Reverse Engineering the Returns for YC Partners

Given Paul Graham's calculation of $575 billion, and combining it with these additional comments in the thread:

Q: Paul, how much money did you actually invest in these companies?
A: Good question. That grew over time, but the biggest ones tend to be older, since they've had the most time to grow, which means the average investment would have been on the low side. Less than $50k on average I'd guess.

Q: Does YC dispose of stakes once companies go public or does it keep and manage that risk?
A: As with most investment firms, the stock gets distributed to individuals and they can decide.

Q: What’s the value of YC’s stake? Could you assume 1-2% with dilution?
A: I'm not sure, but 2% sounds about right.

Some very back-of-the-envelope calculations:

• YC top 30 companies = $575B
• YC owns approximately 2% = $11.5B
• Average investment = $50k
• Approximately 3000 companies = $150M
• Given that few of the recent startups would be in those top 30, but also that YC now puts in $125k per startup, let's just go with those assumptions for this calculation.
• $11.5B (YC's stake) / $150M = 76x

76x.

Interestingly, given Paul Graham's comment that $5.75B would have astounded him, clearly YC's partners wouldn't have funded 3000 companies if that initial appraisal had held true. That is because 2% of $6B is $120M, which would be less than the total funding amount (wildly full of assumptions) mentioned above.

Plus, it's been 16 years. For Techstars, it's been 15 years. For a business growing at 25% annually, that implies a doubling every three years or so. Enormous growth can happen in the second decade of a dominant company.