Balance Analysis of Operational Risk Through the Aggregate Method in the Loss Distribution Approach

Abstract

Operational risk is defined as the risk of loss resulting from negligence or failure in an entity's internal processes or due to external problems. Companies (especially financial institutions) also face these risks. Recording operational losses in insurance companies is often not done correctly, resulting in limited data regarding operational losses. In this research, the focus is given to operational loss data recorded from claim payments. In general, the number of insurance claims can be resolved using a Poisson distribution, where the expected value of a claim is proportional to its variance. On the other hand, the negative binomial distribution has an expected value that is definitely smaller than its variance. The analytical method used to measure potential losses is through a loss distribution approach using the aggregate method. In this method, loss data is categorized into frequency distribution and severity distribution. By performing 10,000 simulations, a total claim loss value is generated, which is the accumulation of individual claims in each simulation. Then from the simulation results, the potential loss value (OpVaR) at a certain level of confidence is determined.