By Alex White, Head of ALM Research, Redington
Schemes who de-risked may have ‘locked in’ a portion of the strong asset performance seen over the last few years and avoided some of this year’s pain. For all schemes though, reducing the absolute size of liabilities lessens the size of any contribution strain or risk on the sponsor. This may make buyout more attainable for schemes whose sponsors are prepared to pay up to a fixed pound amount to get a buyout done.
For hedged schemes, the effect of the change in rates on assets would broadly offset the change in the liabilities, but this is in present value (PV) terms. In cashflow terms, it can differ. Firstly though, it’s worth remembering why schemes hedge. Broadly, if rates fall, the expected returns on assets, in absolute terms, go down. That means a scheme needs more assets to pay its liabilities (fundamentally, a liability PV is the asset value needed to pay cashflows if assets earn equivalent to the discount rate per annum). If rates rise, fewer assets are needed, and the scheme can sell growth assets accordingly.
So, if rates rise, the asset value decreases. And due to the size of rate moves over the last six months, assets and liabilities have fallen materially – many will be roughly 20% smaller. This is not a problem for a scheme with plenty of collateral and liquidity. However, for a scheme running tight collateral buffers, this can lead to a requirement to deleverage the LDI portfolio.
We looked at two roughly equivalent hypothetical schemes to put some colour around this.
Scheme A is a cashflow-matching scheme with 75% buy and hold UK corporate bonds (all stocks index) and 25% LDI.
Scheme B is approximately 80% LDI, 20% equity. Both start 100% funded and 100% hedged. The expected returns (in our model) are aligned as at December, and, while equities fell further than corporate bonds, Scheme B has lower exposure to growth assets, so both portfolios fall by around 27%. This reduces both funding levels to c.95%.
In collateral terms though, it’s quite different. With no rebalancing, the more leveraged LDI holding in Scheme A has been proportionately more affected (even accounting for the fact that the net PV01 is lower as the corporate bonds offer some hedging). The LDI holding would have fallen to approximately 14% of the remaining assets, giving a buffer only large enough to cover an interest rate move of c.130bps instead of c.250bps. Scheme A may well have to rebalance, selling bonds when spreads are wide.
Scheme B, however, has no such problem. The relative strategic asset allocation would still have shifted (from 81:19 to 79:21 in our example), but it still holds a considerable cash buffer – enough to cover a move in rates of well over 14 percentage points – and so would not be forced to sell assets. Ironically, the hypothetical scheme with a cashflow-matching strategy aimed at avoiding cashflow-related difficulties now has a cashflow problem, while the cashflow-agnostic scheme doesn’t.
This is a theoretical issue I’ve discussed before ; the difference is that now we’ve seen it happen. It’s also second-order in that performance would have been similar. CDI strategies are viable, sensible and appropriate in many cases, and corporate bonds have a natural pull-to-par effect through time. The two schemes here are also fictional, and many cashflow-matched schemes will have kept larger buffers, or mitigated these issues earlier in smaller increments. Nonetheless, the point remains: the cashflows from liability payments are much smaller than those that can be needed to rebalance an LDI allocation after a large move in rates. This means that holding lots of cash and bar-belling your portfolio can be a more effective cash management approach than precisely matching cashflow timings with bonds.
|
