For all long-term investors, there is only one objective— maximum total real return after taxes.
—John Templeton, founder of the Templeton Group
As you have learned, much of the conventional wisdom about investing in general, and fixed-income investing specifically, is wrong. Unfortunately, the conventional wisdom that municipal bonds are appropriate investments for only those investors in the highest tax bracket falls into that category of bad advice. In fact, there are times when even investors in the lowest tax brackets should consider owning municipal bonds. This chapter will provide you with the knowledge needed to determine whether municipal bonds are the appropriate choice and to decide on a prudent investment strategy.
We have already covered a substantial amount of information related to investing in municipal bonds. We begin, therefore, with a review of that material. We will then move on to discuss other issues that need to be understood in order to develop an appropriate strategy. The following is a summary of the issues that we have covered to this point.
There are several factors contributing to the fact that the yield curve for municipal bonds is generally much steeper than it is for Treasury bonds. The first is the tax, or regulatory, risk discussed in chapter 2. The second is that municipal bonds entail credit risk, while Treasury bonds do not—and the longer the maturity, the greater the credit risk. The third reason is that there is a supply-demand imbalance in the municipal bond market. The supply of long-term debt is much greater from issuers (who seek to match the maturity of the debt with the long-term nature of their assets) than there is demand from buyers. The result is that the “sweet spot” of the municipal bond yield curve is generally farther out for municipals than it is for Treasuries.
Having completed our review of the material previously covered, we will address other important issues. We begin with a discussion of the characteristics of the municipal bond market that make it unique and differentiate it from the Treasury and corporate bond markets.
One of the main differentiating characteristics of the municipal bond market is that there are fewer participants than in the taxable bond market. For example, while international investors are a significant part of the taxable market, they play no role in the tax-exempt market. This insulates the municipal bond market from the volatility that can be created by “hot money” flows. For example U.S. Treasury securities, being viewed as the safest instruments in the world from a credit perspective, are generally the main beneficiary of capital inflows during “flights to quality” that can occur during a financial or political crisis. Corporate bonds, on the other hand, might be subject to outflows at the same time. Of course, rapid inflows and outflows tend to reverse just as quickly when the crisis is resolved. The absence of these participants from the municipal bond market makes it less volatile than the taxable bond market.
A second difference is that there are no large institutional tax-exempt investors (e.g., pension and profit-sharing plans) in the municipal bond market. Their tax-exempt status makes the purchase of municipal bonds unnecessary. Thus the municipal bond market is not subject to speculative trades as investors shift assets between market segments in an attempt to outperform their benchmarks. For example, if pension fund managers believed that the economic outlook was likely to improve, and thus the risk of corporations defaulting on their bonds would decrease, they might sell Treasury bonds and purchase corporate bonds. If they thought the economic outlook was going to deteriorate and default risk would rise, they would sell corporate bonds and buy Treasuries. Fund managers also shift assets between the various other segments of the taxable bond market (e.g., MBS, emerging-market bonds). The absence of these participants results in lower volatility.
Another factor in the lower volatility of municipal bonds is that it is illegal to short a municipal bond. Going short involves bor-rowing a security and immediately selling it. The seller expects to eventually purchase the security at a lower price and then return it to the lender. Going short entails the taking of much greater risk than does the ownership of a security. When you own a security your losses are limited to the amount of the purchase. With a short sale losses are unlimited. Thus the practice of selling short is generally limited to speculators (e.g., hedge funds) and traders. Since these investors generally have very high turnover rates, their absence is a third reason why the municipal bond market is less volatile than it is for Treasury or corporate bonds.
