Why One ETF Is Not Enough

With hundreds of funds to choose from and more new ones coming out every day, it's easy to think that with ETFs, more is better. But given just how closely linked many ETFs' returns are to one particular benchmark, buying a bunch of different ETFs may not do as much good as you think -- and could in fact leave you dangerously overexposed to stock market risk.

The rise of ETFs
From their humble beginnings in the 1990s, ETFs have exploded onto the investment scene. Now, you can find investments tailored to just about any investment strategy you can think of. Want to harness the power of dividend stocks for your portfolio? You can find many ETFs that track dividend-paying stocks. Are you ready to jump into foreign stock markets for greater return potential and to protect against a downdraft in the U.S. dollar? International ETFs have made overseas investing a lot easier than it used to be.

The idea behind owning different ETFs is that you want a variety of different types of investment exposure. The hope is that when one ETF goes down, the others will go up, giving your overall portfolio a gentle but substantial rise over time.

Too much correlation
Unfortunately, many ETFs don't offer as big a difference in the way they perform as investors think. Returns of many ETFs turn out to be very highly correlated with the returns of the S&P 500. Have a look:

Source: Morningstar.

At first glance, you wouldn't expect to see this list of ETFs so closely correlated with the returns of the S&P 500. Rather, you'd probably buy these ETFs specifically so they wouldn't be correlated with the S&P's returns, expecting to see them zig when the S&P zags.

But as you can see, lots of ETFs targeting vastly different asset classes have produced similar returns lately. Here are some reasons why:

• Not every big company pays a dividend. But when you look at the top holdings of U.S. dividend ETFs, you'll often find top S&P 500 companies near the top of the list. Similarly, ETFs based on fundamentals like earnings often gravitate toward the top stocks in the S&P 500, with the only difference being minor differences in how the stocks are weighted within the ETF portfolio.
• Focusing on either growth or value stocks has led to much different performance over long periods of time. Recently, though, both styles have produced roughly similar results.
• Even disparate markets like mid-cap and global stocks are starting to trade more in synch with each other. Although mid-cap U.S. stocks have outperformed their larger counterparts, they still tend to move in the same direction.

All of this suggests that you might be just as well off buying a single S&P 500 ETF than trying to build a falsely diversified ETF portfolio. But before you give up on diversification, there's still one thing left to consider.

Past performance is no guarantee of future results
The danger from any observation like this is that you apply it too broadly to your own investing strategy. Because correlation changes over time, assuming that high correlations will continue could be a big mistake.

In fact, recent correlations within the U.S. stock market have been near record levels. In other words, even inside the S&P 500, constituent stocks are moving more in lockstep than ever. Historically, though, that hasn't always been the case, and there's no reason to think that it will remain the case going forward. Different industries have different dynamics that affect them, and so their stock performance is likely to diverge over time.

The key lesson to take is that while diversification is useful, you won't always get the positive effects of diversification that you expect. In some market environments, it's very difficult to find investments that will perform differently. Over time, though, you'll find that close correlations tend to ebb and flow -- and that makes diversifying your portfolio well worth the effort.

Great companies don't always reward their shareholders. Morgan Housel recently found five companies that doubled earnings while their stocks went nowhere.