Complex Parsimony in Equity Valuation - An Empirical Assessment of Model Design and the Incremental Effects of Complexity
Abstract: When deciding on a valuation model, an investor must be attentive to the juxtaposition between the usefulness, driven by complexity, and the simplicity of that particular model. This thesis employs an examination of four parsimonious equity valuation models (dividend discount model [DDM], residual income valuation model [RIV], abnormal earnings growth model [AEG], and the Ohlson-Juettner-Nauroth model [OJ]) and their usefulness in relation to the Nordic stock exchanges. Adding to previous studies, it also evaluates the impact of two different payoff schemes, analysts' estimates and martingales. To elaborate on model performance, this study further investigates both the separate and conjoint effects from adding three complexity adjustments, i) extending the forecast horizon, ii) adjusting for bankruptcy risk, and iii) excluding transitory items from the earnings measure. The analysis is carried out using a proprietary measure, which considers a comprehensive view on accuracy and spread. In general, it is found that RIV consistently outperforms the other models, regardless complexity adjustments. Notably, the impact from adding complexity to the models is largely dependent on the combination of model specification and model inputs, but the models' overall performance is increased. Nevertheless, the increased model performance from complexity adjustments must be assessed in light of the additional effort that such adjustments entail, as their marginal benefits are questionable.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)