Another factor impacting the pricing of municipal bonds is that there is a greater supply of long-term bonds from issuers than there is demand for long-term municipal securities from investors. As we discussed in chapter 4, municipalities are generally trying to match their liabilities with the long-term nature of their assets (e.g., highways, bridges, buildings, parks). On the other hand, most investors demand securities with short to intermediate maturities. In order for the market to absorb all the supply, prices adjust downward (yields rise). This supply-demand imbalance is one of the factors that results in the municipal bond yield curve being steeper than it is for Treasury bonds. There are, however, other factors. First, municipal bonds entail credit risk that increases with time. Second, municipal bonds carry the risk of a change in the tax law that could result in the loss of their tax-exempt status. Third, tax-exempt investors (who buy only taxable securities) such as pension plans and foundations need to buy longer-maturity bonds in order to match their long-term liabilities. This demand puts downward pressure on yields for long-term taxable bonds. There is no such effect in the world of municipal bonds. Thus the “sweet spot” on the municipal bond yield curve is generally farther out than it is for Treasury and corporate securities.
We now turn to a discussion of issues related to the taxation of municipal bonds.
While the interest on municipal bonds is generally exempt from federal taxation, there are exceptions. The first is the aforementioned rule that applies to interest on bonds that are bought at a significant discount to par: Unless subject to the de minimis rule, the amount of the discount is subject to ordinary income taxes. The second is the AMT that applies to interest on what are called private activity bonds (PABs). PABs are issued for the purpose of providing special financing benefits for qualified projects such as airports, housing, or industrial development. Interest on all PABs issued after August 7,1986 must be included in the calculation of the AMT.
The calculation of the income tax for individuals must be done on a parallel tax system. The tax due for individuals is the greater of the tax due under regular tax rates or the tax due under the lower tax rate of the AMT. The AMT is designed to prevent tax-payers from avoiding significant tax liability by taking advantage of exclusions from gross income and certain deductions. One of the tax preferences that must be added back into income in terms of calculating the AMT is the interest income from PABs. For those who are subject to the AMT, the value of the tax-exempt feature is thus reduced. Because of this feature, PABs generally carry slightly higher interest rates. The higher rates make PABs attractive to investors not subject to the AMT. This increases the demand for such bonds. The result is that the incremental yield (historically about ten to fifteen basis points, though it can be greater) is not sufficient to offset the loss of the tax benefit for investors subject to the AMT. Thus it is crucial that investors check before purchasing any municipal bond whether they themselves are subject to the AMT and whether the interest on the bond is subject to the AMT.
There are two other important points for investors to be aware of related to the AMT. The first is that mutual funds are the largest buyer of PABs. The attraction is the higher yield—which they advertise in an attempt to attract investors. Investors subject to the AMT may be attracted by the higher gross yield. Unfortunately, they may be unaware that the after-tax return may actually be lower than could be achieved by a fund that had a lower gross yield (because none, or a lower percentage, of its bonds are subject to the AMT). The following example will illustrate this point.
Consider two municipal bond funds. Fund A and Fund B both buy securities of the same maturity and credit risk. The only difference between the funds is that Fund A buys no bond that is subject to the AMT. Its average yield is 5 percent, and thus its after-tax return is also 5 percent. Now consider Fund B. Fund B invests 20 percent of its assets in bonds that are subject to the AMT. These bonds yield 5.3 percent, or 0.3 percent more than non-AMT bonds. The remaining 80 percent of its assets are yielding the same 5 percent as the bonds in Fund A. Thus the average yield of Fund B is 5.06 percent (80 percent x 5 percent plus 20 percent x 5.3 percent, which equals 4 percent plus 1.06 percent). Thus Fund B looks to be the more attractive fund as it provides a 0.06 percent greater yield than does Fund A. And it is for investors not subject to the AMT. The problem for investors who are subject to the AMT is that the 5.06 is a gross return, not a net return. Assuming an AMT rate of 28 percent, the net return to investors subject to this tax will only be 4.76 percent. The calculation is as follows. The 80 percent of the portfolio that is not subject to the AMT yields 5 percent, contributing 4 percent (80 percent x 5 percent) to the total return of the portfolio. The net return of the remaining 20 percent that is subject to the AMT is just 0.76 percent (20 percent x 5.3 percent x [1–0.28]). Thus Fund A is clearly the superior choice for investors subject to the AMT.
Investors considering the purchase of a municipal bond fund should carefully review the prospectus, which must disclose whether the fund can purchase bonds subject to the AMT, and what, if any, limitations there may be (e.g., a maximum percentage).
