Vấn đề tập trung (clustering)
This explainer uses an illustrative dataset of 1.2M anonymised stop-loss placements on EURUSD and XAUUSD over a single quarter. The result is a map of crowd behavior: stops pile up exactly where every textbook says to put them — just beyond round numbers and the most recent swing high/low.
Retail đặt stop ở đâu
% trong số 1,2 triệu lệnh stop
Clustered stops are visible liquidity. Price is repeatedly drawn to these pools — the “stop hunt” traders complain about is mostly the market mechanically filling where the resting orders are.
Clustering tốn gì
Comparing outcome distributions, structure-hugging stops were tagged-then-reversed (stopped out within 10 pips of the eventual turn) 2.3× more often than ATR-scaled stops on the same entries. The fix is mechanical:
- Scale stops by current volatility (e.g. 1.5 × ATR-14), not by chart structure alone.
- If the logical level is 1.1000, the wrong stop is 1.0995 — and so is everyone else’s.
- Wider, volatility-aware stops with smaller size carry identical dollar risk and dramatically better survival.
Chỉ nội dung giáo dục. Số liệu, benchmark và biểu đồ là ví dụ minh họa, không phải dữ liệu sản xuất IUX hay mô tả kết quả giao dịch. Không phải tư vấn đầu tư.