基于Agent的人工股票市场构建及其仿真研究

基于行为金融学和市场微观结构理论的部分研究成果,建立了一个基于Agent的人工股票市场,这在人工股票市场建模领域是全新的尝试。在文中构建的人工股票市场基础上,揭示了市场的信息传递原理,回答了在怎样的条件下市场可以达到弱势有效,同时,揭示了收益率分布尖峰厚尾的原因与其与市场效率的关系,指出了各类交易商在市场中占优的条件。在此基础上,本文将文中模型与中国的股票市场进行了类比,提出完善我国股市的三条建议。     

Exploring issues of market inefficiency by the role of forecasting accuracy in survivability

Whether markets are efficient or not has been broadly discussed in the empirical literature since the efficient markets hypothesis was proposed by Fama and others in the 1960s. Unfortunately, they did not come to a consistent conclusion. Besides, while these studies show whether a specific market is efficient or not, little has been done to explore the issues regarding the degree of market inefficiency. This paper attempts to resolve the puzzle of the inconsistent conclusions in the empirical literature by adopting a bottom-up approach which takes market participants’ interactions and coordination into consideration. By simulating an agent-based artificial stock market, this paper concludes with three main findings. First, agents’ survivability is mainly decided by risk preference, and not forecasting accuracy. Survivors may have diverse forecasting accuracy. Second, because market prices are not decided by agents based on accurate predictions, markets can not be efficient. What may exist is only the difference of the degree of inefficiency between markets. Third, the more relevant to survivability the forecasting accuracy in a market is, the less inefficient the market will be. Therefore, this paper suggests that it may be better to view the divergent empirical results regarding market efficiency as a fact that markets are inefficient to a variety of degrees.     

On the cointegration of international stock indices

This paper investigates the cointegration relationship among a group of international stock indices in light of new developments of econometric methods. Kasa (1992) first documented strong evidence for cointegration relations among five national stock indices, which suggests that there exists a common trend among those stock indices. Using Johansen multivariate cointegration test, we find that his findings are persistent in a sample of longer periods and more countries. In order to investigate whether these results are driven by statistical biases related to the sample size, we apply to our tests the Johansen’s small sample correction factor. The results still point toward the existence of a cointegration relationship but the evidence becomes much weaker. We next examine the empirical patterns emerged from different lag specifications and argue that Kasa’s findings are more likely due to the size distortion in extreme long lag VAR models. Indeed, when we employ a newly developed non-parametric test that does not require estimation VAR models, the null hypothesis of no cointegration cannot be rejected for the original sample of Kasa’s five-country stock indices from 1974 to 1990, nor for the extended period from 1970 to 2003.     

基于EVA的中国纺织业上市公司股价泡沫分析

本文使用基于EVA的企业价值评估方法,对2006年年末以及2007年第三季末中国纺织业上市公司的股价泡沫状况进行了实证研究。通过对沪深两地纺织业上市公司的绝对泡沫以及泡沫度的计算发现,2006年年末中国纺织业上市公司的股票价格两极分化严重：大部分股票价格偏离价值形成泡沫,还有部份股票价格低于价值而被低估。而2007年第三季度末股价泡沫则非常明显,绝大部份公司股价存在泡沫。     

Tender-offer premia as a management signal

In this paper we test whether tender-offer premia for share repurchases can be viewed as an outgrowth of optimizing behavior by managers in a signalling environment. This is in contrast to previous work on the signalling hypothesis which focuses on the stock market's reaction to the announcement of tender offers. Our empirical results indicate that premia are systematically related to the price of the firm's stock and the level of the stock market but are not related to either management compensation, inside holdings of stock or the ratio of shares sought to total shares.     

