An imbalanced market
The UK has the most stringent pension indexation requirements of all the OECD countries. Joseph Mariathasan here points out the pressures that are pushing pension funds towards low-yielding assets such as inflation linkers
Real yields on 50 year index linked bonds in the UK were around 65bp in April and had hit a low point of 38bp in January 2007, as Kevin Adams, fixed income portfolio manager at Henderson Global Investors points out. He adds that, if a comparison is made with the nominal yields on gilts, this implies that the break-even inflation rate at which the yields on both types of bonds would be equal, "is around 3.7% for 30-year bonds, implying that the Bank of England will fail to keep inflation within their target range for the next 30 years".
Yet as David Hooker, fixed income manager at Insight comments: "Demand from pension funds for long dated index-linked sterling assets continues; will it go away? In the imminent future there is no reason to think so." As Adams adds: "The market does not behave in a fundamental way. Index-linked bonds were expensive some time ago and just became more expensive. The standard metrics do not work."
Determining the factors that explain the asset allocation behaviour of the UK's institutional defined benefit pension schemes has often been likened to unravelling a plate of spaghetti and it is no wonder that most companies have thrown in the towel and simply shifted to defined contribution schemes for new members.
The debate over liability driven investment (LDI) strategies reflects the complexities of an environment with multiple stakeholders and a slow tightening of legislation whose aim is to convert promises by pension sponsors into legal liabilities, without a detailed map of where such a road leads. This leads to the argument that pension liabilities are no different from other corporate liabilities and should be brought onto the balance sheet. Having seen immense growth in equity holdings by pension funds during the last half century, the arguments are now swinging back in favour of more bonds as pension funds seek benchmarks relevant to their future liabilities rather than arbitrary market average or peer group comparisons.
Matching liabilities with assets that behave in a similar manner has led to some advocating a complete replacement of equities by a combination of conventional and index-linked bonds. Indeed, leaving aside the uncertainties in mortality assumptions, it is possible to conceptualise such a portfolio of index linked gilts and conventional gilts that would exactly match a schemes liabilities. Whilst extra return could be added through the use of futures and swaps that would give exposure to other asset classes, this would be at the expense of increasing the volatility of the tracking error against the benchmark. In today's world, the derivatives markets have grown to such an extent that it is easier to match the liabilities with derivatives and them seek extra return by investing some of the cash in a diverse range of asset classes.
Economically, the two cases should be equivalent in markets with efficient arbitraging. LDI approaches using derivatives to reduce or eliminate liability volatility are therefore not costless, but entail locking in real yields that many would argue are ridiculously low, whilst generating extra return comes at the expense of introducing additional volatility. "In the last 6-9 months, the demand for LDI has abated," finds Adams, perhaps not surprisingly. "Firstly because equity markets have fallen so it becomes expensive to switch from equities and secondly, because pension funds and consultants are finding the price of inflation protection very high."
For the UK's pension funds, Dawid Konotey-Ahulu, a founding partner of Redington Partners warns: "There is a perfect storm brewing of pressures from the accounting standards boards, the regulatory authorities and the UK's Pension Protection Fund (PPF), all contributing to force pension funds to manage risk to much tighter levels." Moreover, he adds, the credit crunch is driving the authorities to increase the standard of risk management in order to avoid a similar fall-out to the current banking debacle. At the same time we are seeing serious shortages in the supply of inflation-linked debt assets. Yet like many concepts in the economic sphere, what may be reasonable and sensible for an individual entity, can collectively, give rise to sub-optimal behaviour for everyone. Arguably, this is very much the case for the UK's index-linked marketplace. Whilst pressures to hedge exposures to inflation have been strongly increasing over the last year, the credit crunch effectively dried up the origination of infrastructure or utility driven inflation needed to supply the inflation swap market.
