Auto Insurance
Public comment closed
Comment Due Date

Thank you for providing your feedback on FSRA’s Loss Trend Benchmarks for Private Passenger Automobile Major Rate Filings.

We appreciate the comments and questions received to date. This consultation is now closed. 

FSRA uses ratemaking Benchmarks to evaluate if an auto insurer’s proposed rate changes are just and reasonable when an insurer seeks approval for those changes. 

Benchmarks are updated twice a year to identify industry cost trends and provides insight to the regulator’s outlook on future loss costs trends and forecasts. 

Benchmarks analysis, consultation input and FSRA’s selection of new benchmarks will be made public as part of our commitment to transparency. 

Useful Links

#By submitting your content, you agree to have your materials posted on our engagement portal, used in reports and other materials prepared by Financial Services Regulatory Authority of Ontario (FSRA) that may be shared with the public. Content is moderated so that all posts are respectful and professional. The Freedom of Information and Protection of Privacy Act, R.S.O. 1990, c.F.31, applies to all online content.

Use left and right arrows to navigate between tabs.
Sector Comment Date posted Sort ascending
Auto Insurance
[2020-006] Jingwen Li - TD Insurance
A. Selection of ultimate losses for BI & AB
a. According to IBC loss development exhibit, insurers have implemented various changes on case reserve protocol and claims handling process for BI and AB, which makes the reporting pattern unstable and makes it difficult to estimate the ultimate losses for the most recent years. Therefore, the most recent AYs should be given less weight in the trend selection.
B. Trend model selection for BI & AB
a. Longer term trends should be considered for long tail coverages. The split trend method used by OW could potentially overfit the most recent AY and does not generalize to future AYs.
b. A sensitivity test should be done with different data points being excluded or included to assess the stability of the trend selection.
i. The most recent data points should be excluded when selecting the loss trend for long-tailed lines, because most claims have not been reported or developed and the reporting pattern is unstable (see comment A).
c. Reduction of BI loss cost after 2016 reform does not indicate that the BI loss cost will continue to decrease by 7.5% annually beyond 2019.
C. Trend model selection for Collision & Comprehensive
a. In general, using a 7 year period for short tail coverages looks like a too long time frame. There is less volatility in the ultimate loss cost projections and using and 5 years looks like a better balance between credibility and responsiveness.
D. Stability of benchmark selection and rate indication
a. Since OW methodology heavily relies on the reported claims in the most recent AYs, particularly on the long-tailed coverages, we expect the selected benchmark to vary significantly year over year. This will introduce significant volatility into the rate indication if adopted by the industry, which will eventually translate into large premium swing in the market and will not benefit the customers.
b. We recommend FSRA to adopt a more prudent approach to ensure the stability of loss trend and reform benchmark selection, which will benefit both the industry and the customers.
c. We recommend FSRA to adopt a long term view when selecting benchmark methodology.
E. Individual insurer could have loss experience that differs from the industry experience, which could justify different selection of loss trend and reform factors.
F. Given the stability of the data has been impacted by changes in economic environment, past reforms and companies' claims practice, FSRA should consider different point of views from other independent actuarial consultants in addition to Oliver Wyman's study. For example, IBC retained an independent consultant to perform the loss trend analysis using GISA data and the outcome was drastically different from OW selection. If Oliver's Wyman's assumptions do not materialize as expected, mandating the industry to adopt those assumptions could cause some insurers to need higher rates increases or limit insurance availability.

Auto Insurance
[2020-006] Laura Ni - Farm Mutual Reinsurance Plan
Given the material changes in assumptions and approach between FSRA Analysis of Reform Cost and Loss Trend Rates published in October 2019 and June 2020, we would appreciate FSRA providing supporting information to fully explain the rationale and impact of these changes. For example, it appears that June 2020 report did not apply reform adjustment factors to adjust historical costs to a common cost level before performing statistical regressions, which is contrary to the standard procedure.

We were unable with our current set of assumptions to identify a decreasing or neutral trend in BI and AB loss cost after the 2015/2016 reforms within Ontario Mutuals (Company 789) Private Passenger Auto data that is commensurate with predictions set out by this report. We are prepared to submit detailed justification for selected trend rates should they differ from FSRA recommended ones in future PPA filings.
No questions have been asked about this consultation yet.