Macroeconomic implications of mortgage loan requirements: an agent-based approach

It is a well-known fact that the housing market, with its associated mortgage securities, plays a crucial role in modern economies. The recent crisis of 2007, triggered by the U.S. real estate bubble, confirms this key role and suggests the importance of regulating mortgage lending. This paper investigates these issues by designing a housing market with a linked mortgage lending instrument in the Eurace agent-based model. Our results show that the presence of a housing market in the model has relevant macroeconomic implications, driven mainly by the additional amount of endogenous money injected into the economy by new mortgages. This additional money generally helps to support and stabilize aggregated demand, thus improving the main economic indicators. However, if the regulation of mortgage lending is too lax, involving an increase in the debt-service-to-income ratio (DSTI), then the additional supply of mortgages no longer enhances macroeconomic performance, and the stability of the economic system is undermined. Based on a number of recent discussions, a regulation of stock control that targets households’ net wealth (a stock), rather than income (a flow) is designed and analyzed. The results show that regulation of stock control can be combined effectively with DSTI to increase the stability of the housing market and the economy as a whole. Interestingly, the regulation based on stock control also directly affects mortgage distribution among households, avoiding excessive concentration. From a policy perspective, our results suggest that the use of a mild flow control regulation, coupled with a stricter stock control measure, fosters sustainable growth and eases first-time buyers access to the housing market, encouraging homeownership.

     

Information dissemination in an experimentally based agent-based stock market

This paper builds an agent-based model to reproduce the results of an experimental stock market that studies how the market aggregates private information. The aim is to use experiments and agent-based modeling to analyze the trading behavior in experimental stock markets. Using the experimental environment and results, it is possible to formulate a hypothesis about the subjects’ behavior and thereby formalize (algorithmically) the trading behavior in an agent-based model. This may lead to a better understanding of how the market converges to an equilibrium and of the mechanism that allows dissemination of private information in the market.     

Firm–bank credit network,business cycle and macroprudential policy

We present an agent-based model to study firm–bank credit market interactions in different phases of the business cycle. The business cycle is exogenously set, and it can give rise to various scenarios. Compared to other models in this literature strand, we improve the mechanism according to which the dividends are distributed, including the possibility of stock repurchase by firms. In addition, we locate firms and banks over a space and firms may ask credit to many banks, resulting in a complex spatial network. The model reproduces a long list of stylized facts and their dynamic evolution as described by the cross-correlations among model variables. The model allows us to test the effectiveness of rules designed by the current financial regulation, such as the Basel III countercyclical capital buffer. We find that the effectiveness of this rule changes in different business cycle environments and this should be considered by policy makers.

     

Transition, the Evolution of Stock Market Efficiency and Entry into EU: The Case of Romania

In this paper we demonstrate that the measurement of stock market efficiency is an important activity in establishing whether eastern European countries satisfy the Copenhagen Criteria for EU membership. Specifically, we argue that developing an efficient stock market should be an important policy focus for countries with aspirations to join the EU as it helps to demonstrate the existence of a functioning market economy. We illustrate this issue by examining the evolution of stock market efficiency in the Bucharest Stock Exchange from mid-1997 to September 2002. We use a GARCH model on daily price data and model the disturbances using the Student-t distribution to allow for ‘fat-tails’. We find strong evidence of inefficiency in the Bucharest Stock Exchange in that the lagged stock price index is a significant predictor of the current price index. This result is robust to the inclusion of variables controlling for calendar effects of the sort that have been observed in more developed stock markets. The level of inefficiency appears to diminish over time and we find evidence consistent with stock market efficiency in Romania after January 2000.     

Time series properties of an artificial stock market

This paper presents results from an experimental computer simulated stock market. In this market artificial intelligence algorithms take on the role of traders. They make predictions about the future, and buy and sell stock as indicated by their expectations of future risk and return. Prices are set endogenously to clear the market. Time series from this market are analyzed from the standpoint of well-known empirical features in real markets. The simulated market is able to replicate several of these phenomenon, including fundamental and technical predictability, volatility persistence, and leptokurtosis. Moreover, agent behavior is shown to be consistent with these features, in that they condition on the variables that are found to be significant in the time series tests. Agents are also able to collectively learn a homogeneous rational expectations equilibrium for certain parameters giving both time series and individual forecast values consistent with the equilibrium parameter values.     

Volatility clustering and herding agents: does it matter what they observe?

