Housing Market Variables and Predictability of State-level Stock Market Volatility of the United States: Evidence From a GARCH-MIDAS Approach

Afees A. Salisu, Rangan Gupta, Oguzhan Cepni

Research output: Working paperResearch

Abstract

This paper utilizes the generalized autoregressive conditional heteroscedasticity-mixed data sampling (GARCH-MIDAS) framework to predict the daily volatility of state-level stock returns in the United States (US), based on the monthly state and national housing price returns. We find that housing price returns generally tend to affect state-level volatility negatively. More importantly, the GARCH-MIDAS model, supplemented by these predictors, outperforms, in a statistically significant manner over short-, medium-, and long-term forecasting horizons, the benchmark GARCH-MIDAS model with realized volatility (GARCH-MIDAS-RV) for 90% of the states, with the performance of state and national housing returns being virtually inseparable. Such superior forecasting performances continue to hold when housing price returns is replaced with housing permits and housing market media attention indexes, suggesting an overwhelming role of housing market variables: traditional and behavioural, in forecasting state-level stock returns volatility. Our findings have important implications for investors and policymakers.
Original languageEnglish
Place of PublicationPretoria
PublisherUniversity of Pretoria
Publication statusPublished - Oct 2023
SeriesWorking Paper Series / Department of Economics. University of Pretoria
Number2023-30

Cite this