative for the market leader. And we aren’t even talking about the Semi’s international potential. There are two models, a 300 mile and a 500 mile range model. My ASP estimate tries to blend these two models in as they cost $150,000 and $180,000, respectively, bringing me to an ASP of $170,000 as I believe the majority of Tesla’s sales will go toward the higher range variant. Similar to the Cybertruck, it remains hard to tell where margins likely go because of two factors: (a.) no teardown analysis and (b.) no idea in terms of battery pack size. Some estimates peg the 300 miles version at a staggering 500 kWh, with the 500 mile version at 900-1100 kWh. Assuming that $69.63/kWh calculated earlier, a 500 mile $180,000 Semi could cost ~$76,600 at the pack alone, leaving slightly over $100K for the remainder of assembly and profits. So, again, I’m going to be conservative with a 10% margin estimate until we get some more details.
The Roadster will likely be a small contributor volume wise. For a $250,000 car, I see units of ~1,000 per year globally. It’s not a large market. As hinted at, I anticipate ASPs of $250,000 for the Roadster. Margin-wise, we have a little bit more clarity for the Roadster than products like the Semi and the Cybertruck. We are looking at a vehicle with a 200 kWh battery pack, so ~$14K in battery costs. Because battery costs will likely take up such a small portion of total vehicle costs, I am more comfortable assigning a higher gross margin estimate closer to ~25%, though this too could prove conservative.
Tesla had slightly north of $4 billion in OpEx in 2019. My model has this more than doubling to $10 billion by 2025 as Tesla ramps up all aspects of their auto business. I’m using a multiple of 25x on operating profit. Why? Because there could be significant upside to these estimates. Keep in mind, this model doesn’t factor in a $25,000 car whatsoever, a car that could be Tesla’s biggest hit. In addition, there could be more upside in Cybertruck and Semi volumes/margins, and Roadster margins. Combined with an overall growing EV market and Tesla’s technological lead in the backdrop, a 25x multiple seems very justified.
I elected to use an 8% discount rate for two reasons: (a.) we’re and will continue to be in a low rate environment, meaning discount rates will fall and (b.) Tesla has continued to de-risk themselves financially with an improving balance sheet and consistent cash flow. Overall, I assess the auto business’ valuation at ~$348.27/share by the end of next year.
What about Energy?
One of the most overlooked parts of the story here is likely Tesla’s Energy business. Tesla’s strategy is multi-pronged. They are one of the few companies I have ever heard of with an actual mission statement. The mission is to accelerate the world’s transition to sustainable energy. There are three vectors Tesla must attack: transportation, generation, and storage. The auto business is only one of these three vectors, transportation. The energy business is attacking the other two vectors: generation and storage. In my view, Tesla remains in the early innings of developing this business. It is a business that has admittedly been a drag on the overall business. One key catalyst for the ramp up of the energy business will likely be the Tesla Solar Roof. The roofing business alone is massive, with an estimated market value of $119.7 billion. Now that Tesla has sufficiently ramped cell production for the Model 3 and Model Y, I would anticipate Solar Roof production will no longer be cell-starved. And we’re only talking about the generation side of the business. Additional growth drivers could come from the Megapack and commercial utility storage products. These could provide future growth as the grid moves more and more towards clean energy. Here’s what I’m modeling for the energy business right now:
|TESLA ENERGY VALUATION||Estimates|
|2025 Revenue||$5.96 billion|
|2025 Gross Margin||15%|
|2025 Gross Profit||$894 million|
|2025 Valuation||$53.64 billion|
|2021YE Valuation||$35.193 billion|
Keep in mind that in 2019, Tesla reported $1.531 billion in energy, a Y/Y decline. So, this model could be interpreted as aggressive considering the inconsistency in Tesla’s Energy business’ financial results. That being said, with the rollout of the solar roof and more PowerPack and MegaPack deployments coming in the following years, this is a business that could eventually become very large. Piper Sandler is forecasting ~$200 billion in revenue from their Energy segment by 2033 after hitting $12.4 billion in 2023. While I am not as outright bullish as Piper Sandler is in terms of scaling up that fast, I remain very constructive on Tesla’s ability to grow to $5.96 billion in revenue in 2025. This is just short of 4x growth by 2025. Gross margins should be relatively stable at 15% in my model as the more cyclical solar leasing business decreases as a percentage of revenue. As such, margins should remain fairly stable at ~15% long term. With this in consideration, I’m willing to use a very high premium multiple of 60x earnings considering the segment’s long-term potential to be very similar in revenue generation to the auto business and potentially even larger. Considering the fact that energy is a higher risk segment than auto, a higher discount rate is likely warranted relative to the automotive segment. Overall, 2021 year-end valuation of energy comes in at ~$37.70/share.
