As my summer vacation dragged into its last week I had run out of beach-time reading, but fortunately I had packed a copy of “Asset Allocation” by Roger Gibson. The fifth edition of the 27-year-old book is just riveting. I can’t put it down.
Almost 30 years ago, Gibson explained better than anyone the concept of Modern Portfolio Theory and the wisdom of diversifying equity investments across different asset classes, mixing winners and losers, to reduce risk. Although counterintuitive, this made it possible to be fully committed to stocks while experiencing less volatility than when allocating a portion of the portfolio to bonds. If you could avoid bonds, you avoid the drag on returns that the bond portion caused over time.
The trick is to use asset classes that are “inversely correlated.” This means choosing a combination that will create comparative winners and losers over time. The more extreme, the better. The book illustrates what happens with an equal mix of large-company U.S. stocks, non-U.S. stocks, real estate and commodities. The four asset classes are rebalanced to equal percentages once a year. The book’s illustrations typically include the 1972 to 2011 period.
Bottom line: U.S. stocks alone averaged about 9 percent and the four-component mix earned about 11.5 percent. The 2 percent difference over 40 years meant that a dollar invested in the diversified mix grew to $74 over the period versus $43 for U.S. stocks alone — illustrating how important an incremental percentage increase can be. Most people would think that to earn twice as much money, a 5 percent return would have to be double, or 10 percent. Not so, thanks to the “magic” of compound interest.
A further reward is the reduced volatility of the mix. Known as the “standard deviation,” there is a range of how much an investment return can vary from the average while falling within this range two-thirds of the time.
For the diversified mix, the standard deviation figure is 14 percent. Two-thirds of the time, then, the maximum loss would be calculated as follows: 11.5 percent (the expected return) minus 14 (standard deviation), or a net loss of 2.5 percent. The standard deviation for the U.S. stock investment is 18 percent, so the bottom of the range would be 9 percent minus 18 for a possible loss of 9 percent.
But it’s not as simple as just dividing up these equity investments and returning to an equal 25 percent holding of each once a year. We have to deal with “frame of reference” risk. Investors don’t mind losing money as much, as long as everyone else is losing money, too. How well the overall market does becomes our frame of reference, and we tend to compare our results to this.
In periods like now, when the market is on a tear, there’s a side of all of us wanting to put our collective thumb on the scale of the bandwagon. But it takes nerves of steel to have half our money in something like real estate and commodities that are today’s relative losers. So, there’s room to deviate from established dogma and bet against history if it makes us more comfortable — at least inasmuch as our equity investments are concerned.
But to protect an all-U.S. stock portfolio against the downside, there is still room for bonds. One-third of the portfolio in bonds brings the expected return down to around 9 percent with the standard deviation at about 12 percent. This brings the bottom of the range to a net 3 percent loss — about the same as the four-asset mix above.
So, for the same volatility risk, we sacrifice about 2.5 percent of annual return. But we’ll be happier being part of the crowd.