Wednesday, November 17, 2010

Is There a Point to Diversification?

A friend I’ll call Jake was analyzing his investments and posed the following question (lightly edited):
I’ve got a standard mix of ETFs, including XIC, XIN, XSP, XBB, and XRB, each with a target percentage of my portfolio and a plan to rebalance when things get out of whack. I plotted the value of my portfolio against the TSX. Guess what? All three lines are almost identical. The correlation isn’t perfect, but close enough over any time period.

So what has my “diversification” and “balancing” bought me (aside from extra transaction fees)? Are markets so tightly interconnected as to make “diversification” impossible/meaningless? When was the last time you saw the S&P go one way but the DOW and/or NASDAQ go the other? Is there any advantage to carving off a chunk of cash and investing in a sector or part of the world? Logic says yes, but the results say no.

I won’t give up on my diversification just yet, but if I was giving advice to a newbie it might be “buy XIC and some bonds and diversify later (much later!).”
Jake has a point that the world’s stock markets are highly correlated. The fact that including some fixed-income investments in the mix has made little difference is a coincidence that won’t continue indefinitely. However, the interconnectedness of our world usually makes different stock markets move together.

The short answer to the value of diversification is the protection it gives from extreme events. Stock markets tend to move together when all is running smoothly, but there is always the possibility of major events that affect one part of the world more than others. I’d hate to have my investments concentrated in one sector or country and have that country experience some severe crisis.

As for the advice to a newbie, I don’t see anything wrong with starting out with just one or two ETFs while the portfolio is small. What is the value in having $2000 diversified among a half-dozen ETFs? There is nothing wrong with buying one ETF with the first $2000 and then buying some other ETF six months later with another $2000 that has been saved. Another direction for small portfolios is low-cost index mutual funds.


  1. Well, in extreme global events, the value of diversification to different markets (not talking stock/bond diversification) seems minimal.

    For example, go to Google finance and pull up a 10-year chart of the TSX composite and the S&P500 (these aren't total returns, but good enough for my point).

    If you flip back and forth between them, you'll see the shapes are nearly identical: a drop off at the beginning for the tech wreck 10 years ago, a bottoming around early 2003, then a steady rise up until the 2008/2009 crash, with a sharp rebound from March '09. There's a lot of correlation in there, especially when things get ugly on a global scale.

    But look at the overall 10-year returns: the TSX is up ~37% while the S&P500 is down ~14% (and in Canadian $ terms I think it's even worse, ~-35%?). So despite what looks like a large amount of correlation to the eye, even over a long period of time, you could be getting very different long-term returns out of each of these components. Even just the Dow vs the S&P500 has a 10-year difference (again, not including dividends) of over 20%.

    Since you can't predict the winners and losers in advance, it's best to go with the diversification route.

  2. @Potato: I agree with your point that different stock markets can drift apart over time even if their price charts have a similar shape.

    Some major events affect the entire world, but it is possible to have major events that affect a single country. For example, if the world decided that they don't like the environmental effects of oil sands, there might be a world-wide boycott of Canadian oil. I don't consider this likely at all, but if it happened, the effect on Canada would be substantial.

  3. It's a very interesting point and I tried to evaluate it. A perfect diversification would pretty much keep the investment completely flat. Trying to diversify, at the same time geographically, by industry and type of business size etc. I believe would probably even out the gains with the losses. That is why one uses ETFs - as a quick way to bet in a less risky way on geographical area or a sector. Without suggesting to try to time the markets, sometimes it is visible that a certain area and/or industry would perform better. After all, it was Warren Buffet, I believe, who said that a mediocre player in an excellent industry will perform much better than an excellent player in a mediocre or industry (and this could work for geographical areas).

  4. @Andi: The critical question is: do you believe that you can figure out which industries and businesses are mediocre or excellent? I certainly can't.