For most consultants in Europe today, there is little use beginning a conversation on asset allocation by talking about pure alpha. Liabilities are the ship and alpha merely the wind which takes it to the longed-for destination.
Consultants are spending more time explaining the nature of that ship to the client-owners so they understand what is necessary, what is possible, and what risks lie on the horizon. Finding the favourable breeze of alpha is secondary to avoiding squalls, rocks and icebergs.
For Franz Schumacher of PPCmetrics, fiduciaries need to be aware that some risks drive assets and liabilities. Risks are like floating ice because the total picture is not evident from regulatory or traditional actuarial measurements. The ship could be headed for danger while seemingly steady in the water, as all around asset prices undulate with market rates.
This point is pertinent to the Swiss market where the technical discount rate for calculating liabilities in defined benefit (DB) plans is a constant 4%. Schumacher would like trustees to concentrate more on the real liabilities. “Using market rates instead of the technical rate has an important impact on the assessment of the plan’s situation and investment strategy,” he says. In Switzerland the long-term interest rates are currently about 3.25%, much lower than the technical rate. This significantly affects the real situation of defined benefit plans.
Another difficulty when considering asset allocation is that the rule for deciding the discount rate for defined contribution plans in Switzerland (yes, there is a minimum required return which makes these plans de facto DB plans) is not clearly stated. This makes it difficult for such plans to select suitable investment strategies and
Once trustees understand the interaction between interest rates and liabilities, Schumacher sees the next step as defining a risk-minimising bond portfolio as the true benchmark to beat. Bonds, mortgages and partly property can then be used to match the liabilities. They can be held because their risk premia can be absorbed by the fund. Around that basic structure, alpha and beta risk can be added in order to achieve higher returns. “Of course, any risk-taking should occur only when there is a chance of gaining an appropriate risk premium,” concludes Schumacher.
In nearby St Gallen, Michael Brandenberger, chief operating officer of Complementa, has a different philosophy. He is not so critical of the technical interest rate because he sees it as a useful standard for pension funds.
When it comes to determining asset classes’ medium- to long-term returns, Complementa uses reverse optimisation methods, which derive risk premia from the world’s entire portfolio based on market capitalisation. To assess asset allocation risks, stress tests are used to see what the fund would lose in extreme circumstances, although final allocations focus on minimum risks. All this is communicated ‘live’ in workshops, not merely by report, according to Brandenberger.
One change he has noticed among clients generally is appetite not only for more funds of hedge funds, but slightly more aggressive risk profiles at some organisations that have had long-term hedge fund expertise. “Half of our clients use hedge funds. The average allocation is 5-8%, but that is now going up to 10% in some pension funds and one client goes up to 15%.”
For Karel Stroobants, president of the board of Antwerp-based Akkermanns Stroobants, the raison d’etre of a pension fund is to pay pensions and so an asset allocation strategy has to reflect liabilities. “I came from a banking and insurance background. When I started in pensions 15 years ago was the first time I heard of index benchmarks and I never understood them,” he says. Now liability-matching is in fashion but Stroobants does not see this as a bad thing.
It is certainly better than concentrating just on asset returns, which he describes as merely the techniques to arrive at paying pensions. He gives the example of a core/satellite equity approach: “This is neither fish nor fowl to me. A core portfolio of liability-matching assets which stabilises duration I understand, but people mistake index-tracking as being riskless.”
On the other hand, he counsels against full immunisation, that is buying an entire portfolio of bonds. For him, this only works if you are prepared to lose the 2-3% extra that riskier asset classes like equities can provide, which requires a generous sponsor able to pay the extra, or convincing social partners of the necessity of lower benefits. In conclusion there is a mix of bonds, equities and real estate – the latter typically accessed via products such as investment funds, typically Sicavs, or real estate investment trusts (Reits) unless the fund is of a size to run its own property portfolio.
Within the major asset classes, however, Stroobants believes there is still a lot of work for pension funds to do on risk diversification, by going into small and mid-caps, for example. When running the doctors and dentists’ fund in Belgium, he oversaw a strategy where six active equity managers exhibited in aggregate lower risk than the aggregate index, measured always by value-at-risk. “If anything I would look for a large tracking error from an active manager; then I know it is not closet-indexing.”
He advocates value-at-risk as a measurement device because it is easy for trustee boards to understand and calculates risk on an ongoing basis, thereby avoiding the drift away from liability-matching from which some funds with fixed asset allocation suffer. The dangers of value-at-risk, Stroobants says, are that the figures are really based on the past and it is so easy to believe that they can be trusted completely and so one’s critical faculty falls away. He compares its role to the dashboard in a car. “I don’t understand anything about the engine, but the speedometer tells me if I am doing 60 kilometres an hour and if I am getting faster or slower.”
Stroobants’ advice on asset allocation is to avoid tinkering based on fashion, especially techniques in vogue in asset management. Hedge funds for example, he sees merely as techniques within the major asset classes: “You can have value, growth or long/short equity, for example. Or arbitrage as a technique within fixed income.” He sees the promised returns more as a mirage than the Holy Grail. “The black boxes are too black and the fees too high.”
When asked what the worst thing a fund can do to its asset allocation is, Stroobants responds in two words: “Change it.”
Frits Bosch, head of Nuenen-based Bureau Bosch, is wary of the rush to adopt complicated liability-matching strategies. He says it is a trend because of the new accounting standards which come into effect next January 2006 in the Netherlands. The annual Bureau Bosch survey on trends in institutional investment management in the Netherlands notes that the trend towards liability investing leads to “more attention for hedging liability risk; cash-flow matching and duration hedging; reduced equity allocations; additional asset classes to improve diversification; structured forms of active management including derivatives and synthetic products.” These are on top of a growing risk management, including use of value-at-risk.
But Bosch believes that the kinds of derivative instruments these techniques require, notably swaps and swaptions, are not necessarily transparent and so little understood by traditional advisers, consultants and possibly even the new pensions regulator. “We are entering terra incognita,” he says. He wants investment banks and asset managers which execute liability-matching strategies to be independent of each other. He thinks this is one means for pension fund clients of ensuring proper
The Bureau Bosch survey elsewhere finds that equities holdings among Dutch funds are being reduced while exposure to longer-duration bonds and alternatives is on the rise.
“Within the search for alpha we see a trend to give asset managers more room to move within a total risk budget. Sometimes even the traditional guidelines are abolished, as long as the manager makes more return and total risk is kept within the boundaries,” the report notes.