Best practice in governance has been shown to produce above-average investment returns, say Manuel Ammann and Christian Ehmann 

At a glance

• Cases of fraud and mismanagement have highlighted the need for fund governance.
• The authors surveyed Swiss funds, investigating governance architectures and investment performance from 2010 to 2012.
• The analysis showed that governance was positively related to the plans’ realised performance in the years covered.
• Funds with best-practice governance pertaining to target setting, investment strategy definition, and risk-management design were more likely to report above-average investment returns.

Severe financial losses as well as fraud and mismanagement in several occupational pension plans around the world have highlighted the importance of good pension fund governance. 

In Switzerland, cases of management misconduct, such as misappropriation of assets and cronyism in appointing asset managers, have threatened public trust in the second pillar over the past two decades. 

Switzerland has one of the largest occupational pension systems in Europe, insuring more than four million members. Total assets held by registered Swiss pension schemes were in excess of CHF800bn (€750bn), equivalent to 123% of the country’s estimated GDP for 2016, according to the Swiss Federal Statistical Office 2016. Consequently, pension fund managers and trustee boards are have wide-reaching responsibilities, not only towards their plan members. Governance weaknesses can have systemic implications. 

In theory, sound governance structures should be associated with better plan performance. Pension funds with efficient management organisations, structured investment processes and comprehensive risk management systems should be able to achieve superior investment performance for their members, for example, through well-qualified trustees, superior asset manager selection procedures, lower asset management costs, and better risk management systems. 

We conducted a survey of Swiss pension funds, investigating their governance architectures and investment performance for the years 2010-12. We focused on six different governance areas: organisational design, management incentives, target setting and investment strategy, investment processes, risk management, and managerial transparency.

Our proprietary dataset contains data from 139 funds, covering about 43% of total assets in the Swiss pension universe. Whereas previous studies have measured pension fund governance subjectively, for example by canvassing the opinions of senior managers, trustees or CEOs, our survey solely included assessment criteria that were based on objectively quantifiable facts. 

To assess the sample plans’ governance structures in the most objective way, we created a scoring system consisting of 60 governance factors and six individual sub-scores that cover the above-defined areas – the Swiss Pension Fund Governance Score (G-SCORE). This score metric drew from previous empirical research findings and, to some extent, best-practice considerations from academic literature. 

For each answer that was deemed to be theoretically desirable from a governance point of view, a pension fund received a point on the G-SCORE. Otherwise, a fund received no points. Consequently, a high score is associated with a more comprehensive or ‘theoretically desirable’ governance structure.

In a second step, using multivariate regression analyses, we relate the G-SCORE and its sub-scores to the surveyed plans’ average excess return, passive benchmark outperformance, individual benchmark outperformance, and Sharpe ratio. We thereby controlled for institutional factors such as fund size, legal form, plan type, or coverage ratio.  

Our analysis showed that pension fund governance was positively related to the plans’ realised performance of the years 2010-12. It found that the pension funds of the top G-SCORE quartile outperformed those of the bottom quartile by about 1% in terms of average excess returns and passive benchmark deviation. 

For our analysis, we drew on four different portfolio performance measures: average excess return, passive benchmark outperformance, individual benchmark outperformance, and  Sharpe ratio. Regardless of which particular measure we employed, we found both statistically and economically significant positive effects for the composite G-SCORE metric, as well as for the sub-scores that cover target setting and investment strategy and risk management. 

The investigation of those individual sub-score components furthermore showed that independent external experts that participated in investment strategy meetings are associated with superior performance. Risk management design was also found to be an important factor that was related to net investment performance. 

Interestingly, we also found that pension funds which attempted to improve their investment performance by engaging in tactical asset allocation exhibit, on average, lower investment performance than pension funds which do not adjust their portfolio allocations tactically.

To test whether governance structures were related to asset allocation decisions (and therefore indirectly to performance), we furthermore regressed the sample funds’ effective asset allocation weights on our governance scores and control variables. 

Our investigation indicated that asset allocation decisions were mainly independent of the prevailing governance architectures. The asset allocation weights of our sample pension portfolios were primarily related to institutional factors such as size, legal form, or the ratio of active plan members to pensioners. 

Our results have important implications for the industry as they show that the establishment of comprehensive and up-to-date governance structures might directly benefit plan members. Although our study did not give any indication of the direction of causality, it showed that pension funds with best-practice governance pertaining to target setting, investment strategy definition, and risk management design were more likely to report above-average investment returns, both on a raw and a risk-adjusted basis. 

Manuel Ammann is professor of finance and Christian Ehmann was a research assistant and PhD student, both at the University of St. Gallen. For the details of study, see Ammann & Ehmann (2017): Is Governance Related to Investment Performance and Asset Allocation? Empirical Evidence from Swiss Pension Funds, Swiss Journal of Economics and Statistics, forthcoming