A LASSO-BASED ANALYSIS ON VALUE RELEVANCE OF ACCOUNTING INFORMATION IN THE BULGARIAN CAPITAL MARKET

Abstract

The aim of this study is to examine the value relevance of accounting information, specifically the influence of accounting information on the share prices of non-financial public companies listed on the Bulgarian capital market. The data set contains twenty accounting variables and the stock prices of twenty public Bulgarian companies. The analysed period is 01.01.2015 – 31.12.2022. Descriptive statistics and Least Absolute Shrinkage and Selection Operator (LASSO) regression model are applied. The empirical analysis results show that accounting information affects share prices of the examined companies from 2015 to 2022. Addionally, LASSO regression model is an appropriate model for identifying significant accounting information variables that influence stock prices. The findings of the study offer practical implications for various stakeholders, including investors, policymakers, and financial analysts.