If your clients portfolio is following the tired your age in bonds axiom that is, that the percentage allocated to bonds should equal your clients age you may want to reassess your strategy.
Bonds present less volatility than stocks, of course but thats not the only consideration one should make when building a portfolio. The optimal portfolio needs to serve two goals: controlling downside risk while also achieving a reasonable rate of growth.
The order in which returns occur has a dramatic impact on the longevity of a retirement portfolio. Market-based losses (or very low positive returns) in the early years can be disastrous to the long-term solvency of a portfolio.
Thus, the portfolio needs to be sufficiently diversified so as to minimize the timing-of-returns risk. Building a portfolio that has a very large allocation in any one asset class is simply asking for trouble.
In truth, the age-in-bonds strategy has always been a questionable idea for older clients.
There are four core asset classes we can measure back to 1926: U.S. bonds, U.S. large- and small-cap equities and cash. To perform a long-term analysis, I used these asset classes to construct various model retirement portfolios.
Bonds have done a good job of avoiding losses (in nominal terms) over the past eight-plus decades. From 1926 to 2013, U.S. bonds averaged an annualized return of 5.4%, with a standard deviation of 5.7%. The worst one-year return for bonds was a loss of 2.9% in 1994.
Over these 88 years, there were nine years in which bonds had a negative return. In those years, the average loss was 1.3%.
Yet bonds have also experienced protracted periods of very low returns, and those create a distinct challenge in a retirement portfolio, particularly if the return is below the withdrawal rate.
For instance, from 1941 to 1969, bonds experienced a 29-year period in which the average annualized return was a mere 2.2%. During those same years, large-cap U.S. stocks were averaging 12.8% and small-cap stocks were averaging 18.6%. This simple observation reminds us of the value in diversifying, particularly during a clients retirement years.
Large-cap U.S. stocks (as represented by the S&P 500) have produced an average annualized return of 10.1% from 1926 to 2013, with a standard deviation of 20.2%. Large-cap stocks endured 24 negative years, or losses 27% of the time, in this period. The largest loss was 43.3% in 1931, and the average loss was 13.6%.
Small-cap U.S. stocks have an even more colorful past. Over the past 88 years, small-caps have produced an average annualized return of 11.5%, a standard deviation of 31.9% and a worst one-year loss of 58% in 1937. Small-caps have experienced a one-year loss 28 times since 1926, or nearly 32% of the time; the average loss in those years was 16.8%.
Finally, we consider the performance of cash (defined here as U.S. 90-day Treasury bills). From 1926 to 2013, cash had an average annualized return of 3.6%, a standard deviation of return of 3.3% and a worst one-year return of a loss of 0.02%, in 1938. That was the only year with a nominal loss (ignoring inflation).
The performance of three different retirement portfolios is illustrated in the Can This Portfolio Survive? table above. Shown here is the ending balance in each portfolio when the client hits age 100 (assuming retirement began at age 65). If the portfolio did not last for 35 years, the age of the investor at the point of portfolio insolvency is shown in red.
The first portfolio consists of 100% U.S. bonds. The second portfolio is an age-in-bonds model that allocates 65% to U.S. bonds at the start of retirement and increases the allocation to 66% the next year, 67% the next year and so on. The remaining allocation each year is committed to large-cap U.S. stocks.
The third portfolio is an equally weighted, annually rebalanced mix of U.S. bonds, U.S. large-cap stocks, U.S. small-cap stocks and cash.
The objective here is to determine how often each portfolio remained solvent for a full 35 years over 54 rolling periods, from 1926 to 2013. I believe that a portfolio that remains intact until an investor is 100 years old represents a worthy achievement.
The first simulation of portfolio durability assumes an initial withdrawal rate of 3% of the starting portfolio balance (which I arbitrarily set at $250,000), with an annual cost-of-living adjustment of 3%. Thus, the first years withdrawal is $7,500; the second years withdrawal is increased by 3% to $7,725 and so on. The withdrawal in the 35th year is $20,489. A total of $453,466 is withdrawn over the 35-year period.
As shown in the table, a 100% bond portfolio only had a success ratio of 69%. That is, it remained solvent in 37 out of 54 rolling 35-year periods, or 69% of the time. And the 100% bond portfolio at a 3% withdrawal rate lasted 34 years during the period from 1931 to 1965 the portfolio made it until the investor was 99 years old, not 100.
Starting in 1933 and lasting until 1948, there were 16 periods in which an all-bond portfolio failed to last a full 35 years. That said, in every case the portfolio did last until the investor was at least 94 years old.