There is a second issue related to investors in AMT bonds. While the first issue relates to investors who use mutual funds, this one relates to investors who purchase individual bonds. The secondary market for AMT bonds is less liquid than it is for nonAMT bonds. The result is that if investors need to liquidate a bond prior to maturity (or wish to do so in order to harvest a loss for tax purposes) then they are likely to find that the bids for their bond will be lower than they would have been otherwise. The price difference is likely to reflect a yield differential of perhaps five basis points.
The AMT is an important issue for two reasons. The first is that historically about 10 percent of all municipal bond issues have been subject to the AMT. The second is that because income subject to the AMT is not subject to an inflation adjustment (unless Congress changes the law), more and more taxpayers will become subject to the AMT. Given that it is likely that more and more investors will become subject to the AMT, the price differential between AMT and non-AMT bonds is likely to widen, with negative implications for current holders.
There is one other tax related issue we need to cover.
In general, interest expense on funds that are borrowed to purchase investment securities is tax deductible. The exception is that any interest expense incurred to purchase or hold securities, the interest on which is exempt from taxation, is not tax deductible.
In the same spirit, some investors who use investment advisors may take a miscellaneous deduction for advisory fees. However, the amount of the fees that can be attributed to the advice given with regard to municipal bonds is not deductible.
We will now cover issues related to the credit risk of municipal bonds.
There are two types of municipal bonds, general obligation bonds (GOs) and revenue bonds. General obligation bonds are backed by the full faith, credit, and taxing power of the issuer. The credit worthiness of revenue bonds, on the other hand, is determined by the success of the particular project (e.g., hospital, road, water system). The credit support difference between the two bonds, unfortunately, leads to the great misperception that GOs are safer than revenue bonds. The supposed safety of GOs comes from the phrase “full faith and credit” and the theoretically unlimited taxing authority of the issuer. This is in contrast to revenue bonds that get their only support from the revenues generated by the project.
If, in reality, all GOs were safer than all revenue bonds, all GO bonds backed by the full faith and credit of the issuer would be AAA-rated. But this is not the case. The reason this is not true is that in the real world the power to tax is limited. For example, the ability to tax may be limited by state constitution. It can also be limited by economic conditions or even the general mood of the public. Thus the credit rating of a GO will depend on such issues as the strength of the issuer’s tax base, the strength and diversity of the economy of the issuer, the total level of outstanding debt of the issuer and its debt coverage ratio (ratio of revenue to debt), and the issuer’s historic commitment to prudent fiscal policy.
The ratings of revenue bonds will be based on the issuer’s ability to generate revenues sufficient to cover both operating expenses and interest. The stability of revenue will play a major role in determining the rating. In addition, other factors, such as the dependence on governmental support or reimbursement can play significant roles in determining the creditworthiness of a project.
As was mentioned in the review at the beginning of this chapter, municipal bonds have significantly less credit risk than do similarly rated corporate bonds. For example, Standard & Poor’s found that over a fifteen-year period a single-A-rated municipal bond was one-tenth as likely to default (0.16 percent default rate) as was a similarly rated corporate bond (1.8 percent rate of default). 1 And, in general, municipal bonds have experienced many fewer defaults than have corporate bonds. However, there are some sectors of the municipal bond market that do not have a good credit history—and thus prudent investors are best served by generally avoiding them. Those sectors are housing, health care, and industrial development. A September 1999 Fitch IBCA study on municipal bond default risk found that 7.52 percent of all industrial development bonds issued between 1979 and 1994 had defaulted by 1999. For multifamily housing bonds the number was 4.78 percent, for health care 1.60 percent, and for electric power, 1.46 percent. Contrast these default rates with the following ones: 0.05 percent for water and sewer revenue, 0.04 percent for education (school districts and the like), and 0.01 percent for transportation (highways, public transportation systems). There are two reasons that go a long way toward explaining the high default rates in these areas. The first is that hospitals have a strong dependence on government reimbursement programs that often do not fully cover costs. The second is that many development projects (especially nursing homes) are built as speculative projects by private businesses.
