Hull Moving Average (HMA) Explained: Smooth Trend Signals with Less Lag

Every moving average trades lag for smoothness. The Simple Moving Average (SMA) is smooth but reacts slowly to price changes. The Exponential Moving Average (EMA) is more responsive but can be noisy. The Hull Moving Average, developed by Alan Hull in 2005, solves this trade-off: it's nearly as smooth as the SMA while reacting to price almost in real time.

The secret is a two-step weighted average construction that effectively cancels out most of the built-in delay. Traders use the HMA not just as a trend-following line, but as a slope-based signal — when the HMA curves upward, you're long; when it curves downward, you're short.

The HMA Formula

Alan Hull's construction starts with Weighted Moving Averages (WMA), which give heavier weight to recent candles than the SMA does. The key insight is to compute two WMAs of different lengths, then combine them to project price forward and eliminate lag.

Working through a concrete example with period 21: compute WMA(10) on price (the fast half), compute WMA(21) on price (the full period), double the fast WMA and subtract the slow one to get the raw "de-lagged" series, then smooth that raw series with a final WMA(√21 ≈ 5). The result is a line that hugs price tightly without the whipsaws you'd get from a raw fast average.

Why WMA instead of EMA?

The WMA's linear weighting (most recent bar gets weight n, second-most-recent gets n−1, etc.) makes the double-WMA subtraction cancel lag algebraically in a clean way that EMA's exponential decay doesn't allow. EMA-based versions exist but produce slightly less smooth results.

Reading HMA Slope as a Signal

Unlike most indicators, traders don't use HMA crossovers with price or with another MA. The signal is simpler: is the HMA currently rising or falling?

The slope signal

If HMA[current] > HMA[1 bar ago], the slope is positive — the market is in an uptrend and you look for longs. If HMA[current] < HMA[1 bar ago], the slope is negative — downtrend, look for shorts. Many implementations colour the HMA line green when rising and red when falling so you can read trend direction at a glance.

A slope flip is the entry trigger. When HMA transitions from falling to rising (i.e., the current bar's HMA value first exceeds the prior bar's), that's a potential long entry. The opposite flip signals a potential short. Because the HMA is so smooth, these flips are relatively rare compared to noisy fast EMAs, meaning fewer false signals.

Slope-flip confirmation

A single slope flip can still be premature. Many traders require the HMA to close two consecutive bars in the new direction before acting. For example, on the 1h chart with HMA(21): if the HMA was falling for several bars and then prints two consecutive rising bars, that's a confirmed slope flip. Enter on the close of bar 2 with a stop below the recent swing low.

HMA vs. EMA vs. SMA: How Different Are They?

All three use the same period but behave very differently in trending vs. choppy markets.

In a strong trend

All three will eventually point in the right direction, but HMA gets there first. In a sustained BTC uptrend, an HMA(21) might slope upward from bar 3 of the move while the SMA(21) still points sideways until bar 10. That's 7 bars of missed trend — significant on a 1h chart where each bar is an hour of price action.

In sideways markets

This is where the HMA's responsiveness becomes a liability. Because it reacts fast, it whipsaws more than a slower SMA when price is choppy. In a 2% ranging market, an HMA(21) might flip slope direction 6–8 times in a session versus 2–3 for the SMA. The solution is to combine HMA with a volatility or choppiness filter: only trade HMA slope flips when the ADX is above 25 or the Choppiness Index is below 50, indicating genuine trend conditions.

Using HMA on Strategester

Strategester's HMA Slope strategy applies exactly this logic to live market data. The strategy computes HMA(21) on whatever timeframe you select — 5m through 4h — and generates a signal whenever the slope flips direction and holds for at least one additional bar to reduce false entries. You can see the signal status right now across BTC, ETH, SOL, and BNB without creating an account.

The confluence scoring system also weights the HMA slope signal alongside RSI, MACD, and other indicators. A slope flip that aligns with RSI emerging from oversold territory (below 35) and MACD crossing positive is a much stronger setup than a slope flip in isolation — the confluence score quantifies exactly that stack.