On Nimb, a wallet's position in a market is its net exposure: Yes shares minus No shares. A wallet that holds both sides has matched Yes-and-No pairs that pay $1 whatever happens, so Nimb nets them out and leaves them out of the holder lean and the P&L. Only the side the wallet is net long counts.
Netting keeps market reading honest: without it, a wallet with no stake in the outcome can look like the largest holder on both sides at once.
What is net exposure?
Net exposure is the number of Yes shares a wallet holds in a market minus its No shares: the part of its position that gains or loses depending on the outcome. A hedged position is the matched part: for every Yes share there is a No share, and each pair pays exactly $1 at resolution, whichever way the market resolves.
Such a pair is a mergeable set: on Polymarket, one Yes and one No share of the same market can be merged back into $1 of collateral at any time, and $1 of collateral can be split into one of each.
Why would a wallet hold both Yes and No?
- Market making. A wallet quoting both sides ends up with inventory on each.
- Splitting. Turning $1 into a Yes and a No share, then selling one side, leaves the other; partial sales leave both.
- Changing its mind. Buying No later instead of selling Yes has the same effect as selling, at a different price.
- Rewards and liquidity programmes that pay for resting orders on both sides.
None of these is a view on the outcome for the matched part.
What does netting look like in numbers?
Take a market at 70%. A wallet holds 1,000 Yes shares and 600 No shares.
| Shares | Price | Value | |
|---|---|---|---|
| Yes held | 1,000 | $0.70 | $700 |
| No held | 600 | $0.30 | $180 |
| Gross | 1,600 | $880 | |
| Hedged: 600 Yes + 600 No sets | 600 sets | $1.00 | $600 |
| Net exposure: 400 Yes | 400 | $0.70 | $280 |
Now Yes moves from 70% to 80%. Counted gross, Yes gains $100 and No loses $60, and a table would show a large winner and a large loser. Counted net, the wallet gained $40: 400 shares × $0.10. The $600 in sets did not move. On Nimb, this wallet appears once, as a Yes holder with a $280 position, and its P&L reflects the 400 net shares.
Why do gross numbers mislead?
Nimb's own tests found the problem on real data. In the 2028 US presidential market, one wallet held 4.75 million Yes and 4.75 million No shares of "JD Vance". Counted gross, it showed about ±$1.49 million in P&L, a large gain on one side and a matching loss on the other, for a net position of $0. Drawn as two positions, it was the largest holder on both sides of a market in which it had no stake.
The pattern was not rare. Across the snapshot reviewed (2026-11-07), 93 wallet-market pairs held both sides, and their 58.3 million gross shares carried only 0.55 million shares of net exposure. About 99% of that gross was locked $1 sets.
Gross numbers inflate the size of holders, distort the split between Yes and No, and turn hedged wallets into apparent whales. That is why the How Nimb works page says of such a position: "It is not a bet, so Nimb leaves it out of the leaning and the P&L."
How does Nimb net positions?
- For each wallet and each market, add up its Yes shares and its No shares from Polymarket's public holder data.
- Net = Yes − No. A positive net is a Yes holding; a negative one is a No holding of the same size.
- Drop the wallet from that market if the net position is worth less than $500 at the current price.
- Value the net shares at the current price of their side, and compute unrealised P&L on those shares from their entry price, instead of using the API's P&L figure, which can include the hedged leg.
- Use only the net side in the map's lines, the "Who is positioned" split and the concentration checks.
In holder tables, the panel note "Positions that hold both sides are netted" marks this. The market pages under /markets say: "Wallets holding both sides of the same market are left out: their positions cancel and are not a bet." See top holders and wallet lines for how the tables read.
How does netting work across many-outcome events?
In a multi-outcome event, such as an election with ten candidates, each candidate is its own Yes/No market, and Nimb nets each one separately. That misses overlaps across outcomes. In a neg-risk event, where exactly one outcome can win, No on one candidate is economically close to Yes on all the others, and Polymarket lets holders convert between them. A wallet holding No on candidate A and Yes on candidate B is therefore partly hedged, even though each market looks one-sided.
Catching these overlaps needs netting at the level of the whole event. Nimb nets per market only, so it does not catch them. Read large holders of several outcomes in one event with care. Multi-outcome mechanics are on multi-outcome markets.
What can this not tell you?
Netting works within the data Nimb sees. It cannot see positions a trader holds in another wallet, on another venue or off-chain, so a wallet that looks one-sided on Nimb may be hedged elsewhere. It works on the 25 largest holders fetched per side, not on every holder. Other gaps are listed on known limitations; terms are in the glossary.
From reading to execution
Net exposure is what a holder actually has at stake; read it with price and flow before you use the order ticket. The workflow is in read and execute.
Last reviewed 2026-11-07 · Data from Polymarket · Not financial advice · Nimb is not affiliated with Polymarket.