Bootstrap Methods with application in Econometrics and Finance*

This paper studies both the traditional and the state of art aspects of Bootstrap Resampling Procedures. The theoretical asymptotic performances of such procedures are explored by an intuitive way of Edgeworth Expansion. Several refinements of standard bootstrap methods are advanced by a brief introduction, with suggestions for further research frontiers. Further more, empirical experiments are carried out in terms of bootstrap confidence intervals and hypothesis tests to evaluation the managerial performance of several representative funds in the financial market of People’s Republic China. Another contribution of this paper is that it develops an integrated numerical method to analyze the overall performance of each fund instead of relying on the traditional four types of performance measures respectively. Concluding remarks are forwarded in an attempt to explicit that any experimental analysis should be carried out with great caution because a minor intrusion of theoretical assumptions may result in catastrophic failure in understanding the intrinsic properties of the data.