The Empirical Research On Abnormal Cash Flows' Predictability On Accounting Restatements

In recent years, the credibility of financial reporting in capital market has been severely shaken by a string of accounting scandals surrounding large, public companies such as Yin GuangXia, Qiong MinYuan and KeLong event. Concurrently, there has been a significant increase in the number of firms that have had to restate their previously reported financial statements which is crucial for a well-functioning capital market and for the efficient allocation of resources. Given the substantial social and economic costs associated with accounting scandals, it is important to develop diagnostic tools to identify which firms are likely to occur accounting manipulations so that one can take measures to prevent undesirable outcomes.In this paper, we study firms that have restated their financial statements .The listed companies having the corrections of accounting errors in the period from 1999 to 2004 were collected as research sample. We focus on opportunities for earnings management by studying accounts that have been restated and on incentives for earnings management by estimating abnormal cash flows from operations. In the first step, we analyze the database of accounting restatements and compare the restated financial statements with their original numbers in balance sheet and income statement. This process identifies the accounts that were restated and therefore were likely to have been manipulated (opportunity). Then we classify all samples into different types, such as revenue recognition, cost or expense-related. We find that different types have different opportunities. With revenue–recognition restatements, a substantial number of adjustments involves a decrease in accounts receivable. We finally find that each restatement type has its particular accounts that are likely to have been manipulated, which helps the listed firms manipulate earnings.In the second step, to identify pressure (incentives) to manipulate earnings, we rely on abnormal cash flows from operations (AbCFO). Learning from results in existing literature, this paper uses the modified model of accruals, operating cash flows and accounts receivable. Based on these models, this paper decomposes accruals into normal accruals and abnormal accruals, operating cash flows into normal cash flows and abnormal accruals, the change in accounts receivable into changes in normal and abnormal accounts receivable. Finally the listed firms that have restated their financial statements have significant lower abnormal cash flows than normal firms. In other words, the poor performance of restatement firms is almost all reflected in abnormal cash flows.At last, this paper incorporates various measures of business model which include leverage, profit margin and capital intensity. A model is obtained for predicting accounting restatements by combining these three steps, focusing on the candidate variables identified in the first step and abnormal cash flows as well as abnormal accruals from the second step. A formal Logit model is formed based on the data for predicting accounting restatements. For all types of restatements and revenue-recognition restatements, the abnormal cash flows and abnormal accruals have negative and significant coefficients. It proves both of them can predict accounting restatements. But abnormal cash flows is more powerful for predicting. The coefficient on abnormal cash flows is also significant at conventional levels for cost-related restatements. Overall, abnormal cash flows is more powerful and stable for predicting accounting restatements. However, contrary to conventional wisdom, abnormal accruals have a negative effect with accounting restatements. That is low abnormal accruals are associated with restatements. Similar to low abnormal cash flows, low abnormal accruals reflect poor operating performance, which tends to induce earnings management, resulting in restatements in the later periods.