HMA, DEMA & TEMA: Low-Lag Moving Averages Explained

The eternal complaint about moving averages is lag — by the time the average turns, the move is half over. Three classics attack lag head-on: the Hull MA (HMA), the Double EMA (DEMA), and the Triple EMA (TEMA). All three hug price far tighter than an EMA of the same length while staying smoother than simply shortening the period.

How Each One Works

• HMA (Hull): combines weighted averages of the full and half period, then smooths by √period. The smoothest of the three for its lag — turns are visibly crisp.
• DEMA: 2 × EMA − EMA(EMA). Subtracting the "average of the average" cancels most of the lag.
• TEMA: extends the same trick to a third level — the fastest of the trio, and the most nervous.

Speed ranking: TEMA > DEMA ≈ HMA > EMA > SMA. Smoothness ranking is nearly the reverse — except HMA, which cheats and gets both.

How to Trade Them

1. Direction/slope filter. HMA's specialty: trade only in the direction the HMA is sloping. Its crisp turns make the slope test unusually reliable.

2. Price cross. Price crossing DEMA/TEMA fires much earlier than crossing an EMA — good for fast systems, mandatory to pair with a regime filter.

3. Fast/slow crossover. A TEMA-over-DEMA or HMA pair crossover keeps the classic two-MA system but compresses its lag dramatically.

Building Them in Setup.Cash

All three are available in the strategy builder with period and source inputs — use them anywhere a standard moving average fits. For adaptive alternatives that change speed with the market, see KAMA; the extended library adds ZLMA (Zero Lag), JMA (Jurik), T3 (Tillson), ALMA, and the Super Smoother Filter.

Tuning

• HMA 20 / DEMA 20 / TEMA 20: sensible starting points.
• Less lag = more whipsaw in chop. Budget for it with a volatility filter or ADX gate.

Low-lag averages don't predict better — they react faster, and pay for it in ranges. Backtest against a plain EMA baseline to see whether the speed is worth it on your market.