Aurora – Shirtless in Mountain View pt. II

Part I can be found here

Part II – Aurora demands the shirt off the investor’s back.

• The valuation of this deal is even more divorced from reality than Tesla and the EV SPAC crowd that it has spawned as I value this business at $1.1bn which is 90% below the $11bn being asked.
• Typically, these SPACs have revenues beginning in the low tens of millions this year, reaching around $1bn in revenues in 2025 as well as generating a small profit and cash flow in that time.
• This kind of profile has typically generated valuations in the order of $1.5 – $2bn and many of these companies have managed to just about hang onto these valuations so far.
• From a fundamental standpoint, this is a tough pill to swallow but Aurora has outdone the other SPACs handsomely in terms of ask.
• Aurora’s plan (assuming it all goes perfectly) pushes this revenue profile out to 2027 and demands 5x the valuation for the privilege of buying the shares now.
• This suggests that the risk appetite for Aurora, an also-ran in the autonomous driving world, is much greater than it is for any of the other SPACs and makes them look cheap on this basis.
• The valuation benchmarking offered by the company to justify the deal at $11bn, defies logic and reality in my view.
  • First, Cruise and Waymo: Aurora argues that because these companies are valued at $30bn (page 52), that it too should attract a big number.
  • I have already argued that the industry is very overvalued (see here) and these two are the leaders.
  • Aurora is an also-ran and so has a high risk of being competed out by its bigger, stronger, and better peers.
  • Hence, there is no logic in pegging Aurora’s value to that of Waymo and Cruise even if their valuations were fair.
  • Second, comparative valuation: Aurora benchmarks its gross margins (page 53) against software companies that have no effective competition in their space.
  • There is every indication that this industry is going to be brutally competitive meaning that this analysis has no merit.
  • Aurora then goes on to use the multiples of these companies in 2022 to make its 2027 revenues estimates look low.
  • There is no basis for this in any proper fundamental analysis that I have ever seen as:
    • First, time: multiples of revenues or profits need to be compared at the same point in time in order to have any fundamental basis.
    • Aurora is comparing revenue multiples that are 5 years apart rendering any comparison utterly meaningless.
    • To be fair to Aurora, it is trying to indicate what its multiple might be when it hits that level of scale.
    • This has the advantage of making the deal look cheap on a brief inspection but looking deeper quickly reveals that it ignores any of the substantial risks that I discussed yesterday.
    • Second sector: Aurora is comparing itself to enterprise software companies, semiconductor vendors, and Tesla none of which are direct comparables for its business.
    • Hence, the conclusion that its fundamentals will be similar to those of these companies is no more than guesswork.
  • In my opinion, these flaws in logic and reason fail to provide any support whatsoever to the $11bn being asked.

RFM Valuation

• The only to look at this company with any degree of rigor outside of throwing darts at a dartboard is to look at discounted cash flow (DCF).
• Aurora has provided free cash flow forecasts to 2027 which I have used and then assumed that the $150m of cash generated in 2027 grows to $5bn by 2032 and $8bn by 2037.
• From then on, I have assumed growth at 5%.
• I think that this would constitute massive success on Aurora’s part.
• Given how competitive this market is likely to be (see here), margins are likely to be quite constrained meaning that 25% EBIT margins are probably a fair estimate.
• This means that including a 20% tax rate, Aurora will need to report revenues of at least $40bn in 2037 to support free cash flow of $8bn.
• Aurora’s most bullish expectations (very unlikely in my view given the likely price pressure) give $24.4bn of revenues in 2030 from which I think $16bn is more realistic if everything goes swimmingly (page 48).
• From there it has to get to $40bn by 2037 which using its own (very optimistic) estimates, is possible.
• This allows me to create a free cash flow profile out to 2037 and a continuance value from there based on Aurora’s own expectations.
• With no trading data from which to calculate the discount rate, one has to rely on investments that have a similar risk profile.
• Looking at very high-risk investments like biotechnology and mine development, discount rates of around 20% are not uncommon but I think that the risk being assumed here is even higher given where its technology is.
• Hence, I would be comfortable with a discount rate of around 25%.
• Using this discount rate and Aurora’s own cash flow forecasts, I would value this company at $1.1bn some 90% below what is being asked.
• In order to get to $11bn, I need to drop the discount rate to 15% and this would still leave nothing on the table for investors other than the implicit 15% return being assumed to compensate for the risk of failure.
• To get to something interesting, I need to drop the discount rate to 10% which then gives me a current valuation of $39.8bn.
• The problem here is that 10% is not even close to fair compensation for the outsize risk being assumed, leading me to conclude that this is one of the most overvalued transactions I have ever seen.
• This is on par with the $7.25bn that was given to Uber’s autonomous driving business (see here) which I have argued was worth $0 (see here).
• The irony here is that it is that Aurora now owns these same dreadful assets, although I suspect it may have done this deal solely to win Uber as a customer for its offering.
• I think that it may have quietly binned whatever it got from Uber other than the employees.

Conclusion

• The problem that Aurora has is that when it acquired Uber’s self-driving assets, the deal valued Aurora at $10bn (see here) meaning that to keep investors who came in at $10bn happy, the listing has to be higher.
• I don’t think that the deal would have been put together at a lower number.
• Unfortunately, this requirement has meant that the boundaries of reason have long since been crossed leaving a proposition that, in my opinion, is as outlandish as many of the internet IPOs in 1999 and 2000.
• I would not pay more than a $1bn for this company and in my opinion, even that leaves very little upside without a major improvement in expectations beyond that which Aurora has already aggressively laid claim.
• The risk being assumed, and the return being offered are way out of balance and I would not countenance participating in this transaction under any circumstances.