Polymarket Holding Rewards: The Complete 3.25% APY Guide
Strategy · 5 min read · published 2026-04-20 · updated 2026-07-25
Polymarket pays holding rewards - an annualised yield, around 3.25% on the headline programme, distributed to wallets that hold qualifying shares. On its own that is a savings-account return. Stacked on top of the discount built into near-certain markets, it turns a boring position into something closer to short-duration credit. This guide covers the mechanics, the arithmetic, and the four ways the trade goes wrong.
How holding rewards work
For each eligible market, a daily reward pool is distributed pro rata to the wallets holding qualifying shares. Three variables decide what you actually receive:
- Eligibility. Only specific markets are in the programme - typically high-profile, longer-dated events with real liquidity. A market being large is not enough; it has to be included.
- Dilution. The pool is fixed and your slice is proportional. When capital floods into a market, the advertised rate is not what the last person in earns.
- Time held. Rewards accrue continuously while the shares sit in your wallet. Selling ends accrual on that lot immediately - there is no credit for having held it last week.
Holding rewards are distinct from liquidity provider rewards, which pay for quoting inside a maximum spread, and from maker rebates, which return a share of collected fees to resting orders. It is possible to earn more than one of them at once, and worth knowing which is which before you attribute a payment to the wrong source.
The arithmetic that matters
A near-certain market carries two separate returns, and you need both.
The price discount. Buy at 99.5¢ and you make 0.5¢ per share at resolution. The return is (1 − price) / price, annualised by 365 / days remaining. Over 30 days that 0.5¢ is roughly 6% a year. Over 90 days the same half-cent is barely 2%.
The reward yield. Roughly 3.25% annualised on the value held, subject to dilution, paid daily while you hold.
Together, a 30-day position at 99.5¢ in an eligible market runs at about 9% a year. The headline is attractive; the sensitivity is the point. Move the same position out to 120 days and the discount contributes about 1.5%, so almost all of your return now comes from a reward programme that can be changed at any time. Time to resolution is the dominant variable, and it is the one people ignore.
Worked example
$10,000 into NO at 99.4¢ on a market resolving in 45 days:
- Shares bought: about 10,060.
- Payout at resolution: $10,060. Profit $60, or 0.6% over 45 days - about 4.9% annualised.
- Holding rewards at 3.25% for 45 days: roughly $40.
- Combined: about $100, or 8.1% annualised - before fees and before any slippage on entry.
Now assume you crossed a 0.4¢ spread getting in. That is $40 gone, and your annualised return falls to around 4.9%. The execution is not a detail; on this kind of trade it is most of the outcome.
The playbook
- Screen by annualised return, not price. Our 99%+ page ranks by annualised return precisely because 99.5¢ tells you nothing without the end date.
- Check reward eligibility separately. The holding rewards page lists markets currently paying, with the live rate. A cheap market that is not in the programme is a different trade.
- Read the resolution rules. On a 99% market you are being paid half a cent to accept the tail. Understanding exactly what would trigger the other side is the whole job.
- Enter with limit orders. Books at the extremes are thin above the top level. Crossing the spread can cost more than the trade earns.
- Size to the book, not to the balance. If clearing your position would move the price 50 basis points, the position is too big.
- Diversify across 5-10 markets. The failure mode here is not a slow bleed, it is one position going to zero. Spread the idiosyncratic risk.
- Track end dates. Rewards stop at resolution and the capital sits idle until you redeem and redeploy it. On a 5% strategy, a week of idle capital is a meaningful share of the year's return.
What can go wrong
- Resolution risk. The big one. A 99% market that fails costs you 99¢ to have earned half a cent - you need roughly 200 successful cycles to absorb one failure. Ambiguous wording and disputed outcomes, not surprise events, are the usual culprits.
- Programme risk. Reward rates are set by the platform and have been adjusted before. Nothing here is contractual.
- Dilution. The advertised rate is the pool divided by current participation. It falls as others join.
- Execution cost. Spread and slippage routinely exceed the entire expected profit on these trades. This is the most common reason a strategy that looks good on a screener loses money in practice.
- Opportunity cost. Capital locked for months at 8% is not available for the mispricing you find next week.
Operational details worth knowing
- Merging beats waiting. Holding matched YES and NO in the same market? Merging converts the pair straight back into collateral without waiting for resolution.
- Redeem promptly. Winning shares are not cash until redeemed, and they earn nothing while they wait.
- Approvals are per adapter. Binary and neg-risk markets settle through different collateral adapters, each needing its own approval - see the CLOB V2 reference.
- Watch fees at the edges. The taker fee formula scales with uncertainty, so it is small near 99¢ - one of the few things working in this strategy's favour. Details in the fee guide.
Is it worth it?
Honestly: at single-digit annualised returns, this only makes sense at size, with disciplined execution, and with genuine diversification. It is not a way to turn $500 into anything interesting. It is a way to park capital you would otherwise leave idle, at a return that beats most on-chain alternatives, in exchange for accepting a small, well-understood tail risk - and for doing the unglamorous work of reading resolution rules.
The live list, sorted by current effective rate, is on the holding rewards page. Cross-reference with 99%+ bets to find the markets where both returns apply. This is a description of a strategy, not financial advice.