Moving Averages Explained

At a glance

What it measures: The average price over a lookback window
Primary use: Trend context and regime filtering
Key trade-off: Shorter means responsive and noisy; longer means smooth and late
Most cited setting: The 200-day, widely watched and therefore self-reinforcing

Key takeaways

A moving average is a smoothed version of price with a known lag of roughly half its lookback.
The most valuable use is as a filter: permitting long trades only above a long-term average removes exposure to sustained declines.
The choice between simple and exponential matters far less than the lookback and than what you do with the signal.
Moving average crossovers are widely tested and produce modest results on their own; they work better combined with volatility sizing and diversification.
The 200-day average is partly self-fulfilling because so many participants watch it, which is a legitimate reason to use it.

The main types and what differs

Type	Calculation	Character	When it matters
Simple (SMA)	Equal weight to every period	Smooth; each old value drops out abruptly	Long-term filters, widely watched levels
Exponential (EMA)	Exponentially declining weights	Responds faster to recent data	Shorter-term systems
Weighted (WMA)	Linearly declining weights	Between SMA and EMA	Rarely materially different
Hull (HMA)	Weighted combination to reduce lag	Faster, but more whipsaws	Short-term, high-noise contexts
Adaptive (KAMA, VIDYA)	Lookback varies with volatility	Smooth in noise, fast in trends	Adds parameters; test carefully
Volume weighted (VWAP)	Weighted by volume	Intraday execution benchmark	See VWAP

Studies comparing types on the same strategy generally find small differences relative to the effect of the lookback. Choosing EMA over SMA is a refinement; choosing 50 periods over 200 changes the strategy entirely.

What each lookback captures

Lookback	Captures	Typical use
10 to 20 days	Short-term swings	Exits for swing trades, short-term trend
50 days	Intermediate trend	Pullback context in swing strategies
100 to 150 days	Medium-term direction	Trend filters with fewer whipsaws than 50
200 days	Long-term regime	The standard bull and bear filter
10 months / 40 weeks	Same as 200-day on slower data	Position trading and asset allocation

Five legitimate uses

Regime filter. Trade long only when price is above a long average, and stand aside or reduce size when below.
Trend direction for entries. Require the average to be rising as well as price being above it, which filters out flat markets.
Dynamic support and resistance. In established trends, pullbacks frequently find buyers near a rising average, which gives a structured entry area.
Trailing exit. Exiting on a close below a moving average is a simple, testable trailing rule for position trading.
Baseline for other indicators. Bollinger Bands, MACD, and Keltner Channels are all built on moving averages, so understanding the base clarifies the derivatives.

Crossovers: what the evidence says

A crossover system buys when a faster average crosses above a slower one and exits or reverses on the opposite cross. It is the most tested rule in technical analysis, and the results are consistent: modest positive expectancy in trending markets, poor performance in ranges, and heavy dependence on diversification and position sizing rather than on the specific parameters.

Property	Typical value
Win rate	30 to 40 percent
Average win to average loss	2 to 3 times
Trades per instrument per year	4 to 12
Worst characteristic	Repeated whipsaws in range-bound periods
What improves it most	Diversification across uncorrelated markets and volatility sizing
What improves it least	Optimising the two lookback values

Typical characteristics of a daily crossover system on a diversified basket.

The full treatment, including specific rule sets, is in moving average crossover strategies.

Pitfalls

Whipsaws in ranges. In a sideways market, price crosses the average repeatedly, generating losses each time. A volatility or ADX filter reduces this.
Optimising the lookback. The best historical value is usually not the best future value. Prefer conventional settings and require a performance plateau.
Treating the average as a price level. It is a computed line that moves every day, not a level where orders rest.
Using adjusted versus unadjusted data inconsistently. Dividend adjustments change historical averages, so signals differ between data sources.
Applying equity conventions to other markets. The 200-day filter is a well-studied equity rule. Its behaviour in commodities and currencies differs and should be tested separately.

Frequently asked questions

Which is better, SMA or EMA?

Neither consistently. EMA responds faster to recent data, which helps in fast-moving conditions and produces more whipsaws in choppy ones. SMA is smoother and is what most market participants watch, particularly the 200-day. Differences in tested results are usually small compared with the effect of the lookback period.

What is the best moving average period?

It depends on the holding period you intend. As a general guide, 200 days for regime filtering, 50 days for intermediate trend context, and 10 to 20 days for short-term exits. Any strategy that only works at one specific unusual value has been fitted to history rather than to a market property.

Why is the 200-day moving average so important?

Partly because it captures roughly a year of trading and separates sustained uptrends from sustained downtrends reasonably well, and partly because it is watched by a very large number of participants, which makes reactions around it partially self-fulfilling. Both reasons are legitimate; the second is why arbitrary alternatives are not equivalent.

Do moving average crossovers actually work?

They produce modest positive expectancy when applied across a diversified basket of instruments with volatility-based position sizing, and poor results when applied to a single instrument without diversification. The crossover is a way of expressing trend persistence; the money is made by the portfolio construction around it.

Should I use moving averages on intraday charts?

They work mathematically, but intraday price contains far more noise relative to trend, so whipsaws are frequent and costs accumulate. Intraday traders more often use VWAP as a reference, since it reflects actual traded volume and is the benchmark institutional executions are measured against.

Test this idea before you trade it

Describe the rules in plain language and AlgoTrader AI turns them into a structured strategy blueprint with a configurable historical backtest, cost assumptions, and exportable code.

Build a backtest

Keep reading

Referenced by

Educational use only. This guide explains how a strategy works. It is not investment advice, not a recommendation, and no result described here is a forecast. Test any approach on historical and out-of-sample data, size positions conservatively, and never risk money you cannot afford to lose.