RSI and MACD Divergence Explained: Beginner to Pro

Transcript

Gold, four hour chart, July 2026. Price falls to a new low, right into the floor from late June. Sellers look in control. But look under the chart. Price made a lower low. RSI made a higher low. The new low came with less force behind it. That's divergence. The same candle closed above the last small high. From that close, gold climbed 88.4 points, back to where the drop started. The stop was never touched. In this lesson, you'll learn what RSI and MACD really measure, regular and hidden divergence, how to confirm it, where the stop and the target go, why divergence fails in strong trends, and how pros grade it before they trade.

Chapter one. What RSI and MACD really measure. RSI is a speedometer for price. It compares the size of the up candles with the size of the down candles, over the last 14 candles, and puts the answer on a scale from 0 to 100. Above 50, the up candles have been bigger. Below 50, the down candles have. A high reading does not mean price is too high. A fast car is not about to reverse. It's just going fast. Many traders draw lines at 70 and 30. In a range, those mark stretched prices. In a strong trend, RSI can live up there. MACD uses two runners. A quick one, the 12 candle average, and a slow one, the 26 candle average. The MACD line is the gap between them. The signal line is a smoothed copy, and the histogram shows the gap between those two lines. When the quick runner pulls away, momentum is growing. When the gap shrinks, the push is fading, even while price still moves. Here they are, under the July chart. Both are made only from closing prices. They add nothing new. They just show the speed of price in a way that's easy to compare.

Chapter two. Regular divergence: a warning that a move is tiring. Divergence means price and momentum disagree. Regular divergence shows up at the end of a move. Bearish: price makes a higher high, but RSI makes a lower high. The new high was reached with less push. Bullish is the mirror: a lower low in price, and a higher low in RSI. The rule: only compare clear swing points, two highs or two lows you can see without squinting. Then read the indicator at those same two candles. October 2025, four hour chart. Gold ran to a high, with RSI above 80. Two trading days later, a slightly higher high. But RSI was much lower. And the MACD line made a lower high too. Price said stronger. Momentum said weaker. Both indicators agreed. That's regular bearish divergence. But on its own, it's only a warning, not a trade. MACD divergence works the same way. Compare the MACD line at the same two swing points. Some traders use the histogram instead. It warns earlier, but it's noisier. Pick one, and stick to it.

Chapter three. Hidden divergence: the trend taking a breath. Hidden divergence is a different idea. It shows up inside a trend, on the pullbacks. In a downtrend, price makes a lower high, while RSI makes a higher high. The bounce had plenty of push, and still failed to reach the old high. In an uptrend, it's the mirror: a higher low in price, a lower low in RSI. 31 August 2026, one hour chart. Gold was falling hard. A bounce made a high. The next bounce made a lower high, while RSI and MACD both made higher highs. The sellers were still in charge. Price broke the last low of the fall, and kept going. 181.3 points down from that second high, by the next day. So, regular divergence warns of a turn. Hidden divergence says the trend should carry on. Many traders prefer hidden divergence, because it trades with the trend, not against it.

Chapter four. Divergence is a clue. Here's how to confirm it. Use three checks. One, a level: the divergence forms at support, resistance, or a round number. Two, a structure break: price closes through the last swing point. Three, a session: it forms while London or New York is trading, not in a quiet hour. Back to July. The level: the floor from late June, where gold had already bounced twice. The structure break: the last small high after the first low. The divergence candle closed above it. The session: that candle formed in the New York morning. Three checks out of three. Without the break, divergence is just a guess about the future. With it, price itself has started to turn.

Chapter five. Where the stop goes, and where you get out. The stop goes beyond the extreme that made the divergence. Above the second high for a sell. Below the second low for a buy. Plus a small buffer. If price goes past that point, the idea is simply wrong. Now look what that means in October 2025. By the time a four hour candle closed below the swing low, price had already fallen a long way. The stop above the high was 250.4 points away. The divergence was right. The trade was hard. Gold did fall, but it never paid two times that risk. Pros fix this by dropping a timeframe. Same idea, from the four hour chart. But the entry comes on the one hour chart, when it closed below its last small low. Same stop. Now the risk was 61.5 points. Targets: first, the swing low between the two highs. Then the next big level, the round number 4,000. Both were reached, and the stop was never touched. A simple rule: if your first target is closer than your stop, skip the trade, or find a better entry.

Chapter six. Why divergence fails in strong trends. In a strong trend, momentum stays high for a long time. RSI can sit above 70 for days. So divergence can form again and again, while price keeps going. January 2026, four hour chart. Gold made a high. Then a higher high, with lower RSI and a lower MACD. Divergence. Then another higher high, with weaker momentum again. A second divergence. Then a candle closed below the swing low. It looked confirmed. After the weekend, the stop was gone. And gold kept climbing. Another divergence formed on the way up. It didn't matter. At the peak, price was 973.3 points above that stop. What went wrong? The bigger trend was strong, and every high was a new high, with no old level above to stop it. Divergence against a trend like that is early. And early costs the same as wrong. The fix: in a strong trend, don't sell divergence at all. Trade hidden divergence with the trend, or wait for a lower high first.

Chapter seven. The pro level. Grading a divergence, start to finish. Pros grade every divergence before they trade it. Is it with the trend? At a level? Do both indicators agree? Is there a structure break, in an active session? And is there room to the target? December 2025, four hour chart. The trend: higher highs, and higher lows. Check. On 16 December, a pullback made a higher low, just under the round number 4,300. Level. Check. RSI and MACD both made lower lows. Hidden bullish divergence, on both indicators. Check. Now the one hour chart, for the entry. In the London session, a candle closed above the last small high. Structure break, in session. Check. Stop 3 points under the higher low. Target, the last high of the trend. 64.6 points of room, for 28.8 points of risk. Room. Check. Two days later, the target was hit. And the trend carried on for another week. Notice: the divergence alone didn't make this trade. The checks did. A few pro habits. Read the trend on the bigger chart, and enter on the smaller one. Skip divergence just before big news. And log every divergence that failed, not just the winners.

Chapter eight. Six divergence mistakes. One. Selling just because RSI is above 70. In a trend, that's strength, not a signal. Two. Comparing random candles. Use clear swing highs and swing lows only. Three. Entering before the structure break. That's a guess. Four. Fighting a strong trend with regular divergence. Five. A stop that's too tight. It goes beyond the extreme, or the idea is never tested. And six. Ignoring the room. A huge stop and a small target is a bad trade, even when the divergence is right.

Chapter nine. Should you trade this divergence? Any no, and you skip it. Are you comparing two clear swing points, on price and on the indicator? If not, skip. Is it with the trend, or at a strong level? If not, skip. Did price close through the last swing point? If not, wait. Is it an active session, with no big news due? If not, wait, or trade smaller. Can your stop go beyond the extreme, with room to the target? Then write down entry, stop and target. And only then, place the order.

Let's recap. RSI and MACD measure the speed of price, not whether it's too high or too low. Regular divergence warns of a turn. Hidden divergence points to the trend carrying on. Compare clear swing points only. Confirm with a level, a structure break and a session. Put the stop beyond the extreme, and check the room before you enter. And in strong trends, remember: divergence can form again and again. Next lesson: VWAP on gold. The full written guide is on xauusdglobal.com. Subscribe so you don't miss it, and come trade with us in the XAUUSD Global community. It's free during early access.