🎯 In 30 seconds: A moving average is most useful when you give it one clear job. A long average and its slope make a decent direction filter. A shorter one gives a reference point for pullbacks. Crossovers catch trends late and lose small amounts in ranges, so they need a filter. The distance between price and an average shows how stretched a move is. A ribbon of several averages shows at a glance whether the market is trending or tangled.
Give each average one job
Charts covered in averages are hard to read, and each line tempts you to find a reason to trade. Decide what you need first, then pick the average that does it.
| Job | A common choice | What you read |
|---|---|---|
| Direction filter | A long simple average, such as 150 or 200 days | Its slope and which side price is on |
| Pullback reference | A 20-day exponential average | Where dips have tended to stall |
| Signal | Two averages, fast and slow | When one crosses the other |
| Stretch gauge | Distance from a 50-day average, in percent | How far price has run from its recent norm |
| Exit | A close below a short average | When to step aside from a trend |
The basic math of simple and exponential averages is covered in the Moving averages chapter of this guide. This chapter is about using them.
Read the slope, not only the side
“Price is above the average” says less than it sounds. Price can sit above a falling average for weeks during a bounce. The slope tells you what the average itself is doing.
A simple slope measure: the percentage change of the average over the last 10 bars. Hypothetical example: a 50-day average is $102.00 today and was $100.00 ten bars ago. Its slope is 2.0% per 10 bars. You might call anything between minus 0.5% and plus 0.5% flat.
A filter that needs both conditions, price above the average and the average rising, removes many of the false starts that happen when the average is flat.
Pullbacks to a rising average
In a steady uptrend, dips often stall near a short or medium average. That gives a structured way to join a trend without chasing it.
- The 50-day average is rising, and price is above it.
- Price pulls back and touches or dips just below the 20-day EMA.
- Entry trigger: the first close above the prior bar’s high after the touch.
- Stop: just below the lowest low of the pullback.
Hypothetical numbers: the trigger bar closes at $52.40 and the pullback low is $50.90. With the stop at $50.80, the risk is $1.60 per share. If you risk $80 on the idea, that is 50 shares. The trade plan chapter covers sizing in full.
The setup fails often. Some pullbacks keep going through the average and turn into a new downtrend. The stop is what makes those failures small.
Crossovers and the whipsaw cost
A crossover gives a clean, testable signal. Its weakness shows up when the market goes sideways: the averages twist around each other and each cross reverses soon after.
In this hypothetical chart the sideways stretch produced 3 crosses, and the trend produced 1. Every one of the range signals would have cost a small loss plus trading costs. A crossover system depends on the occasional long trend to pay for those.
Ways people try to reduce the damage, each with a cost:
- Require a slope. Act only when the slower average is also rising or falling. You miss the earliest part of new trends.
- Add a buffer. Require the fast average to clear the slow one by a set percentage. Signals arrive later.
- Check trend strength. Skip signals when a trend strength reading is low. The trend strength chapter covers ADX and choppiness, which are built for this.
How stretched is price?
Distance from an average measures how far a move has run. It is easy to compute:
Distance = (price minus average) / average x 100
Hypothetical example: price is $118.80 and the 50-day average is $108.00. Distance is $10.80 / $108.00 x 100 = 10%. Whether that is a lot depends on the stock. If this stock has been more than 10% above its 50-day average only a handful of times in five years, it is unusually stretched. For a fast-moving stock, 10% may be ordinary.
Stretched does not mean about to fall. Strong trends can stay stretched for weeks. Many traders use the reading to avoid buying after a big run, rather than as a reason to sell.
Ribbons: many averages at once
A ribbon plots several averages of steadily increasing length, such as 10, 20, 30, 40, 50 and 60 days. You do not read any single line. You read the shape of the whole group.
- Spread and in order (shortest on top in an uptrend): the trend is steady.
- Narrowing: the trend is slowing.
- Tangled: no trend. Most signals here are noise.
- Reordered the other way: a new trend has had time to show itself, which also means much of the turn is already behind you.
Other averages in the chart tool
| Average | In one line |
|---|---|
| WMA | Weights recent bars more, in a straight-line pattern |
| Hull | Combines weighted averages of different lengths to cut lag, at the cost of more overshoot |
| DEMA and TEMA | Combine an EMA with EMAs of itself to reduce lag further |
| KAMA | Kaufman’s adaptive average, which moves faster when price travels efficiently and slows in choppy stretches |
| VWAP | The average price weighted by volume since a starting point, often the session open |
Faster averages turn sooner and also turn wrongly more often. No version escapes that trade-off. Pick one and learn how it behaves on the stocks you follow.
What the evidence suggests
Moving average rules are among the most studied ideas in this field. Brock, Lakonishok and LeBaron (1992) found informative signals in a long history of U.S. index data. Sullivan, Timmermann and White (1999) showed the edge looked much weaker after accounting for how many rules had been searched and in later years. Park and Irwin (2007) found many favorable results in the literature along with frequent testing weaknesses. Treat any single average or crossover as a tool for structure and risk, and test it with costs before trusting it.
- A 50-day average was $80.00 ten bars ago and is $79.80 today. Is it rising, falling or flat by this chapter’s measure? Answer: Flat. The change is minus 0.25%, inside the minus 0.5% to plus 0.5% band.
- Price is $66 and its 50-day average is $60. What is the distance? Answer: $6 / $60 x 100 = 10% above.
- Why do crossover systems lose money in ranges? Answer: The averages twist around each other, so each signal reverses soon after, and each reversal costs a small loss plus costs.
- What does a tangled ribbon tell you? Answer: There is no clear trend, so most signals in that stretch are noise.
Sources
- Gerald Appel, Technical Analysis: Power Tools for Active Investors (2005)
- Perry J. Kaufman, Trading Systems and Methods (for the adaptive average)
- John J. Murphy, Technical Analysis of the Financial Markets (1999)
- Brock, Lakonishok and LeBaron (1992), “Simple Technical Trading Rules and the Stochastic Properties of Stock Returns,” Journal of Finance
- Sullivan, Timmermann and White (1999), “Data-Snooping, Technical Trading Rule Performance, and the Bootstrap,” Journal of Finance
- Park and Irwin (2007), “What Do We Know About the Profitability of Technical Analysis?” Journal of Economic Surveys
- Moneydesks Technical Analysis tool (moving averages, distance from MA, Chart events)