Presenting a Prediction Model for Stock Targeting Through Block Trades

The logistic regression technique was used to propose a model for predicting stock targeting through block trades. Those characteristics related to the likelihood of companies turning into a target for block trades were investigated. For this purpose gathered data on 117 Tehran Stock Exchange members, whose target of block trades with a trades volume of over 5% and 117 companies that did not target trading, was selected during the period of 2009-2017, using Logit method And probit were studied. The results showed that financial leverage and the change in assets negatively influenced the (block trades) commercialization frequency of the studied companies. It was also found that companies having a greater free cash flow, a higher market share, and a more distributed ownership, as well as companies with state organizations as principal shareholders, were more likely to be turned into commercial blocks. In addition, a comparison was made between the proposed logistic regression model and other well-known prediction models, namely, the artificial neural network and the fuzzy neural network models. The obtained results showed that the fuzzy neural network approach provided a more accurate prediction in terms of stock targeting than other techniques.