Abstract
I first address the initial part of an earlier Hawkins article where he demonstrates the use of regression analysis to determine a subject company's value. Herein, it is shown that even a simple, basic application of regression analysis requires more fundamental analysis than Hawkins put forth. My analysis presents both flaws in the data sets and a conceptual flaw in the use of revenue alone to predict value. I propose and demonstrate corrective methodologies for both of these errors. I then deal with the use of regression analysis to determine the amount of active versus passive appreciation in the growth in value of a company during a marriage. I explain why this is a misuse of regression analysis and demonstrate better regression techniques for forecasting revenues using time trend and housing starts as independent variables, and, at the same time, I transform the model to correct for the serial correlation problem raised in Trout's letter to the editor.
