| Abstrakt: |
In the last decade, actuaries' activities have expanded beyond insurance risk assessment to include enterprise risks in a broad sense: credit risks, market risks, liquidity risks, operational risks, project risks, etc. Increased demand has led to the introduction of a new international credential - Chartered Enterprise Risk Actuary (CERA). Today, CERA's professional education and practical activities are based on methods and models for assessing insurance risks, which assume the availability of large amounts of statistical information. However, enterprise risks are not always so stable and universal that extensive statistical series of past observations can be used to predict future losses. There is a need to develop a more flexible mathematical apparatus that would allow a variety of information for constructing the distribution function of a random value of future loss. The article presents the concept of generalized actuarial valuation, which allows solving the problems of assessing business risks. This concept involves combining all available information into an actuarial risk basis – the totality of all acceptable risk assumptions. The actuarial risk basis may include different scenarios, loss models, and different estimates of parameter values. The article describes an algorithm for statistical modeling based on the Bayesian approach, which allows converting such an actuarial risk basis into a discrete approximation of a hypothetical cross-section of a random loss process in the form of central estimates of quantiles of all necessary levels. The article contains a practical example of assessment of the enterprise risk in the exploration and operation of gas and oil fields in a given area of the subsurface. An actuarial model of the random value of the future loss based on the cash flow from the exploitation of the subsurface area has been developed. Based on this model, a loss distribution function is constructed, which is used to assess the level of risk. [ABSTRACT FROM AUTHOR] |
