Robinhood trade journal: what your fills look like as behavior stats

A Robinhood account keeps a complete record of what you did: every fill, with a timestamp, a side, a size and a price. A trade journal is usually where a trader adds the part that record cannot hold — the reason, the mood, the plan. This page is about the layer most journals skip: the statistics a Robinhood fill history already contains before anyone writes a word about it.

How to export your fills from Robinhood

Robinhood’s help center describes an account activity report. The steps, as written there (checked 2026-08-26):

Select Account activity report
Select Generate new report
In Customize your report, select the account type, the start date, and the end date for the report, and then select Generate report

The export arrives as CSV. A few sentences from the same page are worth reading before you plan around the file: “Most reports take 2 hours to generate, but may take up to 24 hours.” “These reports include all your transactions within your brokerage and retirement accounts.” “Your futures, crypto, and spending account activity aren’t included in these reports.”

Source: https://robinhood.com/us/en/support/articles/finding-your-reports-and-statements/

What the export can and cannot tell you

A fill history is a sequence. It knows when each position was opened and closed, how large it was, and at what prices. From that alone you can measure how long positions were held, how quickly a symbol was re-entered after an exit, and where exits landed relative to the path the position took while it was open. What the file cannot know is why. The plan, the news you were reading, the state you were in — none of that is in the export, and that is the part a written journal exists for.

The two layers answer different questions. A journal answers “what was I thinking?” The fill history answers “what did I actually do, and how often?” Most traders have a version of the first and have never looked at the second, because building the second by hand means sorting hundreds of rows into pairs and computing durations for each one.

One caution that applies to every number on this page: a small sample is a sketch. A distribution built from a handful of fills has a shape that will change with the next few trades. The shape only starts to hold once there are enough closed positions for the median and the spread to stop moving around.

What a behavior view measures on that file

biaX is a trading-behavior analytics app. It is not a P&L tracker, not a signal service, and not a course. It reads fills — the same rows in the export above — and describes patterns in how the account was traded.

The measurements are descriptive. Each one is a distribution, not a verdict:

Holding time by outcome. How long winning positions were held versus losing ones, shown as two distributions side by side rather than a single average.
Re-entry after an exit. How often, and how quickly, the same symbol was re-entered after a stop-out or a close, and how the re-entered positions compare with the originals.
Exit timing versus the position’s peak. Where each exit landed relative to the highest point the position reached while it was open — the same idea the app’s store listing calls “exit vs. peak.”
Trade frequency. Fills per day and per week, as a series, so a change in pace is visible as a number rather than a feeling.
Tagged entries. If you tag fills with what was going on at the time, the app shows how the tagged group’s holding-time distribution differs from the rest — an association between a tag and a behavior, not a claim about what the tag caused.

None of these say what to do next. They show what the record shows, and they show it the same way for every account.

The research behind the patterns

These measurements exist because the patterns are documented, not because they are clever. Two anchors carry most of the weight:

Kahneman & Tversky (1979), “Prospect Theory: An Analysis of Decision under Risk” (Econometrica 47). Losses loom larger than equivalent upside moves, and outcomes are valued relative to a reference point rather than in absolute terms. In a fill history the reference point is usually the entry price, which is why holding-time and exit-timing distributions tend to split around it.
Odean (1998), “Are Investors Reluctant to Realize Their Losses?” (Journal of Finance 53). In discount-brokerage account data, investors realized winning positions at roughly 1.5x the rate of losing ones. That asymmetry is the disposition effect, and in an individual account it shows up as a skew in holding times — not as a verdict on any single decision.

The research explains why a behavior happens. It does not say what will happen to you, and neither does the statistic. A distribution is a description of a history; it carries no forecast.

Reading a distribution without a verdict

The useful habit with these charts is to read them the way an analyst reads any two histograms: where is the middle, how wide is the spread, and do the two groups overlap or separate? A holding-time chart where the losing-position curve sits far to the right of the winning-position curve is a pattern with a name in the literature above. It is not a grade. Two accounts can show the same skew for entirely different reasons, and the fill history cannot tell those reasons apart — which is, again, where the journal comes back in.

Seeing it before touching your own data

If you would rather look at the measurements on a constructed sample first, there is a browser-only demo at https://biax.app/demo?ct=blog — no account needed. When you are ready to look at your own Robinhood history, the export above is the raw material: the same timestamps, sides, sizes and prices, read as a distribution instead of a list.