Auto Insurance Industry Donates $15.4 Billion

Transcription

1 Auto Insurance Industry Donates $15.4 Billion Premium Leakage a Significant Hit to Underwriting Profit A Report by Quality Planning November 2011

2 In 2010, premium leakage cost auto insurers $15.4 billion. Most of it could have been prevented. In 2010, rating error reduced private passenger auto insurance industry premium revenue by $15.4 billion. The reduced revenue was essentially a multi- billion- dollar industry donation handed directly to policyholders although no charitable tax deduction accompanied the effort. The estimate is based on nationwide auto policy audits conducted by Quality Planning, a Verisk Analytics company that specializes in the validation of policyholder information. Premium rating error represents 9.36 percent of a total $ billion in personal auto premium written. Without action, insurers will continue to lose billions more over the coming years all due to premium leakage that could be stemmed if appropriate action were taken. In this report, Quality Planning has aggregated and summarized audit results of roughly 5 million policies from multiple carriers. The data includes substandard to preferred books of business, all distribution channels, and national and regional carriers. 1 State- level data is then weighted to reflect the total national private passenger auto line. In this report 1. Symptoms: The Effects of Rating Error... page 3 2. Causes: What s behind Rating Error and Premium Leakage? Data: Losses Broken Down by Rating Factor Commentary: Comparison of 2008 and 2010 Statistics The R x for Relief: How to Reduce Premium Leakage Data used in this study was limited to audits for which Quality Planning retained contractual rights to aggregate data for industry analysis. Page 2

3 1. Symptoms: The Effects of Rating Error Direct Premium Loss For individual carriers, reducing rating error is a significant opportunity for profit gains. The majority of auto- rating error reduction drops right to the bottom line. In a small margin business where, in a good year, individual carriers might have average profits of 5 percent of premium, adding 1 or 2 percent of loss- free premium can make a substantial difference to an auto writer s combined ratio. Risk Management Costs Rating error leads directly to failures in risk management. As an example, policies with unrated 16- year- old male drivers in the household experience an average loss ratio of more than 200 percent. Moral Hazard Costs An often- overlooked cost of rating error is moral hazard. Analyses repeatedly demonstrate that individuals who misreport policy- rating information are associated with high loss experience. For example, an individual driving 20,000 miles per year but reporting an annual mileage of 5,000 will, on average, have greater claim costs than an individual driving and reporting 20,000 miles. Misreporting results in low- risk drivers subsidizing high- risk drivers. Similarly, the majority of agents who work to accurately determine premium have a strong interest in rating integrity. In the absence of meaningful controls, however, the honest agent is placed at a competitive disadvantage by a minority of agents willing to rate a policy inaccurately to obtain a cheaper quote and close a sale. Business Management Costs The modern insurer relies heavily on rating and underwriting data in all primary areas of corporate management. Policy data provides key inputs to marketing, sales, business segmentation, financial planning, corporate planning, and staff compensation, among others. Errors or worse still, systematic biases in underwriting data deteriorate performance in all management functions. Page 3

4 2. Causes: What s behind Rating Error and Premium Leakage? Rating error can happen at any stage of the underwriting cycle: sales, risk analysis, policy servicing and renewal. While significant error occurs at initial application, analysis shows that the majority of rating error and massive premium leakage arise through changes in rating factors over time. Misrepresentation Every auto insurer must deal with those individuals who intentionally misrepresent their circumstances and then rely on the fact that most companies are either indifferent or incapable of detecting the misrepresentations. Although such policyholders make up a very small proportion of the entire customer base, their actions cause significant disproportionality with respect to premium and loss. The corollary to this dilemma is that companies making the most basic effort to reduce premium leakage can often see substantial and immediate gains in profitability. Unfortunately misrepresentation is not limited to individual policyholders. Some agents purposely input incorrect data into their policy management systems to obtain lower rates and secure new accounts. Companies whose agent incentive compensation agreements are heavily weighted to reward new business and place less emphasis on profitability are particularly susceptible to this problem. Change Another factor that complicates auto rating is the fact that Americans lead more dynamic lives than ever before. Every hour there are: marriages and 121 divorces 167 drunk- driving arrests 4 traffic fatalities More than 2,800 auto insurance claims paid 152 new driver s licenses issued to 16- year- olds 4,235 people who move 6,119 people who change jobs As a result, the risk profile of auto policies is constantly changing, making personal auto insurance risk management a moving target. Consider job changes: The time has long passed 2 Marriage and divorce statistics from Census Bureau Current Population Reports; move estimates based on U.S. Census 2009 Current Population Survey. Job change estimates from Bureau of Labor Statistics. Drunkdriving arrests sourced from MADD website. Traffic fatalities from Department of Transportation, and auto claims paid from National Association of Insurance Commissioners and National Association of Independent Insurers Fast Track. Page 4

