So in a continuation of the festive seasonal spirit, here are some quiz questions for pension fund trustees:LDI stands for ‘Liability driven investing’, which was the hot topic of 2004. What acronym best describes the preceding regime? Was it:
PCMI (Pension Cost Minimization Investing)?
IMDI (Investment Manager Driven Investing)?
or PGI (Peer Group Investing)?
And here’s another:
UK pension funds have moved huge amounts of money from equities to bonds in recent years. Which UK individual or group of individuals has done most to encourage this move? Was it:
Chancellor Gordon Brown – by simultaneously hitting the return from equities, and giving the Bank of England its freedom to keep inflation (and therefore interest rates) down?
Elroy Dimson, Paul Marsh and Mike Staunton through their book ‘The triumph of the optimists’, which convinced us all that the equity risk premium is nowhere near as high as we thought? Or Paul Myners, John Ralfe and Cliff Speed for telling us all we really should operate on a risk budget, and that squandering the whole budget on one country’s equity market is plain gambling?
I confess; they were trick questions. Obviously the answer to both is ‘All of the above, and then more’. After all, when we began the current decade most pension funds were enjoying a surplus of assets over liabilities (as they were then estimated). Now we are half way through the decade, and most funds are in deficit. Unsurprisingly, we are rapidly adopting a new regime. But it is not just the bear market in equities.
Pushing strongly in the same direction are the regulatory changes facing pension funds in 2005. In particular, the minimum funding requirement is to be abandoned after just 10 short years. It is to be replaced by a scheme-specific funding standard, and most importantly, a published transparency statement that includes explanations of what the volatility of assets and liabilities could do to the funding level, and a justification as to why this is acceptable to the trustees. And when a regime shift occurs, that means winners and losers. Let’s ask ourselves who is likely to fall into which category.
Firstly, we should define LDI.
LDI is an approach to strategic asset allocation that determines asset class weights primarily on the basis of their correlation with the change in value of the scheme’s liabilities. Since pensions in payment in the UK are subject to limited price indexation, that part of the liabilities is most closely matched by some combination of indexed-linked bonds and conventional bonds.
Secondly, under FRS17, pension liabilities are to be calculated using the AA corporate bond yield. An experienced commentator, Kurt Winkelman of Goldman Sachs in New York, cautions that pension liabilities are ‘bond-like’ in nature, but they are not bonds. The fit is far from perfect, but one thing we can be sure of – equities they ain’t.
LDI is a move towards lowering surplus volatility, and therefore contribution volatility, by better matching, though not perfectly matching, of assets and liabilities. The problem is that there are very few securities that come close to matching the liabilities!
This is not just a UK issue. Dutch pension funds have assets of about E425bn, but the whole of the euro area’s government bond market is worth only E470bn. Worse still, the fixed income markets that do exist in size are predominantly in the wrong currency, dollars and yen, have the wrong duration, below 20 years, and lack the index-linking feature. The announcement that the UK’s Debt Office is considering issuance of 50-year gilts is unlikely to make much of a dent, and will be snapped up if and when it comes by the annuity providers.
Once, an absence of suitable assets would have been a problem, but not nowadays. Step forward the investment banks and their derivative desks. The idea is that a
pension fund can invest its physical assets in liquid and deep markets, with a wide choice of instruments and securities, and adjust the
economic exposure through swaps, to match better the duration, currency, and inflation-linked character of the liabilities.
George Henshilwood of UK consultants Hymans Robertson provides a simple example in the diagram of an investor using an RPI swap to create a synthetic index-linked corporate bond out of a conventional corporate bond portfolio. As well as holding the bond portfolio, the investor pays away the cash flows from a conventional gilt portfolio and receives the cash flows from an index-linked gilt portfolio. The investor receives the default risk premium and is hedged against negative inflation surprises. The counterparty is
usually an investment bank, but not one acting in any fiduciary capacity, but simply on the other side of the trade. Other investors are already creating synthetic sterling corporate portfolios by investing in the much bigger US corporate bond market and eliminating the foreign exchange risk with a currency swap.
Interest in the swap market and its potential is increasing. In November last year, the UK’s NAPF issued a paper proposing that the interest rate swap curve be used instead of the AA corporate bond rate to discount pension liabilities under the new international accounting standards. It helps that the 20-year swap rate is not only lower than the AA rate; also the market is much more liquid, and spreads have narrowed towards government curves, as the market has grown.
