Principles of Best-Estimate Actuarial Assumptions

What is a "Best Estimate" Assumption?

At its core, a best-estimate assumption is an attempt to predict the future without any intentional bias, conservatism, or optimism. It represents the expected value, or the average outcome, if we could run the future thousands of times. This concept is so fundamental that it is embedded in major international financial reporting standards.

To formalize this, let's turn to professional guidance.

As the guidance states, the "best estimate" is generally considered to be the mean outcome across a range of potential future outcomes. In some cases, such as when dealing with a highly skewed distribution (like large insurance claims), the median might be a more appropriate measure of the central tendency. The key takeaway is that the best estimate is not a guess; it's a statistically grounded central estimate.

A Framework for Assumption Setting

Setting and maintaining assumptions is not a one-off task but a continuous, cyclical process. The American Academy of Actuaries provides a useful high-level framework that illustrates the lifecycle of an assumption. While this resource was developed for a specific US regulatory context (Principle-Based Reserving), the process it outlines represents a robust and logical approach applicable to any area of actuarial work.

The eight steps in the cycle are:

1. Identify Assumptions: Determine which assumptions are needed for the model.
2. Select Timing: Decide on the frequency and priority for reviewing each assumption.
3. Analyze Experience: Gather and analyze data to form a view of the future.
4. Determine Margins: Add layers of prudence for specific purposes (we will cover this in the next lesson).
5. Review Reasonableness: Check that the assumptions make sense individually and collectively.
6. Document Results: Record the rationale, data, and methods used.
7. Implement Decisions: Load the approved assumptions into the models.
8. Monitor Experience: Track actual outcomes against the assumptions, which feeds back into step 1.

This lesson focuses on the principles that govern step 3, Analyze Experience, to arrive at a best estimate. The other steps, particularly determining margins and monitoring experience, will be the subject of subsequent lessons in this module.

The Core Principles of Setting Best-Estimate Assumptions

When an actuary analyzes experience to form a best-estimate assumption, they are not simply extrapolating the past. They are guided by a set of professional principles that ensure the assumptions are robust, defensible, and fit for purpose.

Let's return to the Actuarial Guidance Note to explore these fundamental principles.

Let's synthesize these principles:

1. Appropriateness to Purpose: An assumption must be suitable for the specific task at hand. Assumptions for pricing a new product might differ from those used for valuing long-term liabilities for regulatory reporting. The context dictates the level of detail and the specific factors to consider.

2. Consistency: Assumptions must be consistent with each other. For example, the assumed rate of salary growth for a pension scheme should be logically related to the assumed rate of general price inflation. Similarly, assumptions should be consistent over time and across different but related lines of business, unless there is a clear and documented reason for them to differ.

3. Grounded in Experience, Adjusted for the Future: This is perhaps the most important principle.

   • Start with Data: The process always begins with historical data. The best source is the company's own experience, as it is most relevant. If internal data is insufficient or not credible, actuaries may look to industry studies, reinsurer data, or population statistics.
   • Assess Credibility: As a statistician, you know that the "law of large numbers" is key. If a company has only sold a handful of policies, its claims experience is not statistically credible. Actuaries must assess the credibility of the data and may need to blend company-specific data with broader industry data.
   • Look Forward: This is where actuarial judgment is paramount. History is only a guide. The actuary must consider all available information to form a view of how the future might differ from the past. This includes changes in:
     • Company strategy (e.g., new underwriting rules, a different sales approach).
     • The external environment (e.g., medical advances affecting mortality, economic changes affecting policyholder behavior).
     • Product features or the mix of business being sold.

4. Regular Review and Monitoring: Assumptions are not static. The world changes, and assumptions must be periodically reviewed and updated to ensure they remain appropriate. This is the "feedback loop" in the assumption-setting cycle.

Putting Principles into Practice: Drivers of Key Assumptions

To make these principles more concrete, let's briefly consider the major categories of assumptions actuaries set and what drives them. Understanding the drivers is the first step in applying the principles of analysis.

We can broadly classify assumptions into two groups:

• Economic Assumptions: These relate to the financial markets and the broader economy. They include things like future interest rates, inflation rates, and investment returns on assets like equities and bonds. Setting these assumptions involves analyzing long-term historical data, current market conditions (like yield curves), and the views of economic experts.

• Non-Economic Assumptions: These relate to the demographic and behavioral aspects of the business.

  • Mortality and Morbidity: The rates of death and sickness. These are based on population data, industry tables, and—most importantly—the company's own experience, adjusted for factors like underwriting standards and expected future trends (e.g., mortality improvement).
  • Policyholder Behavior: This is a complex area covering lapse and surrender rates, premium payment patterns, and the utilization of options within a contract. The factors influencing these decisions are multifaceted, involving financial incentives, product understanding, and personal life events. Your background in social research provides a valuable perspective on the human element that underlies these assumptions.
  • Expenses: The costs of running the business, including acquisition costs (commissions, marketing) and maintenance costs (administration, claims processing). These are derived from a company's own accounting data and budgets, with an allowance for future expense inflation.

For each of these, the actuary must apply the core principles: analyze relevant historical data, consider if the past is a good guide for the future, make justifiable adjustments, and ensure internal consistency.

Conclusion

This lesson has laid the groundwork for our entire module on assumption setting. We have established that a "best estimate" is the actuary's unbiased, central estimate of a future outcome, typically the statistical mean. We've seen that setting these assumptions is not an arbitrary exercise but a structured, cyclical process governed by a core set of principles: appropriateness, consistency, a forward-looking analysis of experience, and regular review.

This concept of a "best estimate" is the neutral starting point. However, in many situations, actuaries are required to be deliberately prudent or conservative. In our next lesson, we will explore why we might want to deviate from the best estimate. We will introduce the concepts of prudent assumptions and margins for adverse deviation, and discuss their critical role in ensuring a company's financial soundness.

Key Takeaways & Cheatsheet

Here is a summary framework of the principles for setting best-estimate assumptions.