Analysis of models for claim reserves in insurance industry

Abstract

Insurance contracts are formed such that the insurer (the insurance company), agrees to pay the insured party a fixed sum of money when a certain event happens, in return for a fixed amount or stream of amounts called premiums. Premiums form the revenue stream for the insurance company; where as the cost side consists of sums which are contingent on certain events occurring. Hence, a part of the cost side is essentially random in nature. Insurance claims are raised when the insured exercises his or her right to get paid when the uncertain event, to protect against which he or she had bought the insurance product, has occurred. The insurance company will check the veracity of the claim to decide how much payment the insured is entitled to. The event of payment of the insured is called settlement of the claim. To correctly ascertain the magnitude of expenses incurred, the company needs to estimate how much would be the amount of claims that are raised on the company. They need to determine how much would be the amount of future payments that they will have to make, that are associated with claims that are incurred so far, but are not yet settled. Further, regulations also require the insurance company to set apart sums of money that are reasonable estimates of the future claims obligations that might arise. Claims are the main operating expenses for insurers. Claims reserves is the money that is set aside for eventual payment of incurred claims that have not yet been settled. They are recorded in the liability side of the balance sheet. The reserve amounts can be arrived at either subjectively using actuarial judgment or by using statistical methods. This project looks at various statistical methods used in the insurance industry for the prediction of claims reserve.