Academic journal article International Journal of Business

Measuring Fund Performance Using Multi-Factor Models: Evidence for the Portuguese Market

Academic journal article International Journal of Business

Measuring Fund Performance Using Multi-Factor Models: Evidence for the Portuguese Market

Article excerpt


Since the CAPM-based measures of Sharpe (1966) and Jensen (1968), performance evaluation measures have evolved considerably. One of the major improvements in the area has focused on models that control for some stock market anomalies. For instance, the Fama and French (1993, 1996) three-factor model controls for size and book-to-market effects, and the Carhart (1997) four-factor model adds a stock-momentum variable. Another line of research suggests conditional models of performance evaluation as being able to provide more reliable estimates of performance. Indeed, Ferson and Schadt (1996) and Christopherson, Ferson and Glassman (1998) have argued that it is important to take into account the public information available to investors at the time the returns were generated, allowing both performance and risk estimates to be time-varying.

In spite of the developments mentioned above, most of the recent empirical studies in the financial literature, which use these more sophisticated performance evaluation techniques, conclude that portfolio managers are not able to out-perform the market (e.g.: Blake, Lehmann and Timmermann, 2002; Otten and Bams, 2004; Ferson and Qian, 2004). Besides their poor overall performance, fund managers do not seem to be able to exhibit any market timing abilities as well (e.g.: Ferson and Qian, 2004; Byrne, Fletcher and Ntozi, 2006). This type of evidence is consistent with the earlier studies on fund performance evaluation and come in support of the efficient market hypothesis.

On the other hand, research on mutual fund performance has been somewhat geographically limited (Khorana, Servaes and Tufano, 2005), and essentially centred in the US and the UK markets. Recent developments in fund performance evaluation techniques have not yet been explored in many other markets, especially if we consider studies that employ the theoretically superior conditional multi-factor models. The studies of Kryzanowski, Lalancette and To (1997) in the Canadian market, Otten and Bams (2002) in some European markets and Bauer, Otten and Rad (2006) in the New Zealand market are some of the few exceptions.

Furthermore, given the well known "data mining" problem, one of the attractive features of our study is that it focuses on a largely unexplored European fund market: the Portuguese. In fact, we are aware of only two published studies that focus on the performance evaluation of Portuguese equity funds. Cortez and Silva (2002) evaluate the overall performance of a sample of 12 National stock funds, during the period April 1994--March 1998, using two single-factor models: an unconditional model and a conditional model with time-varying betas. Their results are supportive of the conditional framework, with fund managers presenting, on average, neutral performances. Romacho and Cortez (2006) focus on the selectivity and timing abilities of 21 Portuguese-based funds, during the period January 1996--December 2001. Using unconditional versions of the Henriksson and Merton (1981) model, their results show that managers do not exhibit selectivity or timing abilities and present evidence of a strong negative correlation between these two components of overall performance.

We investigate the performance of a survivorship bias-controlled sample of Portuguese-based mutual funds, with different investment objectives (National and European Union stocks) and during a longer and more recent period (January 2000 through December 2007). We contribute to the international mutual fund performance literature by providing a comprehensive analysis of the performance of Portuguese mutual funds investing locally as well as in the European market. Besides using more robust performance evaluation techniques (conditional multi-factor models) to analyse overall performance as well as timing and selectivity abilities of fund managers, the analysis is complemented with robustness tests to check for possible effects that may arise from survivorship bias, management fees and spurious regressions. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.