'Decision Tree' Approach Takes Multiple Business Interruption Risks Into Account

Losses attributable to business interruption can account for a significant proportion of total commercial claims following a catastrophe. Indeed, for certain businesses--such as chemical plants and off-shore oil and gas platforms--business interruption losses can be significantly higher than property damage claims.

Historically, catastrophe models estimated these losses primarily by calculating the time needed to repair different degrees of building damage--and making the assumption that at low-to-moderate levels of damage, most businesses remain in operation without incurring any significant business interruption.

However, common sense and underwriting experience would dictate that this is not true for many types of businesses.

To achieve reliable estimates of probable business interruption loss, catastrophe models must go beyond simulating business interruption that results solely from direct property damage. These models must take into account both the wide variety of business activity and the range of decisions business owners and managers must make in the aftermath of a catastrophic event.

The central variable in estimating business interruption losses is "downtime." In general, downtime is understood as the number of days it takes before a business can return to full operation following a catastrophic event.

Downtime attributable to physical damage is categorized as "direct business interruption." The insurance industry categorizes other contributors--damage to a dependent building (a building the business depends on for its own operation), infrastructure lifeline and utility disruptions, loss of market share, etc.--as "indirect business interruption."

Modeling business interruption losses for an individual policy or a portfolio of properties must include a characterization of the hazard where each property is located (wind speed, in the case of hurricane, or ground motion for earthquakes, etc.).

Similarly, the vulnerability of the commercial property (characterized according to construction type, occupancy and other physical characteristics of the building) must be estimated with respect to the hazard. The resulting physical damage to the insured structure could be sufficiently severe to cause a suspension of business activities--and loss of business income because of the suspension.

However, business interruption can result from a wide range of circumstances, including when there is no physical damage to an insured structure. Thus, to estimate insured business interruption losses, the full amount of downtime attributable to all causes must be estimated.

In addition to levels of property damage, business interruption downtime will be affected by other property characteristics such as size, complexity and function. For example, large buildings generally take more time to repair than smaller buildings.

Also, the structure and contents of some enterprises can be more difficult to repair or replace, and thus require more downtime. For example, a historic building will take longer to restore to its pre-catastrophe conditions than a typical storage facility.

Similarly, every business has an intrinsic "resiliency" to the impact of business disruption. A software company, for instance, might be able to relocate relatively easily without suffering significant business implications, while a hotel cannot.

Beyond the characteristics of the business itself, management decisions play a very important role in determining the extent of business interruption decisions. Two businesses of the same size and complexity that have suffered the same level of property damage can respond differently. One business owner could make the decision to re-open their business early, while another owner could make the decision to re-open only after 100 percent of the restorations have been made.

This uncertainty can be modeled along the lines of the accompanying decision tree diagram, which lays out a formulation of the flow of decisions that management must make under conditions of uncertainty. At every branch juncture, the decision that is made can impact downtime--in some cases quite dramatically.

Additional business characteristics that need to be taken into account in estimating downtime include:

o Whether a business has multiple locations or a single location.

o Its degree of dependence on local or regional resources.

o Its contracts with suppliers and distributors.

o Its dependence on infrastructure lifelines.

The contingencies noted above, as well as additional ones, can be captured within the modeling framework using the same decision-tree approach. This uncertainty must be captured in a model for risk to be assessed comprehensively.

Business interruption is a significant component of commercial insured losses in the aftermath of a catastrophic event. To accurately estimate those losses requires a modeling framework that allows underwriters to do two things:

o Take into account both building and business characteristics.

o Capture probabilistically the effects of the wide range of decisions that can be made by business owners.