5 when a worker got a job soon out of school and stayed with the same company throughout his or her career. In fact, the average worker has held ten jobs by the age of 36. Overall, 25 percent of workers change jobs every year. Why should auto underwriters take note of this trend? Because if a person s job changes, it is also likely that he or she will have changes in vehicle usage, commute distance, and annual mileage. Unlike homeowners insurance, the basic facts of a personal auto insurance policy change frequently: More than 50 percent of household auto policies experience a vehicle or driver change every year. Nearly 30 percent of households replace vehicles every year. Newly acquired vehicles are often the impetus for change in other important areas such as: Who is driving what car? How far is each car driven? Which car is now the family commute vehicle? While insurers provide their policyholders with multiple methods to report changes, many changes aren t reported at all. Not surprisingly, policyholders are significantly more likely to report life changes that reduce auto premium than report changes that increase it. In fact, analysis shows that policyholders are five times more likely to report mid- term policy changes that lower annual mileage or commute distance than to report changes that increase mileage or commute. Every day, in the course of conducting policy audits, Quality Planning uncovers examples of this behavior. When asked about other household members who might drive a policy s listed vehicles, one mother told Quality Planning, My son doesn t drive my car. He only uses it to go to school each day. Verifying students away at school can also garner interesting responses, such as the policyholder who told Quality Planning, My daughter doesn t need to be added to my policy. She just has one of the cars at the university in case she needs it. As indicated in these examples, most premium leakage is not the result of subversive plans to defraud the insurance company. Rather, it is more likely the profit- sapping consequence of busy people leading busy lives. Unfortunately, the ongoing economic turmoil has exacerbated the frequency of change to underlying policyholder information so much so that Quality Planning statistics may actually understate the true magnitude of the affect change has on total auto premium leakage. Job losses and job switching are the most influential, but household relocations and household make- up are also important contributors. Page 5

6 3. Data: Losses Broken Down by Rating Factor Insurers lost about $15.4 billion in 2010 because of auto premium rating error. It s important to understand how this error is broken down, revealing areas where leakage can and should be prevented. Types of Rating Error The study of premium rating error for 2010, conducted by Quality Planning, found substantial rating error exists in all common factors used to determine auto premium. Table 1 presents estimated premium loss by rating factor. Table 1: Estimated Premium Loss by Rating Factor Private Passenger Auto Vehicle Rating Factors Percent of Premium Total Error Cost ($ Billions) Commute 1.03% 1.7 Annual Mileage 0.87% 1.4 Vehicle Usage 0.88% 1.4 Vehicle Characteristics, Discounts % 0.5 Rated Territory 0.82% 1.4 Vehicle Subtotal 3.90% $ 6.4 Driver Rating Factors Unrated Drivers 1.66% 2.7 Vehicle-Driver Assignment 1.12% 1.8 Driver Characteristics, Discounts % 1.9 Violations/Accidents 1.20% 2.0 Driver Subtotal 5.16% $ 8.5 Other Rating Factors % 0.5 Total Rating Error 9.36% $15.4 Note: Rating error costs were found to vary greatly by individual insurer. The amount and type of rating error varies by many factors, including characteristics of the book of business, geographic location, distribution channels, rating plan, rate pursuit history, regulatory environment, agent relations, and underwriting standards. 3 Includes symbol, safety discounts such as alarms, and vehicle body type discounts. 4 Includes years driving experience, age, marital status, student discounts, affinity group membership, driver identification such as DL and SSN. 5 These factors vary greatly by carrier, including multi-car discounts, years insured, credit score, and multiple products. Page 6

