(Image credit: Dollar Photo Club.)
What if a good driving FICO-like score helped you obtain better insurance rates or made you more desirable to potential employers, much like a good credit score can help you obtain a loan at a favorable rate? Fair Isaac Corp.’s FICO score is used by more than 90 percent of lenders. It makes sense that a similar score measuring driving could be used to enable insurers to better assess risk and help drivers obtain better insurance rates.
In the past, automotive and fleet insurers have relied on loss history to determine risk and renewal strategies. There is no standardization of driver behavior and related risk assessment. To date, even usage-based Insurance (UBI) programs lack defined and uniform standards for assessing driver behavior, apart from miles driven. Ironically, the actuaries who evaluate risk by using statistical data are using proxies of risk exposure that represent only directional representations of actual driver behavior. In the credit world, FICO is essentially used to determine “the likelihood that you will pay all of your obligations (debt) on time for the foreseeable future.” A FICO-like score for driving would essentially be determining “the likelihood that you will be involved in an at-fault accident sometime in the foreseeable future.”
Think about this for a minute: A leading factor in pricing vehicle insurance being used by actuaries today is “how many miles you drive annually.” With that in mind, think about your credit score. It is not based on “how much money did you spend.” Your credit score factors in payment history, amounts owed, length of credit history, types of credit and new credit. In essence, FICO is interpreting your behaviors around money management and the likelihood you will make a payment. Better behavior reaps better rewards.
What if there was a way to truly assess a driver’s likelihood to avoid situations that result in a collision, in addition to those miles driven? Imagine being able to switch from one insurance company to another and being able to provide a specific driver score that is meaningful across all insurance platforms. Imagine also a level of transparency that allows a driver to see the factors—essentially the choices they have made—that affect their driving score. So why aren’t actuaries for vehicle insurance following this model?
In part, the reason is that the risk of an at-fault collision depends on more than just choices that the individual driver makes. We share the road and the risk with millions of other drivers. The state of technology is another factor—there has not been enough computing power, until very recently, that could consume the quantity of behavioral information related to driving.
Insurance companies require this type of standardized information for peer comparison and predictive modeling to forecast future outcomes and results. But absent uniformity in this area, and because actuaries typically require five years of historical data, it has been hard for insurers to keep pace with technological advancements and the volume scale of the data. So why is a uniform standard driving score important? Because it puts every driver on a level playing field when assessing their risk.
How It Would Work
A standardized driver scoring system would consist of both actual driving behavior and contextual elements. Behaviors such as aggressive driving are a choice a driver makes, while distracted driving results in reactions the driver must make to correct a position of risk in traffic. Drivers and fleets would pay insurance rates that reflect their actual driving behaviors. These driving choices are not made in a vacuum, though. To fully understand risk requires a contextual view and data on the driven environment—road service, accident history, time of day, local weather, traffic, congestion and many other circumstantial variables.
Determining a driver or fleet’s underlying risk based on unprecedented visibility into fleet and individual driving data provides better data for better decisions. Understanding a driver’s behavior along with the impact of the driven environment will give insurers the ability to apply rating models to verified driving and vehicle data that they have never had before. It will also create an environment where insurance companies can quote based on the actual safety results of both driver and vehicle in a more timely fashion and reduce time to market. Over time it will also allow insurance companies to reward consumers who invest in active safety systems such as lane departure warning and automatic braking.
For the commercial fleet insurance, there are many different telematics service providers that may offer independent scoring systems or event notification systems; however, with no accepted standard scoring methodology insurance companies can’t uniformly implement these varied scoring approaches. Furthermore, these scoring systems rely on a very limited number of variables which haven’t proven predictive of collisions, at-fault or otherwise. They have only proven useful in finding cohorts of businesses who are lower risk due to use of the TSP system over time and self-selection.
Similar to a credit score, factors in a FICO-like driving system would normalized and weighted so that different vehicles and different scenarios have a uniform impact on a drivers’s score. From the credit rating agency you can receive a copy of your score in advance of your mortgage application and thus have insight into how you will be viewed, or normalized. Transparent and effective normalization of data is a key requirement to produce a commonly accepted driving score that provides significant value to institutions that rely on them.
All drivers, despite differences in driving responsibilities, types of vehicle, speeds driven, etc., must ultimately yield a standardized and comparable score. Just like you can trust a credit score of 750 to mean the same thing to different mortgage lenders, the vehicle insurance world could use that same concept of uniformity to rank drivers, assess risk exposure and streamline underwriting decisions based on a uniform scoring system. Different driving behaviors should be weighted differently which gives insurers the predictive analytic advantage of where the risk is and which vehicles are most likely to be in an at-fault collision.
A FICO-like driving score would need to go beyond traditional methods used to rate drivers and analyze actual driver behavior every second of every trip. The parameters for defining risk assessment using a FICO-like system would have to be multi-dimensional. Using considerably more relevant variables would result in a much more easily explained score for driver feedback.
Some systems offer “event” notifications but are frequently discredited by the driver’s themselves because those events only capture a very small sliver of the overall driving picture. To pinpoint which drivers are the most at risk, the system must catch all the driving behavior between events.
Usage -Based Insurance
For Consumer Vehicles: Vehicle insurance underwriting is experiencing a paradigm shift to UBI. In Personal Lines consumer auto insurance 17 million people worldwide are expected to have tried UBI auto insurance by the end of this year. Much of the traction achieved in the UBI market, thus far, has been due to the nature of self-selection; namely, that good drivers are more eager to participate for the opportunity to pay less insurance. Insurers must now go beyond attracting good drivers only and implement tools with greater predictive capabilities to access broader segments of the market.
