Artificial Intelligence in Finance

Chapter 3. Normative Finance

The CAPM is based on many unrealistic assumptions. For example, the assumption that investors care only about the mean and variance of one-period portfolio returns is extreme.

Eugene Fama and Kenneth French (2004)

[S]ciences that involve human beings rather than elementary particles have proven more resistant to elegant mathematics.

Alon Halevy et al. (2009)

This chapter reviews major normative financial theories and models. Simply speaking and for the purposes of this book, a normative theory is one that is based on assumptions (mathematically, axioms) and derives insights, results, and more from the set of relevant assumptions. On the other hand, a positive theory is one that is based on observation, experiments, data, relationships, and the like and describes phenomena given the insights gained from the available information and the derived results. Rubinstein (2006) provides a detailed historical account of the origins of the theories and models presented in this chapter.

“Uncertainty and Risk” introduces central notions from financial modeling, such as uncertainty, risk, traded assets, and so on. “Expected Utility Theory” discusses the major economic paradigm for decision making under uncertainty: expected utility theory (EUT). In its modern form, EUT dates back to von Neumann and Morgenstern (1944). “Mean-Variance Portfolio Theory” introduces the mean-variance portfolio (MVP) theory according to Markowitz (1952). “Capital Asset Pricing Model” ...