Designing a Model for Predicting Valuation of Block Trade Transactions with a Focus on GRU Artificial Neural Network in the Industry

Abstract

Predicting the valuation of blockTrade transaction allows the market to evaluate control over companies in an efficient manner.In this research, by measuring the indicators affecting block transactions in three active industries in Tehran Stock Exchange During the period of 1390 to the end of 1400, on a daily basis with utilization a deep learning neural network, specifically the GRU model. The study focused on industries with a significant number of market participants, namely basic metals (steel), automotive and parts manufacturing (Khodro), and pharmaceuticals (Darou). The results of the hypothesis testing indicate that, at three industry level, Nine variables significantly affect blockTrade transaction valuation: stock returns, block size, trading volume, company size, price fluctuations, industry returns, market returns, institutional ownership, and market-to-book ratio It affects the valuation of block transactions. At the separate level of industries, the results of the effect of financial indicators on the valuation of blockTrade transaction in each industry are different from other industries, which indicates the independence of industries from each other. The findings of this research will help the managers of industries in the stock market and the users of the valuation indices of blockTrade transactions in better evaluation of pricing.

Keywords

• stock returns
• block trade transaction valuation
• industry
• deep learning neural networks
• gru model