Written evidence from the Institute and Faculty of Actuaries LDI0046

Key points

Most defined benefit (DB) pension schemes were affected by the recent market instability, but the scale and direction of the impact will have varied from scheme to scheme. The impact will have depended on the profile of each scheme’s assets and liabilities, as well as the governance arrangements in place and the ability of schemes to react quickly and appropriately to changes that occurred with unprecedented speed.
Due to the nature of a DB scheme’s liabilities, many schemes will have seen a reduction in the overall size of their scheme, with both assets and liabilities reducing. For many schemes, liabilities reduced more than assets, leading to an improvement in funding levels. Others may not have seen an improvement in the funding level, but may still have been left with a smaller overall monetary deficit to be funded by the employer, resulting in an improvement to the security of members’ benefits and making buy-out a more achievable option. A minority of schemes have seen a deterioration in their funding level, due to hedging and liquidity profiles, which meant that they were forced to reduce hedging levels at an inopportune time.
We do not understate the importance of the Bank of England’s intervention, without which the outcomes could have been very different (both favourable and unfavourable for pension schemes, with more extreme outcomes and potentially having a wider impact beyond the pension scheme population). However, in practice, the crisis for schemes pursuing LDI strategies was a short-term liquidity crisis, and for many schemes a complex operational challenge, but it was not a funding crisis, nor did it lead to the collapse or insolvency of any DB schemes. Nevertheless, the IFoA believes there are lessons that schemes and their advisers will need to take on board going forward.
The IFoA is conscious that fast movements in market conditions cast a spotlight on the governance arrangements that support LDI strategies, with schemes that were able to act quickly generally having better outcomes. In practice, this meant that schemes with segregated LDI arrangements (typically used by larger schemes) generally fared better than those invested in pooled LDI funds (typically used by smaller schemes), who were less able to act quickly, and/or control the actions taken so as to preserve their hedging position, leading to a loss of hedging for some schemes at the height of the yield rise. This meant that these schemes saw a deterioration in their funding positions as yields fell again.
The issues faced by schemes pursuing LDI strategies highlights that good governance to support such strategies is critical. The IFoA expects to contribute in its role as the professional body of one of the various professional groups advising in this area, to ensure that future stress testing of LDI, and liquidity margins held in future, reflect the new understanding based on recent events of the extent and speed at which it is possible for yields to move, and the issues around lack of liquidity and depth in the relevant parts of the bond markets themselves that contributed towards this.

It is clear that recent events have also highlighted systemic concentration risk – the risks of encouraging all schemes to invest and manage risks in similar ways, potentially leading to the spiral of selling gilts to meet collateral calls which further pushed yields up, as we saw recently, and which ultimately led to Bank of England intervention. As we commented in our recent response to the DWP’s consultation on draft DB funding regulations, there are risks in herding all schemes towards a common gilts-based cashflow matching strategy, and we encourage DWP to reconsider the relevant provisions in the proposed new funding regulations.
The IFoA believes that ultimately LDI strategies remain an important risk management tool to help DB schemes manage their inflation and interest rate risks. Including some leverage in their LDI strategy allows schemes to more quickly reduce interest rate and inflation risk whilst seeking some modest return which may still be required to reach full funding. While we believe such strategies have served many DB schemes, their sponsors and their members well over the last decade or so, it is clear that such strategies, and in particular the level of leverage used, need to be reviewed to take into account the updated calibration of the risk and reward trade-offs as a result of the changes to liquidity and collateral requirements. The IFoA envisages improvements to governance, knowledge and understanding in future, as well as increases in liquidity margins, and notes that many investment managers and advisors have already taken some steps to address this.

The Institute and Faculty of Actuaries (IFoA) welcomes the opportunity to submit evidence to this inquiry by the Work and Pensions Committee on Defined Benefit pensions and liability driven investments (LDI). Our key points are set out above. For the purposes of our more detailed submission, the IFoA has grouped material under each of the focus areas listed under the Committee’s terms of reference.

Should you want to discuss any of the points raised please contact Caolan Ward, Policy Manager (caolan.ward@actuaries.org.uk), in the first instance.

