Capital modeling: a conversation about risk, capital and better decision-making

Capital modeling is often viewed as a highly technical actuarial exercise. In reality, it can become one of the most valuable strategic tools available to insurers and captive owners. Frank Huang, CEO of Insurance Solutions North America, sat down with Michael Atkinson, Principal, Davies Actuarial Consulting, to discuss what capital modeling really is, why organizations often overlook it and how it can help businesses make more informed decisions.

What exactly is capital modeling?

Frank: Let’s start with an easy one. What is capital modeling? And no fancy actuarial terms.

Michael: Challenge accepted.

At its simplest, capital modeling is the process of analyzing all of an organization’s risks together to determine how much capital it needs to hold.

Frank: That’s it? Just quantify the risk so you know how much capital to hold? Sounds pretty easy.

Michael: That’s probably too simple.

Think about a captive or insurance company. It’s exposed to a range of uncertain risks. Some are forward-looking, like underwriting risk. Others look backwards, like reserve risk. Each carries its own level of volatility.

Capital modeling brings those risks together into a single, holistic view. Rather than looking at each one in isolation, it analyzes how they interact and then produces a range of possible outcomes. From there, the organization can determine an appropriate level of capital based on its own appetite for risk.

One of the biggest advantages is diversification. Risks don’t all move together, so when they’re modelled collectively, the overall risk profile is typically lower than simply adding up each individual component. The extent of that benefit depends on how the risks interact and the assumptions used in the model, but analyzing them together definitely provides a more realistic view than assessing each risk in isolation.

Frank: That sounds really interesting, even if I’m not an actuary.

So the benefit comes from looking at everything together, rather than treating each risk separately?

Michael: Exactly.

If you simply added every individual risk together, you’d effectively be assuming they’re all perfectly correlated—that every bad outcome happens at exactly the same time. That’s an extremely conservative assumption and, in reality, rarely reflects how risks behave.

The model captures those interactions, giving organizations a much more realistic picture of their capital requirements.

Why isn’t every company doing it?

Frank: If it’s that valuable, why isn’t every company already using capital modeling?

Michael: I wouldn’t necessarily say companies actively avoid it. More often, it just isn’t at the top of the priority list.

Most captive owners naturally focus on premiums, claims and financial performance. Those are important measures, but they don’t necessarily answer another critical question: How much risk are we actually comfortable taking?

Many organizations haven’t formally defined their risk appetite.

What makes it particularly interesting is that different stakeholders often have very different perspectives.

A risk manager responsible for the day-to-day running of the captive may naturally favor a very conservative capital position. Treasury, on the other hand, may prefer to minimize capital held within the captive so those funds can be deployed elsewhere in the business.

Neither viewpoint is necessarily wrong. They’re simply optimizing for different objectives.

Frank: So the model almost becomes the starting point for that conversation?

Michael: That’s exactly what happens.

There are always competing views on what constitutes “enough” capital and where that capital is best allocated.

A good capital model provides an objective foundation for those discussions. It helps everyone understand the trade-offs, agree what level of risk is acceptable and make decisions based on evidence rather than assumptions.

Ultimately, it’s about reaching consensus before difficult situations arise—not during them.

More than just a number

Frank: Is there much difference in how actuaries present this work? Can one capital model make it easier for a business to reach consensus than another?

Michael: Absolutely.

The technical modeling is only part of the engagement. Communicating the results is just as important.

Very few organizations begin the process with complete agreement on their risk appetite. In fact, many haven’t really discussed it in any structured way before.

Seeing the results often changes the conversation.

Once stakeholders understand what’s actually at risk financially—and how that changes as their tolerance for risk changes—they’re in a much better position to agree what’s appropriate for their organization.

Sometimes we’ll also use a formal risk appetite survey to understand how different stakeholders view risk before we even begin discussing the results.

That often highlights where opinions align and where they don’t.

Frank: So the model isn’t telling people what they should do. It’s helping them make better-informed decisions together.

Michael: Exactly.

The model provides the analytics. The business provides the judgement.

Beyond capital: a strategic decision-making tool

Frank: Does this ever influence bigger strategic decisions? Things like growing into new lines of business or changing the way the captive operates?

Michael: Quite often—and that’s one of the biggest reasons organizations commission a capital model in the first place.

Sometimes the commercial insurance market changes and companies find themselves retaining more risk than they’d originally planned.

Other times, organizations deliberately want to expand their captive, add new lines of business or increase retentions.

In every case, the question becomes the same:

“Can we do this within our current risk appetite?”

That’s where capital modeling becomes incredibly valuable.

The first step is establishing a baseline—a clear picture of the risks the captive is already carrying.

From there, organizations can test different scenarios and understand the impact before making strategic decisions.

For example, the model might reveal that although the overall capital position is comfortable, the captive has become disproportionately exposed to catastrophe risk. That insight alone could prompt changes to retentions, limits or reinsurance structures before problems emerge.

Equally, if the analysis demonstrates the captive is significantly overcapitalized, it may support returning surplus capital to the parent company or adopting a different investment strategy that generates higher returns.

Frank: So this is much more than answering, “Do I have enough capital?”

Michael: Much more.

Once the model exists, clients use it to answer all sorts of strategic questions.

