Moving Averages: SMA, EMA & WMA Explained

What a moving average actually is

Stock prices are noisy. Reliance can close at ₹2,950 on Monday, ₹2,920 on Tuesday and ₹2,975 on Wednesday without anything fundamental changing. Looking at three sessions in a row, the trend is invisible.

A moving average smooths that noise out. Add up the closing prices of the last 50 sessions and divide by 50. You get a single number that sits roughly where Reliance has been trading on average.

Plot that number every session and you get a line that slopes gently up or down with the broader direction of the stock.

That line does two simple jobs. It tells you whether the stock is trending up or down, just by the slope. And it acts as a moving floor in an uptrend and a moving ceiling in a downtrend — what chartists call dynamic support and resistance — because prices in a trend tend to bounce off their own average.

The 50-day and 200-day SMA are the two most-watched lines on Indian charts. Many traders use the 200-day SMA as a rough line between long-term strength and weakness, while the 50-day marks the medium-term trend.

That 200-day line is a widely watched gauge of the long-term trend — a convention, not a literal wall between bull and bear markets.

On every large-cap — the big, heavily traded companies like HDFC Bank, Infosys or Tata Motors — those two lines quietly frame how most big institutions and serious traders read the stock.

The moving average does not predict anything. It is a lag indicator: it shows what has already happened, smoothed out. It is useful anyway because trends in stocks tend to persist for weeks or months, so the rear-view mirror still tells you something about the road ahead.

How SMA, EMA and WMA are built

All three averages take the closing prices of a set number of recent sessions and combine them into one number. The difference is how much weight each session gets.

The simple moving average is the easiest. Take the last 50 closing prices of Infosys, add them up, divide by 50. That single number is the 50-day SMA — every session in the window counts equally.

The "moving" part is just a rolling window. Tomorrow, you drop the oldest price from the list, add the newest close, and recalculate — so the average creeps forward one session at a time.

The exponential moving average — the EMA — is the one most short-term traders prefer. It averages the same recent sessions, but it does not treat them equally: it leans heavily on the latest prices and gives the oldest sessions almost no say at all.

There is a formula behind it, but you never have to touch it. The EMA uses a fixed multiplier of 2 ÷ (lookback period + 1) to decide how much weight the newest close gets.

For a 20-period EMA that works out to 2 ÷ (20 + 1) = 0.0952 — about 9.52%, or roughly a tenth of the whole line riding on the most recent close alone. Today's price counts the most, yesterday's a little less, the session before less again, fading out as you go back.

The practical result is that the EMA reacts faster than the SMA. If Reliance gaps up two percent in a single session, the 20-day EMA jumps that same session.

The 20-day SMA, by contrast, barely twitches — it spreads that one jump thinly across its whole window. The faster line catches a real turn earlier, which is exactly why short-term traders lean on it.

The weighted moving average — the WMA — is the in-between option. It also gives recent sessions more weight, but in a steady, evenly stepped way using whole-number weights.

In a 5-day WMA, the most recent close counts five times, the one before it four times, then three, two and one for the oldest. Those weights add up to 15 (that is 5 + 4 + 3 + 2 + 1), and you divide by 15 at the end.

That makes the WMA more responsive than the SMA but calmer than the EMA. In practice, most traders skip it — once you understand the simple and the exponential, the weighted rarely earns a place on the chart.

In every charting app, you just pick the type, set the length, and the line appears. You will never compute these by hand. What matters is knowing which one you are looking at, and why a faster line is not always a better line.

So in practice, most traders on Indian markets settle on the SMA for the big-picture, higher timeframes and the EMA for the faster, lower ones. The WMA sits quietly in the menu, rarely chosen.

How to read a moving average

A single moving average on a chart does three things, and most beginners only notice the first one.

The first is slope. If the line is sloping up, the trend over its lookback period is up. If it is sloping down, the trend is down.

The slope of the 200-day SMA on Nifty is the single most useful read on the overall regime — the market's current mode, whether it is in an uptrend, a downtrend, or going sideways. A flat line means the stock is going nowhere.

The second is position relative to price. Price above a rising moving average is a healthy uptrend. Price below a falling moving average is a healthy downtrend. Price chopping around a flat moving average is a range, and that is exactly when the indicator is least useful.

The third is the line acting as dynamic support or resistance. In an uptrend, pullbacks often pause at the 20-day or 50-day moving average and bounce off it. In a downtrend, rallies often stall at the same lines on the way up.

It is not magic. So many traders watch the same line that orders tend to pile up around it.

When you add a second moving average, things get more interesting. Plot the 50-day and 200-day SMA together on the Nifty daily chart. By the most commonly used convention, when the 50-day crosses above the 200-day, technicians call it a golden cross and read it as the start of a new long-term uptrend. The reverse, the 50-day cutting below the 200-day, is the death cross.

Both signals lag the actual turn in price by weeks. The golden cross on Nifty formed in late August 2020 — months after the March low.

That is not a flaw. It is the price of certainty. You give up some of the move in exchange for confirmation that the regime has actually changed — and because these are lagging signals, they can still throw the odd false alarm.

For shorter timeframes, traders stack three EMAs instead. A common combination on the Bank Nifty hourly chart is the 9-period, 20-period and 50-period EMA. (On an intraday chart we say "period" rather than "day", because each session is one candle — an hour or 15 minutes — not a full day.)

When they are stacked in order, with the 9-period on top, the 20-period in the middle and the 50-period below, the trend is up. When they flip and stack with the 9-period at the bottom, the trend is down. The order of the stack is the regime, and crossovers between the lines mark the change.