The impact on DB schemes of the rise in gilt yields in late September and early October.

To set the context for this submission, the IFoA notes that LDI strategies are an important and genuine risk management tool to help DB schemes manage their inflation and interest rate risks. Some schemes use leverage within their LDI portfolio, which allows them to target a higher level of interest rate and inflation protection whilst maintaining some allocation to assets providing some modest return, often with a view to meeting modest funding shortfalls. In spite of recent events, we believe such strategies have served many DB schemes, their sponsors and their members well over the last decade or so. If these strategies had not been in place during periods of yields falling, sponsors would have been required to pay higher contributions, which may have affected their ability to invest in their business or even the viability of their business.
However, it is clear that changes will be needed going forward, and we already see some of those changes happening. Most schemes and providers have reviewed and in many cases reduced their levels of leverage and increased liquidity buffers. Schemes are also more generally reviewing their risks and governance arrangements to ensure they are appropriate. There is more to do, but the IFoA would caution against any knee-jerk reactions.
To understand what happened, it is helpful to breakdown the impact across the population of DB schemes operating in the UK – in what circumstances were schemes most affected, and why? What proportion of schemes were affected and to what extent? It should be noted that the observations below are based on our knowledge to date – confirmed asset and liability valuations that reflect the funding position of schemes after the effects of the market turmoil will emerge over time.

While many schemes with LDI strategies were affected by the recent instability, the extent of the impact and the ability of schemes to deal with it was mixed. A significant number of schemes with LDI strategies reduced their hedging levels, with a minority being forced to do so (in particular smaller schemes using pooled funds and schemes that had a high allocation to illiquid assets outside their LDI portfolio). These schemes were affected most due to being less able to quickly access liquid assets outside of the LDI fund to meet collateral calls as yields increased. Some schemes were forced to reduce hedging levels when yields were at their highest, and then were not fully protected when yields reduced again. A minority of these schemes may have suffered relatively material falls in funding levels as a result.

In contrast, a large number of schemes with LDI strategies were able to do enough to meet liquidity requirements without any forced changes to hedging levels. These schemes will have emerged with funding levels broadly intact.

Finally, some schemes that had not fully hedged their positions prior to the market changes are likely to have seen improvements in their funding position – for some schemes these improvements could be substantial.

It is worth noting that even those schemes whose funding levels reduced may not necessarily be worse off. This is because with assets and liabilities both decreasing, the nominal amount of any deficit may also have reduced. Recovery Plans (which usually provide for set nominal contributions from the sponsor) will be unchanged, and are therefore likely to make good the shortfall faster than before. Similarly, if deficits are now smaller in nominal terms, it is easier for sponsors to continue to support the scheme.

Whilst in practice there were winners and some losers among the schemes with LDI strategies, along with significant operational issues for many, this was ultimately not a DB pension scheme funding crisis, but rather a short term liquidity crisis. Having said that, we do not understate the importance of the Bank of England’s intervention, without which the outcomes could have been very different (again both favourable and unfavourable for pension schemes, with more extreme outcomes and potentially having a wider impact beyond the pension scheme population).

The impact on pension savers, whether in DB or DC pension arrangements.

For the purposes of this question, we have looked at the impact of changes in gilt yields, rather than the impact of LDI strategies. The two impacts are entirely distinct – gilt yields affecting both DB and DC schemes, and LDI only DB schemes. Further, the impact on DB schemes tends to be in relation to the overall security of members’ benefits and is less likely to have a long-term impact on an individual’s pension scheme benefits other than the terms for member options. In contrast, the impact on DC schemes will directly affect members’ benefits, albeit this impact may be temporary and only directly affect benefits which are crystallised whilst asset values are depressed.

This is indicative of the transfer of risk from trustees and corporates to individual scheme members as benefits transition from DB to DC – a feature the IFoA highlighted in its 'Great Risk Transfer' campaign.

