Measuring and Evaluating Earnings Management Practices on the Commercial Banks Listed on the Egyptian Exchange According to the Loans Loss Provisions (LLPs): An Analytical Multi-Level Empirical Study

This research addresses Measurement and evaluation of earnings management in the fourteen Egyptian banks listed in the Egyptian exchange over the period 2007-2020. The main objective of this research is to formulate and test a model for measurement and evaluation of earnings management in these banks.
The Methodology of this research includes both inductive approach and deductive approach. Whereas the inductive approach is used to formulate and develop the hypotheses and the deductive approach is used to test these hypotheses. The research relied on the most acceptable model in this field, which is a two-step “Binary” model. 
The first step aims to determine discretionary estimations which are included in loans loss provision throughout the residual of regression function for the effects of three variables on loan loss provision. These three variables are specified in non-performing loan rate, change of non-performing loan rate and loans change rate. This step ends with formulating discretionary estimations for the loans loss provision.
Whereas the second step includes the effect of six pre-determined variables on discretionary estimations. These variables are as follows: Loans to the deposits rate, owners’ equity to the total assets rate, earnings before tax and provisions to total assets rate, Bank size, return on assets rate, and liabilities to the owners’ equity rate.
The main findings of this research are the absence of earnings management on the level of sectorial analysis. In contrast, the earnings management existed in nine banks out from the fourteen.
The testing of classification fitting has been relied on two models of discriminant analysis. Whereas the first one is a multivariate linear discriminant analysis with a discriminant power of 99.6%. Whereas the second is multivariate logistic discriminant analysis with a discriminant power of 90.3%, as well as identifying the most influential variables in classifying the banks into two groups, the first of which is free from earnings management, and the second is proven to manage earnings. These variables are: bank size, non-performing loans, and change in a non-performing loan rate.
The researcher recommended that more studies in the field of investigating earnings management practices in the financial institutions should be conducted, in particular, comparative studies; whether at the level of comparing models, or a comparison between financial institutions listed on the stock exchange and non-listed with it.
The research raises many related future research prospects, including the impact of earnings management practices on the market value of the financial institutions.