基于计算实验方法分析程序化交易对股票市场的影响

发布时间：2019-04-19 03:36

【摘要】：程序化交易起源自美国70年代的组合交易技术,至今发展已久。它经历过人们的追捧,也曾一度被视为灾难的根源而被带上枷锁。国际学术界对程序化交易的风险争议至今都还能听见声音。过去,学者采用传统金融研究方法来研究程序化交易的风险问题,有一种基于逻辑推断提出的“程序化交易中的组合保险策略在悲观市场情形下将使市场形成一种‘链式崩溃’的效应”说法被许多学者及业内人士接受,但是事实如何却很难定量,因为一样的情形在现实市场中很难再次相同地发生。应用传统的研究方法研究股指期货市场对现实市场价格波动性影响的最常见方法是对比一个现实市场在加入股指期货市场后的价格波动性变化;研究程序化交易对股票市场的价格波动性影响时最常见的方法是以市场中程序化交易策略的资金比重为自变量,市场的价格波动性为因变量分析程序化交易对股票市场价格波动性的影响。这样的方法的确都可以得出较有说服力的研究结果,然而它们都受制于样本的数量,且无法剔除其他因素的影响,因此在不同市场及不同时期的研究中往往得出相悖的结果。这也是过去几十年里关于程序化交易的风险争议,以及股指期货市场究竟是有助于平稳现货市场价格波动还是增加了现货市场的下行风险等问题的争议一直不断的一个主要原因。解决此类复杂性问题的一个好方法是近年来兴起的计算实验方法。本文正是采用计算实验方法,通过建立人工股票市场,进而严格调控各种实验条件的方式来分析程序化交易对股票市场的风险。由于程序化交易中非常重要的组合保险策略及套利策略均依托于股指期货市场,本文在构建的人工股票市场中加入了股指期货市场并简要分析了股指期货市场的加入对股票市场波动性的影响。通过多次对比实验,本文发现程序化交易的加入短期内的确会造成股票市场的异常波动,但长期来看,其影响却是很微小的;股指期货市场的加入,从总体上确实降低了现货市场的价格波动性。此外,本文通过对市场中的所有投资者施加一个共同的价格下降预期,进行一个有组合保险策略和无组合保险策略的对比试验,发现悲观市场下组合保险策略的确会形成一种自我加强的卖空预期,从而加剧市场的崩溃。
[Abstract]:Programmed trading originated from the United States 70's portfolio trading technology, has been developed for a long time. It experienced popularity and was once shackled as the root of disaster. The international academic circles still can still hear the sound of the risk dispute about the procedural transaction. In the past, scholars used traditional financial research methods to study the risks of programmed transactions. It is accepted by many scholars and insiders that the "portfolio insurance strategy in programmed transactions will cause the market to form a 'chain collapse' effect in the pessimistic market situation", which is based on logical inference. But how the facts are difficult to quantify, because the same situation in the real world is difficult to happen again the same. The most common method to study the effect of stock index futures market on the price volatility of the real market is to compare the price volatility of a real market after joining the stock index futures market. The most common way to study the effect of programmed trading on the price volatility of stock markets is to take the proportion of funds of programmed trading strategies in the market as an independent variable. The price volatility of the market is the influence of dependent variable analysis programmed trading on the price volatility of the stock market. However, they are all subject to the number of samples and cannot be removed from the influence of other factors, so the results are often inconsistent in different markets and in different periods of research. This is also the risk dispute over procedural transactions over the past few decades. And whether the stock index futures market will help smooth spot market price fluctuations or increase the spot market downside risks and other issues have been a major cause of controversy. A good way to solve this kind of complexity problem is the computational experiment method which has been developed in recent years. In this paper, we analyze the risk of programmed trading to stock market by establishing artificial stock market and strictly controlling various experimental conditions. Because the combination insurance strategy and arbitrage strategy, which are very important in procedural trading, rely on the stock index futures market. In this paper, the stock index futures market is added to the artificial stock market, and the influence of the stock index futures market on the volatility of stock market is analyzed briefly. Through a number of comparative experiments, this paper finds that the introduction of programmed trading will indeed cause abnormal fluctuations in the stock market in the short term, but in the long run, its impact is very small; Stock index futures market, in general, has reduced the price volatility of the spot market. In addition, this paper carries out a comparative experiment between portfolio insurance strategy and non-portfolio insurance strategy by applying a common price reduction expectation to all investors in the market. It is found that portfolio insurance strategies in pessimistic markets do create a self-reinforcing short selling expectation, thus exacerbating the collapse of the market.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F830.91;F224

【相似文献】