Let’s assume that LDI grows. How will that affect the market for investment management services? We can only speculate, but there do seem to be some clear pointers already emerging. Firstly, the idea of using derivatives to re-shape economic exposures, overlaid on a physical portfolio invested elsewhere, will sound exotic to many trustees; many of them will equate ‘exotic’ with racy or speculative.
In practice however, the underlying or physical part of the liability-matching portfolio is more likely to be passively managed according to Professor David Blake of Birkbeck College, London in a 2003 article1. If alphas are sought, it will more probably be in an underlying portfolio of global bonds, than in equities, and with tightly controlled tracking errors. The reason is that in a swaps programme requires margining and roll-overs. Too much of a mismatch could exacerbate the costs and difficulty of running the overlay portfolio of derivatives.
Where a surplus exists, or for that part of the portfolio not targeted at matching liabilities, advisers will probably encourage more aggressive investing, in hedge funds,
private equity, or so-called focus funds. Bob Litterman, also of Goldman Sachs in New York, has described what he calls the active risk puzzle. Under the pre-LDI regime, pension funds had a bizarre risk budget. According to Litterman, the mismatch between assets and liabilities gave rise to a policy risk of typically between 8% and 12% annual standard deviation. Call it 10% on average. Active risk is the mismatch between the policy-determined benchmark portfolio, and the actual portfolio as managed by active managers.
Because pension funds usually run with several managers in each asset class, each unwilling to deviate too far from the benchmark, active risk for the whole fund is
typically between 0% and 2%. But we must square these figures to look at their net contribution. So active risk increases the variance from roughly 100 percentage points to 101! (102 plus 12 = 101).
In other words, although trustees fuss around changing managers and discussing their portfolios, only the policy risk really matters, and this arises almost solely from the preponderance of equity assets against bond-like liabilities.
Litterman thinks that this paradox will disappear in the US under the cold light of his own clear rationality.
On this side of the Atlantic, it is changing regulation that will drive us in the same direction. Litterman believes that alternative investing, particularly hedge funds and real estate, will continue to grow in popularity, along with passive approaches for the liability-driven core. The room for aggressive alpha-seeking investing will expand precisely because the policy risk will shrink by the adoption of more liability-driven cores. We might see policy risk reduce to say 6% per annum, and active risk increase to say 4% per annum of total assets, but because they are orthogonal risks, the total is still approximately 7% for a 30% reduction on the previous position (62 + 42 ² 72).
An emerging consensus is therefore that we continue to see a polarisation of assets into the passive vehicles and particularly into fixed income on the one hand, and hedge funds on the other. This will be at the expense of traditional equity approaches.
I see two problems with this consensus, however. Firstly, Litterman points out that trustees seem to be much more averse to active risk than they are to policy risk2. It is not clear why this differential risk aversion should disappear so quickly. We are not sure why it exists; is it a principal-agency issue? Or is it simply that trustees have much more confidence in the equity risk premium over the long term than they do in the promised fruits of active management?
The second problem with the consensus is this: where trustees retain long-only equity managers, it is thought more likely to be for focus funds holding a couple of dozen positions.
Already the big investment firms are preparing for this3, but they will face some difficult challenges. While they can (just about) manage to run several hundred client portfolios with tracking errors of 2% per annum and with tolerable dispersion of results between clients with the same mandates, they will quickly run out of capacity if the search for alpha, not beta, pushes them into focus funds with no more than two dozen holdings. It is not at all clear that the pendulum will start swinging back to the bigger houses and away from the boutiques.
As for trustees, they clearly face new demands on their competence to master some new terminology and must take some difficult asset allocation and manager selection decisions. Let’s hope they are up to it.
Mark Tapley is a former asset manager and currently a visiting fellow at the Cranfield School of Management

1UK Pension Fund Management After Myners: The Hunt for Correlation Begins, March 2003
2 See The Active Risk Puzzle by Bob Litterman, Goldman Sachs Asset Management, March 2004
3 See, for example, Investment Trends: Understanding Liability Led Investment, published as a supplement to Pensions World in May 2004