The fixed-income portion of retirement portfolios is just as important as the equity allocation, yet far less research has been devoted to it. Advisors must decide whether to pursue active or passive strategies and which types of bonds to recommend. I'll address those strategic choices and argue that the best approach is the simplest, lowest-cost one.
Active versus passive
Data availability is different for stocks versus bonds. For stocks, there is good historical information on the performance of active versus passive funds, persistency of performance (i.e., do good funds repeat?) and whether those selecting funds, such as pension fund managers, are able to successfully choose funds that beat the indices.
For fixed income, there is less to go on, and the research that is available is subject to interpretation.
Each year, S&P publishes the "S&P Indices Versus Active Funds (SPIVA®) U.S. Scorecard," which provides detailed comparisons of fund performance versus indices for the past one, three and five years. The scorecard runs 28 pages and provides a lot of detail by specific categories within both equity and fixed-income investments.
But despite the level of detail, there are problems of interpretation for fixed-income investments. Here's an example. For the five years through the end of 2013, 63% of investment-grade intermediate-term active funds outperformed the Barclays Intermediate Government/Credit benchmark (AGG), and active funds produced an average annual return of 5.81% versus 3.96% for the index. These results might be interpreted as solid evidence favoring active management in this fixed-income category.
However, let’s we dig into the numbers a bit deeper. The AGG index uses a mix of fixed-income investments reflecting the size of different segments of the bond market. The bond market is dominated by Treasury bonds and government-backed mortgage pass-throughs, so the AGG holds mostly these types of bonds. Many investors have been attracted to the higher yields offered by corporate bonds. This Vanguard report shows that, at the end of 2012, the AGG index consisted of 23% corporate bonds, while funds benchmarked to the index held an average of 44% corporate bonds. This discrepancy raises the question, discussed in the Vanguard report, of whether the funds' investment mix represents manager skill in tilting toward corporate bonds or whether it represents investors’ desire for higher yield.
A longer history
Five years is not long enough to evaluate active versus passive performance, particularly when that time period featured a consistently improving economy. We need to see how managers have performed through the economy's ups and downs.
In this report, Rick Ferri and Alex Benke analyzed 16 years of data. They compared the performance of indexed Vanguard equity and bond funds to actively managed equity and bond funds. For the bond category, they used the Vanguard U.S. investment-grade fund for comparison. Over the period 1997-2012, the Vanguard fund beat the randomly chosen active fund 91.5% of the time. When the active fund did worse, the median loss was an annualized 0.99%, compared to an average 0.23% advantage when the active fund won. The 91.5% winning percentage in the fixed-income category compared to 77.1% for U.S. equities and 62.5% for international equities. Expense ratios had a bigger impact on the results for fixed income (0.7% for active versus 0.2% for passive) than equity funds because gross returns for fixed income are more tightly bunched.
Including more years of data provides quite a different picture and underscores the difficulty advisors face when selecting a fixed-income fund that will outperform an index.
Choosing the best active managers
It might be argued that the numbers presented from the Ferri and Benke study do not paint a fair picture because advisors will not choose managers randomly, but instead will be selective and choose the best managers. This leads to the question of whether performance persists. Will a superior fixed-income manager of the past five years be likely to continue the performance in the next five years?
Unfortunately, there less large-sample evidence about bond-fund performance persistence than there is about stock-fund persistence, and the studies that exist are dated. In 1995, Ronald Kahn and Andrew Rudd of Barra published a study that examined stock funds and bond funds separately. They found that fixed-income funds did indeed demonstrate more performance persistence than equity funds, but "the average underperformance of fixed-income funds more than cancels out the benefits of being able to choose above-average funds through persistence alone.” Anecdotal information is mixed. There are managers who have outperformed for multiple years, but there are also managers who produced superstar performance for many years and then lost the magic. The well-publicized problems of the largest bond fund, PIMCO Total Return, serve as an example.
In terms of the active versus passive choice for fixed income, there is evidence, but it is sketchy in places. Clients whose advisors choose the active route will likely pay annual investment expense charges of 50 to 75 basis points, plus perhaps a similar level of advisory charges to cover the work involved in selecting and monitoring managers. So choosing an active strategy requires overcoming an expense hurdle of 100 to 150 basis points compared to a passive strategy, with which the cost to clients can be held to 20 basis points or below.
My weighing of the available evidence favors passive fixed-income strategies.
Choosing bond categories
Even if the choice is to follow a passive approach, there is still the issue of which categories of fixed-income investments to choose. Among the choices are Treasury bonds, Treasury inflation-protected securities (TIPS), agency bonds, investment-grade corporate bonds, mortgage-backed securities, international bonds, high-yield bonds and, for taxable funds, municipal bonds.
In his book Unconventional Success, David Swensen, manager of the Yale Endowment, argues for a simplified fixed-income approach for regular investors consisting of an even split of Treasury bonds and TIPS. He argues against adding corporate bonds to the mix. His view is that fixed income should be for safety, and risk-taking should be left to equities. He says Treasury bonds offer risk reduction, because when stock markets are in turmoil, investors seek the safe haven offered by Treasury securities and drive prices up. He also notes that corporate bonds typically fail to deliver going-in yields, not only because of defaults but also because of optionality, such as call provisions, built into the bonds, which favors borrowers at the expense of the bond investors. For example, a call provision gives the borrower the right to pay off bonds early when interest rates decline, which forces the bond owners to reinvest at lower rates. Antti Ilmanen, in his book Expected Returns, makes a similar point about corporate bonds and also mentions the impact of downgrading and trading. He estimates ex-post realized credit spreads of 30 basis points compared to ex-ante spreads averaging 120 basis points.
