each discriminate individually, but how much they would add to the discriminating effectiveness of the system as a whole. In Exhibit VI, the nine recommended tests are compared with a system using the adjusted tests. In the latter, adjusted earnings has been substituted for net gain; change in adjusted surplus has been substituted for change in stated surplus; and the ratio of liquid assets to adjusted liabilities has been added (making a total of 10 tests). The system with the adjusted tests would be marginally more effective than the recommended system, particularly in the last year before insolvency. Thus, use of the adjusted tests could improve the effectiveness of the Early Warning System slightly. More importantly, we feel the adjusted tests are logically more adequate than the similar statutory tests. On the other hand, including the adjusted tests has two disadvantages. Most importantly, the adjusted tests would significantly complicate the users' problem of understanding and using the Early Warning System. All of the recommended tests are relatively simple in concept and can be readily calculated by hand. The adjusted tests, however, would introduce an entirely new level of complexity into the system. We feel this added complexity might prevent some potential users from gaining the maximum benefit from the Early Warning System. Second, a data gathering problem would result from the increased period of time for which annual statement data would be required (five years as opposed to two) and the fact that some of the data required would be "new" information. Obtaining this information might depend upon the cooperation of the insurance departments in all the states and would present numerous opportunities for errors and omissions. Thus, primarily because we believe the added complexity of the tests outweighs the very modest improvement in effectiveness, we recommend that the three adjusted tests not be employed in the Early Warning System. USING THE LIFE AND HEALTH EARLY WARNING SYSTEM The Early Warning System for Life and Health Insurance Companies has been developed by an NAIC committee of state Insurance Department personnel. The purpose of the system is to help the various state Insurance Departments identify the life and health companies most likely to experience financial difficulties, so that the departments' resources for in-depth analysis and on-site examination of companies can be focused on these companies. The system consists of nine ratios or tests that measure various aspects of the company's financial condition and stability. These tests are quite simple, yet historical analysis shows that they are effective in distinguishing between troubled and sound companies. For each test, bench marks for determining "exceptional values" - those most likely to indicate difficulty have been established by comparing past test results for insolvent and sound companies. Depending on the number of test results that are in the exceptional value range, each company is placed in either a "priority" or a "non-priority" category. Although the system is effective in distinguishing between troubled and sound companies, it is by no means foolproof. Therefore, important decisions such as licensing decisions should not be made on the basis of test results without further analysis and examination of the company concerned, and no state should rely completely on the Early Warning System as its only basis for identifying companies for priority attention. The purpose of this manual is to assist state Insurance Department personnel in understanding how the Early Warning System operates and in gaining the maximum benefit from it. The manual consists of three chapters, covering: 1. The early warning tests 2. The priority company system 3. Suggestions for further analysis. 1 THE EARLY WARNING TESTS The Life and Health Early Warning System is based on nine tests that have proven effective in distinguishing troubled from sound companies. The effectiveness of these tests was determined by compa ring the test results for a group of insolvent companies with the results for a matched sample of solvent companies. The insolvent company sample included 50 companies that were liquidated or came under state control during the past decade. The sample includes half of the total number of insolvent companies during that time - all of those for which annual statements were available. A sample of 85 solvent companies was matched with the insolvent sample in state of domicile (Exhibit I) and premium volume (Exhibit II), and was screened by the chief examiners in the states of domicile to eliminate clearly troubled companies. The solvents were also similar to the insolvents in product mix (Exhibit III). Of the insolvent companies, 8 percent were mutuals; six percent of the solvent companies were mutuals. Based on the test results for these two samples of companies, bench marks for determining exceptional values were established for each test at the point that identified the highest percentage of insolvent companies without identifying an unmanageable percentage of solvents. Exceptional values thus represent the test results that are most likely to indicate possible financial difficulty. Work sheets for calculating test results will be found in the Appendix. These work sheets may also be helpful in clarifying the details of how the test results are calculated. The nine early warning tests fall into two groups: Financial tests Stability tests. |