Since my last post 5 days ago, I've grown my portfolio with various bets from $26,664 to $35,632. That's about a 34% ($8,968) increase. This is great, but it could have been a lot more if I had used proper bet sizing:
I noticed that I don't have any positions larger than 5% (~$1,750) of my port on 25usdc right now. Of course $1,750 is a lot of money, that's more than the gross average monthly wage in Greece, but I should be betting a lot more when I'm confident a market is mispriced.
So how much should I bet?
There's a great formula called the Kelly criterion. It can be used to calculate the optimal position size for each bet while maximizing expected value in the long term. All that is needed is the market price and your estimated fair price. Plug these two into the formula and you get the fraction of your portfolio you should allocate to the bet. It's usually wise not to go full Kelly, maybe half or quarter Kelly, so that we can emotionally deal with swings, losses, and avoid wiping out if we were way off with our fair percentage.
Additionally, for many bets, this isn't even possible for larger traders, there's simply not enough liquidity. It's also often not desirable because we can't fairly value everything. Some things just can't be estimated that well. Keep in mind there can also be information asymmetry, for example, if you're buying against insiders. Don't take huge bet sizes, especially when you're starting out.
Most successful traders don't use strict Kelly criterion for all their bets but they have a good intuitive sense of it, adjusted for their personal risk preferences. It's still a great value to go off of though.
For me personally, it's hard to intuitively adjust my bet sizing because my portfolio is growing so fast. I basically have to take bigger trades every day, but I'm not, I'm still trading as if my portfolio were at $10k. I've decided to use Kelly more and train myself with it.
For that, let's take a look at the Will Donald Trump win the Nobel Peace Prize in 2025? market. At the time of writing, No can be bought for 95ct but my fair value for No is 99ct - based on the Norwegian committee reportedly finalizing its decision on Monday, before Trump helped secure the ceasefire in Gaza, as well as their past decisions and general political tendencies. Plugging these two values into the Kelly formula gives us an optimal allocation of 80% of our bankroll. I won't go that far, but using quarter Kelly, I'll allocate 20% of my portfolio, or about $7,000. My expected return is 4.2% in less than one day. Of course, there's a chance I could lose, but I'm happy to take that risk for the expected return.
I'll see you tomorrow redeeming my biggest position yet.