No CrossRef data available.
Published online by Cambridge University Press: 19 October 2009
The API metric, as initially formulated by Ball and Brown [1], has been used to examine the relationship between stock prices and accounting numbers. Its use in this manner has raised at least three difficulties: (1) the metric does not utilize all of the information portentially available from accounting numbers, (2) the statistical significance of the API metric and of differences between API's has not always been satisfactorily considered, and (3) the meaning of the metric has been questioned. This paper will attempt to resolve these difficulties by reformulating the API approach. Two nonparametric statistical procedures are considered, both of which provide a test of the statistical significance of the relationship between accounting income numbers and stock prices. The first procedure is particularly appropriate when alternative accounting income numbers are considered. Both procedures provide a conventional unambiguous interpretation of the results.