Auto Claims Frequency, Severity: What’s Behind the Rise?
By John E. Cantwell and Dorothy E. Kelly
At the 2016 shareholders meeting for Berkshire Hathaway, Chairman and Chief Executive Officer Warren Buffett was asked about the declining financial performance of one of the holding company’s insurers, GEICO. Buffett responded, “[L]ast year both frequency, how often people had accidents, and severity, which is the cost per accident…both of those went up quite suddenly and substantially.” With the insurance industry facing ever-tighter margins, this Verisk analysis looks at some factors driving those changes, as well as their potential implications for policyholders and insurers.
Since the early 1980s, miles driven by U.S. drivers had been climbing steadily. But the Great Recession of 2008 brought an unprecedented decline in mileage exposure. The downturn in aggregate miles outlasted the recession, lingering until 2012 before drivers’ mileage began to rise again.
As the economy improves and gas prices decrease, the trend is that more people are on the roads, and more often. Miles driven set new records in 2015—and on a seasonally adjusted basis that number is continuing to climb even higher in early 2016. Americans drove a combined 3.1 trillion miles over the 12 months through March 2016.
Rise in accidents
The increase in miles driven carries an unfortunate component. Accident frequency is at its highest level in a decade, rising steeply in 2014 and 2015 in strong correlation with the spike in miles driven—and bringing with it shrinking margins for auto insurers. To demonstrate this trend, Verisk examined two common non-bodily injury coverage types of losses for property damage: liability property damage and collision physical damage. For both categories, the rise in frequency mirrors the increase in miles driven.
As demonstrated in the above graph, as of the end of 2015, frequency for property damage climbed to 3.55 percent, while frequency for collision reached 5.96 percent. Amid the recession and the resulting drop in miles driven that lasted through 2011, many auto insurers generally benefited from frequency rates that were on average almost 0.15 of a point lower for property damage and almost 0.35 of a point lower for collision than today’s rates. Multiplied across millions of policies, this dip in frequency—in combination with lowered severity—contributed to insurers keeping premium increases relatively flat in a competitive market.*
A Surge in Severity
The recession also led to an increase in the average age of vehicles on the road. In what has been labeled by industry insiders as “carpocalypse,” new car sales took seven years to rebound to 2007 levels. The recession meant fewer new cars were being driven; so, on average, accidents cost insurers less.
The average age of light vehicles on U.S roads climbed to 11.4 years at the end of 2012, but as more new vehicles were sold from 2012 to 2015, the rate of increase in vehicle age slowed. One business management consultant report forecasts the average age will inch up to 11.6 years in 2016 and not reach 11.7 years until 2018.
With better times boosting sales of new cars, severity is setting new records, generally due, in part, to the high cost of repairing modern safety technology. Since 2007, severity for collision increased at an average annual rate of 1.68 percent, but that rate has been accelerating since 2012.
Between 2007 and 2011, when fewer new vehicles were being sold, average annual severity for this coverage increased only 0.27 percent. As more new vehicles hit the roads between 2011 and 2015, average annual severity jumped to 3.10 percent (or 1.75 percent adjusted for inflation, while severity in 2015 dollars actually declined during the recession and the continuing downturn in miles driven).
Property damage shows a similar pattern, with average severity increasing 1.26 percent between 2007 and 2011, but jumping to 3.93 percent between 2011 and 2015 or 2.57 percent adjusted for inflation, while severity in 2015 dollars declined during the period between 2007 and 2011.
For bodily injury, a number of factors can influence the likelihood of injury. For example, misalignment of bumpers between passenger cars and sport utility vehicles may raise the risk, while the improving safety features of modern vehicles are often a mitigating factor. The frequency of bodily injury claims has been relatively stable before and after the recession, but severity continues to rise, largely affecting losses for auto insurers. While bodily injury has had a significant impact on profitability for years, the rise in frequency and severity on the physical damage side is exacerbating the impact on margins across the industry.
Deaths per 100 million miles driven had been declining markedly since the mid-1930s, from 15 to just over one. In 2015, this trend reversed, and the National Safety Council (NSC) reported an 8 percent spike in road fatalities between 2014 and 2015, the largest increase in 50 years.