In the following charts, I'll compare corporate to Treasury bonds to provide some insights about Swensen's recommended Treasury-only approach for fixed income. A more complete analysis would require separate evaluations for all the other bond categories.
Chart 1 provides a long-term comparison of annual returns for corporate and Treasury bonds based on Ibbotson® data. The categories are long-term corporate bonds (LTCB), long- and intermediate-term government bonds (LTGB and ITGB) and Treasury bills.
Besides the average returns and standard deviations, I also show Sharpe ratios calculated as the spread over the T-bill return, divided by the standard deviation. This measures the return premium in relation to risk undertaken.
Chart 1: Historical Fixed Income Returns 1926 - 2013
|
|
LTCB |
LTGB |
ITGB |
T Bills |
Avg. return 1926-2013 |
6.29% |
5.89% |
5.41% |
3.54% |
Standard deviation |
8.38% |
9.88% |
5.68% |
3.11% |
Sharpe ratio vs. T bills |
0.33 |
0.24 |
0.33 |
0.00 |
LTCB premium over LTGB |
0.40% |
|
|
|
Sources: Ibbotson(R) 1926-2011, various after 2011
Comparing long-term investments, the corporate bonds have a Sharpe ratio advantage over government bonds. The return premium for corporate is 40 basis points, close to Ilmanen's estimate, though there is a lot more fine-tuning in Ilmanen’s research than the simple difference I show here. The chart does show a more attractive Sharpe ratio for ITGB than for LTGB, reflecting the substantial difference in standard deviations. Let’s look at the intermediate-term category more closely.
Chart 2 compares intermediate-term corporate to Treasury bonds, but over a shorter period because there is less historical data available.
Chart 2: Intermediate-term returns 1972 - 2013
|
ITCB |
ITGB |
Avg. return 1972-2013 |
8.40% |
7.68% |
Standard deviation |
8.68% |
6.71% |
Sharpe ratio vs. T bills |
0.36 |
0.36 |
ITCB premium over ITGB |
0.72% |
|
Sources: Ibbotson(R) and St. Louis Fed (FRED)
We now see equal Sharpe ratios for the two categories, with the spread of 72 basis points over this particular period. Higher returns were achieved by investing in corporate bonds, but there was a cost in terms of higher volatility.
Chart 3 focuses specifically on the risk-hedging quality of Treasury bonds. Using the Ibbotson® data, I selected good (return over 20%) and bad (losses more than 10%) years for stocks and compared the performance of corporate and government bonds. Good hedging would entail government bonds doing better than corporates in down years for the stock market and the opposite in good years. Prior to 1950, the Great Depression and World War II disrupted the data, but since 1950 there have been 31 years when the stock market made big moves, and the hedging relationship worked in 22 of them (71%). In the recent data shown below, the hedging worked every year except 2001.
Chart 3: Recent years with big stock market moves
Year |
Lg. co. stk |
LTCB |
LTGB |
ITGB |
2001 |
-11.89% |
10.65% |
3.70% |
7.62% |
2002 |
-22.10% |
16.33% |
17.84% |
12.93% |
2003 |
28.68% |
5.27% |
1.45% |
2.40% |
2008 |
-37.00% |
8.78% |
25.87% |
13.11% |
2009 |
26.46% |
3.02% |
-14.90% |
-2.40% |
2013 |
32.42% |
-5.87% |
-13.03% |
-3.09% |
Sources: Ibbotson(R) 1926-2011, Various after 2011
Treasuries provided particular comfort in 2008.
Choosing a strategy
Based on the evidence I have presented, advisors should favor the simplest and lowest cost fixed-income strategies – passive investing and limiting fixed-income investments to government bonds (and considering municipal bonds for high-tax-bracket investors with taxable accounts). For government bonds, there's a question of regular Treasury bonds versus TIPS. Swensen recommends a 50/50 mix, and I agree. Returns are not perfectly correlated, so there is some diversification advantage from a mixed strategy. Also, the market for regular Treasury bonds is much bigger and less subject to disruption in times of market turmoil, such as at the end of 2008 when TIPS surprisingly lost value. But TIPS do provide the advantage of inflation protection, so I favor a mix.
Another consideration is duration. Based on the Sharpe ratios for the historical Ibbotson® numbers, there is no need to invest long-term to achieve the greatest advantage. Intermediate-term investing works fine.
Finally, there is the question of purchasing individual bonds (or bond ladders) versus bond funds. This subject is worth a separate article (or book) on its own, but my view is that the main focus should be on choosing the approach that keeps expenses the lowest. The decision to choose a fund or individual bonds is secondary.
A broader view
I've discussed fixed-income strategy in terms of bonds only. Earlier articles I've written have discussed the advantages of delaying Social Security and purchasing single-premium immediate annuities (SPIAs). Both of those strategies have the effect of increasing one’s allocation to fixed-income-like investments. I've shown SPIAs to be very competitive to bonds as fixed-income investments, and they also have the advantage of reducing longevity risk. While I recommend simple, low-cost bond investing, investors should also consider Social Security delay and SPIA purchases as part of an overall fixed-income strategy.
Joe Tomlinson, an actuary and financial planner, is managing director of Tomlinson Financial Planning, LLC in Greenville, Maine. His practice focuses on retirement planning. He also does research and writing on financial planning and investment topics.
Read more articles by Joe Tomlinson
