Causal Variation: Exploring the Non-Random in Stock Price, Stock Index, Sales, and Accounting Time Series

Abstract

We show two methods for measuring non-random observations using statistical control software, specifically SigmaXL, across diverse time series. First, we use an error counting method based on eight non-random rules of statistical control charts, and subsequently we assign a dollar value to each of those non-random observations to evaluate non-random to random rates. The computed error rates are also compared to a randomly generated sample of 100K and the corresponding probably of occurrence. Finally, these methods, coupled with a new indicator, the Taylor-Kiymaz multiple, allow for the comparison of stock prices, market indexes, sales metrics, chart of accounts across many time periods.