4 ms·
Matt's mostly right, but the mathematical conclusions are, depending on how generous you want to be, minorly misleading or flat out wrong. Leaving aside the in
by aston 18y ago
Matt's mostly right, but the mathematical conclusions are, depending on how generous you want to be, minorly misleading or flat out wrong.
Leaving aside the incorrect graphs and potentially wrongly chosen distributions mentioned elsewhere, Matt missed the real reason VC's and entrepreneurs have different risk profiles.
Entrepreneurs have one company they're rooting for--their own. VC firms have tens or, over the course of their lifetime, perhaps hundreds. The law of large numbers says that as you sample more (with more companies), variance around the mean reduces, and you're quite likely to end up with your expected value at the mean. The entrepreneur has no such safety blanket and so would clearly choose to reduce variance.
- mattmaroon 18y agoYou still wouldn't want more variance even with a larger sample unless with it came higher EV too. I know my graph is wrong (think I mentioned that in the post) since 0 should probably be the likeliest result.
- aston 18y agoYeah, the reason VC's can stand higher variability is the aggregation, but the reason they want it is approximately your argument, that the distribution makes it so that their worst case is not that bad, but their best case is great (due to the clamp at $0). The wide range of big payoff possibilities must raise the EV, but the individual entrepreneur doesn't care because of the variablity he can't take.
- mattmaroon 18y agoVCs also can stand the variation because they get paid even if they fail. Were VCs paid only a % of the profits, and had to reimburse for failures at the same rate, you'd see their risk tolerance drop greatly.
- aston 18y agoAlso true, though their LPs do get paid only on the profits (less the aforementioned management fees). If they weren't comfortable with the risk, they wouldn't invest, right?
- mattmaroon 18y agoYeah, though they can invest in many high risk VC funds much the way VCs invest in many high-risk startups. Harvard's endowment, for instance, is ~$35 billion. They're risk-tolerant on a level that Sequoia can only have wet dreams about. So though VCs will be less risk-tolerant due to sample size than entrepreneurs, LPs will be that much less than VCs. Even the best VC firm is only a few poorly-performing funds away from the scrap heap.