Nowcasting Japan's GDP

Area of Science:

• Economics
• Econometrics
• Macroeconomic Forecasting

Background:

• Accurate real-time economic monitoring is crucial for timely policy decisions.
• Existing forecasting models may not fully capture the nuances of data revisions and real-time shocks.
• Benchmarking against market consensus provides a vital perspective on forecast performance.

Purpose of the Study:

• To backtest a nowcast model for Japan's real Gross Domestic Product (GDP) growth using genuine real-time data.
• To introduce a novel revision analysis method linking nowcast revisions to both new observations and data revisions.
• To benchmark the nowcast's accuracy against market consensus forecasts at monthly intervals.

Main Methods:

• Utilizing genuine real-time data for backtesting macroeconomic nowcasts.
• Implementing a new revision analysis framework to assess forecast adjustments.
• Conducting a comparative analysis against monthly market consensus forecasts.

Main Results:

• The nowcast exhibits forecast accuracy comparable to, but not identical to, the consensus across most monthly horizons.
• The nowcast effectively captured the impact of the 2011 earthquake on car production, unlike the consensus.
• Both the nowcast and consensus showed distinct responses to the COVID-19 pandemic, influenced by the timeliness of "soft" versus "hard" economic indicators.

Conclusions:

• The developed nowcast offers a valuable tool for monitoring Japan's real GDP growth, particularly in its reaction to specific economic events.
• The revision analysis highlights the importance of considering both new data and data revisions for accurate forecasting.
• Differences in response to shocks underscore the complementary roles of nowcasts and consensus forecasts in economic analysis.