YouTube thumbnails: "Scalp GBPUSD M1 for $500/day." Reddit threads: "Just trade the 1-minute chart, it's simple." The promise of lower timeframes is speed — more trades, faster profits, compound growth. The reality is something no course will tell you.
I backtested the same pullback-to-EMA strategy on GBPUSD across 4 timeframes: M3, M5, M15, and H1. 50 trades each. Same rules. Same pair. Only the timeframe changed.
| Metric | M3 | M5 | M15 | H1 |
|---|---|---|---|---|
| Win Rate | 31% | 48% | 54% | 58% |
| Profit Factor | 0.57 | 1.21 | 1.72 | 1.96 |
| Avg RR Realized | 0.8:1 | 1.3:1 | 1.7:1 | 1.9:1 |
| Net Pips | -1,840 | +620 | +1,890 | +3,120 |
| Spread Cost (as % of avg win) | 28% | 14% | 7% | 5% |
Look at that last row: spread cost. On M3, the average winning trade made 12 pips. The spread on GBPUSD is ~0.8 pips roundtrip. That means 7% of your gross profit goes to spread on every winning trade — and on losing trades, spread still costs you. Add the 4-5 pip SL typical for M3 scalping, and a single 2-pip slippage on stop-out eats 40% of the risk budget.
On M3, the average candle body is 4-6 pips. The spread is 0.8 pips. That means 15-20% of each candle's movement is pure spread noise. On H1, the average candle body is 20-30 pips. Spread is 0.8 pips — 2-4% of movement. The signal-to-noise ratio improves by a factor of 5 when you move from M3 to H1.
M5 showed a 1.21 profit factor — technically profitable but fragile. One bad week wipes months of gains. The reason isn't the strategy. It's the wick-to-body ratio on lower timeframes. M3 candles are 70% wick — random noise, stop runs, algorithmic micro-scalping by institutions. You're not trading structure. You're trading noise and calling it structure because your brain pattern-matches even in randomness.
On H1, the same pullback strategy works because the candle bodies are large enough to absorb spread costs, and the wicks represent actual levels where price tested and rejected — not random noise. The edge is real above M15. Below M15, you're not trading your strategy. You're paying spread to a broker while algorithms scalp the noise faster than your finger can click.