Quick answer: A moving average smooths price by averaging the last N closes, redrawn on every new bar. The SMA weights all N closes equally; the EMA weights recent closes more heavily, so it turns sooner. Traders use the resulting line to read trend direction, filter signals and mark crossovers — accepting that every version of the tool lags price by construction.
Almost every chart-based method begins with one question: is this market trending, and in which direction? Moving averages are the oldest working answer. The line compresses a noisy series of prices into a single line whose slope and position summarise the trend — at the cost of always arriving slightly late.
This guide covers what the line actually computes, the SMA vs EMA difference with the arithmetic shown, how to choose periods for a timeframe, what golden and death crosses mean, and the failure modes that crossover systems inherit.
What a Moving Average Shows
A moving average is the mean of the last N closing prices, recalculated as each new bar completes — the “moving” part. Plotted on the chart, it turns a jagged price series into a smoothed line: price holding above a rising line describes an uptrend, price below a falling line describes a downtrend, and price weaving through a flat line describes a range.
Three properties follow directly from the definition:
- Smoothing and lag trade off. A longer lookback produces a smoother line that reacts later. There is no setting that smooths without delaying.
- The line is a summary, not a forecast. It describes where price has been over the window. Any predictive use rests on trends persisting, which they sometimes do and sometimes do not.
- Slope carries most of the information. A rising 50-period line says more than price touching it once. Position plus slope is the standard reading.
Closing prices are the convention because the close is the bar’s settled value, though platforms allow the average to be applied to opens, highs, lows or median price.
SMA vs EMA: Calculation and Character
The two dominant types answer the same question with different weighting, and the difference is visible on every chart.
The SMA — equal weights. A 5-period simple moving average adds the last five closes and divides by five. Using EUR/USD closes of 1.0842, 1.0855, 1.0861, 1.0849 and 1.0868:
(1.0842 + 1.0855 + 1.0861 + 1.0849 + 1.0868) ÷ 5 = 1.0855
When the next bar closes at 1.0872, the oldest close (1.0842) drops out of the window and the SMA becomes 1.0861. Every close inside the window counts the same; a spike affects the line once when it enters and once more, in reverse, when it leaves.
The EMA — recent bars count more. An exponential moving average adds a fraction of each new close to the running value. The fraction is 2 ÷ (period + 1): 0.1818 for a 10-period EMA, 0.0952 for 20, 0.0392 for 50, 0.0100 for 200. One step, with a 10-period EMA at 1.0850 and a new close of 1.0870:
1.0850 + (1.0870 − 1.0850) × 0.1818 = 1.08536
| Property | SMA | EMA |
|---|---|---|
| Weighting | Equal across the window | Exponentially favours recent closes |
| First reaction to a move | Slow — 1/N of the change per bar | Fast — k of the gap immediately |
| Behaviour on old spikes | Jumps when a spike exits the window | Old data fades gradually, never exits |
| Typical use | Long-term trend reference (100, 200) | Shorter, reactive signals (9–50) |
The step response makes the character difference concrete. If price jumps and then holds flat, a 10-period SMA closes the gap in equal tenths and reflects the full move after exactly 10 bars. The 10-period EMA covers 18% of the gap on the first bar — far ahead of the SMA — but after 10 bars has covered only 86.6%, and technically never finishes. The EMA leads early and trails at the tail; the SMA plods and completes on schedule. Neither is “more accurate” — they answer the recency question differently.
One platform note: MetaTrader‘s Moving Average indicator, checked on a live MT5 build in August 2026, offers four methods — Simple, Exponential, Smoothed and Linear Weighted — plus a shift parameter that displaces the line horizontally. Smoothed and Linear Weighted are further weighting variants; the SMA/EMA pair covers the behaviour range most traders need.
