How Wall Street 'Innovations' Cost Taxpayers Millions
As citizens and taxpayers, we should be asking our city council members and school board trustees tough questions whenever they depart from the time-tested model.
What do the City of Chicago and Jefferson County, Alabama, have in common with Riverside, California, and a school district north of San Diego? These local governments have lost millions of dollars by using creative municipal finance. And if citizens around the country aren’t vigilant and outspoken, their city, county or school district may become the next victim of an unnecessarily complex bond deal.
Perhaps the worst victim of municipal financial “innovation” was Jefferson County, Alabama, which filed for bankruptcy after its financing arrangements known as interest rate swaps blew up. With an interest rate swap, a borrower can issue variable rate bonds while still paying a fixed rate of interest — a transformation achieved through an arrangement with a bank. After suffering a series of rating downgrades, the City of Chicago paid $270 million to close out swaps and convert its variable rate debt to fixed.
Why bother with such complicated deals in the first place? Bankers who promoted interest rate swaps argued that municipal issuers would have lower interest costs overall by borrowing at variable rates. But fees banks collected for arranging the swaps offset these savings. Also, because bankers are more knowledgeable about swaps than politicians and government finance staffers, the terms and conditions of these deals often protected banks at the expense of borrowers.
Here’s another example of how innovative finance benefitted the financiers more than taxpayers. One product sold as a way of lowering interest costs was called the Auction Rate Security (ARS). This instrument allows cities to meet long-term financing needs in the money market. Every four weeks the bonds are rolled over to the money market participant willing to receive the lowest interest rate over the next four weeks. Since short-term interest rates are usually lower than long-term rates, the city saves money.
This seemed like a brilliant idea — until it stopped working. Money market funds don’t want credit risk because they have to be ready to redeem shareholder funds at any time. Thus, they invest only in AAA-rated securities. Most U.S. local governments aren’t rated AAA. The way they could get into the auction rate market was to buy a municipal bond insurance policy from a specialized, “monoline” insurer like Ambac or FGIC.
These insurers were rated AAA, but, by early 2008, investors realized that they were no longer safe because the companies had also insured toxic mortgage-backed securities. Demand for ARS dried up and many auctions had no bidders. But the banks that created ARS and ran the auctions had protected themselves: When auctions failed, they were entitled to receive a penalty rate from the city of as much as 20 percent. To avoid paying this usurious rate month after month, many cities refinanced their ARS with traditional municipal bonds — paying additional origination costs to financial intermediaries in the process. Riverside, a city east of Los Angeles, lost over $12 million when the ARS market collapsed.
Ultimately, Ambac and FGIC went bankrupt, despite their AAA ratings. Every municipal bond insurer failed or suffered large credit rating downgrades during the financial crisis. In retrospect, we can see that the whole municipal bond insurance industry arose from misaligned credit ratings. The cities had less credit risk than the insurers, yet they had lower ratings. Some observers — most notably Bill Ackman — realized that monoline insurers were overrated well before 2008, so this rating agency error cannot be dismissed as a case of 20/20 hindsight.
The municipal bond insurance industry started in the 1970s and peaked shortly before the financial crisis, when more than half of all new municipal bonds carried insurance. In 2007, government bond issuers paid more than $1.5 billion in bond insurance premiums — for an insurance product that often proved worthless.
By 2010, only a handful of new bonds were being insured, but then the industry began to recover. Two insurers have achieved AA ratings from S&P and AA+ ratings from Kroll, an industry upstart. In 2015, over 6 percent of new municipal bonds were once again being insured.
For a new Mercatus Research paper, Dr. Kenneth Kriz and I calculated “All-in Total Interest Costs” for general obligation bonds issued by California school districts rated AA and below (All-in TIC is the municipal bond market equivalent of Annual Percentage Rate, considering both upfront origination costs and periodic interest expenses). About half the sample was insured. After adjusting for ratings, deal size, market interest rates and other factors, we found that districts purchasing insurance did not achieve statistically significant cost savings. The insurance premiums paid by the school districts offset the benefits of lower interest rates, so buying insurance was not worthwhile.
Since there is no record of a California school district defaulting on general obligation bonds, I question whether there is any default risk that needs to be insured against. Debt service payments for the bonds are generated from a separate, dedicated property tax levy. If a district gets into financial trouble, it cannot use bond tax revenues to balance its books.
Dr. Kriz and I found one factor that greatly influenced APRs paid by these districts: whether they used Capital Appreciation Bonds (CABs). CABs are zero coupon instruments, meaning there are no annual interest payments, just a balloon payment at maturity. But because these bonds accrue interest on interest, they are more expensive over the long term.
School districts and other borrowers use CABs to lower financing expenses in the years after a bond deal launches, at the cost of much higher payments years in the future. This is not merely a way to kick the can down the road, but a way to maximize the size of bond issuances.
California state law limits the amount of debt service a school district can incur as a percentage of assessed property values. When districts and their financial advisors structure bond issues, they assume that property values will rise rapidly over the life of the deal. Thus, by backloading more debt service toward the end of the bond’s life, they can maximize the amount of money they borrow — violating the spirit but not the letter of debt service limits. In the process, they effectively mortgage the homes of future residents. Poway Unified School District, north of San Diego, is perhaps the worst victim of CABs: A few years ago, the district issued $179 million of bonds carrying lifetime debt service payments of $1.27 billion.
When it comes to municipal borrowing, the KISS (Keep It Simple Stupid) principle should normally apply. Plain vanilla structures made up of uninsured, fixed rate, current interest (as opposed to zero coupon) bonds have proven themselves across decades of municipal finance. They are readily understood, have relatively low upfront costs and generally do not produce nasty surprises. As citizens and taxpayers, we should be asking our city council members and school board trustees tough questions whenever they depart from this time-tested model. If they can’t provide convincing answers, it may be time to elect someone else.