OPENING AUCTIONS AND THEIR SEQUENCE DOES IT MATTER?
CMCRC research reports the efficiency of opening prices is associated with the stock- specific open-auction sequence on the ASX. The sequential nature of the Australian opening auction implies that some stocks always open before other. For a stock that opens down sequence, the price-relevant information may be extracted from the trading process of comparable stocks in earlier batches in addition to the preopening order evolution. Investors who submit their orders on late opening stocks can revise their expectation of equilibrium value by observing the order flow information from stocks that commence continuous trading earlier. Australia equity markets have the distinctive feature of a single price auction where stocks open for trading with a delay in alphabetical sequence rather than simultaneously. Our analysis focuses on the learning process that could potentially occur across stocks before opening. Investors who trade stocks in later batches could exploit trading information of the stocks in earlier batches. Our sample includes constituents stocks of the S&P/ASX 500 index at any time from November 2004 to December 2006. We find that the sample stocks are relatively evenly distributed across batches (i.e., A-Z firm names) for the whole sample and industry. Utilising order book data from the Securities Industry Research Centre of Asia-Pacific (SIRCA) we calculate indicative opening prices by replicating ASX’s four principles-based ‘open price’ algorithm. To allow for a meaningful comparison of early opening stocks and late opening stocks, we count back ten minutes from the time when the actual open-auction trade occurs, and create ten one-minute intervals for every individual stock. A procedure of unbiased regression is repeated for every one-minute interval to measure the price efficiency of each stock. We find the mean coefficient estimates of late stocks is significantly higher than that of early ones for every one-minute interval, with strongest results appearing at two, three, and four minutes prior to normal trading hours. The cross-sectional variation of opening price efficiencies disappears when the market approaches continuous trading session. Overall, our results indicate that opening prices of late stocks are more efficient compared to earlier ones. This supports our conjecture that indicative opening prices of late stocks reflect learning from not only the fundamentals of themselves but also the trading information of early stocks. Our study also examines the structural change when ASX switched to an anonymous trading system on November 28, 2005 to evaluate the impact of this exogenous event. We find that the cross-stock learning is less prevalent in an anonymous market. Our analysis is useful for regulators in evaluating the effectiveness of policy making and designing the optimal opening mechanism. Our study is also useful for traders in formulating their trading strategies.