損害保険アクチュアリーのレポートは、アクチュアリー以外の方からブラックボックスであるとみられることがよくあります。こうしたレポートは通常、簡単な説明で始まり、それに続いて大半の読者が見る気のしない多数の小さな表示による多くの図表で構成されます。レポートに徹底的に取り掛かる際、目は直ぐにごまかされますが、少数の重要項目をレビューすることでインサイトは得られます。リスクマネージャーおよびその他の読者が、アクチュアリーのレポートで確認すべき４つの重要な情報を以下に紹介します。これにより、アクチュアリーがデータで見ていること、変化のドライバー、またフォローアップの質問や内部調査が適当であると考えられる箇所をよりよく理解できるはずです。これらの項目は、エグゼクティブ・サマリーまたはレポート冒頭付近のサマリーレベルの図表に掲載されていることが多く、何十ページもめくって関連図表を探す必要はありません。

Property & casualty actuarial reports are often seen as a black box by non-actuaries. The reports consist of a brief narrative that is usually followed by a large number of exhibits in small print with more numbers than most readers care to review. It’s easy for eyes to gloss over quickly when delving into the report, but insights can be gained by reviewing a few key items. Below are four key pieces of information that risk managers or other readers should look for in an actuarial report to gain a better understanding of what the actuary is seeing in the data, drivers of change, and where follow-up questions or internal investigation may be appropriate. Each of these items is often found in the Executive Summary or summary-level exhibits near the front of the report, so searching through page after page of supporting exhibits may not be necessary.

1. Change since prior analysis

The first, and potentially most obvious, metric to review is the change since the prior analysis. Rather than focus on the change in reserve, focus should be on the change in ultimate losses. As a reminder, the ultimate losses are the combination of the reported or known losses and the unknown losses, usually referred to as incurred but not reported (IBNR) losses. Generally, this information will be provided in a summary-level exhibit, similar to the table in Figure 1.

While there often won’t be detailed information on the drivers of a change in the exhibit, the drivers will likely be discussed in the report with supporting reasons for the change in estimates. Reasons for a change in estimates may be one, or several, of the following:

• Changes in actuarial assumptions or methods
• Changes in underlying exposures
• Shifts in broader market trends
• Changes in the legislative environment
• Significant difference between actual paid or incurred activity and the expected activity

Regardless of the driver of change, if the change in estimates is material, it is best to discuss with the actuary. It may lead to uncovering assumptions the actuary is making that may not be accurate or applicable, or perhaps it could lead to potential concerns within the business that should be investigated further. Understanding the change in estimates is part of good risk management practices.

2. Actual vs. expected

Actual versus expected (AvE) activity was mentioned above as a potential driver of changes since the prior actuarial analysis, but may be worth reviewing even if the overall actuarial indications are unchanged because it could be a leading indicator of the direction of future changes.

Risk managers may need to request the actuary to include AvE exhibits in the report if they aren’t already. When they are included, AvE will generally be provided for both paid activity and incurred activity, and will show the actual activity and expected activity by accident period. Depending on the number of lines of coverage included in an actuarial report, an AvE exhibit may be shown as a single summary-level exhibit, or may be shown for each individual line of business. AvE exhibits will be displayed differently depending on the actuary putting the exhibits together, but they often will include information similar to what is shown in Figure 2.

It may seem like a lot of information to digest, but from a simplified standpoint the only metric that needs review is the difference between the actual incurred losses and expected incurred losses, or column (8) in Figure 2. If actuarial estimates have recently changed, a review of this column can be useful in pinpointing whether changes are the result of a single line of business, individual accident periods, or simply that the business in total is performing better or worse than expected. In Figure 2, we can see that actual activity has been significantly better than expected activity. It’s likely that the actuary would have lowered estimates in this analysis, with the 2020 accident year as a main driver of the decrease, given that the actual incurred activity of negative $184,000 was favorable compared to an expected incurred emergence of just under $800,000.

