A0431
Title: Fundamentals of financial forecasting: Simplicity vs. complexity
Authors: Dimitrios Thomakos - National and Kapodistrian University of Athens (Greece) [presenting]
Abstract: Forecasting of financial time series is a topic of broad interest to academics and industry practitioners alike. One of the purposes of financial forecasting is to construct quantitative trading strategies which are essentially attempts at market timing: The forward-looking decision of whether to buy or sell a financial asset. This original form of speculation, as old as the financial markets themselves, has a number of characteristics which make it amenable to either a simplistic or a complex form of analysis, for traders can use, for example, simple moving averages or complicated machine learning methods. A number of arguments are offered in favor of simplicity vs. complexity, and a number of suitable yet simple models are presented that are interpretable, understandable, easily computable and (backtested to be) profitable. If one can devise such forecasting models based on simple concepts that are fundamental to financial forecasting, then the need or usefulness of large and complicated models for this kind of academic or professional activity is called for a re-examination.
