Additional risk does not pay off over very long term, study shows
The Swiss pension funds with the highest share of so-called “risky” assets – including equities, private equity and other alternatives – also produced the lowest average performance, according to a controversial study by local consultancy PPCmetrics.
Among the schemes included in the study were the CHF14.6bn (€11.9bn) pension fund for the city of Zurich (PKZH), the CHF5.2bn multi-employer fund Profond and the CHF8.8bn Aargauische Pensionskasse (APK) for the canton of Aargau.
Hansruedi Scherer, a partner at PPCmetrics, told IPE the consultancy’s comparison had been well founded and based on data published by the Pensionskassen themselves between 2008 and September 2014.
However, he acknowledged that the data became skewed where Pensionskassen changed their strategies “massively” in recent years or followed ALM strategies “too strictly”.
The PKZH, for example, changed its approach to adjusting the equity allocation in early 2014, no longer linking it solely to the funding ratio but also to market trends.
Now the allocation is increased in rising markets and decreased in falling ones to allow the Pensionskasse to participate in bull markets regardless of its funding status.
For Scherer, the main problem is that the expected returns have dropped considerably from 5.67% in 2000 to 2.17% in 2014.
He said not all trustees had “accepted this changed world yet” and warned that “realistic expectations” on returns would remain relatively low in the coming years, based on several indicators.
He added that “the only certain thing about long-term prognoses is that they are uncertain”.
Similarly, Heinz Zimmermann, a professor at the University of Basel, said many Swiss pension funds performed their long-term calculations on a “wrong basis”.
For longer periods, analysts should use the root mean square (or quadratic mean) and not the arithmetic mean, he said at a recent PPCmetrics conference in Zurich.
He said there was no certainty, only “reduced uncertainty” in long-term prognoses with better calculations.
He also noted that, with cumulative returns, the volatility increases over time along with the return, and that, while mean reversion exists, it is “too weak to counteract this volatility over time”.
Scherer said it was difficult to take mean reversion into account, as it was almost impossible to say what the mean was.
According to Zimmermann, the strength of the prognosis is best over a horizon of 5-6 years.
The academic predicted an average equity return of 7% over several years, while Scherer said he was convinced the risk premium would remain positive over the long term.