Making Sense Of Sentiment Data

This is the weekly version of The Prime Wave. Here you will find items of general interest to active stock market traders. The weekly is free for all subscribers, daily updates of the Swingex Index are reserved for paid subscribers only.

THE BIG IDEA

Investor sentiment is reliably negative when the stock market hits bottom. Look at any important market bottom, and you will have a hard time finding any data that shows many people feeling bullish. But there is more to the story than just “buy when everybody is bearish”.

The weekly AAII sentiment survey of individual investors (the “II” in AAII) is known to many traders. Earlier this year, a matter of days after the S&P 500 made a new all-time high, there was a shocking result: 19.4% bullish, 60.6% bearish (and 20.0% neutral). This ranks amongst the most negative readings ever in the history of the survey. In the table below we collected data from previous times when bears outnumbered bulls by more than 40 percentage points.

Yes, there were some big positive returns in the future. But all is not as it seems. To understand it better, let’s group the previous instances of big bearish readings into three time periods:

1990 - 5 times

2009 - 1 times

2022 - 4 times

The lone case in 2009 accurately marked a bottom in the stock market during the Great Financial Crisis. But what about the others? The August 1990 survey was merely the first in a series of bearish results over the following three months and the market continued lower during that period. If you went all-in the first time there was a big 40% difference between bears and bulls you would have been underwater until January 1991.

Same thing in 2022. The first episode of super-bearishness occurred in April, but the market didn’t even hit bottom until October.

Returning to 2025, it turned out that individual investors were not wrong to veer in a bearish direction at the end of February. The S&P 500 plunged 16% in six weeks and still has never once been above the level from February 26th. Once again, a strategy of immediately taking the opposite side of the trade did not work out well.

Tom McClellan, keeper of the McClellan Oscillator and other market indicators, often reminds us that sentiment data and other similar statistics should be taken as a “condition” and not a “signal”.

What we have found at swingex.com is that retail traders are not necessarily wrong when they start to see dark clouds on the horizon. The problem for the “dumb money” is that they can remain bearish too long.