Insurance as a product thrives best in environments where risk can be understood, predicted, and priced across populations with widely varied individual risk exposures as best determined by risk experience within the population or application of risk indicators. Blah, blah, blah. Insurance is a long-standing principle of sharing the ultimate cost of risk where no one participant is unduly at a disadvantage, and no one party is at a financial advantage- it is a balance of cost and probability.
There is not much new in that introduction, but what if the dynamic is changed, where the understanding of risk factors is not retrospective, but prospective? Maybe that approach is nigh, where the availability of data is broad enough, and the ability to understand and determine causative actions or combinations of actions can be analyzed and applied to predict outcomes.
Seems it was simpler then – risk predictions over a few pints
The group sitting in the Lime St pubs in the seventeenth century would not have aggregated risk predictions for ships’ voyages where the ships’ captains were inexperienced cabin boys, the risk of failure would be obvious. Flash forward to Philadelphia in the eighteenth century- in similar fashion Ben Franklin would not have recommended fire insurance as a concept unless there were considerations such as no forges allowed in an insured dwelling, or no open campfires in the parlors of houses under an insurance contract. The risks would be obvious.
As time passed and underwriting sophistication improved carriers began to rely on experts (actuaries) who could through mathematic and statistical methods determine risk exposures through analysis of past experiences and through that develop predictions of future risk exposures. Applying those methods across large populations ‘smoothed’ the probability of individual risks and by extension provided macro outcomes that could build pricing models. Experience, understanding, prediction, pricing. Law of large numbers applied to a pricing model.
No longer just looking backwards- are other behaviors involved in predicting risk?
Sophistication and the advent of other sources of data, e.g., credit performance and scoring, took the predictive model further, applying seemingly unrelated data to risk predictions for insurance. A correlation of credit performance was proven in terms of driving performance and carriers began uniform inclusion of those data in determining insurability and/or pricing. Forward-looking risk assessment based on data analysis/application of seemingly unrelated variables.
Now with the exponential growth of artificial intelligence tools, machine learning, data cleansing techniques, and access to huge data lakes, how will the insurance industry proceed? At this time there are many initiatives involving AI and many applications of machine-learned behavior. These include improved virtual claim tools, dynamic underwriting that supports insurance on-demand, direct application of AI in assessing symptoms and diagnosing ailments, smart devices that in concert with AI work to prevent or mitigate risk, and feedback to underwriting algorithms that step past traditional actuarial science and begin to break down the requirement of large populations of risk probabilities in order to understand and price risk.
AI is a cool approach, but others can play, too
What prevents the capital markets from applying AI methods (through design or purchase) in predicting or betting on risk outcomes? The more comprehensive and accurate risk prediction methods become the more direct the path between customer and risk financing partner also becomes. If the folks sitting with a pint on Lime St knew that seas would be calm, and voyages would be smooth even a cabin boy could have been captain of an insured ship, and insurance would be an easy bet.
The risk-sharing/risk financing industry has evolved through the application of available technology and tools, what’s to say AI does not become a double-edged sword for the insurance industry- a clever tool in the hands of insurers, or a clever tool in the hands of alternative financers that serves to cut away some of the insurers’ business?