As the municipal bond market is one of the largest markets in the world, there is little reason for a prudent investor to even consider the purchase of bonds from sectors that have experienced such high rates of default.
There are many municipal issues that do not qualify for a high investment rating of AAA or AA. Their lower credit quality has a direct impact on the cost of borrowing. There is, however, also a secondary, indirect effect. Since the demand for municipal bonds is strongest for those in the highest investment grades, a lower credit rating limits the number of potential buyers. This negatively impacts the liquidity of lower-rated bonds, with a corresponding negative impact on interest costs. In order to minimize total costs, the issuer can purchase a credit enhancement, typically in the form of insurance. The cost of insurance to the issuer has historically been in the range of ten to forty basis points per annum—though it can be much greater for highly risky credits.
Weaker credits are not the only ones that can benefit from the purchase of a credit enhancement. Government entities from smaller communities that do not come to market frequently are likely to suffer a liquidity premium. The reason for this liquidity premium is these issuers are neither widely known to the market nor are they likely to bring to market a large enough issue to create sufficient liquidity on a stand-alone basis. For a smaller municipal issuer, buying insurance may even be simpler and less expensive than applying for a credit rating from the major rating agencies. A third type of issuer might also consider buying insurance. When a transaction is very complex, the market for that issue will be negatively impacted. Many potential investors might not even want to consider evaluating the risks. Thus issuers of securities with complex features may find it advantageous to purchase insurance in order to achieve broader market acceptance.
A municipal bond insurance policy may result in significant interest cost savings that are attributable to the higher bond rating as well as the enhanced liquidity of insured bonds. The benefits of insurance are so great that about 50 percent of all newly issued bonds each year are insured.
The municipal bond insurance industry is dominated by the four leading insurers, each with approximately a 25 percent share of the market: AMBAC Assurance Corporation, Financial Guaranty Insurance Company, Financial Security Assurance Inc., and MBIA Insurance Corporation. They are all AAA-rated insurers. All municipal bond insurers are monoline insurance companies— they can only insure municipal bonds and are thus not exposed to risks to which property and casualty insurers are exposed. They are also highly regulated by state insurance departments.
The insurance policies written by municipal bond insurers guarantee that interest and principal will be paid as scheduled if the issuer defaults. Each guarantee is unconditional and irrevocable and covers 100 percent of the interest and principal for the entire term of the issue. The risks covered may include natural disasters (e.g., earthquakes, floods, hurricanes) and environmental hazards. In the event of default, the insurer would step in to make timely (not immediate) payments. Insurance provides value to the investor in several ways:
Safety: A secondary source will meet the issuer’s obligations in the event it cannot. Of special interest to investors is that insured bonds have performed better than comparable uninsured bonds during difficult economic times.
Liquidity: Bonds carrying AAA-rated insurance have greater liquidity. Greater liquidity results in lower trading costs. This is important for investors who may need to sell prior to maturity in order to raise capital or may have a need to sell in order to harvest a tax loss at some point.
Note that the market does not consider a bond that receives its AAA rating based on the purchase of insurance to be as strong a credit as a bond that is a natural AAA credit. This reflects the market’s perception that the monoline insurance companies are not as creditworthy as those issuers that earn a AAA rating on their own. Thus, insured AAA bonds generally carry slightly higher yields than do natural AAAs. The incremental yield might be in the neighborhood of five to ten basis points for short to intermediate maturities and a bit more for longer maturities. In fact, insured AAAs generally trade even a bit cheaper (higher yield) than do natural AAs (perhaps two to four basis points higher yield).
In summary, buying insured bonds can be part of a prudent investment strategy. Investors, however, are best served by not relying solely on the credit rating of the issuer (most insured bonds carry a dual rating from the rating agencies—a rating without the insurance and a rating with it). They should also diversify credit risk across both issuers and insurers. Keep in mind that the municipal bond insurance industry has only been in existence since the early 1970s, and thus its claims-paying ability has not been stress-tested by a severe economic condition such as a depression.
In addition to insurance, there are two other types of credit enhancements that issuers can purchase, a letter of credit (LOC) and a line of credit.