By comparison, at a 3% withdrawal rate, both the age-in-bonds and four-asset portfolios lasted a full 35 years in all 54 rolling periods. Moreover, the retirement account balances at the end of the 35th year were dramatically larger than the all-bond portfolio in both cases.
The average ending balance for the all-bond portfolio (at age 100) was just over $821,000; it was $1.4 million for the age-in-bonds portfolio and $3.6 million for the equally weighted portfolio.
When the withdrawal rate was increased to 4%, the all-bond portfolio survived for a full 35 years in only 43% of the periods, and had an average ending balance, at age 100, of about $494,000. Furthermore, in 16 of the rolling 35-year periods, the portfolio was out of money before the investor was 90 years old. By comparison, the age-in-bonds portfolio had an 82% success rate, and the four-asset portfolio a 98% success rate.
As noted by the abundant red text in the upper portion of the table, the all-bond retirement portfolio was particularly ineffective in persisting for a full 35 years during the 1940s, 50s and 60s, when bond returns were anemically low. The average ending balance at age 100 for the age-in-bonds portfolio was just over $911,000, while it was $2.8 million for the equally weighted portfolio.
At a withdrawal rate of 5% rarely considered prudent the all-bond portfolio survived for 35 years only 30% of the time, the age-in-bonds portfolio had a success ratio of 54% and the four-asset portfolio survived 89% of the time. The average ending balance at age 100 for an all-bond portfolio was $255,000, while it was $452,000 for the age-in-bonds portfolio and $1.9 million for the four-asset portfolio.
The importance of a diversified portfolio during retirement is clearly illustrated here. For retiree clients seeking an aggressive initial withdrawal rate of 5% or higher, it is incumbent to build a diversified portfolio that has growth potential combined with prudent downside protection. An all-bond or age-in-bonds approach ignores the virtues of diversification when it is arguably needed most during the retirement years.
TIMING IS CRUCIAL
In a retirement portfolio, when money is being systematically withdrawn, the timing (or sequence) of returns is crucial. Look at the 35-year period from 1929 to 1963. At a 3% initial withdrawal rate, all three portfolios survived for the full 35 years, to age 100. At a 4% withdrawal rate, all three portfolios failed before the retiree reached the age of 94.
At 5%, meanwhile, the all-bond and age-in-bonds portfolios actually performed slightly better than the diversified four-asset portfolio the only period during which that occurred but all three portfolios failed prior to age 88.
The reason for the demise of the four-asset portfolio by the age of 83 during this particular period was the disastrous impact of U.S. stock returns from 1929 to 1932. Large-cap stocks had losses of 8.4%, 24.9%, 43.4% and 8.2% over this four-year period. Small-caps performed far worse, with losses of 51.4%, 38.2%, 49.8% and 5.4%.
If a clients portfolio encounters large losses in the first few withdrawal years, the impact upon its durability can be especially devastating. Even so, given such huge losses, it is somewhat remarkable that the four-asset portfolio still lasted until the investor was 83 years old, well beyond the average life expectancy of a retiree during that era.
The all-bonds and age-in-bonds portfolios failed to survive because bond returns during that 35-year period were too low to sustain either of the higher withdrawal rates. The age-in-bonds portfolio was hurt slightly more than the all-bonds portfolio, due to its exposure to the stock market declines of 1929-1932.
Now consider the 35-year retirement period from 1933 to 1967. These results are markedly different: At a 3% initial withdrawal rate, the four-asset portfolio had an ending balance of over $5 million, compared with just over $526,000 during the period from 1929 to 1963. We see a nearly ninefold difference in outcome due to beginning retirement a mere four years later.
The differences in outcome are equally dramatic for the four-asset portfolio at a 4% and 5% withdrawal rate. This illustrates sequence-of-returns risk. True, the Great Depression was a dramatic and rare event, but it does demonstrate the importance of the initial returns experienced by a portfolio.
By their nature, bond returns are less volatile and expose investors to less timing of returns risk. But the painful trade-off is the inability of an all-bond or age-in-bonds portfolio to generate sufficient growth to sustain a withdrawal rate adequate to meet the needs of a retiree for 30-plus years.
There is no perfect retirement portfolio. Every investment faces some type of risk, including equity volatility risk, interest rate risk, inflation risk and currency risk. The key is to build a portfolio that addresses each unique risk while maintaining adequate exposure to needed portfolio growth. Diversification across asset classes is one such way.
Craig L. Israelsen, a Financial Planning contributing writer in Springville, Utah, is an executive in residence in the personal financial planning program in Utah Valley Universitys Woodbury School of Business. He is also the developer of the 7Twelve Portfolio.
- Veres: The Next Great Market Meltdown
- Tech Survey 2014: What's New Now?
- Uh-Oh: The Economists Are Wrong. Now What?