Basically your thesis is that as an asset class, late stage private technology companies are underpriced, since the presumed employee option pool will be smaller than previously assumed, and thus dilution will be smaller than previously assumed. Thus you can bid a higher price than your competitors and still come out ahead in your investment in this asset class.

What I don't understand is how you get around the fact that you would still have to "pick winners".

> What I don't understand is how you get around the fact that you would still have to "pick winners".

If you didn't lead investments and instead diversified substantially amongst late-stage companies, you could probably get sufficient overall exposure to the class so as to not be driven by the performance of individual companies.

In reply to your other comment, VC investors tend to have a "thesis" about a particular market but PE firms can and do have a much broader thesis. "Changing conditions have led to late-stage equity being undervalued as an asset class" would definitely qualify.

>If you didn't lead investments and instead diversified substantially amongst late-stage companies, you could probably get sufficient overall exposure to the class so as to not be driven by the performance of individual companies.

Oh that was the missing link. I forgot that you can "not lead" a round, allowing you to avoid committing too much of the fund in a given company. Thanks for pointing this out.

My impressions is that late stage investors generally have lots of protection here, so that the company is forced to IPO by certain dates to avoid penalties, and if the IPO isn't at an agreed upon number, the company gives more shares to the investor to make up for it. IIRC square had such a clause, although I don't know how that turned out.

Picking the winners isn't so hard if you can protect your investment like this. The lower bounds on their expected returns isn't $0.

To make it a classic arbitrage you would hedge by shorting an "equivalent" asset. For example, if 1/1000th of JP Morgan Chase's assets are those also in your portfolio, then for every 1000 long shares you would short 1 share of JPM.

Since JPM is far from equivalent and most VC capital you don't have access to short, in practice I assume that a) you would assume that any startup receiving financing / investments at unicorn valuations is already a winner, and b) go long a sufficiently large and diverse basket to try to eliminate risk.

Unfortunately (a) is so far from true I'm not sure this would be an effective investment thesis.