If you've spent any time looking at charts, you've met RSI and MACD. They're the two indicators that get bolted onto almost every trading screen, and the pitch is seductive: a number tells you when a stock is "overbought" or "oversold," and a crossover tells you when momentum is turning. The real question most people are actually asking is simpler — do these things tell me when to buy and sell, or do they just look like they do?
The honest answer is: they measure something real, they're useful inside the right conditions, and they will hurt you badly outside those conditions. This guide explains how each one works, what momentum and mean-reversion actually mean, how divergence is supposed to warn you — and the one failure that quietly wrecks more accounts than any other: in a strong trend, these oscillators stay pinned, and the textbook rule sells you out far too early.
What RSI actually measures
The Relative Strength Index (RSI) is an oscillator that compares the size of recent gains to the size of recent losses and squeezes the result onto a scale from 0 to 100. The most common setting is a 14-period lookback. The intuition: when price has been rising hard with few down days, RSI climbs toward 100; when selling dominates, it falls toward 0.
The classic thresholds are 70 for overbought and 30 for oversold. The textbook reading is that above 70 a market has run too far, too fast and is due to pull back, while below 30 it's been beaten down and is due to bounce. Notice the assumption baked into that sentence — it assumes price reverts. That assumption is the whole ballgame, and it's only true some of the time.
RSI is best understood as a measure of how lopsided recent price action has been, not as a verdict on value. A reading of 75 doesn't mean "expensive." It means "the buyers have dominated lately." Whether that's a warning or a green light depends entirely on the kind of market you're in.
How MACD works
MACD (Moving Average Convergence Divergence) is built from moving averages, so it leans on the same foundation covered in our guide to moving averages. It has three parts:
- The MACD line — the difference between a fast exponential moving average (typically 12 periods) and a slow one (typically 26). When the fast average pulls away above the slow one, momentum is building to the upside.
- The signal line — a 9-period average of the MACD line itself, used as a smoothed trigger.
- The histogram — the gap between the MACD line and the signal line, drawn as bars. It's the rate of change of momentum: bars growing means momentum accelerating, bars shrinking means it's fading.
The headline signal is the signal-line crossover: when the MACD line crosses above its signal line, that's read as bullish; crossing below is bearish. The histogram gives you an early heads-up — it flips direction before the lines actually cross, because a shrinking histogram means the two lines are converging. Many traders watch the histogram peaking or troughing as the first hint that a move is losing steam.
A second, slower signal is the zero line. When MACD is above zero, the fast average is above the slow average — the market is in an uptrend by that measure. Below zero, a downtrend. Crossovers that happen on the "right" side of zero (bullish crosses above zero, bearish crosses below) tend to be more reliable than ones fighting the larger trend.
Momentum versus mean-reversion
Here's the conceptual fork that decides whether either indicator helps you. There are two opposing ways to read a strong move:
- Mean-reversion says an extreme reading is a stretched rubber band — price has overshot and will snap back. Buy oversold, sell overbought.
- Momentum says a strong move is evidence of more strength — things in motion tend to stay in motion. A high RSI is confirmation the trend is healthy, not a sell.
The same RSI reading of 75 is a sell signal to a mean-reversion trader and a confirmation to a momentum trader. The indicator hasn't changed — the regime has. Knowing which one you're in matters more than the number itself.
This is why traders argue endlessly about these tools and both sides can show you winning examples. RSI's overbought/oversold logic is fundamentally a mean-reversion bet. MACD's crossover logic is fundamentally a momentum bet. They can flatly contradict each other in the same window — and that's not a malfunction, it's the two philosophies disagreeing about what the price is telling you.
Divergence: the warning sign worth knowing
Divergence is the most genuinely useful pattern these indicators produce, because it looks at the relationship between price and momentum rather than a single threshold.
Bearish divergence happens when price makes a higher high but the indicator makes a lower high. Price is still climbing, but with less force behind each push — fewer buyers, smaller gains. It can hint that an uptrend is tiring. Bullish divergence is the mirror image: price makes a lower low while the indicator makes a higher low, suggesting selling pressure is drying up even as price drips lower.
Both RSI and MACD can show divergence, and it's often more reliable than a raw overbought/oversold reading because it requires two things to disagree in a specific way. That said, treat divergence as a yellow flag, not a trade trigger. Markets can diverge for a long time before anything happens — a trend can lose momentum and still grind higher for weeks. Divergence tells you a move is getting tired, not that it's over. It pairs well with confirmation from other evidence, which is one reason traders cross-check momentum against volume analysis before acting.
