Entry and Exit Rules That Can Be Followed

The short answer

A rule tells you what to do before you know what happens. An intention tells you what you would like to have done. The difference is not tone or length — it is whether a stranger holding the same chart would act the same way you did.

A rule that cannot be checked cannot be tested, and a method assembled from untestable rules cannot be improved, only re-felt.

Rule versus intention

Two sentences, both taken from the same family of plans:

• I will enter on a strong breakout with good volume.
• I will enter long on the first 15-minute close above the prior session's high, when that close is also above the session volume-weighted average price.

Both describe buying a breakout. Only the second constrains you. The first constrains nothing, because "strong" and "good" are set by whoever is reading, and the person reading is you, at 14:32, with a position you already want.

An intention describes a disposition. A rule describes a condition. Dispositions survive contact with a moving chart by bending; conditions survive by being met or not met.

The practical test, and it is the only one you need: hand the sentence to someone competent who was not there, with the same data, and see whether they mark the same bar. If they cannot, you have written a preference. This is the same standard the anatomy of a trading plan applies to the document as a whole, applied here at the level of the individual sentence.

Three tests a rule has to pass

1. Unambiguous

Every term in the rule resolves to something you can point at, and the resolution does not contain another soft word.

"Break of the swing high" is ambiguous until swing high is defined — a candle whose high exceeds the three candles either side, say — and until "break" is defined as a close beyond, or a trade through, or a close beyond by some minimum amount.

The quickest way to find ambiguity is to search your own plan for adjectives. Strong, clean, obvious, healthy, clear, decent, proper. Each one is a place where your judgement will move with your mood, your P&L for the day, and how long you have been waiting.

2. Observable before the fact

The condition has to be evaluable at the moment you must act, using only information that exists at that moment.

This is where otherwise careful rules fail. "Enter when the pullback has ended" cannot be evaluated when you need it — the end of a pullback is identifiable only once price has already moved away from it. "Enter when volume confirms" has the same defect if the bar is still forming, because the bar's volume is not final until it closes.

This is the live-trading version of the look-ahead bias described in backtesting and forward testing. In a backtest it inflates results; in live trading it produces a rule that appears to work in review and never quite triggers in the moment.

3. Falsifiable

There must be states of the world in which the rule does not fire.

A practical check: open the last fifty bars and mark where the rule would not have triggered. If you cannot find those places, the rule is not selecting anything — it is describing the market. Rules like "enter when momentum favours the direction of the trade" tend to fail here. Almost any chart, read generously, can be made to satisfy them.

How a vague rule becomes hindsight

A vague rule does not fail loudly. It gets interpreted, and the interpretation happens after price has started to move.

The sequence is consistent. Price moves. You look at the chart. The move supplies the meaning that the rule left blank — of course that was a strong breakout, look at it. You enter, and you record in your journal that you followed your rule. You did, in the sense that you can no longer show otherwise.

Two things are now broken:

1. The record is not a record of the rule. It is a record of fifty separate interpretations, each shaped by what price was doing at the moment of interpretation. Any win rate or expectancy you compute from it belongs to nothing you can repeat.
2. Failure becomes unattributable. After a losing month you cannot separate "the method does not work" from "I applied it loosely." Both explanations fit, and confirmation bias will pick the more comfortable one for you.

Precision is not about discipline for its own sake. It is what makes a losing month informative instead of merely unpleasant.

Rewriting three vague rules

The numbers and levels below are invented for teaching. They describe no real instrument and are not a trade idea.

Vague: "Enter when the trend is strong."

Rewritten: Enter long when the 4-hour close is above the 200-period moving average, and the last two 4-hour swing lows are each higher than the one before, and the entry bar is the first 1-hour close above the most recent 1-hour swing high.

What was added: a timeframe for context and one for execution, a definition of trend as a structural sequence rather than a feeling, and a specific triggering bar.

Vague: "Take profit into resistance."

Rewritten: Exit the full position with a limit order placed 0.2× the 14-period average range below the nearest higher-timeframe level identified before entry and recorded in the journal.

What was added: the level must exist before the trade, not be discovered during it; the order type is stated; the offset is defined in units of the instrument's own movement rather than a fixed price, so it transfers across instruments. See volatility and average range for why the unit matters.

Vague: "Cut it if it isn't working."

Rewritten: Exit at market if price trades at or below the invalidation level recorded at entry, or if the position has not reached +1R within eight hours, whichever comes first.

What was added: a price condition and a time condition, both fixed at entry. "Not working" almost always means "not working yet," which is a judgement that gets more generous the longer you hold.

Exits deserve more of this than entries

Most plans specify entries in detail and leave exits as a disposition. The asymmetry is backwards: the entry determines whether you are in, but the exits determine the entire distribution of outcomes.

Each setup needs four exit conditions written before the trade exists:

• Invalidation — the price at which the idea is wrong, and why it is that price. See stop losses and invalidation.
• Target or trail logic — the rule that takes you out in your favour, stated as a condition and an order type.
• A time condition — what happens if neither level is reached.
• A session condition — what happens at the close, or at the weekend, or before a scheduled event.

If any of the four is missing, it will be decided live, by the version of you who is currently in the trade.

What to do with this

Take your plan tonight and run the three tests on every rule in it, in writing.

For each rule: circle the adjectives and replace them with definitions. Ask whether the condition could be evaluated at the moment you must act — if not, move it earlier or drop it. Then mark, on recent data, the places where the rule does not fire. If there are none, the rule is describing rather than selecting, and it needs a constraint added.

Whatever survives is a rule. Whatever does not is an intention, and it should either be rewritten or removed, because the third option — leaving it in and calling it a rule — is the one that quietly corrupts everything downstream.