💥 Zero Bullshit
Chapitre 06

Voodoo Valuations

Because several members (and non-members) of Hakisition had indeed raised an eyebrow at the fact that it offered a way to invest in SpaceX, Perplexity and Mistral AI.

Very quickly, Mistral AI’s general counsel issued a denial, but probably a bit too fast. While it’s quite surprising that an investor would be looking for liquidity in this company in particular, it’s much less so for SpaceX and Anthropic.

But what we have here is a double misunderstanding of investing in private companies.

Because it’s true that no private individual gets into a SpaceX primary funding round, and you don’t get Perplexity shares with $2,000. But quite a few funds syndicate investors into SPVs, and those investors may want to get in or out.

In this case, for SpaceX, it’s about funding liquidity for investors in a vehicle run by Force Investments30, and for Perplexity a vehicle run by E131, which is raising money to buy secondary shares from last September’s $200M round (at a $20bn valuation).

But above all, what matters isn’t getting in, it’s understanding the terms.

Take Perplexity, valued at $20bn in the September round. To reach the 10x multiple within a few years that Benotmane has announced several times, it would have to hit a $200bn valuation. Except that according to my information, the valuation at which the SPV is selling to Hakisition is around $30-40bn. So 10x means $300-400bn. Roughly Netflix32.

On SpaceX, the numbers stop making any sense at all. Because once again, Benotmane talks about 50x33.

Screenshot of the Hakisition website

The problem is that the calculation makes NO sense. To exit at a $300bn valuation, as stated in the deck, you would have had to get in at $6bn.

From a Hakisition deck

The proto-secondary market that exists for SpaceX has been valuing the shares (and we could talk about the huge price gap between share classes) at between $200 and $220 since this summer, so around a $400bn valuation. At the last round, in December 2024, which sent Musk’s fortune soaring, SpaceX was valued at around $350bn34.

According to information from Forge, the SPV’s liquidity is being sold to Hakisition at a valuation above $475bn. So a 10x means aiming for a $5tn valuation. More than Nvidia ($4.3tn), Apple ($4tn), Microsoft ($3.5tn), Amazon ($2.4tn) or Tesla ($1.3tn). Even a 5x makes no sense over a few years given the valuation.

In other words, even assuming he got in at a bargain valuation of $400bn, which is unlikely, his calculation leads to a valuation of $20,000bn+. That’s almost twice the Nasdaq 100.

With 50% carried interest (yes, really), Benotmane would therefore pocket €12M of carry on €500K raised.

The madness is the same on Perplexity, which shows 65x over three years, so a valuation of $2-2.6tn. As a reminder, OpenAI was valued at $500bn in its last round35, which is already a huge record, and IPO rumours hover around $1tn. Once again, unheard of.

Screenshot of the Hakisition website

As for Mistral AI, at a mere 157% a year, you’d barely reach 17x. At the valuation of the ASML round, it would have to reach $240bn+. Yet the website talks about a $100bn valuation, which implies Hakisition would get in at $5.9bn, the valuation of May 2024.

Better still, the deck explains that Mistral would make €100M of net profit and €500M of revenue in the medium term. When annualised ARR in mid-2025 is only $70M (vs €30M in 2023), according to several concurring sources, and the company is racking up astronomical losses.

Screenshot of the Hakisition website

And all that is before the fees the SPVs take. In a similar deal, Huntler (administered by Sydecar) was offering 8% in upfront fees and 10% carried interest. More amounts the multiples have to make up for.

In fact, in that same deal, which has been doing the rounds for weeks and mostly seems to attract people who want to sell dreams to retail investors, the central case talks about a 5x (€60bn at exit), 10-50x (€120-240bn) for the very favourable case, and 1-2x (€12-24bn) for the worst case.

From the investment document sent by Huntler Capital to distributors

Clearly, Hakim, the maths ain’t mathing. And nobody does 50x on companies like these. Doing late stage or pre-IPO isn’t wrong. Nor is it banned. But as usual, return is proportional to risk, and nobody serious expects multiples that high on de-risked, established companies.

Which doesn’t mean there’s no risk.
It means the reward can’t be at the levels given.
And that’s normal.