Commodity Futures Pricing Under The Condition Of The Bounded Rationality

Containing the function of price discovery and risk aversion, futures markets play the more and more important roles in the international financial market. Since the pricing of commodity futures has been relative complete, for there are lots of mathematical models about it, but these models generally have strict assumptions, which are somewhat unrealistic that lead to great deviation from the actual market. As an emerging mainstream disciplines, behavioral finance put forward a series of questions to traditional finance, and proposed lots of new theories combined with the various market anomalies. This paper intends to apply behavioral finance to the commodity futures pricing with the point of view of bounded rationality, researching the effect bounded rationality applied to the commodity futures price, then gaining a appropriate futures pricing model.The main contents of the paper are as follow:in the first section, we conduct a comprehensive exposition of the literature on bounded rationality, bounded rationality Origins, existing research methods; in the second section, we expand the traditional futures pricing theory,relaxing its assumptions, then we propose a semi-martingale futures price model with the existence of market jump, using multiple regression model to verify its effectiveness; in the third section we focus on the loss aversion which is a important part of bounded rationality, firstly we propose the trading strategy of loss averse investors, then we measured loss aversion degree by the EGARCH model, and examine the evolution of the loss aversion coefficient in the different stages of the price curve; in the fourth section we concentrate our study on herding, firstly we discuss the investment strategy of investors who are in herding, then exploring the existence and dynamic development of herding by the use of the threshold pre-average volatility model which is suitable to the analysis of high-frequency data; in the last section, we intend to establish a noise-evolution-game model which contains the loss aversion and herding.According the research, the main conclusions of this paper are as follow:(1) Bounded rationality psychology is an important factor affecting the futures price volatility, especially in extreme market, bounded rationality psychological will exacerbate price volatility;(2) The traditional futures pricing model could not explain the irregular phenomenon on the market, and the assuming relatively relaxed semi-martingale futures price model is more in line with actual market, exploring the fluctuations in futures prices on the actual market effectively. Built on the semi-martingale futures prices model, overall regression model and GARCH model results show that the main factors affecting the futures market price are the external factors, such as similar foreign futures market, the domestic stock market, and internal factors such as the degree of rational investors, information factors, the reference point factors will have impacts on futures prices accordingly, and the extent and direction of various factors change with the stages of the futures prices;(3) Futures investors’investment strategy is based on the utility function, one important factor affecting the utility function is loss aversion coefficient. There exists significant loss aversion in the futures markets. The empirical results show that when the price fluctuates between two units, the loss aversion coefficient is about2.15. The other empirical results show that the greater the price volatility, the higher the risk-free rate, the more opaque the information, the large the loss aversion coefficient,whereas the smaller;(4) There are various theories of the causes of herd behavior, but one consensus exists that herd behavior always presents in the trading action of investors, while the incomplete information and investors’tendency to imitate the generation are both plays major roles on herd behavior. This paper describes the generating process of information flow herd behavior, proposing that the previous decisions of the two investors are the major factor affecting the following investors.The empirical results show that significant herd behavior on the market is passed from the large investors to the small and medium investors, and herd behavior is a main factor affecting the market obviously. However, The impact of herding is relative weak when the market is stable, but it will increase when the price volatility increases;(5) Over-lapping-Generation model containing loss aversion and herding shows that, factors affecting their investment decisions are different between complete information investors and incomplete information investors, loss aversion and herd behavior are both the main factors, and loss aversion has a more significant effect on complete information investors while herd behavior on incomplete information investors.