In actuarial reporting, it is tempting to think that the real work begins when the model results are produced. The spreadsheets are refreshed, the liability runs are completed, the capital numbers are calculated, and only then do we start to analyse.

My experience from a recent BMA year-end reporting cycle reminded me that this is the wrong order.

The real analytical work should begin before the final numbers arrive. Before opening the result file, an actuary should already have a reasonable expectation of what the result ought to look like.

Not a precise answer. Not a substitute for modelling. But a structured mental range: which direction should the balance sheet move, which components should dominate the movement, and which results would be surprising enough to require immediate challenge.

That habit, simple as it sounds, can make the difference between reviewing numbers and understanding them.

Seeing The Whole Picture Before The Result

Before the year-end BMA results were finalised, I spent time summarising the major changes that had taken place during the year. They were not small items.

The economic environment had changed. Discount rates moved because of changes in risk-free rates and illiquidity premium assumptions. Credit spreads, foreign exchange rates, and market levels all had potential implications for both assets and liabilities.

There were also operating assumption updates. Lapse assumptions, critical illness assumptions, renewal expenses and other experience-study conclusions needed to be understood not as isolated methodology changes, but as drivers of liability movement.

On the asset side, the investment strategy had also changed materially. A de-risking programme reduced equity exposure and increased bond holdings. That meant the year-end capital position should reflect not only market movement, but a deliberate change in the risk profile of the balance sheet.

There were also specific business events. A financial reinsurance arrangement was terminated during the year. A block of short-term endowment business matured. These items had direct implications for EBS surplus, required capital, liability balances, and the interpretation of year-on-year movement.

Taken separately, each item is just a note in a reconciliation. Taken together, they form an expectation.

Lower discount rates should generally put pressure on liability values. Assumption strengthening should move the BEL in a predictable direction. Equity de-risking should reduce equity risk capital. A reinsurance termination should have identifiable impacts across ceded liabilities, deposits, risk margin and surplus. Maturing short-duration savings products should reduce the liability base.

By the time the model output is available, an actuary should not be looking at a blank page. The result should be compared against a mental map that has already been built.

Expectations Are A Control

This is not only a matter of professional elegance. It is a practical control.

Models can run successfully and still produce wrong results. Workbooks can calculate without error messages and still be incomplete. A result can look neat, formatted and internally consistent while being based on a missed data update.

The most effective first-line control is often not a complicated validation tool. It is a reviewer who can say, calmly and quickly:

This does not look right.

That sentence is powerful only when it is grounded in expectation. Without expectation, every result becomes plausible. With expectation, unusual results have to earn the right to be accepted.

For example, consider a solvency stress test. If equity exposure is broadly unchanged, the equity stress level is the same, and the capital base is higher than before, then the percentage reduction in the solvency ratio should not suddenly become materially worse without a clear explanation. If it does, the first response should not be to rationalise the output. The first response should be to check whether the input data, balance sheet, capital base and scenario settings were updated completely.

That kind of challenge does not require aggression. It requires ownership of the full picture.

From Calculation To Judgement

Early in my career, I worked with a chief actuary who could look at a result for a few seconds and say, with a smile, “Chen, this can’t be right.”

At the time, I sometimes found it frustrating. I had done the calculations. I had followed the steps. I wanted the spreadsheet to speak for itself.

But again and again, he was right.

What I did not fully appreciate then was that he was not guessing. He had expectations. He knew the business, the products, the market environment, the model sensitivities, and the typical scale of movements. The spreadsheet was not the beginning of his thinking. It was a test of his thinking.

That is one of the transitions from technician to actuary.

A technician asks, “Did the model run?”

An actuary asks, “Does the result make sense?”

A stronger actuary asks, “What result should I have expected before I saw it, and why?”

Managing Expectations Is Also Communication

There is another side to this lesson. Expectations should not stay inside the actuary’s head.

For a year-end process involving finance, investment, audit and appointed actuary review, it is valuable to communicate the expected story early:

This helps auditors and senior reviewers understand the result as a coherent movement analysis rather than as a pile of tables.

It also reduces noise. When expectations are created early, the review process becomes more focused. People know where to look, what to challenge, and what explanations are needed.

In that sense, “create and manage expectations” is not only an analytical discipline. It is a reporting discipline.

What I Would Carry Forward

My main takeaway from this year-end cycle is simple:

Do not wait for the final EBS or solvency result before beginning the analysis.

Start with the full picture. Build expectations from economic conditions, investment actions, assumption changes, product events and management decisions. Then use the model result to confirm, refine or challenge those expectations.

When the result is consistent with the story, the explanation becomes clearer.

When the result contradicts the story, the contradiction becomes visible faster.

That is where actuarial judgement lives: not in accepting the output, and not in rejecting it instinctively, but in knowing what the output had to overcome before it could be trusted.

For me, this is one of the most useful lessons from the BMA year-end reporting process. Good actuarial work is not just about producing numbers. It is about having a view before the numbers arrive, and then being disciplined enough to test both the numbers and the view.

Note: This article is based on practical reporting experience. Company-specific figures, product names and confidential details have been intentionally omitted or generalised.