Intelligent System for the Detection of Insider Trading in Indian Stock Market

Abstract

Insider trading is a pervasive stock market malpractice that has existed since the inception of the security market. Insider trading is notoriously difficult for regulators worldwide to crack down on. India has a dismal track record when it comes to prosecuting insider traders. In the last three decades of Sebi's existence, there hasn't been a single conviction for insider trading. In this study, we mainly inspectthe features of insider trading by examiningkey indicators during the time length before the release of insider information. In our investigation, we proposed a methodology for detection of insider trading in Indian stock market. To start with, the insider trading cases that happened in the Indian financial exchange we collected corporate filing data from NSE website for each company of NIFTY 50, which has different columns related to price, action and person or organization doing that action from 1stJanuary to a day prior to the publicationof financial results of December quarter offiscal year 2020-21. On doing as such, we have seen that enormous exchange have been done prior to publicationof financial results in some companies, which can be suspected as insider trading. At that point, themachine learning algorithms were utilized for preparing and for foreseeing Insider trading. Then, the algorithms were used for training and for predicting insider trading. Finally, their performance wasmeasured, compared and accuracy was calculated.Experiments revealed that the recommended method successfully achieved the best accuracy. This could be amazingly helpful for detecting insider trading in future, not only in Indian stock market, but also in other stock exchanges. The proposed approach and results in this examination is of incredible importance for market controllers to improve their oversight proficiency and precision on insider trading..