The break-even arithmetic
Every article about scalping says costs matter. Almost none does the arithmetic, and the arithmetic is the whole story.
Assume a symmetrical strategy — target and stop the same distance — and a round-trip cost of spread plus slippage plus commission. The win rate you need just to break even is:
| Target | Round-trip cost | Break-even win rate |
|---|---|---|
| 20 pips | 1 pip | ~52.5% |
| 10 pips | 1 pip | ~55% |
| 5 pips | 1 pip | ~60% |
| 3 pips | 1 pip | ~66.7% |
| 5 pips | 2 pips | ~70% |
Read the last two rows together. A 5-pip target with a 2-pip round trip needs you to be right seven times out of ten before you make a single unit of profit. Very few directional signals are that good, and the ones that claim to be are usually measuring a backtest with costs left unset.
That is the real constraint on scalping, and it is why the doubling test matters so much at this horizon — see the cost sensitivity test.
Why the cost side is worse than the table suggests
The table assumes a fixed round-trip cost. In practice it varies in exactly the wrong way:
- spreads widen outside liquid hours, and widen sharply around news
- slippage is largest when the move is fastest — which is when your signal most often fires
slippagein Pine is counted in ticks, so a number you carried over from another instrument means something different here
The third is worth checking in your own export. A slippage = 3 calibrated on
an index CFD is not 3 of the same thing on EURUSD, because a tick is not the
same size — the general point is in
the cost parameters.
What the arithmetic implies for the build
It pushes you toward a larger target, which is a different strategy rather than a tuned one:
| Field | Value | Why |
|---|---|---|
| Timeframe | 5–15 minutes | fewer, larger moves per trade |
| Target | at least 10× the round-trip cost | keeps the break-even win rate near 55% |
| Stop | ATR-based, not fixed pips | scales when volatility changes |
| Session filter | London and New York overlap only | the hours where the spread is tightest |
| Costs | set from your own broker's typical spread | not a default |
The session filter is doing real work here rather than being a refinement: it removes the hours in which the cost side of the table is worst. The timezone trap in building one is covered in the London session breakout.
The measurement to run before you build anything
Take your broker's typical EURUSD spread at the hours you intend to trade, add your expected slippage, and put it against the target you have in mind. If the break-even win rate that falls out is above 60%, the idea needs a bigger target or a cheaper execution route, and no amount of indicator tuning changes that.
This is the same reasoning that makes most published scalping results unverifiable: the surveyed academic evidence on technical trading rules weakens sharply once transaction costs are included [3], and scalping is the horizon where costs are largest relative to the move.
Tactix AI — Studio vs Guide
Tactix AI is AlfaTactix’s product assistant brand (open Tactix AI).
- Tactix Studio turns a one-sentence strategy description into a draft across Timeframe, Signals, Filters, and Risk in the visual Strategy Builder. You review and edit every field before Code Generator writes MQL5 or Pine Script.
- Tactix Guide explains the step you are on — what to fill, what a control means, or how to phrase a rule — without dumping untested source code.
That is form-first automation: the LLM never replaces Code Generator, and you keep plan limits and real-time validation.