| China's stock market is an emerging capital market.The Shanghai A-share index can rise from 998.23 to 6124.04 points in a short span of 30 months,and can drop to 3500 points during less than 6 months.Then,how should we understand the market?In the participation of many populations,the stock market is a complex system,with the characteristics of emergent,metastable,cyclical,parasitic,virtual,and high-risky.The demand for the stock is formed by the transactions of numerous investors which can stimulate the emergence of the different collectivity of the investors.Just starting from the microplane,abandoning the macroplane,only has a limited view of something as looking at a leopard through a bamboo tube.Therefore,this article seeks about the "behavioral traits" of the stock market from the perspective of the population.As the virtual of stock,in the role of the self-organization,emergence may lead to short-term market failures.Moreover,the emergence of the market shows a nonlinear feed back features,and short-term shock may have an adverse effect in the long term.In order to grasp the market,it should pay attention to the market emergence caused by the c of the collectivity,investors,which need our think much of population.The researches of the population in the finance include psychological population and objective population.The research to the psychological group focuses mainly on herd effect,and it reflects the interaction between the different populations,which can obtain high-level organizational emergence.According to the system science theory,the paper combines overall analytic method and synthetic microanalytic approach.Through different levels of investor collectivities,it studies,at the macro-level and micro-level,on the population preferences and the interaction of heterogeneous populations,to grasp some aggregate laws of the stock market.Here it regards the populations' traits of the stock market as the research target.And the populations don't aim at certain stock or certain investor,but the whole market.Therefore. we cut to the theme by high-level emergent populations,and the populations our main line and the object of the paper.â… .Overall Analysis.It regards all of the investors as one population and research on the market aggregate preference.The preferences of investors are the most starting point and basic proposition of the study of Which is better,expected utility theory or other utility function theories? Many experiments in the laboratory,can't get a consistent result,which even can't reflect the real preferences of the investors.In addition to non-satiation can be recognized widely. economic theory also did not give any guidance for the specification.At present,the risk preferences of investors show diversity.Under conditions of uncertainty.the risk attitudes may be aversion,loving or neutral.When shifting the view from the micro to the macro. the population behavior is not the simple sum in of the individuals that often shows some new characteristics because there is complex interaction among different units.Clearly the utility function of different populations may be different,so it should develop a framework of utility function which can reflect the diverse preferences.To determine the utility function of a specific population,it can study on the all units of the collectivity as a group,from the perspective of mass.and simplify the research by cutting out the superfluous.Facing with various risk preferences,most of the risk measures and asset selection methods have their drawbackes.So the paper has adopted Stochastic Dominant(SD) analysis,a non-parameter method,which can let the data "speak for themselves" and show the variety of the preferences.Through the analysis of the third-order derivative of the utility function we can exposure some properties of risk attitude,degree of risk preferences and the trading strategies.On this basis,combining the SD criteria based on the traditional expected utility,the prospect SD criteria based on the Prospect Theory and the Markowitz SD criteria based on the Markowitz utility function.And in accordance with Whitmore's (1970)ideas,we construct an analysis framework,namely,Diverse Stochastic Dominance (DSD)criteria,which contains different risk attitude including aversion and loving, different degree of risk preferences including mild and severe,and different trading strategy inclines including momentum and reversal.Moreover we strictly define the third-order SD criteria of non-decreasing concave utility function,S-shaped utility function and reverse S-shaped utility function.The DSD framework can contain the SD criteria with the different risk preferences in various segments.Researchers can build new SD criteria with high-order and multi-segment.At the same time,according to the restrictions of utility function,it can get most abundant information about risk decision-making of investors by testing SD criteria from lower-order to higher-order gradually.Extending the test of Post(2003),Post and levy(2005),we can give the DSD framework of portfolio test.Based on the Bootstrap method,the paper proposes Level Adjust Mean Bias(LAMB)statistic which adjusts biases by means with various magnitudes for different confidence level adjusted,and also put forward Entire Distance Bias(EDB)statistic which adjusts biases by entire distance deviations directly.Through the numerical simulation,it compares the powers and sizes of the above two statistics, Mean Bias(MB)statistic of Post and Levy,(2005),approximate statistic of Post and Vilet (2006).The results show that the approximate statistics may lose identification ability The probability of Typeâ…¡error is a bit high for the MB statistic,even when the sample size is increased.The test power and size of LAMB and EDB statistics is better. Especially when the sample size is small,the probability of Typeâ…¡error for the two statistics is also less than the confidence level.With Wind China A indices,we do a empirical test to the aggregate preference of China's stock market.At 5%confidence level,it refused the assumption that Cnina's stock market obeys traditional expected utility.It also declines that the market conforms to the Markowitz utility function,but accepts the S-shaped utility function which is in accordance with the Prospect Theory,Moreover.it satisfies the restriction that third-order derivative is less than zero in the gain segment corresponding to MGP-S-TSD(Mild and Generic Prospect S-TSD)criteria.This shows that aggregate preference of China stock market are not global risk aversion,but risk averse for gains and risk loving for losses. And for gain the market is mild risk aversion and apt to regard the risk asset as norma, commodity.Further in bull market,the market inclines to positive feed back trading strategy and pursue rising.