Part 2 of 4
By Ronald T. Kuehn FCAS, MAAA, CPCU, ARM, FCA, Kim Piersol FCAS, MAAA, & Alan Pakula ACAS, MAAA of Huggins Actuarial Services, Inc
In Part 2 of this article we discuss examples of predictive modeling usage in claims leakage studies in order to mitigate the adverse loss development that would have otherwise occurred.
Case Study #1: A single-state property and casualty insurance. Lines of business include workers’ compensation, general liability, property and auto.
The Problem: Predictive modeling and data mining algorithms identified deterioration in the workers’ compensation medical line of business. Average outstanding losses per claim and average incurred losses per claim were increasing over time at rates much higher than the industry.
In early 2011, the insurer retained a claims consultant to perform a comprehensive review of the practices and methods of the claims department, as well as of all open claims. This review identified the following issues:
· Extremely long claim reporting lag – 10 days to report from insured to the insurer
· Limited medical management “best practices” used
· Too many poor quality physicians and providers on the medical current panels
· Lack of partnership with employer on return to work
· Adversarial model of claims handling resulting in 80% of lost time claims being litigated
· Extremely low success rate on litigated cases
· Stair-step reserving
· No red flag methodology for repeaters and problem claims
The Solution: The insurer implemented an injury management model which is considered to be best practice in workers’ compensation claims handling. This model focuses on care and concern for injured workers in order to provide the best medical care for a full recovery and return to work. The idea is simple: reduce the number of claim denials and the number of litigated claims since non-litigated claims cost significantly less than litigated claims.
The claims consultant retrained claim staff in all aspects of claims management. A review of all recent open claims was performed with an increased emphasis on claims settlement. The claims staff was instructed to focus on closing and containing higher exposure lost time and complex medical claims using action plans including Medicare set asides (MSA), compromise & release (C&R), aggressive legal intervention or full investigation.
The insurer established claims management controls including director approval required for increases of reserves above $20,000, new claim denials, new litigation, and all independent medical exams and surveillance. Case reserves were set to most the probable outcome.
The Result: Within a year of establishing these new claims department processes, the insurer experienced a decrease in the number of lost time days on open claims in 2011. The average number of litigated cases has dropped from an average of 18 per month in 2008 to 0 in the first half of 2011. Legal expenses are significantly lower.
Case Study #2: Regional Medical Malpractice Insurance Company (RMMIC) is a regional Property Casualty Insurance Company that specializes in the writing of Medical Malpractice Insurance and operates in several mid-western states. RMMIC writes a book of predominantly physician medical malpractice, some hospital medical malpractice and general liability. Their business is produced by a combination of their own direct sales people and independent agents. They handle their own claims and have a number of claims offices that are located throughout the states in which they do business.
The Problem: RMMIC has an internal actuarial department headed up by a Chief Actuary named Roger Goodcalc, who routinely performs a quarterly reserve analysis for the company. In Mr. Goodcalc’s analysis using data as of 06/30/09, his estimates for the reserves rose by approximately $25 million for prior years (2008 and prior) on a total indicated reserve of approximately $500 million. This rise was noticed by management and when asked the reason for the rise in indicated reserves, Mr. Goodcalc responded by saying “that is what my calculations showed”. Three months later, using data as of 09/30/09, Mr. Goodcalc’s estimates for the 2008 & prior years rose by another $25 million for a total of a $50 million increase on an original base of $500 million of reserves. Now RMMIC’s management was very concerned and called in an Actuarial Consulting firm (Actuarial Firm A) and asked them to review Mr. Goodcalc’s analysis as of 09/30/09. Actuarial Firm A performed a full review of Mr. Goodcalc’s work and then issued a report saying they agreed with Mr. Goodcalc’s conclusions. Management then called in a second Actuarial Consulting firm (Actuarial Firm B) to study the situation. Actuarial Firm B reviewed Mr. Goodcalc’s analysis, reviewed the report of Actuarial Firm A, and performed its own independent analysis. Actuarial Firm B then issued a report saying that they fully agreed with Actuarial Firm A and with Mr. Goodcalc.