DC savers

In terms of DC savers who are further from retirement, the impact they suffer may not have been acute. This is predominately due to the market reversion that has taken place since the ‘mini-budget’. However, there is a tranche of DC members (for example, those who may have wanted to draw their benefits in September/October 2022) who could have seen their ‘pot’ weaken significantly if they were heavily invested in gilts. Most DC members are invested in “Lifestyle” strategies which adjust their invested assets over time depending on whether members will ultimately use their DC pot to take cash, use drawdown or purchase an annuity. In some cases, members may have previously made a choice to use a lifestyle option targeting an annuity, or this may be the default option in their DC scheme, but it is not actually their intention to purchase an annuity. Annuity targeting lifestyle strategies tend to have a high allocation of assets invested in gilts and so these members may have seen a significant adverse impact on retirement outcomes.
Conversely, for members who sought to purchase an annuity during the market turmoil, it is likely that annuities were much more attractive, and it would not be surprising to see an uptick in annuity purchases. Previously members saw annuities as an expensive option, but they may now appear better value to them.

The direct impact on DC savers’ benefits suggests there is scope to reconsider the impact of the transferring risk from DB to DC schemes – please see IFoA 'Great Risk Transfer' report for full details.

DB savers

Members of DB schemes are likely to have seen the amounts of cash equivalent transfers fall significantly over the last year and particularly the last few months, as yields increased. Most schemes update pricing no more than monthly, so the values will not have been as volatile as the market itself. These changes in yields may also affect other member options, such as commutation, in future.

As noted above, the security of DB benefits has generally not changed. Some schemes have had funding level shocks, but employer insolvency has not emerged as a real risk, and for many employers remaining deficits are much smaller in nominal terms than they were previously.

Given its responsibility for regulating workplace pensions, whether the Pensions Regulator has taken the right approach to regulating the use of LDI and had the right monitoring arrangements.

Whilst LDI strategies have been successful in hedging pension scheme risks over the last decade or so and were recognised as such by trustees and the Pensions Regulator, recent events highlighted a form of systemic concentration risk that exists over and above the risks to individual schemes.

In other words, the impact of sudden and significant gilt yield movements would generally have had a self-contained and manageable impact on a single scheme with an LDI strategy, requiring collateral and rebalancing, but not leading to a wider gilt market shock. It was the cumulative impact of increasing gilt yields on many schemes running similar strategies, as well as the lack of liquidity and depth in the long-dated gilt market itself, which led to a spiral of gilt sales, increasing yields and further collateral calls, until the Bank of England intervened to stabilise the market.
As we commented in our recent response to the DWP’s consultation on draft DB funding regulations, there are risks in herding all schemes towards a common gilts-based cashflow matching strategy, and we encourage DWP to reconsider the relevant provisions in the proposed new funding regulations. However, we would stress that we do not see a need for tighter regulation of scheme funding, and it is not obvious to us that the Pensions Regulator is best placed to regulate LDI.

The Pensions Regulator has generally welcomed the use of LDI strategies to hedge risks and has rightly encouraged trustees to monitor risks to their schemes using an ‘Integrated Risk Management’ framework, which considers the interrelated risks to pension scheme funding, investment and employer covenant, with a view to ensuring risks remain within an acceptable tolerance and are mitigated where necessary. Monitoring of risks by trustees is therefore commonplace, but the IFoA believes this monitoring and understanding of risk will need to be extended going forward.

With this in mind, we believe that it is essential to clarify where responsibility should lie for monitoring and mitigating systemic risk associated with LDI. This does not naturally fall to the Pensions Regulator since it relates to financial stability. Rather, we believe that a joint regulatory approach is the way forward, to achieve a comprehensive overview and appropriate regulation of LDI from all perspectives (considering, for example, providers, investment advisers and managers, actuarial advisers and systemic risk), whilst ensuring that LDI remains a useful tool to manage pension scheme risks.

Whether DB schemes had adequate governance arrangements in place. For example, did trustees sufficiently understand the risks involved?

The IFoA believes that governance around LDI will need to be reviewed to consider the key learnings of the last few weeks. In particular, schemes adopting LDI strategies will need to ensure that they have in place a governance structure to support those strategies by enabling them to react quickly enough when large, unexpected market changes occur.

