Recent events show that the tide is turning in regards to Quantumscape’s Path to Market. The PowerCo licensing deal, and the subsequent earnings call where Siva doubled down on his preference for “capital light” partnerships going forward (quoted below), established a turning point for QS’s approach to commercialization in a major way.
We are very carefully evaluating every customer's needs so as we can develop a model specific to them with us having the preference of the capital-light approach.
~ Siva Sivaram (2024 Q2 Earnings Call)
Why the change of heart? Building factories is capital intensive, for sure, but capital intensive is one of the best paths towards building strong moats as a business.
The obvious answer is risk - and how building a giga factory will drain much of the remaining capital that Quantumscape currently has left on the books. But there were plenty of other ways they could have approached the initial joint venture. Quantumscape could have simply taken a similar cash advance from that deal structure instead.
You can read more on my summary of the SEC filing for the highlights. I think the deal specifics have been pretty well covered by now. I’d like to use this an opportunity to extrapolate; and expand on why this was the right call for Quantumscape.
Point blank, this new approach may have been a tactical move by Quantumscape. And I have some “tinfoil hat” ideas on why Quantumscape now prefers the capital light model (beyond the simple explanation of saving capital). These are all very much conjecture at this point, so bear with me.
QSE-5 isn’t the Endgame
QS’s first generation of cells is simply a stepping stone towards an even better product.
Simply by increasing the size of their cell - increasing their active to inactive material ratio - should offer another 20% gain in energy density above QSE-5.
Why spend hundreds of millions of dollars, now, building out a commercial production line that will essentially be deprecated in less than 5 years?
That could be the calculus Quantumscape is making right now.
There’s an argument to be made that QS-0 and QS-1 could have been made in such a way that would allow for modular cell designs, such that ‘turning a dial’ would make larger form factors with more layers at different cathode loadings. But at the end of the day, a production line is very complex, and altering cell construction won’t be as easy a turning a few knobs.
In any case, a QSE-10 or a QSP-5 would need to be completely revalidated anyways; running through the full gauntlet of tests to clear B and C sample stages before being cleared for commercialization. So a modular factory is already moot. Once a production line is set up, that’s probably the cell that’s going to made for the foreseeable future.
So, being that QS doesn’t have the processes in place for giga scale production to begin with, why not let someone else foot the bill to figure out scaling? It’s a win-win. I’d hazard a guess that QS’s split of profits (via the royalty payment) will be much less than a 50-50 split (initially projected from the JV) given that Quantumscape gets to keep the joint IP for free (per the agreement).
I originally guessed that Quantumscape could receive upwards of $10-15 per kWh for licensing the IP. It may be that the royalty payment works out to be closer to $5 per kWh, with the delta being made up by the IP gained that’s needed to actually bring the design to scale. In this case, everyone wins: QS is taught a masterclass on scaling with no risk to them and receives rights to IP that they can take with them to the next deal; and PowerCo gets a sweet deal on the next generation of battery technology at a discount.
And, the deal limits the degree to which Quantumscape loses market share - only by 40 GWh (or 80 GWh if agreed upon at a later date).
Future Development
Going a step further, it could also be that Quantumscape management has their sights set on much bigger aspirations. See Right Tails. Here’s a quote from the recent Evercore ISI webcast.
We get to focus on what we do best, which is technology development and creating a technology that is manufacturable. So we are not a one-and-done company. This lithium metal battery is at the infancy of it’s ramp.
Siva Sivaram (Evercore ISI webcast)
As stated above, QSE-5 isn’t even the final form factor to leverage the full potential of the technology.
Quantumscape also has other side projects going as well.
Work is still being done on bringing down the pressure requirements for the cell: A-0 samples required 3.4 ATM of pressure, A-2 cells require 1.7 ATM, and they’ve tested single layers with 1 ATM (absolute), which equates to zero external pressure required from the cell packaging. The obvious advantages to this work is market expansion into consumer electronics as well as improving cell-to-pack energy density (primarily saving weight).
They also have efforts going towards dry-cell technology and all-solid state cells (per recent job postings).
And of course, there’s always work to be done improving their separator technology; whether it’s on reliability, performance, or cost.
