How Robust Is Your Pricing Process?

In Pricing, you input data and output prices. That these prices rely on the data inputted is apparent. But you also make use of quite a few assumptions and rules during that pricing process. But how strongly do your prices depend on these assumptions and rules? And by how much are your prices off, if your assumptions are slightly wrong?
We will see that typically a more complex pricing process is less robust with respect to slight calibration errors. For example, if you decide to just issue a flat street price, then your profitability is rather stable with respect to mild drift in the technical price. But if you use individual optimisation and underestimated the price elasticity of your clients, your portfolio likely will operate on a loss.
With ever improving machine learning models and ever more data being available, it is ever more relevant to be aware of the impact of small deviations to the insurer and the insured.
So, let’s discover how to explain deviations between expected and observed margins, understand its robustness w.r.t the used models and discuss mitigation strategies to combine state of the art pricing approaches with sustainable profitability.