Williams %R Indicator: Complete Trading Guide

Williams %R is calculated using a single formula: W%R = (Highest High − Close) / (Highest High − Lowest Low) × −100. The default period is 14, meaning the indicator scans the last 14 bars for the highest high and lowest low. The result always falls between -100 and 0.

The negative scale is the first thing that trips up new users. A reading of -5 means the close is near the top of the 14-period range — a high-energy, potentially overbought condition. A reading of -95 means the close is near the bottom — oversold territory. This is the inverse of how most oscillators present data.

Functionally, W%R behaves as an inverse stochastic. The standard stochastic %K formula is (Close − Lowest Low) / (Highest High − Lowest Low) × 100. Multiply that result by -1 and shift the scale, and you have Williams %R. The two indicators are mathematically equivalent; the difference is presentation.

The default overbought threshold is -20. The default oversold threshold is -80. These levels define the outer 20% of the price range on each side. A 14-period setting on H1 data covers approximately 14 hours of price action — enough to capture intraday momentum cycles without excessive noise.

Three signal types drive most W%R strategies: threshold crossovers, failure swings, and divergence.

Threshold crossovers are the baseline. A buy signal occurs when W%R crosses back above -80 from below — the close has moved from the bottom of the range back toward the middle. A sell signal occurs when W%R crosses back below -20 from above. The key word is 'crosses back.' Entering while W%R sits inside the oversold zone without confirmation adds unnecessary exposure; the indicator can remain below -80 for extended periods during trending markets.

Failure swings offer higher-probability setups. In a bullish failure swing, W%R drops below -80, rebounds above it, pulls back without reaching -80 again, then breaks above the prior rebound high. This pattern signals that selling pressure is exhausting. The inverse applies for bearish failure swings.

Divergence is the highest-conviction setup. When price prints a lower low but W%R prints a higher low, buying momentum is building beneath the surface. In EUR/USD during Q3 2022, multiple bullish divergences on the H4 chart preceded 150–200 pip recoveries within 3–5 sessions. Divergence does not predict timing precisely — it identifies structural shifts in momentum before price confirms them.

Avoid using W%R signals in isolation during strong trends. During a sustained uptrend, readings can hover near -20 for dozens of bars. In those conditions, oversold readings near -80 function better as pullback entry points than reversal signals.

The default 14-period setting performs differently across timeframes, and the adjustment is not purely mechanical.

On M15, a 14-period W%R covers 3.5 hours of price data. This produces frequent signals — useful for scalping setups, but noise increases substantially. Reducing the period to 10 sharpens responsiveness. Overbought/oversold thresholds can be tightened to -15 and -85 to filter marginal signals. Average signal frequency on M15 with default settings runs approximately 8–12 actionable crossovers per trading day on a major pair.

H1 is the most balanced timeframe for W%R. The 14-period default covers 14 hours — roughly one full trading session plus overlap. Signal frequency drops to 3–6 per day, and false positives decrease measurably. This is the timeframe where divergence setups produce the most consistent follow-through, historically.

H4 favors a longer period — 20 to 21 bars covers approximately 80–84 hours, or one full trading week. This setting reduces noise significantly and aligns W%R signals with weekly momentum cycles. Threshold adjustments to -25 and -75 slightly expand the neutral zone and reduce whipsaws during range-bound weeks.

Period settings above 21 on any timeframe tend to smooth W%R to the point where it lags price action by more than it leads — negating the indicator's primary advantage.