I'm building a tool that predicts the value of a tax strategy, the 83(b) election, and am trying to understand the return distribution for vc-backed startups at various stages so I can ballpark how valuation evolves the 5 years following when someone joins.

Assuming Gaussian price movement (for Brownian motion) is probably very wrong unless the startup is late stage.

Some research from Angellist suggests power law distribution but does not include failed startups, only those with positive returns.

YC companies seem like a biased but good proxy for what I want but I'm not seeing any consolidated analysis of survival and exit.

I'm looking for high quality analysis, not random blog posts about how 90% of startups fail without attribution.