Weekly recap: July 6–13 — 25 trades, an 11-13 split, and a lesson in exit timing

The headline numbers

Between July 6 and July 13, the virtual engine opened and closed 25 trades across 16 symbols. Eleven closed in profit, thirteen closed at a loss, and one closed flat. That's a 45.8% win rate on trades with a nonzero result — below coin-flip, and the week's total P&L reflects it: -$224.81.

This is a losing week, reported as one. A weekly recap that only publishes when the numbers look good isn't a recap — it's marketing wearing a recap's clothes. The point of doing this on a fixed schedule is that the losing weeks show up in the archive exactly as often as they actually happen.

The six best closes

The six worst closes

The one pattern in the data worth naming

Average win: +0.93%. Average loss: -2.40%. Losses ran roughly 2.6x the size of wins this week. That asymmetry alone is enough to turn an above-50% win rate into a losing week, and this week's win rate wasn't even above 50%.

There's a second number that sharpens the picture. Winning trades were held for an average of 3.7 hours before closing. Losing trades were held for an average of 9.4 hours — roughly 2.5x longer. Winners got taken off the board quickly; losers were given more time to develop, which in practice mostly means more time to get worse. This is the same shape as the disposition effect covered in the profit-taking article on this site: a bias toward locking in small wins fast while letting losing positions run, hoping they turn around before the stop is forced.

Twenty-five trades is not a large sample, and one week is not a trend — the same caveat this site's own backtesting-biases article makes about small-sample noise applies here just as much as it would to anyone else's numbers. A single week where losers ran longer than winners could be a real behavioral pattern in the exit logic, or it could be a handful of trades that happened to take longer to hit their stop this particular week. What makes it worth flagging rather than ignoring is that hold-time asymmetry is a specific, mechanical, checkable thing — not a vibe. If it shows up again next week, and the week after, it stops being noise and starts being an exit-timing problem worth fixing directly.

What this recap is and isn't

This is a plain readout of closed virtual-engine trades for the week, not a signal, not a track record being sold, and not investment advice — the usual disclaimer on this site applies in full here too. The numbers come straight from the trade log; nothing here is smoothed, cherry-picked, or reframed after the fact. Next week's recap will report whatever next week's numbers actually are, win or lose.