Jan Baars and his employer, the PGGM pension fund, have been among the fiercest critics of the Z scores. In a series of articles in the Dutch press, PGGM has at times sounded almost apocalyptic, warning that the Industry-wide fund in the Netherlands is doomed. Complaints levelled against the Z scores come in two shapes- one criticising the approach wholesale and the other criticising the statistical formula.
One of PGGM’s complaints is that it is difficult to establish indices to measure private equity and other unlisted investments. In addition, the indices must somehow represent market value, a measure notoriously tricky for private equity and real estate. Baars says that they have had to use a long term interest rate plus one as a measure, something he describes as a very poor proxy.
This lack of indices is likely to make private equity seem extremely risky. The alternative is to get out of private equity but, as Baars points out there’s a good reason that you’re there in the first place. “The Z score is an incentive to stick to the index and might be a reason not to enter asset classes that cannot be indexed. This, in turn, might have consequences for the volatility of the markets.” PGGM claims that the five year testing period is statistically insufficient to produce a valid test. According to Baars, the Z score should be measured over 14 years in order to produce a significant sample. “At the moment, the power of this test is very small,” he says. Part of the reasons for these shortcomings are that the measures were determined by the politicians and not by mathematicians, something not lost on the VB, the association for Industry-wide funds: “It’s very hard to explain to the public and to the politicians that it’s better to have a longer period,” says Joos Nijtmans of VB.
The scores also apply the same tracking error opposed to a fund specific tracking measure. “Tracking error really is specific to a fund, it depends on the way you act and the way you do your allocation,” says Baars. The government could change the means of calculating it but one of the problems is that the funds would have to supply their own data that they would be able to manipulate themselves. “There would also be a lot of work involved.”
More serious than this though is that the fixed measure used at present was based on data from between 1992 and 1996, a period considerably less volatile than 1996 to date. Nijtmans fears that as a consequence of this outdated figure, the chance of a fund failing the Z score even when it performs well is likely to be higher than the 10% originally predicted.
Another shortcoming according to Baars is that you’re only allowed to change the portfolio on which the score is based once a year. If you carry out a study or want to change your portfolio, you can have to wait up to a year to do so. For example, last year PGGM completed an ALM study in April but had to wait until the beginning of this year to introduce the changes.
Sceptics believe that the Z scores could kill off the Industry-wide schemes if they are applied rigidly. “In the end there will be none left because the measure is like a rolling horizon. There’s a chance that you will fail when you shouldn’t do. In the long run everyone will fail,” says Baars. The real test will be the first of April next year when the number of failures becomes known. When this figure becomes clear, it will then be possible to determine the validity of the complaints.
Neither the VB, PGGM nor others unhappy with the measure are likely to see any changes for a while. The ministry for social affairs has said it will review the measure in 2005. “On some major points we think it is necessary that they change before 2005. We cannot wait until 2005, we need changes quicker than this,” says Nijtmans.

Topics