The compounding arithmetic
Suppose all-in cost averages a modest slice of your typical risk, say a tenth of an R per trade. At three trades a week it skims a manageable levy. At fifteen trades a week the levy is 1.5R weekly, most of a decent strategy's entire expectancy, paid before any market edge counts. Frequency multiplies the tax; the tax caps viable frequency, the quiet reason most high-frequency retail journals bleed despite decent ideas, and the arithmetic behind the scalping trade-offs.
Costs are not constant
The average hides the spikes: news minutes, rollover, Sunday opens, thin hours all multiply spread and slippage, per the mechanics guide. A strategy whose signals cluster in expensive windows pays a structurally higher rate, one more reason event-time discipline is cost management wearing a risk costume.
Cost-aware habits
The preserving behaviours: prefer planned limit entries over chased market orders; consolidate marginal signals into fewer, better trades, selectivity is a cost strategy too; measure your actual all-in cost from journal fills and charge every backtest that rate; and let target size respect costs, a ten-tick scalp target paying two ticks of cost is a business running on twenty-percent margins before the market says a word. On the XAU/USD live chart the structure is drawn for you, so the reading above takes a few seconds.
FAQ
How do I measure my real cost per trade?
From journal data: average the difference between intended and actual fills, add spread paid, divide by typical risk to express it in R terms.
At what frequency does scalping stop working?
When frequency times per-trade cost approaches expected gross edge. Each trader's numbers differ; the arithmetic is universal.
Do costs matter for swing trading?
Proportionally far less per trade, but swap on multi-day leveraged holds becomes its own line item worth checking.
ⓘ See these ideas on real price: open the free XAUUSD live chart.