George E. Pinches

Sophisticated capital budgeting selection techniques and firm performance

Firms using sophisticated capital budgeting techniques (i. e., those that employ present value analysis and account for risk) should theoretically perform better than firms using naive models such as the payback period or accounting rate of return. However, previous empirical work examining this question has produced mixed results. To correct for limitations in these studies, several tests were conducted on firms that adopted sophisticated selection techniques versus a control group of firms that employed naive techniques. After controlling for differences in systematic risk, industry effects, and size, interrupted time-series tests of relative market returns were performed. Based on the results of this study we conclude that the adoption of sophisticated capital budgeting selection techniques will not, per se, result in superior firm performance. It is possible that the adoption of sophisticated selection techniques is one of many policies the firm pursues in the face of economic stress, and this, in combination with other policies, may help to bring about economic recovery for the firm.

Testing for Market Efficiency: A Comparison of the Cumulative Average Residual Methodology and Intervention Analysis

During the past decade considerable empirical evidence has been accumulated suggesting the stock market adjusts to the arrival of new information in an efficient manner. The studies providing this evidence consist of announcement tests of new publicly available information (such as earnings, stock splits, accounting changes, etc.) on the risk-adjusted return of securities. The specific methodology employed is crucial since it directly affects the results of a test for market efficiency. Following the pioneering work of Ball and Brown [1] and Fama, et al. [15], many researchers [6, 12, 21, 22, 27] have employed a similar methodology in order to test for market efficiency. This cumulative average residual (CAR) methodology consists of: (1) estimating the parameters of the market model based on data in a time period prior (and sometimes subsequent) to an announcement, and (2) analyzing the residuals derived from applying this model to a time period which includes the announcement date.

Factors influencing classification results from multiple discriminant analysis

Abstract In recent years the use of multiple discriminant analysis has gained wide acceptance in applied business research. This increasing utilization is due, in part, to the availability of discriminant analysis routines in a variety of statistical packages (see Appendix). However, few applied business studies using discriminant analysis are completely free of methodological and statistical problems. These problems have been caused by 1) the lack of in-depth knowledge by business researchers of the statistical properties of discriminant analysis; 2) problems associated with the data typically employed in business studies; and 3) the lack of complete answers by statisticians to many problems associated with discriminant analysis. The purpose of this paper is to provide a partial solution to the first of these problems—the lack of understanding by business researchers of the many statistical considerations associated with classification when discriminant analysis is employed. Specifically, ten factors or items (in two groups) that may directly influence the reported classification results will be investigated. Factors in the first group are those that arise from the variables and/or sample employed. The second group of factors encompasses itmes that are directly under the control of the researcher (and, hence, require explicit consideration) when classification is undertaken. These ten items are those that a) Arise from the variables and/or sample 1. 1. multivariate normality 2. 2. number and independence of predictor variables 3. 3. sample size 4. 4. missing values 5. 5. initial misclassification b) Require classification decisions 1. 1. error rates 2. 2. equal versus unequal dispersion matrices 3. 3. a priori probabilities 4. 4. costs of misclassification 5. 5. reduced (or discriminant) space. Implicitly or explicitly every applied researcher addresses these ten items; the purpose of this paper is to focus specific attention on these factors so further research will not suffer some of the statistical and methodological problems that have plagued many recent research efforts. This paper is written from the viewpoint of an applied researcher who has to be concerned about the practical problems associated with using multiple discriminant analysis with sample-based data.

Additional Evidence of Heteroscedasticity in the Market Model

Sharpes market model [29] is widely used both by academic researchers and practitioners in finance, but it cannot be accepted with complete confidence until some of its basic assumptions are tested more thoroughly. The applicability, usefulness, and reliability of the model are functions of its conformity to real data, which in turn depends partly on the unresolved question of heteroscedasticity.

An Analysis of Portfolio Accumulation Strategies Employing Low-Priced Common Stocks

The recent studies of Fisher and Lorie [10 and 11], Brigham and Pappas [2], and others have contributed significantly to our knowledge about overall rates of returns on common stocks. There is, in addition, a growing body of empirical evidence investigating the performance of alternative portfolio maintenance strategies [4, 8, 9, and 13]. However, to date, this evidence for rates of return under alternative portfolio strategies has been obtained for a “one-time†investment decision (with allowances for reinvestment of dividends and possible intraportfolio reallocation). In this paper the effects of alternative portfolio accumulation strategies on subsequent portfolio rates of returns are examined.

The response of capital markets to the R&D process

Abstract We examine the response of the capital markets to research and development (R&D) project announcements by firms along three stages of the R&D process: initiation, continuation, and new-product introduction. Using event study methodology, conventional in financial economics and strategic management, we examined 525 R&D project announcements over the 1977–1989 period. Our analysis suggests that investors respond favorably to R&D announcements during the continuation and new-product introduction stages. In the biotechnology industry, however, the greatest response occurred in the initiation and continuation stages. Significant gains in wealth were observed for relatively smaller firms, and in the case of continuation announcements when R&D was viewed as a way of stimulating growth. After accounting for firm size and the effect of the biotechnology industry, the frequency of R&D announcements by firms does not lead to greater stock market effects. Our data paint a picture of rational and sophisticated investors who understand and respond to R&D project announcements — a portrait that stands in stark contrast to the current criticisms of a myopic stock market.