Beginning in 2009, the financial press trumpeted that "asset allocation" and "diversification," the watchwords of strategic portfolio management, had failed. For example, on July 9, the online edition of The Wall Street Journal published an article called "Failure of a Fail-Safe Strategy Sends Investors Scrambling." These articles vary somewhat in their presentation, but the central arguments are always the same:

1. Diversification is supposed to mean that some assets will go up when others go down, protecting me from (substantial) losses.
2. Everything went down in 2008 to 2009, except for government bonds.
3. I lost a lot of money, so the model must be broken.

This argument shows a lack of understanding of what diversification can and can't do. There are apparently substantial misconceptions over what diversification and strategic asset allocation could reasonably provide. To judge the success or failure of any theory, you must first define what it's supposed to do. Remarkably, no "asset allocation failed" article provided concrete examples of how much risk reduction it expected from a diversified portfolio.

If investors and advisors used good portfolio analysis tools to calculate the risk they expected from their well-diversified portfolios, and the actual losses vastly exceeded what the models projected, there may be grounds for saying that asset allocation (and by extension, portfolio theory) failed. But for investors to conclude that portfolio theory failed because they lost more than they expected (or more accurately, hoped) is not a rational basis for any judgment of the effectiveness of asset allocation.

ASSET ALLOCATION MYTHS

There are a number of misconceptions about what diversification through asset allocation is realistically supposed to provide. Let's look at some of them.

Diversification means low risk. A diversified portfolio can generate more return with less risk than a portfolio that contains just one or a few highly correlated asset classes. This is the central tenet of portfolio theory. Conversely, diversification can allow a portfolio to generate the same return as a single asset class (like large-cap stocks, as represented by the S&P 500) with less risk.

No credible advisor suggests that a diversified portfolio won't lose money. Furthermore, a diversified portfolio can either be high risk or low risk. Being diversified does not mean that you have a low-risk portfolio. The WSJ article and its ilk are suggesting, however, that diversification is supposed to be synonymous with low risk—and this is simply not true. If this notion comes as a surprise, I recommend a basic primer on the efficient frontier, such as William Bernstein's The Intelligent Asset Allocator.

Combining asset classes may not lower risk. Diversification means some assets in a portfolio will go up when others go down, thereby mitigating total portfolio losses. This is true, to an extent, but it is not true all the time.

Diversification can raise returns without raising risk because different asset classes are not perfectly correlated. A correlation of 100% means that two assets move in lockstep; less than 100% means they tend to move somewhat independently. But any correlation that is greater than -100% means there will be times when two assets move in the same direction—and in 2008 to 2009, almost all asset classes moved downward. Before we can intelligently judge whether these losses violate portfolio theory, we need to look at what was supposed to happen.

For example, let's look at the diversification benefits of combining foreign and domestic equities. The now-discredited notion of "decoupling" between emerging and developed markets was supposed to mean that the emerging markets could withstand a decline in developed markets, but that did not play out. Is this a failure of the theory of asset allocation?

When you look at the numbers, the correlations between developed foreign markets and their U.S. counterparts have been high for years. Back in November 2006, I analyzed the correlations between the S&P 500 (SPY), the EAFE index (EFA) and the MSCI Emerging Markets Index (EEM). This was a period in which investors were rapidly increasing their allocations to international assets and the narrative around emerging markets was positively giddy.

Here's how correlations looked back then: The 75% correlation between EFA and SPY, while less than perfect, nonetheless suggested that EFA and SPY would move in the same direction much of the time-not always, but often. Even the correlation between emerging markets and the S&P 500 was quite high at 73%. Based on these correlations, the idea that combining these asset classes would substantially reduce risk was ill-founded. We can easily make this conclusion obvious.