Non-ergodic Investing Advice
Curve Finance founder Michael Egorov may be the latest very smart person to learn that a large number multiplied by zero equals zero.
Egorov is in danger of having his founder's share of Curve tokens liquidated because he didn't think they could ever get anywhere near zero and now, following Sunday's exploit, it suddenly looks like they can.
I'm guessing you have not borrowed $100 million from AAVE and Frax, so you may not be able to relate.
But there are nonetheless some lessons to be learned.
For one, taking out leverage because you think zero is impossible is a bad idea: Zero is possible! (And much more so with leverage.)
And for two, If something has 1 in 10 odds of paying off, you have to stay in the game long enough to play at least 10 times.
In physics, this is called "ergodicity."
In investing, you want your portfolio to be more ergotic.
That may sound like a $5 word for a 50-cent concept: Don't let yourself be zeroed!
But investors are constantly letting themselves get zeroed.
And most of the rest of us achieve suboptimal returns because the volatility of our non-ergodic portfolios panics us into making bad decisions.
Going a little deeper on the $5 word might help us make better ones.
The best ability is availability
In physics, ergodicity is defined as a condition where the "time-series average" of a process is equal to the "ensemble average" over many iterations of that process.
This is best illustrated with examples:
- If you roll one six-sided die 1,000 times, you'll get a time-series average of 3.5.
- If you roll 1,000 six-sided dies 1 time, you'll get an ensemble average of 3.5.
The averages are the same, so rolling dice is an ergodic process.
Now consider a game of six-chambered Russian roulette:
- If 6 people play 1 time, the ensemble average is an 83.3% survival rate.
- If 1 person plays 6 times, the time-series average is a 100% chance of being dead.
The averages are different, so Russian roulette is a non-ergodic process.
You want your investing portfolio to be more like rolling dice and less like playing Russian roulette.
Not everyone agrees: Sam Bankman-Fried would play Russian roulette with the entire planet: He told Tyler Cowen he'd accept a bet that had a 51% chance of replicating planet Earth against a 49% chance Earth is destroyed.
And he'd keep making the bet until he lost.
(Which may be all you need to know about FTX).
That's maybe not quite as crazy as it sounds: A bet with a 51% probability of winning has "positive expected value," so it's a risk that should be taken.
But betting it all — and doing so repeatedly —- is poor risk management. (Never SBF's strong suit.)
What's the right amount to bet then?
Gamblers and traders sometimes use the Kelly Criterion as a rule of thumb: A system of sizing wagers according to their perceived risk/reward, while limiting the risk of being zeroed by a series of losers.
Kelly does not work for Russian roulette, because there's no coming back from even one loser.
And it doesn't really work for trading, either, because it's impossible to know what the real risk/reward of any particular trade is — there's no card counting in financial markets.
But it can help dampen the volatility of your investing returns by making your portfolio more ergodic.
You can't make your portfolio perfectly ergodic. This is trading, not physics.
But you can manage your risk with that goal in mind — starting with avoiding things that have a reasonable probability of going to zero.
Because, in investing as in Russian roulette, there's no coming back from zero.
Do the math
Crypto investing is the very definition of non-ergodic: Even crypto-enthusiasts will tell you most tokens will go to zero.
You probably think you know the ones that won't. But you might be wrong!
Even the biggest cryptos have some chance of going to zero — likely much more so than a large-cap stock.
That doesn't make crypto uninvestable: Even Russian-roulette-style bets can be worth taking, as long as you can survive losing enough of them.
But it does mean you should size your crypto bets differently than your stock bets.
Don't go all-in on the first trade you see with positive expected value, as SBF would.
And don't leverage a bet because you think it's impossible to get margin-called, as Michael Egorov has done.
Because, in crypto especially, you might soon find yourself multiplying by zero.
No comments:
Post a Comment