• Should we retain more risk or transfer it? For example, a captive may be evaluating whether to increase its workers compensation retention from $500,000 to $1 million per claim, or whether a quota share, excess-of-loss or stop-loss reinsurance structure provides the most efficient balance between capital requirements and insurance cost.
• Are our retention levels still appropriate? Modeling can help identify an optimal retention level by comparing the cost of holding additional capital with the cost of transferring that risk through insurance or reinsurance.
• Are we buying the most effective reinsurance structure? Different structures – including XOL, QS and stop-loss – can be modeled to understand their impact on volatility, capital requirements and overall cost.
• Can we support expansion into another line of business? Adding risk doesn’t necessarily increase overall volatility proportionately. Modeling can show the impact on capital requirements and whether additional diversification benefits are created.
• Are we pricing risk appropriately? The analysis can support risk-based pricing by quantifying both expected losses and the capital required to support the risk, helping organizations rely less on broad market benchmarks and more on their own experience.
• How much diversification are we actually achieving? The model can identify where risks are highly correlated and where adding different risks may improve the captive’s overall diversification.
• How resilient would we be under different stress scenarios? Organizations can test the potential impact of events such as a major Cat loss, economic downturn or significant liability event on their capital position.
• Can we provide the Board with objective evidence to support major strategic decisions? Modeling gives stakeholders quantitative evidence to support discussion and help build alignment around risk and capital decisions.

That’s where capital modeling becomes much more than an actuarial exercise. It becomes a decision-making framework that supports the wider business.

How long does a capital modeling engagement take?

Frank: This sounds incredibly comprehensive. I’m guessing it’s also a pretty long engagement. How long does a typical project take?

Michael: The biggest factor is complexity.

The more risks we’re modeling, the more work is involved. That means understanding each underwriting risk, reserve risk and, where relevant, credit or asset risk. The underwriting and reserve work generally accounts for most of the effort.

Another important consideration is whether there’s an existing actuarial report we can build from. Ideally, the capital model should align with the assumptions underpinning the company’s financial reporting. If those analyses already exist, we can focus on modeling the volatility around those expected outcomes rather than recreating the underlying actuarial work.

For most organizations, I’d expect a typical engagement to take around eight to ten weeks, allowing a little flexibility for data quality issues or particularly complex lines of business such as Cyber. Of course, every organization is different, and timelines can vary depending on the number of risks being modelled, the availability of data and the complexity of the business.

Frank: So it’s not a six-month consulting project.

Michael: Not at all.

It’s a structured engagement with a clear objective. The real value comes from what the organization does with the model once it’s built.

Is this only relevant for captives?

Frank: We’ve talked a lot about captives, but could other organizations benefit from capital modeling as well?

Michael: Absolutely.

The principles apply equally to insurance companies. Larger carriers often have dedicated capital modeling teams, but many smaller insurers don’t have those in-house resources and can benefit from external expertise.

The same thinking can also be applied to organizations that self-insure, even if they don’t own a captive.

The questions may be slightly different, but they’re still asking things like:

• Are we retaining the right level of risk?
• Are we deploying capital efficiently?
• How resilient is our insurance program under different scenarios?

For self-insured organizations, the capital requirement itself may be less significant because the insurance program sits within the wider corporate balance sheet. Even so, modeling can provide valuable insight into risk exposure and complement a broader enterprise risk management program.

Frank: So really it’s less about the legal structure and more about helping organizations understand risk.

Michael: Exactly. The methodology is highly transferable. The decisions it informs may vary, but the underlying principles remain the same.

Can capital modeling ever point you in the wrong direction?

Frank: Slightly left-field question. Have you ever seen capital modeling lead a company to the wrong conclusion?

Michael: Personally, I’ve not seen a well-designed capital model lead a company to a poor decision because of the model itself. More often, challenges arise when assumptions are misunderstood, data quality is weak, or risks emerge that weren’t contemplated in the analysis.

That doesn’t mean the decisions are risk-free.

Take a retention analysis. A company may decide—based on robust analysis—that increasing its net retention is the right strategic move. If a major claim occurs shortly afterwards, that doesn’t necessarily mean the decision was wrong. It simply means the event happened.

Capital modeling is no different.

The purpose of the model is to help organizations make informed decisions based on the information available at the time—not to eliminate uncertainty altogether.

The real risk comes from events that sit outside the assumptions of the model.

Extreme events are difficult to model because they often fall outside historical experience or involve multiple risks occurring simultaneously.

The important thing is being clear about what the model is designed to answer—and what it isn’t.

Frank: So it’s about improving decision-making, not predicting the future.

Michael: Exactly.

No model can remove uncertainty. A good model helps organizations understand it, quantify it and prepare for it.

When should a company start thinking about capital modeling?

Frank: Last question. If someone is reading this and thinking, “Maybe we should look at capital modeling,” what are the signs that it’s the right time to have that conversation?

Michael: The good news is that organizations don’t need to have everything figured out before the first conversation. The initial discussion is typically focused on understanding the organization’s objectives, risk profile and available information, and identifying how capital modeling could support decision-making.

There are, however, a number of situations where capital modeling becomes particularly valuable.

For example:

• you’re experiencing rapid growth or increasing volatility in exposures;
• your mix of business is changing;
• loss experience is deteriorating;
• you’re reviewing retentions, limits or program structure;
• you’re considering adding new coverages or expanding your captive;
• you’re planning a dividend or reviewing surplus capital;
• you’re rethinking your investment strategy;
• or conditions in the commercial insurance market are prompting greater use of alternative risk solutions.

Those are all points where organizations benefit from having a clearer understanding of how strategic decisions affect both risk and capital.

Frank: Listening to all this, I came into the conversation thinking capital modeling was about one question: “How much capital do I need?”

It turns out that’s only the starting point.

Michael: Exactly.

A good capital model doesn’t just tell you how much capital to hold. It helps you understand why, what options you have, and how different decisions could change your risk profile over time.

That’s why I see it less as an actuarial exercise and more as a strategic business tool.

Capital modeling can provide much more than an answer to how much capital an organization should hold. By bringing different risks together and testing the impact of different decisions, it can give captive owners, insurers and other risk-bearing organizations a clearer view of their exposures – and a stronger evidence base for decisions around risk, capital and future strategy.