Choosing Moving Average Periods
Periods are conventions, not laws — their value comes partly from how many other participants watch the same lines.
| Period | Typical timeframe | Standard reading |
|---|---|---|
| 9–10 | Intraday charts | Fast signal line, short-term momentum |
| 20–21 | Intraday to daily | Roughly one trading month on a daily chart |
| 50 | Daily | Medium-term trend reference |
| 100 | Daily to weekly | Longer trend filter |
| 200 | Daily | The institutional benchmark for bull vs bear territory |
The 200-day line earns its reputation from visibility: enough funds, desks and media reference it that price behaviour around it becomes partly self-fulfilling. On lower timeframes the same number means something entirely different — a 200-period line on a 5-minute chart summarises about 17 hours of trading, not 10 months.
Shortening a period produces earlier signals and more false ones; lengthening it filters noise and gives back more of every reversal before turning. That is the same smoothing-versus-lag trade the tool started with, and no combination of settings escapes it. Test any change on a demo account across a trend and a range before trusting it.
Golden Cross, Death Cross and Other Crossovers
A crossover strategy reads the relationship between two averages of different lengths. The fast line crossing above the slow line signals strengthening upward trend; crossing below signals the reverse.
- Golden cross: the 50-day average crosses above the 200-day — the classic long-term bullish reference on daily charts of indices and majors.
- Death cross: the 50-day crosses below the 200-day — the bearish mirror.
- Price crossovers: price itself closing across a single average (often the 20 or 50) — the fastest and noisiest variant.
Crossovers confirm trends rather than predict them. By the time a 50/200 cross prints, a substantial part of the move has already happened — the signal’s value is filtering direction, not timing entries.
The noise floor deserves a number. In August 2026 we simulated 300 trendless random-walk series of 250 bars each and counted 10/50 crossovers: an average of 7.1 crossings per series using SMAs and 7.5 using EMAs. On a daily chart with no directional information at all, a 10/50 crossover rule still fires roughly every seven weeks. A crossover system has to beat that base rate before it can claim an edge — and backtests that show it doing so are not evidence about the future. Past performance does not guarantee future results.
MACD formalises the crossover idea — it plots the gap between a 12- and a 26-period EMA rather than the lines themselves. The MACD guide covers that construction and its own failure modes.
Three More Ways Traders Use Moving Averages
Trend filter for other signals. The most defensible use: take long setups only while price holds above a rising 50- or 200-period line, shorts only below it. The average contributes direction; entries and exits come from structure, a risk-reward plan and a stop.
Dynamic support and resistance. In steady trends, pullbacks often stall near a widely watched average — the 20 EMA in fast trends, the 50 SMA in slower ones. The behaviour is real but irregular: price respects a line until it does not, so the level is a zone of interest for a setup, not a reason to enter on touch.
Baseline for envelopes and bands. Bollinger Bands are a 20-period SMA with standard-deviation bands around it; Keltner Channels wrap an EMA in ATR multiples. Understanding the average underneath explains most of how those tools behave.
Where Moving Averages Fail
Ranges are hostile territory. A sideways market drags every average flat and pushes price back and forth across it, generating whipsaw signal after whipsaw signal. The simulation above puts numbers on it — crossovers fire regularly with zero trend present. Most of a crossover system’s losses come from ranges, and most of its profits from the minority of persistent trends.
Lag is structural, not fixable. Every average is built from past prices. Reversals are confirmed after they begin; tops and bottoms are never signalled at the extreme. Settings tune where on the lag-noise curve a trader sits — they cannot leave the curve.
Self-fulfilment cuts both ways. Widely watched levels like the 200-day attract orders, which strengthens reactions — and also attracts stop-hunting wicks straight through the line in thin conditions.
A line is not a risk plan. An average gives no stop distance and no target. Position sizing, a stop set from volatility, and a defined risk-reward ratio have to come from elsewhere in the plan; the average only votes on direction.
Adding a Moving Average on MT4, MT5 or WebTrader
On MetaTrader, open Insert → Indicators → Trend → Moving Average, set the period, the method (Simple or Exponential for the versions covered here) and the price applied — close is the default and the convention. The line draws directly on the chart, and multiple averages can be layered for a crossover view. On WebTrader the indicator sits in the same trend group.
