12 January 2021Accounting & tax analysis

Tracking and analysing loss development for captives

Actuaries field many types of questions relating to actuarial analytics, but one topic that arises time after time is about loss development. What are loss development factors? Why are they needed? How can I use loss development triangles to improve my risk management efforts?

Questions such as these reveal a desire to gain a deeper understanding of loss experience, but also a potential lack of direction in doing so.

As actuarial professionals, we are immersed in the analytical world, but that is often not the case for many insurance professionals, including those within the captive insurance industry. It’s vital that topics such as loss development are communicated effectively to ensure the benefits of analytics can be understood and used by as many key decision-makers as possible.

This is especially important for captives because the proper use of analytics is a significant factor in the long-term viability of any alternative risk mechanism. Moreover, there are several captive-specific issues that can be alleviated or overcome through analytics.

Estimates of future loss

We first need to establish a common understanding on topics relating to loss development and their relevance to captive insurance. To start with, let’s discuss loss development factors, which are commonly defined as an estimate of future loss development based on historical loss experience and the age of the loss in question.

In an actuarial report, these factors are applied to recently evaluated losses in order to derive an ultimate loss estimate. The method of performing such a calculation is often referred to as a “development method”, and is commonly found within an actuarial analysis, along with other risk-appropriate methods.

“The proper use of analytics is a significant factor in the long-term viability of any alternative risk mechanism.”

Loss development factors themselves are calculated using loss development triangles, wherein loss data is formatted in such a way that loss experience (or development) can be measured over time. These triangles can be established using any regularly cadenced interval, but annual intervals are the most common. Often, loss development triangles are used to measure a company or captive’s unique loss experience, but in some cases they may be used to track industry-wide trends for the creation of benchmark factors.

Generally, incurred and paid losses are tracked through development triangles, but numerous other items can also be analysed this way, including total claim counts, open claim counts, average claim severities and case reserves. Broadening the scope of development triangle use provides an opportunity to obtain a better, more nuanced understanding of a company’s risk portfolio and risk management efforts. It also allows a captive’s decision-makers to base risk-related decisions on a firm, analytic foundation.

Identifiable trends in loss emergence, claim lag, and similar types of loss activity can play a large role in the timing and process of changes to a risk portfolio. The captives industry experienced one such example of analytically-based decision-making during the COVID-19 pandemic. Captives professionals who were already adept at tracking and identifying loss-specific changes were at an advantage when faced with such unprecedented uncertainty, whereas those who were less adept could react based on only a limited amount of information.

Development triangles may be considered as a primary starting point in any captive analytic efforts. However, there are a few issues regarding development triangles that captives may encounter at various stages of their lifespan, such as changing third party administrators (TPAs) during captive formation.

When a captive is formed, the captive owners may choose to use a new, independent TPA moving forward, or a fronting carrier may administer claims for them. If such a change occurs, it may impact the manner in which claims are administered. Awareness regarding the potential impact this could have on development triangles will be important when reviewing the captive’s analytics in the future. In some cases, it may even be advisable to hold discussions with the new claims administrator to cover this change and the ensuing impact on development.

Emerging risk

Another possible issue surrounds the placement of emerging risks into a captive. Such risks often lack a measurable loss history, or the loss experience may not be credible, as the amount of claim frequency or overall volume of losses is relatively small. This lack of history may make it difficult to create useful development triangles on an initial basis, but that doesn’t necessarily mean analytics can’t still be obtained.

Any available information, such as claim report lags, market quotes, or various types of internally maintained data, may be used to establish a basic understanding of the risk. From there, risk-specific development can be tracked on an ongoing basis to supplement current information.

As an example, consider reputational risk, the placement of which in captives has grown significantly in recent years. While “loss history” in the traditional sense may not be available for a specific captive owner, various risk management or risk-adjacent departments may be able to identify similar situations in the company’s past. Once these situations have been established, information could be gathered pertaining to their impact or severity on a monetary scale, such as changes in sales or revenue.

“Information could be gathered pertaining to their impact or severity on a monetary scale, such as changes in sales or revenue.”

These data points may not exactly be transferable to a development triangle, but they could be used as a form of internal benchmarking for when losses do arrive in the future.

Finally, a common issue encountered by ongoing captives is related to Accounting Standards Update (ASU) disclosures in the US. We are often asked by our captive insurance clients how the triangles used in those disclosures relate to the development triangles provided in the actuarial report. One such triangle used in ASU disclosures is for ultimate losses, and while it is commonly labelled as an “incurred loss and allocated loss adjustment expenses” triangle, it actually differs from the incurred loss development triangle.

In essence, ultimate loss triangles are measuring how the ultimate loss estimates for historical periods have changed over time. For regulatory purposes, they are used to determine the predictive accuracy of the methods used in an actuarial report, which may include the aforementioned development methods. While the ultimate loss triangle may not immediately appear useful to a captive owner, it can be used in discussion with their appointed actuary to cover the methodology of their analysis and determine if any changes are needed over time.

Differences such as these also highlight the importance of understanding the terminology surrounding actuarial analytics, and how various terms may differ depending on the audience. Being able to navigate situations where terminology is used in multiple ways may help prevent confusion and potential regulatory issues. Naturally, this will ensure ongoing the captive’s operations proceed in an efficient, undisrupted manner.

The captive insurance landscape is generally growing more analytically sophisticated as it matures, as are potential captive owners. While it may have been accelerated by the uncertainty surrounding the COVID-19 pandemic, such a shift was always likely to occur. Long-term captive strategies are more reliant than ever on analytics and, as we’ve discussed, a key component of that analytic foundation is the use and understanding of development triangles.

The importance of this analytical topic, as well as many others, must be impressed not only on day-to-day captive employees, but also on members of the captive board. Once the efforts of a full, top-down push for analytical education have been realised, those involved with individual captives, as well as the industry at large, can rest assured that their futures are based on solid data and reliable analytics.

Enoch Starnes is an actuarial analyst at SIGMA. He can be contacted at:

Michelle Bradley is a consulting actuary at SIGMA. She can be contacted at: