Strategies are usually judged on whether the idea is right. They fail more often on whether the idea is right by enough. Costs are the hurdle every trade clears before it can be called a winner, and short-term strategies live or die on them.
The spread, in money
The spread is the gap between bid and ask, and you pay it on entry: buy at the ask, and the position is immediately worth the bid.
Convert it before judging it. On EUR/USD at one standard lot, a pip is about $10, so:
- 0.8-pip spread → about $8 per round trip
- 1.5-pip spread → about $15
- 3.0-pip spread → about $30
Now scale it. One trade a day for a year at 1.5 pips is roughly $3,800 in spread on a single lot. The same year at 0.8 pips is about $2,000. Neither figure appears anywhere on your chart, and the difference between them exceeds most beginners' annual P&L.
Why the hurdle is relative, not absolute
A 1.5-pip cost is trivial or fatal depending entirely on your target.
- Targeting 100 pips: cost is 1.5% of the move. Irrelevant.
- Targeting 20 pips: cost is 7.5%. Noticeable.
- Targeting 5 pips: cost is 30%. The strategy is now mostly a fee-generation machine.
This is the honest reason most scalping systems fail. The edge may be real, but it is smaller than the toll, and the toll is charged on every single trade whether the idea worked or not.
Spreads are not constant
Quoted spreads are typically the best case — major pair, deep liquidity, quiet market. They widen predictably:
- Around news. In the seconds surrounding a major release, spreads can multiply several times over as liquidity providers step back.
- Thin hours. Late New York into the Asian open is the quietest stretch of the day.
- Rollover, around 21:00 UTC, on already-thin books.
- Weekend edges. Friday's close and Monday's open.
- Exotic pairs, always and structurally.
The trap is that these moments are exactly when a beginner most wants to trade. "Trade the news" often means entering at the precise moment the cost of entering is at its maximum.
Slippage
Slippage is the gap between the price you requested and the price you received. Your order takes the best available price when it arrives, and in a fast market that has already moved.
Two points about it:
It should be symmetrical. Genuine slippage helps you as often as it hurts. If yours is reliably negative, that is worth recording and raising — over hundreds of trades it is a meaningful cost with a cause.
It applies to stops, and this is where it bites hardest. A stop loss is not a guaranteed price; it is an instruction to exit at the best price available once the level trades. In a gap or a news spike, that price can be well beyond your stop. Your actual risk on a trade is therefore not perfectly capped, which is an argument for sizing with a little room rather than at the theoretical maximum.
Test it properly
A backtest run on mid prices with no costs is not a test of a strategy; it is a test of an idea in a frictionless universe. Before trusting any system:
- Subtract a realistic round-trip cost from every trade, not an optimistic one.
- Widen that cost for trades taken around news or in thin hours.
- Assume some negative slippage on stops.
A system that survives honest costs is worth trading. One that only works at zero cost was never a system.
What to take from this
- Convert spreads into money and annualise them — the number is larger than it looks.
- Cost matters relative to your target; small targets cannot carry normal spreads.
- Spreads widen exactly when beginners most want to trade.
- Slippage should be two-sided; one-sided slippage is a broker question.
- Stops can slip. Real risk is slightly wider than the stop distance implies.
Next: the order types that control where and how you enter and exit — and which ones protect you when the market moves faster than you do.