Combining the two without fooling yourself
RSI and MACD are often used together, and the logic is reasonable: RSI gauges whether a move is stretched, MACD gauges the direction and acceleration of momentum. A common approach is to use MACD to define the trend's direction and only take RSI signals that agree with it — for example, treating RSI dips toward oversold as buying opportunities only while MACD says the larger trend is up.
But be honest with yourself about what you're doing. Stacking two indicators that both ultimately derive from price doesn't give you two independent opinions. They share the same input. When you add a third and a fourth indicator until the chart finally "agrees," you're usually just curve-fitting your way into a decision you'd already made. More indicators feel like more confirmation; often they're the same information wearing different hats.
The best way to build intuition for this is to stop reading about it and watch it move. The RSI Oscillator Playground lets you see for yourself how the readings behave across calm ranges and violent trends — push the settings around and watch how often "overbought" is a warning versus a green light.
Where it falls short
This is the part most tutorials skip, and it's the part that actually costs people money.
In strong trends, oscillators stay pinned. This is the cardinal failure. In a powerful uptrend, RSI can hit 70 and then simply stay above 70 for an extended stretch while price keeps climbing. Traders call it "overbought stays overbought." If you mechanically sell every time RSI crosses 70, you'll exit near the start of the best moves and watch them run without you. The same trap works in reverse during crashes: RSI parks below 30 and price keeps falling, so "oversold" buyers catch a falling knife again and again.
The reason is structural. Overbought/oversold logic is a mean-reversion tool, and trends are the one environment where mean-reversion doesn't apply. These indicators suit ranges, not trends. In a sideways, choppy market, RSI's bounces off 30 and rejections at 70 can work nicely because price genuinely is oscillating around a center. Point that same logic at a runaway trend and it fires sell signal after sell signal into strength.
A few more limits worth holding in mind:
- Crossovers lag. MACD is built from moving averages, so its signals arrive after a move is underway. In fast reversals you get whipsawed — a bullish cross right before price rolls over, then a bearish cross near the bottom.
- Whipsaws in chop. When a market goes flat and indecisive, MACD crosses back and forth across its signal line, generating a stream of conflicting trades that bleed you on fees and frustration.
- Thresholds aren't universal. 70/30 is a convention, not a law. Volatile names blow through those levels constantly; sleepy ones rarely reach them. A fixed threshold means different things on different instruments.
- Divergence can persist. As noted, momentum can fade long before price does. Acting on the first divergence is a reliable way to fight a trend that isn't finished.
None of this means RSI and MACD are useless. It means they're conditional tools. The skill isn't reading the indicator — it's recognizing whether you're in a range (where the textbook rules help) or a trend (where they actively hurt). Rather than trust a number you read somewhere, generate your own and watch how it behaves on the assumptions you actually care about in the RSI Oscillator Playground.
Frequently Asked Questions
Is RSI or MACD better for timing trades?
Neither is "better" — they answer different questions and suit different conditions. RSI's overbought/oversold readings are a mean-reversion tool that works best in ranging, sideways markets. MACD's crossovers are a momentum and trend tool that holds up better when a market is actually moving. Using them together can help, but only if you accept that they share the same price input and won't always agree.
Why does my RSI keep saying "overbought" while the stock keeps rising?
Because you're in a strong trend, and that's exactly where overbought/oversold logic breaks down. In a powerful move, RSI can stay above 70 for a long time while price climbs — "overbought stays overbought." It isn't a malfunction; it's the indicator being used in the one environment it isn't built for. In trends, a high RSI is closer to confirmation of strength than a sell signal.
What is divergence and should I trade on it?
Divergence is when price and the indicator disagree — price makes a new high but RSI or MACD makes a lower high (bearish), or price makes a new low but the indicator makes a higher low (bullish). It can flag a move losing momentum. Treat it as a warning to pay attention, not as a trigger by itself, because markets can diverge for a long time before anything changes. Look for confirmation before acting.
What settings should I use for RSI and MACD?
The standard settings — 14-period RSI, and 12/26/9 for MACD — are the defaults for a reason: they're widely watched, so a lot of traders are reacting to the same levels. There's nothing magic about them, and shorter settings react faster but generate more false signals while longer ones lag more. Rather than hunt for a "perfect" setting, it's more useful to test how a given configuration behaves across different market conditions yourself before relying on it.