Cecilia Torrejan, AVP and Chief Underwriter, Life Reinsurance
Underwriting, simply put, is the method by which insurance companies assess individual life risks for the likelihood of a claim. Over time, the practice of underwriting has evolved due to market influence, improvements in data gathering, and technology. In modern times, certain subjective aspects of underwriting are being replaced by computer algorithms and predictive analytics. Since individual underwriting has been implemented as a means of risk assessment, has the approach to underwriting changed? To determine this, let’s look back at the infancy of the practice to where it is today.
In the early 20th century there was very limited medical underwriting, there being one single risk class, with premiums varying by age only. In the 1940’s, gender distinct rates were introduced to the industry, and were considered to be innovative at the time. Then, in the 1970’s, Smoker/Non-Smoker risk class rates were introduced as the notion that lifestyle had an impact on longevity started to take hold.
The 1980’s brought several changes. One of the most important was the introduction of blood testing at low face amounts in response to the HIV/AIDS surge; in those early days, there was the perception that it had the potential to cripple the life insurance industry. The learning curve got steeper for underwriters as they had to adjust quickly for a virus with no known treatment at the time and which the medical community was only starting to understand. In addition, underwriters and medical directors alike had to deal with tests included in the blood work that traditionally were not part of the routine testing. The dilemma was how to sort and look at this new information and differentiate between the meaningful and trivial deviations to arrive at a fair underwriting decision.
Looking back, the insurance industry and the underwriting/medical community rose to the challenge and managed to adapt to the risk of a new disease. The industry did not crumble as many had predicted because it reacted swiftly and decisively.
Another innovation in terms of risk class categorization occurred through the introduction of preferred products which offered new groupings of price differentiation to the consumer in acknowledgement of those who embraced healthier lifestyles. In the 1990’s, the number of preferred classes increased and Term products became very inexpensive as companies competed fiercely to gain market share.
Tools of the Trade and Processes
With all these changes, underwriting became more interesting but also more challenging. Technology at the time was more limited; the internet did not yet exist. The majority of companies worked in a paper environment, performing most processes manually. Additional support staff were hired to handle the tasks required to complete the underwriting/issue process.
Face amounts were much lower than what we see today, and an application for $1,000,000 was considered a huge amount. The underwriting process was expensive and time consuming, and relied on paper applications, insurance medical exams done by doctors, EKGs, doctor reports, chest X-rays, etc. An applicant with a slightly more complex medical history might end up waiting months for a policy to be issued.
Trends in Distribution and Underwriting
With the change in the financial environment in the 1990’s, the increase in mergers and acquisitions and the sustained low interest environment, reducing costs and increasing productivity became a must in order to survive. During this period, the first expert underwriting systems started to appear and on-line underwriting manuals were introduced. Technology also had a significant impact on the public, who demanded faster service and accessible products as it became more knowledgeable.
In the 2000’s, there has been a liberalization of the preferred criteria and routine age and amount requirements, and an increase in the use of information obtained electronically from vendors databases.
The trend in underwriting in the last few years has been to move to less invasive testing (more Attending Physicians Reports, i.e. APS’s, fewer MD exams, EKGs and chest X-Rays, and Simplified Issue (rapid or instant issue of policies based on information obtained electronically from various sources).
Simplified Issue
There is great interest in these type of products as an effective way to capture the under-insured middle market. An underwriting approach may involve a completely electronic process aimed at the instant issue of a policy. Technological advances have made possible the use of electronic applications and underwriting systems that facilitate the utilization of database driven risk assessment. The typical databases and tools used to process simplified issue cases may be some, or all, of the following: MIB, MVR, Pharmaceutical databases, Identity Checks, Tele-Underwriting with
extensive drill-down questions, and the use of scores for financial and non-financial data. Once the required information is obtained, underwriting rules engines kick in and a decision is made. The decision could be just a simple Approve/Reject or, in other cases, a third option “Refer to Underwriting” is added.
To counteract increased anti-selection, margins have been built into the rates. It is also important to monitor experience and distribution channels on a regular basis to adjust questions, e-requirements and rules engines as necessary. On the other hand, with all the technology advancements and additional tools that may be available in the future, risk assessment for simplified issue products may eventually evolve to be similar to fully underwritten products, at least for a segment of the population.
Predictive Modeling and Life Underwriting
The volume, complexity and variety of data available today is challenging organizations to find new ways to grow and innovate. Life insurers are starting to use predictive analytics to transform their data into useful insights for better decision making and to gain a competitive advantage. A developing trend is to apply predictive models to fully underwritten business using innovation in technology, scoring models, new media and big data.
Today’s customers are well-informed and expect fast and reliable services and products at the right price. Real-time data analysis used in simplified underwriting, combined with predictive models that incorporate traditional underwriting data, have stronger protective value and the potential for significant cost savings as well as reduction of cycle time that will benefit the client. Models can be used for selecting a risk class, requesting/triggering additional information and determining ratings.
This doesn’t mean that underwriters will no longer be needed. Increased underwriting efficiency may help generate more business and reduce wastage. Thus, underwriters can concentrate on the complex cases that have been identified by the model at an early stage, something a machine can’t do.
Developments that will facilitate and speed up the implementation of underwriting models are electronic health records and the much-talked-about wearable health devices and mobile-health apps. As electronic health records are universally adopted by the medical establishment and data is structured, life companies will have an increased ability to track and determine impairments, lab results, prescription histories and diagnostic testing over the years. Development of more predictive scores based on all this data will also increase.
Many of the mobile-health apps in the market, such as Fitbit and Nike FuelBand, are devices that use body sensors to track the wearer’s physical activity, monitor exercise, calorie expenditure and so on. However, the information provided by these devices has not yet made its way to the risk selection process.
There are many wearable medical devices that are available today, such as glucose sensors, blood pressure monitors, as well as devices in development that can predict a heart attack/stroke minutes before happening that may in the future become available to the public. The challenge for the industry will be how and if this information will be accessible, as this is new territory and regulations will have to be formulated as the technology develops.
Conclusion
Change is here to stay and it will continue to evolve due to data availability, and advances in technology and medical industry space. Underwriting practices as we know them today may well disappear in 5 to 10 years from now. Underwriters will be involved in interpreting large amounts of data, helping in the creation of predictive models and rules engines and evaluating
new technologies. The traditional skills required to be an underwriter – analytical skills and technical medical expertise – will become of paramount importance. A background in statistics may also be useful. The future will present enormous challenges as well as great opportunities and the question is if we are ready to face them. As history has shown, the insurance industry will continue to adapt and rise to the occasion to find suitable solutions that will mutually benefit the customer and the risk carriers alike.