Adaptive Moving Average

Explanation & Rationale

The Adaptive Moving Average (AMA) strategy dynamically adjusts to market conditions by incorporating an efficiency ratio that measures trend strength relative to volatility. Unlike traditional moving averages, the AMA responds more quickly during strong trends and smooths out noise during sideways markets, making it more adaptive to price changes. This approach helps traders capture meaningful trends while avoiding false signals in choppy or low-volatility environments.

Code

'''
Adaptive Moving Average (AMA) Strategy.

Buy when the AMA turns upward.
Short when the AMA turns downward.
This strategy adjusts dynamically to market volatility using an efficiency ratio.
Learn more @ docs.ubacktest.com/examples/moving-averages/ama
'''

import pandas as pd
import numpy as np

def calculate_ama(series, window=10, fast_ema=2, slow_ema=30):

    # calculate efficiency ratio from price diffs / volatility
    price_change = abs(series.diff(window))
    volatility = series.diff().abs().rolling(window=window).sum()

    efficiency_ratio = price_change / volatility.replace(0, np.nan)
    efficiency_ratio.fillna(0, inplace=True)

    smoothing_constant = (efficiency_ratio * (2 / (fast_ema + 1) - 2 / (slow_ema + 1)) + 2 / (slow_ema + 1)) ** 2
    ama = series.copy()

    # build a series of ama values to correspond with the smoothed ama
    for i in range(1, len(series)):
        ama.iloc[i] = ama.iloc[i - 1] + smoothing_constant.iloc[i] * (series.iloc[i] - ama.iloc[i - 1])

    return ama

def strategy(data):
    data['AMA'] = calculate_ama(data['close'])

    # Generate signals based on AMA trend
    data['signal'] = 0
    data.loc[data['AMA'].diff() > 0, 'signal'] = 1  # Buy
    data.loc[data['AMA'].diff() < 0, 'signal'] = -1  # Short

    return data