Noise and Efficient Variance in the Indonesia Stock Exchange

In this study we applied the realized variance based estimator to extract the information from noise and efficient variance from the Indonesia Stock Exchange (IDX). The stocks in the sample are stratified by trading frequency every six months from 2000 to 2007. The standard deviation of noise variance has changed to a lower level after the first half of 2004 implying an improvement of market quality in the Indonesia Stock Exchange. Using Bandi and Russell's (2006) method, it is found that the average optimal sampling frequency to estimate the efficient realized variance is 9-minute. The relation between the standard deviation of the noise variance and the square root of the efficient realized variance is positive and significant. From the information asymmetry hypothesis, the positive and significant relationship implies that the higher uncertainty about the fundamental value of asset increases the risk of transacting with traders with superior information. Furthermore, the variance ratio of the average daily efficient realized variance to the daily open-to-close variance reveals that the private information is a significant trading component in the Indonesia Stock Exchange.