About Us
Karen Clark & Company helps senior executives and boards of directors make sure their companies have in place effective risk management processes that conform to best practices. All aspects of the process, from preparing the exposure data, to generating model output, to interpreting and using the catastrophe model results are reviewed and evaluated using independent benchmarks. Additionally, Karen Clark & Company provides unique executive briefings to inform senior management and company boards on the specific information they need to know about catastrophe risk, catastrophe models, and using model results to manage the risk.
Karen Clark & Company was formed by Karen Clark, internationally recognized expert in the field of catastrophe risk assessment and management. Ms. Clark has been helping major corporations estimate and manage potential losses from catastrophes since 1983. She developed the first hurricane catastrophe model and in 1987 founded the first catastrophe modeling company, Applied Insurance Research (AIR), which subsequently became AIR Worldwide Corporation after acquisition by Insurance Services Office in 2002.
Ms. Clark has spent over 20 years working closely with meteorologists, seismologists, engineers, statisticians and other experts to develop, expand and enhance the most scientifically advanced catastrophe models. She has developed processes to validate catastrophe models for all types of natural hazards in 50 countries. She has led the development of the models and the software applications that are used globally as standard tools for catastrophe risk assessment and management.
According to Ms. Clark, "while the catastrophe models and software applications will continue to evolve and improve over time, the bigger challenges now are in helping companies get full value from these tools. This requires a few additional steps in the risk assessment and management process."
Catastrophe Risk
Every part of the U.S. is exposed to catastrophe risk from one or more natural hazards such as earthquakes, hurricanes, tornadoes and other types of windstorms. Every geographic area could be the location of a man-made catastrophe such as a terrorist attack.
While it's virtually impossible to predict when or where the next catastrophe will occur, companies with exposure to loss need to be prepared for the types of events that could occur. They need to know the full range of possible future loss scenarios. They need tools to assess and manage their catastrophe risk.
Catastrophe Models
Many companies rely on catastrophe models to assess and manage catastrophe risk. Catastrophe models are very detailed and complex and they incorporate the science underlying the occurrences of catastrophes and the engineering knowledge to estimate the damage caused by catastrophic events. The models use statistical techniques to generate large samples of hypothetical future events. For each of these events, the models simulate the intensities by location and then estimate the damage to the exposed properties at each affected location.
Hundreds of scientists and engineers have worked on these models over the past 20 years, and the catastrophe models have become the standard tools for assessing catastrophe loss potential from all types of hazards for most parts of the world. While the models are the most sophisticated tools currently available to assess risk, the models do have certain limitations.
The models are based on many assumptions and each assumption has not one 'correct' value, but rather a range of scientifically valid values. The scientists and engineers who work on the different models make different scientific judgments at different points in time about how to implement these assumptions. This means there is quite a bit of variability and uncertainty inherent in the models and hence the model loss estimates. Catastrophe models do not produce deterministic answers, but rather ranges of possible outcomes along with estimated probabilities of those outcomes. The simulated outcomes along with their estimated probabilities will naturally differ between the different models.
The catastrophe models have another important requirement, which is detailed data about the exposure at each location that could be affected by a catastrophe. For residential and commercial properties, this includes the replacement value, construction type, occupancy and other building specific details. Without this information or if this information is inaccurate, the models will not provide reliable results.
Finally, the currently available catastrophe models do not incorporate all sources of loss caused by catastrophes. This was demonstrated clearly by the significant events of the past few years.
All of this doesn't mean the models should not be used. The models are still the best and most sophisticated tools for catastrophe risk assessment. It just means the models need to be used with the limitations clearly in mind. The models are just one part of the risk assessment and management process, and they need to be supplemented and validated using independent information and benchmarks.
Using Model Output to Manage Catastrophe Risk
The catastrophe model output is the estimated loss distribution frequently called the exceedence probability (EP) curve. The exceedence probability curve provides estimates of the probabilities of various levels of loss being exceeded. These curves can be generated by peril, by geographical region and all the way down to the individual location level. Because the models produce so much detailed output, these numbers can easily be misinterpreted and misused.
Many model users focus on point estimates such as PMLs and long return period losses. These numbers are subject to a high degree of uncertainty and each point on the curve has uncertainty around the loss and the probability.
The uncertainty around the EP curve increases with the size of the loss and as the probability of loss decreases.
Additionally, the higher the resolution of the model output, the more the variability and uncertainty around the output. This means extra caution should be applied before using the high resolution numbers taken directly from the models. If the uncertainty in model output is not taken into account, your decisions and business strategies will be subject to instability as the models are updated and improved.