Return Persistence in the Indian Real Estate Market

Abstract

Over the last decade, numerous factors including robust economic growth, population pressure, and the mounting need for office space among growth sectors such as information technology have placed significant upward pressure on Indian realty prices. The easing of government restrictions on foreign investments and venture capital into Indian real estate have provided an additional fillip to the real estate market in the country, and the confluence of such factors appears to have contributed to a speculative bubble in Indian real estate equities in the latter part of the decade. By using this bubble period as a case study, we test for the existence of long memory among real estate equities. For the January 2006-December 2008 period, we employ three self-affine fractal analysis techniques (classical rescaled range, roughness-length, and the variogram/structure function methods) to estimate the Hurst exponent, and find significant evidence of long memory in the Bombay Stock Exchange (BSE) Realty Index. Return persistence is further confirmed by the more powerful Lo’s modified rescaled range analysis (MRSA), which is robust to short-term dependence. In addition to potential regulatory policy implications for this emerging market, our results have ramifications for modeling and forecasting returns, as well as for technical trading rules.