🧵 What is Split/Merge on Prediction Markets (And how to make money with it)

Here’s what you are probably doing on prediction markets right now: Say you think Trump will meet Kim Jong Un this year. The market prices YES at 1.8¢. You buy 100 YES shares at and hope it resolves YES, or sell it if the odds move up.

There is another way to make money by providing liquidity to the markets: Recall - the price of one YES and one NO share always add up to $1. If you have 1 YES + 1 NO, then at resolution one goes to $1 and the other to $0 -> a fully hedged position.

Now let’s say you want to provide liquidity on the market “Will Trump Media buy ETH in 2025?” To start making markets, you need inventory, which means both YES and NO shares.

Where are you gonna get these shares? Buy them off the markets? -> NO! Look at the spread on this market. People are selling Yes shares at 31c and selling No Shares at 93c. If you want 1 Yes + No, you need to spend 124c - terrible!

What is the better way? Use the Split function. You can take your 100 USDC and split it into 100 YES + 100 NO at no premium. Now you can place limit orders like: – Sell YES at 30¢ (below current ask) – Sell NO at any 80 ¢ (above current bid)

As long as they add up to more than 100c, you make money if the orders get filled. This is how liquidity is provided in prediction markets. But you do hold inventory risk -> if the market moves in one direction suddenly, you sold YES too cheap and are stuck holding devalued NO shares.

Merge is simply the opposite -> you combine 1 YES + 1 NO back into 1 USDC. Now, going back to this market "will Trump Media buy ETH in 2025", the spread is a whopping 24.3c! Newly created markets are often inefficient and ripe with opportunities.

Do it wisely and you could be making money being a maker instead of taker on prediction markets.