The 70/30 lines get treated like tripwires. Cross above 70 and half the traders on a Lagos WhatsApp group call it overbought and start hunting for the top. Cross below 30 and someone's calling it oversold and backing up the truck. Both reactions skip the part where you ask why the number got there in the first place. A stock can hit 70 because it's putting in a genuine breakout on real demand, and it can sit above 70 for weeks while it keeps running. If you sold the first time it crossed that line, you left the entire move on the table for someone with a plan and took a small profit or a loss instead.
Overbought does not mean expensive, and it does not mean due for a fall. It means the recent gains have been large relative to the recent losses over whatever window you're measuring. That's a description of momentum, not a verdict on value. A stock in a strong uptrend will spend long stretches overbought, because that's what strong uptrends look like on a momentum indicator. The 70 line isn't a ceiling the price bounces off. It's a threshold you crossed because buyers showed up in size, and buyers showing up in size is usually a reason to pay attention, not a reason to sell into it.
Same logic in reverse. A stock grinding lower can sit under 30 for a long time in a real downtrend, and every bounce off that line looks like a bottom until it isn't. Treating 30 as an automatic buy signal means you keep catching a falling stock on the way down, averaging into weakness because a number told you to, while the trend tells you something different and gets ignored because it's less exciting than a clean signal.
What RSI is actually good for is spotting divergence: when price makes a new high but the momentum behind it doesn't. A stock pushes to a fresh high, RSI makes a lower high than it did on the previous peak. That's the move losing steam even while the price still looks strong on the surface. It's not a sell signal on its own, but it's a flag: the buying that's driving this last leg up is weaker than the buying that drove the one before it. Pair that with something else: volume drying up, a level being tested for the third time. Now you have a case. Divergence by itself is a question, not an answer.
The other place RSI earns its keep is confirming what price is already telling you. A breakout on a chart looks more convincing when RSI is climbing alongside it and hasn't gone anywhere near overbought yet. That's a move with room to run before it gets stretched. A breakout where RSI is already at 78 is a move that's borrowed a lot of its momentum already. Same price chart, different RSI reading, different amount of conviction you should have in the follow-through.
The number itself changes meaning depending on the timeframe you're reading it on. RSI on a daily chart and RSI on a weekly chart are answering two different questions: one about the last few weeks of trading, one about the last several months. Treating a daily overbought reading as a reason to exit a position you took based on a weekly trend is mixing timeframes that don't talk to each other. Decide which timeframe your trade lives on and read RSI on that one, not whichever one happens to be flashing the signal you want to see.
None of this means ignore the indicator. It means stop asking RSI to do a job it was never built for. It's a momentum gauge, and momentum is genuinely useful information: a stock running hot can keep running hot, and a stock that's exhausted its momentum can chop sideways for a while even after price stops falling. Read it for what it's telling you about the strength of the move you're already looking at. Don't hand it veto power over a trade you took for reasons that had nothing to do with a line at 70 or 30. The traders who get burned by RSI aren't the ones using it. They're the ones who stopped looking at the chart once the number crossed a threshold.