When the Dutch pensions market adopted the STAR legislation in November 1997 - heralding the measurement of supplementary scheme performance through the now infamous ‘Z-Score’ system – talk was of the knock-on effects such testing could have on investment.
STAR had been placed on the political agenda by insurance companies balking at what they felt was unfair competition and corporates within industry-wide funds concerned about mandatory affiliation to what they saw as ‘costly’ industry-wide plans.
Upon its acceptance by the OPF association for Dutch company schemes, VB association for Dutch industry schemes and the insurers union (Verbond van Verzekeraars), noises were heard in various quarters that STAR would oblige funds to invest conservatively or go passive as a result of being in the performance headlines.
Others though argued that STAR was not stringent enough and that performance really had to bottom out dramatically before the Z-Score level was broken.
The truth appears to be a blend of the two.
Recommendations for STAR from the Dutch labour foundation (Stichting van de Arbeid) began in June 1995. The intention, the foundation noted, was to improve the market based operations of the country’s supplementary pensions system.
In essence STAR stipulates that companies applying for exemption from the relevant industry pension fund (bedriifstak pensioenfonds -BPF), can do so if there are objective grounds for concluding that over several years the whole of the BPF’s investment performance, administration costs, and quality of service remain substantially below par when compared to other collective pension schemes.
A BPF is not immediately penalised for any short-term performance dip, but allowed the opportunity to take measures to improve performance over the measurement period.
The structure of the regulation is a performance test. This reflects the fixed commitment structure of the fund as well as the fund’s asset position on the basis of a norm portfolio.
This is an annually defined preferred portfolio composition based on an ALM analysis, current portfolio composition and anticipated cash flow. The review also tests whether a norm portfolio has been defined in a careful and prudent manner.
Subsequently, the scheme must explain why the investment approach taken would lead to stable and low contributions for employer and employees.
Actual yield from the fund portfolio is then compared with the projected yield of a ‘norm’ portfolio over a period of five years.
The initial implementation period for STAR, however, was three years (98-2000), this switched to four this year (98-2001) until reaching the five year level.
All the above criteria are established within a long-term investment policy.
The policy’s actual implementation is then evaluated in order to see whether performance has been substantially lower than could objectively have been expected on the basis of market indices.
Calculation is made on the permissible deviation for an entire portfolio and determined on each fund individually based on the weighting of fixed-income and other investments (to be optimised on the basis of market returns). Internal administration costs are also taken into account.
Yet, while the approach is in line with internationally applied standards for the publication and measurement of investment performance, STAR has thrown up issues.
Joos Nijtmans at the VB association for industry-wide pension funds says one observation by pension funds is that the measurement period of five years is too short.
“ If you have a bad year or two then it is very hard to compensate for this by the fifth year and it is especially hard at the moment with the four year level. “At one of our committee meetings in December there will be presentations and discussions on the Z-Score to discuss these issues. One fund, for example, will be looking at the issue of tracking error.”
Furthermore, Nijtmans says that in essence if a fund has two bad performance years it is almost obliged to strongly consider a passive approach to investment to ensure it doesn’t dip below the Z-Score level.
“ It’s not a trend that all pension funds are following, but most pension funds are coming to us with this theoretical observation and now we are examining whether it is correct.”
However, Nijtmans says he believes it will be hard to suggest a longer time span for the Z-Score test: “ If we talked about seven years then the public would not be for it I think, because you should be able to make the performance good by then. “I think we will rather be concentrating on changes in the calculation methods”.
Nevertheless, Nijtmans says he also fears that with no funds as yet failing the Z-Score test, public demand might in fact be in favour of the implementation of a tougher regime.
John van Markwyk, director of investment at Relan Pensioenen, which manages in excess of NLG20bn (e9.1bn) for Dutch industry-wide pension scheme assets - including those of the tobacco, agriculture, butchers and retail schemes, says that not every fund is unduly disturbed by the Z-Score.
“ We have some special clients in that the Z-Score is the not the only thing in life for them. They tend to look to the longer game and not invest in the index just because of the risk of a bad Z-Score.”
Markwyk believes this is the important because Z-Score preoccupation can make pension funds take their eye off the ball. “One year the performance is very good, the next year it might be very marginal, but if you look to the Z-Score you only look to the relative risk of investing against the index.
“ That’s not the case with a big part of our clients which look at absolute risks, so that if the IT sector goes up that this is a risk for the index if you don’t buy it, but when they think it is overvalued they don’t let us buy it.”
Yet Van Markwyk says he thinks the Z-Score is a good instrument: “ It forces you to be more disciplined in your investing and to think better about short term action.
“ It does have certain limitations at the moment though, which we shall have to discuss with the Ministry of Social Affairs in the coming years.”
These, he says, include the fact that pension funds are now looking solely at the short-term: “This is very bad. Another important issue is that here has to be competition. “The market has to do its thing, but the opposite is in fact happening with everyone investing in a defensive fashion”.
And while there has not been a Dutch fund with an insufficient score so far, Van Markwyk notes there could be problems ahead.
“There is a certain risk, because one of the issue is that the sectors in the formula for risk limitation are taken from, I believe, the period 1990-1995, which was a period of low market risk. So the current Z-Score has risks in it which are based on an old period.”
He also suggests a longer time frame: “ An improvement could be to have a longer time period. This wouldn’t be 10 to 15 years or anything because that is too long to wait for insurers and the government.
“ It is four years at the moment but that is still short term when you look at the market risk over the last two to three year.”

Topics