Fishalive: Two Markets, $9 Million, and a Step-Shaped PnL Curve

Fishalive (0xed64a7bf029040aa331abc87902434d815ef217d) ranked seventh on Polymarket's all-time PnL leaderboard on 8 November 2026 with $8,998,446.90 in reported profit on $13,281,460.37 of volume. Reported PnL is about 68% of reported volume, the highest ratio among the seven names in this series.

The profile explains why. It says two markets traded. The account joined in June 2026, recorded a biggest win of $4.7 million, and held no open positions on the snapshot date. Both visible markets are legs of the same 2026 World Cup match: Spain versus Cape Verde.

This is not a diversified record. It is one event, expressed twice. That makes it a clean case study in concentration, liquidity, and why a headline PnL is not a strategy.

The numbers

The Data API row showed $8,998,446.90 of PnL, $13,281,460.37 of volume, PnL rank 7, and volume rank 2,205. Figures change; these are point-in-time readings. The volume rank is the lowest of the seven. The profile page showed +$9.1 million, 75,200 profile views, two markets traded, a biggest win of $4.7 million, and $0.00 in positions value. The profile total is a rounded display; the Data API figure is the precise leaderboard value.

The biggest win is about 52% of the all-time total. The average PnL per listed market is roughly $4.5 million. With two markets, that average is not a statistic; it is a description of two outcomes.

Two markets, one match

The public activity is dated June 2026 and centers on Spain versus Cape Verde. The visible settlements are:

• $8,540,656.48 on "Spread: Spain (-2.5)" with the Cabo Verde side.
• $4,740,803.89 on "Will Spain win on 2026-06-15?" with the No side.
• A maker rebate of $2,463.11 and two small rewards of $1.09 and $1.82.

The two redemptions are not independent bets. They are two ways of expressing a view about the same match: Spain failing to cover a 2.5-goal spread, and Spain failing to win outright. The outcomes are correlated. If Spain wins by three or more, both legs lose. If Spain wins by one or two, the spread leg wins and the outright leg loses.

That correlation is the first thing a copy trader should notice. A profile that says "two markets" sounds more diversified than a profile that says "one match." It is not. The effective sample size is one event.

The volume coincidence

The two visible redemptions sum to $13,281,460.37. That is exactly the reported volume on the leaderboard. The match is precise to the cent, which is unusual enough to be worth flagging.

This does not prove that the leaderboard's volume field is calculated from redemptions. It could be a coincidence, or the volume could be dominated by these two settlements because there were few other trades. But it does mean the volume figure is not an independent check on the PnL. When volume and settlement value are the same number, the PnL-to-volume ratio is not measuring trading intensity in the usual sense. It is measuring profit relative to the size of two payouts.

That matters for interpretation. A 68% PnL-to-volume ratio looks extraordinary next to swisstony's 1% or RN1's 0.9%. But those accounts turn over a book many times; fishalive's volume is essentially the settlement value of one event. The ratios are not comparable. This is why the PnL-to-volume ratio should be used as a description of the account's activity, never as a return on capital.

Flat, then a step

With two markets held to resolution, the realized PnL path is step-shaped. There is no visible compounding, no series of small wins and losses, and no evidence of how the account behaved between entry and settlement. The realized curve is flat, then a step, then flat again.

That shape has different risk properties from a high-frequency book. A high-frequency trader accumulates many small results and can estimate an edge from the distribution. A two-event trader has one draw from a binary distribution. The result is large, but the confidence interval around the underlying skill is enormous.

It also means the public record contains no drawdown information. A two-market record can show no visible losing streak, even if one leg lost money. That does not mean the account never lost; it means the sample is too small to show a loss. The absence of a drawdown in the data is not evidence of risk control.

Liquidity and concentration

The redemption sizes imply a large position. $8.54 million on a Spain -2.5 spread is not a retail-sized bet. The visible fills in the activity feed are much smaller, including buys of 9.43, 377.36, 18.30, 283.02, and 18.87 shares at around 47 cents. Those fills total a few hundred dollars, not millions. The redemption therefore reflects a position accumulated across activity that is not fully visible on the page, or a settlement of a much larger holding.

The public data does not show the order book at the time of entry. An $8.5 million position in a national-team spread market may have required working an order over time, or it may have been built in a market with more depth than the visible fills suggest. We cannot tell. What we can say is that the position size is far beyond what most copiers could replicate, and that the price impact of a large order is invisible in the final PnL.

Concentration cuts both ways. Two correct legs produced $9 million. Two incorrect legs would have produced a loss. The concentration is the source of the result, not a detail beside it.

Copy-trading caveats

There is nothing to copy now. The positions are closed, the account had no open positions on snapshot day, and the newest activity is dated 16 June 2026. The event is over.

Even during the event, copying would have been difficult for four reasons:

1. The two legs were correlated, so copying both would have concentrated exposure in the same match rather than diversifying it.
2. The position sizes were large relative to typical retail capital, so proportional copying would have been impossible for most users.
3. The entry prices are not fully visible, so a copier could not know the cost basis or the expected return.
4. The settlement value is not the same as profit. A copier who mirrored the payout without the cost basis would overstate the edge.

The broader lesson is that a two-market record is a sample of one event, and one event is not a strategy. It can be a very good outcome. It cannot support a claim about repeatability.

What the data cannot tell you

The public record does not show the bankroll, deposits, withdrawals, or the full order history. It does not show whether the account had other losing positions in the same match or in other matches. It does not show whether the two legs were part of a hedge with a third position. It does not show the maximum drawdown, because there is no time series with enough observations.

It also does not show whether the result came from a superior read of the match or from a favorable outcome. The data records the settlement, not the reasoning. Public PnL is not proof of skill, and in a two-market record the distinction between skill and outcome is especially hard to draw.

Leaderboard PnL can include unrealized gains when positions are open. Fishalive had none on this date, so the reported figure was settled.

Fishalive is the purest concentration case on the leaderboard. The number is large, the sample is tiny, and the PnL curve is a step. Read it as a case study in event risk, not as a template.