Date of Award
Bachelor of Science
Dr. Dean Croushore
This paper builds on recent research utilizing real time datasets in order to assess the forecasting utility of consumer sentiment indexes in the United Kingdom. Academic researchers have consistently found that consumer confidence indexes accurately predict consumer spending in the near term. Few of these examinations, however, have utilized out of sample forecasting and only one has incorporated real time data. In an effort to recreate the exact dataset that is available to economic forecasters in real time, this paper utilizes the recently published Gross Domestic Product Real-Time Database from the Bank of England in order to produce forecasts of consumer spending growth. The results of the root mean forecasting error analysis indicate that the inclusion of a consumer sentiment index in a VAR forecast does not improve the accuracy of the model.
Smith, Bradford M., "Forecasting utility of UK consumer sentiment indexes in real time : do consumer sentiment surveys improve consumption forecasts in real time?" (2009). Honors Theses. 640.