Some portfolio managers suggest that in many ways, the answer is a clear no. They say the returns on short-term paper are so low that they offer virtually no protection if short rates jump. And if interest rates do climb, it is short-term rates that could swing the most.
Meanwhile, the slope of the yield curve is much steeper than usual, but investors often fail to take the opportunities there.
The muni team at DWS Investments suggests much of the misallocation in investors’ municipal portfolios is a result of overreliance on duration analysis, which too many buyers blindly apply to the tax-exempt market.
Duration is a somewhat tricky concept measuring a bond’s sensitivity to interest rates. Annoyingly, it’s expressed in years, even though duration is not a measure of time. (Time is the main determinant of duration, so it is often linked with maturity — the number of years until principal is repaid — but the two are not synonymous.)
Each year of duration generally implies a 1% change in price when interest rates move 100 basis points.
Let’s say you own two bonds. One has a duration of 4.5, the other is 8. If interest rates jumped 100 basis points, the first bond would lose about 4.5% of its value in price; the other would lose about 8%.
Higher durations imply riskier investments, so all else being equal, bonds with longer duration are more volatile. Longer maturity bonds tend to have longer duration, so investors seeking safety understandably crowd into short-term paper.
But not all durations are created equal, DWS pointed out last week. This basic math works best for Treasuries, but munis are far more stable. Using duration assumptions in the municipal market overstates the volatility of tax-exempt paper and thereby persuades investors to avoid longer-term debt.
DURATION & ASSET CLASSES
To show the limits of duration analysis for munis, DWS looked at how muni and Treasury returns varied from their average returns over the past decade.
This measure of volatility shows the 10-year Treasury return, as measured by Citi’s benchmark 10-year, deviated from its average by 8.22%. The Barclays Capital 10-year municipal bond index, with roughly the same duration, deviated by 5.05%.
A 10-year muni has roughly double the duration of a five-year Treasury. Relying solely on duration analysis would imply the shorter-term Treasury would be less volatile, but over the last decade, the higher-yielding 10-year municipal bond experienced roughly the same volatility.
“People are seeing Treasury rates bounce all over the place, but that doesn’t mean other asset classes behave in the same way,” said Ashton Goodfield, head of muni trading at DWS. “Munis have a much lower degree of volatility even if they have the same duration.”
Barclays’ municipal index, which tracks investment-grade munis with maturities greater than two years, has an average duration of 6.2. Its volatility over the past decade was less than 5%.
Barclays’ emerging market index has a similar duration but experienced roughly twice as much volatility.
Credit Suisse’s high yield index, which tracks speculative corporate bonds, has an average duration of just 3.8, yet it registered volatility of almost 10% from their average returns.
The evidence suggests longer durations don’t necessarily mean more risk.
“For Treasuries, duration is the dominant form of risk,” DWS said in a note published Sept. 12. “For more speculative bonds, duration risk is just a small portion of their overall risk.”
Popular misconceptions of duration abound in the muni market, Goodfield said, because it’s so dominated by retail.
“Maybe they don’t use the term duration, but individual investors will invest in short and intermediate parts of the yield curve thinking that will keep them out of trouble,” she said. “But the value is elsewhere. The longer part of the yield curve — you could have better performance there even if some interest rates are rising.”
According to Municipal Market Data, a top rated muni bond maturing in five years yields just 0.90%. Locking into such a low yield for half a decade, DWS suggests, fails to offer much protection in case short-term rates rise. And if they don’t, 0.90% offers no real return; in fact, it’s below inflation.
“It’s not the safe haven people think it is,” said Anthony Valeri, investment strategist at LPL Financial, speaking of short-term munis. “There just isn’t as much yield there to provide a buffer and it could end up being counterproductive over long periods of time.”
Valeri said duration is a decent starting point, but retail buyers show too much bias towards it and interest rate risk. He calls that emphasis a prime reason why the muni curve is steeper, historically, than other asset classes.
MMD shows the spread between two-year and 10-year debt has averaged 179 basis points this month. Four years ago, it averaged 26 basis points.
Stepping out the yield curve this way made little sense in 2007, but an investor could now sextuple his cash flow. Even after the Federal Reserve pancaked the yield curve by announcing Operation Twist last week, MMD’s two-year spot yielded 0.32% on Thursday, while the 10-year offered 1.97%.
Investors choosing to buy short munis despite ultra-low yields are likely doing so on the assumption that when rates rise, they will have cash on hand to stuff into higher-yielding paper.
This could be a swell move if rates do take off. But the strategy comes at the cost of missing out on opportunities now.
J.R. Rieger, president of fixed-income indexes at Standard & Poor’s, noted on Friday that the S&P AMT-Free Muni Series 2020 Index yielded 2.46% tax-free. The taxable equivalent of that, for investors in the 35% tax bracket, is 3.78% — more than double the 10-year Treasury rate of 1.83%.
