Research & Publications
Fixed Income Articles

The following article is reprinted from the April, 1997 issue of On the Edge,
the Interactive Data Fixed Income Analytics bimonthly newsletter.

How to (Mis)manage a Municipal Bond Portfolio

Wesley Phoa, Ph.D.
President of Research



Over the past decade, tax-exempt bond funds have - on the whole - underperformed their performance benchmarks. The following graphs compare the average performance of municipal bond funds, as reported by Morningstar, with the relevant Lehman Brothers municipal bond index returns. Three fund categories have been analyzed:
  • "Municipal bonds - national"
  • "Long-term municipal bonds"
  • "Short-term municipal bonds"

For the first category, the fund universe was compared to the aggregate muni bond index. For the latter two categories, a performance benchmark was constructed by taking a weighted average of maturity-specific Lehman indexes, in order to match the average reported maturity of the funds in each category.

What is the reason for this consistent underperformance? It is unlikely that all muni bond fund managers are poor market timers. The problem appears to have been systemic, and is unlikely to be explained by manager-specific factors. The likeliest explanation is that muni bond portfolios are not, in general, being managed correctly: the early redemption option, a feature of most muni bonds, is incorrectly valued.

Most muni bond portfolios are managed on the basis of yield-to-worst and duration-to-worst. This ignores the dynamic nature of the options, whose values change in complex ways as interest rates and market implied volatilities shift. This phenomenon affects both the value and the duration of a bond portfolio. As a consequence, sophisticated option models are required to analyze option-embedded bonds correctly; yield-to-worst and duration-to-worst greatly understate option risk.

Investors who cannot analyze embedded options correctly run the risk of making two kinds of errors. Firstly, in using duration-to-worst they are mis-estimating the interest rate sensitivity of their portfolios; this introduces significant tracking error. Secondly, they are tempted to reach for yield without recognizing that they are exposing themselves to a high degree of volatility risk in the process. These errors, in combination, ensure underperformance in market environments where the embedded options are valuable.

The performance figures are consistent with this explanation. Muni funds outperformed in 1994; and indeed, yields rose sharply in that year, reducing the value of the call options embedded in the bonds. In 1988, short bond yields rose (making the options worth less) and short muni funds outperformed slightly; but long bond yields fell (making the options worth more) and long muni funds underperformed significantly. Of course, for any specific fund, there will be other more specific factors contributing to underperformance or outperformance.

In fact, there is some evidence that the municipal bond market is becoming more efficient, and that more sophisticated investors have begun to use fully dynamic, option-adjusted analytics; these methods have long been employed by leading taxable bond fund managers to ensure that portfolio option risk is correctly valued. BondEdge allows managers to compute effective durations and other dynamic risk measures for muni bond portfolios, using a fully dynamic option model. Interactive Data Fixed Income Analytics is currently carrying out more detailed research on muni bond performance, based partly on monthly index performance data; the results of this research will be reported as they become available.