Oscillators: RSI and MACD in Practice

RSI and MACD are the two indicators every platform ships and every beginner misreads in the same two ways: treating an extreme reading as an instruction, and treating divergence as a prediction. Both come from not knowing what the calculation actually does — which takes about five minutes to fix.

What RSI measures

Relative Strength Index compares the size of recent up-closes to the size of recent down-closes, and expresses it on a 0 to 100 scale.

Read it literally and the standard interpretations come apart:

• RSI 70 means recent up-moves have been substantially larger than recent down-moves. That is a description of a strong uptrend.
• RSI 30 means the reverse — a strong downtrend.
• RSI 50 means up and down moves have been about equal in size.

So a high RSI is evidence of strength, not of exhaustion. The word "overbought" imports a conclusion the arithmetic does not contain, and it is the most expensive word in retail technical analysis.

What MACD measures

MACD is the difference between two exponential moving averages, usually 12 and 26 periods, with a 9-period average of that difference drawn on top as the signal line.

It is two moving averages, subtracted. Every caveat from the previous lesson therefore applies to it in full: it lags, it whipsaws sideways, and its crossovers arrive after the move that caused them.

The histogram is the more useful half, because it changes before the crossover does. That makes it earlier and correspondingly less reliable — the same trade-off in a new place.

Overbought is not a sell

In a strong trend RSI can sit above 70 for weeks, and every day of that is a day when the "overbought" reading was a sell signal that lost money.

This is not a flaw in the indicator. A sustained trend is a period where up-moves exceed down-moves, so RSI reporting a high number is RSI working correctly. The error is the translation from "strong" to "about to stop".

What to do instead

1. Use the extremes as classification. RSI persistently above 60 is a trending market; oscillating between 40 and 60 is a range. That is a condition, and lesson 4 already told you what conditions are for.
2. Change the thresholds with the condition. In an uptrend, pullbacks often bottom near RSI 40 rather than 30. Waiting for 30 in a strong trend means waiting for a pullback that never comes.
3. Require a level. An extreme reading at a resistance zone you drew beforehand is a reason to look. The same reading in open space is not.
4. Never fade an extreme on its own. "RSI is 78" is not a trade. It is a fact about the last fourteen candles.

Divergence and its failure rate

Divergence is price making a higher high while the oscillator makes a lower high, or the mirror at lows. The story is that momentum is fading beneath a continuing move.

The story is sometimes true. The problem is what the observation is worth on its own, and there are three structural reasons it is worth less than it looks.

• It is arithmetic, not prophecy. A second high reached more slowly than the first must produce a lower oscillator reading. Divergence often reports nothing except that the move decelerated.
• It repeats. A trend can diverge three or four times and keep going. Each instance looks identical to the one that eventually marks the turn.
• It is only visible afterwards. Divergence needs the second peak to be complete, and while price is making that peak you cannot know whether it is the second one or the third.

Using it as a question

Divergence earns its place as a filter on a trade you already had a reason for — never as the reason itself.

Used that way an oscillator does the only job it is equipped for: describing how the recent move compares with the moves before it. Everything else people ask of it is asking a fourteen-period average to know the future.