“Retail investors flock to the short-term maturities and they really should be doing the opposite,” Valeri said. “They would have been better off in intermediate and long-term bonds the last two years.”
MMD also shows the spread between five-year and 10-year muni yields is rarely greater than the spread between one-year and five-year debt. Since early 2010, however, that’s been the norm as fearful investors crowd the short-term market.
The one- to five-year spread is currently 68 basis points, while the spread between five years and 10 years is 121 basis points. The extra pickup means investors are rewarded more than usual for each step beyond the five-year spot.
“A lot of retail investors think they are always safer in the short, two-to five year or two- to 10-year part of the curve,” Goodfield said. “But with interest rates the way they are and the shape of the curve the way it is, there really is more value going out longer.”
John Dillon, chief muni strategist at Morgan Stanley Smith Barney, has been advocating a search for yield to longer-term, single-A rated muni paper for months.
He reiterated the strategy in his Sept. 9 report, noting that three-quarters of the yield available on the muni curve is captured within 14 years of maturity. Because shorter yields are abysmal, the key bet is in the six- to 14-year range, he said. Extending beyond that is less attractive.
“Our allocation to short term is 10% or in some cases zero,” Valeri said. “We just don’t think you’re adequately compensated there and the opportunity is much greater on longer-term bonds.”
Key to this strategy is an outlook that sees interest rates remaining relatively stagnant. But even if rates jump, greater cash flow from longer-term bonds could offset any price weakness.
A study earlier this year by Florida-based investment advisers Wasmer, Schroeder & Co. showed that 12 different capital bond indexes maintained by Barclays had positive total returns in each of the three rising rate environments of the past 20 years.
The study contradicts the belief that bond portfolios lose value when rates rise and highlights the opportunity costs of parking in money market funds for lengthy periods.
“Income, over time, is the most important component,” the study found.
RISING INTEREST RATES
One problem for investors is a lack of a precision when thinking about “rising interest rates,” DWS said. Investors often wrongly assume that when rates rise 100 basis points, short- and long-term rates all jump in tandem in what’s called a parallel shift. In fact, this rarely happens.
Over the last 20 years, for instance, the two-year muni rate swung from 5.20%, in 1995 to 0.30% earlier this month. The 20-year rate fluctuated from 6.85% in 1994 to 3.24% last week. The standard deviation on the two-year year note was 1.30, nearly double the 0.69 standard deviation of the 20-year note.
This implies that not only do longer-maturity bonds yield more, but they are also sensitive to the part of the yield curve that is less volatile.
“Investors trying to reduce risk by moving from long-maturity to short-maturity investments won’t necessarily reduce the volatility of their investments,” DWS wrote.
True, investors in individual bonds could just sit out a volatile period until redemption. But those holding short-term funds, which have no maturity dates, could see unexpected volatility on what they thought was a safe-haven investment.
Another error investors make: assuming the duration of a bond fund works just like an individual bond. It can, but it might not.
A bond fund with an average duration of five years might be comprised of bonds with four- to six-year maturities — a bullet portfolio — but it could also be comprised of 15-year bonds balanced with cash — a barbell portfolio.
“Both portfolios have the same duration, but they act differently when intermediate rates move,” DWS said in its note. “If just the intermediate rates go up, the bullet portfolio would decline while the barbell portfolio would not.”
The limits of duration analysis have become more pronounced in the last three years as the municipal market transforms to a credit-driven market from a rates-driven one.
In the era of monoline bond insurance, when more than half of all new volume was enhanced to triple-A credit ratings, buyers often ignored underlying credits — foolishly, as it turns out — and purchased almost solely on yield.
With the collapse of the insurers and the resurgence of deep credit research, it’s been widely remarked that munis are less homogenous and are now driven more by credit concerns.
Less emphasis on interest rates means that duration, while still important, is more limited than before.
One consequence is that municipal bonds are no longer so correlated with the Treasury market.
In the three years leading up to the credit turmoil in September 2007, the correlation for 10-year maturities was nearly 93%. When the muni market lost its credit supports, the correlation between the two assets declined to just 45% for the next three years.
In the past year, thanks to ratings recalibrations and a market becoming comfortable in the post-monoline era, the correlation has climbed to more than 75%.
“We’re still well below the levels when munis were considered more of an interest-rate product,” said Dan Berger, senior market strategist at MMD. “As munis become more comfortable being a credit market, we’re approaching pre-crisis levels, but we’re not there yet.”
Whether the two asset classes re-link with each other or not, duration can still be a useful starting point in looking at the potential performance of a muni bond or fund.
But Goodfield stresses that muni performance is now going to come from a wider variety of sources than before.
“Duration is still going to be a main component,” Goodfield said. “But you can’t just invest in the market, look at how interest rates are going to change, and predict how a fund will do. A lot more of the performance now is going to come from credit issues.”