The world of finance has been in constant pursuit of ways to understand the underlying processes that operate in the financial markets. Motivated by the desire to increase profit, model-building procedures have been developed by researchers to represent and fully interpret the behavior of finance-related factors.
One such factor is the pricing of financial assets which is covered in the doctoral study “Estimation and Evaluation of Asset Pricing Models with Habit Formation Using Philippine Data” (2004) by Dr. Raymund R. Abara of the University of the Philippines, Diliman College of Business Administration.
Financial assets are non-physical possessions which are of economic value such as securities and certificates. In the study, Dr. Abara focuses on one of the most popular financial asset pricing models today, the Consumption Capital Asset Pricing Model (C-CAPM). The C-CAPM is a financial model which considers the risk or probable loss of expected returns of an asset and the link between asset returns and consumption in pricing bonds, stocks, and other financial assets. It is simpler than other asset pricing models for it has fewer statistical assumptions on asset returns and it has flexibility when dealing with various types of utility functions that represent an investor’s risk preferences. The power utility function, also called constant relative risk aversion (CRRA), is the usual choice of researchers to represent the risk preferences of individuals in the C-CAPM due to its simple functional properties. However, several studies have shown that the only way C-CAPM would reasonably predict financial returns is if the CRRA is extremely large, implying that the investors are highly averse to uncertain payoffs unless properly compensated. The risk aversion is suspiciously too high compared to the established principles in standard models of financial markets, a situation researchers cannot explain. This phenomenon is known as the equity premium puzzle. In his study, Dr. Abara evaluates another kind of utility function in the Philippine setting that can be an alternative to the CRRA but offers a probable solution to the equity premium puzzle. This utility function is the habit formation utility function.
The habit formation utility function models an individual’s habit of consumption based on past consumption and other external factors. Based on current research, the habit model may be divided into two: the internal habit model and the external habit model. The internal habit model is based purely on the investor’s past consumption decisions. The external habit model, on the other hand, represents the habit as dependent not only on the investor’s past consumption decisions, but also on other external factors such as envy and the bandwagon effect.
Using Philippine equity and Treasury bill data from 1990 to 2003, the author’s analysis shows that, similar to previous habit formation studies, the equity premium puzzle is resolved when the C-CAPM is estimated with the habit formation function. Furthermore, the external habit model has been found to provide better forecasts or prediction values for financial returns than the internal habit model. Habit is so complex it cannot be represented by past consumption decisions alone. Other external factors must be considered to predict accurately the behavior of financial returns.
The external habit model has also been found to perform best among other models, such as the power utility function combined with C-CAPM, the random walk with drift model and the traditional Capital Asset Pricing Model. This assessment, however, should still be carefully investigated in future research.
Raymond Abara is an Associate Professor (on leave) in financial economics and quantitative methods in the University of the Philippines, College of Business Administration (UP-CBA). He teaches finance subjects in the graduate and undergraduate levels and his teaching portfolio consists of financial derivatives, corporate finance, and CFA review courses. He has also handled statistics and operations research subjects.
