How To Go Broke While Being Right

**By **@allquantor

The magic number

In our previous episodes, we covered Base Risk (being wrong), Liquidity Risk (being stuck), Resolution Risk (being cheated), and Time Risk (being bored).

Let’s say you have solved all of them. You are the smartest person in the room. You found a market on Polymarket: *"Will the Sun rise tomorrow?" *trading at 90 cents. You know for a fact the real odds are 99.9%. You have an Edge.

Now you have a problem. You have to decide how much money to bet.

If you bet $1, you are wasting your time. If you bet your life savings, and the Oracle gets hacked, you are homeless.

Luckily, there is a formula for this. It is called the Kelly Criterion. It was invented by a physicist at Bell Labs in the 1950s who liked horse racing. It is mathematically proven to grow your wealth faster than any other strategy. It is the "Correct Answer."

It is also the fastest way to go bankrupt.

Here is the funny thing about the Kelly Criterion: It is extremely aggressive.

If you have a coin that lands Heads 60% of the time, and you get paid even money ($1 wins $1), Kelly says you should bet 20% of your entire net worth on the next flip.

Not 20% of your "gambling budget." 20% of your money.

If you do this, and you are right, you get rich exponentially fast. You will own the world.

But there is a catch. The "Kelly Curve" is not a gentle hill - it is a cliff.

• 1.0x Kelly: Maximum Wealth.

• 1.5x Kelly: Massive Volatility (you get rich, but you have heart attacks).

• 2.0x Kelly: Guaranteed Ruin.

If you bet twice the optimal size, your long-term expected growth rate is negative. Even if you have an edge! The math of variance guarantees that you will eventually dig a hole so deep you can never climb out.

The problem with the Kelly formula is that it requires an input variable called [ *p -- * your probability of winning ].

The formula assumes that when you type in "p = 0.8$," you actually know that the probability is 80%.

But you don't know that. You are just a guy who read a Substack.

This is called Model Error

• You think the odds are 80%.

• The Universe (or the rigorous market makers) knows the odds are 60%.

If you bet based on 80%, but the reality is 60%, you are accidentally betting 2x Kelly. You have walked off the cliff. You are betting like a genius, but you are actually betting like a suicide bomber.

In prediction markets, where the "True Probability" is inherently unknowable (there is no reference class for "Will Kanye West run for President in 2028"), your estimate of p is almost certainly wrong. And since humans are overconfident, it is almost certainly wrong in the dangerous direction.

Quarter-Kelly: The Strategy for the Rich Coward

Because of this, professional gamblers, hedge fund managers, and people who like staying rich do something very specific.

They calculate the "Optimal" Kelly bet. And then they divide it by four.

This is called Quarter-Kelly.

• The downside: You get rich slower. You leave money on the table. You feel like a coward.

• The upside: You survive.

A Quarter-Kelly bettor has a "Moat of Safety." You can be wrong about your edge. You can be wrong about the odds. You can even be wrong about the payout. And you will still make money.

The difference between a "Legendary Trader" and a "Bankrupt Trader" is usually not that the Legend is smarter. It’s that the Legend bet 5% of his portfolio, and the Bankrupt guy bet 20%.

To get rich, you have to be right. To stay rich, you have to be a coward.

If you find a "Sure Thing" on Polymarket and you bet your whole portfolio on it, you aren't an investor. You are just a guy holding a lottery ticket that hasn't expired yet.

We have built the Checklist of Doom. Before you click "Buy," ask:

Base Risk: Am I wrong?

Liquidity Risk: Can I leave?

Resolution Risk: Is the judge Steve?

Time Risk: Is the money dead?

Sizing Risk: Am I betting like a robot or a human?