Swisstony: $1.85 Billion of Volume and a 1% PnL Margin

Swisstony (0x204f72f35326db932158cba6adff0b9a1da95e14) ranks second on Polymarket's all-time PnL leaderboard with $18,482,390.76. The same row ranks first for volume, at $1,850,584,673.06. That combination is the whole story: the account has processed an enormous amount of turnover and kept a reported margin of about 1%.

This is a different animal from a concentrated event bettor. The public profile says 203,104 markets traded, a biggest win of $1.2 million, and no open positions on 8 November 2026. The visible trades are small, frequent, and spread across lower-profile sports markets. The number is big; the edge per trade is thin.

The scale

The Data API snapshot showed $18.48 million of PnL on $1.85 billion of volume, with volume rank 1 and PnL rank 2. Figures change; these are point-in-time readings. The profile page displayed a higher total, +$23.5 million, on the same 1.8 billion of volume, along with a past-day figure of -$34.75. The profile also showed the account joined in July 2025, had 1.4 million profile views, and carried the bio line "So long, and thanks for all the fish."

The gap between the two PnL totals is worth pausing on. The leaderboard and the profile disagree by roughly $5 million, or about 27% of the smaller figure. Different surfaces can use different windows, treat rebates or rewards differently, or update at different times. The lesson is not that one is wrong. The lesson is that "the PnL" is not a single immutable object. Pick one surface, record the snapshot date, and disclose it.

What the activity looks like

The public trade feed is a stream of small positions in markets that most casual Polymarket users never open. Recent examples include:

• A $7.50 buy of Preston North End +1.5 against Everton in an EFL Cup spread market.
• A $92.37 buy of Lea Boskovic at 92 cents in an ITF W50 women's tennis match.
• Several fills in a K-League total, including 1,164.67 shares at 33 cents for about $397.
• Redemptions in Premier League halftime markets, a KBO baseball game, and a USL Championship soccer match.

The sizes are modest per order. The count is not. A profile counter of 203,104 markets traded implies an industrial process: many small decisions, many fills, and presumably many positions that are opened and closed or held to resolution. The most recent trade in the feed is dated 26 August 2026, so the account was active in the months before this snapshot even though it had no open positions on the snapshot date.

That matters for interpretation. A high-volume account with no open positions is flat at a point in time, not inactive. The all-time PnL is the sum of a very large number of completed outcomes.

The arithmetic of a thin margin

Reported PnL is about 1.00% of reported volume. Put differently, the account's reported profit is about 1% of its reported turnover. Theo4, by contrast, shows PnL equal to about 51% of volume. Those ratios are not comparable returns on capital, because volume is a turnover measure and capital is not public. But they do describe two very different operating models.

A 1% gross margin is vulnerable to costs that a concentrated bettor never faces. Slippage of a cent on a dollar-priced contract consumes the entire margin. A systematic error repeated across tens of thousands of fills compounds in the wrong direction. Adverse selection in fast sports markets, where the price moves while an order is in flight, can turn a small theoretical edge into a loss. Polymarket's fee schedule and rebate programs matter here: a strategy that looks profitable before costs can be marginal after them, and a strategy that relies on rebates is not the same as one that relies on forecasting.

This is also why the PnL/volume ratio should not be read as a "return." If an account turns over the same dollar ten times, volume rises tenfold while capital does not. The ratio measures how much profit the account extracts per unit of trading activity. It does not measure how much profit it extracts per dollar of equity.

Risk profile

The visible risk profile is high-frequency and diversified by market count, but not necessarily diversified by underlying exposure. Sports markets cluster: a set of Premier League totals, spreads, and exact-score markets on the same afternoon are correlated through the same matches, the same weather, and the same team news. A hundred small positions across ten games can be less diversified than ten positions across ten asset classes.

There is also operational risk. An account trading 200,000 markets depends on connectivity, order management, and fast data. A latency problem or a mispriced feed can produce a burst of bad fills that a slower trader would never experience. None of that appears in a leaderboard row.

The snapshot does show one useful risk datum: a past-day PnL of -$34.75. That is tiny relative to an $18.5 million all-time total, which is consistent with a high-volume, low-variance process, but a single day is not a drawdown measurement. We do not know the worst day, the worst week, or the largest peak-to-trough decline.

What the numbers do not prove

The public record proves that the wallet generated a large cumulative PnL and an even larger volume. It does not prove:

• That the PnL came from forecasting rather than from market making, rebates, or liquidity rewards.
• That the account has positive expectancy after all costs.
• That the edge is stable across sports, seasons, or rule changes.
• That a copier could achieve the same fills at the same prices.
• That the account is one person, one strategy, or one wallet.
• That the all-time PnL is fully realized. The snapshot shows no open positions, so there is no visible paper gain, but that is a point-in-time fact.

Most importantly, high PnL is not the same as copyable edge. A 203,104-market operation is not something a retail copier can mirror manually. Even an automated copier faces a structural disadvantage: it sees the trade after the wallet's fill, and it competes for the same liquidity with a slower connection. The original account's edge may be partly the ability to be first, which is precisely what a copier is not.

For an account with open positions, the same leaderboard figure could include unrealized gains. Here the snapshot shows none.

How to use this case

Swisstony is the best available example of the volume archetype at the top of the leaderboard. The right questions are not "what did they buy?" but "what does one cent of margin require?" and "what happens to that margin when I add my own slippage?"

If you are evaluating a high-volume trader, ask for the ratio of PnL to volume, the concentration of PnL by market, the share of PnL that comes from rebates or rewards, and the worst rolling drawdown. The public leaderboard gives you the first number and none of the others. Treat the rank as a starting point for investigation, not a verdict.

Public PnL is not proof of skill. It is a record of outcomes, and in this case the outcome is a very large turnover business with a very thin reported margin.