Matching investments to needs and liabilities can lessen need for smoothing. Kirk Loury is CIO of Concord Wealth Management, a Matawan, N.J.-based wealth-management platform provider.

The concept of asset diversification is widely known. It's also understood that its ability to "smooth out unsystematic risk events in a portfolio so that the positive performance of some investments neutralize the negative performance of others" holds up -- as Investopedia tells us -- "only if the securities in the portfolio are not perfectly correlated."

Still, execution often falls short and results in a false sense of security for both advisors and their clients.

Asset classes are highly correlated

There is general agreement that diversification's benefits are possible only if there is negative correlation among the securities in the portfolio. |image1|Our grasp of this concept has not changed significantly over the decades, but the challenge it presents to today's advisors has become more formidable.

The average correlations of 341 benchmarks linked to the S&P 500 have risen from 0.38 in 1996 to 0.59 in 2006, a 55% increase. Mathematically, 0.59 is in the correlation continuum's top quartile, and is considered highly correlated. Bonds (as represented by the Lehman Aggregate index), the presumed uncorrelated asset class, have increased almost two diversification quartiles the past five years ending in 2007, compared to the previous five-year period ending in 2002.

The "relative correlation" myth

In response, advisors have introduced the idea of relative correlation. This questionable concept is especially prevalent when advisors compare U.S. and international stocks. The case for relative correlation goes something like this: "While large and small U.S. stocks have a correlation of 0.9931, international stocks have a correlation of just 0.89." Unfortunately, proposals, presentations and performance reports suggesting that international stocks help smooth returns are statistically specious.

|image2| To understand why combining U.S. and non-U.S. stocks may fail to achieve meaningful smoothing, it is helpful to view correlation not as a number (and to avoid the temptation of making distinctions where none exist), but to view correlation as a pattern. The close correlation between U.S. and International stocks is clearly displayed in the graph immediately above. As the S&P 500 moves up and down, so does the international line (represented by MSCI EAFE).

Another fallacy is that returns will be similar among securities with very high correlations. However, the fact that highly correlated asset classes mixed together produce different return and standard deviation values is a function of mixing the ingredients, not the form. Think of it this way: varying the ingredients will make cakes look and taste different, but, in the end, you still have cake.

Smooth returns the natural way

Typically, diversification principles have been used to smooth equity volatility, which is much higher than the volatility of a traditional bond portfolio. This use of diversification reveals a misplaced understanding of the role that equities play in a portfolio.

Funding a client's short-term needs and liabilities is paramount; failure to do so makes an investment plan unviable (just as a business not paying its current liabilities leads to its demise). Cash primarily funds liabilities coming due in the next 12 months and bonds are used to fund needs falling in a two-to-three-year horizon. On the other hand, equities fund longer-term needs seven or more years hence.

Just as matching uncorrelated asset classes tends to smooth volatility, the passage of time can help attain the same goal.

|image3|

The Ibbotson chart above (based on 1926-2004 historical returns) shows that for a one-year holding period, small company growth stocks (blue column, far left set) are considerably more volatile than Treasury bills (blue column, far right set).

However, over a twenty-year holding period, the volatility between these vastly different asset classes becomes closely aligned, even though the compound return differential is meaningful: 12.7% versus 3.7%. Over time, the mathematical link of a series of rising and falling returns accomplished a return smoothing result similar to combining two uncorrelated asset classes. Unlike failed execution of negative asset class correlation, this happens regardless of ever tighter integration of the global economy.

Clearly, this long-term perspective requires an advisor to manage clients' expectations. Clients must appreciate the application of horizon-based portfolios in order to interpret volatility properly.

Asset-allocation implications

The presumed need for volatility smoothing is largely mitigated when investments are matched to specific need and liability groups. Needs in the short-term horizon require income-producing, low-volatility assets; needs with a long-term funding horizon are matched with equities; needs in the mid-term have assets that produce both income and growth.

Also, dealing with highly correlated asset classes doesn't obviate the need for a variety of different structural exposures. Instead of return diversification, using a process called "structural diversification" allows portfolio building with a view to the future instead of the past. We'll take this theme up in the next Critical eye. -FWR

Purchase reproduction rights to this article.