SMA vs EMA: Which Moving Average Suits Your Trading
A simple moving average gives every close in its window the same weight. An exponential moving average gives recent closes more weight, so it turns faster. That single difference explains everything else about how the two behave.
Neither is better in general. The EMA reacts sooner and therefore signals sooner, but it also reacts to noise and produces more false signals. The SMA is slower and steadier. Which one suits you depends on your timeframe and how many whipsaws you can tolerate.
Both formulas, precisely
The SMA is an arithmetic mean of the last n closes.
SMA = (C1 + C2 + … + Cn) divided by n
Each close carries a weight of 1 divided by n. A close that drops out of the window stops affecting the value completely, which is why an SMA can move purely because an old bar left the calculation.
The EMA is recursive. Each new value is a blend of today’s close and yesterday’s EMA.
EMA today = (Close today x k) + (EMA yesterday x (1 minus k)), where k = 2 divided by (n + 1)
For n equal to 10, k is 0.1818. For n equal to 20 it is 0.0952, and for 50 it is 0.0392. The first EMA value is usually seeded with an SMA of the first n closes.
One consequence of the recursion is that no close ever fully leaves an EMA. Its influence shrinks daily but never reaches zero, so the EMA carries a long tail of old data at tiny weight.
A worked comparison of a price shock
Take an illustrative stock, not real data. It trades flat at Rs 1,000 for a long stretch, then gaps to Rs 1,100 on a results day and stays there. Both averages use n equal to 10.
| Session | Close | SMA(10) | EMA(10) |
|---|---|---|---|
| Before the gap | Rs 1,000 | Rs 1,000.0 | Rs 1,000.0 |
| Day 1 after | Rs 1,100 | Rs 1,010.0 | Rs 1,018.2 |
| Day 2 | Rs 1,100 | Rs 1,020.0 | Rs 1,033.1 |
| Day 3 | Rs 1,100 | Rs 1,030.0 | Rs 1,045.2 |
| Day 5 | Rs 1,100 | Rs 1,050.0 | Rs 1,063.3 |
| Day 10 | Rs 1,100 | Rs 1,100.0 | Rs 1,086.6 |
Read the table carefully, because it corrects a common misconception. The EMA moves faster at first, sitting Rs 8 above the SMA after one day and Rs 13 above after five. But by day 10 the SMA has fully absorbed the gap and reads Rs 1,100 exactly, while the EMA is still at Rs 1,086.6 and will approach Rs 1,100 without ever quite reaching it. The EMA is quicker to respond and slower to finish.
The lag versus whipsaw trade-off
Every moving average is a compromise between responding late and responding to nothing.
| Feature | SMA | EMA |
|---|---|---|
| Weighting | Equal across the window | Heavier on recent closes |
| Reaction to a shock | Slower and steady | Immediate, then tapers |
| False signals in a range | Fewer | More |
| Behaviour when old data drops out | Can jump for no current reason | Smooth, no drop-out effect |
| Common uses | Long-term levels, 50 and 200 day | Faster systems, MACD, TRIX |
In practice, longer lookbacks reduce the gap between the two. On a 200-period average the difference in daily values is often smaller than a couple of ticks, and the tick size on NSE equities is Rs 0.05, so the choice barely matters. On a 9-period average on a 5-minute Bank Nifty chart, the choice changes your signals materially.
Choosing between the two
- Intraday or short swing trading, where a day of lag is expensive: an EMA usually suits better.
- Position or long-term trend following, where you want fewer decisions: an SMA is easier to live with.
- Widely watched reference levels, such as the 200-day average that market commentary quotes: use the SMA, because that is the convention.
- Inside another indicator, follow whatever the indicator specifies. MACD and TRIX are defined with EMAs.
Whichever you choose, test it on the instrument you actually trade. A setting that works on Nifty 50 futures may whipsaw badly on a midcap stock with wider spreads.
Frequently Asked Questions
Is a 50-day EMA the same as a 50-day SMA with a different name?
No. Both use 50 in the name, but the EMA applies a smoothing factor of about 0.039 and includes data from before the 50-day window at small weight. The two lines can differ meaningfully after a sharp move.
Does the EMA give better returns in backtests?
Sometimes, and usually by a margin small enough to be explained by the specific period tested. Results flip between instruments and date ranges, which is the honest sign that neither has a general advantage.
What about the weighted moving average?
A WMA weights closes linearly, so the newest close gets weight n, the next n minus 1, and so on. It sits between the SMA and EMA in responsiveness and adds little that either does not already offer.
Which is better for spotting a crossover?
EMA crossovers trigger earlier, which helps in trends and hurts in ranges. If you use crossovers, choose based on how often you can accept being stopped out rather than on which one looks better on last year’s chart.
Key Takeaways
- SMA weights every close equally. EMA uses a smoothing factor of 2 divided by (n plus 1), so recent closes count more.
- After a price shock the EMA moves first, but the SMA fully reflects the new level once the old bars leave the window.
- Faster response means earlier signals and more whipsaws. That trade-off cannot be removed, only chosen.
- Neither average is superior in general, and the difference shrinks as the lookback lengthens.
- The worked figures above are illustrative and rounded, meant to show behaviour rather than real prices.
