EMA (Exponential Moving Average): Formula and Use

Definition

An exponential moving average (EMA) is a type of moving average in which recent candles count more than older ones. Every past price still contributes, but its influence shrinks exponentially with age. As a result, an EMA follows price more closely than a simple moving average of the same length.

Formula

First compute the smoothing factor:

α = 2 ÷ (N + 1), where N is the EMA length.

Then on every new candle:

EMA_today = α × Close_today + (1 − α) × EMA_yesterday

Example for a 9-period EMA: α = 2 ÷ 10 = 0.2. If yesterday's EMA was 100 and today's close is 105, today's EMA = 0.2 × 105 + 0.8 × 100 = 101.

The very first EMA value needs a starting point. Some platforms seed it with an SMA; TradingLead's TLScript-Pine ta.ema seeds from the first valid sample. Early values can therefore differ slightly between platforms, then converge.

How traders read it

• Trend filter: price above a rising 50 or 200 EMA is often treated as an uptrend; below a falling one as a downtrend.
• Crossovers: a fast EMA (for example 9 or 20) crossing a slow EMA (50) is a common momentum signal.
• Pullback zones: in strong trends on Nifty or large-cap stocks, traders watch the 20 EMA as a possible pullback area.
• Building block: MACD is built from the difference between two EMAs.

Common settings

• 9 / 20 — short-term and intraday momentum.
• 50 — medium-term trend.
• 100 / 200 — long-term trend on daily charts.

Pitfalls

• Faster reaction also means more false turns in choppy markets.
• EMA crossovers still lag — by the time they confirm, a large part of the move may be over.
• Values depend on how much history is loaded; very short histories give less stable early EMAs.
• Never use one EMA rule alone as a trading decision; test it first.

How to add an EMA on TradingLead

1. Open a chart at tradinglead.in/charts.
2. Click Indicators and search Moving Average Exponential, then set the length.
3. For four EMAs at once, add EMA 20/50/100/200 Multiple and change any length in its settings.

To build your own in Indicator Studio (documented Pine v5/v6 subset, validated before execution):

//@version=6
indicator("EMA pair", overlay=true)
plot(ta.ema(close, input.int(20, "Fast")), "Fast EMA", color=color.teal)
plot(ta.ema(close, input.int(50, "Slow")), "Slow EMA", color=color.red)

In Python, df["close"].ewm(span=20, adjust=False).mean() gives a similar EMA — see Write a Python indicator.

Educational content only — not investment advice.