Regulators eye big data impact on ratemaking

Regulators eye big data impact on ratemaking

Risk Disclaimer >>
Ad disclosure Fintech-Insight stands firm in its mission to facilitate sound financial decisions for you. We forge alliances with specialists to provide the latest in news and facts. Engagement with designated links, sponsored entries, products and/or services, leading transfers to brokers, or promotional content might entail financial recompense for us. We pledge to protect our users from any negative repercussions arising from utilizing our site. Be informed that no content hosted here should be interpreted as authoritative in legal, tax, investment, financial matters or any expert counsel; it is meant for informational purposes exclusively. Should there be any concerns, securing the guidance of an independent financial consultant is recommended.

In insurance, big data connotes unstructured and structured data that influences underwriting, rating, and pricing. Structured data gets stored as pre-defined fields and tables, while unstructured could be social media, recorded videos or documents. Using predictive technologies, insurers use big data to discern trends and forecast future events. According to SNS Telecom, carriers’ spend on big data technologies is expected to reach $3.6 billion this year.

The advent of big data in insurance has spawned growth of predictive modelling in ratemaking. Despite significant benefits to insurers and customers, the relative complexity and rapid growth in new models pose a challenge to regulators, tasked with approving such models. They now demand more insight into what data is available, how it’s used and whether it is beneficial or harmful to consumers.

In the US, the P&C Insurance Committee recently adopted the Regulatory Review of Predictive Models White paper at the National Association of Insurance Commissioners (NAIC), which seeks to identify best practices for review of predictive models filed by insurers with regulators to justify rates.

Though predictive models are being increasingly embraced by traditional carriers, insurtechs are often at the cutting edge of how to leverage big data and predictive models. Even before the big data phenomenon, insurers used data mining techniques, within relevant regulatory frameworks. Pricing or premium would be reviewed at the time of contract renewal. But now, insurers tweak premiums in real time resulting in savings for customers (e.g. pay-as-you-drive motor insurance). This increased use of models helps avoid adverse selection, better serve customer needs and rationalize cost structures.

Europe has no specific regulation on big data. However rules exist that are relevant to its use, such as General Data Protection Regulation (GDPR) and Insurance Distribution Directive (IDD). The GDPR has a legal framework for processing data and provides insurers with guidance to mitigate risks from big data use. Consumers benefit from strengthened rights to protect personal data. Additionally, consumers have the right not to receive a decision solely based on automated processing. IDD regulates the distribution of insurance products, preventing poor selling practices enabled by use of big data analytics. Its provisions on product oversight regulate new insurance product design.

The NAIC’s Executive Committee will consider adoption of the white paper during their Spring Meeting. Earlier, during draft reviews, parties had deliberated about new rating standards that extend the statutory scope of rate reviews. Some voiced that these requirements might stifle innovation. Many had concerns about providing the volume of data requested and how it would affect their proprietary algorithms.

The white paper has identified such best practices as would ensure rating factors produce rates that are not excessive, inadequate or discriminatory. It seeks a clearer understanding of underlying data with assumptions and adjustments.  These are apparently not binding on state insurance departments but intended as guidance. The guidance is focused on personal automobile or home insurance ratings, but is transferrable to other LOBs, as outlined.

The availability of more data isn’t automatically a safe passage to pricing improvements. Inappropriate use of data can cause serious reputational issues, including ethical issues around using customer data. One insurer, following backlash from customers, was forced to stop using data from Facebook to price business. While these technologies can enhance data enrichment, left unchecked, unintended biases creep in, distorting pricing models or excluding customers. The oversight goes beyond setting up the algorithm as these are right at a set point and degrade over time or with a different population.

More than ever, it is becoming critical for insurers to take sufficient care of not only how a model will be used, but how it will be received by stakeholders (including regulators).

Cover Image Credit

You get 3 free articles on Daily Fintech. After that you will need to become a member for just US$143 a year (= $0.39 per day) and get all our fresh content and our archives and participate in our forum.

Risk Disclaimer

Fintech-Insight is dedicated to delivering unbiased and dependable insights into cryptocurrency, finance, trading, and stocks. However, we must clarify that we don't offer financial advice, and we strongly recommend users to perform their own research and due diligence.

Leave a Reply