Distributing the inherited estate of a closed with-profits fund

What the estate actually is

The starting point is a realistic fund balance sheet: assets, less asset shares, less the realistic cost of the guarantees and options, less the cost of running the fund to run-off (expenses, tax, investment management), less the capital the fund is required and chooses to hold. Only the excess over all of these is genuinely distributable. The £600m is the fund's own figure and should be independently re-derived.

Constraints to quantify before sizing anything

• Run-off capital. SCR and risk margin on a shrinking, closed fund, plus any internal target. Fixed costs per policy rise as the fund shrinks, so the expense overrun is a real first claim on the estate.
• Guarantee costs. A stochastic projection of the GCA and any minimum annuity or surrender guarantees at a defensible confidence level.
• The investment strategy is part of the calculation. The estate earned 4.1% against a 3.0% crediting rate. De-risking into credit and gilts lowers the expected return, so a distribution sized on the old asset mix may leave the estate short later. Run the model on the post-de-risking mix.
• Crediting policy. If the estate is being built by systematically under-crediting asset shares, part of it arguably already belongs to policyholders, and that is a fairness question as much as a solvency one.
• Tax, unclaimed assets, and the 10% shareholder share.

When a distribution is unfair

• Policyholders' reasonable expectations. What past communications, the PPFM and past practice led people to believe. A distribution that contradicts them is unfair even if it is actuarially defensible.
• GCA versus non-GCA policyholders. GCA holders paid for their guarantee through lower asset shares or higher charges. A distribution delivered purely as an asset-share uplift gives them nothing extra while removing the fund backing their guarantee; they are effectively cross-subsidising the rest.
• Generations. Long-standing policyholders whose charges and under-crediting built the estate versus policies taken out shortly before closure, which contributed little but would share in it.
• Exited policyholders. Those who surrendered may already have subsidised the fund through surrender value reductions.
• The test is comparative. Each group should be no worse off than under a defensible alternative, and the reasoning must be written down and defensible to the with-profits committee, the board and the PRA.

Recommendation: a capped special distribution plus a targeted buy-out - not one or the other

• Retaining the whole estate is hard to justify. The fund is closed, the estate exceeds the central run-off requirement, and retention merely defers the same fairness question while exposing the surplus to further investment risk with no new business to diversify it.
• The £250m special distribution is affordable on the numbers - roughly 42% of the estate, and with cover at 158% it should leave headroom above the 140% floor - but only after the run-off model is re-run on the post-de-risking asset mix and the guarantee cost is re-stressed. Size it from the model, not from the board's headline figure.
• The £150m buy-out of the 35,000 smallest policies should be justified on cost grounds, not as a distribution. Removing small policies cuts per-policy fixed costs, shrinks the expense overrun and reduces the number of policyholders requiring communication. The buy-out provider will price the guarantees, which also crystallises their cost - that is a benefit, not a drawback.
• Doing both reduces future estate strain and delivers something to policyholders now, which is the combination most likely to satisfy the fairness test.

Governance and communication I would insist on

1. Re-derive the estate, the guarantee cost and the distributable amount on one consistent basis, with an independent actuarial opinion on the size and on fairness between groups.
2. Review and reissue the PPFM before announcing anything; the smoothing policy is stale and must be reset prospectively so the distribution is not read as a smoothing change.
3. Formal consultation with the with-profits committee, including a written record of how GCA holders were protected.
4. Test against Consumer Duty: evidence that the outcome is fair value for each group of policyholders, not just in aggregate.
5. Engage the PRA early on methodology, not on the announcement.
6. Give policyholders clear advance notice with worked examples of the effect on their own policy, and a route to challenge.
7. Check whether the distribution mechanism needs a court-sanctioned scheme or formal notification, and whether the buy-out triggers a Part VII transfer timetable.
8. Publish the basis of the distribution so it can be tested later - the second distribution is always harder than the first.