MACRO FACTOR, MARKET VOLATILITY, AND STOCK-BOND CORRELATION: A DYNAMIC MIXED DATA SAMPLING FORECAST

Abstract

This paper explores how macroeconomic fundamentals affect the long-run dynamics in the volatility of and correlation between the stock and bond market. We use China as a laboratory by employing the GARCH-MIDAS and DCC-MIDAS models, which possess proven ability to capture the long-run component of second-moment market performance in a rapidly-growing economy. With actual data and predictivity evaluation, the finding is that the industrial production growth rate accounts for 32.04% of variations in the total conditional volatility for Chinese stocks, and the industrial production growth volatility drives 29.63% of movements in the total conditional volatility for Chinese bonds. Our proposed model is particularly advantageous in long-run volatility forecasts. Moreover, the present study complements the literature concerning the impacts of macro factors on the time-varying correlations between China’s stock and bond markets. We find weak evidence in this regard, probably due to the absence of multi-market macro-strategy investors that prevail in more developed markets.