Mispricing of US Municipal Bonds


  • Fideres finds that underwriters systematically misprice US municipal bonds (“Muni Bonds")
  • We estimate that Muni Bond issuers between 2006 and 2015 municipal bond issuers may have overpaid up to $28bn due to this mispricing

The Muni Bond Market

Outstanding Muni Bond debt in the United States numbers in the trillions of dollars ($3.8tn as of Q3 2017). The average daily trading volume of Muni Bonds between 2007 and 2017 was approximately $9bn. The chart below shows the volume of new Muni Bond issuance for the past five years.

Unlike similar securities (such as US Treasuries), Muni Bonds are not priced in a competitive bidding process. Municipalities typically ‘outsource’ the pricing of their bonds to underwriters, who generally come from a small group of large banks. The chart below estimates underwriters’ market share, based on a sample of trade and issuance data for Muni Bonds provided by Bloomberg:

Are Muni Bonds Systematically Underpriced?

Issuers will typically price their bonds generously to attract investors, so it is normal for the price of newly listed securities to rise in the secondary market after they begin trading. Nevertheless, Fideres’s findings show that the price of Muni Bonds increases far more than securities in similar markets.

We analysed a Bloomberg sample of price increases for Muni Bonds, issued between 2006 and 2015, and compared them to US corporate bonds and US Treasury bonds.The results are shown in the charts below: 

Weighted by issuance size, the average Muni Bond price increase is 163 bps, compared to 55bps for underwriters’ bonds, 33bps for Treasuries, and 64bps for corporates. This implies that there is sufficient market demand for municipalities to issue bonds at lower yields, and save significantly on borrowing costs. Underwriters are also much better at pricing their own bonds than Muni Bonds. 

While the numbers in the chart represent national averages, these conceal glaring individual cases of mispricing. Fideres has identified states whose Muni Bonds increase in price by up to 12% on their first day of trading.

Why Muni Bonds are Mispriced 

Before a Muni Bond is issued, underwriters are tasked with researching demand for the bond and pitching it to prospective investors. They then give the issuer feedback and agree final terms for the bond’s price. The underwriters also allocate the initial issuance to investors and typically retain a portion, in order to provide secondary market support. When the bonds are issued, investors pay cash to the issuer and receive the bonds, after which they trade in the secondary market.

Several reasons may explain the systematic mispricing of Muni Bonds: 

  • Conflicts of Interest – if underwriters know that the issue price is artificially low, they have an incentive to hold the bonds when secondary market trading begins, and then sell them at a profit once the price rises.
  • Use of Comparables – underwriters make their pricing recommendation based on set of comparables, such as issue prices for similar bonds. Thus, if Muni Bonds are systematically underpriced, the trend becomes self-perpetuating. 
  • Information Asymmetries – issuers often do not have the market knowledge to price their bonds accurately, and hence have to rely on the underwriters’ recommendation.

Food for Thought

Fideres' research raises important questions regarding the pricing of Muni Bonds:

  • Why are Muni Bonds generally sold through negotiated sales instead of competitive bids, when our work and other studies have shown that negotiated sale prices are more mispriced than competitive?
  • Why, in the age of electronic exchanges and exchange traded funds (ETFs), are bonds in such a large market not issues through electronic auction? 

In light of this evidence, particularly given its implications for taxpayers, it is surprising that very few issuers have taken action to improve the pricing of their own securities. 

The details of our report were also recently covered by Bloomberg and their coverage of the issues presented can be found here.

 

1The sample is restricted to non-taxable, fixed coupon bonds, with notional sizes of $50m or greater.