Indicators That Earn Their Place
The simplest useful indicator, and the one most often misused.
intermediate · 3 min read · 10 XP
A moving average is the average closing price over the last N periods, redrawn every bar. That is the whole idea — and understanding exactly what that does, and does not, tell you is the point of this lesson.
A 20-period moving average adds up the last 20 closes and divides by 20. Next bar, it drops the oldest close and adds the newest. The line you see is that calculation repeated across the chart.
Two variants matter:
Neither is better. An EMA responds sooner to a genuine change and also sooner to a false one. That is the entire trade-off, and no choice of period escapes it.
What a moving average is *for*
It answers one question well: which way has price been going, on average, recently? It does not know about levels, news, or what happens next. Every reasonable use follows from that one capability.
New traders complain that moving averages "lag". They do — necessarily. An average of the last 20 closes cannot respond to the 21st before it happens.
Lag is the price of smoothing. You are trading responsiveness for reliability. A 200-period MA is slow and almost never gives a false trend reading; a 5-period MA is fast and gives them constantly. Choosing a period is choosing where on that spectrum you want to sit, not finding a setting that avoids the trade-off.
1. Trend direction. Is the MA sloping up or down? Is price above or below it? On a daily chart, price above a rising 50-period MA is a reasonable, mechanical definition of "uptrend" — not because the line is magic, but because it summarises the last fifty days in one glance.
2. Dynamic support and resistance. In a strong trend, pullbacks often stall around a well-watched MA (the 20 and 50 EMA especially). This is partly self-fulfilling — enough traders watch those lines to create orders there — and self-fulfilling is still tradeable.
3. Volatility and regime. When price crosses a moving average constantly, you are in a range. When it stays on one side for weeks, you are in a trend. The number of crossings is a cheap regime detector.
The classic beginner system is "buy when the fast MA crosses above the slow MA, sell when it crosses below". It is the first thing everybody tries and it disappoints reliably. Here is why.
A crossover is a lagging confirmation of a move that already happened. In a strong trend it enters late but profitably. In a range — where markets spend most of their time — it produces a stream of crossings in both directions, each one entering just as price turns back. The wins in the trending 20% rarely cover the losses in the ranging 80%.
The optimisation trap
The natural response is to search for better periods: "9 and 21 didn't work, let me try 11 and 34." On historical data you will find a pair that looks excellent. That pair is fitted to noise that will not repeat. Curve-fitting is covered properly in the Strategy path — for now, treat any parameter you found by searching for the best backtest as guilty until proven innocent.
Not useless — insufficient. A crossover is fine as one condition inside a system that also knows whether the market is trending, where the levels are, and how much to risk. It fails as a whole strategy because it answers only one question and pretends it answered all of them.
What to remember
A moving average averages the last N closes and answers one question: which way has price been going recently. Its lag is the necessary cost of smoothing. It works as a trend filter, a dynamic level and a regime detector — and fails as a standalone crossover system, because markets range most of the time.