Superannuation guarantee and annuity buyout: valuing an in-the-money option

Value the option as the excess cost of buying the promised annuity at market rates.

• At 5.45%, ₹1 of fund buys an annuity of ₹0.0545. To deliver the promised ₹0.059 annuity at market rates requires 0.059 / 0.0545 = ₹1.0826 of fund. The uplift is 8.26% of the fund for each member who exercises.
• The near-retirement group holds 34% of ₹640 crore = ₹218 crore, so the option cost is about 8.26% × 218 = ₹18 crore if rates stay unchanged. Across every member retiring before 2031 the cost is higher, but longer-dated members may see rates move either way, so their option value is smaller and more uncertain.
• The option is not a fixed cost; it is a call on interest rates. A 100 basis point fall in annuity rates takes the market rate to about 4.45%, and the uplift to 32.6% of the fund, roughly ₹71 crore on the same group. That tail is the real risk.
• Take-up matters. If only a third commute and the rest buy annuities through the trust, the full 8.26% applies to the annuity portion; if many choose the maximum commutation, the cost falls proportionately. Check whether the rules allow the trust to vary the GAR.

Who bears it. The trust, and behind it the sponsor if the trust cannot meet the promise, because the GAR is a scheme rule, not an insurer obligation. Members bear it only if the trust reduces benefits, which would be a breach of the rules. The cost should be recognised as a liability in the scheme's actuarial valuation and funded, not left as a contingent promise.

The quotation is expensive, but so is running the risk.

• ₹700 crore against a ₹640 crore fund is a 9.4% shortfall, about ₹60 crore for the sponsor. Decompose it: a best-estimate cost of buying the annuities of roughly ₹655 crore, plus insurer expenses, a risk margin and a profit load of about 6 to 7%. The insurer must also hold capital against longevity and credit risk, and it is pricing an option that is already in the money.
• Self-annuitisation looks cheaper on paper: gilts at 7.05% paying annuities at 5.90% leaves a 1.15% spread for expenses and longevity. But the trust keeps the longevity risk for perhaps 30 years with only 2,100 members, and the near-retirement group is only about 150 lives. Mortality volatility on a pool that small is high, and a longevity reserve of 10 to 15% of the fund, ₹64 to ₹96 crore, may be needed. That is more than the buyout shortfall.
• The trust would also need investment governance, administration and actuarial valuation indefinitely, and the sponsor would still be exposed if the fund ran short. The buyout removes longevity, investment and sponsor risk in one transaction and gives members a contract with an insurer regulated for solvency.
• Accounting: a buyout normally produces a settlement gain or loss under Ind AS 19 and removes the liability from the balance sheet, which is what the sponsor wants. Self-annuitisation keeps it there.

Recommendation. Run a competitive process with at least three insurers and consider a phased approach: buy out the near-retirement cohort and those over 55 first, at a price close to the best-estimate cost, and retain the younger members, whose GAR is far out of the money and who can be funded by future contributions.

A three-part plan: reserve, match and monitor.

• Reserve. Recognise the GAR option as a liability in the scheme's valuation, using a market-consistent annuity rate and a take-up assumption, and fund it through a small increase in the sponsor contribution or a one-off contribution of about ₹20 crore. Do not rely on the 7.05% gilt yield continuing.
• Match. For members within five years of retirement, invest their share in long-dated government securities matched to the expected annuity purchase date, so the option cost is hedged against interest rate moves. For younger members, keep a growth-oriented portfolio but review the allocation annually.
• Trigger. Set a buyout funding-level trigger. When the fund reaches 100% of the buyout quotation, the trustees must take a decision within three months either to buy out or to document why they are retaining the risk. This prevents drift into a large uncovered position if rates fall further.
• Monitor longevity. Use the current Indian assured lives mortality table with an explicit improvement scale, review annually against the scheme's own experience, and disclose the sensitivity of the liability to a one-year improvement in life expectancy.
• Governance. Trustee training on funding and longevity, an investment policy statement with permitted ranges and a prohibition on illiquid assets, an annual actuarial valuation, and clear member communication that the GAR applies only to retirements before 2031 and is subject to the rules. Consider joining a multi-employer pension pool if one is available, because a single 2,100-member scheme cannot diversify longevity risk on its own.