نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشیار مدیریت مالی دانشگاه تهران
2 عضو هیئت علمی دانشکده مدیریت دانشگاه تهران
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
This research is an attempt to detect and discover manipulated trades using mathematical algorithms eventually leading to discovery of price manipulation based on pump and dump model. This study utilizes send-level datasets which are not accessible for ordinary users and studies specifications and algorithms of suspicious stocks. Manipulations such as pump and dump during the manipulation period result in short-term oscillations of price thereby threatening the performance and reliability of the market. In this research the dataset manipulated companies between 2013 and 2016 were used and to assess the performance of the algorithm, an anticipatory neural network was used. Also, Econometric methods and data were utilized to ensure the reliability of the algorithm. Econometric methods concerning price manipulation such as stationary test, autocorrelation, kurtosis, skewness, Run test and duration dependence were used. The econometric test validate the results of the designed algorithm. The algorithm is successful in 91.5% of cases.
کلیدواژهها [English]