What Is the Stochastic Momentum Index (SMI)? SMI vs. Stochastic Oscillator

Search Fintrev

Search our guides, reviews and glossary.

The Stochastic Momentum Index (SMI) is a momentum oscillator that measures where the current close sits relative to the midpoint of the recent high/low range, then smooths that distance twice with exponential moving averages. The result normally moves between +100 and -100. A positive reading means the close is above the midpoint of the range. A negative reading means it is below. William Blau developed the SMI as a refinement of the classic stochastic oscillator, aiming for a line less prone to false swings.

How the SMI Is Calculated

The standard stochastic oscillator asks where the close sits within the range, measured from the bottom. The SMI asks how far the close sits from the center of that range. The calculation uses a lookback length q and two smoothing lengths r and s:

1. Find the highest high (HH) and lowest low (LL) over the last q periods.
2. Compute the midpoint: M = (HH + LL) / 2.
3. Compute the distance of the close from the midpoint: D = Close − M. Also compute the range: R = HH − LL.
4. Smooth both values twice with exponential moving averages, then divide:

SMI = 100 × EMA(EMA(D, r), s) / (0.5 × EMA(EMA(R, r), s))

Dividing by half the range scales the result so that a close at the highest high produces +100 before smoothing and a close at the lowest low produces -100. Many charting packages also plot a signal line, which is an EMA of the SMI itself. It plays the same role as %D on a standard stochastic.

Worked Example (Hypothetical)

Assume a 10-bar lookback on a currency pair, with no smoothing applied yet so the arithmetic stays visible:

• Highest high: 1.1000
• Lowest low: 1.0800
• Current close: 1.0950

The midpoint is 1.0900, so the close sits 0.0050 above it. Half the range is 0.0100. The raw SMI value is 100 × 0.0050 / 0.0100 = +50.

For the same bar, the standard stochastic %K is 100 × (1.0950 − 1.0800) / (1.1000 − 1.0800) = 75.

The two readings are linked: before smoothing, SMI = 2 × %K − 100. In this example that gives 2 × 75 − 100 = 50. On a live chart the two lines look different because the SMI applies double EMA smoothing to the numerator and denominator separately, while the stochastic oscillator uses simple averaging of %K.

SMI vs. Stochastic Oscillator

George Lane popularized the stochastic oscillator in the 1950s on the premise that prices tend to close near their highs in an uptrend and near their lows in a downtrend. The SMI keeps that premise and changes the reference point and the smoothing.

The extra smoothing trades some responsiveness for a cleaner line. Whether that helps depends on the market and timeframe; it does not make the SMI a more accurate predictor by default.

How Traders Read the SMI

Overbought and oversold zones. Like the stochastic oscillator, the SMI mainly flags stretched conditions: readings above +40 are often treated as overbought, and readings below -40 as oversold.

Trend bias. The same thresholds also serve as trend filters: a value holding above +40 suggests a bullish trend, and one holding below -40 a bearish trend. A reading above +40 can therefore mean a market due to pull back or one with strong upward momentum. Price structure and the higher timeframe usually decide which applies.

Signal line crossovers. When the SMI crosses above its signal line, momentum is turning up; a cross below means it is turning down. Many traders give more weight to crossovers inside the extreme zones than to those near zero.

Zero line crosses. A move from negative to positive means the smoothed close has shifted from the lower half of the range to the upper half.

Divergence. When price makes a new high and the SMI makes a lower high, upward momentum is weakening. The bearish mirror image applies at lows.

Cross-checks. Because the SMI is built only from prices, traders often pair it with volume indicators, the RSI or MACD.

Settings and Platforms

The SMI has three inputs: the lookback length (q), the first smoothing length (r) and the second smoothing length (s), plus a signal line length. Shorter values make the line more reactive and produce more crossovers. Longer values smooth it further and delay signals.

No setting is best across markets. A configuration that suits a daily stock chart may whipsaw on a five-minute forex chart. Testing candidate settings on historical data, or in a paper trading simulator, shows how often each one produces signals and how late those signals arrive.

The SMI is available on TradingView, MetaTrader 5 and NinjaTrader as a built-in study or a custom script. Implementations differ, so check the formula and default inputs before relying on one. The separately named “SMI Ergodic” is another Blau tool derived from his True Strength Index rather than the high/low midpoint, so its readings are not interchangeable with the SMI.

Limitations

• Lag. Double EMA smoothing delays turns, so signals can arrive after much of a move has passed.
• Extremes can persist. In a strong trend the SMI can stay beyond +40 or -40 for long stretches, so fading those readings means trading against momentum.
• No volatility or volume information. The SMI normalizes the close to the range but says nothing about how wide that range is in absolute terms or how much participation is behind a move. Pairing it with the Average True Range and a volume tool covers both.

The Bottom Line

The SMI suits traders who want a smoother version of the stochastic oscillator and can accept some lag in exchange. Its signals work best when confirmed by price structure, the higher timeframe, and tools that add volume and volatility information. Settings should be tested on historical data or in a paper trading simulator, and each platform's formula and default inputs should be checked before relying on it.