Pro Access

Williams %R Reversion is part of the expanded Pro strategy library. The guide stays public, but running it in backtest, compare, paper tracking, and source-code view is a Pro capability. Pro currently unlocks 8 additional pre-built strategies.

The Intuition

Williams %R was developed by Larry Williams and published in his 1973 book How I Made One Million Dollars Last Year Trading Commodities. Williams was a commodity trader who needed a fast, readable indicator of whether a market was near the top or bottom of its recent range — not a trend indicator, but a position indicator within the current trading environment.

The indicator is expressed on a scale from 0 to -100, which is the inverse of how most oscillators present. A reading near 0 means the close is at or near the highest price seen in the last N sessions — the market is overbought relative to its own recent range. A reading near -100 means the close is at or near the lowest price in the window — the market is oversold. This inverted presentation can be unintuitive at first, but the logic is deliberate: Williams oriented the scale so that extreme readings feel directionally aligned with the condition (very negative = very depressed).

Mathematically, Williams %R is the inverse of the Stochastic %K before smoothing. The formula divides the distance from today's close to the recent high by the total recent range, then multiplies by -100. This means the indicator reacts instantly to new price extremes: if today's close is the highest close in the window, %R immediately reads 0; if it is the lowest, it reads -100. Unlike RSI, which accumulates path-dependent gains and losses, %R has no memory beyond the lookback window's high and low.

The practical interpretation for systematic traders: a Williams %R reading below the oversold threshold (commonly -80 or -85) suggests the close has sold off hard relative to its recent range. A reading above the overbought threshold (commonly -20 or -15) suggests the close has rallied aggressively into the top of that range. The assumption is mean reversion — these extremes will be temporary. But that assumption holds only in non-trending regimes. In a persistent downtrend, %R can spend weeks below -80 while price continues to fall.

The distinction that separates effective use from overuse: Williams %R identifies position within a range, not the end of a trend. Traders who use it most effectively treat it as a filter for entry timing within a broader view — not as a standalone signal for calling reversals. When the broader regime is neutral and the indicator reaches -90, that is a different situation from the same reading in a market that is making consecutive new lows. The number is the same; the context is entirely different.

The Math

Read this as a compact model summary: what the signal sees, what it ignores, and where fragility can creep in.

HighestHigh(t) = max(High[t-n : t])
LowestLow(t)   = min(Low[t-n : t])
%R(t)          = -100 × (HighestHigh(t) - Close(t)) / (HighestHigh(t) - LowestLow(t))

Signal(t) = +1  if %R(t) < oversold threshold   [for example, -80]
          = -1  if %R(t) > overbought threshold [for example, -20]
          =  0  otherwise

Parameters

ParameterTypeDefaultDescription
window int 14 Range lookback window
oversold float -80.0 Williams %R threshold for long entries
overbought float -20.0 Williams %R threshold for short entries

Source Code

Source access for this built-in strategy is included with Pro.

Further Reading

  • Williams, L. (1973). How I Made One Million Dollars Last Year Trading Commodities.
  • Murphy, J. (1999). Technical Analysis of the Financial Markets. NYIF.
  • Pring, M. (2002). Technical Analysis Explained. McGraw-Hill.

When It Works / When It Fails

Works
  • Range-bound, non-trending markets with oscillation
  • Timing filter within a broader directional view
  • Short-term snapbacks in liquid instruments
  • Confirmed by a separate trend-neutral regime filter
Fails
  • Strong directional trends — indicator stays pinned
  • Gap-prone single-stock equities with event risk
  • Without trend filter in sustained bull or bear runs
  • Small-cap, low-liquidity names with wide spreads

Regime Fit

Transition / Calm vol
Best environment. Range oscillation is clean and the indicator reads full swings reliably.
Transition / Normal vol
Workable. Some noise but mean-reversion signals remain valid.
Trend / Calm vol
Borderline. Slow-moving trends may still produce false reversal signals.
Trend / Stressed vol
Worst case. Indicator stays pinned at extremes; signals are unreliable.

Compared to Alternatives

vs Stochastic
Algebraically equivalent to %K with an inverted sign. The practical difference is whether %D smoothing is applied on top.
vs Bollinger
Bollinger adapts band width to volatility; Williams %R uses a fixed lookback range. Bollinger is more stable when volatility regime shifts.
vs RSI
RSI is path-dependent via cumulative gains/losses; %R uses only the range position. RSI is smoother; %R responds instantly to new price extremes.
Unlock This Strategy →