The American Recovery and Reinvestment Act created a new kind of municipal debt instrument called a Build America Bond. BABs were designed so that issuers, if they chose, could raise capital through the broader and deeper liquidity of the taxable market.
Constrained liquidity in the traditional tax-exempt market made BABs, which carry a 35% interest subsidy to the issuer or a 35% tax credit for the investor, a desirable financing alternative for municipal issuers. BABs thus ensured that a greater number of new-money capital projects can be financed at interest rate levels acceptable to the municipal issuer.
Our analysis shows that as this market matures the real option-adjusted spread between BABs and tax-exempts is widening, making BABs an increasingly attractive borrowing alternative.
Since issuers have the choice of funding new-money projects through issuance of BABs or a tax-exempt bond, comparison of the relative merits of each alternative is critical to their decision making.
Originally, a major complicating factor in making that assessment was the fact that most BAB deals were being issued with make-whole call provisions, while most tax-exempt issues are sold with 10-year optional calls.
However, as a steady stream of new BAB deals have come to market, the traditional municipal call has become very common. In June, 22 of 24 BAB issues were sold with an optional 10-year call at par. While the new trend towards using a traditional 10-year call on BABs reduces the complexity of evaluating the two alternatives, problems remain.
The uncertainty regarding whether BAB deals are advance refundable, and whether the subsidy payments will be maintained, create a concern that the 10-year call on the BABs is worth less to the issuer than the 10-year call on a tax-exempt issue. So far the advantage of BABs over tax-exempts is sufficiently large for most issuers that the benefit of a BAB financing overwhelms concern about whether the calls are properly valued.
In order to properly evaluate the two types of bond issues, an issuer needs to be able to place a BAB financing on an equal plane with a tax-exempt financing. After subtracting the 35% subsidy received from the federal government from the taxable coupon rate, the issuer must then compare the net BABs rate - including the value of its 10-year optional call - to the tax-exempt rate, including the value of its the 10-year optional call, which is possibly the more valuable call. Only then can the two financing alternatives be properly compared.
Determining the value of the call given the uncertainties regarding refunding regulations is the biggest stumbling block to reviewing BABs and tax-exempt bonds on an "apples for apples" basis.
Our option-pricing model draws heavily from the Black Derman Toy model. The most unique aspect of the Loop option pricing model is that, as an input, it uses a volatility that is derived solely from the municipal market. This is in contrast to the standard practice of choosing volatility based upon an arbitrary percentage of a taxable swaption. Using this methodology, we evaluate two issues, one sold in early April when BABs were just getting started and one sold in June, to determine what we can learn about the relationship between BABs and the traditional tax-exempt market, and any valuation trends.
The ideal laboratory for evaluating BABs and tax-exempt bonds occurs when the same issuer prices tranches of each type at the same time. This has occurred on a number of occasions.
We found two great test cases, one in April and one in June, where the same issuer sold both types of bonds. This month, a transit authority issued BABs and tax-exempt bonds. Since in this case both are callable in 10 years at par, we can directly compare yields on these two issues. The BAB issue matures in 2034, has a 7.00% coupon and was sold to investors at a 7.14% yield to maturity. Including the 35% federal subsidy, the issuer will have to pay a 4.55% coupon. With the discount, this corresponds to 4.66% YTM for the issuer, 11 basis points above the Municipal Market Data curve.
At the same time the tax-exempt bond portion was issued with 5.00% and 5.125% coupons, structured with a 2032 maturity, and was priced with a 5.20% YTM, 74 basis points above MMD. The spread differential between the two issues at the time of issuance was 63 basis points, indicating that BABs were a much better deal for the issuer.
In this case it was possible to compare the two issues directly since they conveniently had the same call feature. This is not always the case, however. To compare BABs and tax-exempts with different call features on an equal basis we need to use the option adjusted spread, which is the spread versus a benchmark for a callable bond after the value of the call has been removed. The OAS spread allows issuers or investors to compare bonds with different call features on equal terms.
Focusing on the issue from April when the BAB market was just beginning, a turnpike authority sold BABs with a 7.41% coupon, a make-whole call feature, and a maturity in 2040. Deducting the 35% federal subsidy from the coupon gave the issuer a 4.82% coupon cost. The authority also issued tax-exempt bonds with 5.25% coupon, callable in 10 years, and also maturing in 2040, at 5.35% YTM.
By issuing BABs with the issuer subsidy the authority nominally saved 53 basis points. However, to calculate true savings, the subsidized yield on effectively non-callable BABs (since make-whole call is of little value) should be compared with the option-adjusted (noncallable) yield on the tax-exempt issue. Using the Loop option pricing model we determined that the issuer's tax-exempt bonds were 27 bps cheaper than they would have been without the call. Thus, we estimate true savings from issuing BABs as 53 bps less 27 bps, for a final savings of 26 bps.
Based on these two examples, and analysis of other deals, we conclude that the spread between BABs and comparable tax-exempt bonds has widened on an option adjusted basis since April, making BABs more attractive for issuers now than initially when the market began in April. This improvement in the advantage of BABs over traditional bonds makes sense given the tendency for new markets to be less efficient, forcing issuers to pay an "early adopter" penalty.
One complication in the analysis is that since BABs cannot be issued after 2010, refunding outstanding BAB issues would presumably be undertaken with a tax-exempt refunding issue. Significant advantage in pre-refunding would only occur in the case where interest rates in general are lower and tax-exempt rates have at least kept pace with taxable rates, if not outperformed. At the present time it is unclear if the Treasury would continue the subsidy on the pre-refunded BABs to the call date. If Treasury does not continue the subsidy, the value of that option is clearly lower.
Ivan Gulich is a research associate at Loop Capital Markets, where Chris Mier is a managing director and municipal strategist.