In the same article, Morgan Stanley provides this observation:
"With $500 billion-plus of capital ... we expect the (property and casualty) industry is once again well prepared to pay all Frankenstorm insured losses," Morgan Stanley analyst Gregory Locraft said in a report on Monday, using the nickname for the Sandy-nor'easter combo.
Rather than predicting what's about to happen, we'll point you to some reading on events in the rearview mirror. You can learn a lot about insurance just through how the industry has evolved following natural catastrophes.
- David Sanders frames up the challenge of insuring natural catastrophes in his 2006 paper, "The Price of Civilization." Sanders builds off of something Voltaire said after witnessing the wreckage following the 1755 Lisbon earthquake: "Is this the price mankind must pay for civilization?" Sanders tries to answer this question by looking at how we pay the price that natural catastrophes extract and examining who bears the brunt of that expense. Here's a helpful excerpt:
To assess how dangerous an insurance risk is, it is often convenient to apply the Pareto parameter. This rule--commonly known as the 80/20 rule--states that 20% of the claims in a particular portfolio are responsible for more than 80% of the total portfolio claim amount. With the Pareto parameter as a baseline, we can assess a portfolio's vulnerability. If a single event can spell financial ruin, there may be a problem.
Hurricane data in the Caribbean indicates that insurers can make profit for a number of years, and then find themselves hit by a "one-in-1,000-year" hurricane, which swallows up 95% of the sum insured in one go. For example, when Hugo hit the U.S. Virgin Islands, the total cost of the loss for residential property insurers was equal to 1,000 years' worth of premiums.
The regulators of the insurance industry generally target a one-in-100-year to one-in-200-year insolvency level. They do not cater to the one-in-1,000-year event. Typical solvency levels for major developed insurance markets that cover catastrophes are on the order of three to six times the cost of a once-in-a-century event. However, Katrina-type losses are not one-in-100-year events. Recent history indicates that they are more like one-in-five-year events, which means every five years the insurance industry can expect a $50 billion loss [most recently Katrina4].
- Joel Chansky's informative piece, "Liability cat bonds: You won't see this on Animal Planet" examines the evolution of catastrophe bonds as a response to natural disasters, starting with the birth of cat bonds following Hurricane Andrew. This sums up the evolution:
There is a finite amount of reinsurance capacity, with billions available worldwide. A company might find a reinsurer willing to insure a $100 million dollar loss, but can they find one willing to cover single-year losses that exceed that threshold? It is difficult to adequately spread around risk and fill a reinsurance dance card when aggregate losses reach ten digits, which is why "securitizing" insurance risks in the capital markets has become an attractive option. Cat bonds were born in the early 1990s following Hurricane Andrew--a seminal event because it revealed the limits of reinsurance. Since then, companies, insurers, and reinsurers have used cat bonds to provide another layer of insurance, often protecting against an insurer's unlikely third or fourth hundred-million-dollar loss--the one that finally exhausts insurance or reinsurance capacity.
- Companies looking to triage the flood of claims that will come their way in the next two to five days might look to text mining as a way of comprehending Sandy's claims. Phil Borba's article on text mining shows how insurers can analyze claim notes to better screen and triage their claims.
- Matt Chamberlain's recent article examines how something called geocoding can lead to more precise pricing for homeowners policies in hurricane-prone parts of Florida.
Although it may seem like defining the "coastline" is clear-cut, it is actually quite ambiguous when considering a property's exposure to a hurricane. Does the coastline follow bays, such as Tampa Bay? Does it follow barrier islands? Does it follow rivers and, if so, how far? After a company decides that it should organize its territories based on distance to the coast, that company's first instinct may be to use an existing coastline. However, such a coastline may not be suitable for the purpose. Off-the-shelf coastlines may follow many small-scale features that do not, in fact, affect hurricane risk.
Maybe we'll have to publish a follow-up article analyzing the potential for geocoding in Lower Manhattan.
Wherever you are, stay safe and dry. And if you are lucky enough to maintain electricity and Internet, feel free to post related reading below.