What the demonstration plan looks like
A plan of representative size and structure, valued at 31 December 2025 on a projected unit credit basis.
The twelve elements, and where each is delivered
The Survey Report Guidelines specify twelve elements. Each has a section of its own in the specimen, in the order a reviewer expects to meet them.
| Section | What it contains in the specimen | |
|---|---|---|
| 1 | Report highlights | Purpose, scope, the five findings that matter, and comparison against the prior review |
| 2 | Certifying actuary and qualifications | The named actuary, the basis of qualification, and a table dividing responsibility between firm and actuary |
| 3 | Basis of valuation | Data relied on, what was verified against what was relied on as supplied, member reconciliation, plan provisions |
| 4 | Actuarial assumptions | Each assumption with adopted value, derivation, published source, and the alternatives rejected |
| 5 | Actuarial methods | Projected unit credit, the two valuations kept separate, treatment of the statutory guarantee, and the engine controls |
| 6 | Valuation results | Obligation by member class, vested split, cost components, benefit cash-flow projection, movement reconciliation |
| 7 | Assets and investment performance | Asset basis stated with its limits, and the contract guarantee tested against the statutory minimum year by year |
| 8 | Financial conformity review | Every ledger line re-derived from plan terms, exceptions quantified and aged, remedies, and what conformed |
| 9 | Sensitivity analysis | Tornado on the approved basis, which assumptions carry the risk, and the combination worth budgeting against |
| 10 | Belgian legal compliance | Provision-by-provision test, the three 2015 Act changes taken separately, anticipated changes quantified |
| 11 | Conclusions and sustainability | Surplus or deficit on each basis, threats to solvency, and six specific measures |
| 12 | Actuarial certification | The signed opinion, stating exactly what the actuary stands behind and what he does not |
One engine, three plans
The specimen above reports on D-B, because a Survey Report concerns one plan. But a model that only works on the plan it was built for is not a model. These three were valued on the same code, the same assumption framework and the same valuation date. They differ the way real plans differ: in size, in maturity, and in whether they are still open.
- Members
- 167
- Obligation
- EUR 1,896,733
- Duration
- 22.2 years
- Discount rate read
- 4.61%
- If rates fall 100bp
- +25.4%
Open to new entrants, few retirees, long horizon.
- Members
- 612
- Obligation
- EUR 15,408,210
- Duration
- 15.0 years
- Discount rate read
- 4.45%
- If rates fall 100bp
- +16.8%
Open, three decades of history, a settled membership.
- Members
- 902
- Obligation
- EUR 36,382,954
- Duration
- 8.0 years
- Discount rate read
- 4.28%
- If rates fall 100bp
- +8.7%
Closed to new entrants, pensioner-heavy, short horizon.
Each plan reads its own discount rate
This is the part worth pausing on. Nobody chose the three rates below. The engine measured each plan's own liability duration first, then read the published euro curve at that point.
A young open plan is paying benefits decades away, so it sits far out on the curve at 4.61%. A closed plan mostly in payment sits near the front of it at 4.28%. Applying one convenient index rate to all three would misstate two of them, and the error compounds through every figure that follows.
It also explains the last row of each card. A 100 basis point fall moves D-A by +25.4% and D-C by only +8.7%. That is duration doing what duration does, and it means the same review produces genuinely different advice for each employer.
Why we published it before award rather than after
You can judge the deliverable, not the promise
Every quotation asserts that its report will be thorough and auditable. There is no way to test that assertion from a proposal. There is a way to test it from a report. The specimen is here so the assessment can be made on evidence.
It shows the analysis, not the template
The specimen carries five real conformity exceptions, a year-by-year comparison of the contract guarantee against the statutory minimum, and six specific remedial measures. It shows what our findings look like when we have something to report, which is the only useful demonstration.
Nothing in it is typed in by hand
The numbers come from the same projected unit credit engine that runs live in the Valuation Lab. Change an assumption there and the results move exactly as they would in the report. On award, the Embassy's census replaces the demonstration data and the machinery runs unchanged.
One finding the specimen demonstrates, because it is the one that matters here
Belgian law imposes a statutory minimum return on employer and employee contributions, and that guarantee binds the employer, not the insurer. It moved from 1.75% to 2.50% on 1 January 2025, its first change in nine years, and it applies by contribution generation rather than as one flat rate across the fund.
In the specimen, the post-2024 contribution generation was credited at the superseded 1.75% floor. The engine finds it, quantifies it at EUR 38,400, ages it, and pairs it with a remedy. It also finds that the insurer's contractual guarantee has sat below the statutory minimum since 2019, which is an exposure that never appears on an insurer statement because it is not an insurer obligation.
That is the class of finding this review exists to produce, and it is why the guarantee is tracked generation by generation rather than as an average.