# Types of Moving Averages: SMA, EMA, WMA, Hull, DEMA, TEMA and VWMA

Web page: https://uptickflow.com/en/learn/types-of-moving-averages  
Updated: 2026-10-29

Types of moving averages explained: the weighted moving average formula, Wilder’s SMMA, DEMA vs EMA, TEMA, the Hull moving average and VWMA vs VWAP.

The main types of moving averages are the simple (SMA), exponential (EMA), weighted (WMA), smoothed or Wilder’s (SMMA/RMA), double and triple exponential (DEMA, TEMA), Hull (HMA) and volume-weighted (VWMA); they differ in how much weight each past close gets, which sets how much they lag.

Below: each formula with a worked example, why the faster types overshoot, VWMA vs VWAP, a short look at the adaptive KAMA, and a side-by-side test of lag and smoothness at the same length. The SMA and EMA basics are in [moving averages explained](https://uptickflow.com/en/learn/moving-averages). Prices are hypothetical, and trading carries real risk.

## What are the types of moving averages in trading?

Every moving average is a weighted average of past closes. The type decides the weights, and the weights decide the trade-off: put more weight on recent bars and the line turns sooner but follows more noise; spread the weight evenly and the line is calmer but later. No type escapes it; the faster ones only move the place where you pay.

The two everyone knows take a sentence each. The simple moving average (SMA) gives each of its N closes the same weight. The exponential moving average (EMA) gives the newest close a weight of 2 ÷ (N + 1) and lets older closes fade without ever dropping them. The drawing at the top shows four 20-day types on the same hypothetical closes: after the lowest close, the Hull reached its own low 2 bars later, the WMA 7, the EMA 8 and the SMA 11.

## What is a weighted moving average (WMA), and what is its formula?

A weighted moving average gives the newest close a weight of N, the one before it N − 1, and so on down to 1 for the oldest close in the window, then divides by the sum of the weights, N(N + 1) ÷ 2. The weights fall in a straight line, so the WMA leans toward recent prices more than an SMA of the same length and, unlike the EMA, drops each close completely once it leaves the window.

Worked example, five hypothetical XYZ daily closes, oldest first: $50.00, $51.00, $53.00, $52.00 and $56.00. Weighted: 50 × 1 + 51 × 2 + 53 × 3 + 52 × 4 + 56 × 5 = 799. The weights add up to 15, so the 5-day WMA is 799 ÷ 15 = $53.27. The 5-day SMA of the same closes is 262 ÷ 5 = $52.40. The WMA sits closer to the last close because the $56.00 counts five times as much as the $50.00.

A detail most explanations miss: the newest close gets the same weight in a WMA and an EMA of the same length, 2 ÷ (N + 1), or 33.3% for five days. The difference is the tail. The WMA’s weights reach zero at bar N, while the EMA keeps a sliver of every close it has seen. On a steady trend that tail costs the EMA lag: a 20-day SMA and a 20-day EMA both trail a straight-line trend by 9.5 bars, a 20-day WMA by about 6.3.

- **Weighted moving average (WMA)**: A moving average whose weights fall in a straight line, from N on the newest close to 1 on the oldest, divided by the sum of the weights, N(N + 1) ÷ 2.

## SMMA or RMA: Wilder’s smoothing inside RSI and ATR

The smoothed moving average (SMMA), which TradingView’s code calls RMA, is J. Welles Wilder Jr.’s method: each new value keeps (N − 1) ÷ N of the previous one and adds 1 ÷ N of the new close. It is an EMA with a smaller weight on the newest bar, 1 ÷ N instead of 2 ÷ (N + 1), so an SMMA of length N behaves exactly like an EMA of length 2N − 1. A 14-period RMA is a 27-period EMA under another name.

You rarely plot it on price, but you use it all the time: Wilder built it into the [RSI indicator](https://uptickflow.com/en/learn/rsi-indicator) and the [average true range](https://uptickflow.com/en/learn/average-true-range), both with a default of 14 periods.

