|
The trm Report - May 2007
Liability-driven investment: a tight definition is elusive but this article considers a range from the extreme of buying out a scheme’s liabilities in full to a swap overlay designed to protect against interest rate and inflation risk in tandem with an aggressive absolute return mandate. Why? Many final salary schemes are now closed to new entrants or new accrual with the consequence that their liabilities are rapidly becoming crystallised. Even those schemes admitting new entrants tend to be more mature with the cost of new accrual a small part of the overall liability. Of course significant uncertainties remain on the timing and nature of benefits (when will members retire, how long will they live for, how much pension will be surrendered for lump sum and is it realistic to assume stingy commutation factors will be retained forever?). Nevertheless the future shape of benefit outgo has never been clearer and with this increasing certainty the opportunity to reduce risk becomes ever more attractive. The volatility of investment markets (the world changed for many between 2000 and 2002 when it was realised equity markets could go down and not recover immediately) has led both trustees and employers to look for alternatives to the conventional investment strategies with a view to reducing risks. Liability-driven strategies offer this, but come with a flip side of reduced participation in any equity upside that may emerge in future. When? There are trailblazers, wary followers and those who shake their head convinced all will end in tears (and are often right). Timing is everything. After three years of negative returns between 2000 and 2002 many groups of trustees remained in equities or made only token switches to bonds, in both cases recognising that such a switch would lock in the deficits that had emerged. In hushed conversations the “S-word” (seven letters – sounds like clerical garment) is increasingly mentioned. The time is therefore approaching when the resolve of the “switch, but not yet” brigade will be tested. Can they remember how in 2002/3 they vowed to hold firm and switch once their deficit had been eliminated or will they, forgetting good intentions, carry on like the gambler reluctant to cut a winning streak? Of course the appropriate time for a switch will vary according to circumstances. Many others will conclude liability-driven investment is simply not for them. How? The hardest bit! We consider three types of liability-driven investment. These include buyout, bond solutions and swap-based solutions in tandem with absolute return strategies. As in so many areas in pensions there is a paradox. The strongest employers and/or those cases where the financial significance of the pension scheme relative to the employer is smallest are those most able to adopt and least in need of a solution which will eliminate or significantly reduce risk. That said even the mightiest have toppled and even here trustees should be open to exploring all ideas that reduce risk. Buyout The most fundamental liability-driven solution is full buyout in association with winding up the scheme. This will be appropriate for schemes where the intention is to end final salary provision altogether. New entrants to the buyout market abound with an expectation that buyout costs will eventually reduce. However as yet it is not clear how these savings will be achieved. Will the new providers be pioneering innovative and sophisticated investment strategies or have they simply entered a market which had previously over charged its customers (and in the latter case how can anyone possibly know)? Traditionally, buyout has been considered expensive (although what this means merits some consideration). At one level it ought to be expensive – the insurance company only has one opportunity to get it right as, unlike trustees, they do not have the opportunity to return, cap in hand, to ask for more money. So not only do they have to cover the expected benefit costs resulting from a conservative investment strategy imposed by regulation (and common sense) but also manage the risk of adverse experience while meeting expenses, profit margins and any future surprises regulators may have up their sleeves. Currently buyout costs arising in actuarial valuations can be up to double the ongoing liability (the latter historically being a perceived “realistic” estimate of the cost of providing benefits on the twin associated assumptions that a sponsoring employer will remain available to pick up the tab of adverse experience and that equities will outperform bonds over the long term). It is this gap (between the buyout and ongoing cost) which is most commonly used by as the measure of “expensive”. However the gap is likely to narrow over time. In particular the new scheme-specific funding legislation requires schemes are funded on the basis of prudent assumptions. The extent to which funding becomes more prudent remains to be seen. Early signs are that trustees are taking their responsibilities under the new legislation seriously and that a climate is developing in which employers are accepting prudent funding. This of itself will narrow the gap. There is also a concern that scheme funding may still not make sufficient allowance for increasing life expectancy. While actuaries have recognised the accelerated improvements observed amongst those born between the two World Wars the most commonly adopted table (the so called medium cohort) truncates the accelerated element of improvements at the year 2020 (a relatively short time frame when one considers that most schemes will still be paying benefits in 70 years time). The argument most commonly used against buyout (albeit often not consciously) is that the scheme will achieve investment returns greater than those underlying the buyout cost. If this proves to be the case then the effective cost of buyout (measured as the cost in excess of the ongoing cost) will get cheaper every year as the ongoing cost and the buyout cost converge (with buyout effected when the gap no longer looks