A flexible model for spatial volatility with an application to the Chicago housing market

Existing volatility models normally emphasize the behaviour of prices in a temporal sense and comparatively few studies have explicitly analysed spatial variation volatility. This paper proposes flexible model for squared returns using Box–Cox transformation that includes linear log-linear forms as special cases, thus providing unified framework simultaneously testing space-varying its functional form. The use is illustrated by substantive application to housing price data US city Chicago. estimation results suggest Chicago show exhibits strong dependence form appropriate. In final model, new practical indicator, called neighbourhood elasticity, proposed determines how one linked surrounding neighbourhoods.