Momentum: RSI and MACD, used honestly

Momentum indicators try to answer a slightly different question from moving averages. Instead of 'which way is the trend?' they ask 'how forcefully is price moving, and is that force fading?' The two most popular — the RSI and the MACD — are worth understanding precisely, because they're also the two most widely misused. Most of the damage comes not from the math but from the words people attach to the math, especially 'overbought' and 'oversold'.

What the RSI actually measures

The Relative Strength Index (RSI) is a number between 0 and 100 that compares the size of a stock's recent gains to the size of its recent losses, over a chosen window (14 periods is the common default). Crudely: over the last 14 days, were the up-moves collectively bigger than the down-moves, and by how much? If gains dominated, the RSI is high; if losses dominated, it's low; if they roughly balanced, it sits near 50. That's all it is — a normalised ratio of average gains to average losses, squeezed onto a 0-to-100 scale. It does not measure a stock's strength against other stocks (despite the confusing name), it does not know anything about the company, and it is calculated entirely from the very price series you're already looking at. A high RSI literally means 'price has risen a lot relative to how much it's fallen, lately'. Read it that way and you'll never be fooled by it; read it as a buy/sell button and you will be.

Why 'overbought' and 'oversold' mislead in a trend

By convention, an RSI above 70 is labelled overbought and below 30 oversold, with the dangerous implication that overbought means 'due to fall' and oversold means 'due to rise'. In a sideways, range-bound market this can loosely hold — price oscillates, and extremes tend to revert. But in a strong trend, these labels are actively harmful, and this is the single most important thing to understand about the RSI. In a powerful uptrend, a stock can be 'overbought' — RSI above 70 — for weeks while it keeps climbing. The RSI is high because the trend is strong; that's not a sell signal, it's a description of strength. Traders who reflexively short every 'overbought' reading in a bull run get repeatedly run over, because they've mistaken a thermometer for a forecast. The thermometer says it's hot. It does not say it's about to cool. The same applies in reverse: stocks in brutal downtrends stay 'oversold' for long, painful stretches.

Divergence: the one RSI use with real teeth

The most defensible use of the RSI is spotting divergence — when price and the indicator disagree. If a stock makes a new high in price but the RSI makes a lower high than it did at the previous price peak, that's bearish divergence: price pushed higher, but with less internal momentum behind it than before. The buying force is weakening even as the price keeps rising. The mirror case — price makes a lower low but RSI makes a higher low — is bullish divergence, hinting that selling pressure is exhausting. Divergence is more interesting than the overbought/oversold lines because it's about the rate of change of momentum rather than its level, and a fading rate of change genuinely can precede a turn. But be honest about its limits: divergence can persist for a long time before anything happens, a strong trend can simply absorb it, and you'll only ever know which divergences 'worked' in hindsight. Use it as one whisper of caution among several inputs, never as a standalone trigger that overrides the actual trend.

MACD: two moving averages in a trench coat

The MACD (Moving Average Convergence Divergence) sounds exotic but is, underneath, just moving averages again. The core MACD line is the difference between two EMAs of price — typically a fast 12-period EMA minus a slow 26-period EMA. When the fast average is above the slow one, the MACD line is positive; when below, negative. A signal line (commonly a 9-period EMA of the MACD line itself) is plotted on top, and the histogram shows the gap between the two. So everything the MACD 'tells' you is a restatement of moving-average behaviour. The MACD crossing above its signal line is essentially a faster, repackaged version of a short average crossing a longer one — a momentum-flavoured crossover. The MACD line crossing zero just means the fast and slow EMAs themselves crossed. The histogram growing means the two averages are spreading apart (momentum building); shrinking means they're converging (momentum fading). Once you see that the MACD is two moving averages and their difference, its signals stop looking mysterious and start looking like exactly what they are.

• MACD line above signal line — short-term momentum turning up relative to the trend; a faster cousin of a moving-average crossover.
• MACD crossing the zero line — the fast and slow EMAs have crossed; a slower, more significant shift.
• Histogram expanding — the averages are diverging, momentum is accelerating.
• Histogram contracting — the averages are converging, momentum is fading even if price still rises.
• MACD divergence vs price — like RSI divergence: momentum disagreeing with price, a whisper of a possible turn.

Everything here is a derivative of price

Step back and notice the pattern that unites this entire module. The RSI is computed from price. The MACD is computed from price. Moving averages are computed from price. Bollinger Bands, next chapter, are computed from price. Every one of these indicators is a mathematical transformation of the same price series — a derivative of price, in both the calculus sense and the everyday sense of 'downstream from'. None of them contains information that isn't already in the price chart; they only rearrange it to make certain features easier to see. This has a hard consequence people resist: an indicator can never lead price, because it's made of price. It can summarise, smooth, and reframe — and that's valuable — but it arrives a step behind by construction. Stacking five indicators does not give you five independent opinions; it gives you five rephrasings of the one thing you already have. They agree with each other and with price most of the time precisely because they're all the same data wearing different hats. The illusion of confirmation from 'multiple indicators all agreeing' is one of the most seductive errors in technical analysis.

How to actually use momentum

Used honestly, momentum indicators have a modest, real role. They help you gauge whether a move has force behind it or is running on fumes. They flag divergences worth a second look. They give mechanical, repeatable triggers you can write into a rule rather than eyeballing. What they cannot do is tell you the future, override a strong trend, or substitute for risk control. Treat the RSI and MACD as descriptive instruments — momentum thermometers — and they'll serve you. Treat them as oracles and they'll quietly bankrupt you while you congratulate yourself on your 'system'.