PRA's Funded-Reinsurance Plan Contains, Does Not Align

Johan Landman, writing in Insurance Asset Risk, sets out the expected response to the PRA's funded-reinsurance consultation: the calibration is too stiff, the notching too harsh, the transition too compressed. The comment period ended on 31 July, and Landman doubts many letters argued the other way. That is predictable, because the firms with standing to respond are precisely those whose capital treatments move.

His more consequential reading is of CP8/26's own numbers. The paper repeats its aim: funded reinsurance should be treated more like economically similar assets. Landman argues the figures it publishes do not support that promise. They support limiting the advantage, not eliminating it.

Paragraph 3.5 contains the three numbers that frame the debate. The average funded reinsurance transaction already on the books carries capital equal to 2% to 4% of the annuity liabilities behind it. A direct holding of economically similar assets attracts 11% to 15%. Under the proposal the average lands near 10%, which the PRA describes as a material step toward fixing the inconsistency while granting that differences will persist.

Paragraph 4.11 says what the 10% comprises: valuation and capital charges, the counterparty default adjustment, the SCR and the risk margin, summed together. It is the full bill for holding the exposure, so the comparison with the direct range is straightforward. The plan leaves one to five percentage points of the gap in place.

The reason for the original gap is in the accounting, not in model assumptions. Set up £100 of annuity liabilities backed by credit assets. Held directly, the fundamental spread is taken out of the matching adjustment held in technical provisions, spread and downgrade risk draw an SCR charge, and the risk margin is layered above. If the same assets are ceded, the insurer's balance sheet shows a reinsurance recoverable. A counterparty default adjustment is applied — in today's regime that deduction is anything from under 0.5% up to roughly 2% — and counterparty SCR adds about 1% of the asset value, by the PRA's estimate. The credit risk is the same either way; the provision beneath the liabilities is roughly ten points lighter.

Landman describes containment as a legitimate policy choice. A regulator could sensibly conclude that full alignment carries more complexity than benefit, or that a partial narrowing is enough. CP8/26 does not present that reasoning. It stops at 10% even though the direct range starts at 11%.

The PRA is sorting through the comments and preparing a policy statement. That document will explain the final numbers. If it holds near 10%, funded reinsurance keeps a capital edge of one to five percentage points over direct investment. The gap is thinner than it was, but still there. The policy statement will be where the PRA says whether that residual is the point of the exercise or a compromise it is willing to live with.