Home Articles Abstract
Research Article

A Study on the Additional Information Value of Consolidated Financial Statements

Jung, Gyueon · Kim, Jeongwon

Published: January 1994 · Vol. 23, No. 3 · pp. 83-108
Full Text

Abstract

Recently, there have been numerous arguments that the current consolidated financial statements based on parent-subsidiary relationships should be replaced or supplemented with business group consolidated financial statements, given that in Korea's unique economic environment—characterized by stock ownership structures in which the de facto owner-manager controls affiliated companies within a business group through shared controlling stakes among the owner-manager, related parties, and affiliated companies—the usefulness of such statements is questionable. As part of an empirical examination of these claims, this study empirically tested the usefulness of current consolidated financial statements. The usefulness of consolidated financial statements was verified by examining the incremental information value of consolidated financial statements over individual financial statements through stock price reactions at the time of financial statement disclosure. Specifically, the incremental information value of consolidated financial statement information was tested through regression analysis using abnormal returns at the time of financial statement disclosure as the dependent variable, with individual financial statement earnings and sales, consolidated financial statement earnings and sales as explanatory variables, and with the difference between consolidated earnings and individual earnings as an explanatory variable. Abnormal returns were measured using a two-factor model that added an industry index to the market model to reduce the cross-sectional correlation of market model residuals. The empirical results were as follows. First, when consolidated earnings information was measured as the difference from prior-period consolidated earnings, the incremental information value of consolidated earnings could not be verified due to the high correlation between consolidated and individual earnings. However, when consolidated earnings information was defined as a dummy variable based on the difference between consolidated and individual earnings, it was found to have incremental information value over individual earnings. Since individual financial statements use the cost method for investment stock valuation, the operating performance of subsidiaries is not reflected and unrealized gains from intercompany transactions are not eliminated, whereas consolidated earnings are calculated with both of these items incorporated. Therefore, the difference between consolidated and individual earnings primarily represents the operating performance of subsidiaries and unrealized gains from intercompany transactions, which constitutes the incremental information that consolidated earnings possess over individual earnings. In this regard, verifying the usefulness of the difference variable between consolidated and individual earnings implies the usefulness of consolidated earnings. Second, neither consolidated sales information nor individual sales information significantly explained abnormal returns, either individually or incrementally over earnings. The failure of individual sales information to demonstrate information value appears to be due to the measurement method for sales information and the selection of the verification period—the timing at which information reaches the market. Unlike earnings information, which is calculated through the complex process of settlement, individual sales information is generated through simple aggregation of sales amounts, and since each quarter's corporate sales are reported to the tax office by the 25th of the following month through value-added tax filings, sales information can easily leak to the market even during the accounting period. Therefore, when sales information is measured as the difference from the prior period, it would be measured very differently from the sales information that the market actually encounters. Additionally, for December fiscal year-end companies, fourth-quarter sales are aggregated early in the following year and reported to the tax office by January 25th, so sales information should be considered as having leaked to the market by January at the latest. Consequently, in this study, which measured sales information as the difference from the prior period and set the verification period as February and March, individual sales information failed to explain abnormal returns at all.