March 01, 2007

Indexing and Risk

Investment academics for the last few decades have preached the religion that the only way to win in investing is to diversify away your risk. In other words, they tell everybody to use indexing (and that sort of thing) instead of stockpicking.

To me, this seems a little bit defeatist, maybe even intellectually lazy, a way of saying "just put your money where everybody else has already put it." So my question is always, "what happens when everybody indexes their money?" What happens when everybody stops thinking and just starts to mimic the average as closely as they can?

In "Ahead of the Tape" in today's WSJ, Justin Lahart muses on the rocky market in the last day or two. And he proposes a hint of an answer to my question:

By deliberately investing in assets that have shown little correlation in the past, investors may be making those assets more correlated, says Craig Asche, executive director of the Chartered Alternative Investment Analysts Association. If investors find that, say, Pet Rock and corn-dog prices tend to be uncorrelated and decide to invest in both as a result, they will end up driving prices for the rocks and dogs up together.

Once everybody diversifies, is it a surprise when diversification doesn't work as well as it is supposed to?

Maybe investments shouldn't go where the money already is. Maybe us investors need to be thinking about where the money will have the biggest future impact.

Posted by David at March 1, 2007 08:42 AM
Comments

That sounds like a promising argument and there's probably some truth to it. However, if stock prices are at least partially based on fundamentals, it seems like unrelated stocks should move independently at least some of the time due to the differing fortunes of the underlying companies. (If they didn't, someone would take advantage of it.)

I'm no expert but I think this is a plausible counter-argument: It seems like indexing implies a division of labor between professional stock pickers and investors who just want to let the professionals decide. Instead of hiring a professional and suffering from their limited judgement and conflicts of interest, by using an index, you take advantage of all professional judgements for free. If the quality of the pool of professionals diminished enough, we would be in trouble. But there's no sign of that happening. As amateurs get smart and use indexes, the professionals will have to make their money by trying ever-harder to beat the market and thus making it more efficient, and everyone (hopefully) benefits from that.

Note that lots of interesting computer algorithms are already being used by the professionals to make money from tiny discrepancies and it seems pretty plausible that this trend will continue to the point where the computers take over, and competition will be in getting better data into the computers and using better algorithms to take advantage of it.

Posted by: Brian Slesinsky at March 1, 2007 12:36 PM

I agree in principle that a smart investor who spends the time to learn about market sectors and wonder about where the money will have the biggest future impact will generate hgher returns than an indexed fund. However, this implies a few things:
- Proportionally higher risk for possibly higher returns (would you/I be willing to invest 50% of our life savings on our research?)
- Time and effort to research and differentiate enough to earn the higher return

I always look at the indexing argument as an argument against actively managed mutual funds. Burton Malkiel makes a good argument in "A Random Walk down Wall Street" that
a) past returns of the mutual fund manager are no guarantee of future returns
b) what kills returns in the end for a mutual fund is the management fees
c) what you lose in returns, you make up in lower risk.
In that context, I'd rather put my money in a very low cost index fund rather than a higher cost mutual fund.

Posted by: Vamshidhar Kommineni at March 1, 2007 03:59 PM
Post a comment









Remember personal info?