expensive). Recognising the market’s reluctance to buy out in full, providers have sought partial solutions, such as buying out only existing pensioners. However this particular approach has not been attractive to trustees. They consider themselves able to manage the risk for existing pensioners but seek an inexpensive solution for deferred pensioners which of course (given the greater investment risk of longer duration liabilities) the market is unable to provide. Buyout of part of the liability may also cause practical problems if the scheme is ultimately wound up with an employer unable to fund all benefits to buyout level. Bonds Traditional theory holds that bonds are the best match for pensioner liabilities. Most schemes probably now hold them as a broad match against liabilities for existing pensioners (and sometimes for those approaching retirement). So bonds are a useful tool for reducing risk but not a complete answer. Given this inability to match fully trustees are reluctant to move entirely into bonds with the attaching loss of potential for upside. There is some scope for upside from bonds from a number of sources (for example moving to corporate bonds, bonds in other currencies or the management of duration). Nevertheless trustees tend to be unconvinced and set unambitious targets for their bond portfolios (often just matching the return on one of the longer-dated gilt indices). This reflects a view that outperformance from bonds requires running very hard in order to, if not stand still, at least not move very far, combined with a belief that such outperformance is more efficiently achieved through higher-risk assets. It may therefore be the case that the bonds solution remains appropriate for a scheme which is close to the finishing line. If it is concluded the bond solution is not appropriate now it would nevertheless be useful to identify circumstances (eg specific combination of rises in equity markets and/or increases in gilt yields) when such a switch might be justified. This information could be used to identify (and act quickly upon) opportunities for switching which might be implemented on a staggered basis. Swap-based and absolute returns The fact that few schemes are ready for a full bonds solution has left many looking for a liability-based solution with scope for meaningful additional returns. This desire has been fed by memories of investment managers appearing at trustee meetings proudly trumpeting large losses on the basis they had outperformed targets and other managers. Something felt wrong (although a little reflection would remind the trustees who had set those targets). The problem of negative returns could be addressed by implementing absolute return mandates (designed to give positive returns). However this still leaves schemes vulnerable to liability-related issues (eg changes in interest rates and or movements in inflation). This in turn has led to the adoption of absolute return mandates with swap overlays. The swaps and the absolute return mandate are constructed with the objective that the assets should move in line with the liabilities together with an outperformance target. The greater the outperformance target set to the absolute return manager, the greater is the risk required to achieve it and therefore the likelihood that it will not be achieved. Great care needs to be taken in setting an appropriate target. In addition the greater the extent of any swaps, the greater is the need to keep some of the absolute return manager’s assets available to act as security for the swap so inhibiting scope for outperformance. The absolute return managers will usually claim to be operating on a reduced risk budget (so higher returns for the same level of risk or the same returns at a lower level of risk). The returns will typically be relying on hedge funds and/or achieving returns in inefficient markets. The lower risk will derive largely from diversification. The obvious questions arise as to whether such managers can demonstrate a track record and/or whether as in so many other investment markets the scope for outperformance is sustainable as new entrants come onto the market. Another key decision is the extent of any swap. Should this be full or partial, at each year or only a few years, over the full duration of the liabilities or only the early years? These discussions will encourage full consideration of the risks the scheme faces (is this for example necessarily high inflation or is the risk of deflation greater?). These issues should be discussed openly between trustees and sponsors (the latter may have a particular perspective to bring and wish to consider the impact of the same risks on their core business). Consideration should be give as to how performance is measured (this is particularly the case where investment managers advise on swaps). Those embracing the concept of liability-driven investment should do so wholeheartedly and resist the temptation to measure against (and dream of) what might have been had they not done so. The most obvious benchmark is the liabilities of the scheme as they would be measured at any subsequent valuation (recognising, in the spirit of fairness, that investment managers cannot be expected to anticipate demographic experience and changes in valuation methods). Where? Kipling was writing before the invention of the internet and e-mail. It follows location was more important in his day and “where” is included here for completeness only. Nevertheless many of the new absolute return managers are based in the West End as opposed to the traditional managers in the City. This opens up the possibility of new restaurants to be explored. Who? We believe most schemes could benefit from actively considering the issues outlined above. Historically the bias has been towards larger schemes who probably have first call on the consultancy resources familiar with these solutions. However increasingly pooled solutions are being delivered which may justify smaller schemes or parts of larger schemes going down this route. John Branford |
 |
|
|||||||||
|
|||||||||||
