Why peak LDI could be further off than we think

Stephanie Baxter
clock • 9 min read

Key points

At a glance

  • There are many risks to not hedging on the basis that LDI will peak in a few years
  • Lower pension scheme demand for gilts would not necessarily lead to higher yields
  • Some predict the peak could easily go out beyond 2021, even as far as 2033

Predictions that LDI flows could peak as soon as 2021 have led to hopes of higher gilt yields. However, Stephanie Baxter finds there are many variables at play.

Liability-driven investment (LDI) has become one of the biggest investment trends for defined benefit (DB) pension funds. Emerging around two decades ago as a specific strategy to hedge their liabilities, it has continued to grow year-on-year.

An additional £61bn of liabilities across 312 new mandates were hedged using LDI in 2017, according to XPS Pensions Group's 2018 survey - more than in any other year. All-in-all, 49% of the UK's private sector DB liabilities have now been hedged, based on a £1.9trn liability figure on a gilts plus 0.5% basis and using data from the Pension Protection Fund's 2017 Purple Book.

This poses the question of whether the market has almost reached 'peak LDI' - the last year of large flows into LDI before flows slow substantially. And if it has, what does that mean for pension schemes?

Research by Hymans Robertson and Nomura published in the spring argued that given the high level of hedging already in place, the current pace of increases in LDI hedging can only continue until 2021 - which could have big implications for gilt prices. The report suggested there could be lower pension scheme demand for gilts, which could in turn lead to higher yields.

Price dynamics

River & Mercantile Derivatives managing director Mark Davies says: "If you buy into the argument that pension schemes are the ones moving the long-dated interest rate market, if they stop buying LDI assets - whether derivative or physical - then you might expect an increase in rates. That creates an interesting price dynamic."

Though the report did not forecast a large move in global real yields over the medium term, it said there is room for "considerable cheapening" in the UK as a result of the transition to peak LDI or loosening of LDI philosophy.

There has been talk that this is bolstering the arguments of those financial directors and trustees who have not bought into LDI or put off increasing hedging levels on a belief that yields will rise soon.

KPMG head of LDI research Barry Jones says while it is a "really well-written paper with insightful analysis", he believes its conclusion - namely that the next three years will see a trend by UK DB pension schemes that causes a material sell off in gilts - is "maybe pushing too far", especially given some of the wider geo-political events over this timeline.

He also believes that the strength of the conclusion could sway trustees into action that isn't appropriate from an integrated risk management perspective. He said: "There are trustees that have held back on de-risking despite covenant uncertainty in the belief that yields must sell off at some point. This paper could keep those decision makers at the roulette table for another spin."

Jones explains: "LDI has always been an emotive subjective and, within trustee boards there are some very powerful figures with strong opinions - typically financial directors. I'm sure some are saying 'This has proved me right'."

There are many risks to not hedging on the basis that LDI will peak soon. Jones says: "You could argue the peak at 2021 is right, but it would be a sad moment for a scheme to not hedge on that basis alone and then real yields collapse further following a bad Brexit, for example. Especially given how little capital is required to remove this risk."

XPS Pensions Group chief investment officer Simeon Willis agrees: "The extent to which the report's suggestion that the age of peak LDI will arrive by 2020 has persuaded people to hold off hedging has probably been an undesirable result and unhelpful for schemes in question.

"Where it has led to informing people's understanding of the market's overall dynamics isn't a problem, but where it has persuaded people to take a different course of action has been very unhelpful."

The report said there is evidence of distortion in the UK relative to the US, suggesting this is due to supply/demand factors given the low prevalence of LDI in the US versus the UK.

Just because DB pension demand for gilts is high and will ultimately fall, does not necessarily relate to a buying opportunity. There are lots of unknowns in the gilt market, which is moved by many factors.

Willis does not think gilt prices are distorted. "The economics term of market distortion is where you have queues on the demand or supply side where people aren't able to interact at the price the market's prevailing because of some constraint introduced, such as rent caps. The argument the gilt market is distorted is simply not right."

Also, a lot of schemes will be buying out their liabilities, which will be above the gilts basis for funding. "As those insurers will need to access the LDI market to hedge, those liabilities still stay in the market," says Willis.

Peak beyond 2021?

It could also be argued that the age of peak LDI is still some way off.

"There are lots of reasons why schemes might continue to employ a greater level of LDI, and there's plenty of scope for it to rise from here," says Willis.

Davies found that by just tweaking some of the research's assumptions in a fairly reasonable way and looking at them in a slightly different way, peak LDI can be pushed out beyond 2021, to as far out as 2033.

While not suggesting the report's analysis itself is incorrect, he says by doing this, "you end up with something that's far less of a panic situation than if you look at the three-year peak LDI headline of the paper" (see box below).

BMO Global Asset Management head of LDI client portfolio management Simon Bentley agrees the peak could easily go out beyond 2021 just based on the volumes coming through.

"Most of the £1bn plus schemes already have a manager in place, with a reasonably high level of hedging at around 50%-75%. But there is a very long tail of medium and small-sized schemes that have either done very little or no hedging at all.

