If You Trade on Polymarket or Kalshi, You Already Know How to Bond

A market that settles a binary question on a hard deadline, where the edge belongs to whoever has better information than the crowd. If you trade on Polymarket or Kalshi, that sentence is your day job, and it is also a complete description of bonding.

Every Piece is a tiny market on one question: will backers hit this goal before the clock runs out? A bonder is the trader holding YES. So this post is written in your language: the price, the breakeven, and the two places where the math genuinely differs from every market you have traded.

The exchange rate

On Polymarket, you buy YES at a price, and the price is the whole trade: it fixes your payout and your breakeven before you commit.

A bond has the same property, computed instead of quoted. A funded Piece pays a fixed kickback, the kickback percentage times the raise, split pro-rata across the bond pool. The pool itself is capped at that same percentage of the funding goal. Both numbers are public before you stake, so your payout is arithmetic: your principal back, plus your share of the kickback. Divide stake by payout and you are holding YES at a price.

Run it on a $500 Piece with a 20% kickback, which caps the pool at $100 and pays a $100 kickback if it funds:

• Full pool: your $50 of the $100 pool earns $50 of the $100 kickback. Out with $100. On this Piece, that is YES at 50 cents.
• Pool at half its cap: your $50 is the only claim on that same $100. Out with $150. YES at 33 cents.

The crowd’s absence is what cheapens your entry, and the cap protects it: the pool cannot grow past the kickback it pays out, so later bonders can dilute your share only down to the full-pool split and no further. We walked those scenarios through dollar by dollar in the bonder’s field guide.

Before you go looking for underfilled pools, though, note the rule that kills the obvious play: bonding closes the instant a Piece funds. You cannot wait for a campaign to look certain and then buy YES cheap, because certainty and availability are mutually exclusive here. Every bond is placed while the ending is genuinely unknown.

One more familiar comfort: settlement has no oracle. Nobody adjudicates whether the goal was hit, no resolution gets disputed, and there is no UMA drama. The funds live in smart contracts on Base, the contract can see its own state, and the outcome pays out mechanically in both directions.

The twist: you’re allowed to push

Here is where the math stops being a translation and becomes a different game.

On Kalshi or Polymarket, the event is outside the market. You forecast an election; you do not get to vote a swing state into existence, and nudging the outcome you bet on is the thing regulators and resolution rules exist to prevent. Your probability estimate is a photograph of a world you cannot touch.

On Pieces, your stake changes the probability it is priced on. A funded bond pool is a public signal that people with money at risk expect this to succeed, and backers read it. The pool also funds the refund bonus, which is what makes backing safe enough to be the dominant move. (The mechanism is called a dominant assurance contract, and the name is literal: backing becomes the dominant strategy.) And nothing stops you from sharing the campaign, rallying the creator’s audience, or backing it yourself. Every one of those pushes YES toward resolution.

Think about what that does to expected value. You are buying YES at a price you computed yourself in a market where the purchase itself moves the true probability above whatever it was when you sized the trade. There is no other venue where that is legal, because everywhere else, someone is holding NO against you.

Which is the deeper difference: nobody holds NO. A prediction market is zero-sum by construction, a transfer from bad forecasts to good ones. Here, the other side of a funded Piece is not a losing trader. It is a creator getting paid up front and a finished work going public, for everyone. And when YES loses, the stake is not claimed by a sharper counterparty; it becomes the bonus paid to the backers who showed up. Money changes hands in both endings, but in neither one did your win require someone else’s read to be wrong. We keep the full comparison on the prediction markets page.

The honest ledger

What you gain and what you give up, plainly.

Pros:

• Your worst case is computable before you stake. The kickback and the pool cap are public, so the least a winning bond can pay is arithmetic, not a fill you hope for. Later bonders can dilute your share only until the pool is full, never past it.
• Your edge does not get arbitraged away. No market maker reprices the pool against you. If the crowd has not shown up, the whole kickback is yours to claim.
• You can move the odds. Promotion, signal, and backing all push your own position toward paying. Participation, not manipulation.
• Mechanical settlement. On-chain on Base, no oracle, no resolution disputes, gas covered, wallet custody stays yours.
• Defined window, set in advance. Every Piece has a deadline the creator picks and everyone can see, and the position resolves on it.
• Both endings fund something real. A win pays a creator and releases work. A loss pays the backers. There is no ending where the money just moved between speculators.

Cons:

• No exit. There is no selling YES at 92 cents on good news. A bond is locked until settlement, with no mark-to-market in between and no way to unwind a position you regret.
• Binary total loss. A campaign that fails at 92% of goal costs you exactly what one that failed at 5% would. The single exception: a Piece that ends with no backing at all returns bonds in full, because there is no one to pay the bonus to.
• Your share is not fixed when you stake. Until the Piece funds or the pool fills, every later bonder dilutes your slice of the kickback, and with no exit there is nothing you can do about it except have priced in the full-pool split from the start.
• One-sided market. You cannot short a creator you think will miss. Bonding can only express confidence.
• Capacity is capped, and small. The bond target is the kickback times the goal, so a modest kickback on a modest goal may accept only a few hundred dollars. Conviction does not scale past what the Piece needs.

Who all of this helps

Trace where the money lands and the roles line up instead of opposing each other. The creator is paid the raise up front, at their price, before the work is public, which is the entire reason the market exists. The backers get a market that is safe to enter early precisely because your stake is behind it. You get paid the kickback for reading a creator’s audience correctly, the same skill you have been renting out to order books, pointed at an outcome worth causing. And when it all resolves YES, the settlement is not a number on a chart. It is a piece of work that exists now, public, for everyone.

That is the trade. Same forecasting muscle, a market you are finally allowed to be part of, and a resolution you can listen to. See how the whole mechanism fits together, or go find a market you can read better than the pool can.