Financial ratio time series models in defense industries

Abstract

This thesis examines the forecasting ability of financial ratio time series models as applied to defense industry firms. Using previously developed descriptive models of financial ratio time series behavior, this thesis identifies plausible financial ratio forecasting models. The ability of seven different models to predict future values of financial ratios is then tested with data from defense industry firms. The results are used to answer questions concerning the accuracy and bias of forecasts and the appropriate applications of specific forecasting models. The thesis concludes that the ability of time series models to forecast future values for financial ratios depends on the specific ratio being forecast, and that the simplest model, a random walk model, is among the most useful for forecasting.