Tuesday, August 31, 2010

Tapping Into Investment Sentiment from Millionaires

Tom Bradley at the Steadyhand blog wrote a piece pointing out that millionaires are bearish right now according to one measure. It turns out that Spectrem Group maintains an index of investment sentiment among millionaires by doing 250 interviews each month. So, do these millionaires know something we don't know?

Bradley wasn’t endorsing this index in any way, but it got me thinking. If there is any group that may know something about investing, maybe it is millionaires. Perhaps that’s how many of them came to be millionaires. Perhaps this index has some predictive value.

The scale of the index is -100 to +100. Five years ago it was hovering around +20. However, in late 2007 it went negative. This was nearly a year before the market meltdown. This seems like a vote in favour of millionaires having useful insight into short-term stock market moves.

The next test is whether this index predicted the huge stock market rebound in 2009. Unfortunately, millionaires were bearish through all of 2008 and 2009. This index completely missed one of the biggest single-year stock market surges in history. So, it seems that millionaires aren't infallible.

This millionaire index may be useful for something, but there seems to be little evidence that it is useful for predicting stock market moves. I suspect it is more useful for financial advisors to know what sorts of investments to pitch to wealthy people.


  1. Interesting post.

    People are always interested in the myth of a secret club of millionaires who know all of the insider tips and can predict the future.

    But I guess if they quoted the poor persons investment sentiment index, nobody would really care :)

  2. @Echo: The more plausible conspiracy theory is a secret club of billionaires who create the future and don't answer silly survey questions :-)

  3. "by doing 250 interviews each month"

    There are apparently 250,000 millionaires in Canada. Interviewing 250, means they sampled 0.1% of the population. Must be a truly representative sample!


  4. @CC: I wouldn't be surprised if they actually got the statistical part right. If you do the right things, it is possible to get reasonable information about a population of 250,000 with only 250 samples. I think they claim to be plus/minus 6%. The more difficult problems to deal with are whether the respondents are saying what they really think and whether they actually know anything useful.

  5. I think it is widely known that taking broader surveys of investors (not just millionaires) is a good contrarian indicator.

    When individual investors are bearish, it's a good time to invest. This indicates there is a lot of money on the sidelines ready to rush back in when the market is set to rise.

    I was listening to a podcast from late July which stated that retail investors are bearish while institutional investors are bullish. Another podcast stated that bond funds are seeing massive inflows. Seems to me that once there is some positive news and stocks rise a little, we could be in for a strong rally.

    While I suspect this may happen, I'm not doing anything different. As usual, I'm fully invested in equities. I don't see anything cheap enough that I want to take on margin, which I rarely do anyway, and only in very conservative amounts.

  6. @Gene: I find myself in a position similar to yours frequently. I learn some information that causes me to have some opinion about the short-term direction of stocks. Then I don't act on this opinion. I used to act on these opinions, but I've come to realize that I'm wrong as often as I'm right.

  7. @MJ:

    You make a good point. I find 100% of my predictions are correct when I make them, but half of them turn out to be wrong later on.

    This reminds me of a scene from the show "Corner Gas", set in the town of Dog River, with about 500 residents:

    Lacey: Yeah, right. What are the chances of there being a riot in Dog River?
    Karen: Fifty-fifty.
    Lacey: What? Fifty-fifty? How do you figure?
    Karen: Well, either there will be a riot, or there won't be. Gee, Lacey you really suck at math.

    So, in a way, I have a 50-50 chance of being right.

  8. @Gene: I've always liked using the 50/50 bit with lottery jackpots. A perverse truth about compounding is that if you take action on your guesses and are right 50% of the time, you'll lose to the index.

  9. I prefer to look at surveys like this one that take consider everyone's opinion and computes weighted average based on how much they have invested.

  10. @Patrick: You're right that this type of "survey" is tough to beat. They sometimes get it wrong, but I never know when.