How to configure the Arnaud Legoux Moving Average? (ALMA settings)

Understanding ALMA Core Parameters

1. The Arnaud Legoux Moving Average relies on three primary input variables: period, sigma, and offset.

2. Period defines the number of bars used in the calculation — typical values range from 9 to 100 depending on timeframe and asset volatility.

3. Sigma controls the width of the Gaussian distribution curve applied during smoothing — lower values produce tighter curves, higher values broaden responsiveness.

4. Offset determines where the center of the Gaussian kernel falls relative to the most recent bar — expressed as a fraction between 0 and 1, with 0.85 being the standard setting.

Default Configuration in Major Platforms

1. TradingView sets ALMA’s default parameters to period=9, sigma=6, offset=0.85 for intraday charting.

2. MetaTrader 4 and MT5 implementations often ship with period=25, sigma=6, offset=0.85 for swing trading contexts.

3. Some crypto-focused charting tools adjust sigma upward to 7 or 8 when applied to high-noise assets like meme coins.

4. Exchanges with built-in technical analysis modules — such as Bybit and OKX — expose ALMA under custom indicators with editable fields matching the original formula.

Adapting ALMA for Cryptocurrency Volatility

1. Bitcoin daily charts commonly use period=34, sigma=5.5, offset=0.88 to reduce whipsaw while preserving trend alignment.

2. Ethereum 4-hour charts frequently deploy period=21, sigma=6.2, offset=0.82 to balance lag and sensitivity amid frequent pump-and-dump cycles.

3. Altcoin pairs traded on decentralized exchanges benefit from shorter periods — often period=13 with sigma=4.8 — due to compressed liquidity windows.

4. Stablecoin-denominated pairs like USDT/BTC sometimes apply period=50 and sigma=7.1 to filter micro-fluctuations caused by arbitrage bots.

Mathematical Implementation Details

1. ALMA computes weighted sums using a Gaussian distribution: w(i) = exp(-((i - offset × (period - 1))² / (2 × sigma²))).

2. Each weight is normalized so that the sum across all i equals one before multiplying by price values.

3. Unlike SMA or EMA, ALMA does not assign equal or exponentially decaying weights — it intentionally peaks near the offset point.

4. The algorithm requires full recalculation per bar; no recursive shortcut exists, making it computationally heavier than traditional moving averages.

Frequently Asked Questions

Q: Can ALMA be used as a standalone entry signal in spot BTC trading?Yes, but only when combined with volume confirmation and candlestick structure — isolated ALMA crossovers generate excessive false signals on low-cap tokens.

Q: Does ALMA repaint when applied to real-time tick data?No, ALMA is non-repainting because it uses only historical close prices and fixed kernel parameters — no future data points are referenced.

Q: How does ALMA behave during exchange outages or missing candles?Missing candles cause gaps in the series, resulting in shifted weight centers — backfilling with interpolated OHLC values restores alignment but introduces estimation bias.

Q: Is ALMA compatible with Heikin-Ashi smoothed candles?Yes, though smoothing cascades — Heikin-Ashi already filters noise, so pairing it with ALMA may over-smooth price action, especially on low-volume altcoin pairs.