My P&L Doesn't Match Polymarket
Why your Polycopy portfolio value can differ from Polymarket, what was fixed on 2026-08-21, which small differences are normal, and how to compare the two numbers.
Last updated August 21, 2026
Your headline portfolio value on Polycopy now comes from Polymarket's own valuation, so the two numbers should closely match. Small differences in per-trade figures are normal and explained below. If you saw a large mismatch before 2026-08-21, it was real, and it has been fixed.
What was wrong before 2026-08-21
Before that date, two defects could overstate your losses. Winning positions in resolved markets could be booked as total losses, and very large portfolios could be only partially counted. Both are fixed. Polycopy now reads your portfolio value from Polymarket directly, so the corrected number is also the current one. There is nothing you need to do to get the corrected figures.
Small differences are normal
Polycopy's per-trade records cover trades made through Polycopy, meaning bot copies and quick copies. Positions you opened directly on Polymarket count toward your total value but have no Polycopy trade history behind them. So expect the headline value to match closely, while trade-level P&L on Polycopy describes your Polycopy activity rather than everything you have ever done on Polymarket. Prices also move constantly, so two pages loaded a minute apart can differ by a little.
How to compare the two numbers
Open both portfolios
Open your Polycopy Portfolio in one tab and polymarket.com/portfolio in another, then refresh both so they price at the same moment.
Compare the totals
The headline values come from the same source, so a small drift from live price movement is all you can expect to see between them.
If the gap is large, send us the details
Use the Get Help form (the ? icon in the top navigation). Include both totals, screenshots of both pages, the time you compared them, and any specific market whose value looks wrong. That lets us check the exact positions instead of guessing.