بررسی اثرات عدم تعادل جریان سفارش در استراتژی کوتاه مدت در بازار سهام استرالیا

We use Lo and MacKinlay's [Lo, A.W., MacKinlay, C., 1990. When are contrarian profits due to stock market overreaction? The Review of Financial Studies 3, 175–205.] contrarian portfolio approach to examine the profitability of short-horizon contrarian strategies in the context of the Australian Stock Exchange. The results show that simple contrarian strategies lead to small but still statistically significant profits when applied to daily and intra-day portfolio formation. However, the profits are not sufficient to cover transaction costs for institutional investors. The source of contrarian profits is also analyzed leading to the conclusion that stock market overreaction is found to be the primary source of contrarian profits. We also examine the relation between the degree of return reversal and order flow activity after abnormal price changes. We find that the degree of return reversal is positively related to the level of order flow imbalance. Larger profits are generated from order flow based contrarian strategies when the order flow imbalances are high.

Some empirical studies suggest price movements in the stock market are to some extent predictable based on past price history. This view is largely based on empirical evidence that a contrarian strategy of buying previous losers and selling previous winners generates significant profits. For example, Lehmann (1990) has documented evidence of statistically significant profits using contrarian strategies over weekly returns, even after corrections for plausible transaction costs.1 However, the interpretation of the sources of contrarian profits have been heavily debated in the finance literature. Some research such as Conrad et al. (1997) and Boudoukh et al. (1994) argue that the majority of contrarian profits are due to market microstructure biases like nonsynchronous trading and bid–ask spreads. They find contrarian profits would disappear once these errors are accounted for. Another stream of research hypothesizes contrarian profits come from stock market overreaction. For example, Jegadeesh and Titman (1995a) suggest that investors overreact to information, and consequently a subsequent correction generates negative correlation in stock returns. Contrarian strategies could then be profitable because in the presence of negative autocorrelation, current losers would become future winners and current winners then become losers. The suggestion that stock market overreaction is the only source of contrarian profits has been questioned by Lo and MacKinlay (1990). They have proposed return forecastability across securities is another important source of contrarian profits. They have argued that even when there is no stock market overreaction and an individual security's returns are serially independent, a contrarian strategy can still be profitable simply because of price lead–lag effects among securities. The first objective of this paper is to adopt Lo and MacKinlay's (1990) contrarian portfolio methodology and to investigate both the profitability and the sources of contrarian strategies on daily and hourly horizons in the context of the Australian market. With the majority of prior research based on the US market, there are relatively few papers that have examined contrarian strategies in a non-US context.2 Prior research examining the profitability of contrarian strategies over time scales shorter than weekly returns is also relatively scarce. The second objective is to investigate the role of trading activity and liquidity in explaining return reversals and overreaction. Some researchers have argued that return reversals and overreaction are results of irrational trading behaviour. Another possibility is that as the market lacks sufficient liquidity to dissipate unexpected price pressures, this results in short-term return reversals. We analyze the relation between return reversals and order flow imbalance, which is a measure of liquidity pressure and trading activity first proposed by Chordia et al. (2002), by examining order imbalances following large price changes. Trading and liquidity measures can also convey information that cannot be deduced from share prices. Traders are generally divided into informed and uninformed traders in the market microstructure literature. Since informed traders would want to trade larger quantities when they have valuable information, uninformed traders would interpret excess liquidity demand as an indication of private information. A number of authors have studied whether trading information can help explain stock returns. For example, Blume et al. (1994) show that “volume provides information on information quality that cannot be deduced from the price statistic.” Conrad et al. (1994) find price reversals for heavily traded securities and price continuation for low volume securities, suggesting there is a significant relationship between lagged volume and the current returns of securities. Campbell et al. (1993) also find that “price changes accompanied by high volume will tend to be reversed”. The organization of this paper is as follows. Section 2 describes the contrarian portfolio strategy and presents empirical results addressing contrarian profits. Section 3 analyzes the relationship between return reversal and order flow imbalance by conditioning abnormally large price changes on order flow imbalance. The profitability of order flow based contrarian strategies is discussed in Section 4. Section 5 concludes the paper.

This paper has shown short horizon contrarian strategies that are constructed from 200 securities on the ASX generate small but statistically significant profits. However, after execution costs are considered, these strategies become unprofitable even for large institutional investors. We also find that the degree of return reversal is positively related to order flow imbalance. Our results suggest that market overreaction can be caused by temporary liquidity imbalance. The temporary lack of liquidity causes the price to overshoot, and subsequently moves in the opposite direction once the temporal liquidity imbalance is dissipated. We also investigated the dependence of the profitability of contrarian strategies on the liquidity conditions of each individual security. Specifically, we find contrarian trading during high levels of order flow imbalance generates higher paper profits. A possible extension of this work is to investigate whether the optimal lag of contrarian strategies is dependent on the size of securities, and also on the liquidity conditions. Some securities will react to information faster than others, and some are more resilient to change in liquidity conditions than others.