by Scott Burns
Craig Israelsen, an associate professor at Brigham Young University, wants us to know that it pays to build “Steady Eddy” portfolios for retirement. Having lots of asset classes, he says, is a good start. His recent article “The Benefits of Low Correlation” in the Journal of Indexes tells us a lot about a question that vexes most investors.
The question: If common stocks provide a 10 percent return over the long run, why can’t I spend 7 percent a year and leave 3 percent to compensate for inflation? Why do all these planners tell me I can spend only 4 percent? Where does that lost 3 percent go?
For most people, it just doesn’t seem right.
The statistical answer, which is less than illuminating, is that the lost return went down the variance sink. Another answer is that the 3 percent is the price of market volatility.
Build a more stable portfolio and you’ll pay less for market volatility--- the inevitable ups and downs of asset prices. You’ll sleep better. You’ll raise the odds you can spend more each year. And you’ll be able to do it without going broke at some inconvenient time, like your 72nd birthday.
To demonstrate the idea, professor Israelsen tested the impact of increasing diversification on retirement portfolios. The portfolios were set for an initial withdrawal rate of 5 percent, with the original dollar withdrawal amount increasing by 3 percent each year thereafter. Using the period 1970 to 2006, he examined both the statistical stuff (standard deviation and internal rate of return) and the scary stuff (worse single-year loss and frequency of one-, two- or three-year losses of 10 percent or more).
A portfolio with seven asset classes not only has a higher return than a simple domestic equity portfolio, it also has about half as much risk. More important, this Joseph’s Coat portfolio had no periods with losses greater than 10 percent. The simple portfolio had worst-year losses three times larger, a daunting 30 percent.
In a telephone interview, professor Israelsen was quick to point out that some asset classes make better additions than others. Adding foreign stocks, for instance, didn’t do much for your portfolio. Adding intermediate bonds and cash lowered risk, but it also lowered return.
“The real kicker came with the addition of REITs and commodities. Commodities were the seventh asset added, but it had the lowest correlation of any of the asset classes,” he said. “You really need an asset class that moves in the opposite direction (from other asset classes). It’s good to have something that goes up when everything else is tanking.”
His two-asset portfolio--- 50/50 U.S. large- and small-cap stocks--- produced an annualized internal rate of return of 10.74 percent. But it lost a deadly 30.8 percent in its worst year. His seven-asset portfolio--- equal portions of U.S. large- and small-cap stocks, international stocks, U.S. intermediate fixed-income, cash, REITs, and commodities--- provided an 11.25 percent return. But the worst-year loss was only 10.2 percent.
When he observed that some portfolios appear more diversified than they are, I asked for an example.
“We can have a portfolio with U.S. large-cap and EAFE (international stocks), and it will look diversified. But it isn’t in terms of timing--- the timing of return is too similar. Diversification is really like an engine. Each asset class is a piston. It’s important that they ‘fire’ at different times.”
The Holy Grail of portfolio design would be a combination of asset classes that perfectly offset each other, providing a smooth return of 9 percent or 10 percent rather than a bumpy one. Such a portfolio, if it could be built, would allow an annual inflation-adjusted spending rate starting at 6 percent or 7 percent instead of the frequently prescribed 4 percent.
The benefit to retirees would be massive. It would amount to a 50 percent to 75 percent increase in portfolio spending power for retirees. Many retirements would work a lot better if the safe withdrawal rate could be raised to 5 percent.
So here’s an embarrassing question: How close are most investors to having the amount of diversification that professor Israelsen suggests is a good start?
Well, we have a long way to go. But the data is encouraging. Collectively, we’re a lot more diversified than we were only a few years ago. The Hewitt 401(k) index, for instance, shows that we not only own the traditional large-cap domestic stocks (21 percent), but we also own small- and mid-cap domestic stocks (about 8 percent), as well as international stocks (10 percent) and emerging market stocks (1.55 percent). We still have significant risk in our employers’ stock (16.6 percent), but the monthly transfers routinely show we are moving money out of company stock and into international stock.
Basically, all that’s missing is a REIT allocation and a commodities allocation. Unfortunately, those are the two asset classes that do the most to reduce portfolio risk.
On the web: