Yield Farming as Growth Hacking
Web3 yield farming is basically Silicon-Valley/Uber/Moviepass style highly-subsidized-usage growth hacking, except it subsidizes users through newly issued equity instead of discounts paid out of VC's pockets. It's as if, instead of Moviepass charging you $10 to watch $30 of movies each month, you paid $30 to watch $30 of movies, but every month Moviepass prints $20 of Moviepass stock and give it to you. You're free to sell the stock, so in the end you still pay $10 for $30 of movies.
Yield farming is possible in web3 because tokenization allows issuance of stuff that looks like equity — has governance rights, and sometimes stuff that looks and feels kind of like cash flow rights — with much lower costs than traditional startups. Imagine hypothetically trying to do this as Moviepass. You'd have to have each of your users sign a micro-SAFE for like $20 a month of equity. I don't think you're even allowed to do this since your users likely aren't accredited investors (though I’m not a lawyer, this is not legal advice!)
Is yield farming "better" than VC-subsidized growth hacking? Under a naive Modigliani-Miller capital-structure-irrelevance benchmark, the two should be literally the same. $20 is $20 whether you pay it in cash, discounts, equity, or free popcorn.
But in reality? When Moviepass sells $30 of movies for $10, eventually Moviepass shows that it loses money and looks stupid. When Moviepass prints equity instead, Moviepass looks like it doesn't lose money, which I suppose is a benefit, but the equity holders pay through dilution.
At some level, the puzzle as to why yield farming works is that, if everyone expects yield farming to happen, why would anyone hold equity they know is going to become highly diluted quickly as part of the business model? Shouldn't everyone just dump the equity rewards the moment they get them? And if they do, the equity becomes worth less, meaning the protocol has to sell more new equity and dilute more to amount to $20 of movies, giving people even more incentive to dump the equity, and so on…
So at some level the puzzle is why yield-farming protocol equity prices remain high enough, for yield farming to work at all. Obviously one hypothesis is this is supported by a bubble/systematic overvaluation of protocol equity tokens, relative to fundamentals. But if this is the case, then yield farming is very smart, from the protocol's perspective, as a growth hacking strategy! Equity sales are cheap for you as a funding source, because for some reason everyone is willing to hold your equity at stupid high prices. If everyone wants your equity even at stupid high prices, of course it's worth it for you to give them equity, rather than giving them cash. So you end up with a relatively cheap way to ride the bubble and growth-hack adoption.
Is yield farming "good for society"? Or, to be more precise, who wins and who loses? Theoretically:
Protocol "users" who use the protocol, but don't hold equity, win by getting essentially paid to use protocol, so long as they sell their equity before it gets diluted too much.
Protocol owners/VC's sort of win: they get more adoption, increasing value, but they also dilute themselves through issuance. Kind of a balancing act. If they dilute too much/exit too late/bubble breaks, they lose.
To the extent you believe that the yield farming model works because tokens are systematically overvalued, the losers - as in any bubble - are the guys holding on to the overvalued equity tokens and supporting the high prices.
So, under this view, ultimately this whole thing works, and yield farming is more viable than cash growth-hacking, because there is some dumb money in the market pushing token equity prices up, thus making equity cheaper than cash for protocols to issue. Eventually, as in any such episode, the dumb money loses. The problem, of course, is it is not always easy ex-ante to tell who is the dumb money!