Your journal's P&L never matches the broker statement. Check these six things, in this order.
Six reasons a trading journal disagrees with the statement, in the order to check them: fees, missing times, option legs, duplicates, currency, partial fills.
Published by TradeFlow Quantum. Filed under reconciliation, import, brokers, measurement.
A journal that does not match the statement is not a journal with slightly wrong charts. Every statistic built on it — expectancy, drawdown, hour of day, setup comparisons — inherits the gap silently, and the one number a trader can check against reality is the one that is off. The good news is that the gap is almost always one of six causes, each with a thirty-second test. Run them in this order; the first three catch nearly every case.
- Fees and commissions are missing. Most broker exports and activity endpoints carry the fill price and quantity and leave out commission and exchange fees, so the journal computes gross P&L while the statement shows net. Test: compute (exit − entry) × quantity × multiplier for one trade by hand; if the journal agrees with your arithmetic but not the statement, fees are the gap. In TradeFlow Quantum 9 of the 10 direct connections import gross because the broker's endpoint carries no fees; Coinbase deducts the commission it reports. A CSV with an exit price on each row keeps the realised-P&L column the preset maps. A per-fill export gets P&L derived from its paired fills, even when it carries a P&L column, so it is gross.
- The export has dates but no execution times. A date-only row is stored at midnight UTC, which is the previous evening in US Eastern, so every trade files a day early and every hour-of-day chart shows one spike at 7 PM. Test: if every trade is exactly one day early, the file had no times. TradeFlow Quantum files date-only rows on the written date and says how many there were; the full account is in the date-only export post.
- Option legs were recorded as separate trades. A vertical or an iron condor is one position with one P&L, but the export lists two to four fills, and a journal that keys trades by ticker shows legs that each look like a win or a loss and leaves a phantom position when a closing leg is dropped. Test: count the journal's trades for one spread you opened and closed; more than one means legs were not grouped. Since November 9, 2026 the Alpaca, Charles Schwab, Tradier, Tastytrade, Questrade and E*TRADE connections import options through a grouping engine — each fill read from its OCC symbol, matched to its own contract, legs opened together grouped as one spread; the other direct connections leave option fills out rather than import them as mis-sized stock.
- The same file was imported twice. Trade count and P&L roughly double. Test: look for pairs of identical rows. TradeFlow Quantum keys every row on user, symbol, open time, side, entry, quantity and an id — the broker's execution id on a direct connection, one derived from the row's content on a CSV — and reports repeats as already in your journal instead of inserting them.
- The account's currency is not the journal's. A CAD or EUR statement against a journal assuming USD disagrees by the exchange rate; forex P&L quoted in pips disagrees by the lot size. Test: confirm the account currency on the per-broker page before comparing a single number.
- Partial fills were merged differently. Three fills at three prices show as one average on the statement and up to three rows in a journal that does not merge them. Test: compare one partially filled order's row count. Whether a direct connection's partial fills arrive merged depends on the broker's history: TradeStation's and Schwab's report a transaction per order, others one per fill. CSV executions stay as the file lists them.
What to do once it matches
Then the rest of the product means something. The features that read closed trades — the P&L calendar, the weekly digest, tags, the mistake detector — are only as good as the rows under them, and the prop copilot measures drawdown, daily loss and the consistency rule against the same rows, so a reconciled journal is the difference between a cushion you can act on and a number you have to second-guess.
The long form of this list, with the vendor help pages that admit each failure mode, is the answer page Why doesn't my trading journal P&L match my broker statement?.
$15/month or $150/year, one tier, 7-day trial with a card at signup and $0 charged during it. The ten direct broker connections and CSV import work during the trial.
Related posts
Not financial advice. This post reflects the author’s opinion based on publicly-available information at the time of writing. Mention of third-party products is not an endorsement; product features and prices change over time. Past performance does not guarantee future results.