Written evidence submitted by The Investment Association LDI0081

Summary

• The Investment Association (IA) represents the asset management industry operating in the UK. Our members, acting as agents for their clients, manage the investments of UK defined benefit (DB) and defined contribution (DC) schemes, as well as other retail savers and institutional investors. They are responsible for the management of around £10 trillion of assets in the UK on behalf of domestic and overseas investors. For DB pension schemes, working closely with scheme trustees and with investment consultants, our members collectively manage Liability Driven Investment (LDI) mandates that hedge £1.6 trillion of UK pension liabilities.

• We are responding to this second call for evidence in order to offer our views on two important regulatory developments highlighted by the Committee since it began its work: the publication of TPR’s draft DB Funding Code; and the Bank of England’s Financial Policy Committee (FPC) recommendations on improving the resilience of LDI strategies.

• TPR’s draft DB Funding Code enshrines the existing approach of DB schemes to asset-liability management, which we regard as positive, given the benefits this approach has brought to scheme members and their sponsors. At the same time, there is sufficient flexibility within the regime for less mature schemes to take more investment risk where this can be supported by the corporate sponsor.

• Regarding the possibility of whether the Code encourages herding of scheme investment behaviour through the ‘Low Dependency Investment Allocation’, while it is true that, in aggregate, mature DB schemes have similar investment strategies, this issue is one that pre-dates the Code. Concentration of ownership of long-dated and index-linked gilts by DB schemes most likely did exacerbate the issues caused by the September 23rd fiscal event. However, the solution here may be to look at ways to diversify the pool of ownership of these assets, rather than seeking to fundamentally change the way that DB schemes invest.

• Although not part of the Code, we also briefly discuss the issue of accounting standards, which have been raised by some commentators as a contributor to the LDI market turbulence through their impact on pension scheme investment decisions. We do not share this view: the aggregate shift of the DB sector into bonds is driven primarily by increasing maturity of DB schemes combined with the effects of TPR’s DB funding regime, with its focus on asset-liability matching.

• In relation to the FPC’s initial recommendations on enhancing the resilience of LDI strategies, we agree with the sentiments and the broad recommendations outlined: indeed, industry action in the wake of the Bank of England’s intervention is consistent with the initial FPC recommendations. Going forward, while we defer judgement until the FPC’s next statement on the matter in March, we urge regulators to proceed with caution on defining steady state resilience. While it is important to increase resilience, any policy recommendations should bear in mind the trade-off between overall resilience of the system and the ability of DB schemes to continue to be able to effectively match their assets and liabilities in a cost-efficient manner.

TPR’s DB Funding Code, herding of pension scheme investment strategies and the impact of accounting standards

TPR’s draft DB Funding Code

1. As we stated in our first round of written evidence to the Committee, it is not obvious to us that the events of September/October 2022 mean the entire DB funding and investment framework needs review. Rather, in the context of an unprecedented market shock emanating from outside the pension system, the lessons to be learned are more on the side of governance and operational issues in the context of significant systemic stress, as we have previously discussed.

2. There is no debate over the issue of pre-funding for corporate pension plans (i.e. putting aside money to invest toward the payment of pension benefits as they accrue). When the payment of pension liabilities decades into the future is conditional upon the continued existence of a sponsoring employer, it is only sensible to reduce the reliance on the sponsor covenant over time, by seeking to move the scheme to a position where it is fully funded.

3. Consistent with the principle of avoiding unnecessary/unrewarded risks, is a need to reduce investment risk as the scheme becomes mature and the payment of pension benefits comes closer. In that regard, the DWP’s draft Occupational Pension Schemes (Funding and Investment Strategy and Amendment) Regulations 2023 and TPR’s draft DB Funding Code, which provides detailed guidance for trustees under those regulations, seek to formalise the existing approach towards asset-liability management by DB pension schemes.

4. It is worth recalling the history here, that the introduction of the 2004 scheme funding regime was in part a response to the previous DB investment paradigm of significant equity allocations, aggressive assumptions on assumed rates of return by which liabilities were discounted, and a consequent reduction in sponsor contributions, which collectively led to big deficits opening up when global equity markets fell in the early 2000s. As stated in our earlier written evidence, the asset-liability management approach embodied in the current regime has delivered benefits for sponsors and scheme members by stabilising funding levels and increasing the security of benefits. The draft regulations and Code’s focus on codifying the existing approach is therefore helpful.

