Financial Statements Manipulations Using Beneish Model and Probit Regression Model. A Case of Banking Sector in Kenya

The main objective of the study was to establish whether the banks in Kenya were involved in financial statement manipulations. The study involved all the banks registered and operating in Kenya and whose financial statements are published for public consumption. Beneish five-variable model was first used to categorize the banks as likely non-manipulators and likely manipulators. The probit regression model was used to determine non-manipulators and manipulators based on the averages derived from non-manipulators using Beneish five-variable model. The results obtained showed that 78.8% of all banks were not involved in financial statement manipulations while 21.2% were involved in financial statement manipulations. The study concluded that some banks that were involved in financial statement manipulations. The study recommended that both internal and control auditors should compute individual indices to determine whether the preparers of financial statements were involved in manipulations. Further it was recommended that the organizations should enhance and strengthen the ICS to seal the loopholes utilized in financial statement manipulations