A test for multivariate normality in stock returns
Article Abstract:
A general test procedure is developed for calculating multivariate test statistics that considers cross-sectional dependence with regards to normality in stock returns. The testing procedure for the presence of a multivariate normal distribution is based on L. Hansen's generalized method of moments approach. The tests employed in the study exploit data found in both the marginal and joint moments of asset returns. Ease in calculation with asymptotic distributions is another advantage of the test statistics in this study. The results indicate strong evidence of the nonnormality of both the stock returns and market-model residuals. The nonnormality is also present in both the marginal and joint distributions of asset returns. These findings show that, empirically, the multivariate normal hypothesis cannot be supported.
Publication Name: The Journal of Business
Subject: Business, general
ISSN: 0021-9398
Year: 1993
User Contributions:
Comment about this article or add new information about this topic:
Empirical tests of the bias and efficiency of the extreme-value variance estimator for common stocks
Article Abstract:
The extreme-value variance estimator of Michael Parkinson performed poorly in comparison to the close-close estimator in the absence of an outlier screen, but was more efficient in a continuous-trading model. Statistical tests confirm that application of a screen for high and low price errors results in a downward bias in data. An inverse relation of this bias is seen with trading volume. For nine of 12 price-volume groups, Parkinson's estimator outperforms the close-close estimator in efficiency. A more precise variance measure is made with Parkinson's estimator than with the close-close estimator in applications involving event study. The sample data for the study was taken from Cornell University Price Volume tapes containing daily prices and volumes of stocks traded in NYSE and AMEX over a 16-year duration.
Publication Name: The Journal of Business
Subject: Business, general
ISSN: 0021-9398
Year: 1991
User Contributions:
Comment about this article or add new information about this topic:
Naive trading rules in financial markets and Wiener-Komolgorov prediction theory: a study of "technical analysis."
Article Abstract:
The preference of stock market analysts for informal prediction rules, commonly referred to as technical analysis, is studied. The reasons for analyst preference for technical analysis over formal Wiener-Kolmogorov prediction theory is explored by investigating two key issues. These are: whether the various ad hoc rules of technical analysis can be represented in formal algorithms, and whether technical analysis can explain stock price movements in a way that linear models of Wiener-Kolmogorov prediction theory cannot.
Publication Name: The Journal of Business
Subject: Business, general
ISSN: 0021-9398
Year: 1991
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: A comparison of futures pricing models in a new market: the case of individual share futures. Pricing Commodities When Both Price and Output Are Uncertain
- Abstracts: The role of the accounting rate of return in financial statement analysis. Investment decisions and the equity accounting standard
- Abstracts: Correlates of charismatic leader behavior in military units: subordinates' attitudes, unit of characteristics, and superiors' appraisals of leader performance
- Abstracts: An alternative to free trade or protectionism: why corporations seek strategic trade policy. Strategic use of technology
- Abstracts: An alternative to free trade or protectionism: why corporations seek strategic trade policy. part 2 Productivity planning and strategy in retailing