A Method of Testing Loss Reserves [Discussion]

After reading Mr. Comstock's paper on testing loss reserves, it strikes me that possibly we are worrying ourselves unnecessarily over the problems of the actuary. Current loss reserves are the product of a continuous whirlpool of transactions involving : First, setting up the original estimates; second, changes in estimates as facts develop; then reduction of the amounts by subsequent payments ; and lastly, the elimination of settled items. Each step in the evolution of the total is a distinct operation, and the companies' reserves will be accurate only to the extent that the individual system provides for a prompt and efficient completion of these operations. Notwithstanding this continuous movement throughout the entire set-up, the trend of the totals is generally consistent with the surface indications; but the figures always carry with them the general feeling that, as it is easier to be liberal than cautious, and as putting up reserves is a more direct operation than taking them down, the final figure contains more than ample margin of safety.