Why trailing stops aren't free
A break-even stop feels like it can only help — "once I'm in profit, protect it." Then we tested it on your own roster and it lost money. Here's the mechanism, in your real numbers, and the rule it taught us.
1The intuition that feels airtight
It goes like this: once a trade is up by , move the stop to your entry. Now the trade "can't lose" — worst case you scratch at break-even, best case it runs to target. Free downside protection. Who wouldn't take that?
The flaw hides in three words: it doesn't capture the profit. Moving your stop to entry doesn't bank the +1R — you never took that money off the table. All it does is remove your downside. The trade still has to travel all the way to its real target (+2R, +3R) to actually win. And in exchange for that downside protection, you've added a brand-new way to get thrown out: any dip back to your entry now closes the trade.
That matters because price doesn't travel in straight lines. A winner routinely pushes to +1R, pulls back through your entry on an ordinary wobble, then continues to the target. With a fixed stop that's a full win. With a break-even stop, you got tapped out at zero on the wobble and watched it run without you.
2The same trade, two exits
Neither exit is wrong in the abstract — it depends entirely on how often, for a given strategy, price wobbles back to entry and then still reaches target. You cannot reason that out from an armchair. You have to count. So we did.
3The proof, in your roster's numbers
We added a break-even/trailing option to the backtest engine (opt-in, so no existing result moved) and ran every roster strategy across its full history with real spread and swap — fixed bracket versus break-even. The clearest verdict came from Sentinel, the strategy with the deepest evidence on the whole roster. The number to watch is how many trades reached the full +3R target:
fixed bracket → 104 full-target hits · net +$1,881
break-even @+0.5R → 88 full-target hits · net +$1,560
The break-even stop destroyed 16 of the big winners — trades that would have run to +3R got tapped out on a pullback instead. Sentinel's entire edge is those big runners (it wins only ~35% of the time; the size of the winners carries it), so losing sixteen of them dwarfs every small loss the stop saved. Return dropped. On the roster's most trustworthy strategy, "protection" was a wealth transfer away from you.
4So when does it help?
Same experiment, opposite answer for RSI — the gentle break-even actually improved it:
| strategy · pair | fixed PF / ret | + break-even @0.5R |
|---|---|---|
| RSI · USDJPY | 1.12 / +13.6% | 1.21 / +18.8% |
| RSI · GBPUSD | 1.13 / +19.8% | 1.17 / +22.5% |
| Sentinel · USDJPY | 1.14 / +17.0% | 1.13 / +13.9% |
Same mechanism, different balance. RSI's break-even also kills a few winners — but RSI is with a nearer 1:2 target and a lot of full-loss trades, so the losses it saves outweigh the winners it gives up. Net positive. It's not that break-even is good or bad — it's a trade, and which side wins is a property of the strategy, not of the stop.
(Even this RSI edge is in-sample — measured on the same history it was checked against. Before it ever touches a live order it has to survive the same sealed out-of-sample exam the roster was built on. Looking good on history is exactly how curve-fit exits fool you.)
5The rule: match the exit to the thesis
Step back and the whole thing resolves. A trailing stop isn't a universal upgrade you bolt onto anything — it's the native exit for one kind of strategy, and a mismatch for the other.
Trailing is really a winner-maximizer with a loss cap, not a loss reducer. The clean ways to shrink losses without needing the market to cooperate are the boring ones: risk less per trade (drop 1% → 0.5% and every loss halves, edge untouched), and the guardrails that already bench a strategy when it turns bad. Tightening the exit is never free — it always trades some winners away, and whether that's worth it has to be measured, per strategy, every time.
6The takeaway
Three things worth carrying out of this one:
1. A stop that "can't lose" isn't free. It buys a smaller downside by adding a way to exit winners early. Sometimes that trade pays; often it doesn't.
2. An exit rule has no fixed value — it's only good or bad relative to a strategy's thesis. Match trailing to trend riders, fixed targets to measured moves.
3. Test it, don't assume it. The same break-even rule helped RSI and hurt Sentinel. The only way to know which you've got is to count on your own data — which is the habit this whole system is built to make easy.