7 4. Commentary: Comparison of 2008 and 2010 Statistics The year 2010 saw a slight decrease (0.41 percentage points) in auto premium leakage compared with 2008 (the last time this report was published), as a result of a combination of factors. There is ample evidence that people, on average, drove somewhat more in 2010 than in This finding is consistent with the Federal Highway Administration report of an upward trend in vehicle miles traveled (VMT) in 2010 compared with the decline in According to the U.S. Department of Transportation, Americans drove about 0.9 percent more, or 26.2 billion additional miles, in 2010 than in However, premium leakage from misrepresented mileage estimates actually declined slightly (0.03 percent). We believe that the overall mileage increase in driving hides the fact that there was a segment of the population whose driving was drastically reduced because of job losses and tougher economic times. Ironically, many of the individuals who had previously understated their mileage ended up properly rated and the premium self- corrected. Many of the drivers who actually increased their annual miles in 2010 did so in pursuit of more pleasurable activities (e.g., more local vacations rather than destinations by plane). This phenomenon, however, likely had a minimal impact on premium because an increase would apply only to those who were not already in the highest mileage- rating category (often 10,000 to 15,000 miles per year). Premium leakage related to the commute- rating factor also increased minimally in Once again, poor economic conditions with widespread job losses and fewer job openings are the most influential factor. Workers demonstrated their flexibility and willingness to commute longer distances for a job. Added to this was the housing crisis, which forced an increasing number of people to move when they could no longer afford their mortgage payments and/or faced foreclosure. According to the Census Bureau, the percentage of people who reported a move in 2010 increased to 12.5 percent compared with 11.9 percent in This equated to about 37.5 million people, or 5.3 million more than those who changed residences in Surprisingly, premium leakage from garaging address misreporting was stable at 0.82 percent. Anecdotal evidence suggests that record numbers of grown children moved back to their parents or grandparents homes typically in more affluent and subsequently lower- risk neighborhoods. The result was more truthful reporting and less premium leakage but lower overall paid premium to their insurer. Finally, there was a very small decline (0.02 percent) in premium leakage from policyholders who use their vehicle for business but report its use as strictly for pleasure. This trend can be traced to the construction work slow- down and the decline in home sales throughout Quality Planning also found a sizable decline in premium leakage due to errors stemming from driver characteristics and discounts (0.27 percent) and a decline in violations and accidents (0.20 percent). Better technology, data sources that provide more accurate information, and increased focus by insurance carriers to update their underlying policyholder data are having an impact. However, there is a small but expensive upward trend in the number of unreported drivers (0.06 percent) and, to a lesser degree, the correct assignment of household vehicles to household drivers. There was an increase in the number of people between the ages of 18 and 24 who were living with their parents due to the economic downturn and job losses. The trend Page 7

8 was most striking in large urban areas where the cost of living is noticeably higher. Nationwide, one to two percent of all policies written include an unrated operator who is most often a high- premium younger driver. Household consolidation has resulted in more opportunities to hide younger drivers and simultaneously misreport who is driving each vehicle. Page 8

9 5. The Rx for Relief: How to Reduce Premium Leakage Auto insurers can prevent multi- billion- dollar premium leakage by leveraging new detection methodologies. These tools address the new dynamics of America changing jobs, locations, life circumstances, and so forth. By adopting new approaches to ensure accurate rating, insurers can reduce premium leakage and increase profits. Here s how: A Three- Step Rating Integrity Process Reduces Leakage by 60 Percent Rating error can be introduced at all stages of the underwriting cycle. While significant error occurs at initial application, Quality Planning s analysis shows the majority of rating error arises through changes in rating factors that occur over time. On average, after initial policy screening, 82 percent of audits uncover policies lacking enough premiums to cover the intended risk. By efficiently identifying errors in rating data, carriers can take the steps necessary to correct costly errors and restore profitability to their book of auto policies. Step1: Get it right at the point of sale. To ensure rating integrity and limit premium leakage, insurers need to employ available technologies and conduct sophisticated data analyses. This should be done on all new business, in real time, from an agent s or customer service representative s desktop. New business audit checks should also be applied to numerous rating variables to identify potential rating errors. These audit checks should look to verify: Vehicle- driver assignment Annual mileage Vehicle garaging territory Unlisted drivers Accidents and violations Identity theft Commute distance VIN identification Farm use Non- owned vehicles Vehicle symbols Salvage vehicles Business use Marital status Commercial use Criminal records With additional analytic tools such as pattern analysis and statistical algorithms, insurers can flag questionable policyholder information for a variety of findings, including: Vehicles garaged at mail- drop addresses Households with unreported youthful operators Incorrect vehicle- driver assignments Underreported commute distances False driver s licenses and Social Security numbers Commercial vehicles insured as private passenger autos Page 9