5. We welcome the discussions in the Code related to the expected characteristics of a mature DB scheme’s investment portfolio, namely the need for cash-flow matching, resilience of the funding ratio to short term adverse market movements and the need for good liquidity management policies. We note that further material for trustees may follow in light of the ongoing work by the Bank of England’s Financial Policy Committee (FPC) on enhancing the steady state resilience of LDI strategies.

The herding of DB investment strategies and systemic risk

6. Some commentators have questioned whether the DWP and TPR’s introduction of a ‘low dependency investment allocation’ – whose characteristics as per the regulations are a portfolio of assets that allow for broad cash-flow matching while simultaneously allowing the scheme’s funding ratio to be highly resilient to short-term adverse changes in market conditions – will create the risk of herding in investment strategies by DB schemes seeking to comply with the regulations as they mature. In this context herding refers to schemes investing in the same way. The concern is that the potential for systemic risks increases when large numbers of DB schemes invest in a similar fashion.

7. The herding dynamic was undoubtedly seen in the autumn 2022 turbulence in the LDI market. With large numbers of schemes simultaneously reducing hedging levels, selling long-dated and index-linked gilts in order to do so, conditions were created under which there were large numbers of sellers, with no buyers. This had the effect of leading to falling prices, further collateral calls, more deleveraging and ultimately, a negative price spiral. It was only the intervention of the Bank of England as buyer of last resort that stopped these market dynamics.

8. Corporate DB schemes do tend to follow similar investment strategies that focus on matching assets and liabilities, with regulation and scheme maturity incentivising them to do so. At the micro level this has worked well for individual schemes, and the principles behind these strategies remain valid. Accordingly, we do not expect the guidance in TPR’s new draft Funding Code to change the risks posed by any herding: if there are any such risks, they already exist. However, it is not clear what the alternative for DB schemes is. With most DB schemes already heavily invested in matching assets and seeking a similar end-game, it is natural that they will end up investing in a similar fashion. Indeed, artificially seeking to reverse this shift may itself trigger systemic events if schemes change investment strategies and exit their matching exposures en masse.

9. It is important to emphasise again the exceptional nature of the September 2022 shock, rather than the DB funding regime being a root cause. It took an unprecedented spike in gilt yields, resulting from an event exogenous to the pension system, to trigger a systemic event in a part of the gilt market that otherwise behaves in a generally stable and orderly fashion. At that point, the Bank stepped in to reduce the risks to financial stability in the UK. In any financial system, a Central Bank will always have a role to play in guaranteeing financial stability and it is impossible to create a system where Central Bank intervention will never be required. Rather, the focus should be on reducing the likelihood of Central Bank intervention in the first place. It is here that the important governance and operational lessons learned from the LDI episode can play their part in making the system more robust.

10. The financial stability concerns arising from the concentration of DB schemes’ ownership of long-dated and index-linked gilts need to be weighed against the benefits of such an ownership profile: the UK government has benefitted from a steady and robust demand for its long-term and index-linked debt from pension schemes and insurance companies. Data from the DMO shows that pension funds and insurance companies owned around a quarter of the stock of gilts by market value in the third quarter of 2021[1]. More recent data from the ONS shows that by the summer of 2022, corporate DB schemes had long-dated (25+ years) and index-linked gilt holdings of just under £400 billion[2]. The influence of the DB and insurance sectors can be further seen when comparing the average maturity of the UK debt stock to other countries: for the UK, this is just over 14 years, compared to 8 years for France and Japan and just under 6 years for the US[3]. A long average maturity of debt significantly reduces the UK government’s exposure to refinancing risk, by enabling gilt issuance to be spread along the maturity spectrum.

11. We also encourage policymakers to consider if it is the characteristics of the underlying market that are also causing a problem here, rather than the behaviour of DB pension schemes (and associated LDI investment strategies) per se.  The concentration of ownership in parts of the gilt market does appear to be a factor in the recent crisis, with one solution being greater emphasis on encouraging a more diversified group of buyers. Concerns over gilt ownership should not prevent DB schemes from using instruments and investment strategies that help them manage their risks effectively and efficiently. 

12. In the longer term, there may be a significant reduction in demand from DB pension schemes for further gilt purchases as these schemes mature and wind-up or transfer their assets and liabilities to insurers.  Furthermore, insurers are subject to a different prudential regulatory environment in the form of Solvency II and are expected to have access to a wider range of assets in their portfolios i.e. they will not be solely restricted to UK gilts and corporate bonds. This may reduce the potential risks emanating from concentrated ownership of parts of the gilt market.

