The sharp spike in interest rates following the chancellor's "mini" Budget pushed many schemes into the eye of the storm, which could have been disastrous for some had the Bank of England not stepped in.
Concerns will relate not only to the security of member benefits but also to DB schemes’ use of and risks embedded in liability-driven investment strategies – paradoxically, the very structures put in place by many schemes to improve risk management.
What is the rationale behind LDI, what went wrong and what needs to change?However, it would be wrong to assume that all schemes will have been badly impacted by recent events. Indeed, many will have seen an improvement in their funding positions because of rising gilt yields, enabling them to reduce risks and create greater certainty for members.
So, what is the rationale behind LDI, what went wrong and what, if anything, needs to change?
To answer these questions, it is useful to look back at the history of DB schemes and what drove the use of LDI.
The rationale behind LDI
DB schemes are designed to pay out benefits to members based on a formula relating to length of service and salary. It is possible therefore to estimate the projected outgoings many years into the future. With these estimates, schemes could hold government bonds such that the income and maturity proceeds matched the projected benefit outflows.
As long as the estimated projections are reasonably accurate, this is a relatively low-risk way of making good on the member benefit promise.
However, in the early days when DB schemes were set up, they were operated in a manner akin to with-profit funds.
So much for the expectation that equities would outperform boring government debt.The core benefit was often a fixed pension without any inflation linkage with the contributions and investment strategy (typically with a heavy equity bias) designed to create an asset pool that would cover the benefits plus discretionary pension increases.
However, if the asset returns fell short, the discretionary benefit increases would not be given, so members effectively bore the investment risk with the employer only covering the cost if the assets fell short of that needed for the core benefits.
Over time, various forms of pension inflation linkage were forced on DB schemes. This resulted in the investment risk being moved from members to the employer (and is a key reason behind the demise of DB schemes).
Further, legislative changes also encouraged most DB schemes to assess how well funded they were by comparing the market value of the assets with the capitalised value of the future benefit payments using a discount rate linked to gilt yields (this ratio it typically referred to as the funding level).
In this new environment, the funding level and, consequently, the contributions the sponsor had to pay were driven by not only how the assets performed but also by how the capitalised value of the future benefits changed.
Leveraged LDI was often presented as the solution to these challenges.This could result in some very wide swings in the funding level if assets and liabilities moved in opposite directions – for example, when Covid hit financial markets in March 2020.
Furthermore, the steady decline in gilt yields over many years meant that liabilities were often rising at a faster rate than the assets, despite periods of strong returns from growth assets. So much for the expectation that equities would outperform boring government debt.
Leveraged LDI was often presented as the solution to these challenges.
Controlling the volatility in the funding level could have more simply been achieved by selling the growth investments and holding all the assets in government bonds. However, this would have reduced the expected investment return and required substantial additional contributions from sponsoring employers.
What LDI allowed you to do was retain your growth assets while also effectively buying bonds. With LDI, you engineered the same economic exposure as if you retained your current investments but then borrowed money and used it to purchase gilts.
The practical consequence of this is that the value of the chosen bonds would move up or down in line with the DB scheme’s liability value and thereby reduce the funding volatility to that of the growth assets held.
The argument for doing this was that funding level volatility could be reduced by implementing LDI without, in principle, affecting the DB scheme’s expected investment return.
What went wrong?
There are various ways of implementing LDI. One approach is to use interest rate swaps whereby the pension scheme commits to paying a floating rate of interest in return for receiving a series of fixed payments. The reference to 'swap' relates to the fact that the two parties are entering into an exchange that at the outset has zero value.
After initiating a swap contract, its value will move up or down depending on movements in yields. If yields fall, then the value of the 'fixed leg' becomes more valuable and creates an investment profit for the DB scheme that will help cover the rise in the liability value. The opposite occurs if yields rise.
It is constructive to consider recent market dynamics from a pooled-fund perspective.The leveraged nature of LDI means values can move dramatically for what may seem like a relatively small change in interest rates.
LDI can be implemented either on a segregated basis – that is, direct ownership of the underlying securities and derivatives – or indirectly via pooled fund structures. It is constructive to consider recent market dynamics from a pooled-fund perspective.
In a falling interest rate environment, which we have seen over much of the course of this century, the profits to pension schemes invested in LDI pooled funds resulted in the investment managers paying out cash to ensure the pooled funds maintained their leverage levels (thereby remaining attractive to new pension scheme clients wanting leveraged exposure).
These cash distributions have proved particularly beneficial to pension schemes, creating efficiencies in the payment of member benefits and rebalancing asset portfolios.
However, as yields started rising this year, LDI began to generate losses, and pooled funds rather than paying out cash started demanding it. The frequency of these cash calls increased rapidly, depleting any contingency cash held. This meant that other assets had to be sold, potentially at depressed prices, to fund the cash call.
A fundamental review of the short-term metrics used to assess pension scheme funding could be years in the making.The situation reached fever pitch following the chancellor’s fiscal statement, such that even if DB schemes had sufficient liquid assets to cover the cash calls, the size and speed of yield rises created real challenges in delivering the cash in time to keep the required exposures in place.
Alternatively, schemes could decide not to fund the cash call and let the exposures reduce. However, this ran the risk that if yields dropped suddenly then liability values would rise further than asset values. This is the situation some DB schemes will have found themselves in if their exposures were reduced just prior to the BoE’s intervention on September 28.
What needs to change?
The fundamental issue underlying this goes back to the history of DB schemes. The benefit and funding arrangements were predicated on long-term investment in growth assets.
However, the measurement of a DB scheme’s progress to delivering the member benefits was structured around short-term metrics related to the current cost of guaranteeing member benefits – that is, via investment in government bonds and low risk credit.
If your scheme is not able to completely de-risk, rethink your investment strategy and how it is expected to evolve.This had the effect of driving investment behaviour, namely the implementation of LDI to better align the longer-term investment needs with the shorter-term risk metrics. However, while this reduced one risk, the introduction of leverage created other risks – the effects of which have now become clear.
So, is LDI the problem? Or do the metrics for measuring the health of pension schemes need to be reviewed? Given recent events, I would contend the answer to both questions is yes.
While LDI managers have immediately begun looking at the operation of their LDI programmes, a fundamental review of the short-term metrics used to assess pension scheme funding could be years in the making.
The good news is that for many schemes, the rise in yields and credit spreads may now provide an opportunity to go back to first principles; buy un-levered gilts and corporate bonds such that the income and maturity proceeds align closely with your projected cash flows without any worries over future cash calls.
Many schemes now have an opportunity to reduce the role that leveraged LDI plays in investment strategies without abandoning focus risk management.If your scheme is not able to completely de-risk, rethink your investment strategy and how it is expected to evolve. Many DB funding and investment strategies assume a gradual reduction in growth assets each year, with the expected holding period of some of these being relatively short.
By reconfiguring the sequence in which assets will be sold down, schemes could hold a lower allocation to riskier growth assets now, knowing that they expect to hold them for longer to generate the expected returns.
This will free up capital to hold more in gilts and bonds and thereby reduce leverage while retaining the ability to closely manage the pension scheme balance sheet.
In summary, we believe many pension schemes now have an opportunity to actively reduce the role that leveraged LDI plays in investment strategies without abandoning focus on either prudent risk management or alignment with the metrics underlying the current regulatory regime.
Ultimately this means questions of whether LDI remains fit for purpose can be side-stepped by many.
Carl Hitchman is head of UK investment consulting and chief investment officer of Buck Consultants
