News effects and structural shifts in price discovery in Hong Kong

This paper examines two questions about the temporal stability of the price discovery relationship using data from Hong Kong. We first study the extent to which abnormally large returns (positive or negative) to securitized real estate are transferred to the returns in the direct real estate market. Our regressions show that price discovery is much reduced in the period following a news event. They show that the estimate of the long run price discovery effect also is reduced once we control for news effects. The second question we examine is whether the price discovery relationship is stable over time. The evidence appears to indicate that the post-1994 period was different from the preceding period. The change in the strength of the price discovery effect may be linked to a change in banking regulations in February 1994 that limited banks' risk exposure to real estate loans by capping them at 40 percent of total lending or to anti-speculative measures introduced by the government in the second quarter of 1994 to curb speculation in the residential market. Our statistical tests of structural stability give mixed results and are therefore inconclusive. These findings suggest that the size of the price discovery effect depends upon the amount of real estate information embedded in the history of securitized returns. The findings further suggest that the securitized return series itself may be an incomplete measure of the quantum of information.