Rethinking yield farming
I'm changing my mind about farming yield in crypto. Not so much because what I've learned was wrong. Or even that the field has changed; it hasn't.
What changed is my sense of risk vs reward.
The reward part is obvious. You have idle assets, it should be put to use or you're not being capital efficient. Meaning my shitcoin is better off being lent out for 18% APR return than sitting in my wallet doing nothing.
After three years of doing that, I'm re-evaluating the extent of risks I've taken on in order to carve that kind of return.
I don't think about it much but now that I do, it turned out I got burnt by yield farms more times than I care to admit. Anchor/LUNA was a big one. There were smaller protocols where either hacking happened or smart contracts went wrong.
You see this whole business is very similar to you putting your money in a bank, except the bank is not licensed and there is no government to bail them out when shit hits the fan. When that happens your entire deposit is at risk. Sometimes they may make it up to you by giving you some kind of IOU, which may or may not be worth anything in the end.
Risks in DeFi comes in many shapes and sizes. Primarily it's divided between smart contract risks and market risks.
For someone who writes code (maybe even good enough to audit them), I seem to have this blind faith about smart contract working as intended. There is an unhealthy lack of fear on this front.
But even smart contract is something visible. What is invisible is market risks, where your deposits gets lent to parties you cannot see. Or that incentive design goes haywire (like Terra Luna). Movies get made about nonsense like these (The Big Short).
To manage these risks, there is the inevitable time labor cost involved when participating in yield farms. You may check on your bank once a year just to make sure your money is still there, which is good enough because the bank will probably stick around.
A yield farm however may go south at any moment, which necessitate me monitoring them every one or two weeks. This cost adds up when there are dozens of farms involved.
Anecdotally, the longer your money stay in one yield farm, the likelier you get burnt.
Given the time, effort and risk cost, the reward simply does not justify it.
While I'm progressively withdrawing from yield farms, there are some guidelines I'm going to follow when farming.
Any inflationary coin should be farmed for yield; for everything I shall refrain. This includes highly inflationary shitcoins and stablecoins.
Inflationary shitcoins should not amount to much of a portfolio anyway. If it's a substantial portion of the holdings, we have a bigger problem than yield farming.
Stablecoins however is the base currency. It's not optional to hold them. If I expect to hold them for a long period (say over a two-year-long crypto winter), then it's prudent to spread them out over multiple stablecoins rather than concentrated in USDT only.