Predicting and Reviewing NYSE Stock Prices by Use of Basic Statistical Analysis and Logic

Authors

  • Craig G. Harms

Abstract

"This paper presents a three week in-class case that empowers the class to predict weekly stock prices for a group of NYSE stocks. After each Friday close, the data base is updated and the process to predict stock prices for the next week begins again. Minimal statistical background is required in this junior/senior level quantitative methods course which develops predictive models using techniques such as regression analysis. After completing several introductory cases, which develop the student’s understanding of normal curve theory, trend analysis, seasonal adjustment, and residual analysis, they attack the very real problem of stock price prediction. Perceived as “real-real” by students, interest soars as students discuss predicted stock price for their chosen stock. A good discussion commences about quantitative models versus guessing the movement of stock prices (random walk theory). In the end, a discussion dealing with how these quantitative models can be used today by the students is enthusiastically joined. Topics included in the paper: model building, simple regression, time series analysis, statistical and residual analysis, graphical analysis, and investor decision making. "

Downloads

Published

1997-03-06