EMA vs SMA: Which Moving Average Should You Use?

The arithmetic behind simple and exponential moving averages, and what Microsoft's 2025 chart shows about lag, whipsaws and crossovers.

Contents
  1. One big day, two different reactions
  2. Microsoft in 2025: crossovers on the way down and up
  3. Four crosses in six weeks
  4. The 20-day line in April: closer, but jumpier
  5. Choosing by job, not by preference
  6. Using averages without letting them trade for you

Put a 50-day simple moving average and a 50-day exponential moving average on the same chart and they look almost identical most of the time. Then a big day comes along and one of them moves twice as far as the other. That gap is the entire EMA-versus-SMA debate, and it is easier to settle with one real day of arithmetic than with any rule of thumb.

One big day, two different reactions

Microsoft reported earnings on April 30, 2025. On May 1 it closed at 425.40, up 7.6% from 395.26. Here is how the 20-day version of each average absorbed that close:

MSFT 20-day SMA and EMA on May 1, 2025
AverageHow it updatesBefore → after
20-day SMAAdd 425.40, drop the close from 20 days ago (382.14), divide the change by 20378.03 → 380.20 (+2.16)
20-day EMAPrevious EMA + 2/21 × (425.40 − 382.74)382.74 → 386.80 (+4.06)

The SMA's move depends on two numbers: the new close and the one leaving the window. That is its quirk. An SMA can jump or sag because of what happened 20 days ago, not today. The EMA never drops anything; it just gives today a fixed share, 2 ÷ (20 + 1) = 9.5%, and lets older days fade gradually. The SMA gives today 1/20 = 5%. So on a day with a big move the EMA reacts roughly twice as much, and on quiet days the two stay close.

Microsoft in 2025: crossovers on the way down and up

MSFT daily candles with 50-day EMA, 50-day SMA and 200-day SMA, February 3 – July 15, 2025350.0400.0450.0500.0EMA 50SMA 50SMA 200SMA ✕ downlowEMA ✕ upSMA ✕ up2025-02-032025-04-232025-07-15
MSFT daily, February–July 2025. The SMA 50/200 cross down came on Feb 25; the closing low was 354.56 on Apr 8; the EMA 50/200 cross up came on May 19 and the SMA version on Jun 10.

The 50-day crossing the 200-day is the most famous moving-average signal: a "golden cross" when the 50 moves above, a "death cross" when it moves below. Microsoft gave both in 2025, in both EMA and SMA form:

MSFT 50/200-day crossovers in 2025
SignalDateClose that day
SMA 50 below SMA 200Feb 25397.90
EMA 50 below EMA 200Mar 3388.49
Closing lowApr 8354.56
EMA 50 above EMA 200May 19458.87
SMA 50 above SMA 200Jun 10470.92

On the way up, the EMA version fired 15 sessions earlier and 12.05 lower than the SMA version. Both still arrived late: the EMA golden cross came 29.4% above the closing low, the SMA one 32.8% above it. On the way down, the order flipped. The SMA death cross came four sessions before the EMA one. That is the drop-off effect again: an SMA's path also depends on which old closes are leaving its window, so it can turn before or after the EMA depending on what happened weeks earlier.

So "EMA is faster" is true on average, not every time. What is always true is that both are late, because they are averages of the past.

Four crosses in six weeks

Crossovers are most misleading when a stock goes sideways and the two averages run almost level. In late 2024 Microsoft's 50-day and 200-day SMAs ran almost level and crossed four times:

  • November 1: death cross at 410.37
  • November 13: golden cross at 425.20
  • November 25: death cross at 418.79
  • December 13: golden cross at 447.27

Anyone treating each cross as an instruction would have flipped four times in six weeks. The chart itself was telling a simpler story: a flat range. When two averages are tangled together, a cross carries almost no information. When they are clearly apart and sloping, it carries more.

The 20-day line in April: closer, but jumpier

Short EMAs hug price, which is useful and irritating in equal measure. On April 9, 2025, a day of broad market relief, MSFT jumped to 390.49 and closed above both its 20-day SMA and 20-day EMA. The next day it closed at 381.35, below the SMA (381.56) but still above the EMA (380.23). By April 16 it was below both again at 371.61, and the actual closing low of the move had already been set on April 8. The EMA kept the "above" reading one day longer; it did not make the signal any more reliable.

Choosing by job, not by preference

When each average tends to fit
If you want to…Lean towardWhy
React quickly to a fresh breakout or gapEMA (9, 20, 21)Heavier weight on the latest closes
Map the big trend with fewer wigglesSMA (50, 200)Smoother, and it is what most charts and news reports quote
Trail a stop in a strong trendEMA (20 or 50)Tightens sooner when momentum stalls
Compare with what other traders watchSMA 50 / 200Widely published levels attract attention

Whichever you pick, stick with it long enough to learn its habits on the markets you watch. Switching between EMA and SMA until one "works" on a past chart is curve-fitting, not analysis.

Using averages without letting them trade for you

  • Read the slope first. A rising 50-day with price above it describes an uptrend; a flat one describes a range, whatever the crossovers say.
  • Treat the average as a zone. Price often overshoots a 50-day line by a percent or two before turning back.
  • Check what else agrees. A pullback to a rising average that is also prior support is a stronger setup than the average alone.
  • Remember that MACD is built from two EMAs. If you already watch it, a separate EMA crossover adds little.

If you want a feel for the lag, hide the next candle and ask whether it will close above or below the 20-day EMA before you see it. Doing that a few dozen times on real historical charts, as the One Candle Ahead simulator lets you, teaches the rhythm faster than reading about it.

Questions readers ask

Is EMA or SMA better for beginners?

Start with the 50-day and 200-day SMA to read the big trend, since they are the lines most charts and news reports quote. Add a 20-day EMA when you want a line that reacts to fresh moves.

Does the golden cross work?

It confirms a trend change late. On Microsoft in 2025 the SMA golden cross came 32.8% above the closing low, and in late 2024 the same pair of averages crossed four times in six weeks during a sideways stretch.

Why does an SMA sometimes move when price barely changes?

Because an SMA also drops the oldest close in its window. If a very high or low close leaves the window, the average moves even on a quiet day.

Sources

Charts and worked examples use daily prices from the One Candle Ahead historical dataset (split-adjusted). Past price behaviour does not predict future results.

Written and maintained by the Studio Solum Editorial Team. How we write and review guides