FXPrimus provides MT4, MT5 and WebTrader across forex, metals, indices and energies, with Negative Balance Protection on every live account. A free PrimusDEMO account is the place to compare an SMA and an EMA of the same period on live pricing — the character difference described above is visible within a few dozen bars, and demo testing costs nothing while the settings are being chosen.
Moving Averages — FAQ
What is a moving average in trading?
It is the mean of the last N closing prices, recalculated on every new bar and plotted as a line on the chart. Traders read its slope and price’s position relative to it as a summary of trend direction. All versions lag price, because they are built from past data.
What is the difference between SMA and EMA?
The simple moving average weights every close in its window equally; the exponential moving average weights recent closes more heavily using a factor of 2 ÷ (period + 1). The EMA reacts sooner to new prices, while the SMA is smoother and jumps when old extremes leave its window.
Which moving average is best for day trading?
No single setting is best in any measurable sense — shorter periods such as the 9 or 20 EMA are common intraday because they react quickly, at the cost of more false signals. The defensible approach is testing a specific period on the instrument and timeframe traded, on a demo account first.
What is a golden cross?
A golden cross is the 50-day moving average crossing above the 200-day, conventionally read as the start of a longer-term uptrend on daily charts. Its mirror, the death cross, is the 50 crossing below the 200. Both confirm moves already underway rather than predicting new ones.
Why is the 200-day moving average important?
Mainly because of who watches it: funds, analysts and financial media use the 200-day line as the boundary between bull and bear territory, so orders cluster around it. Price above a rising 200-day average is the most widely shared definition of a long-term uptrend.
Do moving average crossovers work?
They identify trend direction reliably and time entries poorly. In our simulation of trendless price series, a 10/50 crossover rule still fired about seven times per 250 bars, so crossovers alone carry a high false-signal rate in ranges. Most systems pair them with structure and a volatility-based stop.
What period should moving averages be set to?
Common daily-chart references are 20, 50, 100 and 200 periods, with 9–21 favoured intraday. Longer periods smooth more and lag more. The period defines how much history the line summarises, so the choice follows from the holding time of the strategy, not from a universal rule.
Is the moving average a leading or lagging indicator?
Lagging. It averages past prices, so it confirms changes in trend after they have begun. Shorter periods and exponential weighting reduce the delay but cannot remove it. Traders accept the lag in exchange for the noise filtering the average provides.
The Takeaway
The moving average summarises where price has been, weighted either equally (SMA) or toward the present (EMA), and every use of it — trend filter, crossover, dynamic level — inherits the same smoothing-versus-lag trade. It earns a place in a plan as the direction vote, next to structure for location and a volatility-based stop for risk. Used alone as an entry machine, it produces its base rate of false signals and no edge — the simulation numbers above are the reason.
FXPrimus offers MT4, MT5 and WebTrader with Negative Balance Protection on every live account, plus a free PrimusDEMO account for comparing SMA and EMA settings before committing capital.
[→ Open a PrimusDEMO account] | [→ Compare account types]
Risk disclosure. This article is published for informational and educational purposes and is not financial advice, legal advice or tax advice, nor a recommendation to trade any instrument or to use any indicator or strategy. Trading forex and CFDs carries a high risk of loss and is not suitable for every investor; you may lose more than your initial deposit unless Negative Balance Protection applies. Past performance does not guarantee future results, and simulated or backtested figures do not reflect live execution costs. Performance statistics published by signal providers and strategy sellers are often self-reported and should be verified against underlying statements. Spreads and trading conditions referenced are indicative, self-reported by FXPrimus, and vary by account type. Review the full terms and conditions and the relevant risk disclosure before opening an account or placing a trade. FXPrimus is a trading name of entities regulated in multiple jurisdictions; the entity you contract with, and the protections that apply, depend on your country of residence.