Insurance companies need to understand the concept of risk, and data is the most important factor when it comes to assessing risk potential. In order to derive profits and provide better service, insurers need accurate assessments. In the past, the statistical models that may have taken actuaries weeks to complete can now be performed in minutes and with the benefit of real-time data analysis.
With Big Data tools, insurance companies create scenarios that illustrate what can happen in the future according to multiple variables. The formulation of these scenarios is based upon data collected in the past, which is in turn compared to certain measurements and variables. For example, data related to wildfire season in parts of California can help insurance underwriters determine the minimum that should be charged in terms of annual premiums; this determination is based on the likelihood of a neighborhood being affected by a forest fire during dry season and how much damage a home could sustain.
Predictive modeling is another process used by insurance companies to determine the likelihood of fraudulent claims, which represent a major headache for the industry. Insurance fraud is an issue that is silently reflected in the elevated premiums that affect all individuals who need to purchase policies. Although insurance company executives cannot actually prevent the incidence of fraudulent claims, they can study variables and patterns to detect certain factors that tend to go hand in hand with situations prompted by fraud. These days, behavioral patterns leading up to false claims can be observed in social media, and dynamic modeling of business relationships can also help to flag claims made against commercial liability policies.
The prospect of self-driving technology is very auspicious for insurance companies for various reasons, and telematics happens to be one of the most exciting elements. In countries such as South Africa, auto insurance providers are offering drivers the option to install sensors that report driving conditions to underwriters; the actions being monitored include acceleration, turning, changing lanes, and habitual driving patterns. Armed with this information, the insurance company can use an algorithm to adjust policy premiums. In the near future, telematics could help drivers find cheap auto insurance if they are willing to provide information collected by their smartphones from sensors installed in their cars.
Big Data applications in the insurance industry are certainly promising, but they also worry privacy advocates who are concerned about long-term implications in terms of data collection. One example of such concerns can be related to the ongoing debate about health insurance reform in the United States; the determination of coverage and premiums based on preexisting conditions could be augmented by means of genetic profiling or even behavioral analysis. For some health insurance companies, the prospect of ordering genetic testing or collecting data from wearable devices may seem very attractive, but privacy advocates believe that such practices would be too intrusive. Government regulators may also be opposed to genetic testing unless they can reassured that the data collected will not be used for purposes other than premium and coverage determination.
In the end, Big Data and the insurance industry seem to be made for each other. Aside from privacy concerns, both insurers and their clients stand to benefit from more accurate risk assessments and predictive modeling as well as from a lower incidence of fraudulent claims.
Kevin Faber is the CEO of Silver Summit Capital. He graduated from UC Davis with a B.A. in Business/Managerial Economics. In his free time, Kevin is usually watching basketball or kicking back and reading a good book.
Follow him on Twitter: @faber28kevin