Additional modelling to provide additional information to support the project's objectives

Welcome to the Next Steps: Adding Value Through Extra Modelling

Congratulations! You have built your base model, and it is working correctly. But in the world of actuarial work, a single "best estimate" result is rarely enough. Decision-makers need to know what happens if things don't go according to plan. This chapter focuses on additional modelling—the process of testing your results to provide deeper insights and support the project's objectives.

Think of it like this: If you were planning a long road trip, your "base model" tells you how much gas you need if traffic is clear and the weather is perfect. "Additional modelling" tells you what happens if you hit a traffic jam, if gas prices spike, or if you take a scenic detour. It makes your advice robust and useful.

Don't worry if this seems like extra work; it is actually where the real "actuarial magic" happens! Let’s dive in.

1. Why Do We Need Additional Modelling?

The primary goal of a CP2 project is to provide information that helps a client make a decision. A single set of results is a good start, but it doesn't tell the whole story. We perform additional modelling to:
1. Assess Risk: Identify how "risky" the base results are.
2. Identify Key Drivers: Find out which assumptions have the biggest impact on the final answer.
3. Improve Confidence: Show the client that we have considered various possibilities.
4. Support Decisions: Help the user understand the range of potential outcomes.

Quick Review: Additional modelling isn't just about "doing more math." It’s about answering the question: "How much can I trust these results if the future looks different from my assumptions?"

2. Sensitivity Analysis: Testing One Thing at a Time

Sensitivity Analysis is the process of changing one single input at a time to see how much the output changes. It helps you identify the "sensitivity" of your model to specific assumptions.

The Process:
• Pick one assumption (e.g., the interest rate \( i \)).
• Increase or decrease it by a small, reasonable amount (e.g., \( i + 1\% \) or \( i \times 1.1 \)).
• Record the new output.
• Compare it to the base result.

Analogy: The Salt in the Soup
Imagine you are cooking a soup. Sensitivity analysis is like adding one extra teaspoon of salt. If the soup becomes inedible, your recipe is highly sensitive to salt. If you barely notice the difference, the recipe is robust regarding salt levels.

Common Mistake to Avoid:
Don't just pick random numbers. Your sensitivities should be realistic and relevant to the project objectives. If you are modelling a 30-year pension plan, a 0.01% change in interest rates might be too small to be meaningful, while a 50% change might be unrealistic.

Key Takeaway: Sensitivity analysis helps you identify which individual variables are the most "dangerous" or influential in your model.

3. Scenario Analysis: Testing Events

While sensitivity analysis looks at one variable, Scenario Analysis looks at multiple variables changing at the same time in a way that makes sense together. Scenarios often represent specific "real-world events."

Example: An Economic Recession Scenario
In a recession, you wouldn't just see interest rates fall. You might also see:
• Lower investment returns.
• Higher unemployment (or higher lapse rates in insurance).
• Lower inflation.

A scenario analysis would change all of these variables simultaneously to see the combined effect on your project's objectives.

Did you know?
Scenarios are often given names like "The Optimistic Case," "The Worst-Case Scenario," or "The Stagflation Event." This helps non-actuaries understand the "story" behind the numbers.

4. Stress Testing: Pushing to the Limit

Stress Testing is a form of scenario analysis that focuses on extreme but plausible negative events. It is designed to see if the system (or the company) can survive a "worst-case" shock.

For CP2, you might be asked to check if a fund stays solvent under extreme market conditions. When doing this, keep your focus on the objective: Can the client still meet their goals even when things go very wrong?

Memory Aid: The Three S's
• Sensitivity: Single variable change.
• Scenario: Story-based multiple changes.
• Stress: Shock/Extreme changes.

5. Identifying Limitations and Further Improvements

Part of providing "additional information" is being honest about what your model cannot do. This is a crucial part of the "Next Steps" section in your CP2 summary.

When you finish your modelling, ask yourself:
1. What data was missing? Could better data lead to a better model later?
2. What simplifications did I make? (e.g., "I assumed a constant inflation rate, but in reality, it fluctuates").
3. What is the next step for the client? Should they perform a more detailed stochastic simulation?

Quick Tip for the Exam:
In your summary report, always suggest one or two areas for future modelling. This shows you understand the limitations of your current work and are thinking about the project's long-term objectives.

6. Communicating the Results of Additional Modelling

Once you have run your sensitivities and scenarios, you must explain them clearly. Use the following structure to keep it simple:
1. What did you change? (e.g., "Increased the discount rate by 1%").
2. What was the result? (e.g., "The Net Present Value decreased by \( \$10,000 \)").
3. What does it mean for the client? (e.g., "The project is highly sensitive to interest rate changes, so we should consider hedging this risk").

Example Sentence: "While the base model suggests the project is profitable, the sensitivity analysis shows that a 2% decrease in sales volume would result in a loss, suggesting the project has a narrow margin for error."

Key Takeaway: Results are useless without interpretation. Always link the numbers back to the client's original goal.

Summary of Key Points

• Additional modelling provides a range of outcomes rather than a single point estimate.
• Sensitivity analysis changes one variable to find the "key drivers" of the model.
• Scenario analysis changes multiple variables to reflect a specific event or "story."
• Stress testing looks at extreme, "worst-case" conditions.
• Always identify limitations and suggest further improvements to support the project's long-term objectives.
• Communication is key: Explain why the results changed and what the client should do about it.

Don't be intimidated by the extra steps! This section is your chance to show the examiners that you aren't just a spreadsheet builder—you are an actuary who understands risk and uncertainty.