Yesterday, the big news for US data was poor post-winter retail sales. I believe US consumption growth still supports a 2 or even 3%ish level of GDP growth. It is actually elsewhere where the US economy has shifted down: inventories, capital expenditure and trade. The US economy is, therefore, now at stall speed, susceptible to a decline in borrow to spend-related consumption. We should take a June rate hike completely off the table now. Even a summer rate hike has to be in jeopardy.
Let’s start with the data first. Yesterday, we got two pieces of data that are relevant. The first, retail sales, is where the greatest attention has been – and I believe wrongly. U.S. retail sales in April were flat compared to March. Economists had expected a 0.2% rebound. March, which rebounded nicely after poor data from December to February, was revised up to 1.1% from the previous 0.9% change over February. Just looking at the data versus expectations this is a wash, then. The only negative is the downshift, one that was largely expected.
The Atlanta Fed, which has a widely-followed nowcast, lowered its estimate for Q2 GDP growth to 0.7% annualized from 0.8%, based on these data. But, here’s the important piece to remember: consumption growth is still robust.
The GDPNow model forecast for real GDP growth (seasonally adjusted annual rate) in the second quarter of 2015 was 0.7 percent on May 13, down slightly from 0.8 percent on May 5. The nowcast for second-quarter real consumer spending growth ticked down 0.1 percentage point to 2.6 percent following this morning’s retail sales report from the U.S. Census Bureau.
Yes, the flat retail sales numbers lowered the Atlanta Fed’s numbers slightly. But that is not the source of weakness in their forecasts because consumption growth in the U.S. is still on track. For example, the BEA’s personal income measure through March 2015 is up 3.8% y-o-y while personal consumption expenditures are up 3.0% in nominal terms. That’s a decent story, which I believe supports a 2 to 3% GDP growth number, everything else equal.
But everything else is not equal. Yesterday, we got inventory data for March. And that was the data point that should have been in the news because it showed business inventories barely rising, pointing further to a decline in Q1 GDP. Inventories were up 0.1% m-o-m after a downwardly-revised contraction of 0.2% in February. Moreover, the inventory to sales ratio is still 1.36, which is relatively high, meaning inventory accumulation is not in the cards going forward without an aggressive uptick in consumption. The April ISM report showed inventories at 49.5 versus 51.5 in March, meaning that we should expect inventory accumulation in April to also be weak. The customers’ inventory subindex in the April ISM was even weaker at 44, down from 45.5. The bottom line here is that an inventory purge looks to have begun.
If we combine the inventory numbers with weak trade numbers and weak capital expenditure numbers, we have a picture of a stall speed U.S. economy, susceptible to a breakdown in the borrow to spend model which has got us where we are today.
First, real wage growth has been weak even as household deleveraging has been limited. Most of the deleveraging we have seen has been accomplished through defaults and foreclosures. Actual nominal levels of household debts are only 4% below peak. And household debt to GDP levels are elevated in comparison to every post-World War 2 cycle except the last one. What this means in a zero rate environment is that there is no numerator relief in the household debt to income metric that will prevent another deleveraging when households experience debt stress. Only the denominator can help through additional personal income and wage growth.
Second, the drop in oil prices has meant that a massive cut in capital expenditure is taking place. I see the fracking and shale oil boom as the business equivalent of borrow to spend i.e. it has been borrow to invest. And the question now is whether those investments were too much and simply goosed GDP growth over the near-term or whether there is long-term a long-term payoff. My sense is that some of the debt in the shale sector will eventually be written down, and the growth during this cyclical upturn will come to be seen as phantom growth, much as the housing boom’s borrow to invest dynamic was.
Third, the strong dollar has already had a negative effect on the US trade balance, subtracting from Q1 GDP. And while the US dollar has declined from its highs, the currency’s strength will still be a drag on GDP growth.
Lastly, there does seem to be a renewed push to cut government spending. This is a wildcard that could potentially be the ‘exogenous’ shock that brings this recovery to its knees. Let’s see how the present Congress’ fiscal plans develop.
Overall, the U.S. economy’s baseline is still in the 2 to 3% range because of consumption growth underpinned by decent personal income growth. This is pulled back significantly by the developments in inventories, capital expenditure and trade. What we would like to see is annual nominal personal income growth accelerating to the 5% range, supporting even higher growth and reversing the negative inventory development. The Fed believes this will occur because unemployment is nearing the 5.0 to 5.2% range it sees as long-term full employment. I don’t agree.
Full employment is lower. There is no sign that wage growth is about to accelerate. And even so, there is no empirical evidence that wage growth causes consumer price inflation to accelerate. With US CPI well below target, there is no reason for the Fed to think that inflation is about to rear its head. Unemployment is a lagging indicator anyway. And so to the degree we are seeing the U.S. now move into a stall speed economy, unemployment would be the last indicator to show weakness. Year-on-year changes in jobless claims are more of a coincident indicator in terms of jobs. And they are still looking good as the average fell to a 15-year low today, which is mildly comforting.
I do think the retail sales number was weak enough to prevent a June hike altogether. The summer is still on the table. So July or September could see rate hikes if the unemployment level falls to 5.2%. But, right now the real economy in the U.S. is looking weak, moving decidedly into stall speed as the borrow to spend dynamic fails before wage growth has kicked in. Moreover, with inventories being purged and retail sales weak, I don’t see where the additional pressure on wage growth can come from that would push up wages to make up for the decline in borrowing to spend activity.
The U.S. economy is not in a recession. But it is weak. And it is certainly weaker than a declining unemployment rate would suggest. The reduction in capital expenditure due to the crash in shale oil plays alone was a drag on the U.S. economy, but combined with a drag from inventories and trade, it makes the U.S. economy increasingly vulnerable to recession. If so-called market death becomes widespread in the oil sector in the second half of the year, the U.S. economy will shift down another notch. At a minimum, there is unlikely to be a revival of US GDP growth as there was in 2014. I would expect long-term yields to decline again as this fact becomes apparent.