The Final Piece of The Puzzle: Autonomy
With the valuation of energy and auto finished, the final, and potentially biggest piece of the puzzle remaining is autonomy. Autonomy is the real wild card with Tesla. If Tesla nails it, the stock could still be a multi-bagger from here. Think of it like this: Apple (NASDAQ:AAPL) released the iPhone in 2007. In the subsequent years, software services drove growth. Tesla’s BEVs are the iPhone, and autonomy is the services business. It’s a rough comparison, but it does the trick. Tesla’s combination of advanced proprietary computing hardware, billions of miles of real-world data, and strong software foundation make me believe they are in a leadership role for the industry. This is backed up by Tesla’s plan to deliver a full self-driving beta in roughly a month. If Tesla executes on this (and it is a big if), it will give credibility to the idea that Tesla is in pole position in this market.
My current belief is that autonomy pans out positively for Tesla, with the company launching an autonomous ridesharing service at massive scale over the course of the next few years. Here is a brief rundown of my autonomous ridehailing valuation model:
|2030 Bookings||$50 billion|
|Gross Profit||$9.75 billion|
|Operating Income||$7.5 billion|
|2030 Valuation||$300 billion|
|2021YE Valuation||$90.197 billion|
|2021YE Valuation per Share||$96.62|
Keep in mind, this builds out to a bookings value of $50 billion by 2030, this is where Uber (NYSE:UBER) currently is globally. If Tesla differentiates itself with an autonomous platform with millions of vehicles on the network by 2030. This also assumes a ~30% take rate, higher than current in-person ridesharing take rates which are closer to ~20%. I believe this is justified because of the lack of driver anywhere in the model. I believe we can also see Apple Services-esque margins in the mid 60% region, and $2.25 billion (15% of total revenue) to go to operating expenses (think server and infrastructure costs). Considering the long-term growth trajectory for autonomy to disrupt even further, I am applying a 40x earnings multiple to this operating income. This brings me to a valuation of $300 billion. Discounted at 12.5% (because of the higher risk of autonomy coming to fruition), we arrive at a 2021 year end valuation of ~$90 billion, or ~$96.62/share.
Putting it all together:
Combining the valuation of autonomy, automotive, and energy together gets us to a final valuation of $482.59/share, an upside of ~16% from current trading levels. Considering the lucrative upside profile, and the reasonable assumptions in the valuation model to get to this target, the stock is a buy at this price.
The Only Real Risk
In the past, all of these ambitions have had numerous headwinds, but Tesla’s core headwind comes down to execution. Can they execute on their timeline and really deliver. Particularly when it comes to autonomy, will they really launch a beta in a month, or is that just a wish? Tesla has a track record of executing, just not necessarily on time. I believe that Tesla will continue to execute, continue to prove the bears wrong, and continue to drive value for shareholders.
What About The Delivery Numbers?
Tesla recently announced that they delivered 139,300 vehicles globally, a positive surprise relative to the ~136K unit consensus and a record quarter delivery wise. I believe that, despite the stock’s pullback, the delivery numbers were a validation of the bull case for the stock, and the stock will reset higher. So far, I have been fairly accurate with my predictions for the stock, and with fundamentals improving, I believe the stock will re-rate higher. Rating reiterated at buy, price target raised from $350 to $480.
Disclosure: I am/we are long TSLA. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I am not a financial advisor. This is not financial advice. Please do not interpret this as financial advice. Do your own due diligence before initiating a position in any of the securities mentioned.