Abstract
The estimation of the parameters of logit model is mostly performed with method of maximum likelihood. However, the classical maximum likelihood estimators are biased and inefficient in appearance of small samples. The jackknife maximum likelihood estimator improves the above problems but still includes serious disadvantages. In this article, the Bootstrap Maximum Likelihood Estimator is developed as an alternative advanced method for reducing the bias and correcting the troubles with inefficiency and nonnormality. The importance of the method is shown through its application on data of Greek mergers and acquisitions.
Acknowledgement
The author is granted from the Alexander Onassis Foundation.
Notes
1 23 acquired and 84 nonacquired companies.