Behavior Finance and Agent-Based Computational Finance: Toward an Integrating Framework

One principal objective of financial economists is to understand, to explain and even to predict the macrophenomena that emerged into the financial markets. To do this, economic model builders seek appropriate micro-foundations of traders’ decision-making behavior under risk and uncertainty that are both empirically plausible and analytically tractable. Following the neoclassical economics tradition, the orthodox financial theorists often adopted the assumptions that decision-makers possess von Neumann-Morgenstern preference and are rational expected-utility maximizers. This solid microfoundation of decision-making behavior along with some other key assumptions about the whole structure of models, e.g., rational expectations, representative agents, imposed market-clearing conditions, no-arbitrage conditions, etc., constitute the formal framework of the standard finance theory today. In this vein, financial economists constructed numerous highly successful and influential theories, such as the capital asset pricing model (CAPM), efficient markets hypothesis (EMH), and the Black-Scholes option pricing model, among many others. The common features of these models are their neat structure and analytical tractability. Supported by several empirical studies, the financial economists’ view of the financial markets was based on these models until the mid-1980s. In this Utopia, the asset prices react to any new information prevailing immediately, and thus the asset returns are unpredictable. The volatility of asset prices comes mainly from the effects of the fundamental side. The only forces that drive the asset prices and expected returns are economically meaningful risk factors.1 That is to say, traders can receive return premia solely from bearing market risk. Any irrational or noise traders will lose money to informed rational arbitrageurs and eventually be eliminated from the market in the long run. By further assuming that it is common knowledge that all traders are rational and that traders share common prior beliefs, the notable no-trade theorem holds in these settings. The financial market is complete and thus derivative securities are nothing more than redundant assets. No-arbitrage arguments work pretty well in pricing these assets. The above scenarios are all that we’ve learned in the standard finance textbook as the core of finance theory.