Large numbers of insurers have realized that use of an automated underwriting system has become a must for jet issuance of policies. The industry is struggling to grow trying multiple channels of distribution but the growth is hindered by complex underwriting processes which are either too complex or time consuming thus making the new business acquisition an expensive and protracted affair.
Life Insurers have not really changed the way the policies are underwritten from last few decades. Processes are still people dependent, their experiences, referring to rule documents and other raw pieces of information from different sources. Manual processes may result in inconsistencies in terms of high error rates and cost of labor.
Even though there are technology advancements to create instant issue environment such as paperless offices i.e. no need to float the physical papers across, Web based underwriting systems integrated with advanced analytics engines to support underwriting decisions but most of the companies are still on the old legacy systems.
An automated underwriting System evaluates all criteria relating to an application and the applicant and then return one of the three recommendations decline, approve or review (gray area) or “caution”. These decisions are arrived by the rules configured in the System. A typical Life Underwriting System can underwrite an application or a proposal on following criterion:
1. Insurable interest: The first step in underwriting a life insurance policy is to verify the policy owner's insurable interest in the insured individual, defined as "having an interest in that person remaining alive, or expect emotional or financial loss from that person's death."
2. Medical and Non-medical: Companies usually investigate the health of the proposed insured. Height & weight (Body Mass Index), blood pressure and other health factors are taken into account.
Based on the Sum Assured, Entry Age of the Life Assured (LA), Sum under Consideration (Sum Assured of all policies and proposal including specific riders of the LA with the insurer), and Plan opted by the applicant etc medical tests to be undertaken by the LA can be derived.
3. Female Lives: based on the occupation, age, insurance on husband (if married) etc.
4. Juvenile Lives: health condition of the child/ minor.
5. Occupation & Avocation: A proposal can also auto loaded / rated up the system for Occupation of the Life Assured and the policy holder. Different occupations can be associated with classes.
6. Medical questions filled by the Life Assured and the results of medical tests: the system can be made interactive with reflexive questionnaire based on user inputs. E.g. If Life Assured is a smoker than next question would be “how many cigarettes in a day?” etc.
7. Family Health history: Any medical history in the family and any post critical illness deaths before the age of 60 etc.
8. Alcohol & smoking history.
9. Financial Underwriting: financial stability of the life assured or the policyholder can be judged by annual income, source of income etc.
Cased Base Reasoning
The underwriting process and the rules can be continuously improved based on the experience collected in these systems. The process of underwriting a new proposal based on the underwriting decisions made of similar past cases is called Cased Based Reasoning (CBR). Case-based reasoning is not only a powerful method for computer reasoning, but also a pervasive behavior in everyday human problem solving; or, more radically, that all reasoning is based on past cases personally experienced.
CBR traces its roots to the work of Roger Schank and his students at Yale University in the early 1980s where the effort was made to introduce another method of artificial intelligence. These principles can be extended to Automated Underwriting (author comments).
CBR for the purpose of automated underwriting can be of a four step process.
1. Retrieve: Given the set of data parameters in the application form, relevant set of cases underwritten earlier can be retrieved and the decisions made. E.g. an underwriter who wishes to underwrite a term policy for a pilot with age 55 years may retrieve all term plan cases underwritten earlier with occupation as pilot.
In the process of retrieval also the claim information of similar cases can be retrieved.
2. Reuse: Map the retrieved data to the target proposal which is in the process of underwriting. E.g. in this case the underwriter must adapt his retrieved solution to include the addition of higher age at entry.
3. Revise: Having mapped the previous solution to the target situation, test the new solution in the real world e.g. the decision can be given on vetting by an experienced underwriter or also looking at the previous claims experience of similar type of cases.
4. Retain: After the solution has been successfully adapted to the target problem, store the resulting experience as a new case in memory. E.g The newly underwritten case becomes an addition for the system thus enriching set of stored experiences.
Conclusion:
Moving towards an automated underwriting system adapting to latest technology insurers can create a real time and intelligent environment which can support an instant and jet issuance environment. Policies can be issued in minutes and companies can open more distribution channels such as online, retail, banks, individual agents etc.
