Happy pooled LDI fund managers are alike; every unhappy pooled LDI manager is unhappy in its own way. Except of course that last year there were almost no happy pooled LDI managers.
Following the FT’s coverage of investment advisers Barnett Waddingham and XPS Pensions’ decision to advise their clients to sell specific but unnamed pooled LDI managers, XPS has picked up the Tolstoyan baton to probe the nature of their individual unhappiness in a note to clients (albeit in mostly tabular form).
The investment adviser looks at the experience of six of the main pooled LDI managers during the gilt market mayhem. Annoyingly it’s anonymised, which is doubly annoying as pretty much everyone on the inside knows who each of the managers are. But — you know — litigation risk, relationships, blah blah blah.
Just to give you a sense of the market presence of fund managers in the LDI space, here’s an old chart from Investments & Pensions Europe breaking down the LDI market (it includes non-pooled LDI).
Let’s have a look at how the unhappiness manifested and what this actually means.
OK, so every pooled LDI manager “called capital at short notice”. Yes, larger buffers of excess collateral would’ve prevented this, but things were moving so fast in the gilt market that short notice capital calls were maybe to be expected.
What does this mean for pension schemes? Imagine a situation where the fund manager calls you and says ‘give me £10mn pretty much now if you want to maintain your hedge’. UK pension schemes unable to respond in time lost some or all of their hedges, potentially at the worst possible time.
Second up, “exposure reduction”. I read this as shorthand for ‘we participated in the doom loop’: the mini-budget triggers a bond market rout, requiring more collateral to be posted, pushing prices further down, rinse and repeat.
One of the triggers for that negative hit to prices is that if you’re a leveraged pooled LDI fund, have already posted all your unencumbered collateral, and can’t suck in new cash fast enough to post it as new collateral (ie, you “called capital at short notice” but not everyone can answer this call). Then you need to unwind your positions and reduce exposure: ie dump bonds, and fast.
XPS reckons that this happened to three of the six pooled LDI managers. For some clients this exposure loss was only temporary (still not great!), but for one it was permanent. One of the managers reduced exposure by an average of 70 per cent. That’s a lot.
Two managers suspended either “pricing or capital calls and distributions” during the melee. This would not have been great if, say, you’re trying to make payments to pensioners, process pension contributions or pump in money to maintain hedges. That sort of thing.
One manager applied a “pricing adjustment”. Just for one day mind. Sounds legit? If this was a minor thing (maybe widening the bid-offer or applying swing-pricing on a pooled fund) it would be too meh to feature on the list. Alternatively it could be a very big deal that raises a whole host of awkward questions.
Beyond this list, XPS also say that one of the managers had a fund accounting error that screwed up their pricing for a while. So pension scheme clients didn’t have accurate information on which to base decisions. Unhelpful.
So, what does this mean in pounds and pence for pension schemes? As always, (sorry) it depends.
XPS show the average reduction in hedges experienced by all XPS Investment’s trustee clients that used pooled LDI funds. It shows the impact on scheme hedging level from either the pooled fund manager taking a decision to reduce hedging (aka, caught in the doom loop — in blue), or the hedging lost when trustees were unable to make the super-short notice capital calls made on them (in pink).
Averages hide a litany of outcomes, and XPS state that there is material dispersion in the reduction of hedges across clients using any given manager’s pooled LDI funds. Although Manager 3 should be pretty pleased with themselves. (They are.)
Still, what does it mean to have lost a bit of hedging?
For a typical scheme with, say, a 70% target hedge, a reasonable rule of thumb is that a 10% reduction in hedging would have resulted in a 7% reduction in liabilities hedged, leading to a 1.5% impact on funding level. This reflects the 1% reduction in yields which was roughly the fall seen on 28 September following the BoE’s intervention. For context, the average reduction in hedge across XPS Investment’s pooled LDI fund client base during the gilts crisis was 5%, with 2% of this resulting from forced hedge reductions. This translates to a c. 4% reduction once the target hedge is accounted for.
So if a 7 per cent reduction in liabilities hedged equates to a 1.5 per cent fall in funding levels, a 4 per cent reduction in liabilities (if mirrored across the whole industry) would mean an average 0.9 per cent fall in funding level versus their counterfactual?
We’re told by the Bank of England that pooled LDI accounts for around £200bn — or roughly 10-15 per cent of overall LDI assets. That said, given their leverage, we can maybe infer they were used to implement perhaps a third of the system-wide hedge ratio? So let’s settle for a 0.3 per cent hit to funding levels stemming from pooled LDI.
That sounds pretty small, especially in the context of a large improvement in funding levels that accompanied higher yields. Using this number, and the PPF’s estimate of scheme liabilities*, the loss versus the counterfactual for pension schemes comes to around £3bn.
Smaller than the Bank’s £3.8bn gain on their intervention to stop the doom loop. But not nothing.
* Sure, if you want to get really geeky, the s179 valuation of liabilities will be lower than other measures that schemes use like Technical Provisions, Self-Sustainability, or Buy-Out. So maybe the right liability number is 10-40 per cent higher, depending on the measure.