Keep a Record of the Trades You Chose to Skip

You have a rule that keeps you out of trades during the first half hour of the session. It feels sensible. You have followed it for months, and your journal, which records every position you opened, says nothing at all about whether the rule has helped you, hurt you, or done nothing, because the trades it filtered out never made it onto the page.

That gap is structural. A journal of taken trades measures what happened after you said yes.

Filters work at the moment you say no. Checking whether they help means recording those moments too.

What a skipped-trade entry holds

Keep it short, or you will stop keeping it. Each entry needs:

• The date and the ticker.
• The setup, in the same words you would use for a taken trade.
• The entry, stop and target you would have used.
• The reason you passed, named as a specific rule.
• The outcome, filled in later.

The reason has to point at a rule. “Didn’t like it” cannot be tested; “earnings in four days” can. If you write each trade idea as a one-sentence trade thesis, the setup line is already done, and the skip reason is usually the item from your pre-trade checklist that the trade failed.

Write the skip before you know the ending

This is the step that decides whether the log is worth anything. Record the entry, stop and target at the time you pass, with the chart in front of you, before the stock has moved.

Hindsight is quick and quiet. A trader who fills in skipped trades at the end of the week, after seeing which ones ran, will tend to remember the stop as a little wider on the winners and a little tighter on the losers, and to recall passing on setups that worked while forgetting the dull ones that went nowhere, so the log fills up with evidence that every filter is costing money.

The outcome needs a mechanical rule too. One simple version: whichever the stock reached first, stop or target, decides the result, measured in multiples of the planned risk (R). No judgment calls about where you “would have” gotten out.

A hypothetical month

Here is an illustrative month for a trader with three skip rules. The numbers are invented for the example and describe no real stock or result.

Week	Setup	Skip reason	Outcome if taken
1	Breakout over range high	First 30 minutes	-1R
1	Pullback to rising average	Earnings within a week	+2R
1	Breakout over range high	First 30 minutes	-1R
2	Gap-and-hold	First 30 minutes	+2R
2	Pullback to rising average	Stop too wide for size	+2R
2	Breakout over range high	Earnings within a week	-1R
3	Pullback to rising average	Stop too wide for size	+2R
3	Gap-and-hold	First 30 minutes	-1R
4	Breakout over range high	Earnings within a week	-1R
4	Pullback to rising average	Stop too wide for size	-1R

Reading the month carefully

Ten entries tell you very little.

A handful of outcomes can swing on one or two trades, and a single month catches whatever the market happened to be doing in those four weeks, so the honest reading of this table is a set of questions to keep watching and no verdict on any rule yet. How many entries are enough is not a question with a fixed answer. More is better, and a rule should survive several different kinds of market before you trust what the log says about it.

Some of the patterns are still suggestive. The “stop too wide” rule is the interesting one. Those setups may have been fine; what was wrong was the size. A trader seeing this over a longer stretch might keep taking them at a smaller share count sized to the wider stop.

The half-hour rule looks mildly useful. The earnings rule looks neutral, but it exists to avoid a gap you cannot manage, and a month without a big earnings gap cannot test that.

One more comparison helps. Set each rule’s skipped total beside the taken trades from the same weeks, measured in R the same way, since a filter that removes trades doing about as well as the ones you kept is removing opportunity along with risk, while one that removes trades doing clearly worse is doing its job.

What the log cannot see

Skipped trades get perfect fills. You never face the slippage, the partial fill or the hesitation at the stop, so the hypothetical results will usually look a little better than the same trades would have gone live.

Some filters guard against rare events. A rule that keeps you out of earnings can look pointless for months and then do its whole job in one session.

And the log says nothing about your state of mind. Passing on a trade because you were tired is a real reason. Write it down anyway; a pattern of skips on bad days is its own finding.

Where the log earns less

Not every trader needs one. If your rules are fully mechanical and you can test them against historical price data, a backtest answers the same question faster and over far more examples than a hand-kept log ever will, although it still misses the setups you would have ignored in real time for reasons you never wrote into the rules.

The log also covers only what you saw. A setup you never noticed is a different problem, one about your watchlist and screening, and adding it later from a scan turns the record back into hindsight. Log the trades you looked at and passed on, and nothing else.

Keeping it going

Put the skipped log in the same file or notebook as your taken trades, with a column marking which is which. Review it on a fixed schedule, monthly perhaps, when the market is closed. The taken-trade records are also the ones that matter for what records a trader should keep for taxes; the skipped entries are research, kept for no one but you.