Business tendency surveys are a popular tool for the timely assessment of the business cycle, used by economists and by the public. This article considers survey results in the manufacturing sector in more detail and looks into the question of, whether the analysis of branch results leads to an information gain. The business cycle turning points are identified in the filtered series and average leads to the turning point of industrial production are calculated. In addition to these leads the ratios of the signal variances to the noise variances are calculated to assess the clarity of the signal contained in the indicator series. Apart from assessing the general business cycle course the survey results in manufacturing are often used to forecast moment-to-moment changes of industrial production. Analyses based on wavelets show that the survey balances are useful to forecast the larger scale movements only. Nevertheless, the comparison of out-of-sample forecast errors show that the inclusion of survey results as independent variables in an autoregressive model improves the forecasts.
Business tendency surveys business cycle analysis turning points Grnager causality wavelet cross-correlations