Data_Prediction of Financial Restatement by Managerial Tone- Using Machine Learning Approach

The dataset used in Prediction of Financial Restatement by Managerial Tone: Using Machine Learning Approach. A total of 1,893 financial restatement observations were identified from the Audit Analytics database and matched with firm-level data from Compustat for the period 2010 to 2022. The earning conference call transcripts are obtained from StreetEvent. Consistent with Zhang et al. (2025), only the most recent restatement year for each firm was retained, resulting in a sample of 1,025 distinct restating firms. To examine the association between managerial tone and the likelihood of restatements, a matched sample of non-restating firms from the same time period was randomly selected. This process yielded a balanced panel comprising 2,050 unique firms. All continuous variables were subsequently winsorized at the 1st and 99th percentiles to mitigate the influence of extreme values. Variable Description RESTATE Indicator variable equal to 1 if firm has restated earnings, and 0 otherwise. Tone_LM The optimistic words less pessimistic words of managers scaled by total words in the conference call, based on the Loughran and McDonald (2011) word list. ROA Operating Income Before Depreciation as a fraction of average Total Assets based on most recent two periods. SIZE Natural logarithm of Total Assets. PROFIT Gross Profitability as a fraction of Total Assets. LEV Total Debt as a fraction of Total Assets. IAC Inventory as a fraction of Current Assets CAPEI Price divided by the average of ten years of earnings, adjusted for inflation. IT Inventory Turnover. BS Book Value Per Share. BM Book Value of Equity as a fraction of Market Value of Equity. CR Long-term debt divided by the sum of long-term debt and shareholders' equity. AT Revenue divided by average total assets. PTB Price-to-Book ratio. PEG Trailing Price-to-Earnings to Growth ratio. PS Market capitalization divided by annual revenue.