• Selected
• Entire List

Abstract:
One of the most important developments in the field of finance during last forty years is the mutual fund performance evaluation technique. The traditional techniques use the unconditional moments of the returns. Such techniques cannot capture the time-varying element of expected return. As a consequence Ferson and Schadt (96) advocate a technique called conditional performance evaluation, designed to address this problem. This paper utilizes that technique on a sample of 89 Indian mutual fund schemes, over the period of 1999:1 to 2003:7. The broad based S&P CNX 500 is used in the study as benchmark. The study uses the lagged information variables - T-bill yield, dividend yields, term structure yield spread and a dummy for April-effect. The paper measures the performance with both unconditional and conditional form of - CAPM, Treynor-Mazuy model and Henriksson-Merton model. We examine the effect of incorporating lagged information variables into the evaluation of mutual fund managers' performance in Indian context. The result suggests that the use of conditioning lagged information variables improves the performance of the mutual fund schemes, causing the alphas to shift towards the right and reducing the number of negative timing coefficients. Tech rally of 1999 to 2001 is a major event in the history of Indian capital market. We have also incorporated the impact of the tech rally in the conditional models by introducing a dummy variable indicating the period of rally in tech stocks. We found that fund managers' performance as well as timing skill worsens with the inclusion of this dummy.

Conditional Performance, Mutual Funds

Abstract:
This paper calculates the Performance Change measure (PCM) developed by Grinblatt & Titman (Journal of Business, 1993, vol. 66, no. 1)for a sample of 50 Indian mutual funds over a period of 26 months. PCM as a measure has some advantages compared to the traditional measures, the most important one being - it is free from using a benchmark portfolio and consequently the resulting biases arising out of usage of such a portfolio. So by using PCM as a measure, this paper, without using any benchmark, attempts to asses whether the selected mutual funds are able to provide above-normal return on average - using no more information than what is available to the common investor. PCM has been calculated for one month, one quarter, and one year lag. And using PCM as a measure the study finds that though in the short term, the mutual funds were unable to generate above-normal return but on the average the combined PCM of all the mutual funds is significantly different from zero, which are in agreement with the original findings of Grinblatt & Titman, in this Indian context.

Performance Measurement, Mutual Funds, Benchmark

Abstract:
Do Indian mutual fund managers contribute to the better performance? In this paper, we address this question and measure the performance of Indian mutual funds in the conditional framework advocated by Ferson and Schadt (1996); Christopherson, Ferson and Glassman (1998). We find that when the beta of the fund is conditioned to lagged economic information variables, the fund performance does not change appreciably. However, when fund alpha is also controlled for these information variables, the fund performance on an average becomes significantly negative. This finding has enormous economic significance. The result shows that on an average the Indian mutual fund managers only captures the opportunities from the available economic information, they do not contribute anything beyond it. This paper utilizes a sample of 133 open-ended Indian mutual fund schemes, over the period of 1999:1 to 2003:7 for the study. The broad based S&P CNX 500 is used in the study as benchmark. The study uses the lagged information variables - interest rates, dividend yields, term structure yield spread and a dummy for April-effect.

This paper also examines the evidence of persistence in the performance of the Indian mutual funds. Our approach to measure performance persistence is based on cross-sectional regressions of future excess returns on a measure of past fund performance. Both unconditional and conditional measures of performance are used as measure of past fund performance. We use the methodology of Fama and MacBeth (1973) to test the hypothesis. We find the evidence that conditional measures of past fund performance predict the future fund returns significantly. In the present study, the horizons of future return are considered as 1, 2, 3, 6, 8, and 12 months and the past performance is measured for previous 18 months. Between the two different conditional measures of past performance, time-varying conditional alpha is found to be a better measure in indicating persistence in performance of Indian mutual funds. We estimate the standard errors t-ratios using the heteroskedasticity-consistent and autocorrelation-adjusted estimation techniques of White (1980), Hansen (1982) and Newey and West (1987).

Mutual fund, performance measure, conditional alpha

Abstract:
There are various conflicting evidences in existing literature about the predicting power of different volatility-forecasting models. There are evidences in favor of the simpler regression model (Dimson and Marsh 1990), as well as complex GARCH family models (Akgiray 1989, Pagan and Schwert 1989, and Brailsford and Faff 1996). In this paper, monthly volatility of market indices (Sensex & S&PCNX-Nifty) of Indian capital markets has been modeled using eight different univariate models. Out-of-sample forecasting performance of these models has been evaluated using different symmetric, as well as asymmetric loss functions. The GARCH (1,1) model has been found to be the over all superior model based on most of the symmetric loss functions though ARCH (9) has been found to be better than the other models for investors who are more concerned about under predictions than over predictions.

Forecasting, volatility, GARCH