How DFA Can Help theby Susan T. Szkoda, FCAS
Editor's Note: This is the third in a series of articles on Dynamic Financial Analysis (DFA). Click here to view Part I and Click here to view Part II.
In my previous DFA article on pricing, I noted the importance of being realistic about the fundamentals and the importance of having a "true compass" as to what your real costs are. Nowhere is this more true than in reserving.
Sound reserves form the bedrock of financial stability. In addition to being important to the balance sheet, sound reserves provide that true compass for your real costs and true pricing needs.
As we get drawn further into the current cycle we need to be careful to separate fact from wishful thinking. If we look back on the last cycle in the early-mid 1980s we will note that industry reserve adequacy deteriorated sharply - more or less in line with the deterioration in rate level adequacy.
It is possible that a component of this deterioration was due to overly optimistic assumptions about companies' ability to control costs through judicious underwriting or claims practices. Part of it may have been due to the belief that investment income would more than offset any shortfall in pricing. Part of it may have been due to management pressure.
I came upon an old Actuarial Review letter from the 1984 CAS President Carlton Honebein. I was struck by how timeless his comments were:
"Insurance results are horrendous. Reinsurance results are even worseÉWe had the forecasts and we believed them; why weren't we more forceful in demanding actions that would have avoided an unsavory situation that profoundly affects ourselves, our employers and the insuring public."
The mantra of the last cycle was that high interest rates would bulletproof all pricing and reserving decisions. The mantra of this cycle is that low inflation rates will bulletproof all pricing and reserving decisions. Time will tell.
Are we currently at the point where insurance results are horrendous? No, of course not! Reported results today are in fact extremely favorable. Could we get to this point within the next few years? It would certainly seem to be possible.
How can DFA help a company avoid or minimize a bad outcome?
Let's focus on a single line of insurance - workers compensation - and walk through some specific considerations in a DFA-style reserve analysis.
Assume the actuary has available the basic data elements of premium, exposure units, claim counts, paid losses, case reserves, type of claim (permanent total vs. permanent partial vs. temporary total) and type of loss (medical vs. indemnity vs. allocated loss adjustment expense).
What are some of the items that separate a DFA-style reserve analysis from a "traditional" reserve analysis?
- The item being analyzed can be viewed as the hub of a wheel. The spokes of the wheel are all other relevant areas/considerations that interact with the hub. There are many actions and reactions based on the flow of information between the hub and spokes.
- The reserve analysis itself is multivariate. It is explanatory in nature. The highly summarized "triangle" approach will generally not be considered sufficient.
- The analysis is scenario-based and range-based. Major variables are examined as ranges rather than absolutes. The impact on the "wheel" and strategic actions are determined at a minimum at the major points of the range (low, medium, high scenarios).
Our DFA workers compensation reserve analysis for Hypothetical Company X considers the following items:
Nature of Book of Business/Premium
- Assume 50 percent of the voluntary book is National Account business. Over the past three years this book has converted almost fully to the High Deductible Product. As a result premium for this segment has dropped from $1 billion in 1993 to $250 million in 1996.
- Assume 50 percent of the voluntary book is Guaranteed Cost. Rate levels peaked in 1993 and have fallen by 10 percent per year in each of 1994, 1995 and 1996. In addition to this schedule credits increased from 0 percent in 1994 to five percent in 1995 to 10 percent in 1996. Therefore premium for this segment dropped from $1 billion in 1993 to $625 million in 1996.
- Assigned Risk Servicing Carrier premium decreased by 70 percent between 1993 and 1996 due to the depopulation of the assigned risk pool. Assume 1993 premium was $1 billion while 1996 premium was $300 million.
- Assume underlying payroll unit growth in National Accounts decreased by 15 percent between 1993 and 1996 due to restructurings and layoffs in the Fortune 500 client base.
- Assume underlying payroll unit growth in the Guaranteed Cost segment increased by 8 percent over this period as smaller businesses added staff and grew.
- Assume the bulk of business is traditional manufacturing exposure in the Northeast, Mid-Atlantic (NY & NJ) and Midwest. There is little presence in fast growing South, Southwest and High Tech Silicon Valley/ Pacific Northwest regions. The mix of book by state and industry group is generally stable.
Renewal Retention Rates
- There is high retention (90 percent) on National Accounts.
- Smaller Guaranteed Cost business has average retention (70 percent).
Company X has slashed its expenses by $350 million over the past three years but finds the expense ratio has actually increased from 25 to 34 percent, due to lower premium levels and change in product mix of business.
The 30 Year Treasury yield was around 6.5 percent at year-end 1996. It now appears to be 7.0 percent. Will it head north to 8.0 percent?
The portfolio is 80 percent bonds, 10 percent stocks and 10 percent cash. The bond maturity distribution is:
30 percent 0-5 years
20 percent 5-10 years
10 percent 10-30 years
40 percent +30 years
- Wage inflation over the past few years has been approximately 0-2 percent per year. historical wage inflation has been approximately four percent per year.
- Annual medical inflation over the past few years has been -5 to +5 percent based on where you are in the managed care cycle. Historical medical inflation has been approximately 7 to 12 percent per year.
- The current table used is the Standard 1980 Table. Lifespan has since lengthened by 5-10 percent.
- Impaired Mortality on Permanent Totals is used where appropriate.
- Significant historical reforms occurred over the1992-96 period with significant rate reductions relative to reforms. Pricing is uncertain on many reforms. The impact on claim department case reserves is also uncertain.
Claim Department Changes
- Significant staff downsizing has occurred, including the loss control unit.
- The claim department reports no significant change in case reserving procedures. However, review of data indicates:
Average paid claim inflation running between 0 and +5 percent.
Average outstanding claim inflation running at -10 percent.
Claim settlement rates appear to be flat to perhaps a modest slowdown.
Cost of excess of loss and stop loss reinsurance has decreased significantly and is now widely available. In 1993 this coverage was expensive and difficult to obtain.
The precise impact of many of these variables on the final reserve need is unknowable at the time of the reserve review. Therefore it is necessary to construct scenarios using plausible ranges for these variables. Judgment comes into play. DFA helps us understand and articulate the impacts of the many judgments we routinely make as actuaries.
Our enhanced understanding of the impact of many of these variables helps us to feed back meaningful information to the marketing, underwriting, claims, reinsurance, investment and pricing areas of the company (spokes and hub concept).
Come see how this workers compensation DFA reserving problem is solved at the Casualty Loss Reserve Seminar, September 29-30 in Atlanta, GA. The solution will be published in the Winter Forum.