Study: what happens when RSI > 80 on every time frame


The U.S. stock market’s momentum is INSANELY overbought on every single time frame: daily, weekly, and monthly.
Here’s the S&P 500 and its 14 daily RSI

Here’s the S&P 500 and its 14 weekly RSI

Here’s the S&P 500 and its 14 monthly RSI

Notice how RSI is above 80 on every single time frame. This is an extremely rare event that has only happened twice in history before January 11, 2018.

  1. July 5, 1955
  2. March 14, 1986

Let’s look at these 2 historical cases. Here’s the data in Excel.

July 5, 1955

The S&P began a 10.5% “small correction” in late-September 1955 (2.5 months later).

Its next “significant correction” (as defined by our Medium-Long Term Model) began 1 year later in August 1956.

March 14, 1986

The S&P began a 9.6% “small correction” less than 4 months later in August 1986.

The next “significant correction” was almost 1.5 years away (the crash of 1987).

Conclusion

The S&P 500 followed a similar, logical pattern in both of these historical cases.

  1. The S&P began a “small correction” that’s close to 10% within the next few months.
  2. The next “significant correction” is still far away. When momentum is insanely strong, the next correction will not be a “significant correction” because people will buy the dip.

*If you expand the data (e.g. using >RSI 75 as a criterion instead of >RSI 80), you get a similar conclusion.
These 2 historical cases were followed by very quick corrections (TIME), one of which was a V-shaped correction. This is logical.

  1. When momentum is very strong, everyone rushes to buy stocks.
  2. When the selling begins, all the short term traders get their stops triggered at the same time. This cases a mini-crash (e.g. 8-10% in a few days).
  3. But because the long term outlook is so bright, investors rush to buy the dip, thereby causing a V-shaped bottom.

I expect the next correction to be V-shaped. I also expect it to be closer to 10% than 6% (both are within the definition of a “small correction”).
I think the correction will most likely begin in February. But I could be wrong. It’s insanely difficult to predict the start of a “small correction”, and I don’t have a quantitative model for that.

10 comments add yours

  1. Got info. Do you believe this “small correction” happens wothin a very short timeframe , ie, hours days, or is it a longer drawn out event?

    • No correction happens within a few hours. Almost all corrections last AT LEAST 2 weeks.

  2. Suggestion:
    Can you write posts about Tech and Semiconductor sectors as well? Those are the two I, and most of my friends are most interested in. Basically your thoughts on semi, if overbought, etc.
    Thank you for all your hard work. I truly appreciate it, and as I said before, am addicted to your site!

    • I can. But I’m not sure what value I’d bring Ahmad. You clearly know much more about the fundamentals in tech than I do considering your profession. 🙂
      In the stock market, technicals without fundamentals is useless. So yes, I can show that tech is overbought and that overbought conditions typically need a bearish divergence before topping. But that’s about it.

      • Hi Troy, hope you’re doing well!
        I’m a big believer of 2nd opinions, and frankly me and my friends would love to see your high level thoughts on those major industries as well. It could be similar to your crypto post but more of a summarized version of it; just a high level analysis of what you think of tech and semiconductors, under the equities market post that you make.
        Did I ever tell you that you are awesome?
        Thanks!

        • Ok Ahmad I’ll give it some thought 🙂
          I hope you’re enjoying your long weekend.

  3. The RSI overbought conditions travel into the intraday time frames as well. Yesterday I checked all of the significant SPY time frames down to one minute and there was only 1 small exception. This was an extraordinary fluke of sorts, but I suspect that the deleveraging will eventually have to start slowly from the smaller time frames; unless there is an 1987 type crash. I am in Canada, and in 1987 I was working for a major US company. My stock plan was cut in half during that sudden crash. Fundamentals didn’t matter much.
    Don’t think many of today’s Stocktwit’ers understand this possibility.

    • Yes, it’s insanely overbought on every time frame (I also look at the 4 hour chart). These bearish divergences typically start from 1h to 4h to daily to weekly). Monthly divergences are unlikely because that would be the end of the bull market.

Leave a Comment