By Alex White, Head of ALM Research, Redington
Firstly, what is CDI? There is not one set definition, but for our purposes we consider CDI to be the philosophical approach of managing a pension scheme by trying to match the liability cashflows with secure bonds.
By matching the cashflows, you are then less concerned about the market price of the bonds- you only care that they don’t default.
There are a great many subtleties around the exact shape this takes, but this tends to mean buying investment grade credit as far out as you can get it (about the first ten years), overlaying this with RPI swaps, and using gilts and repo to cover the long-end. So, we can think of CDI as a plan to match cashflows, being aware that this generally results in a strategy composed of credit plus LDI.
For well-funded schemes, a portfolio of investment grade (IG) credit plus LDI is a simple, reasonably secure, and eminently viable solution that boasts several supplementary benefits. For example, credit-based portfolios will typically provide a better match to IAS19 pricing. While this may seem more of a sponsor benefit, there are significant benefits to the scheme in promoting the sponsor’s health, as even when a scheme doesn’t need contributions, sponsor insolvency means either paying the PPF the higher SWoSS levy or being forced into cutting benefits via some combination of the PPF and partial buyout. But is being a good match to the cashflows one of them?
The main issue is a philosophical one- does it make sense to try to match cashflows rather than PVs?
Matching cashflows is somewhat irrelevant if you can afford to hold large cash reserves. This is particularly a problem in today’s relatively low credit spread environment as buying credit means you need to invest more money upfront. For context, at current spreads a strategy targeting gilts + 50 would require a c50-60% investment in IG credit. The same return could be achieved with a c15% allocation to equity (assuming returns of gilts + 3-4%). The equity “bar-bell”, while not matching the cashflows, could easily free up around 30% of the assets as cash (allowing a generous margin to cover the extra hedging). This could be used to cover many years’ cashflows.
The cashflows are also uncertain, due to changes in longevity, inflation, and transfers out, while other cash requirements may be larger than pension cashflows. As a minimum, the scheme will still need inflation protection in the early years (where the cashflows are backed by credit), which means holding collateral. Say you hold enough to cover a 1% move in RPI. For a 70% inflation-linked scheme, needing 10 years’ inflation protection, this equates to approximately a 1% move x 70% exposure x (on average) 5y duration = 3.5%. A typical well-funded scheme may be paying 2-4% of assets/year in pension cashflows- so the lowest collateral requirement is likely to be as large as the cashflow strain. In practice, we would expect leverage in the long-end LDI, meaning the collateral requirement was far higher. Moreover, the availability of UK credit is limited. Larger schemes will typically have to use US credit to get a meaningful pick-up. This incurs FX and basis hedging, which makes it harder to argue the cashflows are matched as well as requiring collateral.
As long as it does not incur sacrificing the hedge, designing a strategy with a CDI mindset will probably result in a simple allocation that is unlikely to go horribly wrong. Moreover, this is not an argument against simple credit-based strategies; a credit-heavy strategy may be right for your scheme whether you reached that conclusion by trying to match cashflows or not. However, while the story of matching cashflows can be seductive, it may not always be the right approach to strategy selection.
|