I have been observing for years how many traders ignore one of the most powerful tools in technical analysis: the RSI and especially the divergences it generates. Most only look at whether it is above 70 or below 30, but that’s not where the real magic lies.

Let’s start with the basics. The RSI (Relative Strength Index) is an oscillator that measures the relative strength between bullish and bearish closes over a given period, usually 14 candles. The interesting part is that it normalizes everything on a 0-100 scale, allowing you to see patterns that pure price does not show. When it’s above 70, we talk about overbought; below 30, oversold. But here’s the important part: these extremes do not mean that the price will reverse immediately.

I’ve seen stocks remain in overbought territory for months if buyers keep entering. What really matters is understanding what happens when the indicator exits those extreme zones and how it behaves at the mid-level (50). If the RSI oscillates between 50 and 70, we are probably in a consolidated uptrend. If it moves between 50 and 30, a downtrend. This is more useful than any point-in-time reading.

Now, what really changes the game is RSI divergence. This occurs when the price makes higher highs but the RSI makes lower highs (bearish divergence), or when the price makes lower lows but the RSI makes higher lows (bullish divergence). When you see this, the market is signaling that something is changing.

I remember analyzing Tesla between 2019 and 2022. In May 2019, the RSI was in oversold territory but started to recover while the price was forming higher lows. That was a clear buy signal. Then in 2020, multiple overbought points without the RSI dropping below its mid-zone meant corrections were opportunities to buy more, not to sell. But in October 2021, something changed. The RSI didn’t reach new highs in the overbought zone while the price was making lower highs. That RSI divergence was the early warning that the trend would break. And it did.

Trading divergence with RSI is especially powerful because the indicator acts as a leading oscillator. A bullish divergence (higher lows in the indicator with lower lows in price) within a downtrend almost always precedes a rebound. The opposite with bearish divergence in overbought during an uptrend. I saw this clearly in Broadcom a few years ago: while the price kept falling, the RSI was already forming higher lows. Two months later, the trend had reversed.

But here’s the secret many forget: RSI only gives you the necessary condition, not the sufficient one. You need to confirm with a breakout of a trendline. If the RSI exits oversold but the price doesn’t break the downtrend, it’s probably just a temporary correction. I’ve lost money ignoring this.

One technique I use is combining RSI with MACD. When RSI reaches overbought or oversold (necessary condition), I wait for the MACD to cross its midline in the opposite direction of the trend (sufficient condition). This greatly strengthens the signals. In the case of Block Inc., we had overbought RSI, then the MACD crossed downward confirming the bearish move. That allowed me to short with much more confidence.

What I’ve learned is that RSI divergence isn’t a perfect indicator, especially on very short timeframes where it generates many false signals. But on weekly or daily charts, when you see it confirmed by breaking a previous trendline, it’s one of the most reliable signals out there.

The key is not to obsess over a single indicator. RSI is just one tool in your arsenal. Combine it with trend analysis, support and resistance levels, and other oscillators. That’s how you build a trading system that truly works. RSI divergence is powerful, but only if you use it in the right context.