Critique : ‘Empirical Evaluation of accounting income numbers’
Title and reference:
Ball, Ray and Phillip Brown, (1968) “An empirical evaluation of accounting income numbers ’, Journal of Accounting Research; Vol 6; Issue 2, PP 159-178; 20p.
This paper is the first attempt to study the usefulness of the accounting income numbers. Traditionally accounting theorists have evaluated the usefulness of accounting practice by the extent of their agreement with the particular analytic model. However their mode of inquiry has generally been restricted to the extent of comparing the existing practices with the more preferable ones or suggesting a normative model. One of the major shortcomings of the normative model is that it ignores the extent to which it can predict or explain the observed behavior. Normative accounting models in past has led theorists to claim that income numbers cannot be defined substantively, they lack meaning and are therefore of doubtful utility. Hence empirical testing becomes mandatory to emphasize that a lack of substantive meaning does not generally imply a lack of utility.
Since accounting income numbers are of special interest to the investors, security prices have been used as the predictive criterion of the investment decision. If an observed revision of stock prices is associated with the release of income report then it can be concluded that the information reflected in income numbers s useful.
Context and theoretical grounding:
The context of the paper can be understood as that of providing an empirical evaluation for the usefulness of the accounting income numbers . This paper derives it theoretical foundation from the vast literature of financial economics. Two models, Market model and conditional expectations, have been used to study the information content of the accounting income numbers.
Discussion and critique about methodology:
The paper is empirical in nature. The first section deals with providing a general overview of the analytic models studied by the accounting theorists and their limitations. It emphasis the need for an empirical evaluation to tests the utility of the accounting income numbers.
Financial income statements are framed to cater to the needs of target audience. The audience broadly includes government, investors, managers etc. This paper implicitly assumes investors to be the main audience of the income statements and furthers its investigation by studying the relationship between the announcement of the income numbers to the changes in the stock prices. However usefulness of income numbers ranges far beyond the stock prices. Hence this article takes a limited view of the usefulness.
The justification provided for selecting the behavior of security prices as an operational test of usefulness is based on the efficient market hypothesis. This proposition states that the capital markets are both efficient and unbiased i.e. if information is useful in forming capital asset prices , then the market will adjust asset prices to that information quickly without leaving any opportunity for abnormal gains. This hypothesis though theoretical sounds promising leads to ambiguous results when tested for various stock exchanges in different settings. Thus this raises questions on the basis of selecting stock prices as an indicator to reflect the usefulness of the accounting income numbers.
The second section of the paper deals with the construction of models and verifying the information content of the income numbers. The authors construct two alternative models of what market expects income to be and then investigating then markets reaction to when its expectations prove wrong. The literature shows that Firm’s earnings per share (EPS) is associated with economy wide factors, thus at least a part of the change in a firms’ income from one year can be expected in the following year. If the firm’s income is related to income of other firms in some way, then this knowledge can result in the conditional expectation model.
The left hand side of the above equation reflects the change in income for firm J while the right hand side f the equation depicts the change in the average income of all firms other than firm J. Uj,t is the forecast error which is assumed to contain the new information which results in the changes in the firms’ income form the other firms in the market.
To explain the markets reaction the authors adopt the market model. The impact of market wide information on the rate of return on the firm J’s stock can be estimated by its predicted value from the linear regression of monthly prices relative of J’s stock on market index return.
However the Fisher’s index is used to represent the market index. The relevance of fisher’s index to represent the market is limited as it is an equally weighted index and thus would cause the results to be erroneous. Further the market wide information relating to all the firms could have been broken into specific factors such as term structure of interest rates, inflation, industrial production etc. this would have led to more robust results instead of taking fishers index as market index representative of the market wide information.
The residual in the above equation measures the extent to which the realized returns differs from the expected returns .Since market adjusts to the new information efficiently and quickly, the residual represents the impact of new information, on the return from holding common stock in the firm j.
Two measures of income have been taken, net income and EPS. The graphs relating the abnormal performance index and the announcement clearly demonstrate the information contained n the annual income number is useful in that if actual income differs from the expected income, the market typically reacts in the same direction. One of the most important results of the paper is that most of the information contained in reported income is anticipated by the market before the annual report is released. The anticipation is so accurate that the annual number does not appear to cause any jump in the abnormal performance index in the announcement month.
Data had been collected for a period from 1957-1966.The selection criteria would have led to further bias. The sample did not include young firms; those who have failed .This exclusion clearly would have resulted in survivorship bias in the results which shows that persistence in exuberant performance might have its basis in the survivorship bias. Further as results have indicated a correlation between sizes of the firm and returns thus the exclusion of the small sized firms could lead to erroneous results.
However one of the most serious limitations remains the assumption of the unidirectional relationship between income and stock prices. It would be naïve to think that only income causes changes in stock prices and not vice versa. Evidence shows that stock prices and income are endogenous in nature thus it would be more relevant if the system of simultaneous equations is used rather than simple regression technique.
None the less this paper had made seminal contributions in the accounting research. This paper was the first systematized attempt to study the information content of the accounting income numbers. This paper had led to paradigm shift in the models that were earlier analyzed by the accounting theorists.
It can be said with much confidence that this paper was one of the building blocks towards the event study methodology which was given by Fama and Fench (1992).