The point of asset allocation — and investing in general — is to achieve an appropriate blend of performance and risk, based on the needs and requirements of the specific investor. But what exactly is risk?
A common definition is the volatility of returns, as measured by the standard deviation. A more useful approach is to gauge risk against the adequacy of the retirement portfolio. What is a large risk for a portfolio that is on target to meet the client’s objectives may be quite small when viewed in the context of a portfolio that will fall well short of expectations.
In this analysis, retirement adequacy will be defined as having a nest egg equal to or greater than 12 times your final annual salary — or a 12x retirement portfolio.
In the case of a person with a salary of $35,000 at age 30 who received an annual salary increase of 3% and retired at age 65, the final salary would be $95,617 and the target retirement amount balance would be $1,147,400.
The 12x logic is based on an initial withdrawal rate of 4% from the portfolio in the first year of retirement, which in the example amounts to about $46,000, or roughly half of the individual’s final annual income of $95,617 at age 65.
Any subsequent increase in the annual drawdown will be up to the retiree, but a typical assumption is a 3% COLA during the retirement years. Any additional retirement income is assumed to come from other sources, such as Social Security. Bear in mind, this is a simplified approach to retirement nest egg sufficiency and is a general guideline, rather than a strict rule that applies to everyone.
- Read more: Kitces: Smart Fix for the 4% Rule
Ideally, we start saving for retirement in earnest by age 30, giving us roughly 35 years to build up our retirement nest egg. Assuming we retire at age 65, we would have a 35-year distribution period if we live to age 100.
The symmetry is interesting: We have 35 years to prepare for what could be a 35-year retirement; no small task. Let’s examine how this accumulation-to-distribution process has played out over the past 89 years.
Let’s start with a summary of 10 asset-allocation models, or more simply, a menu of portfolio designs. As shown in the chart below, the portfolios range from single-asset approaches (such as 100% cash, 100% bonds or 100% large-cap U.S. stocks) to multi-asset models.
In the second column, the first number is the average 35-year annualized rolling return over the 89-year period from 1926 to 2014 (there were a total of 55 rolling 35-year periods). The second number is the average 35-year standard deviation over the 55 rolling periods. Finally, the third number is the worst one-year return during the 89-year period. The models are listed by worst one-year return, from smallest to largest.
The 10 investment models in this analysis are drawn from four asset classes: cash, bonds, small-cap U.S. stocks and large-cap U.S. stock. These four asset classes were used because performance data are available back to 1926. Today, of course, we would use a wider variety of asset classes in building diversified allocation models.
In calculating returns, cash is represented by 90-day U.S. Treasury Bills. Bonds are represented by SBBI U.S. Intermediate Government Bonds from 1926 to 1975, then the Barclays Capital Aggregate Bonds Index from 1976 to 2014. Small U.S. stocks are represented by the Ibbotson Small Stock index from 1926 to 1978, and the Russell 2000 Index from 1979 to 2014. Large-cap U.S. stocks are represented by the S&P 500.
These asset classes are combined to create various allocation models, including two four-asset portfolios. The first is an equal-weighted 50/50 model consisting of 25% cash, 25% bonds, 25% small-cap U.S. stocks and 25% large-cap U.S. stocks. The second is a growth-weighted 65/35 model consisting of 10% cash, 25% bonds, 25% small U.S. stocks and 40% large U.S. stocks.
All models consisting of more than one asset class were rebalanced annually back to the specified allocations. Performance does not reflect inflation or taxation.
A portfolio consisting totally of large-cap U.S. stock had an average 35-year return of 11.1% over the 89 years — the highest among this group of models. But the standard deviation of returns was just over 18%, or nearly 64% higher than the rate of return. Additionally, the worst one-year return for an all-stock portfolio was minus 43.34% in 1931.
The least rewarding asset allocation model consisted of 100% cash, with a 4.5% average 35-year return. But cash had very low volatility — an average 35-year standard deviation of 2.7%. Its worst one-year return was minus 0.02% in 1938.
Now we get to the heart of the matter. The right side of chart shows the rates of success in hitting the 12x target over 55 rolling 35-year periods over 89 years. The analysis is based on the frequency that each asset allocation model achieved a retirement account balance of at least 12x final income after 35 years of saving across 11 different annual rates.
The annual savings rate is the percentage of income that the individual invests in a retirement portfolio each year. The percentages highlighted with the blue background represent combinations of an asset allocation model and savings rate that had less than a 50% historical success rate. Gray highlighting represents combinations with a 90% or better historical success rate.
So, for example, the 65/35 growth-weighted four-asset portfolio combined with a 10% savings rate achieved a retirement portfolio at age 65 equal to at least 12x the investor’s final annual income in 96% of the rolling 35-year periods.
By comparison, a 100% large U.S. stock portfolio achieved the same frequency of historical success at a 9% annual savings rate — but with more volatility along the way. This model had an average 35-year standard deviation of 18.1%, compared with 13.7% for the 65/35 model. This is 32% more volatility.
A 50/50 equal-weighted four-asset portfolio had a 95% success rate assuming a 12% annual savings rate. Thus, a bit more savings effort by the investor enabled a reduction in the portfolio’s standard deviation from 18.1% in a 100% large U.S. stock model to 11.2%.
Said differently, the 100% large U.S. stock model has been nearly 62% more volatile than the 50/50 four-asset model over the past 89 years. The “price” of removing extra volatility was saving more money each year.
The success percentages highlighted by the larger black font represent the sweet spots, or the savings rate/portfolio design “efficient frontier,” which is essentially the gray boxes that are furthest to the left.
WILL IT LAST?
Finally, having arrived at retirement with a 12x portfolio, how long will it last? Shown in the chart below is a quick summary of the rates of success of assuming a 4% initial withdrawal rate. A 3% increase in the cash withdrawal was assumed after the initial withdrawal at age 65.
Success in this analysis is defined as the retirement portfolio lasting at least 35 years, or until age 100. This analysis of portfolio durability was drawn from the same 55 rolling 35-year periods from 1926 to 2014.
A 100% cash portfolio and 100% bond portfolio were unable to persist for 35 years at a 4% withdrawal rate. A high rate of success was not achieved until the portfolio consisted of at least 40% stock.
Even better than simply increasing the percentage in stock was also increasing the diversification. The two diversified four-asset retirement models were clearly superior over the past 89 years.
Bear in mind that for the retirement portfolio models that had a low success rate (100% cash and 100% bonds), the portfolios ran out of money before age 100.
In summary, over the past 89 years, a 12x retirement portfolio was achieved 95% of the time by investing in a four-asset 50/50 portfolio and saving 12% annually from age 30 to 65. For slightly more risk-tolerant investors, a 12x retirement portfolio was achieved 96% of the time by investing 10% of annual income into a growth-oriented four-asset 65/35 portfolio between the ages of 30 to 65.
Once retired, both four-asset retirement portfolios lasted at least 35 years 98% of the time. Moreover, the average ending balance at age 100 for the 50/50 portfolio and 65/35 portfolio were higher than any of the other retirement models.
Again, every investor must start early. Advisors must tell clients that if they start later, they’ll have to save more each year.
Craig L. Israelsen, a Financial Planning contributing writer in Springville, Utah, is an executive in residence in the personal financial planning program at the Woodbury School of Business at Utah Valley University. He is also the developer of the 7Twelve portfolio.