Analytics → Time breaks your closed trades into four trading sessions, and the AI Insights widget on your dashboard can call out a session where your numbers stand apart. This article explains how a trade is assigned to a session, and why some of your trades may not appear in that chart at all.
The four sessions
They are read in US market hours and they don't overlap, so every trade lands in exactly one:
- Asia — 19:00–02:59
- London — 03:00–09:29
- New York — 09:30–15:59 (US regular hours)
- After hours — 16:00–18:59 (US post-market)
A trade needs an entry time
A session is worked out from the trade's entry time — the same field P&L by hour of day uses. If a trade has no entry time recorded, its session genuinely can't be known, so it is left out of the session chart rather than being dropped into a bucket that would be a guess. The card tells you exactly how many trades that is ("based on the 34 of 512 closed trades that have an entry time"), and the same note appears on any session insight on your dashboard.
This matters most on imported and auto-synced books, where the export often carries a date but no time. If your session chart looks thin, that is usually why — and it is fixable: add entry times to the trades you care about and they join the chart immediately. Nothing needs recalculating.
Which clock is used
Entry times are stored exactly as you (or your broker export) recorded them, and are read as US market hours — the same convention the trade chart uses when it places your entry and exit markers. If you record times in your own local zone and you are not on US Eastern time, the session labels will be shifted by that difference. Recording times in market hours keeps the session chart, the hour-of-day chart and your chart markers all telling the same story.
Reading a session result honestly
- Check the sample first. The note under the chart is there so a result from 40 trades never looks like a result from 400.
- Sessions overlap with other explanations. If you only trade one strategy in the morning, a "London edge" may really be a strategy result. Cross-check against Analytics → Performance.
- It describes what happened, not what will. These are your own past records summarised — nothing here is a prediction or a recommendation.
Fixed in v9.0.0
Before this release, session grouping in the dashboard's AI Insights widget used the time a trade was saved to your journal rather than the time it was entered. On a hand-logged trade those are usually close together, so the result looked reasonable. On an imported or auto-synced book they are not related at all: every trade in a batch was written at the same moment, so all of them landed in whichever session the import happened to run in — and the widget could report a "session edge" that was really just the hour you pressed Import.
Session figures are now taken from the entry time only, trades without one are excluded and counted, and the same rule is shared by the dashboard widget and the Analytics card so they can't drift apart again. If you acted on a session insight before this release, it is worth a second look — particularly if most of your journal was imported. Day-of-week grouping was tightened in the same change and now always uses the trade's own date.