Recent agent-based models have demonstrated that agents’ herding behavior causes volatility clustering in stock markets. We examine economies where agents herd on others, yet they have limited sets of information on other agents to imitate. In particular, we conduct experiments on economies with agents with different levels of information sharing where agents can imitate: (1) the strategies of others but with an error, (2) the strategies of only a fraction of agents, or (3) the strategies of others, but update their parameters only by a proportion. In each experiment we change the likelihood that agents make errors to copy the strategy of others, the fraction of agents to herd, or the proportion of the parameter that agents update, in order to examine the effect of the different degrees of information sharing on volatility clustering. We show that volatility clustering tends to disappear when agents have limited information on the strategies of others, and agents need to imitate the strategy details of others in order to generate the clustered volatility.     

Speculative behavior and the dynamics of interacting stock markets

We develop a simple agent-based financial market model in which heterogeneous speculators apply technical and fundamental analysis to trade in two different stock markets. Speculators׳ strategy/market selections are repeated at each time step and depend on predisposition effects, herding behavior and market circumstances. Simulations reveal that our model is able to explain a number of nontrivial statistical properties of and between international stock markets, including bubbles and crashes, fat-tailed return distributions, volatility clustering, persistent trading volume, coevolving stock prices and cross-correlated volatilities. Against this background, our model may be deemed to have been validated.     

Did Vietnam stock market avoid the “contagion risk” from China and the U.S.? The contagion effect test with dynamic correlation coefficients

This paper examines the Vietnamese stock market with an extension of the recent investigation of risk contagion effects. Daily data spanning October 9, 2006–June 19, 2009 are sourced for the empirical validation of the risk contagion between the stock markets in Vietnam, China, and the U.S. To facilitate the validation of contagion effects with market related coefficients, this paper constructs a bivariable EGARCH model of dynamic condition correlation coefficients. First, we examine whether there are contagion effects when there is a financial crisis in the Vietnamese stock market. Next, we verify whether the contagion risk triggered by the crisis can affect the Vietnamese market and examine which market influences the Vietnamese market the most. We find that compared to the U.S. stock market, the Chinese stock market brings more contagion risk to the Vietnamese market, and these effects gain more significance after the sub-prime mortgage crisis.     

The dynamics of the leverage cycle

We present a simple agent-based model of a financial system composed of leveraged investors such as banks that invest in stocks and manage their risk using a Value-at-Risk constraint, based on historical observations of asset prices. The Value-at-Risk constraint implies that when perceived risk is low, leverage is high and vice versa; a phenomenon that has been dubbed pro-cyclical leverage. We show that this leads to endogenous irregular oscillations, in which gradual increases in stock prices and leverage are followed by drastic market collapses, i.e. a leverage cycle. This phenomenon is studied using simplified models that give a deeper understanding of the dynamics and the nature of the feedback loops and instabilities underlying the leverage cycle. We introduce a flexible leverage regulation policy in which it is possible to continuously tune from pro-cyclical to countercyclical leverage. When the policy is sufficiently countercyclical and bank risk is sufficiently low the endogenous oscillation disappears and prices go to a fixed point. While there is always a leverage ceiling above which the dynamics are unstable, countercyclical leverage policies can be used to raise the ceiling. We also study the impact on leverage cycles of direct, temporal control of the bank׳s riskiness via the bank׳s required Value-at-Risk quantile. Under such a rule the regulator relaxes the Value-at-Risk quantile following a negative stock price shock and tightens it following a positive shock. While such a policy rule can reduce the amplitude of leverage cycles, its effectiveness is highly dependent on the choice of parameters. Finally, we investigate fixed limits on leverage and show how they can control the leverage cycle.     

Selecting the portfolio investment strategy under political structure change in United States

This paper investigated whether stock market returns and volatilities were induced by change of long-term political structure. The empirical study finds that the political change is a crucial variable to DJIA and S&P 500 stock returns, but is insignificant to volatilities. But after the 1987 Crash, the political change has a positive effect on DJIA stock returns, and reduced the risk of DJIA and S&P 500. When political structure change, significant economic policies must submit to political realities and those proposed by previous governments often do not get implemented, resulting in market confusion. But following the increasing the consummation of market structure during post-1987 crash, hence, the political change effect increased DJIA stock returns, and reduced the risk of DJIA and S&P 500, and therefore the investors might be able to make a profit when they took active portfolio positions of DJIA.     