DB schemes’ shift to investment in bonds and the impact of accounting standards

13. We would urge caution with respect to the role that accounting standards may have played in driving the aggregate shift by DB schemes to investing in bonds.  One significant achievement of the changes in accounting standards for corporates has been to make the cost of DB promises more visible and changed the lens through which DB promises are viewed by corporates to one of risk management.

14. In our view, scheme investment behaviour is primarily the result of the actions of trustees independently exercising their fiduciary duty over the investment of scheme assets in the context of TPR’s funding regime, combined with a growing maturity of the DB system. These two factors have resulted in a DB system which is naturally more focused on asset-liability matching as schemes become cashflow negative.

15. For those schemes that are less mature and can continue to take higher levels of investment risk, there is sufficient flexibility in the DB funding regime to enable them to do so. Moreover, the use of LDI does not imply that schemes cannot continue to invest in growth assets. Schemes can, and do, maintain both growth and matching components to their portfolios.

16. The overall macro trend on DB provision is not going to reverse: corporate DB schemes are now largely closed and are only going to become more mature. Given these conditions, there is no basis to presume that changing accounting standards will change scheme investment behaviour.

FPC recommended policy actions to improve the resilience of LDI strategies

17. We note the FPC’s analysis of the LDI episode in its December 2022 Financial Stability Report and broadly agree with its recommendations. However, given the high-level nature of the recommendations and the fact that the FPC is working to provide further detail in the coming weeks, a firmer assessment will necessarily be deferred. For now, we provide some initial commentary on each of the recommendations.

Regulators should take action to ensure continuing resilience to the prevailing level of interest rates following the LDI market turbulence

18. This recommendation effectively formalises the work that has already been done by the DB investment industry and its various regulators in response to the immediate market pressures following the September 23rd fiscal event. In that regard, we note the various statements of 30 November 2022 by the FCA[4], TPR[5], the CBI[6], the CSSF[7] and ESMA[8] as signalling their satisfaction with the current levels of resilience as measured by collateral buffers[9] – around 300-400bps on average – to use in the event of further interest rate rises. These statements reflect significant efforts made by pension schemes and their investment managers to recapitalise their LDI hedges in the period during which the Bank intervened in the market. As such, the recommendation already reflects where the industry is today.

Maintain financial and operational resilience – including risk management of any liquidity relied upon outside LDI funds – to withstand severe but plausible market moves

19. This recommendation appears to address the need for good liquidity management by pension schemes so as to be in a position to meet additional collateral calls in the event of buffers being exhausted. This is a sensible step and should already be part of what trustees do as part of their investment implementation. TPR’s interim guidance on maintaining LDI resilience provided some helpful initial points for consideration and the draft DB Funding Code contains further guidance on liquidity management but avoids prescription. We await to see whether the FPC’s further work will provide more specificity in this area, particularly around minimum standards for pension schemes’ management of liquidity.

Regulators should set out appropriate steady-state minimum levels of resilience for LDI funds more broadly in relation to operational and governance processes and risks associated with different fund structures and market concentration

20. We acknowledge the need to ensure that LDI strategies are resilient on a steady state basis. However, in setting these standards, particularly in relation to collateral buffers, regulators must take care to ensure that they are not set at such a high level that the product set becomes unusable for clients.

21. In particular, regulators need to consider the trade-off between resilience to yield shocks, efficiency of the product and risk management of DB liabilities. An objective of minimising systemic risks would lead to maximising the size of collateral buffers. The risk here is that buffer levels that are set too high would demand too much collateral from schemes, leading to a potentially sub-optimal asset allocation to cash/cash-like assets in order to meet buffer requirements.

22. If the required buffers were unable to be met, hedge levels would have to come down (to reduce collateral demands), resulting in increased asset-liability mis-match risks and more volatile funding outcomes. Therefore, when setting steady state resilience standards, regulators should ensure that the risks being calibrated to are not at the extreme end of the outcome distribution.

23. It is also important to allow for sufficient time to implement any new steady state resiliency levels. If the transition is not done properly, there is the risk of forced and co-ordinated action which may destabilise markets, as schemes seek to recapitalise and/or adjust strategic asset allocations to be consistent with the new standards. Schemes should therefore be given sufficient time to transition to any new steady state. 

24. In the new steady state, if and when collateral buffers are once again depleted, there should be sufficient time to replenish them. This may need more forbearance from derivative counterparties, and regulators can help here by ensuring the counterparties recognise the operational constraints of the pension schemes they are transacting with.