â…¡.Synthetic Microanalysis.By classifying the investment populations in the stock market,it studies the herd behavior,the characteristics of different populations and the aggregate preference of the market.On this basis it further discusses the ruke of the price volatility and analyzes the policy.It is the performance of the market emergence that to regard herd behaviors as increased and flocks of positive feedback ones.In order to analyze the market,we should do the research starting from the positive feedback behaviors.Because arbitrage is limited and risky,positive feedback traders can exist in the market not only at present but aiso in long-term period.The bubbles of stock price are inseparable from the positive feedback behaviors and the herd behaviors,especially that the herd behaviors enlarge the noise in the market,which may become a kernel in analyzing the bubbles.On the basis of Sentana and Wadhwani(1992)'s positive feedback model test.we find that there exists significant positive feedback behavior in the stock market of China.At the same time.when introducing a dummy variable test we could see that the herd behaviors don't change significantly since 2003.Tnis result indicates that the effect to stabilize the market is not very obvious by the institutional investors of Social Security Funds and QFII.Furthermore.through the compartmentalization of state of the rise and fall in the earlier market,this paper extends Sentana and Wadhwani(1992)'s positive feedback model,and builds up a Markov State Dependent Model(MSDM)of positive feedback trading to distinct the differences the positive feedback trading behaviors and the herd effect in the market.Empirical researches show that:the positive feedback trading behaviors in the market of China are dependent on the multi-phase changes in the state of stock price and have the non-symmetry characteristic.That is.there exists the significant positive feedback trading characteristic in the rising process and the herd effect is inspired easily by the continuous rising of the earlier stock price,but the positive feedback characteristic is not obvious when the price is falling.At the same time,the changes of stock price do not obey the random walk process.There is high possibility,that the stock price wilt go up (down)furthermore when the price continuously rising(falling)in the prophase.Based on the De Long(1990a)'s DSSW Model,the paper also makes distinguish between different positive feedback traders and introduce the herd behaviors into the model to research normatively.We take the bankers into consideration and establish a Banker Model.The Banker Model has captured the situation of the market price fluctuations when the market is manipulated by the banker,and revealed the banker's the trading strategy.Moreover,we derive that the existence of positive feedback trades and enough herd effect is the necessary condition for the banker to earn money in the market.Furthermore,we point out that,the prerequisite that there is equivalent herd effect in the market is implied in the situation that stock prices which manipulated by the arbitrager is deviated from the fundamental in DSSW Model.Simplifying the Banker Model,we also establish a non-banker model and take main consideration about the role of the arbitragers to stabilize the price.The analysis indicates that if there is positive feedback trading,the stock price may still have bubbles and the herd behaviors will increase the bubbles further even there is no malicious manipulation. At the same time,assuming that the positive feedback trades wilt go on persistently when the liquidation is infinite,if there are the herd behaviors,the changes of stock price will be showing complex characteristics.And when the herd effect is large,the "invisible hand" of the market will have no use and market failure is occurred.According to stock price changes of China,we analyze the bubbles in the market from 2001 to 2007 by using the banker model and the non-banker model.During 1997 to 2004.the government policy has good intention indeed,but in practice it has been used by the speculators to meet their trading manipulation in the stock market of China.In a period of time.under the transactions manipulated by the banker.many investors in the market make investment decisions in line with not the fundamental but the public information.Through the transactions,bankers can promote the market price and stir up market sentiment to form herd behaviors of positive feedback,thus stock prices are elevated and are deviated from the fundamental at every turn.Overall.by 2006 the market environment has improved significantly.To a large extent. the price bubble in 2007 results from the investor's bulish views on the market under much good news,and the rising of prices have strengthened investors' expectations. Moreover,Stock Index Futures,the Olympic economy,the reorganization and merger,and the revaluation - these brilliant themes continuously provided the impetus for the market to rise.and then formed a positive feedback loop for the rising price and triggered sustained herd effect.When continual herd behaviors emerge in the market,the price is likely to stay away from the fundamental and the bubble is forming.Under such circumstances,the market doesn't have any price mechanism.The causes leading to price bubbles are different,but the herd behaviors are key factors to understand the changes of price.On this basis,we propose some relevant policy recommendations.Policy making should pay attention to herd behaviors in the market. and give full play to the market mechanism in the principle by the guideline of doing what should be done,while leaving others alone.The stock market needs "in order to selective". On the one hand.the Government should not intervene in the stock market excessively but intervene in the market as little as possible,especially regular intervention,because this can get certain valuable public information which can affect the decision of investors and make the functions of the market not fully running.On the other hand.market sentiment should be more concerned,which should be fine-turning appropriately by the government when necessary.We introduce new policies not only from the macro-economic but also the situation of the micro main body of financing and investing. Pro-cyclical policies must be introduced carefully,because these policies' signal will be exploited by speculators more easily;Moreover,these may also strengthen the momentum of price,simulate continuous herd effect and lead to market failure.When the market is changing constantly and the herd effect is obvious,we may wish to consider reverse fine-tuning of the market.Based on the obvious tendency of pursuing rising in China's stock market,we should focus on making corresponding policies when the market price is continuous rising.In the long run,the stock price can't be completely divorced from the real economy for the existence of the parasitic characteristic.It is the 'invisible hand' that can take order of the stock market and keep the eunomy and stabilization.In the short term,as the virtual of stock,the market behaves positive feedback characteristic and herd effect in the role of self-organization.Such behaviors of investors lead to a sharp short-term fluctuation and different levels of over-reaction and under-reaction in the stock market.So the market is high-risky.The pursuing-rising trend in China's stock market is evident,and the market will be not orderly or stable when the herd behavior is inspired.This influence won't be eliminated in the short term,even make the market price mechanism failure.In order to minish the price bubble and restore the market mechanism,the herd behavior which reflects emergence should be attended to closely.
|
PDF Full Text Request