The Solution: Management hired Actuarial Firm C to come in and review the entire situation and try to determine why the actuarial estimates had risen so sharply. Actuarial Firm C came in and requested a detailed electronic listing of all claims that were active in the three years prior to and up through 09/30/09. Using a series of analytical techniques that included predictive modeling and data mining, Actuarial Firm C was able to determine that in general, RMMIC’s claim examiners would set mature case reserves that would eventually settle for a value approximately 40% below the mature claims reserve. They also noticed that during the three month period from 4/1/09 through 6/30/09, the historical 40% redundancy on closing claims had started to decrease to a lower redundancy. In performing his analysis, Mr. Goodcalc was correctly projecting his reserve estimates using the lower redundancy level and increasing his estimates by $25 million. Actuarial Firm C further noticed that during the three month period from 7/1/09 through 9/30/09, the redundancy on closed claims continued to decrease. Actuarial Firm C then performed more comprehensive data analysis on all claims closing during the period 4/1/09 through 9/30/09. They subdivided the claims by the closing redundancy on the claims and noticed that they ended up with two distinctive groups of claims. Group 1 claims all closed with redundancies that scatter plotted around 40%, and the Group 2 claims closed with redundancies that scatter plotted around 25%. It was noted that historically, prior to 4/1/09, closing claims had acted similarly to the Group 1 claims. So now it was theorized that the Group 2 claims were a new phenomenon to be studied. Actuarial Firm C brought in their claims auditing specialists to pull the claims files of all Group 2 claims in order to find out if they had common characteristics. In performing claim file by claim file comparisons of Group 2 claims to Group 1 claims, the claims auditing specialists determined that the Group 2 claims were being settled for values that were too high. They found approximately $10 million in “claims leakage” for the Group 2 claims that were settled between 4/1/09 and 9/30/09. It was also determined that the Group 2 claims were settled in a shorter time span than Group 1 claims, and that they were being settled by a broad cross section of RMMIC’s claims offices and adjusters. Detailed interviews were held with the RMMIC’s claims adjusters that had handled the Group 2 claims and also Group 1 claims. In addition, the Senior VP in charge of claims at RMMIC was interviewed. It was then discovered that in order to save $1 million, he had closed one of their claims offices in early 2009. All of the claims examiners in this office were terminated and their claims filed were transferred to other claims offices within RMMIC. The net result of this event was that the Group 2 claims were now being handled by different adjusters that did not feel the same responsibility for these claims as the terminated adjusters at the closed claims office. The adjusters that inherited these claims were not very happy with their increased claims load and had moved to settle these “extra” claims in order to get them off their desks. Senior management of RMMIC had been obsessed with the $1 million expense savings from closing the claims office and had not imposed adequate internal controls to make sure that the Group 2 claims were being settled with the normal cost efficient settlements. The Group 2 claims became “orphan” claims that no one wanted. Those who were assigned these claims became resentful, felt overburdened, and tended to settle them too quickly. Thus the $1 million in expense savings from the closing of the claims office led to claims leakage of $10 million due to improper controls during the process.
The Result: The Management of RMMIC installed proper controls and review mechanisms for the handling and settlement of the remainder of the Group 2 claims. During the quarter ending 03/31/10, all claims settled were back to averaging a 40% redundancy. The actuarial estimates for the 2008 and prior years began dropping and in the next two quarterly actuarial analyses, Mr. Goodcalc’s estimates for the 2008 and prior years dropped by $40 million, back to his original estimates prior to the closure of the claims office (the $40 million drop plus the $10 million in leakage on the closed Group 2 claims accounted for the entire $50 million increase in estimates by Mr. Goodcalc.) As the remainder of the Group 2 claims settled, they were closed at the historical average redundancy of 40%.
In Part 3 of this article we will discuss additional case studies in claims leakage.
Ronald T. (Rusty) Kuehn is a Fellow of the Casualty Actuarial Society and the Conference of Actuaries in Public Practice, a Member of the American Academy of Actuaries and the International Actuarial Association. In addition, Mr. Kuehn holds the Chartered Property-casualty Underwriter (CPCU) designation and the Associate in Risk Management (ARM) designation. email@example.com
Mr. Piersol is a Fellow of the Casualty Actuarial Society and a Member of the American Academy of Actuaries. Prior to joining Huggins, he served as Senior Vice President & Chief Actuary for Crum & Forester Insurance Companies. Prior to that he served as chief actuary for CNA Insurance Companies, and was a consulting actuary for KPMG LLP, Arthur Anderson LLP, and CFO of AIG Risk Management. He has over forty years of experience in the actuarial field. firstname.lastname@example.org
Alan Pakula is an Associate of the Casualty Actuarial Society, and a Member of the American Academy of Actuaries. He has served as a Managing Actuary at the New Jersey Department of Banking and Insurance. Prior to that, he was an actuary for the Reliance Insurance Company. Mr. Pakula has over nineteen years of experience in the actuarial field email@example.com