Letters of credit are irrevocable commitments, typically issued by commercial banks, for a limited period of time (subject to renewal), that allow trustees or fiscal agents to draw on the letters when necessary to make payments of principal or interest on bonds. LOCs present two problems for investors. Thefirstis that they are typically issued for a maximum of ten years (with no guarantee of renewal). Thus there is the obvious potential for a problem to arise if the maturity of the bond purchased extends beyond the term of the LOC. The second problem is that the financial institutions providing the LOCs are generally not as highly rated as the monoline insurers.
The least valuable form of enhancement is a line of credit or other “backup” credit facility. These are even less valuable than LOCs as conditions usually exist for the provider to not have to advance funds. They are also generally short-term in nature. Investors would best serve themselves by avoiding issues that have their credit enhanced by a line of credit.
There is an additional way in which municipal bonds can have their credit rating enhanced. The enhancement comes as an indirect benefit of pre-refunding a bond.
Consider the following situation. In 2000 a municipality issues a $10 million bond with a 5 percent coupon that will mature in 2030. The bond has its first call date in 2010, with the call option at par. By 2005 interest rates have fallen, and a new issue with a maturity date of 2030 is yielding just 3 percent. If the municipality waits until 2010 to call the bond and take advantage of lower rates, rates by then might have once again risen. In order to avoid this risk, and take advantage of the current low-rate environment, the municipality issues a new $10 million bond with a maturity of 2030 (or perhaps even longer). It uses the proceeds to buy U.S. Treasury securities with a maturity of 2010 (matching the call date). The issuer then places the Treasury securities in an irrevo-cable trust. The trust uses the interest it receives to pay the interest on the older debt and the principal it receives to redeem the older, higher-yielding bonds at the 2010 call date. This process is called defeasance. The benefit to the issuer is that they will save on interest costs from the call date to the final maturity of the original issue.
Pre-refunded bonds can provide investors with the highest possible credit quality because payment is no longer dependent upon the revenue stream or tax collections of the issuer. Instead, payment is guaranteed by the collateral (generally a special type of U.S. Treasury instrument). Therefore, pre-refunded bonds, at least those that are backed by U.S. Treasury securities, are rated AAA.
Bonds are generally pre-refunded in low-interest-rate environments. Therefore, they are usually priced at premiums to par value. Because many investors avoid premium bonds, at the time of the pre-refunding these bonds will generally trade at a greater yield to maturity than par or discount bonds of comparable quality. This “market inefficiency” is a free lunch (and there are not many of those in the world of investing) for those willing to purchase premium bonds with calls that have been waived by the issuer. Another benefit of premium bonds is that they provide greater price stability because their higher coupon rates generate higher cash flows than current coupon or lower coupon bonds, thus cushioning the impact of rising interest rates.
It is worth noting that investors in the original bond receive a permanent benefit when a bond is pre-refunded. The reason is that, for all practical purposes, the original issue is now a tax-exempt bond backed by the U.S. Treasury. The increase in the credit quality from the escrow of U.S. Treasury securities provides a benefit in terms of market price. The market recognizes the impeccable quality of the credit and, therefore, pre-refunded municipals typically trade with a yield five to ten basis points lower than non-pre-refunded AAA-rated municipals. A point of note is that most municipal bonds are issued with first call dates of no more than ten years. Therefore, almost all bonds prerefunded to a call date fall within ten years of maturity.
In summary, pre-refunded bonds can play an important role in a municipal bond portfolio. And as long as all calls are waived prior to the refunding date, investors should not be concerned by the premium at which they trade. In fact, as mentioned above, the premium provides a defensive feature. In addition, the bonds are of the highest quality, and they will generally fall within the part of the yield curve investors are trying to access.
In chapter 6, “The World of Short-Term Fixed-Income securities,” we learned that there are both taxable money-market funds and municipal money-market funds that investors can use for liquidity, as opposed to investment, purposes. We also learned how to determine which provides the highest after-tax return.
While money-market funds are the most common instrument used for saving (i.e., emergency funds and funds needed for liquidity purposes) because of their convenience and safety, two municipal bond security types—municipal auction rate securities (MARS) and variable rate demand obligations (VRDOs)—can serve a comparable purpose while adding more value via higher yields.
