A hitchhiker's guide to LDI
Once the decision has been made to implement a liability driven investment (LDI) strategy to provide protection against interest and inflation rate movements, the implementation process still lies ahead.
There are a multitude of different areas in the implementation process that will require simultaneous attention. It is vital for the prospective parties involved, namely the trustees, LDI manager(s), investment consultant, lawyers, sponsoring employer’s treasury department, investment banks and scheme actuary to work together effectively to ensure a smooth and quick implementation process.
It is important to set aside time for regular update sessions in order to allow the different parties to communicate successfully and to reduce the possibility of bottlenecks occurring that can lead to delays.
One of the first decisions to be made is whether to use a pooled or a segregated solution. A pooled solution normally offers an easier implementation process because the due diligence from the legal side is usually a lot quicker as the scheme is effectively entering into a pooled fund.
However, there are two distinct disadvantages to using a pooled solution. The first is a loss of flexibility as the available options will largely depend on the pooled vehicles available from the managers. The second is whether the existing bond portfolio can be effectively incorporated into the LDI solution without being sold and incurring trading costs.
A segregated solution offers two main options: using the manager’s umbrella International Swaps and Derivatives Association (ISDA) master agreements or setting up a scheme’s own ISDA agreements with the individual banks. The former offers the advantages that:
q The ISDA agreements do not have to be negotiated separately with each of the banks, a process that can take months;
q The trustees do not need to sign an ISDA agreement each time a trade is executed;
q The trade is executed in the name of the manager, thereby concealing the identity of the scheme.
The advantage offered by the latter is that the terms and conditions of the ISDA agreements can be negotiated with the banks, and that can allow the utilisation of the expertise of the sponsor’s treasury department, which may already have ISDA agreements with some of the banks.
Some schemes may not be allowed to invest directly in derivative instruments such as swaps and credit default obligations but find that the existing pooled vehicles offered by investment managers lack the flexibility required to implement an LDI solution for their scheme. A possibility here is to look at setting up a pooled life fund policy, of which the scheme is the sole unit holder. However, additional time should be allowed for the legal due diligence to be carried out on the life insurance company issuing the policy and for reviewing the legal agreements. Separate agreements will be required with the manager and the life insurance company.
Agreeing and reviewing the investment manager agreements will probably take more time than with other mandates due to the additional complications of the investment guidelines and restrictions of the swap agreements. It is important to determine the level of flexibility left to negotiate with the manager, as often a tender contains the core outline of the strategy, with scope to expand further with a successful manager.
Discussing the strategy with the chosen manager before drawing up the legal documentation is essential, as value may be added through additional ideas generated at this stage. Time spent here can definitely save time later.
Larger schemes will generally have more flexibility in deciding the benchmark. If a custom benchmark is to be used, additional time will need to be set aside as it is likely to result in a unique benchmark with which the manager is unfamiliar. Additional complications in benchmark discussions may occur if more than one manager is used because of different solutions and expected tracking errors.
A large variety of benchmarks could be considered, for example targeting gilts + x%, corporate AA bond index (FRS17/IAS19 purposes), swaps + y%. The chosen benchmark will implicitly indicate how much active management is required to achieve it, and this will have an impact on the fee negotiations.
A trade-off between providing a manager with an investable benchmark and trying to protect the funding position relative to a specific actuarial valuation basis (eg, FRS17 basis) will most likely need to be made. For example, on a FRS17 basis the current market practice is for the scheme actuary to use a single corporate AA bond yield to value the liabilities.
This presents a problem because in practice the actual yield curve is not perfectly flat, and it does not stretch out as long as the liabilities, so the manager is not presented with an investable benchmark.
Therefore, a trade-off will need to be reached between providing the manager with an investable benchmark and trying to protect the scheme’s FRS17 funding position. As a result, it is crucial to understand how the scheme actuary derives the discount rate used in the valuation.
It is essential to focus on the expected net returns when engaging in fee negotiations, as a passively managed solution may produce a higher expected net return when compared to a more actively managed solution because of the relative difference in fees.
The two most common methods for providing the liability cash flow estimates over the whole term structure of the scheme are by using an uninflated cash flow approach (ie, fixed, RPI and LPI liability split) and through scenario analysis (based on different rates of future inflation). Different parties, notably scheme actuaries, are able to provide the liability cash flow estimates.
It is important to note that these estimates do not allow for future demographic changes, and therefore the LDI solution will not protect a scheme against this risk. The frequency for reviewing these liability cash flow estimates should also be discussed with the provider to achieve an optimal balance between accuracy and costs.
It may be possible to provide both cost savings and more frequent liabilities reviews by ‘piggybacking’ off the annual FRS17 review. In addition, a decision should be made on how to treat future accruals.
Further, the way to treat different LPI liabilities within the LDI solution needs to be settled, while the availability of and the prevailing market price for the appropriate LPI swaps would have to be discussed.
Another important consideration is the level of precision required in attempting to match the liability cash flows. Time should also be spent to decide whether the LDI solution should target the whole term structure of the liabilities and how the ultra-long liabilities (over 50 years) should be treated.
Finally, it is expected that the swap contracts used by approved pension schemes in an LDI solution will usually be tax exempt. However, each scheme will require its own legal team to establish whether or not this is the case. Consequently, it is advisable to consult the tax authorities.
It is important to plan the transition to an LDI strategy thoroughly, including the initial allocation. Different approaches that can be used include:
q An immediate transition, locking into the current market rates (however, due to delays in the implementation process it may be worth considering using derivative instruments to help protect the funding position in the short term until the LDI strategy has been implemented without incurring transactions costs from changing the underlying assets that are held);
q A phased transition over time, including the use of deficit correction payments to fund the LDI portfolio;
q Delaying the transition in the hope of implementing it in more favourable market conditions;
q Employing a dynamic phasing approach based on the funding level;
q Using a variety of derivative instruments to provide the opportunity for implementing it in more favourable market conditions in the future while providing downside protection.
It is extremely important to keep the implementation process of an LDI solution confidential. Swap prices can change considerably when the market knows that a scheme is about to implement such a solution, costing the scheme dearly. Assuming a 20-year duration, a change of five basis points in swap prices can cost a typical scheme about 1% of the assets in initial implementation costs.
Consequently, it is advisable to allow the manager enough discretion and time to implement the LDI solution. This is in addition to giving the manager some discretion for incorporating the existing assets in the solution or using them to secure more favourable swap prices.
In conclusion, while the practicalities of implementing an LDI solution may seem daunting, many schemes have implemented it satisfactorily. But it is advisable to allow adequate time and take expert advice as the implementation process is more complicated than that for traditional bond mandates and time is often the issue that is most underestimated.
Nevertheless, it is clear that an increasing number of pension schemes will pursue an LDI solution. Far from being just the current fashion, LDIs will prove to be the cornerstone for the future.
Martin Kraus is head of the UK investment practice client consulting at Hewitt Associates