For Commercial Fleets: Commercial lines auto insurers are looking for a different scoring system, one that encompasses all of their vehicles and drivers. Overall fleet scores take into account the operating characteristics and installed safety equipment on their vehicles, the environment in which fleets operate (road types, times of day, traffic dynamics, weather, etc.) and the aggregated driving behavior scores of all their drivers.
Heretofore, commercial auto insurers have been collecting a five-year claims history and reported miles on the fleets they insure to determine risk and renewal strategies. A new UBI fleet scoring standard would allow them to compare fleets of similar industries and territory to determine relative risk, leading to a strong desire to elevate the issuance and renewal process with more accurate means of underwriting risk assessment. The combination of direct measurement of driving behavior and contextual analysis of the driven environment significantly outperforms traditional rating variables such as age, gender, and credit scores when estimating risk. As commercial auto insurers well know, the sheer nature of risk is much greater in fleets, compared to the personal lines market, with associated higher premiums and anticipated payouts, often involving larger vehicles with more expensive loads and expanded legal ramifications.
Forward-looking commercial auto insurers are assessing their current business models in light of a more sophisticated risk analysis capability enabled through the use of analytics applied to telematics-generated data. Indicators suggest insurance actuaries are ready to adopt more driver behavior related indices. Such is the underwriting value of more predictive risk identification that many fleet insurers could provide financial incentives to their insureds to give them access to the driver and fleet scoring.
More than ever before, fleets would have the opportunity to monetize the value of their telematics data at an increasingly lower incremental cost thanks to simplified data collection approaches such as Blue Tooth devices and smartphone apps. Further, on the commercial side, limitations and inherent flaws of the self-selection (opt-in) process are eliminated because drivers must adopt management policies, thus creating a more thorough view of insured risk of everyone.
Redefining UBI to RBI: Expanding Beyond Usage-Based to Risk-Based
The concept of UBI could better serve both the insurer and insured by expanding beyond usage-based to risk-based (RBI). Risk-based insurance would have strong correlators to underwriting risk and tell insurers a lot more than a record of how many miles driven, but rather how those miles were driven and what is the associated risk per mile.
With the development of a standardized driver scoring system, RBI would be the perfect counterpart. As insurance companies look for a better way to measure exposure with current day technology advancements, the main selling points of transparency, accuracy and risk mitigation could contribute to the wide-spread adoption of a FICO-like driving scoring standard. Insurers looking to cut costs, improve business practices, and better assess clients’ risk levels, will increasingly invest in tapping into the data analytics that telematics enable.
With a better understanding, insurers will become more effective at identifying higher risk clients, allowing them to price-adjust accordingly or move that risk to another insurer. Carriers that don’t offer RBI will not be able to compete for the safest drivers and will risk losing their best policyholders.
In a new RBI auto insurance world, RBI insurers will attract and retain safer drivers. Consumers will benefit from more accurate policy rating and pricing from insurance companies that use RBI and standardized driving safety scoring. And both consumer and commercial drivers will be able to improve their driving behaviors by knowing their risk score and applying corrective action that will improve safety and positively impact their insurance rates. Implementing incentives for good driving will enable insurers to offer effective risk mitigation and driver improvement programs.
Who Owns the Score?
A FICO-like driving score delivers value to three stakeholders: the individual driver, commercial fleets, and the carriers who insure them. But whose score is it?
In the personal lines market, the consumer owns their personal driving score and the underlying data, and the consumer has the right to disclose the score as needed to obtain more favorable personal auto insurance. For commercial auto lines, where individual driver scores are aggregated to determine a fleet-wide driving safety ranking, the fleet operator owns the data and controls who sees the score and driving data, along with how the score and data are used, whether for safety improvement and loss prevention, insurance underwriting, self-insurance loss reserves, or claims management. The fleet owner may ‘opt-in’ to share driving data with insurers for preferential insurance coverage or service.
Insurers want a standard data set, but they aren’t ready to just accept the score, they need the underlying data in a standard form. Consumers are ready for a standard score if they can be assured of transparency and location privacy. Telematics and safety technology advancements will have huge implications for the insurance industry over the next five years. Research indicates that commercial auto insurance companies are willing to pay 2-10 percent of annual premiums for accurate telematics scores to receive the value of better underwriting efficiency and risk mitigation capabilities. Further proof that a FICO of driving is transforming the auto insurance industry is the high probability this would most likely be a joint endeavor between the telematics service providers and insurance companies who are willing to cost-share in exchange for data.
A valid driving score would elevate underwriting by providing a uniform and portable scoring system that accurately predicts driver risk. Both personal lines and commercial insurers would benefit from being able to price risk more accurately, identifying the riskiest drivers, reducing accident frequency, and reducing exposure to risk.
Telematics and UBI programs are moving into the mainstream, despite numerous technological challenges and deficiencies. Turning raw driving data into actionable correlations is a major challenge facing insurers. If there were the perfect telematics scoring solution, universally applicable to risk underwriters, that could determine exactly how telematics data-points translate into a valid predictor of risk, it would open up a whole new perspective for the underwriting process. Market competition is forcing insurers to consider rating variables that outperform traditional proxies such as age, gender, and marital status. Usage-based insurance (UBI) is at the top of their list. With a uniform standard of driver behavior scoring and analytics, insurance companies will be able to accurately assess driver performance based on actual driving habits on a risk-per-mile basis.
By 2020, over 50 million US drivers will have tried UBI insurance, making a FICO-type driving score almost inevitable. SAS predicts by 2020 over a quarter of all U.S. auto insurance premium income will be generated via telematics, representing more than $30 billion. Carriers that don’t offer UBI or RBI are in danger of adverse selection because they won’t be able to compete for the safest drivers and policyholders.