Williams %R

What Is Williams %R?

Williams %R (Williams Percent Range) is a momentum oscillator created by Larry Williams that measures the level of the close relative to the high-low range over a specified lookback period. It oscillates between 0 and -100, with readings near 0 indicating the close is at the top of the range and readings near -100 indicating the close is at the bottom.

The default lookback period is 14, matching the RSI and stochastic oscillator defaults. The indicator provides a quick visual assessment of whether current price is near the top or bottom of its recent range, which correlates with buying or selling pressure.

How Traders Use Williams %R

The primary signals are overbought and oversold readings. When %R rises above -20, the market is considered overbought. When it falls below -80, the market is oversold. These thresholds correspond to price closing within the top 20% or bottom 20% of its recent range.

Failure swings provide momentum signals. A bullish failure swing occurs when %R drops below -80, rallies above -80, pulls back without reaching -80, and then breaks above the prior high. This pattern suggests building buying momentum even within oversold territory.

Divergence between price and Williams %R carries the same significance as divergence with other oscillators. When price makes a new high but %R makes a lower high, the negative divergence warns of potential price weakness.

Williams %R vs. Other Oscillators

Williams %R is essentially the inverse of the fast stochastic %K, making it functionally redundant if you already use a stochastic oscillator. Its primary advantage is simplicity: a single line with no signal line, making it easy to read at a glance.

Traders who prefer minimal chart clutter may choose Williams %R over the stochastic for quick overbought/oversold assessments. Those who want crossover signals and additional smoothing typically prefer the slow stochastic or RSI. Ultimately, %R, stochastic, and RSI all measure momentum and produce correlated signals, so using more than one on the same chart adds complexity without proportional analytical value.