Soft Market—Hard Choices
By Thomas Ryan, Chair, CAS Committee on Reserves 

As we head toward year-end, many actuaries are preparing to perform loss reserve analyses in support of financial statements and Statements of Actuarial Opinion. The good news for the property-casualty industry is that, according to several sources, carried loss reserves were estimated to have been redundant at year-end 2006. Both Conning Research and Consulting and Morgan Stanley have estimated a redundancy of nearly 5% of carried reserves for the industry at year-end 2006. These estimates are a dramatic change from the deficiencies (generally accepted to have been in the billions of dollars) that existed in the industry in recent years. The view that reserves may be redundant for many companies and for the industry overall has shifted the interest and attention of many reserving actuaries to the variability in these reserves and stochastic reserving methods. It is important, however, that we not lose sight of gathering dark clouds on the horizon. These clouds do not belong to some looming natural catastrophe but rather of the rapidly advancing “soft” market for property/casualty insurance.

The current underwriting market (soft or softening—depending on your viewpoint) is characterized by increased competition for business in many, if not most, segments of the industry. This competition has led to decreasing rates and increasing coverage and will likely result in increasing loss ratios. While analyzing loss reserves is a complex endeavor in any period, estimating loss reserves in a soft market has proved especially problematic for actuaries in the past. The impact of the underwriting cycle on reserves is an area that has received much attention in the U.K., specifically within GIRO, and was the focus of a recent working party. The same level of attention has not yet been given in the U.S. to this issue. Recent results in the U.S. indicate a distinct correlation between the underwriting cycle and initial reserve adequacy.

Table 1 provides a summary of the initial reserve adequacy for the ten most recent accident years (1997-2006). Initial reserve adequacy is measured here as the difference between the carried net ultimate losses and defense and cost containment (DCC) expenses from the first evaluation of each accident year (at the 12-month evaluation) to the latest evaluation (year-end 2006). The accident years shown contain those from the last recognized soft market—roughly 1998-2001—as well as those from the following years in a hardening market. The data presented is taken from the 2006 industry aggregate Schedule P Part 2 for all lines of business combined as provided by Highline Media. The latest estimates of ultimate losses for each accident year are based on the evaluation at year-end 2006 and could continue to change, particularly the more recent years.

     

Table 1

As shown in Table 1, the net ultimate losses for accident years 1998-2001 increased by approximately $56 billion in total since the first evaluation of each year. This increase includes continued adverse development of $4.0 billion in calendar year 2006 alone. For each accident year, we graphed the percentage change from the first year evaluation for that year to the latest evaluation in Figure 1. This graph is an update of work done by Bob Conger and presented at the 2003 GIRO meeting and shows the cyclical nature and correlation of the adequacy of initial booked ultimate losses with the underwriting cycle. As shown in Figure 1, the soft market years (1998-2001) show an increase in ultimate losses since the first evaluation while the hard market years (post-2001) are currently showing a decrease in ultimate losses.   

Figure 2 provides a view on how the net ultimate loss ratios for these same accident years have changed since their first evaluations (at 12 month’s maturity). As shown in this figure and as expected, the soft market accident years (1998-2001) have the highest net ultimate loss ratios (all above 75% at the latest evaluation). The ultimate loss ratios for these years have been almost strictly increasing since the first evaluation. The remaining accident years have lower loss ratios (all below 70%) and have nearly all been strictly decreasing in magnitude since the first evaluation.

In order to avoid a recurrence of the adverse development experienced in the last soft market or at least temper its magnitude, reserving actuaries need to account for the underwriting cycle in their thought processes and reserving approaches, especially in the early years of a soft market. Some basic suggestions to accomplish this follow:

1. Know the Business—Reserving actuaries must talk to underwriters and others in the business unit to better understand the business for which loss reserves are being estimated. In a soft market, the mix of business within a line can change as companies move strategically to write or re-underwrite certain territories or classes. These changes can affect the appropriateness of the use of historic internal or industry development patterns. It is also important to understand how much of the business is new business as compared to renewal business. New business is traditionally “won” through lower prices or broader terms of coverage and often shows higher ultimate loss ratios than a review of historic loss ratios would indicate. Any expected loss ratios (ELRs) used in reserving methods for new business may need to be higher than those for similar renewal business.      
2. Track Changes—In a soft market, it is critical to monitor average rate-level changes and loss trends in order to select reasonable ELRs. It is also important to monitor attachment points and deductibles as they tend to decline in a soft market while coverage limits rise. These changes could impact the length of historical development patterns and the appropriateness of using such development patterns. Also, the impact of changes to policy terms and conditions (which usually increase coverage in a soft market) on estimated losses is difficult to quantify but it is important for an actuary to understand in order to consider judgmental adjustments.   
3. Review Indicators—Actuaries must compare internal data with available market information and attempt to reconcile the two. For example, are internal rate monitoring indications very different from published industry averages for certain segments? While the internal monitored results may be correct, it is valuable to understand why differences may exist. This knowledge will help in the selection of ELRs used in reserving methods.      

These are only three broad suggestions to help actuaries begin to recognize the potential impact of underwriting cycles on the reserving process. For more in-depth discussion on the market cycle and its impact on the reserving process, two good resources to review are (1) “Market Cycle Management: Blunt and Straightforward” by Mark Lyons of Arch Insurance—the keynote address at the 2007 Ratemaking Seminar and (2) “The Cycle Survival Kit” ( http://actuaries.org.uk/files/pdf/proceedings/giro2003/Line.pdf), a working party paper by published by GIRO in 2003. Nonetheless, there is much more research to be done in this area, particularly in regards to the impact of market conditions on loss development patterns.   

Many industry analysts have optimism that the current soft market we are entering will be different from those in the past due to increased discipline from the increasing role of enterprise risk management. Other leaders believe the fundamental economics of the insurance business have not changed enough from prior times and, therefore, the current soft market will end badly. As actuaries, we need to do our best to generate an accurate view of realistic outcomes based on our current knowledge of the business and market. By ensuring that proper impartial information is available, we can help ensure that well-informed decisions are made.

In his presentation at the CAS Ratemaking Seminar, keynote speaker Mark Lyons stated that one of the key objectives of an actuary in a soft market is to become unpopular. Continuing to use existing methods and assumptions without reacting to the underwriting cycle may be an easy choice for an actuary to make. Asking tough questions, digging deeper into the data, and providing potentially bad news early on about profitability is a hard choice. In this case, as it often is, the hard choice is the right choice for actuaries to make.