While the exact cause of this troubling spike is not clear, there are potential contributing factors:
- Density: We've entered uncharted territory in terms of the number of miles driven. Not only does exposure to accidents increase as people drive more, the exposure of more vehicles in proximity to one another also rises.
- Distracted driving: As the chart below illustrates, many drivers admit to unsafe driving behaviors. As interactive technology pervades our society, the distraction it brings behind the wheel may be leading to increased frequency and severity for insurers, with a potentially greater toll for society.
- Defective vehicles: A greater share of newer cars has come onto highways since 2012, but there remains a dichotomy between new and old. Newer cars drive up property damage claim severity, while a small but increasing number of older vehicles with safety issues may explain some of the increase in number of severe accidents. Verisk analyses of data supplied by three large insurers show the presence of branded titles (prior total-loss vehicles and vehicles that are deemed beyond their mechanical limits) has more than doubled over the past five years.
Ways Forward for Insurers
Auto insurance claims frequency has returned to prerecession levels, and severity has hit an all-time high. This is a pain point for the industry. This may correlate with many insurers raising premiums. But these trends could hardly come at a worse time, since they meet a broader disruption in the industry. Competitors using technological and analytical advances are putting pressure on insurers to elevate the customer experience and refine underwriting and rating as never before. And a simple price-based approach to the frequency and severity squeeze may be falling short.
Premiums for personal auto insurers grew 5.5 percent in 2015, although that has not been enough to compensate for rising losses. According to data reported to A.M. Best, auto insurance adjusted loss ratios climbed by a half percentage point in 2014 and almost two-and-a-half percentage points in 2015. Further, a Verisk analysis of this data over the past three years shows that only one in five premium dollars belongs to insurers that are hitting the sweet spot of profitable growth.**
“We estimate premium leakage—revenue lost through misreported or omitted underwriting information—is costing auto insurers an amount equal to between 10 and 15 percent of direct written premium annually,” said John Petricelli, vice president of Verisk – insurance solutions. In fact, the 2016 Verisk Auto Insurance Premium Leakage Survey found more than 80 percent of insurance leaders are at least “moderately concerned” over premium leakage, with nearly half indicating that they were “very concerned” or “extremely concerned.” But there are tools available to tackle this problem. Insurers have access to a number of innovative solutions that can better align risk and premium and keep them aligned over time:
Mileage solutions: Since underestimated mileage accounts for more than 18 percent of leakage and is strongly correlated with frequency, accurately determining the miles a vehicle will be driven and maintaining this record over the life of a policy can help stem rising losses. Many new tools, including sophisticated analytics, smartphone apps, telematics, and connected-car data, can help empower insurers. Collecting mileage at point of sale and over time allows for refined rating.
Rating symbol models: Using predictive modeling that accounts for frequency and severity by make and model can help insurers adjust to the higher costs of repair for modern automobile technology.
Loss history indicators and driver monitoring solutions: The increase in frequency and severity means insurers will see more shopping from prospective customers who are dissatisfied with rate increases stemming from a hardening market. Loss history indicators and driver monitoring are cost-effective solutions that can identify up front in the quote workflow if recent claims or violations may be an issue.
Prioritized pursuit: At the point of sale and renewal, insurers can use sophisticated risk scores and projected losses for drivers and vehicle types—for example, a safe driver with a minor infraction or a vehicle with a branded title—and help inform decisions that balance short-term premium pursuit against potential lifetime policyholder profits.
Verisk has a number of analytics solutions designed to address these and other challenges. Please contact the authors at firstname.lastname@example.org or email@example.com to discuss how Verisk can assist in forming an effective strategy to combat the effects of rising frequency and severity on insurer margins.
*To account for seasonality, a rolling average of the last four quarters was used for measures of frequency and severity. In comparing past frequency with current rates, adjustments for this lag time were made at the start of the recession.
**The analysis used private passenger auto insurance data reported to A.M. Best and defined ”sweet spot” insurers as maintaining or growing market share in combination with either outperforming the industry average or improving upon past performance for adjusted loss ratios between 2012 and 2015.