Early last year, I visited the USS Midway Museum in San Diego. The museum offers tours of the decommissioned aircraft carrier, led by docents, many of whom served on the USS Midway or another vessel. As I toured the bridge and entered the chart room, the docent recalled the various measures for determining the carrier's location and their important purpose of ensuring that the carrier would not run aground. These measures include the use of sonar to measure depth, the use of radar and visual positioning in relation to landmarks, and the use of special instruments to compare the angle of the horizon in relationship to the stars. Today's measures would also include GPS. Each of these measures has its limitations, such as the accuracy of the instruments used, the presence of underwater objects that may not yet be charted, or the dependence on outside sources such as satellites, but, when compared as a whole, give a fairly accurate range of position.

Actuaries use various methods for determining the ultimate value of claims. Development methods based on incurred losses, paid losses and case reserves are typically used to estimate the ultimate value of claims. Loss projections based on pure loss rates may be considered for "green" years. Likewise, Bornhuetter-Ferguson methods, which are a blend of development methods and loss projections, may be used. Each of these methods has its strengths and weaknesses. Thus, one method may be preferable over another in a certain circumstance. For example, development methods applied to case reserves may be used for policy years where most claims are paid and closed, perhaps with overall closures earlier than those of the average year. Additionally, an average of several different methods may be used. Consider an incurred development method which may be biased high due to the earlier-than-normal establishment of large case reserves on a group of claims. A weighted average of various methods, which mitigates the bias created by the incurred method in particular, may be used. The goal is that these methods, with consideration to the strengths and weaknesses of each, give a fairly accurate range in ultimate losses.

The above actuarial methods are based on historical experience and known loss information to date. But what if little historical information is known, as is the case for many low-frequency and high-severity type risks? Or what if the types of risks occurring in the past do not provide an accurate view of the current or future situation? Estimates for these risks may include the use of industry and similar entity data. This process often involves discussions between the actuary, management and other consultants regarding the probability of an event, the potential cost if an event happens, and any nuances in the company's unique risks that may not be reflected by industry sources.

A question recently posed in a discussion was "Doesn't the company's management team have a better understanding of their own risk than an actuary?" In some cases, especially in the situation where limited historical data is available, they may. There is also evidence to suggest, however, that many people over or underestimate their own risk by large margins. An actuary and the management team most likely get the best results when the management team’s risk understanding is placed within an actuarial analytical framework.

It is crucial for companies to review the actuarial results in light of their understanding of the risk and notice areas where internal estimates align with or deviate from actuarial estimates.  This often forms the basis for discussion. Companies must also understand the limitations of the estimates produced from each method. Often a range in estimates is provided by the actuary. This range defines a set of reasonable outcomes, based on the range in methods and limitations of the data.

The discussion between a company and their actuary may be done several different ways. In some instances, an actuary will issue an initial draft based strictly on the numbers, without consideration to any qualitative information provided by the company. Then after discussion, the actuary may find that management's expectations align with initial actuarial expectations or will subsequently refine the initial draft estimates. In other instances, management will choose to discuss particular items with their actuary on the front end. These initial discussion items may include internal changes that were made over the past year that may impact development, items contributing to trends in the data, or particular claims that may need special attention due to their size or nature in relationship to the “average” claim. The actuary will weigh each piece of information, both qualitative and quantitative, in order to determine their best estimate of the ultimate value of losses.

The USS Midway requires specialized methods used by a community of experts trained in specific areas to accomplish its overall purpose. Likewise, companies should depend upon the most reliable information and expertise available to accurately estimate the ultimate loss value for their unique risk exposure. For more information regarding the various methods used by actuaries and the types of qualitative information often considered in forecasting losses, view our video here or contact Lori Ussery at lori@sigmaactuary.com.