While these securities are actually long-term bonds, they have many features that make them similar to money-market funds or other very short-term instruments. For example, in terms of liquidity, some of these bonds offer daily, weekly, or monthly reset schedules—dates when the interest rate on the bond is set. The result it that even though a bond may mature in 2035, it can have a daily, weekly, or monthly interest pay date and an optional redemption provision that allows investors to redeem their principal at par on any of the interest pay dates. Because of the feature of par in and par out liquidity on a very short-term basis there are usually no concerns regarding loss of principal.
Another positive feature is that many of these instruments carry the desired high credit rating. For example, the credit ratings of most MARS range from AAA to A. Because of the unique structure of these bonds, both MARS and VRDOs are commonly assigned both a long-term and a short-term rating. However, if an investor purchases MARS or VRDOs as a substitute for a money-market fund—and thus does not plan to hold the bonds to maturity—he might only need to concentrate on the bond’s short-term rating.
The conclusion we can draw is that MARS and VRDOs have features that make them similar to money-market funds in terms of both liquidity and safety of principal. And the attraction is that they provide a higher yield than tax-exempt money-market funds. In addition, there have even been times when, despite their tax-exempt status, they have provided higher yields than taxable money-market funds. For example, a taxable Schwab money-market fund offered a seven-day current yield of 0.78 percent (as of June 30, 2004). 2 A VRDO with a maturity date of 2040 had a rate of 1.03 percent (as of July 1, 2004). 3
There are several reasons why these instruments are able to provide higher returns. The main reason is that money-market funds have expenses; an operating expense ratio of 0.5 to 0.7 percent is fairly common. A second reason is the very large minimum purchase requirement. MARS and VRDOs generally require a minimum investment of $100,000. Thus the issuer has lower costs in managing a program. A third reason is that these instruments do not come with the check-writing privileges of many money-market funds (again reducing costs).
MARS and VRDOs also have a disadvantage relative to money-market funds in that transactions costs are likely to be charged by the custodian when buying and selling these securities. Thus the benefits of the greater yield must be weighed against the incremental costs and the lack of convenience.
Both MARS and VRDOs provide high credit ratings, high liquidity, and higher yields relative to comparable investment vehicles such as municipal money-market funds. This combination makes them attractive low-risk alternatives for those investors who can meet the minimum-purchase requirement.
Investors who do not have the liquidity needs of a tax-exempt money-market account, but still desire to keep maturities short, have a wide variety of securities to choose from.
Bond Anticipation Note (BAN): Short-term loan repaid from the proceeds of a future bond issue.
Tax Anticipation Note (TAN): Short-term loan to be repaid from future tax collections.
Revenue Anticipation Note (RAN): Short-term loan to be repaid from future revenues, either general revenues or project-specific revenues.
Tax and Revenue Anticipation Note (TRAN): Short-term loan to be repaid from future taxes or other revenue sources.
Grant Anticipation Note (GAN): Short-term loan to be repaid from an intergovernmental grant to be received in the future.
The following are the common characteristics of these short-term instruments:
Because the demand for these securities is very high, depending on the investor’s tax bracket, at times they might actually provide lower after-tax returns than comparable taxable instruments. Thus, as is always the case, investors should compare returns of taxable and tax-exempt securities before investing.
Before completing our discussion on municipal bonds we need to cover one more type of security—a taxable municipal bond.
A taxable municipal bond is not an oxymoron. There are municipal bonds that are taxable. Taxable munis are bonds that are is-sued by the same entities that issue tax-exempt securities. The bonds, however, are deemed to be issued for a private purpose (not for the public good). A good example might be a bond issued to finance a sports facility. The yields on taxable munis are comparable to those issued by taxable corporate bonds. Thus they might be worth considering for tax-exempt or tax-deferred accounts, or for investors in the lowest tax brackets.
We have now completed the section on municipal bonds. The next chapter will cover issues related to developing a fixed-income portfolio, including the development of an investment policy statement.