Synthetic venture exposure?

 A few observations:

1. Venture capital funds at the highest levels is typically access-constrained,

2. Some large VCs are looking more PE-like (e.g. AI roll-up strategies), 

3. Some large public companies have venture arms (e.g. Google Ventures' $10B portfolio and includes some great companies, like Waymo)

My idea/question: is there a way to create an investable "synthetic venture" structure to capture these public companies' venture exposure?

For example: Google Ventures' revenue is classified under "Other Bets" in the 10-K (2024 revenue of $1.5B vs. advertising revenues of $237.9B) ... could you buy GOOG stock, hedge out the search and cloud businesses, and just have (levered) exposure to Google's venture upside? There may not be a financial instrument that can actually capture this ... but it feels like an interesting idea regardless, especially as many other large companies -- like (famously) FTX -- allocate to their own venture portfolios.

(Side note: a similar question I had a while back was if you could replicate PE exposure with leveraged small cap equities -- which is how I ended up finding Dan Rasmussen's 2015 paper, from which he created Verdad Capital. The general idea was that private equity firms in aggregate don't create value above and beyond the public markets, so why sink money into the illiquid asset class if you can get similar exposures in more liquid marketable securities? It turns out "private equity without the illiquid headaches" was good to start a hedge fund but not great marketing materials over the long run, so Verdad quietly pivoted away from that as their core strategy.)

Weird questions like this are where I think AI shines, so I asked ChatGPT. (This is also an area where asking a derivatives expert is helpful -- a creative financial instrument is unlikely to make a Google search. I phoned a friend in finance, and this type of thing doesn't exist today.) Chat GPT's results, summarized:

1. Ventures typically make up such a small fraction of the company's assets that you'd have to short nearly the entire company. For example, Google had $348B non-"Other Bets" revenue vs. $1.5B "Other Bets" revenue in 2024, so even if you valued "Other Bets" at 50x revenue (and everything else at 10x), "Other Bets" is 2% of Google's value. So, you'd go long GOOG, short 98% of it, and lever that 2% long exposure up -- risky, and likely expensive too.

2. Finding the right basket of things to short would be challenging -- you'd want to short the ads business, Google subscriptions, and Google Cloud. This would be hard to get right.

Two ideas I did like: (1) going long a basket of microcap stocks (conceptually similar to Dan Rasmussen's idea), and (2) a recent July 2025 paper that creates a framework for understanding VC returns as a basket of options (a paper I need to reed more thoroughly). (Other ChatGPT references point to better VC benchmarks (e.g. PME) and using R&D spend and Bloomberg innovation indices to give a "venture tilt."

So: still more questions than answers, and more things to read and digest now. But I'm sure this idea will sit on the backburner, waiting for the right time to be realized as an investable strategy.