Interested in buyout? Strategies for estimating the sensitivity of liabilities to interest rates

The PPF estimates that on a full-buyout basis, the net funding position of defined benefit (DB) schemes has  improved from 79.2% in March 2022 to 111.9% in March 2023, with 67% of schemes fully funded as at September 2023[1].

Schemes have a variety of endgame options, but many are ultimately expected to seek a buyout. While they prepare for a such a transaction, DB scheme trustees may wish to hedge their buyout liability basis as they seek to preserve their buyout funding level.

Buyout price interest rate sensitivity

The key economic drivers of buyout pricing are interest rates, inflation, credit spreads and swap spreads. In this blog we focus on the interest rate sensitivity of buyout pricing and in particular, how schemes can approach hedging ‘PV01’. PV01 is a measure that (in this case) estimates how buyout pricing might change for a 0.01% change in interest rates.

Hedging on a buyout basis is more challenging than hedging on a gilts basis. One reason for this is that although consultancies provide broad indications, buyout pricing is only known after obtaining a formal quote from an insurer. Another is that the buyout pricing mechanism is complex and differs by insurer.

Consequently, it’s impossible to precisely hedge interest rate risk on a buyout basis. However, it’s important to consider what the best approximation might be, particularly for schemes targeting buyout in the next few years.

Breaking down pricing

A pragmatic starting point is to consider the key constituents of buyout pricing. A high-level illustration is as follows:

From the insurer’s perspective, they have capital requirements consisting of elements known as the Risk Margin and Solvency Capital Requirement (focusing on a 1-in-200-year event over the next 12 months), but they are likely to write business at a more attractive level compared with this to provide a competitive price. The risk buffer captures risks that are difficult to hedge fully (credit risk), or have a material cost to hedge (e.g. longevity, inflation exposure with caps/floors).

The natural starting point would be to consider the PV01 exposure of the blue element in the diagram above, e.g. the interest rate exposure when discounting the cashflows by the assumed yield on the assets matching these liabilities (we assume a yield of gilts +0.75% p.a. for illustrative purposes). However, this poses a question around what interest rate exposure is for the risk buffer and other elements, and what allowance should be made for this.

Discounting cashflows at gilts +0.75% p.a. may give a present value of £90m, but the buyout premium might be £100m, for example. Often, this is converted to an implied buyout yield, e.g. solving the discount rate applied to the cashflows to give a present value of £100m might result in gilts +0.0% per annum in this example.

This suggests three potential options (without changing the cashflows) for setting a liability hedging target on a buyout basis:

1. A ‘Best Estimate Liabilities’ discount rate (‘BEL’): Hedge 100% of the cashflows based on a gilts +0.75% p.a. discount rate
2. A ‘Best Estimate Liabilities’ discount rate scaled to the price (‘Scaled BEL’): As above, but apply a scaling factor to reflect the higher buyout premium value (scale by £100m/£90m in the illustration above)
3. Am implied buyout price discount rate (‘Buyout discount rate’): Hedge 100% of the cashflows on gilts +0.0% p.a. discount rate (the implied buyout yield)

The question is: which option might be best?

Differences between the options

For Option 1 (BEL), a key question to ask is if the £90m increases in value, would you expect the £10m to increase too? Broadly, yes. Risk buffers increase directionally with the size of liabilities even if not all the smaller elements do. Given the risk buffer would usually be much larger than fixed elements, this points more towards Option 2 (Scaled BEL) or 3 (Buyout discount rate).

As for Option 2 versus Option 3, it might not be immediately obvious why they differ! You might expect the implied level of liability hedging to be the same given they both target £100m. However, this isn’t quite the case as the duration of the liabilities is generally greater under Option 3 due to the lower discount rate (a convexity effect).

It depends on exactly what the cashflow profile looks like but for the pretty typical example we studied, the implied level of interest rate hedging would be 6% higher under Option 3. This difference would typically be larger the greater the gap between the assumed yield on the insurer portfolio versus the implied buyout yield (and vice versa).

Option 2 or Option 3?

This question ultimately boils down to if we expect the risk buffer to have a similar or higher duration than that of the underlying cashflows. Option 2 could be best if the risk buffer duration is similar, and Option 3 (or something in between) potentially better if the duration is longer.

The risk buffer may be expected to have a greater duration than that of the underlying liabilities given that two of the key risks captured by the risk buffer (credit and longevity risk) both broadly scale with time horizon. For example, all else equal, a credit downgrade has roughly double the impact for a cashflow that is twice as far into the future.

That said, this is extremely complex. In the example above, Option 3 implies the risk buffer would have a duration of 21 years. The likely answer is something between the two, but this will be contingent on a number of considerations. Watch this space for a future blog!

Target time

Taking a step back from the detail it’s important to reiterate that it’s impossible to hedge interest rate risk precisely on a buyout basis. However, we still believe a target hedging level should be set, and there is significant potential merit in considering the potential approaches and acknowledging the differences between them.

[1] Source: The Pensions Regulator