Shifts in volatility driven by large stock market shocks

This paper presents an extension of the stochastic volatility model which allows for level shifts in volatility of stock market returns, known as structural breaks. These shifts are endogenously driven by large return shocks (innovations), reflecting large pieces of market news. These shocks are identified from the data as being bigger in absolute terms than the values of two threshold parameters of the model: one for the negative shocks and one for the positive shocks. The model can be employed to investigate different sources of stock market volatility shifts driven by market news, without relying on exogenous information. In addition to this, it has a number of interesting features which enable us to study the effects of large return shocks on future levels of market volatility. The above properties of the model are shown based on a study for the US stock market volatility.     

Market reaction to large bank merger announcements in oligopolies

This paper evaluates the individual and rival stock price reactions to large bank merger announcements and subsequent regulatory rejection in an oligopoly. The results show that the announcements produce significant positive abnormal returns for the merger candidates. Regulatory obstacles and denial of the proposed mergers produce significant negative returns. Analysis of the rivals’ reactions doesn’t produce consistent significant results. This suggests that the market reactions for the merging banks results are driven by expected increases in efficiencies. The rivals’ reaction is explained by the fact that the market would remain contestable after the mergers since the offered products are homogeneous.(JEL G14, G34)     

The relationship between the smoothing of reported income and risk-adjusted returns

This paper uses risk-adjusted returns for the firms in the S&P 500 to test whether the stock market response to accounting performance measures is related to the smoothness of companies’ reported earnings. Three income models, increasing in their measure of smoothness, test the hypotheses using cumulative average abnormal returns. The results indicate that companies that report smooth income have significantly higher cumulative average abnormal returns than firms that do not. When size is considered, market returns are higher for small companies than for large companies. There is also a significant relationship between the type of industry and income smoothing.     

Evolving Market Efficiency with an Application to Some Bulgarian Shares

With new technically advanced methods and computers at our disposal, the efficient market hypothesis is once again being debated. At the same time, we are witnessing an unprecedented growth in both existing and new financial markets. These new markets are often in economies which have just recently embraced free market economics; we term these stock markets infant markets. Such stock markets are obviously not efficient in allocating the supply of savings to productive capital. We do not test whether or not these infant markets are informationally efficient, but instead examine whether and how they are becoming more efficient. We propose modelling the excess returns of individual securities using a multi-factor model with time-varying coefficients and generalised auto-regressive conditional heteroskedastic (GARCH) errors. If the markets are becoming more informationally efficient or the agents are learning, we would expect this to manifest itself as the time-varying coefficients becoming more stable as time increases. We test our model using data on four Bulgarian shares. First, we estimate an AR(2) model and a GARCH-M(1,1) model for the shares. Then, we estimated our AR(2) model with time varying coefficients and GARCH type errors. We find varying levels of efficiency and varying speeds of movement towards efficiency within our sample of four shares. This revised version was published online in July 2006 with corrections to the Cover Date.     

Order aggressiveness, pre-trade transparency, and long memory in an order-driven market

Recent empirical research has documented that the state of the limit order book influences stock investors' strategies. Investors place more aggressive orders when the same side of the order book is thicker, and less aggressive orders when it is thinner. We conjecture and demonstrate that this behavior is related to long memories of trading volume, volatility, and order signs in stock markets. We investigate our conjecture in two types of artificial stock markets: a transparent market, in which agents observe all limit orders on both sides of the book and order volumes at those prices before they trade; and a less transparent market, in which agents observe only the best five bid and ask quotes with the depth available at these limit prices. The first market structure resembles certain actual stock exchanges in the level of pre-trade transparency, such as the Australian Stock Exchange, NYSE OpenBook, and the London Stock Exchange, whereas the second market structure is consistent with stock exchanges such as Euronext Paris, the Toronto Stock Exchange, the Tokyo Stock Exchange, and Hong Kong Exchanges and Clearing. We demonstrate that our long memory results are robust with different levels of pre-trade transparency, implying that the strategy constructed by the state of the order book is key for explaining long memories in many actual stock exchanges.