Accounting For Catastrophe

Accounting For Catastrophe

By Tom O'Brien

Catastropheslow frequency, high-impact events such as hurricanes, earthquakes, tornadoes and terrorist attackscan bring down buildings, disrupt supply chains, devastate a workforce, and cripple or even put a company out of business.

Despite their infrequency, the magnitude of the potential losses from these events obligates corporate risk managers to consider the possibilities.

Historically, the greatest impact to corporations from catastrophes has been on property insurance. However, in this new risk environment, the risk management community has been made painfully aware that some events can lead to catastrophic losses from workers compensation claims and loss of life.

A hard insurance market makes the risk from catastrophes even more challenging to manage, forcing corporations to make tough decisions about how much coverage to buy, whether to increase deductibles, or assume some or even all of the risk internally. Risk transfer decisions such as these can be made rationally, however, only if they are informed by an understanding of the full range of potential losses and their probabilities.

Now, tools are available that provide corporate risk managers and chief financial officers with just that kind of information. Catastrophe models, which have been widely used by the insurance industry for more than a decade, are also the best way for risk managers to understand their catastrophe risk.

The combination of computer modeling technology and advanced scientific understanding enables sophisticated loss analyses that just cant be matched by traditional probable maximum loss studies. Catastrophe modeling helps risk managers prepare for extreme events in the following ways:

(1) Provide detailed loss analysis: A detailed, location-specific loss analysis will provide much deeper insight into property and workforce exposure to extreme events. By quantifying the range of potential losses, a risk manager is better able to plan for, mitigate and transfer risk. Results can also be used to negotiate insurance rates at renewal.

(2) Identify large loss drivers: Catastrophe modeling will let a corporation know if a small percentage of their properties contribute disproportionately to their risk exposure. Consideration can be given to the installation of loss mitigation devices at high-risk properties, minimizing the impact of an extreme event on the property and employees, and potentially reducing insurance premiums.

(3) Understand large loss scenarios: Corporations need to know what their largest losses might be and the chances of those losses occurring. In addition to using this information for insurance buying purposes, it is also good information to have for disaster recovery planing.

(4) Optimize risk transfer: When researching transfer options, it is important to know how different approaches can impact the bottom line. Catastrophe modeling helps corporations evaluate multiple approaches, including consideration of captives, risk pool formation or other means of alternate risk transfer.

When transferring risk through traditional insurance markets, approaching renewals with a catastrophe loss analysis provides considerable negotiating leverage with your underwriter.

By incorporating the latest scientific research about the frequency, location and behavior of catastrophic events, whether natural or man-made, models are able to replicate the range, severity and frequency of potential events.

Applying expertise in scientific, engineering and other disciplines in the computer modeling process enables dramatic reduction in the uncertainty of catastrophe loss estimation.

For the corporate risk manager, a little bit of data will go a long way in replacing the subjective, “rule of thumb” approaches to estimating potential losses. Reliability of the resulting estimates depends on the sophistication of the model and the quality of the input data provided by the corporation.

To successfully perform modeling for a property portfolio, location information, basic structural details and replacement values for each building are required. Additional data is required for workers compensation and life analyses.

Looking at how each of these data elements is incorporated into the study provides a better sense of how catastrophe modeling works:

Location: First the model determines the latitude and longitude for each property address. Using advanced statistical techniques, the model then simulates a set of possible extreme events and overlays the location or path of those events to determine the destructive force on each property.

Also considered at this point are the geographical and geological attributes of property locations to determine what kind of amplifying or dampening effect they could have on the destructiveness of the event. For example, hurricane exposure is exacerbated by proximity to the coast and soil type has a large bearing on loss potential from an earthquake.

Structural characteristics: Mathematic equations known as “damage functions” are developed for each combination of construction type, such as wood, masonry, concrete and steel; occupancy type, such as residential, commercial and industrial; and peril, such as earthquake, extreme wind and terrorism. These “damage functions” are based on engineering studies and historical losses from actual extreme events. By combining an understanding of event severity at a particular location with an understanding of a buildings structural resiliency and responses, realistic assumptions can be made about the range of potential damage, on a percentage basis, that could result from a possible event.

Replacement Values: By applying the percentage of structural damage to a buildings replacement value, the loss from that particular event can be described in dollar terms. But loss assessments are not limited to the structure. Building content and business interruption losses can also be derived from catastrophe modeling.

Beyond property: Using property damage as a foundation and incorporating details about the number of employees and shift rotation for factories, determinations can be made regarding injuries and fatalities resulting from an extreme event to estimate workers comp losses. To fully analyze a life portfolio, additional information is needed that includes assumptions regarding employees whereabouts outside of the office.

One of the primary benefits of the computer modeling process is the ability to run extreme event analyses repeatedly with a level of randomness that replicates reality. The models simulate thousands of event scenarios that might happen each year. With no two years exactly alike, the range of potential losses can be quantified in meaningful terms.

As the hard insurance market continues to put a strain on corporations, risk managers must understand their companys exposure to catastrophic losses; be able to convey that understanding to the broker, underwriter and supervisor; and determine the ideal mix of retention, transfer and mitigation of risk. The output of a catastrophe modeling analysis is an invaluable tool in the execution and communication process.

The level of detail and quantification of the various probabilities of loss enable the risk manager to develop a property, workers comp or life program that truly matches the risk tolerance level of their company. Being able to differentiate between the events and probability of a tolerable loss and the most extreme losses allows the risk manager to select the deductible and coverage limit that is most cost effective.

Catastrophe modeling is a decision support tool that enables the risk manager of the 21st century to identify, quantify and manage extreme event risk. Incorporation of modeling into a risk management program helps the risk manager keep ahead of growing insurance demands and keep the company protected.

Tom O'Brien is a risk-modeling consultant for AIR Worldwide Corporation in Boston. AIR is a catastrophe and weather risk modeling company.

Reproduced from National Underwriter Edition, March 10, 2003. Copyright 2003 by The National Underwriter Company in the serial publication. All rights reserved. Copyright in this article as an independent work may be held by the author.