Actuaries Have Special Role When Explaining Credit Scores and Losses

November 16, 2007

  • November 16, 2007 at 10:23 am
    Dustin says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    I am surprised no one has jumped on this to say how the use of credit score is so terrible, and not an accurate indicator of risk. I only hope that the actuaries are able to explain why we should use credit scoring to rate; however, even if they can give the best explanation that doesn’t even leave one inch of room for doubt, people will moan and complain about it because anything that makes their insurance go up is wrong. While credit score is not the sole determinant in accidents or propensity for taking risks, when used with other underwriting and rating data it provides a much more accurate rating of the risk.

  • November 16, 2007 at 10:52 am
    Lowell says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Credit scoring is the most important rating factor. Auto and homeowners rates are based on risk. Driving record, age, sex, marital status, type of vehicle, choice in past on coverages, territorial riskiness, living near storm areas, are all measures of risk RESPONSIBILITY… Credit scoring is one more measuring tool that separates every level of risk’s responsibility. 10000 persons with no claims history in past five years will have 1000 or so auto claims in the next 12 months. Since none of them have a past claim’s history…the most accurate measure to group those least likely to have a loss as a measure of responsibility…Credit scoring does that best.

  • November 16, 2007 at 12:46 pm
    Lynne says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Lowell I couldn’t agree with you more that the key word is responsibility and that being rich,poor,black,white has nothing to do with. Credit is the best way to measure responsibility.

  • November 16, 2007 at 2:25 am
    Reagan says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Dustin, you could show ten years worth of film of irresponsible people being responsbile for higher losses due to their behavior and because of the fact that most of them are minorities, no one will lend credence to this sudy. I love how they say people attack credit scoring due to not understanding why their is a correlation between it and risk.
    Me thinks they undestand it just fine.

  • November 16, 2007 at 2:34 am
    Dustin says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Is that your argument, that you can make statistics say whatever you want them to say? Hasn’t the argument that irresponsibility in your life, finances, etc leads to irresponsible driving been proven yet? I think the argument makes perfect sense. People who take risks in their day to day lives are more likely to take risks on the highway.

  • November 16, 2007 at 3:15 am
    Patrick Butler says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Here’s a letter I sent to participants of the CAS modeling meeting reported on:
    Dear Participant,
    Re: Exhibit about the FTC Report on Credit-Based Automobile Insurance Scores
    NOW’s seminar exhibit supplies the economic logic noted but not developed in the recent FTC Report to explain what causes the correlation of more claims with lower credit scores.
    Not only lower credit scores but every group indicator of tight budgets (such as lower income zip codes and lower paid occupations) must predict more miles per car for the group and therefore more claims per 100 car years. WHY? Because pay-by-the-car, all-you-can-drive premiums cause financially constrained drivers to insure fewer cars and drive each more miles. This explanation is further documented by these two items:
    1) Texas NOW fact sheet (2003) “Mandatory Automobile Insurance . . . Why high premiums for low-credit-score or low-income drivers can’t be regulated away!” Side 2 describes the free-market, cents-per-odometer-mile remedy. (#736 on website)
    2) Short paper (2007) “Why Low Credit Scores Predict More Auto Liability Claims: Two Theories” (to appear in the Journal of Insurance Regulation and available at the exhibit). (The negligent driver theory discussed in the paper is supported by the biological correlates study by Credit Scoring Update panelists Brockett and Golden.)
    Although the FTC Report presents a truncated version of the economic logic (page 32, citing a 2006 academic paper of mine), it does not consider the inevitability of the correlation of more claims with lower credit scores caused by a need to economize on car insurance. Furthermore, the report states without discussion that “companies find it difficult to capture information on ‘miles driven’ with a great deal of accuracy,” and ignores the study published in the 1993 CAS Forum (referenced in my 2006 paper) on the practical application of per-mile premiums under state insurance law and the 1972 Federal Odometer Act.
    For high car insurance prices there is no such thing as “no cause.” By default, drivers with financial problems are currently assumed to be negligent drivers. NOW posits instead that these drivers are no different from other drivers but are economizing by insuring fewer cars, which, as illustrated by the enclosed fact sheet, must lead to high car insurance prices.
    Please stop by the NOW exhibit for further information. I can also be reached at 202.628.8669, ext. 148, or by email: pbutler@centspermilenow.org.
    Sincerely,
    Patrick Butler

  • November 16, 2007 at 4:22 am
    Hon. Scott A. Adamsons says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    It has been my general experience that the more complex an explanation is, the less credibility it has. Credit scores are based on repayment histories, and therefore make their use in evaluating the risk of future debt repayment reasonable. Now that’s simple. What the actuaries are saying is not close to being simple. In fact, how much of contents in the proverbial black box do they know? Sometimes a correlation is valid and other times it is a post hoc ergo propter hoc – a logical fallacy and an assault to the senses of reasonable people.

