Non-Synchronous Trading and The Investigation of Market Efficiency-The Bombay Stock Index

Most tests of weak form efficiency of stock market prices are conducted on aggregate stock indices. Aggregation however introduces measurement biases when the constituent stocks in the index do not trade frequently. Non synchronous lagged trading and the consequent catching up, is likely to show up as spurious autocorrelation, rendering imprecise the inferences drawn from tests of market efficiency. We conjecture that the failure to correct for infrequent trading may account for the market inefficiency often reported in the extant literature on thinly traded emerging markets. As the observed index may not represent the true underlying index value in these markets, there is a systematic bias toward rejecting the efficient market hypothesis. This paper tests the random walk hypothesis for the Bombay Sensitive Stock Index (Sensex), the oldest stock exchange index in Asia. The tests are first conducted on the observed index and then replicated using the estimated true index corrected for infrequent trading. The observed index is corrected for non-synchronous trading using the Beveridge and Nelson (1981) decomposition. The observed index exhibits significant deviations from a random walk, in marked contrast, the corrected index is weak form efficient. Separating out the effects of infrequent trading reduces the likelihood of spurious rejections of the RWH. and weak form efficiency.