Most oscillators have the same weakness: they watch one lookback period. Tune them fast and they whipsaw; tune them slow and they miss the turn. Larry Williams designed the Ultimate Oscillator in 1976 specifically to attack that problem — instead of one timeframe, it blends three, and it weights them so the short-term reading leads without dominating.
What the Ultimate Oscillator measures
The indicator is built on the idea of buying pressure: how much of the bar's true range did the bulls actually win?
BP and TR are then summed over three windows — classically 7, 14 and 28 periods — and each window produces an average (sum of BP divided by sum of TR). The final value combines them with 4:2:1 weights:
UO = 100 × (4 × Avg7 + 2 × Avg14 + 1 × Avg28) / 7
The result oscillates between 0 and 100. Because the 7-period average carries the most weight but the 28-period average anchors the reading, the line reacts to fresh momentum without flipping on every candle — that is the whole point of the design.
How to read it
The classic Williams entry: the three-step divergence
Williams did not trade the levels on their own. His buy setup has three explicit conditions:
What the Ultimate Oscillator measures
The indicator is built on the idea of buying pressure: how much of the bar's true range did the bulls actually win?
- Buying Pressure (BP) = Close − the lower of (current Low, prior Close).
- True Range (TR) = the higher of (current High, prior Close) − the lower of (current Low, prior Close).
BP and TR are then summed over three windows — classically 7, 14 and 28 periods — and each window produces an average (sum of BP divided by sum of TR). The final value combines them with 4:2:1 weights:
UO = 100 × (4 × Avg7 + 2 × Avg14 + 1 × Avg28) / 7
The result oscillates between 0 and 100. Because the 7-period average carries the most weight but the 28-period average anchors the reading, the line reacts to fresh momentum without flipping on every candle — that is the whole point of the design.
How to read it
- Readings above 70 mark strong buying pressure (overbought territory); readings below 30 mark strong selling pressure (oversold).
- The midline around 50 separates bullish from bearish pressure regimes.
- As with any oscillator, "overbought" is not a sell signal by itself — a strong trend can pin the line above 70 for a long time.
The classic Williams entry: the three-step divergence
Williams did not trade the levels on their own. His buy setup has three explicit conditions:
- Price makes a lower low, but the Ultimate Oscillator makes a higher low — a bullish divergence.
- The first oscillator low in that divergence happened below 30, so the market was genuinely washed out.
- The oscillator then breaks above the high it made between the two lows. That break is the trigger — not the divergence itself.
The sell setup is the mirror image: a bearish divergence whose first peak formed above 70, triggered when the oscillator drops below the swing low between the two peaks. The trigger condition matters — most divergence traders lose money by entering as soon as they spot the divergence, while the market keeps trending against them.
Strengths and limitations
The multi-timeframe construction filters out a lot of the false divergences that plague a single-period RSI or Stochastic, and the buying-pressure math ties the reading to where price closed inside the bar rather than to close-to-close changes alone. The trade-off is signal frequency: the full three-condition setup appears rarely, and in strong one-way trends the oscillator can stay stretched while price never gives the divergence you are waiting for. It also inherits the usual oscillator caveat — it says nothing about how far a move can run, only about the pressure behind it.
Practical tips
- Use it as a confirmation tool, not a standalone system: pair the divergence trigger with structure (support/resistance, a break of a swing level) or a trend filter on a higher timeframe.
- The 7/14/28 defaults suit swing trading on H4-D1. Intraday traders sometimes compress them (for example 4/8/16), accepting more noise in exchange for speed.
- Because the signal is rare, backtest it honestly before sizing up — scarcity makes it tempting to overtrust each occurrence.
Do you use the Ultimate Oscillator or prefer a single-period oscillator like the RSI? Share your settings and experiences below. - Use it as a confirmation tool, not a standalone system: pair the divergence trigger with structure (support/resistance, a break of a swing level) or a trend filter on a higher timeframe.