The muni market was moribund last week in the wake of Superstorm Sandy. Almost all deals were postponed as both buy-side and sell-side participants were sidelined by this terrible storm. We take a look at the muni market post-Sandy from three different perspectives: market technicals, credit, and future developments.
Last week saw most issuance postponed. For example, the state of New Jersey cancelled a $2.6 billion Tax and Revenue Note issue. Many other issuers cancelled deals and are bringing them back this week or next. The week after next is Thanksgiving and is traditionally very quiet.
Market participants tend to be wary of buying credits from affected areas in the period immediately following natural disasters such as hurricanes and earthquakes. Think of this as group apprehension: investors waiting for someone else to make the first move. This is usually reflected in higher yield spreads on bonds issued from affected areas. We witnessed this phenomenon in a number of different incidents: the San Francisco/Oakland earthquake of 1989, Hurricane Andrew in South Florida in 1992, and Hurricane Katrina in the gulf area, especially New Orleans, in 2007. These are only three very large examples; there are many more small ones.
In all cases there was spread widening in almost all credits, and the more affected the credit – think of New Orleans Water Authority – the wider the spread. In most cases the spread widening that initially occurred after the crisis started to narrow as the yield spreads of the affected credits reverted to the mean. Generally speaking, incremental yield in the face of a natural disaster has been an opportunity.
Superstorm Sandy is the most damaging storm to ever hit the New York/New Jersey area. Connecticut, Massachusetts, Rhode Island, Delaware, Pennsylvania, Maryland, and Virginia were affected but much less so. New Jersey is particularly affected, with 50% of north Jersey homes still without power one week after the storm. The long-term record of municipalities dealing with natural disaster is very good. Most of the time there has been little or no impact on bond ratings of issuers. Moody’s states that budget strains are likely to come if recovery costs exceed budgeted contingencies and there are lags in aid from higher levels of government.
There are usually two offsetting effects from natural disasters on issuers. The first is an initial sharp decline in demand and output. Think about fewer people using bridges, roads, tunnels, or subways, as in New York, or fewer people driving and paying tolls or gas taxes, as in in New Jersey. Or the attendant costs of cleanups and restoration being imposed on the many municipalities in the New York area, especially in Long Island and in northern New Jersey.
Generally speaking, in federally-mandated disaster areas, the federal government pays up to 75% or more of emergency costs, with the state having a hand in allocating the non-federal share. Certainly, in the New Jersey and New York areas, the risk is that state and local municipality costs are going to be much higher than original estimates. Fund balances will be drawn down. There is also the prospect of higher mortgage delinquencies in New York and New Jersey, as homeowners who are already somewhat strapped face higher costs of recovery plus likely delays in insurance reimbursements. From a pure bond perspective, bonds that are revenue-oriented (tolls, fare boxes, sales taxes, etc.) will feel the most immediate impact. Moody’s states that debt-service coverage may decrease due to lost revenue during periods of inoperability and high cleanup costs for mass transit systems. Also vulnerable are debt classes such as sales taxes, special taxes, and housing bonds, due to appropriation risk and revenue cyclicality.
The second, offsetting effect is the bounce-back effect. State and local governments rebuild damaged or lost enterprises, resulting in a boost to output, often times in excess of pre-disaster levels. We have seen this time and time again, and there is no reason to think it won’t happen in the areas affected by Superstorm Sandy. One of the issues, of course, is that much of the area was really just starting to recover from the financial crisis.
We maintain a bias toward larger credits with strong insurance risk-management practices and contingency plans. We use as a case study the Metropolitan Transportation Authority (MTA), a public provider of mass transit that services the New York City area. The transit system’s large cash balances and reinsurance portfolio act as a financial cushion for bondholders during prolonged service stoppages. As some of our readers can attest, the gridlock experienced in New York City during the last few days, post-Sandy, highlights the essentiality of this system
Before the storm, MTA made necessary preparations, such as moving 1000 railcars and locomotives to higher ground. Sandbags were deployed, as well. As of now the Brooklyn-Battery tunnel is the only tunnel not in operation; however this accounts for only 6% of revenues. Most MTA train lines are operational, including the Port Jervis line and New Canaan branch, which has had to substitute bus service for rail service. The Metro North line is an exception.
As a firm, preservation of a client’s capital is the framework from which we make all decisions. When analyzing a credit we examine possible future events and their likely outcomes. For MTA, a credit we favor, this entails an analysis of bondholder safety during a possible interruption in system operations. The MTA’s largest revenue bond credit, the transportation revenue credit, is a gross revenue stream pledge with 10x debt-service coverage. The bonds are senior to operations and maintenance expenses. The system has $1.3 billion cash on hand and can operate for five weeks without revenues. Total non-reinsured and non-FEMA exposure is $180mm. The system has a $30mm-deductible pre-reinsurance and $150mm maximum liability, concentrated in Flood Zone A, the area of mandatory evacuation designated by Mayor Bloomberg. These metrics and others that indicate an adequate bondholder safety margin formed the basis of our decision to hold this credit, and possibly add to our position at cheaper levels.
It makes sense that we will see more municipal bond issuance to help rebuild the areas affected by Sandy. In addition, it would not surprise us to see a consensus-building effort in the new Congress, or even in the lame-duck Congress, after the election. Though a return to issuing Build America Bonds is a long shot, there is a case to be made for issuing them on a regional basis to help towns, counties, and other affected jurisdictions rebuild. This could also be done through a larger regional bond bank structure that would allow smaller municipalities to borrow without having to absorb higher issuance costs due to their small size. In addition, Congress could also allow municipalities to advance refund existing higher-coupon REFUNDING issues. This was allowed during the post-financial-crisis period and certainly should be called for now. Allowing banks to reenter the arena and buy CURRENTLY issued bonds and deduct the carrying costs (as with 2009 and 2010 issues) would also help to lower costs for municipalities.
Nationwide, the storm’s effects on economic growth are expected to be similar to those expected on the local level. An analysis by Goldman Sachs points to two important and offsetting effects on nationwide economic activity: sharp declines in demand and output in the short term, followed by a boost back to pre-disaster output and baseline growth. Goldman estimates a ¼-½ point reduction of GDP growth in the 4th quarter. The storm will likely have larger effects on higher-frequency indicators to be released over the next two months, such as retail sales, jobless claims, and nonfarm payrolls. Goldman states that during this time sentiment indicators such as confidence may paint a more accurate picture of the economy’s growth.
A separate analysis by Eric Strobl, The Economic Growth Impact of Hurricanes: Evidence from US Coastal Counties, published by Ecole Polytechnique in Paris, suggests that a county’s annual economic growth rate will fall by 0.45%, on average, in response to a large hurricane strike. This is not insignificant, as economic growth in the NY-northern NJ-Long Island MSA has averaged only 1.43% annually since 2001.
Sandy was quite a wallop for the area, and it will be, at a minimum, weeks and probably months before anything approaching normality returns. But we do feel that higher yields resulting from post-Sandy developments will most likely be an opportunity. We will continue to monitor and keep readers informed of future developments.