ROBUST ESTIMATION OF LOSS MODELS FOR LOGNORMAL INSURANCE PAYMENT SEVERITY DATA

Abstract The primary objective of this scholarly work is to develop two estimation procedures – maximum likelihood estimator (MLE) and method trimmed moments (MTM) for the mean variance lognormal insurance payment severity data sets affected by different loss control mechanism, example, truncation (due deductibles), censoring policy limits), scaling coinsurance proportions), in financial industries. Maximum estimating equations both payment-per-payment payment-per-loss are derived which can be solved readily any existing iterative numerical methods. asymptotic distributions those estimators established via Fisher information matrices. Further, with a goal balancing efficiency robustness remove point masses at certain points, we dynamic MTM claim models above-mentioned transformed scenarios. distributional properties comparison corresponding MLEs along extensive simulation studies. Purely illustrative purpose, examples 1500 US indemnity losses provided illustrate practical performance results paper.