Composite & famousUltimate Oscillator · UO
Larry Williams' three-timeframe momentum blend, built to cut the false divergences of single-period tools.
Works best in ranging marketsEngine-computed on a fixed sample series
What it is
The Ultimate Oscillator (UO) is a momentum indicator created by Larry Williams in 1976 to solve a specific, nagging problem with ordinary oscillators. Single-period tools like a basic Relative Strength Index throw off frequent false divergence signals, because a value tuned to react quickly is easily fooled by short-term noise, while a slow value misses turns. Williams' insight was to combine three different lookback periods, short, medium, and long, into one line so that the signal draws on multiple timeframes at once. The result plots on a bounded 0 to 100 scale, like RSI, and is designed above all to make divergence signals trustworthy. For a beginner, think of it as three momentum readings of different speeds blended into a single, steadier number.
How it is calculated
The building block is buying pressure, defined for each bar as the close minus the lower of today's low and yesterday's close. That buying pressure is measured against true range, the greater of today's high or yesterday's close minus the lesser of today's low or yesterday's close. For each of three lookbacks, typically 7, 14, and 28 periods, the calculation sums buying pressure over the window and divides by the summed true range, producing three averages that each express what fraction of the range was driven by buyers. Those three averages are then blended with a fixed 4-2-1 weighting, giving the fast 7-period reading four times the influence of the slow 28-period reading, because recent action matters most while the longer windows provide stability. The weighted blend is multiplied by 100 to land on the familiar 0 to 100 scale, so a high number means buyers dominated the recent range across all three horizons.
Reading it, step by step
At the simplest level, readings above 70 lean overbought and readings below 30 lean oversold, but Williams intended these as context rather than triggers. The real signal is a strict, three-part divergence pattern. For a buy, first price makes a lower low while the oscillator makes a higher low, and that oscillator low forms below 30; second, you draw the high of the oscillator between the two price lows; third, you act only when the oscillator then rallies above that intervening high. The sell setup is the mirror image: price makes a higher high while the oscillator makes a lower high above 70, and you act when the oscillator breaks below the low formed between the two price peaks. Because the line is already a multi-timeframe blend, these divergences are far more reliable than those from a single-period oscillator, which is the entire reason the tool exists.
Best timeframes and settings
The classic 7, 14, 28 lookbacks with 4-2-1 weighting are well tested and rarely need changing, and they work across daily swing charts and intraday charts alike. On daily charts the tool suits swing traders looking for multi-day reversals, while intraday traders can apply the same periods to hourly or 15-minute bars for shorter setups. Shortening all three windows makes the oscillator quicker and better for scalping but reintroduces the very noise the multi-timeframe design was built to suppress, partially defeating the purpose. Lengthening them makes it slower and better for position-level signals. Because the three-window smoothing is already doing heavy lifting, most practitioners leave the periods alone and instead adjust which timeframe chart they apply it to, which is the cleaner way to change its speed.
When and where to use it
The Ultimate Oscillator is at its best in ranging and mean-reverting markets, where overbought and oversold extremes actually mark turning points rather than the middle of a run. It is particularly useful on liquid instruments with clean price data, since its true-range basis handles gaps well. It excels when you specifically want divergence signals you can trust, because reducing false divergences was its founding goal. Avoid leaning on the raw overbought and oversold levels in a strong, one-directional trend, where any oscillator can pin near an extreme for a long stretch while price keeps running. It is not a trend-following tool, so in a powerful trend use it only for its confirmed divergence setups and defer to trend indicators for direction.
Strategies that use it
The flagship strategy is Williams' confirmed bullish divergence: enter long when price makes a lower low, the oscillator makes a higher low below 30, and the oscillator subsequently breaks above the high it formed between the two price lows, placing your stop below the recent price low and targeting the prior swing high. The mirror strategy shorts a confirmed bearish divergence above 70. A second, simpler approach for range markets is to fade extremes: buy as the oscillator turns up out of the sub-30 zone and sell as it turns down from above 70, but only when price is clearly range-bound and ideally near known support or resistance. A third uses the 50 midline as a trend-bias filter, favoring longs while the oscillator holds above 50 and shorts while it stays below, which keeps you aligned with the intermediate drift.
Combining it with other indicators
Because the Ultimate Oscillator already blends three timeframes, stacking it with another fast oscillator like Stochastics is redundant and can create false confidence, so it is usually better paired with tools of a different type. A trend filter such as a 50 or 200-period moving average tells you whether to favor the bullish or bearish divergence setup, keeping you on the right side of the larger move. Horizontal support and resistance levels give the oscillator's divergence a price location to react from, sharpening entries. Volume tools like On-Balance Volume can corroborate that a divergence reflects genuine shifting participation rather than a quiet drift. Chart patterns such as double bottoms often coincide with the oscillator's bullish divergence, and when structure and oscillator agree the signal is considerably stronger.
Where it fails
The most common failure is skipping the confirmation step and trading a raw overbought or oversold level, which produces a stream of early, losing entries because the full divergence rule with its break of the intervening extreme is what gives the tool its edge. In a strong trend, even this blended oscillator can remain overbought or oversold for an extended period, so counter-trend trades against a powerful move are dangerous. The strict three-part rule is more demanding than a simple cross, and impatient traders often jump the gun before the pattern completes. It can also whipsaw in choppy, directionless markets that produce many minor divergences that lead nowhere. Guard against all of this by insisting on the confirmed break, by respecting the prevailing trend, and by using support and resistance to filter which divergences are worth taking.
A worked example
Picture a stock in a pullback that prints a lower low at 30 dollars. On that low bar the close is 30.50, the low is 30, and the prior close is 31, so buying pressure is 30.50 minus the lesser of 30 and 31, which is 0.50, while true range from a high of 31.20 is 31.20 minus 30, or 1.20, giving a healthy buying-pressure fraction of about 0.42 for that bar. Even though price made a lower low, the blended oscillator makes a higher low, printing 32 versus its earlier trough of 25, and it forms this while below 30. You mark the oscillator high of, say, 41 that formed between the two price lows. A few bars later the oscillator rallies through 41, completing Williams' confirmed bullish divergence, and you enter long with a stop just under 30 and a target back at the prior swing high near 34.