Working Papers
"The Role of Okun's Law in Forecast Accuracy" Abstract: This paper focuses on one way that relationships between variables could play an important role in forecast accuracy, namely Okun's law. The accuracy of forecasts among the SPF forecasters varies as well as the implicit Okun's law. Okun's law is an important input to many forecasting models. A better understanding of the relationship between the unemployment rate and the real output growth rate improves the accuracy of forecasts. In general, if a forecaster's implicit Okun's law model is approaching to the Okun's law estimated from the realized data, then the forecaster tends to provide better forecasts for real output growth rate, the unemployment rate, and inflation rate. The result implies that heterogeneity in forecaster's implicit Okun's law model is a factor in explaining the variation in cross-sectional forecast accuracy. "Are the Fed's forecasts superior to the private sector's in terms of Okun's law?" Abstract: A massive body of literature compares the Fed forecasts and private sector forecasts in terms of the unbiasedness, efficiency, and rationality. This paper addressed a question: are the Fed forecasts superior to the private sector forecasts in term of Okun’s law? The results show that the Fed's Okun's law and the private forecasters' Okun's law are not significantly different at most horizons. The magnitude of the difference is relatively small. Therefore, people who are interested in the Fed's view about the relationship between the unemployment rate and real output growth rate can utilize the median SPF Okun's coefficient as a proxy. This chapter also compares the forecasters' (both Fed and the private sector forecasters) predicted Okun's law to the actual Okun's law. There are two main findings. First, forecasters' predicted Okun's law and actual Okun's law are not significantly different at most horizons. Second, when a recession dummy is introduced, the actual Okun's coefficient becomes significantly smaller than the forecasters' Okun's coefficient in absolute value during expansions. This implies that forecasters tend to believe that a prosperous economy would lead to a large decline in the unemployment rate during expansionary periods, while the actual data may not show the case. “Do the First Announcements Affect Future Output Growth?" Abstract: This paper investigates the role of the first announcement in future output growth. There is a debate in the literature about how the first announcement affects the future output growth rate. Rodriguez-Mora and Schulstad (2007) argue that the relationship between the first announcement and future growth due to agents' reaction after receiving new information, while Clements and Galvao (2010) argues that the relationship is due to the nature of data revision process. The study first proposes a method to disentangle the ``pure'' news from the first announcement and then builds a simple framework to mimic the behavior of agents after they receive the news. The first announcement is divided into two parts: the expected term and the unexpected term. The expected term is expressed by forecasters’ predictions, and the unexpected term is the ``pure'' news since it is the forecast error from the first announcement. In conjunction with saturation techniques, this chapter finds that forecasters will revise their forecasts by incorporating this news. However, this news has a small impact on the future economy. The most informative observations are picked up by indicators. This result suggests that even though the first announcement is a coordination device as described in Rodriguez-Mora and Schulstad (2007), the “pure” news in it does not affect the future economy. Working in Progress "The Effect of News on Professional Forecasts"
|