The Two-Sided Book: How to Make Money When You Have No Opinion

The Trade Where You Don't Need to Be Right

Most of prediction market strategy is about being smarter than the crowd. Better base rates, faster news reads, sharper probability estimates. It's exhausting, and most of the time you're wrong anyway.

There's another way to make money: stop picking sides and start selling both.

This is market making. You post a bid and an ask, you wait, and you collect the spread when someone crosses each side. You don't need an opinion on the event. You need an opinion on the spread.

The Basic Setup

Say a market is trading 62¢ bid, 65¢ ask on YES. That's a 3¢ spread.

A market maker parks orders on both sides — maybe 63¢ bid, 64¢ ask, tightening the spread just enough to be the best price. Now every impatient trader who wants to buy YES hits your 64¢. Every impatient trader who wants to sell hits your 63¢.

Each round trip: you bought at 63¢, sold at 64¢. 1¢ of pure edge, no directional risk.

Do that fifty times a day on a liquid market and the math starts to look interesting.

Why This Actually Works on Prediction Markets

On stocks, market making is a HFT arms race you'll lose. On Kalshi and Polymarket, most books are wide, most markets are ignored, and the competition is a handful of humans and a couple of scripts.

Spreads of 2-5¢ are common. Spreads of 10¢+ exist on anything niche. Meanwhile, event resolution puts a hard cap on how far prices can move — YES can't go above $1, NO can't go below $0. That bounded volatility is a market maker's dream.

The Inventory Problem

Here's where beginners blow up. You post both sides, but the market drifts. Suddenly your 63¢ bid gets filled ten times in a row and your 64¢ ask gets filled zero times. You're long 1,000 contracts of YES in a market that's now trading 55¢.

Congrats, you're not a market maker anymore. You're a bagholder with a strategy deck.

Rules that keep you alive:

Skew your quotes based on inventory. Long too much YES? Lower your bid, lower your ask. Make it easier to sell out, harder to accumulate more.

Set a hard inventory cap. If your max position is 500 contracts, at 500 you stop bidding entirely. You only offer.

Widen or pull when news hits. The fastest way to get run over is leaving stale quotes up when information changes. If you're not watching, don't quote.

Which Markets to Make

Good market-making candidates share three traits:

- Wide spreads relative to volatility. If the market moves 2¢ a day but the spread is 5¢, you're getting paid to warehouse tiny risk.

- Steady two-way flow. You need buyers AND sellers. A market where everyone agrees on direction (95¢ contracts trending up) will just eat your ask side over and over.

- Distant resolution. Markets resolving tomorrow have violent gamma. Markets resolving in three months chop sideways, which is exactly what you want.

Sports markets close to game time? Terrible for making. Political markets four months out with active but non-directional interest? Beautiful.

The Honest Part

Market making isn't free money. Your edge per trade is tiny, so a single bad fill can wipe out fifty good ones. You'll get adversely selected — the trader hitting your ask often knows something you don't, and the trader hitting your bid does too.

You survive by quoting small, adjusting fast, and accepting that some markets will chew you up. The goal isn't to win every trade. It's to be the house — slightly favored, boringly profitable, always in the game.