Trade and inventories are treated differently in the GDP accounts. All other GDP data is the average of the three months data. But trade and inventories are the change from the end of the quarter to the end of the quarter. This is because we not not directly measure output. Rather, we measure consumption and adjust that for the chage in trade and inventories to indirectly estimate output.
But this is a major source of revisions to the data. When the preliminary data is released there is no data for inventories and trade in the final month. That is reported in time for the first revision of the GDP data.
This quarter this should generate significant upward revisions to the reported growth of real GDP in the 4th Q of last year. The real trade data for December was released this morning.
This chart is of the data and the arrow points at the September data, the end of the last quarter.
Note that the November observation –the next to the last data point –was below the September observation. So using the November data it looked like the real trade balance expanded in the 4th Q — a negative for real GDP growth. But now we see that the December data is above the September, so we now know that in the fourth Quarter the real trade balance contracted, a positive for real GDP growth. We only have some nominal data for December inventories, but it strongly implies that the inventory data in the next GDP report will be resived significantly. These two recent data reports imply that when the next real GDP report is released we will see that 4th Q, 2007 real GDP growth was from 0.5 to 1.0 percentage points higher then the 0.6% originally reported.