    As a banking executive who has risen through the ranks from being a part-time teller in my college days to a vice president of a national savings institution, I can tell you that credit scores alone are indicators at best. Good underwriting comes from having a good understanding of the subject, the risk if you will. In the case of the insurance industry, that would be the insured risk. Unfortunately, this concept still evades many people in the banking industry, where credit reports and scores have been used extensively for decades. It is not a far stretch believe that fresh users of this type of information have much to learn before they using this tool responsibly. It takes time to develop this kind of experience.

    From a banker’s perspective, a person whose credit score is average or below average is more often than not a higher risk prospect; however, the credit score alone is not used to make that determination (nor is it in insurance). An underwriter must be responsible and dissect the applicant’s credit history, look for trends, irregularities, and learn about the person behind the report. Here are some factors that can impact a person’s credit score but might not impact their personality as an insured risk: Single income households, parents of multiple children, prior divorce, a billing error that leads to a collection action (happens more people than one would think), identity theft, and mistaken identities on the part of the reporting agency.

    In some of the cases above, individuals and families who would otherwise qualify for a low premium based upon their actual claims history and risk profile would be bumped to a less desirable pricing tier due to their credit score. A case example: In the case of a single-income household, where one parent sacrifices a salary to stay at home and raise their children, it makes a world of difference. In real life, such a family made the right decision. Instead of turning over the second income to a child care provider, they forego the income and eliminate that expense, an expense that would not likely be reported on a credit bureau report anyway (at least for timely payments). However, these households typically carry a higher leverage burden as there is only one income. This impacts their credit score adversely. It’s not quite that simple, but the net trickle-down effect is the same. So in this example, a family doing the right thing, for the right reasons, pays more in premiums than a comparable family with two incomes and kids in daycare. This is but one example where credit bureau reports and scores may fall far short for their use in an insurance application.

    Another thought to consider is that the formulas that the credit reporting agencies’ use are unknown to the general public and will not likely be released anytime soon. It takes a great deal of experience to use this kind of a tool in an underwriting process in a way that is not prejudicial or skewed. Even then, the experts sometimes make poor decisions.

    The only good reason for using the information from a credit bureau report is to hypothecate future risk of debt repayment — that is after all why they were initially created. Repayment history is used to predict future repayment behavior — that is about as apples-to-apples as it gets. Actuaries will have to be more convincing in their position promoting the use of credit scores for risk profiling in their industry. Show me.

  • November 16, 2007 at 4:34 am
    Bill says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Good explanation and theories.
    A couple of things from my own observations. First of all, blanketing all formulas that insurance companies use under the same umbrella is incorrect. Some companies use pretty much a straight credit score. Most use credit as one criteria, but factor in many other (secret) variables.
    The national media has always assumed that because a person is low-income or doesn’t live in the right part of town, they automatically pay more in premium. That is definitely not the case. In fact, I have seen the opposite happen on more than one occasion.
    Credit scoring is nothing more than the latest predicting tool. It will continue to undergo many changes in the future, and will never be a perfect tool. Unfortunately, many/most underwriters are no longer given any latitude when it comes to what a customer pays. If their score says they pay X, then they pay X or go elsewhere.

  • November 17, 2007 at 12:25 pm
    Gill Fin says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    the desire to sell more insurance by better rating risk and then offering the lowest rates to the best drivers.

  • November 17, 2007 at 12:31 pm
    Gill Fin says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    and credit scoring is the latest tool to offer insureds what they want. Furthermore, ability to pay is a factor when considering desirability because policies that lapse for nonpay and then are reinstated are in a different cost and profit category than policies that don’t lapse for nonpay. Also, as pointed out in another posting, not all insurers use the tool in the same way. When a client comes to the office we look at all their policies. 9 out of 10 times their auto premium has gone down, usually by $20 or $30 per policy period. That is a result of their good credit working to their advantage. I keep reading about other companies and other outcomes, but thats how it works with my company.



Add a Comment

Your email address will not be published. Required fields are marked *

*