Most moving averages force you into one tradeoff: smooth the noise and accept lag, or react faster and accept more whipsaws. The Adaptive Moving Average AMA tries to cheat that tradeoff by changing its sensitivity based on what price is doing right now. When price moves cleanly in one direction, AMA speeds up. When price chops around, AMA slows down.
Adaptive Moving Average AMA in plain English
AMA is a moving average that adapts its “speed” to market conditions. It does not predict reversals. It changes how quickly it follows price depending on whether price movement is efficient (directional) or inefficient (noisy).
If you already understand how a classic moving average behaves, AMA is easiest to grasp as a sliding dial between two behaviors. In clean trends it can act more like a faster average that hugs price, and in choppy ranges it can behave more like a slower average that stays flatter and ignores wiggles. That makes it a practical trend filter for traders who want fewer false flips during sideways periods without making the trend line permanently slow.
To anchor your expectations, compare it mentally to a fixed lookback average like a Simple Moving Average SMA guide and a more reactive average like an Exponential Moving Average EMA overview. AMA is still a single line on the chart, but the smoothing is not constant from bar to bar.
The simplest way to understand the AMA formula
Most AMA implementations used by traders are based on the Kaufman style approach, where the key idea is an Efficiency Ratio that measures how straight the move has been over a lookback window.
A simple version of the logic looks like this:
Efficiency Ratio ER
ER = abs Price today – Price N bars ago / sum of abs daily changes over N bars
Interpretation:
If price traveled from A to B in a fairly straight line, the numerator is large relative to the wiggle sum, so ER is high
If price chopped up and down to get to roughly the same place, the wiggle sum is large, so ER is low
Then ER is used to build an adaptive smoothing constant SC that stays between a fast and slow boundary
FastSC = 2 divided by FastPeriod plus 1
SlowSC = 2 divided by SlowPeriod plus 1
SC = ER times FastSC minus SlowSC plus SlowSC then squared
Finally AMA updates like an EMA style recursion
AMA today = AMA yesterday + SC times Price today minus AMA yesterday
You do not need to compute this by hand to use it well. The key takeaway is what the math forces: when ER is high, SC rises and AMA follows price faster. When ER is low, SC falls and AMA becomes slow and stable.
AMA settings traders use
AMA has three knobs that matter in practice: the ER lookback period, the fast boundary, and the slow boundary. Many platforms bundle these into defaults that often look like ER 10, Fast 2, Slow 30. That default is popular because it creates a meaningful spread between fast and slow behavior without requiring constant tweaking.
Instead of searching for one best setting, match parameters to your decision horizon and the instrument’s volatility. Shorter ER lookbacks adapt faster but may react to short bursts of noise. Longer ER lookbacks adapt slower but can stay calmer through messy consolidations. A practical way to think about it is: ER controls when the line changes personality, while fast and slow boundaries control how extreme the personalities are.
Common, reasonable starting points on daily charts include ER around 10 to 20, fast boundary around 2 to 5, and slow boundary around 20 to 50. If you study past winners, you will often find that the goal is not perfect turning points. The goal is a line that keeps a clear slope during the sustained trend phase and only flattens when trend quality truly breaks.
How AMA behaves on charts
On a chart, AMA tends to do three noticeable things that fixed moving averages cannot do as consistently.
First, it changes curvature. When a stock transitions from a tight base into a clean expansion move, AMA can bend upward sooner because ER increases as price starts moving more efficiently. That can help you stay aligned with trend direction earlier, without permanently using a very short moving average.
Second, it can resist range noise. In sideways action where price crosses a fixed average repeatedly, ER tends to compress because the path is inefficient. AMA slows down, which often reduces the number of meaningless crossovers.
Third, it highlights trend quality through slope. A rising AMA with clean, steady slope often reflects persistent directional movement. When slope starts flattening while price begins overlapping prior bars, that is a visual hint that efficiency is fading even if price has not fully reversed yet. Treat this as context, not a trigger by itself.
When AMA tends to work well for and why
AMA tends to help most in sustained directional moves that include pauses and pullbacks, but keep making progress. In those conditions, the ER logic often stays elevated enough that AMA remains responsive, which keeps the line close enough to price to be usable as a trend filter and pullback reference.
It also tends to be useful during transitions from consolidation to trend. When a stock breaks out and starts walking higher with less overlap, the move becomes more efficient and the line adapts. The benefit is not that AMA predicts the breakout. The benefit is that it can shift from slow to faster behavior without you changing settings mid move.
Finally, AMA can be a clean regime tool. Many traders do better when they separate decisions into trending versus non trending environments. AMA gives you a single line that often stays flatter and less reactive during chop, which can reduce the temptation to treat every bounce as a trend.
When AMA tends to fail and why
AMA is still a moving average. It can still whip in noisy markets, especially when price alternates between short bursts of direction and sharp reversals. ER can spike during a fast push, pulling AMA closer, and then collapse during the snapback, creating the feeling that the line changed its mind too late.
It also struggles with gaps and event driven candles because the efficiency logic can be distorted by a single large move inside the ER window. A gap can temporarily make the path look efficient even if the next few bars turn into overlap.
To reduce disappointment, use AMA as a context line and pair it with price structure and risk rules, not as a standalone entry signal. One simple filter that often helps is to demand that AMA slope is clearly rising or falling for several bars before treating it as a trend regime, rather than reacting to the first crossover.
Summary
Adaptive Moving Average AMA is a moving average that changes its sensitivity based on how efficiently price has been moving. It does this using an Efficiency Ratio that compares net progress to total wiggle, then converts that into a smoothing constant that moves between a fast and slow boundary.
Traders use AMA because it can act fast in clean trends and slow down in choppy ranges without changing settings manually. Common starting settings often revolve around ER 10 to 20 with fast and slow boundaries similar to 2 and 30 or slightly wider for smoother behavior. AMA tends to work best in persistent trends and clean post breakout advances. It tends to fail in whipsaw regimes, gap heavy action, and sharp reversal environments. Used as a trend filter and structure reference, it can support more consistent decisions when you study how past winners behaved during their sustained trend phases.