What approach to take for a liability-driven investment strategy (LDI)? How to implement it? How far do you go want to go with LDI? These are the questions of pension fund trustees inclined to put LDI into practice. However, and despite all the talk, it is not so easy for pension plans to get to the implementation stage for LDI.
Much of the discussion about LDI is still rather abstract. Academics, actuaries and investment professionals argue - sometimes dogmatically - about the pros and cons of LDI. Unfortunately, there is a high degree of confusion around the term LDI. The definitions range from ‘revolutionary new risk management' to ‘what you have always been doing'. No surprise that many trustees have remained sceptical as they feel this is just another investment fashion that will fade away.
In a narrow sense, LDI is about exactly matching the cash flows of your future pension payments with cash flows of your assets. In a broad definition, LDI is about taking liabilities into account when setting the investment strategy. In-between the extremes, a broad range of concepts and products have been proposed.
Whatever the definition of LDI, most pension plans in Europe have been forced to take a much closer look at their liabilities than in the past, to review their investment approach and to upgrade their risk management. The reasons for are well rehearsed, including higher pressures from sponsors, members and regulators.
A number of pension plans in Europe have already implemented some sort of LDI strategy in recent times. A bigger number is thinking about how and when to do it best. It is too early to assess the results of such bespoke LDI strategies, but a few trends have already emerged:
o There is a very broad range of approaches taken by different pension funds in different countries;
o The implementation of LDI is very scheme-specific;
o Traditional methods co-exist with new LDI techniques that are being continuously produced.
Pension boards are right to think long and hard about what approach to choose for LDI. It is not a technical issue of secondary importance. Clarity at top level is required about a number of factors:
o Objective: What is driving the decision pro LDI? Will the pension fund keep running or is a buyout at some stage envisaged?
o Regulation: What approach to LDI is most suitable for compliance with new solvency and accounting rules?
o Sponsor covenant: How much volatility in the funding ratio or contributions is feasible for the sponsor?
o Starting point: It makes a big difference, where you start in terms of your funding position, age profile, open or closed fund, etc.
o Risk taking: Is the risk budget for LDI well defined? Do you need to find additional returns from assets?
o Governance: Is there structure and process in place to manage a major change internally and externally?
Contrary to what the term LDI implies, discussions with investment experts often focus 90% on assets and 10% liabilities. Beware of spurious accuracy. There are number of pitfalls in this process:
o Pension liabilities are often reported as fixed number but they are not. They are a fuzzy, moving projection. Actuaries, accountants, insurers, regulators: they are all valuing them in different ways, changing all the times;
o How is your estimate of liabilities affected by changes in the main risk factors such as interest rates, inflation and longevity?
o Over time, pension payments will be (possibly dramatically) affected, eg, by legal and tax changes, takeovers, benefit increases, and other unpredictable factors.
A number of answers need to be found for practical questions. If liabilities are the new benchmark, how do you nail it down with your fund manager? Some people prefer absolute benchmarks like inflation plus x% or LIBOR plus x%. More complex proxies are proposed, eg, a mix of fixed interest and inflation-linked indices of various maturities. How good an approximation will it be? Whose responsibility is the ‘mismatch; between the liability benchmark and real liabilities?
In substance, there is not so much new with LDI, neither in theory nor in practice. A ‘liability-replicating' or ‘cash-flow-matching' portfolio (ideally with a series of zero-bonds) used to be starting point of insurance and pension investing from the outset. Since exact matching is not always possible, practical, or affordable, other ‘immunisation' strategies were developed long time ago. Bond duration-matching, for example, aims at neutralising the risk of asset-liability mismatch arising from changing interest rates.
Unfortunately, some basic principles had been forgotten over the years - in some places more than others. An example is the unfortunate use of market-based bond indices with an average duration of five years or so as benchmarks for long pension liabilities - a very expensive mismatch for many pension funds in the recent years.
To mitigate the interest rate risk, a number of pension funds are currently extending the duration of bonds (and perhaps also increasing their weighting). In a similar way, a better hedge of the inflation risk is sought through the addition of inflation-linked bonds. Longevity bonds would be the logical third element in the management of the trinity of key risks, but they are not yet available on the financial markets.
In a nutshell, pension funds can go a long way with traditional asset classes in order to achieve better liability matching. It never was and is only about bonds: property rents, high-yielding infrastructure or other companies, and other regular income streams have a role to play. It is surprising, how little emphasis this field has received!
hese days, LDI increasingly makes use of derivatives, as their availability and liquidity has enormously improved. Better cash flow matching and/or duration management can be achieved with the help of interest rate swaps. Inflation swaps are useful when future pension payments are inflation-linked. Therefore, a swap overlay has become popular as a second route in LDI. However, you will need cash to pay for the swaps: the collateral of any derivatives overlay needs to be carefully managed.
Pension plans in several countries have to deal with additional guarantees, floors and caps, triggers, recovery periods and similar regulatory constraints. Derivatives are useful in reshaping the risk-return and payout characteristics of your assets. Therefore, some providers offer even more sophisticated programmes, including options, swaptions and futures alongside swaps.
A third major route is now offered by fund managers in the form of LDI funds of some sort. Such pooled products may consist, eg, of a series of ‘duration buckets' or sub-funds of pooled swaps of different maturities. Pension funds may mix them in order to approximate their own cash-flow requirements.
The three major approaches can, of course, be combined and varied.
For those in charge of pension plan investment, there are a number of decisions to be taken:
o Full or partial matching? Some pension plans use LDI only for a part of their funds (eg, a pensioners sub-fund);
o Pure or mixed strategies? The longer the time horizon, the less of a market there is for underlying instruments, and things become expensive. Some plans opt for a combination of full matching (nearer term), duration matching and risky assets longer term;
o Direct or indirect? Do you have appropriate resources for dealing derivatives directly with an investment bank counterparty? Or do you ask a suitable asset manager to manage a swaps overlay?
o Segregated or pooled? A tailored solution has the obvious advantages, but may not always be feasible, in particular for smaller funds with lower resources;
o Extra return? Some providers offer ‘LDI plus x%' solutions that offer additional returns over time. What are the sources of additional returns (eg, alpha or beta)?
o Leverage. Leveraged LDI funds can free capital for other (return-seeking) investments, eg, equity futures. Clear about the risks?
o Costs. Immediate and future costs? How flexible is your structure if you wish to adjust it to new circumstances in future?
o Timing. Will LDI be introduced at once or will it be phased in? What are the benefits and risks of waiting?
LDI may allow a clearer management of certain risks, but may also introduce new risks and uncertainties. One challenge is the control and audit of a new type of transactions. Another one is the management of the counterparty risk that comes with derivatives. Furthermore, are performance/risk measurement and reporting able to deal with the new strategies?
Last but not least, there is an issue of incentives. How active or passive is the LDI approach proposed? What responsibilities are exactly being transferred to the fund managers, investment banks and consultants involved? What is an appropriate fee structure for that?
Georg Inderst is an independent consultant based in London.