25. As a final comment on this recommendation, we disagree with the FPC’s focus on LDI pooled funds as a key source of systemic risk. There were also examples of negative outcomes for schemes invested with segregated portfolios. The driver of outcomes is scheme governance and investment strategy, not the investment structure used to access LDI strategies. There is a long tail of small to medium-sized DB schemes for which pooled funds are the only option to access LDI. These schemes still need the product, their hedging needs are invariant to their size. The risk of singling out pooled funds for additional measures is that it risks making the implementation of LDI less achievable for small and medium-sized schemes.

Bank counterparties to the LDI sector should apply a prudent approach when providing finance to LDI funds, taking into account the resilience standards set out by regulators, and likely market dynamics in relevant stressed conditions

26. Feedback from our members is that LDI counterparties are already showing more flexibility to pension schemes in relation to their posting of collateral[10]. This is another area where more detail is needed from the FPC before any judgment can be made, but as we noted in our initial written response to this inquiry, broadening the range of non-cash collateral that pension schemes can post would have alleviated some of the pressures felt last autumn.

Regulatory and supervisory gaps on the LDI sector should be filled, so as to strengthen the resilience of the sector, and improve governance and investor understanding

27. There is no doubt that regulation of the LDI market is complex as a result of the different regulators involved in regulating DB pension schemes, investment management firms and their products. Overlaid on this is the role of the Bank of England in ensuring the stability of the UK’s financial system. There are good reasons for the existence and role of these different regulators, and a simplification of the regulatory landscape on LDI is not something that we think is necessarily desirable or achievable.

28. However, we do think there is greater scope for consistent reporting of LDI exposures from DB pension schemes and investment managers, coupled with greater data sharing and co-ordination amongst the different regulators, including on a cross-border basis. The various regulators should work together with the pensions and investment management industries to define the necessary dataset and then ensure that between them, they have access to this information. Any information-gathering should be proportionate, decision-useful and not duplicated across regulators. As well as benefitting the industry through proportionate data requirements, a consistent set of data points will ensure there is no regulatory arbitrage between different regulators and regimes.

29. In terms of what that dataset might be, the following aggregate data points could be helpful for regulators:

• DB pension asset allocation (growth and matching assets)
• LDI strategies: aggregate value and composition of physical assets, derivative[11] and repo exposures.
• LDI reporting should be split by pooled funds and segregated portfolios
• Headroom (yield increase) to asset exhaustion in collateral pools supporting LDI strategies

30. We await with interest the further FPC work in this area and are ready to engage with the Bank to help it achieve its objectives on reporting.

Given the role played by investment consultants in influencing the investment strategies of pension schemes through the unregulated investment advice they provide, the FPC supports bringing investment consultants into the FCA’s regulatory perimeter

31. There are two distinct issues here:  whether consultants should be within the FCA’s regulatory perimeter, and whether this would have directly made a difference to the course of events through Autumn 2022.

32. Given the size of pension scheme assets and the influence of consultants’ advice, we agree with the FPC’s recommendation in this area and encourage HM Treasury to set out its plans to implement this recommendation.  As we noted in our previous submission to this inquiry, the role played by investment consultants in the UK pensions market was analysed in depth by the CMA in its 2017-19 investigation[12] of the UK Investment Consultancy market. The CMA noted that some aspects of consultant advice to pension schemes – around investment strategy, asset allocation, manager selection – were unregulated and recommended that these areas of consultants’ activities be brought within the regulatory perimeter of the FCA.

33. While the government has previously committed to considering this recommendation and consulting in due course[13], no timetable has been set out for doing so. In our submissions and representations[14] to the CMA during the course of its work, we argued for this recommendation to be taken forward on the grounds that it would increase regulatory oversight over the investment advice given to pension schemes.

34. However, we think it is unlikely that FCA-regulation of investment consultants’ advice to pension schemes on investment strategy design and implementation would have avoided the issues experienced in September and October 2022. The impact of rising rates on LDI strategies was known to consultants and, based on our discussions with our members, schemes were working well with their consultants and managers to design collateral waterfalls and recapitalise as appropriate, as interest rates rose through the spring and summer of 2022.

35. As with other actors in the investment chain, the speed and scale of the September shock created the challenge for consultants. Given that regulators had deemed a 100bps rate shock as sufficient for stress-testing purposes, we consider it unlikely that they would have been instructing consultants to ensure their advice reflected the need to prepare for larger shocks, had they been responsible for regulating this advice.

March 2023