[HKIMR Working Paper No.04/2011]
(January 2011)

Abstract
This paper applies the factor model proposed by Giannone, Reichlin, and Small (2005) on a large data set to nowcast (i.e. current-quarter forecast) the annual growth rate of China’s quarterly GDP. The data set contains 189 indicator series of several categories, such as prices, industrial production, fixed asset investment, external sector, money market and financial market. This paper also applies Bai and Ng’s criteria (2002) to determine the number of common factors in the factor model. The identified model generates out-of-sample nowcasts for China’s GDP with smaller mean squared forecast errors than those of the Random Walk benchmark. Moreover, using the factor model, we find that interest rate data is the single most important block in estimating current-quarter GDP in China. Other important blocks are consumer and retail prices data and fixed asset investment indicators.

[관련자료]
학술보고서 : Structural change in China: Economic development still reliant on industry for now

학술보고서 : The Innovative Performance of China’s National Innovation System