KAMA: Kaufman Adaptive Moving Average Guide

Key Takeaways

• KAMA Kaufman adaptive moving average uses an efficiency ratio (ER) that compares net move to total path to gauge whether price is trending.
• ER is bounded between 0 and 1; a high ER speeds KAMA up by blending it with a fast EMA, low ER slows it toward a slow EMA.
• The most common mistake is treating KAMA as a faster moving average rather than as an adaptive one.
• Default parameters are KAMA(10, 2, 30): 10-bar ER window, fast EMA period 2, slow EMA period 30.

What It Is

KAMA is an exponential moving average whose smoothing constant is not fixed. Instead, the smoothing constant changes each bar based on a measure of how "trendy" recent price action has been. When the market is moving in a straight line, KAMA reacts quickly. When the market is going sideways, KAMA almost stops moving.

Kaufman's stated goal was to avoid two failure modes that fixed-length averages share: lagging too far behind in real trends and getting whipsawed by noise in flat markets. The adaptive constant lets one indicator behave well in both regimes.

The Intuition

If you walk 10 miles north in a straight line, your net distance equals your total distance. If you walk 10 miles in random directions, your net distance is small even though your total path is the same. That ratio of net to total is exactly Kaufman's efficiency ratio.

In trending markets, ER is near 1 and KAMA acts like a fast EMA. In choppy markets, ER is near 0 and KAMA acts like a slow EMA, almost a flat line. The line itself is the same kind of recursive EMA you have already seen; only the multiplier adapts.

How It Works

KAMA has three parameters: ER period (default 10), fast EMA period (default 2), and slow EMA period (default 30). The formulas, documented by StockCharts ChartSchool and Corporate Finance Institute, are:

ER = |Price_today - Price_{N bars ago}|
   / sum( |Price_i - Price_{i-1}| for last N bars )

fast SC = 2 / (2 + 1)   = 0.6667
slow SC = 2 / (30 + 1)  = 0.0645

SC = ( ER * (fast SC - slow SC) + slow SC )^2

KAMA_today = KAMA_yesterday + SC * (Price_today - KAMA_yesterday)

The smoothing constant SC is squared so the transition between fast and slow is non-linear. When ER is near 0, SC is roughly slow SC squared, about 0.0042, which gives a very slow EMA. When ER is near 1, SC is roughly (0.6667)^2 = 0.4444, which gives a fast EMA.

A useful intuition: when ER is high, KAMA closes most of the gap to current price every bar; when ER is low, it barely moves. The line tightens to price during trends and flattens during chop, by construction.

Worked Example

Suppose over the last 10 bars price moved from 100 to 110 in a near-straight line, with small wiggles. Net move = 10. Total absolute moves = roughly 11. Then:

ER = 10 / 11 = 0.909

Compute SC:

SC = (0.909 * (0.6667 - 0.0645) + 0.0645)^2
   = (0.909 * 0.6022 + 0.0645)^2
   = (0.5474 + 0.0645)^2
   = (0.6119)^2
   = 0.374

If KAMA yesterday was 108 and today's price is 110:

KAMA = 108 + 0.374 * (110 - 108)
     = 108 + 0.748
     = 108.748

Now imagine the same 10-bar window has net move of just 1 (price oscillated between 105 and 110 several times) but total absolute moves of 20. Then ER = 1/20 = 0.05, SC is essentially slow-only and squared (about 0.005), and KAMA barely moves regardless of where today's price prints.

Common Mistakes

1. Reading KAMA as a faster moving average. KAMA is not faster than an EMA in trends and slower in chop, it is both depending on regime. Backtests that compare KAMA to a fixed EMA without separating regimes hide the indicator's actual behavior.

2. Tuning ER period like a normal lookback. The ER window is a regime detector, not a smoothing length. Doubling it does not double smoothing; it changes which window is used to define "trending."

3. Ignoring the squared SC. The square in (ER * (fast - slow) + slow)^2 is what makes the transition non-linear. Some implementations drop it. Without the square, the line behaves more like a linearly interpolated EMA and loses much of the adaptive benefit.

4. Trading KAMA crosses in low-ER regimes. When ER is near 0, the line is almost flat and price oscillates around it. Crosses in that regime are noise. Use KAMA as a trend filter, not a signal source, when ER is low.

5. Using KAMA in isolation. Adaptive moving averages are powerful when combined with a separate regime detector (volatility, ADX, breadth). Used alone, they look great in trends and unimpressive in everything else.

Frequently Asked Questions

What is KAMA Kaufman adaptive moving average in simple terms? The KAMA Kaufman adaptive moving average is an EMA whose smoothing speed changes each bar. It speeds up during trends and slows down during chop.

How does KAMA affect investment decisions? Systematic traders use KAMA as a trend filter that adapts without manual retuning. Its flat behavior in choppy regimes also makes it useful for measuring whether a market is "in a trend" worth trading.

What is a real-world example of KAMA use? KAMA is a built-in study in StockCharts, TradingView, MetaTrader, and most professional charting platforms. Default settings (10, 2, 30) are widely used on daily charts of liquid equities and futures.

How can investors use KAMA effectively? Combine KAMA with a regime gauge (ADX, ATR, or just ER itself), avoid trading crosses in low-ER environments, and respect that the line is intentionally slow when the market is flat.

How is KAMA different from VIDYA? Both adapt their effective length to volatility. KAMA uses Kaufman's efficiency ratio of net to total move. VIDYA, by Tushar Chande, uses the Chande Momentum Oscillator as its volatility index instead.

Sources

1. StockCharts ChartSchool. "Kaufman's Adaptive Moving Average (KAMA)." https://chartschool.stockcharts.com/table-of-contents/technical-indicators-and-overlays/technical-overlays/kaufmans-adaptive-moving-average-kama
2. Corporate Finance Institute. "Kaufman's Adaptive Moving Average (KAMA)." https://corporatefinanceinstitute.com/resources/career-map/sell-side/capital-markets/kaufmans-adaptive-moving-average-kama/
3. TradingView Help. "Kaufman's Adaptive Moving Average (KAMA)." https://www.tradingview.com/support/solutions/43000773012-kaufman-s-adaptive-moving-average-kama/
4. Stock Indicators for .NET. "Kaufman's Adaptive Moving Average." https://dotnet.stockindicators.dev/indicators/Kama/

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.