Data and analytics for captives

The captives space has benefited from the growing interest and reliance on data, due in part to the nature of many risks handled through captive insurance, say L. Michelle Bradley and Enoch Starnes of SIGMA Actuarial Consulting Group, Inc.

"This breadth of industry-wide information can serve as an exceptionally useful tool for risk management." L. Michelle Bradley

The business world’s reliance on data and analytics as decision-making tools has undoubtedly increased over the past two decades. Captive insurance has, of course, been no exception to this shift. One could even argue that the captives sector played a role in data’s rise to prominence within the risk strategy context, just as it has for many other innovations.

With that in mind, the team at SIGMA take this opportunity to reflect on how captive insurance’s relationship with data and analytics has changed and, perhaps, where it might be heading.

One stark example of how this relationship has changed can be found in anecdotal memories of how data was provided 20 years ago. Even in the early to mid-2000s, it was not uncommon for SIGMA’s analyses to rely on boxes of paper loss runs shipped directly to our office door. Even more common was the issuance of our reports in bound, physical form, but these physical formats soon gave way entirely to their digital counterparts, such as PDFs and Excel spreadsheets.

Those PDFs and spreadsheets remain the most popular methods of providing data and analytics, but we’re beginning to see the ascent of new forms, such as cloud-based data warehousing and dynamic (and sometimes interactive) analytical tools. While the accessibility of these emerging forms is still being honed for mainstream use, their benefits are already becoming clear.

You might wonder why anyone would be interested in the method or format of data delivery. How could the way data is provided have any impact on the resulting analytics? The answer can be boiled down to two aspects: speed and detail. The receipt of data has now become near-instantaneous, greatly reducing the time needed to process, discuss, and iterate on datasets for analytical purposes.

We have also found that current and emerging data formats tend to have a higher degree of accuracy because of automated checks and reviews. That’s not to say issues regarding timing and reliability have been fully eliminated, but they’re steadily becoming an exception rather than an expectation.

The captives space in particular has benefited from the growing interest and reliance on data, due in part to the nature of many risks handled through captive insurance. Some of these risks, which can often be esoteric or highly specific, were traditionally viewed as difficult or impossible to quantify. Decisions on pricing and placement were subsequently based on what is generously described as “guesswork”. However, as an increasing proportion of high-level decision-makers have come to understand the benefits of data and analytics, more companies have focused on finding ways of capturing these risks in a consistent, quantifiable manner.

Risk management strategies

One impetus for this change can be found in the rise of enterprise risk management (ERM) techniques. The ERM philosophy of identifying and contextualising the full breadth of risks a company may face almost certainly brought emerging or complex risks to the forefront of C-suite discussions, and with them came the discovery of how difficult certain risks were to quantify. Examples of these include cyber events, product defects, errors and omissions, and weather-related revenue loss.

As the ERM process improved, so too did the methods of collecting and tracking data related to certain risks. Over time, we have seen ERM strategies grow into a collaborative approach in which data gathered from various departments and specialists is used to support more holistic strategies regarding a company’s risks. This data, which may not have previously been viewed as “loss” data in the traditional sense, is now being used to inform the pricing and placement of captive insurance policies.

However, companies which have identified the parameters of ERM-type risks may still not have access to unique, historical data, thus necessitating the use of industry benchmarks. The insurance industry is certainly no stranger to benchmarking information, but it’s the focus and availability of this data in particular that we’ve seen grow in the last couple of decades.

Legal databases, for example, offer useful loss data on specific types of suits and can be used to identify emerging risks faced by certain business sectors. These legal databases have existed for years, but the ability of their users to filter and sort with increasing complexity and speed constantly improves. They can even provide a qualitative understanding of how the risk landscape is impacted by legislative or regulatory changes.

Databases pertaining to more niche risks, such as cyber, D&O, and medical malpractice, are becoming more available. In a similar vein, weather and climate datasets (readily available to the public) are becoming popular tools for determining triggers in parametric policies. When used proactively, this breadth of industry-wide information can serve as an exceptionally useful tool for risk management and financing strategies.

Loss data analysis

Perhaps of equal importance to the changes we’ve covered thus far are the methods used to analyse loss data. These too have evolved and grown more robust, especially in recent years. From an actuarial perspective, traditional approaches are now being increasingly complemented with those which can more effectively leverage technological advancements, such as regression modelling, loss simulation, and even artificial intelligence.

While newer, cutting-edge approaches may not see as much demand for well-understood risks with plenty of historical data, they’ve become vital in the analysis of the emerging or complex risks placed frequently in the captives space.

Alongside these advancements is an increased demand for dynamic reporting. Decision-makers who are more poised to understand and use analytical information are beginning to seek analytics presented in a graphical, interactive format that allows them to explore the results on their own terms. This also means that the tools and skillsets which can facilitate such demands have become a much higher priority than they might have been in the mid-2000s.

Another feature that has seen a rise to prominence in risk analytics is that of iterative reporting, especially for captive insurance companies. The ability to quickly identify emerging or changing risks and pivot a captive’s structure accordingly can be of immense value in the modern risk landscape. Captive owners are more capable than ever of reviewing available policy options concurrently and adapting their captives to fit the parent company’s overall risk strategy in an efficient manner.

Conclusions

Although the focus of this reflective piece has been the evolution of data and analytics, we do not mean to imply that analytical professionals are the only sources of advancement in the world of captive insurance. We gladly serve as one piece in the collaborative process of insurance and risk management.

Regardless of the types of risks a company may face, a wide range of expertise must be utilised to derive a truly robust risk-based strategy. We’re more excited than ever to work alongside our captive insurance colleagues in fostering future innovations, and we hope that these combined efforts will continue to produce an environment that best serves our captive insurance clients.

L. Michelle Bradley is a consulting actuary at SIGMA Actuarial Consulting Group, Inc. She can be contacted at: mb@sigmaactuary.com

Enoch Starnes is a captive and complex risk consultant at SIGMA Actuarial Consulting Group, Inc. He can be contacted at: enoch@sigmaactuary.com