Fixed-spread tenders need better benchmark

The 2025 Bond Buyer Deal of the Year in the innovative category was the tender/refunding transaction by the State of Wisconsin. In a typical tender/refunding transaction the issuer acquires not-yet-callable bonds selling at a premium and refunds them by issuing new high-coupon callable bonds. The distinguishing feature of the Wisconsin transaction was the use of fixed spreads to determine the tender prices.

"The state and its deal team went to new lengths to carve out a more efficient and innovative financing approach," the Bond Buyer article profiling the deal says. It points out that the fixed spread tax-exempt tender approach eliminates asymmetric risk for the issuer ? investors participate if the price declines and abstain if it rises.

For this risk-management benefit to be optimally realized, the fixed spread associated with a tender/refunding should be derived from a robust yield. For corporate bond fixed spread tenders on-the-run Treasuries (e.g. 2-year, 5-year, 10-year, 30-year) are the customary, universally accepted benchmarks. In contrast, the tax-exempt market lacks such benchmark bonds. If fixed-spread tax-exempt tenders are here to stay, the municipal market will need better benchmarks.

While fixed-spread tenders may be a novelty for tax-exempt bonds, they have been around for over 40 years for corporate bonds. I witnessed the historical development of tender/refunding transactions from both the customer side (from within AT&T (T)) in the 1970s and the banking side (while at Salomon Brothers, which dominated the initial tender/refunding activity) in the 1980s.

Although interest rates hit a post-war high in 1974, Bell System telephone companies kept on issuing 40-year bonds callable Year 5, i.e. in 1979. By late 1976 interest rates declined, and in January 1977 a major tender/refunding transaction was executed by four Bell companies for bonds issued in 1974. Three of these carried 10% coupons, and the fourth coupon was 9.5%. Details are discussed in Robert Metz's New York Times (NYT) article dated Jan. 31, 1977, which is available on the internet. These were fixed price tenders conducted over a one-month period. As it turned out, because investors had no incentive to decide early, virtually all activity took place during the final days of the tender period. Fortunately for Bell, rates remained steady, and the tenders were quite successful. Because I worked at the AT&T Treasury at the time, I was able to observe this first-hand.

Interest rates reached new highs in 1981, long-term Treasury exceeded 15%, callable utility bonds were issued with coupons above 18%. What goes up must come down -- by 1984 rates substantially declined, and the bonds issued in 1981 and 1982 were ideal tender candidates. By then, participants understood the risk of keeping a fixed-price tender open for one month. But how to reduce this risk? The obvious answer was to shorten the tender period, from one month to one week. However, a short tender period carried other disadvantages. The creative solution was the fixed spread tender, which mitigates interest rate risk even over a lengthy tender period.

Tax considerations play an important role in tenders for both issuers and investors. Corporations can immediately deduct the premium paid over par, as an ordinary expense. However, such tax incentives are absent for municipalities. In fact, taxes are an obstacle in municipal tenders, because the investors of tax-exempt munis must recognize the resulting capital gains. Not surprisingly, the average acceptance rate of corporate tenders of taxable bonds was considerably higher, around 80%, than that of municipal tenders of tax-exempt bonds, typically well under 50%.

In a fixed-spread tender, the tender price is based on moving benchmark yield plus a fixed spread. Instead of the yields of individual liquid U.S. Treasury bonds used for pricing corporate bonds, the yields of tax-exempt bonds are estimated from benchmark yield curves of dubious foundation. There are several different tax-exempt benchmark yield curves. Tender prices based on the interpolated yields in a curve are questionable, to say the least. The municipal market's persistent absence of a rigorously determined, universally accepted benchmarks poses a challenge for the growth and success of fixed-spread tenders.