## DEMA vs EMA, and TEMA: how do they cut lag?

Patrick Mulloy introduced the double exponential moving average (DEMA) in Technical Analysis of Stocks & Commodities in January 1994, and the triple version (TEMA) the following month. Despite the name, a DEMA is not an EMA of an EMA. It measures how far a second EMA, taken of the first, trails behind, and adds that gap back: DEMA = 2 × EMA − EMA(EMA). The TEMA repeats the trick one level deeper: TEMA = 3 × EMA − 3 × EMA(EMA) + EMA(EMA(EMA)).

The correction works on a steady trend, where both lines lag almost nothing. The cost shows at turns and in noise. Because they project the recent slope, they overshoot: on a clean V-shaped test path, a 20-period DEMA and TEMA both dipped below the actual low before turning up, and on random noise they let through 13.3% and 19.3% of its bar-to-bar jitter, against 6.9% for the 20 EMA. So the honest answer to DEMA vs EMA is earlier turns for more false wiggles, and TEMA vs DEMA vs EMA is the same trade pushed further.

- **Double exponential moving average (DEMA)**: Patrick Mulloy’s lag-reduced average, 2 × EMA − EMA(EMA). Despite its name, it is not an EMA of an EMA.

## Hull moving average: formula, settings and Hull vs EMA

Alan Hull designed the Hull moving average (HMA) in 2005 to cut lag while keeping the line smooth. It is built from weighted moving averages in three steps: take a WMA of half the length, double it and subtract a WMA of the full length, which pushes the line ahead; then smooth the result with a WMA whose length is the square root of the original. In one line, HMA = WMA(2 × WMA(N ÷ 2) − WMA(N), √N), with whole numbers, so a 16-period HMA uses WMAs of 8 and 16, then a WMA of 4.

Hull vs EMA: in the drawing at the top, the 20-day HMA reached its low 2 bars after the lowest close, against 8 bars for the 20-day EMA. It also dipped about $0.50 below that close before turning, and it was the first to turn down on the late pullback. That is the HMA’s nature: fast at turns, prone to overshoot, and jumpier than an EMA when price goes sideways. There is no best setting; [best moving average settings](https://uptickflow.com/en/learn/best-moving-average-settings) shows how to choose a length without fitting it to the past.

- **Hull moving average (HMA)**: Alan Hull’s fast average, WMA(2 × WMA(N ÷ 2) − WMA(N), √N). It turns early and can overshoot at sharp turns.

## VWMA vs VWAP: what is the difference?

The volume-weighted moving average (VWMA) weights each close by that bar’s volume: the sum of close × volume over the last N bars, divided by the volume of those bars. Hypothetical: three closes of $20, $21 and $22 on 1,000, 1,000 and 4,000 shares give a VWMA of 129,000 ÷ 6,000 = $21.50, against an SMA of $21.00, because the heavy bar pulls the line toward its price.

VWAP sounds similar but works differently. It accumulates price × volume from the start of the session, or from an anchor you choose, instead of using a rolling window of N bars; TradingView uses each bar’s typical price, and the session VWAP resets every day. So the VWMA is a trend line for any timeframe, while VWAP is the session’s average price and mostly an intraday tool: see the [VWAP trading strategy](https://uptickflow.com/en/learn/vwap-trading-strategy) and, for multi-day anchors, [anchored VWAP](https://uptickflow.com/en/learn/anchored-vwap).

- **Volume-weighted moving average (VWMA)**: The sum of close × volume over the last N bars divided by their total volume. It equals the SMA when every bar has the same volume.

