Isotonic Recalibration under a Low Signal-to-Noise Ratio
This addresses the need for fair and systematic pricing in insurance, though it appears incremental as it builds on existing recalibration techniques.
The paper tackles the problem of ensuring auto-calibration in insurance pricing systems by proposing isotonic recalibration for regression models, proving that under low signal-to-noise ratios, this method results in explainable pricing systems with low complexity.
Insurance pricing systems should fulfill the auto-calibration property to ensure that there is no systematic cross-financing between different price cohorts. Often, regression models are not auto-calibrated. We propose to apply isotonic recalibration to a given regression model to ensure auto-calibration. Our main result proves that under a low signal-to-noise ratio, this isotonic recalibration step leads to explainable pricing systems because the resulting isotonically recalibrated regression functions have a low complexity.