Editor's note: Hemant Shah is CEO andpresident of catastrophe modeler Risk ManagementSolutions.

Superstorm Sandy was an extraordinary weather event,particularly given the destruction produced by its large stormsurge in relation to the wind damage. Two weeks following theevent, RMS issued its first and only estimate: insured losses wouldfall between $20 billion and $25 billion. So what goes intoproducing loss estimates and what have we learned about the art ofpost-event loss estimation from previous events?

Balancing the need for reliability with timeliness, we issuedour loss estimate two weeks after Sandy made landfall. Our policy,based on extensive client consultation and a number of years ofhard-earned experience, has revealed there is little benefit inrushing to publish and publicize an industry-loss estimate. It canbe unhelpful if clients are forced to react to a modeler's estimatebefore we are even able to support their own internalestimations.

It is not just about the “number,” but rather the insight thatcan help responsibly inform our clients' judgment of their losses.Moreover, each storm poses unique challenges; reliable lossestimation takes time. It is certainly not helpful to issue anestimate only to continuously revise it, or to issue a range sowide that it fails to provide actionable information.

Within hours of Sandy's landfall, it was clear this was not atextbook hurricane: an unusual track; a large, diffuse andtransitioning windfield; and a catastrophic storm surge despiteborderline winds were some of its characteristics. Several dayspassed before floodwaters had receded and power and transportationresumed. Five days after the storm made landfall, somejurisdictions were still pumping water from flooded areas, andpower was still out to tens of thousands of households. Previousmajor storms have shown that lifeline and infrastructure-recoverytime can ultimately have a powerful impact on the total insurancebill. Combined with other drivers of loss, such ascontingent-business interruption, it clearly takes time to fullyunderstand and absorb the impact of these major events.

Putting State-of-the-Art Modeling to theTest

Superstorm Sandy's large and destructive storm surge was farmore damaging than its wind impact. The storm's broad windfieldrapidly diffused as it made its way inland and, technically, Sandywas no longer a hurricane at landfall (hence its popular naming as“Superstorm” Sandy). Yet, the storm surge was equivalent to aCategory 2 hurricane along the New Jersey coastline. This was notthe first time the U.S. had experienced such a devastating andapparently disproportional storm surge: in 2005, Hurricane Katrinawas a Category 3 hurricane at landfall with a Category 5 surge.

We were able to gain early insights into the potential magnitudeof Sandy's surge damage by using the latest state-of-the-arthydrodynamic storm-surge model technology, in real-time, withinputs of the storm's characteristics from days before landfall.This advancement in storm-surge modeling has only been madepossible in recent years.

Complex hurricanes – such as those with extremely largewindfields or those that decay but remain intense prior to landfall— as well as complex coastlines like the Hudson estuary on the NewYork/New Jersey border, have historically challenged the industryfrom a modeling perspective. Old-style surge models significantlyunderstated the hazard and related losses. Learning from similarpast extreme weather events, we incorporated a new way of modelingthese unique physical features into version 11 of the RMS U.S.Atlantic Hurricane Model.

The model encompasses a completely new way of analyzing stormsurge, using state-of-the-art dynamic modeling to accuratelyrepresent a disproportionately large storm-surge risk compared towind impact.

On the Ground Research

In addition to this modeling effort, we examined thousands ofdamage reports and constantly evaluated diverse feedback loops. Inthe immediate aftermath of the event, we deployed two teams ofengineers on the ground to conduct detailed surveys of the impactsalong hundreds of miles of coastline. Two weeks later, a third teamof engineers was deployed to assess the recovery efforts.Meanwhile, our catastrophe modelers in London and California workedaround the clock to analyze the incoming data, integrating it withthe modeling insights to generate information that insurers andreinsurers could run against their portfolios of accumulatedexposures.

Drilling Down to Increase Precision

To calculate a more precise estimate of the losses, wedynamically modeled the surge street-by-street, distinguishing theflood risk by property based on the elevation and proximity of eachbuilding to the water's edge. Complex interrelationships betweenextensive power outages, disruption from flooding, widespreadcoastal property damage and the closure of transportation systemsprovided additional insights, as did other factors drivingpotential post-event loss amplification.

It takes significant time and effort to collect and assess thiswealth of data to ultimately develop a credible loss estimate. Sohow robust was RMS' range? Four months after our estimate wasreleased, our range remains credible. Although, as with all extremeweather events it will take some time for all the claims to beresolved, $20 billion – $25 billion is consistent with currentreporting from re/insurers across the industry.

While we are confident in our estimate, we recognize this is nota time for complacency. Every catastrophe provides crucial insightsand opportunities to learn. Our estimate was accompanied by atechnical whitepaper that not only made our assumptionstransparent, but also provided a roadmap for further learning. Aswe mine the claims data, we will learn more and incorporateimprovements into our models. Ultimately, it is not about'the number'. As with our modeling in general, our response toSuperstorm Sandy reflects our commitment to transparency andresiliency, by empowering our clients to take control of theirmodeling environment and make decisions that are informed bymodels, not dependent on them.