,
13 tweets,
3 min read
Read on Twitter
I'm subtweeting re power laws because I don't want it to look like I'm being combative. I've just thought about this *a lot* so I'm skeptical of "answers" but very very supportive of people working on it
1/ We *don't* know that venture returns follow a power law distribution. We don't have the data. It's far more likely they follow a lognormal distribution, but this is boring to talk, think, or write about
2/ Any model that generates a PLD for VC returns requires some nitpicky fine-tuning that is unlikely to happen in the real world. PLDs do not spontaneously emerge as robustly as Gaussians
3/ If VC returns are PLD, we don't know the mechanism that produces the distribution. My guess would be combination of exponentials or a birth-death process, as per this post reactionwheel.net/2015/06/power-…
4/ A preferential attachment mechanism seems prettier, and that has merit, but I have been unable to create a preferential attachment model that reflects what real world data. I am probably just not smart enough. If you can, let me know!
5/ If VC returns *are* PLD, then it's probably because the group of companies VCs invest in are PLD and that would be because *all* companies are PLD, although with different power-laws (different alphas)
6/ I believe this because every investor is different. For every VC that swings for the fences on every investment, there are a dozen smaller ones that won't take real risks. But survivor bias makes it look like there's only the former reactionwheel.net/2015/01/80s-vc…
7/ If this is so then it's not that VCs are *causing* PLDs, they are simply the victim or beneficiary of them, depending on how you look at it reactionwheel.net/2019/01/why-do…
8/ VCs may prefer to invest in businesses that are winner-take-all, but this is effect not cause. These businesses have always existed. The "Great Narrowing" in the stockmarket is not new, but a function of the business cycle neuvc.com/labs/top500/
9/ Even when people like Peter Thiel tell founders to try to monopolize, he's just channeling Michael Porter ("competition to be the best is destructive") and the idea itself is much older reactionwheel.net/2019/01/schump…
10/ I am sympathetic to the idea that there is a better way. I want there to be a better way. But, as with anything, I don't think we should underestimate the difficulty of finding that way
11/ As with any innovation, we need to ask: why now? What do we now know that we didn't used to know that makes this possible? It's not as if all the others before us were just too dumb or unmotivated
12/ It *can't* be that it's just a decision we make. The world is too diverse to assume no one has ever made that decision before; they have, they just didn't survive to tell /end
