I’ve been using the code to perform nowcasting for other countries. While conducting backtesting exercises, I noticed that information incorporated in the last month of the quarter does not seem to have a contemporaneous effect on the GDP estimate for that quarter.
For example, the nowcast for December GDP remains the same whether I include data up to November or up to December (assuming all historical data up to November is identical in both cases).
Could this be a bug in the code?
I’ve been using the code to perform nowcasting for other countries. While conducting backtesting exercises, I noticed that information incorporated in the last month of the quarter does not seem to have a contemporaneous effect on the GDP estimate for that quarter.
For example, the nowcast for December GDP remains the same whether I include data up to November or up to December (assuming all historical data up to November is identical in both cases).
Could this be a bug in the code?