David Wilson of Bloomberg quotes Pierre Lapointe, a “global strategist” at Brockhouse & Cooper to point out that the weekly RSI for the S&P 500 index is at a critical level. He suggests that because the RSI is at 75.48 (using 14 RSI and the weekly data) the equity market is near a top. Generally speaking, an RSI above 70 is “overbought” as can be seen in this chart:
Lapointe and also Alex Bellefleur write in a recent report: “Sentiment is almost euphoric, with very few skeptics remaining…”. While I completely agree with that and have been reporting this through the various indicators and measures featured in the weekly sentiment overviews, I don’t think I can agree with the RSI indicator argument.
I haven’t read the full report so perhaps I’m overlooking other data that Lapointe and Bellefleur already cover but taking a longer perspective shows that they may be succumbing to the recency effect. Here’s a chart showing the (14 weekly) RSI for the S&P 500 index:
There are definitely several points where an RSI above 70 coincided with a market top. However, there are also many where there was no such correlation. For example, if you look at 1995 we can see an especially exaggerated divergence. While the RSI went above 70 early in the year and stayed there for the rest of the year, the index itself rocketed higher with no noticeable pause. So that was no help because while it indicated a top, there was none.
The next example I want to show you is the major market top in 2000. This was the infamous tech bubble top that is even today, overshadowing the market. The RSI failed to point out this important top. While the S&P 500 was carving out this ceiling for itself, the RSI never climbed above 70 and was in fact not even close. So again, the RSI was of no help because a market top took place while it provided no warning.
The October 2007 market top was a similar scenario. The market topped out and the RSI was once again, below the 70 “line in the sand” level.
If you actually look at the formula for the indicator, this isn’t all that surprising. RSI is simply a momentum indicator so during times of sustained momentum, it will register an extreme. And if the market makes a top while not under a momentum surge, it will not even register on the RSI’s radar.
There are, of course, many different ways of using the RSI indicator than simply waiting for it to rise above 70 (and selling) or waiting for it to fall below 30 (and buying) which is what I assume Lapointe is hinting at. We can look at divergences for example. The 2007 top provides a good example for that. In any case, it is important to remember that no one single indicator provides a silver bullet.