Risk, Return and Market Timing : A Conditional Performance Benchmarking Model

The traditional approach of portfolio benchmarking was developed by Markowitz (1952) with his Mean-Variance (M-V) model. The emergence of Capital Asset Pricing Model (CAPM), however, led to the introduction of indices provided by Sharpe (1966), Treynor (1965) and Jensen (1968). Nevertheless, these models suffer from two major shortcomings — one is the benchmark selection process and the other is the use of linear factor models such as CAPM. Moreover, the asset-pricing models assume constant beta coefficient over the sample period under study. But if fund managers adopt active fund management strategy known as market timing (Treynor and Mazuy, 1966; and Henriksson and Merton, 1981), an estimation bias will occur in case of benchmark models which in turn will make computed measures unreliable. The present study extends the portfolio evaluation framework provided by Sharpe (1966) and Treynor (1965) by including the parameter of market timing with the help of a non-parametric framework. However, the present study departs from the conventional point estimation based Data Envelopment Analysis (DEA) framework. Robust bootstrap based DEA has been used in the present exercise to evaluate the conditional performance of 51 mutual fund schemes operating in India. In the second stage, the linkage between fund efficiency and market timing is explored through truncated regression analysis. The regression result does not show a statistically significant association between bootstrap efficiency scores and market timing indicator