"We continue to see a lot of demand for our pooled funds from smaller schemes, so there's a lot of activity still to come from there. While it's not as exciting as a £1bn scheme coming through the door, they all add up to quite a decent amount. We see very few schemes with 90%-95% hedge ratios."

Another angle, says KPMG's Jones, is that any significant change in life expectations could be a natural source of additional demand for LDI.

Risks of holding off

There is a lot of risk still out there that if real yields fall, schemes will be further away from their long-term targets.

"From our view of the industry, very few schemes are at their long term desired risk / hedge level, and a lot still have significant hedging in order to get there," says Jones.

Jones also points out a number of recent high profile corporate failures where the underlying pension schemes' alternative approaches to de-risking had not offered the desired protection to members.

Willis agrees: "Pension schemes are all in the same boat, and are there for a reason, which is to manage their risk. It's no good holding off for when this isn't an issue for schemes because by that point you will have either benefitted or lost out as a result of taking that risk. But it will be too late to do anything about it."

He equates it to giving up smoking. He says: "You only get the benefit from when you actually do it, rather than having a plan to do it. In general, schemes have recognised that, and so their de-risking plans have focused more on time-based triggers rather than market-based triggers - as the latter never come and the situations where they don't come is when you most need them. When the market goes against you, you don't hedge and are then in the worst position. This by definition compounds the problem."


Why peak LDI could be 15 years away

Predicting the future is always challenging - not least trying to estimate when the age of peak LDI will come and how it would impact the gilt market. The Hymans Robertson/Nomura report acknowledges this, and explains their underlying assumptions.

Having made some tweaks to the assumptions, River & Mercantile Derivatives' Davies believes there are four key reasons why peak LDI can be pushed out further than three years away, to as far as 2033.

"Little changes make small differences but combined together you come up with something nearer a peak being 15 years away," he adds.

The report asked a "wide number" of LDI managers, several self-administered pension schemes, and two life insurers to better understand how much LDI has been done, estimating that £1.2bn of liabilities is hedged, which equates to 55% of gilt liabilities.

Davies argues the £1.2trn figure is too high to start with, and that it is closer to £1trn - which was estimated in the first XPS Pensions Group LDI survey (formerly run by KPMG). That alone could simply add on at least another couple of years to the peak point using the base assumptions, he calculates.

Secondly, he points out the historical rate of demand for LDI is lower than the £100bn per year discussed in the Hymans/Nomura paper. The past three KPMG surveys have estimated growth in LDI to be £32bn in 2015, £103bn in 2016, and £77bn in 2017; this gives an average closer to £70bn and could add on another two years to the date of peak LDI, says Davies.

Thirdly, he finds another five years can be added on by making another tweak to the assumptions. "£100bn of hedging per year is more than the £100bn worth of liabilities that are actually being hedging. This is due to a technicality in the way liabilities work in the background. Let's start with gilt-based liabilities of £2.3trn, and assume hedging 50% so £1.15trn. The value of those liabilities will fall to £2.2trn in a year's time. The report includes a graph of gilts-based liabilities - sloping line starting at £2.3trn and going down over time.

"£1.15trn of hedging today is actually going to turn out to be £1.1trn of hedging next year. If you start with £1.15trn of hedging now and add another £100bn of hedging, you will only end up with £1.25trn next year rather than £1.3trn. The effect of that slows the increase in the rate of hedging. That change in itself adds around five years to the peak LDI point."

One final assumption that further changes the picture is whether schemes hedge up to the asset value or higher. The report suggests schemes will hedge up to their asset level, and that very few go above.

However, Davies says: "You could make an argument that some schemes may want to hedge all the way up to the size of their liabilities rather than up to the size of their assets. Some schemes do that - if that changes then that's where you get to the 15-year point.

He says the most extreme example of hedging up to the value of gilt liabilities could increase the aggregate demand by around £200bn - which equates to additional three years of demand at £70bn per year.

Key points

At a glance

  • There are many risks to not hedging on the basis that LDI will peak in a few years
  • Lower pension scheme demand for gilts would not necessarily lead to higher yields
  • Some predict the peak could easily go out beyond 2021, even as far as 2033

More on Investment

Border to Coast strengthens responsible investment policies

Border to Coast strengthens responsible investment policies

LGPS pool outlines plan to vote for resolutions aligned with the Paris climate agreement

Martin Richmond
clock 29 February 2024 • 2 min read
Schroders posts 'resilient performance' despite 'market headwinds'

Schroders posts 'resilient performance' despite 'market headwinds'

Profits down as a result of restructuring costs but assets under management up by 2%

Cristian Angeloni. Additional reporting by PP.
clock 29 February 2024 • 3 min read
Partner Insight: How effective are impact investments?

Partner Insight: How effective are impact investments?

Impact investing has transformed over the past decade, giving investors the opportunity to pursue both financial returns and social and environmental outcomes.

Sarka Halas
clock 29 February 2024 • 2 min read