|dc.description.abstract||Evaluation of the performance of investment managers is a much studied problem in finance, but results of studies the last thirty years have constantly been revised. Do managers of actively managed portfolios actually add some value? . I address this problem since there are indications they do not.
There has been a tremendous growth in savings put into mutual funds in Norway the last 10-15 years. The evaluation of the management of this wealth has for a long time been characterised as non-professional and non-informative. The last couple of years there have been improvements and this has gained both the investors and managers. Managers have been evaluated on a more justified basis, and investors have got access to more reliable reports of mutual funds performance. In this paper I highlight some of the controversies still existing, and I also try to formulate what can be expected of the managers.
By evaluating fund performance over a long period of time there are possibilities of finding out whether superior performance was caused by luck or skill. Norwegian media often just report return patterns for mutual funds the last six and twelve months, and that hardly tells us anything about persistence of performance. I will also discuss different criteria for plausible rankings.
This paper discusses the performance of 16 mutual funds in the period 1982-1998. I have empirically examined market timing and selectivity performance of a sample of mutual funds with a particular emphasis on the market timing ability of the investment managers. Calculation of results have been done by using Excel and PC-give. I have used a simple regression technique to separate stock selection ability from timing ability. This technique was first suggested by Merton (1981), but was more completely presented by Henriksson and Merton (1981). This model is still one of the most accepted one for the purpose of examining market timing and selectivity. Fund managers are evaluated by their abilities to forecast market changes. The forecasters in this model are not very sophisticated, and follow a qualitative approach to market timing. The forecaster predicts when stocks will outperform riskless securities and when riskless securities will outperform stocks, but does not predict the magnitude of the relative returns. It is assumed that managers will adjust the composition of their portfolios on the basis of forecasts, and they have succeeded in market timing if they have reduced risk in bear-markets and increased risk in bull-markets.
Using the model presented by Henriksson and Merton it is not necessary to have the investment managers market timing forecast themselves. The only data needed for the testing of timing ability is the time series of realised returns on the portfolio. There is a cost of not having the time series of forecasts, which is that the test procedure requires the assumption of a specific generating process for returns on securities. The generating process assumed is the Capital Asset Pricing Model (CAPM).
The results of my empirical tests show little evidence of superior market timing-or stock selection-abilities. Only one fund has significant positive market timing coefficient for the entire period. When splitting into two sub-periods, I found two significant positive market timing coefficients for the first period and none for the second.
My results show some evidence that managers might be better at predicting periods of extreme market movements than others, but these periods were also characterised by very bad stock selection abilities.
On average the funds investigated did have the same risk as the market index, but were underperforming the market index for the period 1982-1998. The funds managers were doing relatively better on average in the second period (1991-1998) compared to first period (1982-1990). In the first period the managers on average had negative return on the portfolios.
Computing the Jensen s alpha did show no superior management, and I can not reject the null hypothesis of zero abnormal return. Evidence about the collective performance of funds is relevant to the efficient market hypothesis. Significant evidence of superior forecasting skills would violate the Efficient Markets Hypothesis. Such violations, if found, would have far- reaching implications on the theory of finance especially with respect to optimal portfolio holdings of investors. My results show no superior management and, thence support the Efficient Market Hypothesis.||nor