Moving Averages, Explained in Plain English
7 min read·Reviewed by the StockTools.ai Research Team
- ▸A moving average is just the average of the last N closing prices, recalculated every day, which turns jumpy price action into one smooth trend line.
- ▸An EMA weights recent prices more heavily than an SMA, so it turns faster — but faster also means more false turns.
- ▸The 20, 50, and 200 length settings are conventions, not laws of nature: short-term pulse, medium-term trend, and long-term regime.
- ▸Golden crosses and death crosses confirm a trend change that already happened — they are lagging signals, not predictions.
- ▸Moving averages describe the trend; they say nothing about how much you can lose, which is why they pair naturally with risk and break-even math.
What a moving average actually is
Strip away the jargon and a moving average is grade-school math. Take the last N closing prices, add them up, divide by N. Tomorrow, drop the oldest price, add the newest one, and divide again. Because the calculation window slides forward one day at a time, the result is a line that moves with price — hence the name.
A tiny worked example makes it concrete. Suppose a stock closes at 20, 21, 23, 22, and 24 dollars over five days. The 5-day simple moving average is the sum, 110, divided by 5, which is 22. If day six closes at 26, the window becomes 21, 23, 22, 24, and 26 — the 20 falls out, the 26 comes in — and the new average is 116 divided by 5, or 23.20. Price jumped 2 dollars in a day, but the average line drifted up only 1.20. That gap is the whole point.
Averaging smooths out the day-to-day noise so the underlying direction is easier to see. A stock can whip around 3 percent in a single session for no lasting reason; its 50-day average barely flinches. Traders and long-term investors alike use that smoothed line to answer one question the raw chart makes hard: is this thing generally going up, generally going down, or going nowhere?
SMA vs EMA: two ways to average
The simple moving average (SMA) treats every day in the window equally. The close from 50 days ago counts exactly as much as yesterday. That makes the SMA stable and easy to reason about, but it also means old news lingers: one big down day from weeks ago keeps dragging on the average until it finally ages out of the window.
The exponential moving average (EMA) fixes that by weighting recent prices more heavily and letting older prices fade gradually instead of dropping off a cliff. You do not need the formula to understand the behavior: when price turns, the EMA turns sooner, because the newest closes dominate the calculation. Picture two people summarizing the same season of games — one gives every game equal weight, the other cares most about the last few. The second one notices a hot streak faster.
Faster is not automatically better. Because the EMA reacts quickly, it also reacts to noise — a two-day bounce inside a downtrend can bend the EMA upward and tempt you into reading a reversal that is not there. The steadier SMA ignores more of that chatter but confirms real turns later. Neither is the correct choice; they sit at different points on the same trade-off between responsiveness and false signals.
The famous lengths: 20, 50, and 200
Open almost any charting app and you will find the same three default lengths: 20, 50, and 200 days. There is nothing magical about these numbers — 20 is roughly one month of trading days, 50 is about a quarter, and 200 is close to a year. They stuck because enough people watch them that they became shared reference points, a little like how everyone agrees rush hour starts around 5 p.m.
By convention, the 20-day tracks the short-term pulse — swing traders watch it to judge the immediate move. The 50-day describes the medium-term trend and is a common health check for a stock in an established run. The 200-day is the big-picture regime line: trading above it is loosely read as a long-term uptrend, below it as a long-term downtrend. Financial media leans on the 200-day constantly, which reinforces its status.
The honest caveat is that these are descriptions, not thresholds with physical meaning. A stock does not know where its 200-day average sits. The levels matter partly because many participants watch them, which can make reactions around them self-reinforcing for a while — but that is crowd behavior, not a property of the number 200.
Golden crosses and death crosses — what the hype leaves out
A golden cross happens when a shorter average, usually the 50-day, crosses above a longer one, usually the 200-day. A death cross is the mirror image: the 50-day crosses below the 200-day. The names alone tell you how the financial press treats them — as dramatic omens of what comes next.