"In the past, a corporate, PFI or utility with inflation linked cashflows would issue debt as a triple-A bond guaranteed by a triple-A rated monoline insurer," Alan James, head of inflation research at Barclays Capital explains. "The debt was modelled on the comparable index-linked gilt, although the structure did not matter as a lot was sold to banks acting as asset swappers who were able to use this to hedge inflation swaps written for pension schemes." Now that many monoline insurers are no longer rated AAA, the utilities are no longer able to issue AAA-rated inflation-linked debt. Indeed, as James points out: "A lot of the existing debt has lost value as it is not triple-A any more. So it has made it difficult to issue any new monoline wrapped paper given that no one knew what would happen to existing issues. Other than FSA, all the monolines had big problems so it is now rare to have a monoline insurer-wrapped paper."
Ultimately, as Hooker points out, utilities will have to fund eventually. "But the investors they have sold to in the past will not be the ones they will be able to sell to in the future. Previously, inflation bonds were sold to banks for asset swaps, now they may have to issue inflation-linked bonds direct to pension schemes. The bottleneck is that schemes prefer to receive inflation in swap form but accounting rules limit utilities ability to issue inflation in this format." Consequently, as James points out, "the UK inflation linked debt market is expensive not just because demand is so high, but because supply has been halved".
What should funds do? "For pension schemes, the risk attributable to long term interest rates and inflation is so great it needs to be reduced even at these high rates," Konotey-Ahulu argues. "The two biggest risks a pension fund faces are falling interest rates and rising inflation and whilst these may well be matched by equities over a 30-year timeframe, developments in regulation and accounting mean that pension scheme deficits need to be managed in real time. They don't have 30 years," he says. "Pension funds need long-term inflation matching assets; equities give them high returns but cannot be relied upon to match the behaviour of inflation linked liabilities. Only one asset can match long-term liabilities in lock-step and that is index-linked assets such as inflation linked bonds or swaps."
Given that inflation swaps are currently expensive, what schemes should do, he argues, depends on their specific risk tolerance - which includes considering the strength of the employer covenant. "Pension schemes to the greatest extent possible, should strip down the liability volatility and then diversify the assets to get the greatest return for the minimum risk." he explains. "Take a scheme 90% funded on an accounting basis. The two questions we typically ask are ‘how well funded, ultimately, do you want to be' and secondly ‘what is your time frame?' Usually, the company and the trustees want to get to the buyout level, ie 125% of accounting valuations over a 7-10 year time frame." Konotey-Ahulu goes on to suggest that the approach should then be: "Freeze liabilities strategically, but tactically, whilst now may not be the best time to implement it all, a scheme could hedge 10% immediately and plan to hedge more if breakeven inflation rates on index linked bonds fall to say 3.5% from 3.8%. The scheme should have a clear framework to govern its implementation levels - both for interest rates and for inflation."
Whilst it is possible to argue endlessly about strategy, there is another intriguing aspect about the UK's pension schemes. The inflation linking in a scheme usually has a 5% cap. Konotey-Ahulu explains: "With 3.5% UK inflation, the 5% cap is so far away that it does not really impact hedging decisions. But as inflation creeps above 4.0% and higher it will become increasingly relevant." So is there a world where if UK inflation is closer to 5%, the cap at 5% means that pension schemes may take the view that future higher inflation does not need to be hedged? The evidence from markets, such as the Netherlands, which do not have such a definite linking to inflation, and real yields on euro-zone inflation linked bonds are over 2%, suggests that this may be true. As Hooker explains: "Inflation bonds for a European fund are far more opportunistic than an LDI driven asset. If you have a nominal benchmark, do inflation bonds stack up? From a European perspective, index-linked gilts look expensive. They could get more expensive, but would you want to invest on that basis?"
"A skilful financial economist can make a sound case for investing pension assets in either equities, bonds or any asset class in between based on corporate finance theory," argue Moshe Milevsky and Mike Orszag, co-editors of the Journal of Pension Economics and Finance. The continued demand for long dated index-linked bonds at yields that no pension fund fiduciary would chose to invest suggests that the UK pension industry may have dug itself into a hole, that is just getting deeper, as it faces up to the inevitable consequence of much tighter risk management requirements.