Entries and exits

Level 4. Assumes market types.

The idea

A rule is two decisions, not one. Beginners spend nearly all their effort on the first and nearly none on the second, and it is the second that decides the result.

The entry chooses when you are exposed. The exit chooses how much of the move you keep, and how much a mistake costs. You can pair a mediocre entry with a good exit and make money. The reverse almost never works.

Four ways out, and a rule usually has more than one:

Exit	Fires when	Gives you
Signal	the condition that got you in reverses	a full ride, with the end given back
Stop	price goes against you by a set amount	a bounded loss, and being taken out of moves that recover
Target	price reaches a set profit	a high win rate, and a cap on your best trades
Time	a period elapses	an answer on positions that go nowhere

Why it matters

The arithmetic. What a strategy earns per trade is its expectancy:

expectancy = (win rate × average win) − (loss rate × average loss)

Two rules, both real shapes you will meet:

	Win rate	Average win	Average loss	Expectancy
A — target exit	70%	1.0	2.0	(0.7 × 1.0) − (0.3 × 2.0) = +0.10
B — trend exit	35%	4.0	1.0	(0.35 × 4.0) − (0.65 × 1.0) = +0.75

Rule B is wrong twice as often as it is right and earns seven times as much per trade. This is not a trick of the numbers — it is the normal shape of trend following, and it is why win rate on its own tells you nothing. Anyone advertising a 90% win rate is telling you about one column of a four-column table.

The exit is what sets those two columns. A target exit raises the win rate and truncates the average win. A trailing exit lowers the win rate and lets the average win run. You are choosing a distribution, not a hit rate.

What it looks like

An asymmetry worth internalising: your stop is a decision you make once, and your exit is a decision you make repeatedly. The stop is the only part of a trade whose worst case you control in advance. Everything else is the market's choice.

Which is why the shape below is the one Arvo refuses outright — quick to buy, slow to sell:

flowchart LR
    A["Enter on 40 bars<br/>of strength"] --> B["Exit on 60 bars<br/>of weakness"]
    B --> C["Gives most of the<br/>trend back before it<br/>admits the trend ended"]

    style C fill:#b62324,stroke:#b62324,color:#fff

It feels patient. It is not patient — it is asymmetric in the wrong direction, and it is the most common beginner shape because holding a loser feels like conviction.

What people get wrong

Trading without a stop, then inventing one. A position with no predetermined exit has an unbounded loss and no position size can be computed for it. "I'll watch it" is not a stop.

Widening a stop that is about to be hit. This converts a known small loss into an unknown large one, and it is the single most destructive habit in trading. The stop was correct when you were calm.

Optimising the entry for months. The entry is the parameter people tune because it is the visible part. It is also where tuning does the most damage: see level 7.

Treating a target as free. Capping your winners while your losers are uncapped inverts the table above. Rule A survives only because its win rate is genuinely high; most beginner targets pair a truncated win with an unbounded loss.

See it in Arvo

• An exit never asks the risk gate. Every check in the gate asks whether to take risk; none may stop you shedding it. A halted account cannot open, but it can always close — an exit path that can be refused is not an exit path.
• A stop is set as stop_atr_multiple in risk.json: stop distance as a multiple of average true range, so "a stop" means the same thing on a $20 stock and a $600 one. Not set runs with no stop, and the Risk tab says so rather than showing a default as if you chose it.
• risk_per_trade needs a stop — without one there is no loss to size against, and a proposal with no stop is refused as NoStop. The arithmetic is in level 5.
• momentum_breakout refuses exit_period above entry_period, with the reason quoted above. The engine will not run the shape in the diagram.
• Every trade's exit reason is on the record, so "why did it sell?" is answerable per round trip, not inferred from a curve.

Watch

• (to be chosen) — expectancy worked through with real numbers.
• (to be chosen) — why a high win rate can lose money.

Next

Risk and sizing — the part that decides whether you survive, which comes before the part that decides whether you profit.