There is one hour of the day quietly costing you money

Here is a pattern that shows up in almost every trading journal with more than a hundred trades in it: there is one hour of the day that loses money consistently, and the trader has never noticed.

Not a bad month. Not a bad instrument. One hour.

Why it hides

Nobody looks, because the hour is not obviously different from the hours around it. You are at the same desk, trading the same instruments, with the same strategy. The difference is usually one of three things:

1. Liquidity has changed. The London/New York overlap ends at 12:00 New York. What worked at 10:00 does not work at 14:00, and the setup looks identical on the chart.
2. You have changed. Four hours into a session, decisions get worse. The hour after lunch is a common culprit, and so is the hour after the first loss of the day.
3. It is the hour you trade out of boredom. The good setup came and went. What is left is the compulsion to have a position on.

None of these show up as "a bad strategy". They show up as an unremarkable hour with a quietly negative total.

How to find it

Group every closed trade by the hour of entry, in UTC, and sum the net P&L. That is it. You need about five trades in an hour before the number means anything, and thirty or more before you should act on it with confidence.

What you are looking for is not the worst hour — one hour will always be the worst. You are looking for an hour that is substantially worse than the rest, with enough trades behind it that variance is not a plausible explanation.

In Alpha Ledger this is the Worst hour window on the Analytics overview, and the coach will name it without being asked.

What to do about it

Stop trading it.

This is the entire intervention, and it is why it is the cheapest change available to you. You are not learning a new strategy, refining an entry, or developing patience. You are closing the platform for sixty minutes.

The traders who resist this usually say some version of "but some of my best trades happen then". Check. The number already includes them.

A worked example

One of the accounts in the Alpha Ledger demo data has this profile:

• 16:00–17:00 UTC: 38 trades, 24% win rate, −$6,340
• Every other hour combined: 241 trades, 47% win rate, +$9,054

That account is up about $2,700 for the period. Remove one hour from the schedule and it is up nine thousand. Nothing else changes — same strategy, same instruments, same risk.

That is the whole argument for measuring when you trade, not just what.