This is one of the those intuitive models that has slowly risen to the surface as I’ve understood the venture industry -
Alex Hormozi talks about aggregating and disaggregating risk as one of the core drivers of human value. Risk is hard to model, but we can imagine it as a bed of lava, where certain bubbles will form and pop, although it is very difficult to predict exactly where those will occur. I’ll also refer to it as a stochastic bed, characterized by the individually random but collectively understandable nature of the bubbles.
The different players of the venture industry can be modeled as value capturing over this stochastic bed.
Founders, for example, concentrate their holding on a specific intersection of markets, personal dynamics, and ideas that might result in a spike. In the event of a spike, they capture a large part of that reward. They are choosing a specific plot and waiting for a bubble to appear there.
VCs diversify across a slightly larger bed, say 15-20 such intersections. If one great company emerges, they capture a smaller portion of the reward, but with slightly less risk, because they have 15-20 shots instead of the founder’s 1. They have a larger plot, which increases their chance of getting a bubble.
Note: this is why people say that every investor wants to be a founder, and every founder wants to be an investor. They are each merely jealous of the other’s risk-reward profile - the founder sees the investor with a lower-risk chance of becoming moderately wealthy, while the investor sees the one founder in their portfolio who has become obscenely wealthy.
Of course, one can go one layer further out, to the LPs of a VC fund. A fund of funds invests in ~10 VC funds, which gives them exposure to ~150-200 potential home runs. Of course, that exposure comes at a greater cost, which decreases their risk but also their reward. They are effectively betting that the entire stochastic bed will yield at least one home run, and by spreading their risk accordingly, they can capture a decent return (though they are unlikely to become moderately or obscenely wealthy except by scaling significantly). Generally, LPs are diversified across many industries and asset classes - trying to spread their risk across the broadest version of the stochastic bed in order to make sure they don’t lose too much money. They don’t know which sector will do well, but as betting on the total growth of the world economy to protect their wealth.
There is an interesting theme throughout: proper aggregation of risk can create something knowable out of the unknowable. This is the basis for a lot of the math behind probability. No one exactly which part of the stochastic bed of startups will outperform (every founder thinks they do, but unfortunately most of them are wrong). VCs don’t really know, either, but they shift their risk/reward ratio to increase the chance of getting the home run. Their LPs do this on an even higher plane. In essence, all of these players are toying with an unknowable thing, and using different strategies to create more or less knowable versions of the future.