## KAMA: what is an adaptive moving average?

Perry Kaufman’s adaptive moving average (KAMA) changes its own speed. It computes an efficiency ratio over the last N bars, commonly 10: the net price change divided by the sum of all the bar-to-bar moves. A clean trend scores near 1, and the KAMA moves almost as fast as a 2-period EMA; choppy back-and-forth trading scores near 0, and it slows toward a 30-period EMA. The usual 10, 2 and 30 are starting points, not rules.

It solves part of the trade-off, not all of it: a sudden reversal out of a quiet range still catches it late, and it is calculated from the same closes as every other type.

## Which moving average should you use? Lag vs smoothness compared

The drawing below tests seven types at the same length, 20, two ways: how many bars each takes to turn up after a perfectly V-shaped low, and how much of a random series’ bar-to-bar jitter it lets through. The RMA turned after 13 bars with 3.6% jitter, the SMA after 10 with 5.0%, the EMA and WMA after 6 with 6.9%, the HMA and DEMA after 3 with 11.1% and 13.3%, and the TEMA after 2 with 19.3%. Pick the type by its job, not its reputation:

- SMA: the convention for slow, widely watched lines such as the [200-day moving average](https://uptickflow.com/en/learn/200-day-moving-average) and the golden cross.
- EMA: the default for trading lines, pullbacks and crossovers, as in the [EMA trading strategy](https://uptickflow.com/en/learn/ema-trading-strategy).
- WMA: slightly less lag than an EMA on a steady trend, with the same jitter; mostly a building block, as in the Hull.
- SMMA/RMA: slow and smooth; you meet it inside RSI and ATR.
- DEMA, TEMA and HMA: early turns for fast trading or exits, with more overshoots and false wiggles; pair them with a slower filter.
- VWMA: a trend line that leans toward prices where volume was heavy. Several lines of one type, stacked, make a [moving average ribbon](https://uptickflow.com/en/learn/moving-average-ribbon).

## Key takeaways

- Every moving average is a weighted average of past closes; the type only changes the weights, and the weights set the trade-off between lag and noise.
- The WMA’s weights fall in a straight line from N to 1, divided by N(N + 1) ÷ 2; the SMMA/RMA is an EMA with a 1 ÷ N weight, the slow line inside RSI and ATR.
- DEMA, TEMA and the Hull cut lag by projecting the recent slope, so they turn early and overshoot; at length 20 they let through two to four times the jitter of an SMA.
- The VWMA is a rolling, volume-weighted trend line; VWAP is a cumulative session average that resets every day. Choose a type for its job and keep it consistent.

## FAQ

**What are the types of moving averages?**

The common ones are the simple (SMA), exponential (EMA), weighted (WMA), smoothed or Wilder’s (SMMA/RMA), double and triple exponential (DEMA, TEMA), Hull (HMA), volume-weighted (VWMA) and adaptive averages such as KAMA. All of them average past closes; they differ in how they weight them.

**What is the weighted moving average formula?**

WMA = (N × newest close + (N − 1) × the close before it + … + 1 × oldest close) ÷ (N(N + 1) ÷ 2). For five closes of 50, 51, 53, 52 and 56, that is 799 ÷ 15 = 53.27.

**Is the Hull moving average better than the EMA?**

It is faster, not better. In a 20-period test the HMA turned up 3 bars after a V-shaped low against 6 for the EMA, but it let through about 60% more jitter and dipped below the low before turning. It suits traders who want early turns and accept more false ones.

**DEMA vs EMA: which is better?**

Neither in general. A DEMA, 2 × EMA − EMA(EMA), has almost no lag on a steady trend and turns sooner, but it overshoots at turns and wiggles more when price moves sideways. Use it where speed matters more than smoothness, and test it on your own charts first.

**What is the difference between VWMA and VWAP?**

The VWMA is a rolling average of the last N closes weighted by volume, so it works on any timeframe. VWAP is a cumulative volume-weighted average of the session’s prices that resets every day, so it is mainly an intraday reference.

---

Educational content, not financial advice. Trading involves risk of loss. https://uptickflow.com/llms.txt