The MTA Procures Storm Surge Protection via the Catastrophe Bond Market

On October 28, 2012, the Metropolitan Transportation Authority was staring down the barrel of a rocket launcher. In the last hours before Hurricane Sandy made landfall, the agency steeled itself against the storm as best it could: Workers moved as many subway cars and buses as possible out of harm's way and fortified the stations and equipment they couldn't.

It didn't make much difference. When the storm hit, seawater flooded into low-lying stations in steady, salty cascades. You could have scuba-dived through South Ferry-Whitehall Street, where the subway platform was entirely submerged and the lobby was a scene from Waterworld: a lone escalator ascending from a cloudy green lake.

In December, when the agency finished surveying its corroded, waterlogged system, the MTA estimated it had sustained damage in excess of $4 billion—including $4 million worth of subway cars, $222.5 million in station restoration (much of that set aside for South Ferry-Whitehall Street), and $215 million for track repairs.

NOAA

James Franklin of the National Hurricane Center: "For New York City, it's once every 175 years for a major hurricane."

Related Stories

• Commuters Are Mesmerized By Fulton Center, Lower Manhattan's 'Futuristic' New Transit Hub
  November 10, 2014
• The G Train Is Back, So We Can All Resume Complaining About It
  September 2, 2014
• Kronos Quartet & Laurie Anderson
  September 17, 2014
• The Brooklynite Who Brought the Joys of A Transit Countdown Timer Into His Home
  July 25, 2014
• Jon Batiste and Stay Human+Hypnotic Brass Ensemble
  June 25, 2014

More About

• Metropolitan Transportation Authority
• James Franklin
• National Oceanic and Atmospheric Administration
• Weather
• Natural Disasters

Sandy was a blow to the MTA, but what happened next was, arguably, worse. As the agency tallied the damage and submitted its claims, it was becoming painfully clear that, under the current conditions, it would not be able to afford the kind of insurance coverage it had maintained when the storm hit.

Its policy was set to expire May 1.

That's how the fate of the subway system ended up in the hands of just 20 investors. If a hurricane were to descend on New York tomorrow, repairs to the subway system would be paid for with money put up by those investors, who bought shares of a so-called catastrophe bond the MTA sold this past summer.

It's not philanthropy; it's an investment. The same 20 bankrollers stand to make millions, provided another Sandy doesn't hit tomorrow or anytime in the next three years.

Before Sandy, the MTA owned an insurance policy worth $1 billion. After the storm, says the agency's director of risk and insurance management, Laureen Coyne, "It was impossible to get that kind of coverage." Even half a billion dollars' worth would have cost twice as much. The MTA was particularly concerned about its protection in the event of another flood.

"We were in a state of panic for several months," says Nora Ostrovskaya, MTA's senior manager of strategic initiatives. A handful of potential solutions were considered and discarded before one of the agency's consultants floated the idea of a catastrophe bond.

"Normally, when MTA contemplates a project, we have certain control over when we execute. With any of our big capital commitments, we can say, 'OK, let's wait until the architectural plan is fully fleshed out. Let's wait for an approval. Let's wait for this; let's wait for that.'" In this case, Ostrovskaya says, "we were under tremendous time pressure."

Without a lot of options, the MTA dove headfirst into the small, strange catastrophe bond market, where an estimated 100 investors worldwide do $16 billion worth of deals. Other wagers available in the market include whether there will be a windstorm in Europe, an earthquake in Japan, a cyclone in Australia, or major crop losses in Mozambique.

In order to sell the bond, the MTA's in-house insurer, First Mutual Transportation, enlisted a law firm to create an offshore entity dubbed MetroCat Re, located—for "various legal and tax reasons," Coyne says—in Bermuda. All the money involved in the bond arrangement goes through MetroCat Re.

Essentially, the bond is a simple high-stakes bet: If a catastrophic hurricane causes a Sandy-magnitude storm surge on or before August 5, 2016, investors lose every cent they ponied up. No hurricane, and they get all of their money back, plus a return that will top 13.5 percent.

The MTA put the bond up for sale in July, expecting to sell $125 million in shares, but the interest was overwhelming. The agency ultimately sold more than $200 million to just 20 investors. (Ostrovskaya describes the buyers not as individuals, but "pretty much specific catastrophe bond funds that specialize in this type of investment.")

The surge protection supplements the $500 million in insurance the MTA has purchased to cover perils such as wind and fire. That policy costs the MTA $46 million a year.

It's easy to see why the catastrophe bond was popular—it's an incredibly lucrative proposition. A $10 million share could pay off more than $1.35 million in just three years.

The money comes straight from the MTA, which makes quarterly payments into a trust, in much the same way that it would pay an insurance premium. At the end of three years, if disaster hasn't struck, the investors cash out.

The question, of course, is whether disaster will strike.

Long-range hurricane forecasts are notoriously imprecise. James Franklin, branch chief of the hurricane specialist unit at the National Oceanic and Atmospheric Administration's National Hurricane Center, says NOAA's seasonal forecasts are only marginally more accurate than "just predicting 'average' every year." This year, for instance, the agency predicted hurricane activity would be above average, but more than halfway through the season, activity has been well below normal.

If it's hard to forecast hurricanes, trying to predict a storm surge associated with a hurricane adds another order of magnitude to the puzzle. Says Franklin: "We don't even begin to make storm-surge predications until we're within 48 hours of landfall."