# A fair pricing model via adversarial learning

**Authors:** Grari Vincent, Charpentier Arthur, Lamprier Sylvain, Detyniecki Marcin  
**arXiv:** [2202.12008](https://arxiv.org/abs/2202.12008)  
**Date:** 2022-02-24  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2202.12008v1/

## Key points

- Insurance business relies on classification of insured individuals into risky and non-risky categories
- Actuarial fairness dictates that riskier individuals should pay higher premiums
- Actuaries use econometric or machine learning techniques to classify individuals
- Fine line between fair actuarial classification and discrimination
- Growing interest in fairness and discrimination within the actuarial community
- Non-sensitive characteristics can serve as substitutes for protected attributes, but may not be enough for accuracy
- Traditional pricing models have limitations in achieving fairness due to potentially biased components such as car or geographic risks
- Authors developed a novel pricing model approach utilizing autoencoders to generate geographic and car pricing components while mitigating unwanted bias according to desired metrics
- Approach can be generalized to multiple pricing factors and is adapted for a fairness context allowing for debiasing of the set of pricing components
- Debiasing predictor alone may not be sufficient for adequate accuracy in predicting insurance prices
- Importance of developing fairer pricing models in the insurance industry

## Summary

The insurance business relies on the classification of insured individuals into risky and non-risky categories, with actuarial fairness dictating that those who are deemed riskier should pay higher premiums than those who are less risky. Actuaries use econometric or machine learning techniques to classify individuals, but there is a fine line between fair actuarial classification and discrimination. As such, there is growing interest in fairness and discrimination within the actuarial community. Non-sensitive characteristics can serve as substitutes or proxies for protected attributes, but using them alone may not be enough to maintain adequate accuracy. Traditional pricing models are built in a two-stage structure that considers many potentially biased components such as car or geographic risks, which has significant limitations in achieving fairness. To address this issue, the authors have developed a novel pricing model approach that utilizes autoencoders to generate geographic and car pricing components needed to predict pure premiums while mitigating unwanted bias according to desired metrics. This approach can be generalized to multiple pricing factors and is perfectly adapted for a fairness context since it allows for debiasing of the set of pricing components. The authors' findings suggest that debiasing the predictor alone may not be sufficient to achieve adequate accuracy in predicting insurance prices, highlighting the importance of developing fairer pricing models in the insurance industry.