strategy-dev
Scalping — fast in/out trades on 1m–5m timeframes
Tiny moves, tight stops, ruthless execution. Why most retail traders should NOT start here.
12 min · expert
What you'll have when finished
- Understand what scalping is (and isn't)
- Know which timeframes + hold times define this style
- See why fee drag dominates scalping profitability
- Identify the exchange + capital + psychology requirements
- Know when scalping makes sense and when it actively destroys capital
Before you start
- Scalping has the lowest profitability-per-unit-of-effort of any retail strategy. Most people who try this lose money even with a positive raw edge — because fees, slippage, and emotional fatigue eat it all.
- Scalping is NOT suitable for the Conditional Bot v1 (it evaluates regime every 5 minutes — too slow for sub-minute decisions).
- This guide explains mechanics, not predictions. No timeframe choice guarantees profits.
Walkthrough
-
What scalping actually is
Scalping = hundreds of trades per day, hold times of seconds to minutes, profit targets of 0.1%–0.5% per trade. You're harvesting tiny micro-trends in order-book imbalance. Win rate is the entire game — you need 60%+ wins to overcome fees at 1:1 reward/risk.
-
The right timeframe + period
Timeframe: 1m or 5m (we offer 5m as our minimum — 1m candles are too noisy for indicator-based strategies). Backtest period: 7-30 days is enough — at 5m, that's 2,000–8,640 candles, which is statistically meaningful.
-
Why fees destroy scalping math
At 40 bps per fill on Coinbase, every round-trip costs 0.8% (entry + exit). If your average target is 0.3%, you need a 3x+ winner-to-loser ratio just to break even. Most retail scalpers don't realize this until they've already lost half their capital.
Success criteria: Round-trip fee = 2× per-fill rate · Net edge per trade = avg win − fees · You need: win_rate × avg_win > loss_rate × avg_loss + fees
-
Requirements before you try this
You need: a low-fee venue (Binance/Bybit at 0–10 bps as a market maker, NOT Coinbase retail at 40 bps), $5K+ capital (so $5 fees aren't 0.1% of your account), a stable home network (latency matters), and the psychological tolerance for 50+ trades per session. Most retail traders fail one or more of these.
-
When scalping makes sense
High-volatility regimes (Bollinger Width > 4%) where the noise itself contains exploitable patterns. Range-bound markets where mean reversion works. Markets with deep order books where your size doesn't move price.
Success criteria: Volatility expansion phase · Range-bound (no sustained trend) · Deep order book (low slippage) · You have a maker rebate or zero fees
-
When scalping destroys capital
Trending markets (you exit winners too early, hold losers too long). Thin order books (slippage doubles your fee cost). Periods of low volatility (no edges exist to harvest). If you're reading this guide as a beginner, the answer is: don't try this yet.
-
Indicators that suit scalping
RSI on 5m (oversold/overbought happens hourly). Bollinger Width (volatility expansion = scalping opportunity). Volume spikes (institutional flow). Avoid EMA50 — at 5m it's a 4-hour smoothed line, way too slow for scalp decisions.
Success criteria: RSI: react to 25/75 extremes, not 30/70 · Bollinger Width: enter on compression < 0.015, exit on expansion > 0.04 · Volume: confirm with > 2x avg
What's next
Most platform users should NOT start with scalping. Build a swing strategy first (1h timeframe), backtest it, paper-trade it, run it live with small capital. Once you understand the full strategy lifecycle, THEN consider whether scalping fits your edge. The fastest path to profitability is rarely the fastest timeframe.