Introduction
The following is a summary of the testimony given by Bryan Rehor, Director of Regulatory Affairs, ZestyAI, at the second CDI wildfire workshop, held on September 28, 2023. Bryan’s testimony from the first Wildfire Workshop can be found here.
ZestyAI's Vision and Role in Rate Assessment
ZestyAI’s vision is to predict and mitigate the impact of natural disasters on society by harnessing cutting-edge technology, such as Artificial Intelligence. In July, we shared the importance of transparency in the rate review process as well as the benefits that new-age risk models, such as our flagship product, Z-FIRE, can offer the market. Today, we will focus on the implications of modeled rates on insurance availability and affordability and will offer our suggestions on how this change can be implemented most effectively.
ZestyAI’s vision is to predict and mitigate the impact of natural disasters on society by harnessing cutting-edge technology, such as Artificial Intelligence.
The Current State of Rate Indication and Review Process
ZestyAI acknowledges that using a model to develop a catastrophe load has the potential to generate more accurate rate indications. However, a higher degree of accuracy will have zero impact on the market if carriers cannot adjust rates in a timely manner. Proposition 103 already lays the foundation for a predictable review process by deeming rates approved after 180 days, while allowing certain provisions for intervenors. Unfortunately, carriers typically face a 12, 18, or 24-month waiting period before they can implement a filed rate revision.
Introducing a new layer of complexity into the rate application without making transformative changes to the review process will only lengthen approval times. The inclusion of models in rate setting will have a significant impact on the process used to review rate applications, and it may risk causing further delays in the review timeline. However, this also presents an opportunity to update and improve the review process with the goal of providing feedback to the market more rapidly. If the CDI wishes to make an impactful change, they must also provide a path to approval in 180 days for rate applications that are justified by the CDI’s own rate application template.
Inclusion of models in rate setting... presents an opportunity to update and improve the review process with the goal of providing feedback to the market more rapidly.
Reviewing Loss-Producing Models
ZestyAI acknowledges and appreciates the CDI’s plan to increase staffing by the end of 2024 as a means of expediting rate filing reviews. This must be combined with a well-thought-out model review framework in order to be successful.
In July, ZestyAI shared the importance of filing a model’s assumptions, inputs, and validation statistics in order to build public trust. ZestyAI recommends that the CDI begin reviewing loss-producing models independent of carrier filings. If the CDI lacks the current resources or expertise to review models, we recommend they review models in conjunction with a third-party actuarial firm, which was the approach taken with ZestyAI’s own Z-FIRE model. Reviewing models prior to, rather than in conjunction with, carrier rate filings has multiple benefits. It can reduce the time to market, increase availability sooner, and provide reassurance to the public that the models have been fully vetted by experts.
Increasing Options for Homeowners
Assuming the CDI addresses the root cause of the insurance availability crisis, ZestyAI believes that the incorporation of models into the rate application can increase insurance options for homeowners while providing additional rate stability in the future. Requiring higher acceptability of homes in distressed areas, as the CDI has announced, is a great step in the right direction. The primary question is whether this exchange with the market will entice carriers to reenter.
Requiring higher acceptability of homes in distressed areas, as the CDI has announced, is a great step in the right direction.
To answer this, we must first understand the reason why carriers need to incorporate models in the first place. Rate indications are intended to ensure that the premiums a carrier charges one year in the future are adequate to cover losses and expenses. That means that the assumptions made must reflect the carrier’s future exposure. If carriers are expected to increase their wildfire exposure, it’s logical that their assumed wildfire losses reflect this increased average exposure. Models can inform a carrier about their future level of risk, based on which exposures they expect to write, including if they were to expand their market share in distressed areas.
That means that a prerequisite to reentering the market is that a carrier’s rate plan already reflects the increased risk. Carriers cannot simply begin assuming more wildfire risk prior to revising their rating plans. This relationship places a high amount of urgency on allowing risk models to be incorporated sooner rather than later if the CDI hopes to act quickly on this crisis.
Carriers cannot simply begin assuming more wildfire risk prior to revising their rating plans.
Future Implications for Consumers and the Market
If combined with risk segmentation, increasing wildfire base premiums does not necessarily mean that all customers will pay more for insurance. In fact, risk-based rating and segmentation means that most consumers benefit from a price standpoint, and all consumers benefit from greater availability. And, with greater acceptability of homes in distressed areas, the admitted market can help depopulate the California FAIR Plan and lower their assessment risk, allowing carriers to remain in the market long-term.
Conclusion: ZestyAI's Call to Action for CDI and Path Forward
To conclude, ZestyAI feels that the CDI is on the right path towards resolving the homeowners' insurance availability crisis in California. We believe that our recommendations can ensure that the public benefits from these changes soon, not years from now.
- First, the CDI must facilitate a timely and predictable rate application review process, especially in cases where the rate requested is justified by the CDI’s own rate application template.
- Next, the CDI must allow rate applications that incorporate models prior to expecting an increase in availability.
- And finally, the CDI should share the burden of model reviews with third-party experts in order to instill public confidence, focusing on inputs, assumptions, and validation on actual wildfire events.
A resolution to this crisis, one that increases availability and improves affordability and rate stability, is within reach.