Moving averages on Indian charts

The clearest way to learn moving averages is to walk through what the lines were doing on familiar Indian charts during well-known events.

Nifty in 2020 is the textbook case. After peaking near 12,400 in mid-January, the index first closed below its 200-day SMA on 26 February as the pandemic hit. Anyone using the simple slope-and-position rule was out of long positions early.

As March wore on, price stayed deep below a falling 200-day line, collapsing to a low of 7,610 on 23 March — roughly 38% below the January peak. The death cross, the 50-day SMA dropping below the 200-day, printed on 19 March, only confirming the danger after the index had already fallen hard.

Then the bounce came, and it was sharp. Nifty reclaimed its 200-day SMA on a closing basis on 17 July 2020, and the golden cross — the 50-day crossing back above the 200-day — followed in late August 2020.

From there the index ran on, printing fresh record highs near 13,980 by the end of 2020 and pushing toward 18,000 over the following year. Both signals lagged the actual turn badly — yet they still kept a patient trader on the right side of a year-long move.

A trader who simply waited for price to close back above the 200-day SMA sat out the panic, sat out the false rallies on the way down, and still caught most of the recovery. That is the slow, lagging power of the line.

Reliance through 2022 is the opposite lesson. For long stretches the stock just chopped sideways, with the 50-day SMA wandering more or less flat. Price crossed back and forth over that flat line every couple of weeks.

Anyone trading every cross lost money. Anyone who read the flat slope as a sign to stand aside saved themselves a quarter of pointless churn.

HDFC Bank in a clean, steady uptrend is yet another flavour. When a stock trends smoothly, pullbacks tend to pause at the 20-day or 50-day line and turn back up, and the trade almost reads itself.

On a chart like that, the SMA and the EMA of the same length look almost identical. The two only really part ways at sharp turns — which is the whole reason the choice between them matters.

Bank Nifty on a short, 15-minute chart, with the 9-period, 20-period and 50-period EMA stacked together, is where a lot of intraday trading happens. On busy days the three lines squeeze tight around price and then fan out as one side takes control.

Trading only in the direction of the stack — and ignoring counter-stack signals, the ones that go against the order of the EMA stack — can help you avoid many counter-trend trades.

The pattern across all four examples is the same. The moving average works when the slope is clean and price respects the line. It fails when the slope is flat and price treats the line as a punching bag.

When moving averages fail

If moving averages are this useful, why do so many retail traders lose money using them? The honest answer is the same one every trend tool runs into. Moving averages are built for trends, and most stocks are not trending most of the time.

This is worth saying plainly: sideways markets make trend tools look foolish, even for experienced traders. If a moving average keeps giving you false signals in a flat market, that is the tool meeting its limit — not you failing. Do not blame yourself for every whipsaw.

A stock stuck in a range produces a constant stream of small crossovers. Price closes above the 20-period EMA, then below, then above again, all in the space of a week. Every signal looks real for a session or two and then flips — a whipsaw, a false signal where price crosses the line and quickly reverses.

Watch out, too, for a dead-cat bounce — a temporary bounce inside a falling market that can briefly lift price back over a moving average before the downtrend resumes. It can look like a fresh signal and be nothing of the sort.

Each loss is small. Ten of them in a quarter, plus the brokerage and taxes you pay on every trade, add up to a real dent in the account.

Shortening the moving average makes this worse, not better. The faster the line, the more it whips. A 5-day EMA on a sideways stock flips direction almost every other day. The signals are louder, but they mean less.

Three simple filters fix most of the damage.

The first is the higher timeframe. Only take signals on the hourly that agree with the daily trend. A bullish moving-average crossover on a stock trading below its 200-day SMA is fighting the larger picture, and most of the time the larger picture wins.

The second is the slope. Skip any signal that fires on a flat moving average. If the line is travelling sideways, the stock is range-bound by definition and the cross is noise. Wait for the line itself to be tilting clearly up or down before paying attention to where price sits relative to it.

The third is a nod from price itself. A push above the most recent high — what chartists call a swing high, a pivot where price last turned down — or a bullish engulfing candle, a candle pattern where buyers clearly overpower the previous candle, is worth far more than a bare moving-average cross. Combine the two and the false signals drop away sharply.

Apply those filters and the number of valid signals on a typical large-cap drops from twenty or thirty a year to a handful. Those few are the ones worth taking.

How to actually use moving averages

The framework, once it is clear, is small. Moving averages are not a system on their own. They are a filter and a context for trades you would consider anyway from price action.

Use them for three jobs, in this order. Read the regime, find the pullback entries inside that regime, and use the line as a soft trailing stop.

For the regime, look at the 200-day SMA on the daily. If price is above it and the line is rising, the bias is long. If price is below and the line is falling, the bias is short or flat. Anything else is a no-trade environment for trend strategies.

For the entries, wait for price to pull back into a faster line inside that trend — usually the 20 or 50 EMA. A bounce off it, in the direction the slow line is already pointing, is the cleanest setup there is.

Pair that bounce with a simple confirmation from the price itself — a strong up-day after the dip, or a push above a recent swing high — and you have a complete entry rule.

For the exit, trail the stop just under the rising 20 or 50 EMA, depending on how aggressive the position. When price closes through the moving average and stays there for two or three sessions, the trend is breaking and the trade is done.

One practical note. Resist the temptation to stack five moving averages on one chart. Two is usually enough; three is fine if you know what each one does.

Five lines means none of them are really driving the decision, and the chart starts to look like a plate of noodles.

The other temptation is to keep tuning the length, hunting for the perfect number. The market does not care. The 50 and 200 are not magic, but they are watched by everyone, which is part of why they work. Pick standard lengths and put your effort into the entry rules around them.