先行指标与宏观经济波动预测

Leading indicators and macroeconomic fluctuation forecasting

文中基于非参数协整检验,建立了基于先行指标预测我国经济短期波动的平滑转换自回归模型,以考察不同先行指标对短期内宏观经济波动的预测能力.结果显示:经济合作与发展组织(OECD)先行指标对季节调整后实际产出水平的预测误差基本在1%以内,国家统计局先行指标对国内生产总值(GDP)同比增长率的预测误差基本在5%以内.对2011年第2季度后GDP增长的样本外预测结果表明,两种先行指标的预测结果都与实际GDP增长率很接近,OECD指标的预测精度略优于国家统计局指标.

Based on the nonparametric cointegration test, this paper develops the smooth transition autoregression （STAR） model to forecast the short-run macroeconomic fluctuations using the leading indicators. It is found that the prediction error of actual output levels provided by Organization for Economic Co-operation and Development （OECD）＇s leading indicators after the seasonally adjusted is within 1%, and the prediction error of year-on-year growth rate of Gross Domestic Product （GDP） provided by National Bureau of Statics of China （NBSC） is less than 5%. Forecasting result of GDP growth in 2011 showed that the forecasting results of two leading indicators are very close to the real GDP growth rate, and leading indicator provided by OECD is prior to that provided by NBSC.