Insurance companies are well-known for investing their resources in GLM loss modelling. They have become so mainstream you’d be hard-pressed to find an actuary without any experience building them. In fact, nowadays, bigger companies even use their interns to keep these models updated.
Although not all companies are developing or leveraging these models the same way; the focus on GLM modeling is natural because it helps to accurately determine the true expected loss cost of a risk. Yet, some insurance professionals may be missing opportunities to improve their business results by not building and leveraging demand models. These demand models aim to predict the probability that a prospect or policyholder will either buy or renew the policy at a given price. While both models have their core strengths, I would like to highlight through an example how demand models may benefit the modern insurer.
Loss cost model vs Demand models:
Loss cost modeling is a core business process that brings tremendous value to an insurer. Loss cost models are more complex than demand models. Whereas loss cost models strive to create the best prediction for each and every risk variable, demand models do not require such granularity.
The main reason for this difference is that loss cost models are trying to explain an unconsciousness behavior; nobody (hopefully!) is purposefully crashing their cars. Therefore, when attempting to calculate the probability of a car accident happening, these models must consider many variables that could have even the smallest indirect impact on such an event taking place.
On the other hand, demand models are trying to predict a very controllable decision; whether someone will purchase or not purchase a policy. Thus, a lot less explanatory variables (with each variable having a bigger influence on the result) are needed to calculate the probability of a policy being bought or not. A simple example of such a variable is price. No other variable has as much importance in risk modeling as price has in demand modeling. It has tremendous impact because the price level will directly affect the willingness of a prospect/policyholder to renew or buy a new policy.
Demand models are still not yet widely used so they can be a differentiator for your company to improve business results. By using demand models, you have the chance not only to imitate the competition but leap frog over them. So, make sure you join the demand modeling game!
Download the Loss-Cost vs. Demand Modeling Whitepaper to learn more about the differences, similarities and opportunities for insurers in Loss-Cost vs. Demand Modeling as well as practical steps for successfully implementing demand models into your organization.