The Federal Reserve released on May 18th “Liquidity Coverage Ratio: Treatment of U.S. Municipal Securities as High-Quality Liquid Assets, a Notice of Proposed Rulemaking with Request for Public Comment”. General Obligation municipal paper would be included in Level 2B assets, although there are some additional proposed limitations that really squeeze down eligible assets. Lobbyists were hoping for 2A and will be disappointed.
Muni issuers and bank investors have lobbied the Fed to include municipal bonds as part of the HQLA bucket when calculating LCR. From a Risk article dated May 28, 2014 “US muni treasurers warn LCR could crimp spending”:
“…It may seem unlikely, but if the residents of Cary in North Carolina notice public verges becoming unkempt, or pilots using Boston’s Logan International Airport start seeing potholes, there could be a single cause – a rule drawn up in Switzerland and implemented by the US Federal Reserve Board…This rule is the liquidity coverage ratio (LCR). If last October’s Fed draft version is introduced unchanged, bonds issued by US states, cities and towns will not count towards the new liquidity buffers that banks are required to hold. This will reduce the value of the bonds and demand for them will drop, the argument goes. With banks currently holding around 10% of the stock of municipal debt, issuers fear they will have no choice but to pay higher coupons…”
The muni bond market has many issuers with individual issues that can be fairly small. While there is an institutional and bank investor component, it is a popular investment for high tax-bracket individuals. According to the Fed there can be high issue investment concentration too:
Given the diffused nature of the issuance – it seems like every county and local water authority has bonds outstanding – and the information to assess the creditworthiness can take a lot of work to unearth, investors often depend on credit ratings to make decisions. We have seen this movie before in sub-prime. Understandably, there are questions about the depth of the muni secondary markets.
Detroit’s bankruptcy and Puerto Rico’s financial woes have been well documented. More recently Chicago has hit the news. The market seems bifurcated into very solid credits and the riskier high yield world.
So what has the Fed proposed? General obligation investment grade munis would be included in Level 2B assets. The bonds will be subject to standard Level 2B 50% haircut and are capped at 5% of total HQLA. Revenue bonds – paper that is backed by a specific project (like a stadium) and don’t carry an obligation to pay from other revenue sources – will not be HQLA eligible.
Assets to be eligible as HQLA must be liquid and readily marketable – a standard that munis did not consistently meet. One test has been an active repo market and that was always a stretch. For example, in tri-party statistics (April 9, 2015) the Fed buckets munis in “other” (along with CDOs, International, Money Market, and Whole Loans) and the category accounts for just 2.1% of tri-party collateral.
Like corporate paper included in HQLA, investment grade GO municipals must show that they have not declined by more than 20% in value during periods of high stress nor have repo haircuts on those assets gone up by more then 20% during the same period. The latter will be tricky. Tri-party haircuts tend to not change, but bilateral repos can jump all over the place. And the data on bilateral repo haircuts is not published.
Due to adverse correlation, bonds that are guaranteed / wrapped by a financial institution are not HQLA eligible.
The Fed has recognized that muni bonds can, individually, be small issues and not be very liquid. But the rules accommodate for it by looking at traded volume in GO obligations for the issuer and not the issue. Still, they limit concentration to 2 x average daily trading volume:
“…The proposal would limit the aggregate fair value of the U.S. general obligation municipal securities of a public sector entity that may be included as eligible HQLA to two times the average daily trading volume of all U.S. general obligation municipal securities issued by that public sector entity because, based on the Board’s analysis, a holding of two times the average daily trading volume could likely be absorbed by the market within a 30 calendar-day period of significant stress without materially disrupting the functioning of the market…”
Banks cannot count more than 25% of their holdings of an individual bond (defined as the same CUSIP) toward HQLA. This calculation is made before the Level 2B 50% haircut.
The headline hides the restrictions that the Fed has proposed. The Fed did not get pushed over by the lobbyists and focused on the realities of the muni market. The rule set seems pretty complex – one of those Venn diagrams with virtually no overlap – and might be a bit over-engineered. We are sure the comments (due by July 24th) will be interesting. Between the analysis that banks have to do and the different levels of liquidity caps, we can’t imagine there will be too much in the way of municipal bonds included in HQLA anytime soon.