10 Step 2: Get it right at renewal. Whether or not a policy is reviewed at new business, there is a significant likelihood that changes in everyday lives will create a different risk profile over time. A change in marital status, job changes, new cars, new houses, and new 16- year- olds all these create very different risks. Annual review of renewal policies can keep insurers on top of these changes. The renewal process is similar to the new business review but should be more in- depth. Intelligent reviews would identify only those policies that have a high likelihood of changes to rating elements. Once these policies are identified, insurers should employ an efficient and effective validation program. The policyholder contact strategy needs to be comfortable yet comprehensive. To achieve optimal household contact, today s active policyholders also need multiple response options, most important of which is an ability to respond via the Internet. With this strategy, it s possible to achieve customer contact rates of more than 80 percent. Contact is important not only for securing customer acknowledgment of rating variable changes but also for improving the overall customer insurance experience. Far too often, a policyholder s only contact with his or her insurance company is the annual or semi- annual invoice. An annual checkup call or contact goes a long way toward maintaining a positive and long- term company- policyholder relationship. Step 3: Develop a long- term approach. Once an insurer has completed steps one and two and cleaned up its auto book of business a baseline of accurate information exists to enable regular maintenance of a premium leakage program. The baseline will help enable affordable rating integrity effectiveness in the coming years. The key is to avoid lapses that could allow rating error to creep back in. Studies show that changes made during reunderwriting have a life span of two to three years. That means there is substantial lifetime value in a rating integrity program that goes far beyond the first- year cost of execution. The bottom line is that a great deal of time and effort goes into a one- year cleanup effort. Significantly less time and effort, though, is required to keep an auto book clean. Smart companies realize this and have integrated rating integrity into their regular renewal processes. Page 10

11 About Quality Planning A Verisk Analytics company, Quality Planning is focused on providing rating integrity solutions to auto and home insurers. Quality Planning works with insurance companies to identify areas of rating errors using sophisticated database management, statistical analysis and modeling, customized survey design, and highly targeted customer interaction. Quality Planning helps clients work within their existing rating plans and charge fair prices to policyholders, based on a true representation of risk. The company was founded in 1985 and is headquartered in San Francisco. For more information, visit Page 11

12 Appendix: Audit Methods The 2010 premium rating error report aggregates and summarizes audit results of around five million policies from multiple carriers. The data includes substandard to preferred books of business, all distribution channels, and national and regional carriers. 6 State- level data are then weighted to reflect the total national private passenger gross written premium. Two primary methods were used to develop the estimates of rating error: statistical risk estimators and direct measures. Statistical Risk Estimators: The first method we employ to estimate rating error is to compare the expected distribution of rating factors to the rated distribution. In the case of annual mileage the expected distribution is the distribution given the characteristics of policies written. We expect the average new Ferrari to be driven an average of 3,500 miles per year and the average new Chevy cargo van to be driven more than 20,000 miles. Based on numerous studies of vehicle use patterns we have estimated and validated equations which develop an expected mileage based on vehicle make, model, and year; number of vehicles in the household; garaging ZIP code; number of drivers in the household; age and occupation of driver and so on. Actual odometer reading data from over 80 million vehicles was used in developing the statistical models. For every vehicle insured, expected mileage is compared to reported mileage to detect any patterns of systematic error. Direct Measures: The second method we use to estimate rating error is direct measurement. For over a million vehicles in the sample we had data for multiple odometer readings to evaluate actual annual mileage. In addition, for multiple carriers, we interviewed over one million insureds concerning their vehicle usage patterns, drivers in the household, address verification, and annual mileage. Results of the odometer and interview data, in turn, were used to validate and refine the statistical models. Statistical and direct measures were combined for each carrier in the sample and contrasted with rated values. These were then consolidated for this industry report. 6 Data used in this study were limited to audits where Quality Planning retained contractual rights to aggregate data for industry analysis. Page 12