But these are small potatoes. Going even further, conversion cell chemistry offers energy densities that potentially double Quantumscape’s highest current projections for their NMC cells, potentially yielding closer to 2,000 wh/L and 600 wh/kg. See Right Tails (linked above).
Quantumscape already has irons in the fire (no pun intended) for this technology with several patents already calling out FeF3 (iron flouride) cathode materials. If that’s their eventual end goal, why bother iterating on Raptors and Cobras and the like when they could start work on this next endeavor instead?
In this case, it makes sense that QS would rather spend their time developing more IP than iterating on small scale production lines.
QS will remain an innovation-focused company, and we will continue to develop new intellectual property and technical know-how.
~ Siva Sivaram (2024 Q2 Earnings Call)
Competing Paths to Scale
The PowerCo deal is very interesting because any new technology or manufacturing processes jointly developed will also belong to Quantumscape. This is disclosed in the licensing agreement.
Now imagine that QS signs a similar deal with Ford (for example). Ford engineers may be able to offer their own insights for manufacturing. All of a sudden, Quantumscape gets to A/B test different processes for production. Lessons learned will be able to be shared across each company, sure, but Quantumscape won’t be footing the bill for either one of them. And then, Quantumscape can take any lessons learned with them to their next deal.
I think this is their primary goal anyways. Scaling manufacturing, especially for a brand new technology, is time-consuming and money-consuming. Neither of which QS is in much supply of.
Watering Down The Product
Pulling the thread from the last point, there are downsides to consider. It’s possible that each OEM will be running a different version of QSE-whatever.
PowerCo will surely add their own spin on the product - probably adding their dry cell technology. This will likely be true of all the OEMs that QS eventually partners with.
One fear that I have is that all these differentiated products across OEMs will spread QS thin in terms of their ability to support different cell designs. It’ll be interesting to see whether QS cells will narrow down to a couple of core form factors, similar to legacy 21700 & 18500.
Another fear is that quality controls won’t be as stringent across all OEMs that are running QS lines. If Ford (for example) builds a QS line that ends up with massive recalls, that will negatively affect the Quantumscape brand.
Even if QC issues don’t lead to recalls, we could still see large discrepancies across performance specs such as cycle life and energy density and etc. between OEMs. Again, that could negatively impact the QS brand.
Having read the licensing agreement, I didn’t see anything regarding auditing production facilities or minimum QC requirements.
Anyways, that’s something to keep an eye on.
What is the Endgame?
Personally, I think Quantumscape needs to pursue building out their own Giga factories. Eventually. I think the risk to their IP is increasingly heightened the wider that data is sourced. But, as of right now, they really don’t have a choice. I’m surprised that they only projected to save $134 million from the dissolution of the Joint Venture. That must have only been their initial commitment for the pilot line at QS-1 because multi-GWh facilities cost billions of dollars. Which, Quantumscape obviously doesn’t have.
IP is useless if they can’t actually scale it. PowerCo should be able to help with that. And naturally, revenue from the royalty agreement will allow Quantumscape to expand on their own. Again, eventually.
We still don’t know the terms of the deal. The royalty amounts are completely obfuscated in the agreement by design. The royalty advance payment of $130 million works out to be $3.25 per kWh if counted against one full year of 40 GWh production. This amount makes sense to me as Quantumscape is compensated in IP as well. But the actual royalty payment is anyone’s guess at this point.
Also, during the last earning’s call, we did get hints as to the “key technical milestones” that are required to receive (and keep) the the prepayment:
In terms of milestones, these are related to QSE-5 B-samples, the Cobra process, and finalizing the target design for the cell. We will then grant the license and receive the royalty prepayment.
~ Siva Sivaram (2024 Q2 Earnings Call)
It seems that a lot of the qualifications for prepayment will be accomplished by the end of this year: namely, B-Sample delivery and target cell design.
Regarding Cobra process requirements, more information on this was actually just revealed in the latest Evercore ISI webcast. PowerCo is bringing their fleet of engineers (to the tune of 150 personnel headcount) to California to assist in the scaling of Cobra. This should both help QS ramp up Cobra more quickly and improve the transition for the PowerCo team on their return to their Giga factory.
Either way, add these to the list of catalysts that should be coming up in the next handful of quarters.