It is clearly also important for trustees to be able to demonstrate that they have received appropriate advice on LDI (as with any investment), and that they have strong and appropriate governance arrangements and risk monitoring tools in place to support their investment decision-making going forward.

The magnitude and speed at which yields changed in late September and early October was unprecedented over similar time frames. This experience, and the liquidity issues it created for schemes, have changed the understanding of the risk/return trade-off for LDI, and the extent of the testing needed to ensure sufficient liquidity. Actuaries and others advising trustees on LDI will be recalibrating their models to take account of this new understanding, helping trustees make the decisions required of them in future.

In our experience trustees routinely get training on managing situations involving collateral and risk, and we expect that most trustee boards understand the risks involved with LDI strategies, particularly in larger DB schemes. However, knowledge and understanding of risks could be further improved, particularly regarding the existence of a systemic concentration risk and its potential impact. Also, for smaller schemes in particular, there may need to be additional support to ensure trustees are encouraged to appropriately balance the advantages of LDI against the need for a strong governance framework to support it and have appropriate delegation in place to act quickly when required.
We are also aware of differences in the way pooled and segregated LDI funds fared and believe there is an under-appreciation of the operational challenges and governance needs for pooled LDI funds, which generally have less flexibility and fewer options to manage collateral quickly, given they are dealing with a large number of pension schemes and may have no immediately available source of liquidity outside of the portfolio. Having said this, the IFoA believes pooled LDI funds play a useful role in providing access to the LDI market to smaller pension schemes, and we understand that some LDI fund providers are already looking to address the issues that have arisen recently, by requiring trustees to hold liquid assets with the provider alongside the LDI fund. Generally, segregated funds are much better equipped to act quickly and flexibly to address changes in yields.

The IFoA believes trustees generally monitor liquidity on a quarterly basis, with limits set up to allow (until recently, at least) adequate collateral. Recent events have demonstrated that there would be value in stress testing more extreme risks. One suggestion for this would be to employ the type of reverse stress-testing used by insurance companies, which considers the magnitude of risks that can be tolerated.
The recent issues faced by schemes pursuing LDI strategies clearly highlighted the importance of good governance, and the IFoA expects to play a role in supporting the changes that need to be made to ensure future stress-testing of LDI, and liquidity margins held in future, reflect the new understanding based on recent events of the extent and speed at which it is possible for yields to move, and the issues around lack of liquidity and depth in the relevant parts of the bond markets themselves that contributed towards this.

Whether LDI is still essentially ‘fit for purpose’ for use by DB schemes. Are changes needed?

The IFoA believes that ultimately, LDI strategies remain an important risk management tool to help DB schemes manage their inflation and interest rate risks whilst seeking some modest return. We also believe such strategies have served many DB schemes, their sponsors and their members well over the last decade or so, although it is clear that such strategies need to be reviewed to take into account the updated calibration of the risk and reward trade-offs needed as a result of the changes to liquidity and collateral requirements. The IFoA envisages enhancements to governance, knowledge and understanding in future, as well as increases in liquidity margins.

It will be necessary for there to be greater regulatory monitoring of systemic concentration risk in future. However, this should not prevent schemes continuing to use this important risk management tool.

Does the experience suggest other policy or governance changes needed, for example to DB funding rules?

As we commented in our recent response to the DWP’s consultation on draft DB funding regulations, there are risks in herding all schemes towards a common gilts-based cashflow matching strategy, and we encourage DWP to reconsider the relevant provisions in the proposed new funding regulations. However, we would stress that we do not see a need for tighter regulation of scheme funding, nor for the Pensions Regulator to regulate LDI, although we do envisage other regulatory bodies considering this.

As we note above, it is clearly important for trustees to be able to demonstrate that they have received appropriate advice on LDI (as with any investment), that they have strong and appropriate governance arrangements and risk monitoring tools in place to support their investment decision-making going forward, and that they are either able to take action quickly when required, or have delegated responsibility appropriately to those who can.

November 2022