Minimize

Welcome!

Is the US slowdown for real?

October 9, 2015

This post takes a look at the numbers behind the overall GDP assessment from the Atlanta Fed. The main focus in inventories as this is predicted to have the biggest drag in Q3.

Click here to visit the posting page over at The Financial Times.

In the aftermath of the supposedly “weak” US employment data published last week, investors seem to have shifted their assessment of the likelihood of the US Federal Reserve tightening interest rates by December — and also of the extent of tightening in the next two years.

Since the data were published, several investment banks’ economics teams have ruled out a December rise. Furthermore, equities have been strong; and the bond market’s implied probability of a 25 basis points rise in the federal funds rate by December has fallen from 76 per cent in mid-September to only about 40 per cent.

Nor is this seen as a minor postponement in the first rate rise. The expected federal funds rate at the end of 2016 implies only two Fed rate hikes in total over that entire period. Clearly, investors increasingly believe that the US economy is now slowing enough to throw the Fed off course.

This big change in market opinion is, frankly, surprising. The rise of 142,000 in non-farm payrolls in September was not all that weak, given the normal random fluctuations in the monthly data. And as John Williams, president of the San Francisco Fed, has pointed out, a slowdown to a monthly rate of increase of under 200,000 was long overdue anyway. Rightly or wrongly, there is little indication so far that important Federal Open Market Committee members share the market’s increased post-jobs-data dovishness.

The crucial question is how much growth in the US has slowed since the middle of the year, and whether this will continue. This is the kind of question that economic “nowcasts” are best suited to answer, so let us examine the recent evidence.

The graph below shows the recent tracking of the gross domestic product nowcasts by the Atlanta Fed — an increasingly popular source of up-to-date US activity estimates — along with the Fulcrum activity nowcast, and the consensus projection from the Blue Chip macroeconomic forecaster.

The Atlanta Fed and the Blue Chip numbers relate to the GDP outcome for 2015 Q3, while the Fulcrum figures relate to underlying growth without specifically seeking to predict the Q3 data release. The numbers suggest that there has been a slight, but only very slight, slow-down in the US economy in the two months since the China policy crisis erupted in early August.

The Blue Chip consensus for the third quarter growth rate in GDP has been reduced by about 0.7 per cent since late July, and the Fulcrum activity estimate has dropped by a similar amount. Both of these sources continue to report that the US economy is growing at or above trend in the face of the commodity shocks that have hit the world economy this year.

The Atlanta Fed GDP nowcast is, however, much weaker than the other two sources; and, in view of the fact that this model essentially nailed the growth rates in both 2015 Q1 and Q2, it needs to be taken seriously. It predicts that the 2015 Q3 GDP growth rate will be only 1.1 per cent, which is well below trend and suggests that the economy has significantly lost momentum since mid-year. If this proves accurate, the Fed might indeed be thrown off course. So why does the Atlanta Fed model come up with such a pessimistic figure?

One of the most useful features of the Atlanta Fed methodology is that it produces forecasts for each of the main expenditure components of GDP, so it is very easy to see how its GDP forecasts are derived. The graph on the right shows its estimates of the major GDP components since early August. It is immediately clear that main source of the low GDP forecast is the very large decline in inventories that the model predicts for Q3. This is expected to reduce the growth rate by 2 full percentage points in the quarter — an unusually large hit, even in a fairly large downturn for inventories.

Other mainstream forecasters also see a sizable drag from inventories in their GDP projections for the third quarter, stemming mostly from the unexpected decline in net exports which has left the manufacturing sector with unwanted stocks of finished goods. But in general these estimates for the inventory drag in Q3 are only about 1.2-1.5 per cent, and even that much seems doubtful given the fairly stable inventory numbers that have been published so far in the quarter.

Why does this matter? Inventory adjustments are, in general, self-correcting, except in the most severe of economic downturns, when they can exacerbate the contractionary forces early in a recession. With the risk of an outright recession still very low, according to the Fulcrum models, it would be surprising if the present inventory correction developed into something more disastrous. Instead, it is quite likely to depress the GDP growth rate for a short while before its effects are reversed.

That, anyway, is what the Fed seems predisposed to conclude. Even on the Atlanta Fed figures, final domestic demand in the US economy has been rising at more than 3 per cent in the past year. This is similar to the pattern seen for much of this year because the beneficial impact of lower oil prices on consumers’ expenditure has largely offset the hit to net exports from the emerging market shock and the rising dollar. The Fed may conclude that this pattern of growth could re-establish itself once the inventory correction is over.

So is the US slowdown for real? Yes, but it is not yet very severe — and some of it is the result of the temporary inventory correction, and some to the rising dollar. Unless it grows worse in the next few weeks, it is unlikely to dislodge the Fed from the path it has now firmly chosen.