Advance Refunding Bonds

An advance refunding bond is a new municipal security issued more than 90 days before the call date of an older bond, with proceeds deposited in an escrow account to pay the old bond at the call date. This structure allows an issuer to lock in rate savings before the call date arrives, extending the period of control and refinancing optionality.

The refunding imperative

Issuers want to lower their cost-of-debt. If a city issued a 5% bond five years ago, and rates have fallen to 3%, it makes economic sense to redeem the old bond and issue a new one at lower rates. The savings accumulate year after year.

But there’s a catch. The old bond usually has a call date—the earliest date the issuer can redeem it without penalty. A typical bond issued 10 years ago might not be callable for another 10 years (or might have call protection for the first 5–10 years). The issuer is stuck paying the higher rate.

Enter the advance refunding: a mechanism to refinance earlier than the call date permits.

The escrow mechanism

Here’s how it works. The issuer floats new bonds—advance refunding bonds—and uses the proceeds to buy Treasury bill and Treasury bond securities. These Treasuries are deposited in an irrevocable escrow account with a trustee.

The escrow is sized and timed so that the Treasury proceeds will exactly match (or exceed) the amount needed to pay off the old bonds on their call date. When the call date arrives, the trustee redeems the Treasuries and uses the proceeds to call the old bonds. The original bondholders are paid in full; the old debt vanishes. The issuer now owns only the new refunding bonds.

This is elegant: the issuer locks in the refinancing outcome years in advance. If rates rise between issuance and the call date, the issuer still calls the old bonds on schedule because the money is in escrow. If rates fall, the issuer might have preferred to wait, but the transaction is already committed.

Why advance refunding bonds exist: the arbitrage window

Advance refunding bonds are usually issued when rates have fallen enough that the issuer can profitably refinance, even accounting for the time cost of waiting.

Example: A city issued $100 million bonds at 4.5% with call protection until year 10. It’s now year 7; rates have fallen to 2.5%. The city could wait three years and refinance directly at 2.5%. Or it could issue advance refunding bonds now at 2.75%, invest in Treasuries yielding 2.0%, and lock in a refinancing in three years.

The arbitrage works if the issuer’s new borrowing cost is less than the old coupon, even after paying the time value of money and the Treasuries’ lower yield. In a steeply downward rate environment, this arbitrage can be profitable by 50–100 basis points or more.

The arbitrage becomes arbitrage—sometimes poorly

The danger is that issuers become too aggressive. If rates haven’t fallen far enough, or if the issuer miscalculates the opportunity cost, the advance refunding can lock in a worse outcome than waiting.

Imagine a bond trading at a 4% coupon with call protection until year 10. An issuer issues advance refunding bonds at 3.5% and escrows Treasuries at 2.0%. If rates stay above 3.5% for the next ten years, the issuer comes out ahead by waiting. But if rates fall sharply (to 2.5% or below), the issuer wishes it had waited. Worse, if credit conditions deteriorate and the new issuer rating drops, the issuer might have issued the new bonds at 3.5% when it could have issued at 3.0% a year later.

These are real risks. Some municipalities, over-eager to refinance, have issued advance refunding bonds in moderately high-rate environments and later regretted it. The market tends to overdo refundings near cyclical rate peaks.

Double-barreled risk: what if the old bonds are not called?

Here’s a structural risk specific to advance refunding: the old bonds might not be called. If interest rates spike, or if the issuer’s credit deteriorates sharply, the issuer might choose not to exercise the call and refinance (instead, it avoids the cash outlay and keeps the old low-coupon debt).

But the escrow trustee has already purchased Treasuries. Those proceeds will eventually mature and must be accounted for. If the old bonds are not called, the refunding bonds remain outstanding—they were issued to fund a call that didn’t happen. The issuer now carries both the old bonds and the refunding bonds.

This is rare, but it has occurred. A municipality’s credit rating plummets (think Detroit, 2013), and refinancing becomes infeasible. The escrow account holds money destined to call debt that will never be called. The refunding bonds become “orphaned,” and investors are left holding unwanted securities.

Advance refunding bonds sometimes include “yield maintenance” or “make-whole” provisions to mitigate this: if the old bonds are not called, the issuer must make bondholders whole by paying a prepayment penalty. This protects investors but imposes a cost on the issuer if rates spike and refinancing is no longer economical.

The tax risk: “arbitrage bonds”

The IRS views advance refunding bonds with a wary eye. If the proceeds are invested in higher-yielding securities than the bond coupon, the “profit” from the arbitrage might be viewed as gross income to the issuer, eroding the tax efficiency of the refunding.

To prevent abuse, the IRS limits the amount of advance refunding on certain bonds. An issuer can engage in only one advance refunding per original bond (with limited exceptions). The purpose is to prevent chains of refundings, where an issuer repeatedly issues new bonds at lower rates to refund older bonds, deferring maturity indefinitely while extracting arbitrage rents.

Additionally, the escrow account is subject to the “negative arbitrage” rules. If the Treasuries purchased for escrow yield more than the refunding bonds, the excess must be rebated to the federal government. This removes the temptation to arbitrage the tax advantage itself.

Market timing and refinancing cycles

Advance refunding bonds are often issued in waves during favorable rate environments. When the Fed is cutting rates (2008–2010, 2019–2020, 2023–2024), issuers flood the market with refunding bonds, knowing that the arbitrage window is open. These waves can be visible in municipal-bond issuance calendars.

Investors should be aware: when refunding volume is high, new-issue concessions (yields) can be generous as underwriters compete for allocation. Later, as the refunding wave passes and rates stabilize, older refunding bonds may trade at a discount if new issues are being sold at wider spreads. The timing risk cuts both ways.

Who holds advance refunding bonds

Because advance refunding bonds are backed by an escrow of Treasuries, their credit risk is minimal—they’re as safe as Treasuries, assuming the issuer doesn’t disrupt the escrow (which would be a major breach and rare). Accordingly, they trade on call-date yield and Treasury spreads, not on issuer credit analysis.

Bond-etf managers and mutual-fund portfolios hold them as part of broad muni allocations. Conservative investors favor them because the escrow-backed structure removes issuer credit risk. But investors should understand: you own the refunding bond until the call date, at which point the investment ends. There’s no optionality for the investor—the call is the issuer’s decision, and it’s effectively guaranteed (assuming the escrow is intact).

The modern landscape: tax law constraints

Tax reform in recent decades has tightened the rules around advance refunding. The Tax Cuts and Jobs Act of 2017 eliminated the ability to issue “advance, advance” refunding bonds—that is, refunding bonds issued more than 90 days before the first call date. This shortened the refunding window and reduced the arbitrage opportunity.

As a result, advance refunding is less common than it was in the 1990s and 2000s, when long refunding windows allowed aggressive refinancing strategies. Today, most advance refunding is limited to the 90-day window before call dates, which constrains both the magnitude of arbitrage and the volume of transactions.

Despite these constraints, advance refunding remains a staple of municipal finance, particularly for large issuers with significant debt portfolios. A state bond authority or a major city might conduct advance refundings quarterly or annually, capturing rate declines and reducing long-term debt service.

See also