Momentum (EMA/SMA)
A moving average smooths noisy price into a single line that follows the trend — but because it averages the past, every signal it gives arrives after the move has already started.
Moving averages are the most basic momentum tools on a chart. They take a stretch of recent closing prices and blend them into one number, then plot that number over time to produce a smooth line. The point is to strip out day-to-day noise so the underlying direction is easier to see. There are two common versions, and the difference between them is just how the past prices are weighted.
SMA and EMA, defined
- A Simple Moving Average (SMA) is the plain average of the last N closing prices, with every price weighted equally. A 10-day SMA adds up the last 10 closes and divides by 10. Tomorrow, the oldest close drops off and the newest is added.
- An Exponential Moving Average (EMA) is a weighted average that gives more weight to recent prices and progressively less to older ones. Because recent prices count for more, an EMA reacts faster and lags less than an SMA of the same length.
Both are descriptions of where price has already been. Neither contains any information about the future; they reorganize the past into a smoother shape.
What traders use them for
Three uses are common. First, smoothing: the line filters out jumpy price action so a trend is easier to read. Second, slope as direction: a rising average says recent prices have been climbing, a falling one says they have been dropping. Third, crossovers: when a faster (shorter) average crosses a slower (longer) one, trend-followers treat it as a marker that the recent balance has shifted. The best-known examples use the 50-day and 200-day averages: a golden cross is the 50-day rising above the 200-day, and a death cross is the 50-day falling below the 200-day. The dramatic names do not make them reliable signals — they are simply two smoothed lines of past prices crossing.
Worked example: a 10-day SMA versus a 10-day EMA on a sudden jump
Imagine a stock sits quietly around 100 dollars for weeks, so both a 10-day SMA and a 10-day EMA read about 100. Then on a single day it jumps to 120 dollars.
The 10-day SMA treats that 120 as just one of ten equally weighted closes, so it ticks up only modestly — the jump is diluted by the nine older, lower prices. It will keep rising over the next several days as the old low prices roll out of the window one by one, but it gets there slowly. The 10-day EMA gives the newest 120 close more weight, so it lifts noticeably more on the very first day and continues to track the new level faster. Same length, same data, but the EMA turns sooner because of how it weights recent prices. That is the whole practical difference: the EMA is quicker to react, the SMA is steadier and slower.
The catch that never goes away
Moving averages are lagging by construction. Because they average the past, their signals always arrive after a move is already underway — a crossover confirms that a shift has happened, it does not call it in advance. They tend to work better in sustained, one-directional trends, where the smoothing keeps you aligned with a move that lasts. They work poorly in choppy, sideways markets, where price drifts back and forth and the averages cross each other repeatedly, producing a run of false signals known as whipsaw.
A moving average can only tell you what price has already done. The same lag that smooths out the noise — the reason the line is readable at all — is exactly what makes its signals late. You cannot keep the smoothing and remove the delay; they are the same property.
Honestly stated: moving averages and their crossovers are widely used, but the academic evidence that any single one reliably predicts returns is mixed and weak. They are best understood as a smoothed description of recent price behavior, not a system that forecasts it. For two momentum tools built on the same averaging idea, see RSI and MACD.
Educational only — not investment advice. Options involve a substantial risk of loss and are not suitable for every investor.