If the actuary hasn’t altered their estimates, it’s still worth a look at the AvE exhibits to see whether any anomalies stand out that may influence discussion with the actuary or lead to further investigation internally. If actual activity differs significantly from expected activity, it could be the result of a number of reasons, including:

• Significant claim activity.
• Random variation in actual activity due to either a very small book of business, or the impact of large claims where timing is difficult to predict (for example a large medical malpractice claim that settles for several million dollars).
• Changes in reserving or payment patterns in a given time period. An example of this could be a slowdown in settling litigated claims due to delays in court dates related to the COVID-19 pandemic. Any changes of this matter should be discussed with the actuary.
• Inaccurate prior ultimate estimates.
• Selected loss development patterns that differ from the underlying business.

3. Loss ratios or pure premiums

Briefly, the loss ratio is the ratio of ultimate losses divided by earned premium. The pure premium (sometimes referred to as “loss cost”) is the ratio of ultimate losses divided by exposures (i.e., the loss per exposure unit). Units of exposures vary by line of coverage, but are expected to be correlated to loss amounts for a line of coverage. For medical malpractice coverage in hospitals, a unit of exposure is typically an occupied bed equivalent. In workers' compensation, the exposure is payroll, and for general liability coverage at a retail store exposures may be square footage or revenue.

When it comes to projecting ultimate losses for future periods, actuaries will typically utilize the loss ratio or pure premium in their calculations. Given its significance in budgeting for future periods, an exhibit that shows historical and selected loss ratios or pure premiums should certainly be reviewed. Either, or perhaps both, metrics will likely be provided in a summary-level exhibit in the actuarial report and can be a quick gauge as to how the business is performing over time. If loss ratios are increasing over time, it may mean that premium rates need to be increased, or perhaps the rating structure needs to be revised. If the pure premium is increasing over time, it may mean there are underlying trends in the line of business that are driving insurance losses up. Frequency, or claims per exposure, may be increasing. Severity, or claim cost per claim, may be increasing due to legislative trends, social inflation, or various other causes.

It may also be the case that loss ratio, or pure premium, for a given accident period is substantially different from any other due to an anomaly such as a large claim or event. If this occurs, it’s worth discussing with the actuary whether they are treating that period any differently from others in their analysis, given the potential impact on future projections.

Lastly, the actuary may provide a comparison of the loss ratio or pure premium to that of a broader industry benchmark. If the industry metrics are materially lower, it may be worth investigating internally what may be driving the losses higher than others in the industry.

4. Frequency and severity

Frequency and severity were briefly mentioned above. When reviewing frequency metrics in the actuarial report, it’s best to refer to ultimate frequency, which would include claims reported to date plus those not yet reported. Similarly, for severity it’s best to review ultimate severity, which would consider both the losses incurred to date (i.e., paid amounts plus case reserves), and the IBNR amounts, which could be additional development on claims already reported, or expected incurred amounts for claims not yet reported.

Oftentimes, you’ll find these statistics in the summary-level exhibits in an actuarial report. If any trends are apparent, or particular accident periods look out of line with others, it could be worth investigating internally to determine whether there are any underlying causes. For example, if frequency appears to be moving higher each year, then perhaps an investigation into the root cause of the increase in claims may reveal that more stringent risk management procedures are needed. If severity is increasing each year, then perhaps there are additional claim handling metrics that could be reviewed to be sure that claims are being adjudicated in the best ways possible. Or perhaps diving into whether the increases in severity are driven by increases in indemnity or legal expenses would be valuable.

Similar to the loss ratio and pure premium discussion above, it may be the case that frequency or severity in a given accident period is markedly different from any other due to an anomaly such as a large claim or event. Again, if this occurs it’s worth discussing with the actuary whether they are treating any claims or occurrences in that period any differently.

Conclusion

Although much of the actuarial report may seem like a black box and can be difficult to interpret, the items mentioned above are generally easy to find, simple to understand, and can provide useful high-level indications for the C-suite and board members of what is happening within the book of business . Any anomalies, patterns, or simple observations can help them formulate additional questions for the actuary to better understand the actuary’s viewpoint on the business, or assumptions underlying the analysis. Further, the insights gained can provide useful information to investigate internally to identify areas of improvement and also verify the success of risk management. Lastly, if any of the above metrics aren’t easily found in the actuarial report, or additional discussion would be helpful to better understand the above metrics or any others, it’s best to ask. Your actuary would welcome the opportunity to share insights and better your understanding of the report.