Here is the framing the headlines skip: by construction, a cross can only happen after price has already moved a long way. For the 50-day to climb through the 200-day, weeks or months of rising prices have already occurred. The cross is a lagging confirmation that the trend changed a while ago, not a prediction that it will continue. Studies of past crosses show mixed results — sometimes the trend keeps going, sometimes the cross lands near the point of exhaustion, and famous signals have fired just before markets reversed the other way.
That does not make crosses useless. As a slow, hard-to-fake summary of trend direction, they help long-horizon observers cut through noise. The mistake is treating them as timing signals with predictive power. A more honest reading of a golden cross is that the market has been strong for months — which you already knew if you looked at the chart.
Support, resistance, and trend filters
In a steady uptrend, you will often see price pull back to a rising moving average, touch it, and bounce — as if the line were a floor. In downtrends the same line can act like a ceiling. Traders call this dynamic support and resistance: dynamic because, unlike a fixed price level, the line moves every day.
Part of the effect is mechanical and part is self-fulfilling. Buyers who missed an early move often wait for a pullback, and a widely watched average like the 50-day gives thousands of them the same spot to act. But the line is not a force field. Averages get sliced through all the time, and a bounce off the 50-day tells you nothing about whether the next bounce will hold.
The other common role is as a trend filter — a rule that sorts the market into environments rather than picking entries. Someone might, for example, only study long setups while price is above the 200-day average, on the logic that fighting the prevailing tide has historically been a losing habit. Used this way, the average is not saying when to act; it is saying which direction deserves the benefit of the doubt.
The chop problem — and pairing MAs with risk math
Every moving average has one well-documented failure mode: sideways markets. When price grinds back and forth in a range, it crosses its average over and over, and each cross looks momentarily like the start of a trend. Traders call this whipsaw — you get flagged in, the move fizzles, you get flagged out, and the small losses stack up. Trend-following tools need a trend; in chop they generate mostly noise.
You cannot fully engineer the problem away. Lengthen the average and you whipsaw less but react slower; shorten it and you catch turns earlier but chop yourself up more. This is the same trade-off as SMA versus EMA wearing a different hat, and it is why no single setting tests as best across all markets and eras.
This is also why moving averages pair naturally with risk arithmetic. An average can describe the trend, but it says nothing about how much a specific idea risks or what a payoff needs to look like to be worth it. If a rising 50-day sits 4 percent below the current price, a risk-reward calculator can show what a stop near that line implies about your downside versus your target. And because whipsaws mean repeated small losses, a break-even calculator shows the other cost honestly: lose 10 percent and you need about 11 percent to get back to even; lose 20 percent and you need 25 percent. Moving averages frame the trend — the risk math tells you what a mistake costs.
FAQ
Which is better, SMA or EMA?
Neither is objectively better. The EMA weights recent prices more, so it reacts to turns faster but produces more false signals; the SMA is slower and steadier. Which trade-off suits you depends on your time horizon — and long-run tests show no setting that wins in every market.
Why do traders use 20, 50, and 200 specifically?
They map loosely to a month, a quarter, and a year of trading days, and they became standard because everyone watches them. Their importance comes from shared attention, not from any mathematical property — 45 or 210 would smooth prices almost identically.
Does a golden cross predict a rally?
No. A golden cross can only form after months of rising prices, so it confirms a trend change that already happened. Historical results after crosses are mixed — some preceded further gains, others arrived near tops. It is a summary of the past, not a forecast.
What is a whipsaw?
A whipsaw is a false signal in a sideways market: price crosses its moving average as if starting a trend, the move fizzles, and price crosses back. Each whipsaw is small, but in a long choppy stretch they stack up — the classic weakness of any trend-following tool.
Can a moving average be used as a stop-loss level?
Some traders place stops just beyond a rising average, treating it as dynamic support. That is one convention, not a rule — averages get sliced through routinely. What matters more is knowing the dollar risk that stop implies, which is a risk-reward question, not a chart question.
Do moving averages work on any timeframe?
The math is identical on 5-minute, daily, or weekly charts — average the last N closes. But shorter timeframes carry more noise, so intraday averages whipsaw more. The 20/50/200 conventions and most of the research around them refer to daily charts.
Put it to work
More to learn
Educational only — not financial advice. Concepts simplified for clarity; markets are messier than definitions.