We construct a novel set of monthly U.S. sector-level economic conditions indices from a small but diverse set of sectoral economic indicators using mixed-frequency dynamic factor models. The resulting indices are driven by a balanced mix of the underlying indicators and display considerable heterogeneity, particularly in the depths, timing and duration of their downturns.
Moreover, the sectoral economic conditions are driven by a common factor that explains most fluctuations in the overall economy and is closely related to aggregate production. Meanwhile, the service-providing sectors are additionally driven by a correction factor that handles the heterogeneous impacts of the financial crisis and covid pandemic. Lastly, sector-level GDP growth nowcasts are constructed, which are found to consistently outperform a simple autoregressive benchmark for almost all sectors, especially during the COVID-19 pandemic.
