Levitating

Levitating. Lev-i-tat-ing. To rise or cause to rise into the air and float in apparent defiance of gravity.

On December 20, 2012 the Bureau of Economic Analysis (BEA) announced the third (“final”) estimate of 2012.Q3 real change in GDP at 3.1 percent, up from the advance/first estimate of 2.0 percent announced the end of October and from the second estimate of 2.7 percent announced the end of November.

Our general view of the U.S. economy is that it has been on a glide path towards another recession – in fact, our forecasts called for the recession to potentially have started by now. As can be seen in the graph below, until these latest revisions, the data of since 2011.Q4 has certainly been consistent with our forecast perspective.



Click on Image to Enlarge

But, as can also be seen, we have been on a similar glide path once before during the current business cycle. Between 2009.Q4 to 2010.Q3 the economy’s momentum was stalling even with over $800 billion of Federal ARRA stimulus (shaded region on the graph above) being fed into its veins. The scheduled reduction in Federal stimulus spending, combined with slowing growth in real disposal income starting in 2010.Q2, seemed to indicate a nearly unavoidable recession was on the near horizon.

In fact, a recession seemed so unavoidable the Economic Cycles Research Institute (ECRI) famously asserted as much in September 2011, “…the U.S. economy is indeed tipping into a new recession. And there’s nothing that policy makers can do to head it off.” As the months passed and “the recession” didn’t arrive ECRI continued to say they ultimately would be proven correct, despite many doubters.

A quick aside: Finally in July 2012 ECRI stated the US economy had already entered recession (and thus “coincidentally” self-confirming their original recession call) and that the National Bureau of Economic Research (NBER), which officially declares the start and end of recessions for the U.S. economy, would ultimately vindicate ECRI’s reading of the economic tea leaves. As ECRI continued to double-down on their recession call a cottage industry sprang up analyzing, and sometimes ridiculing, their proprietary methodology.

Whether ECRI ultimately is proved right or wrong about the U.S. economy being in recession, at least one question screams for an answer: what happened during the latter portion of 2011 so that the economy avoided what seemed to be an almost certain contraction? Perhaps more importantly, could it happen again at this time in 2012 and was the most recent upward revision in 2012.Q3 real GDP change the first sign of the economy gaining altitude?

While there are definitely some parallels between late 2011 and late 2012, we don’t think there are enough to say that lightning strikes twice and the upward revision in 2012.Q3 is the start of a new growth spurt. But let’s briefly examine what happened in late 2011 to see if there are any lessons for today.

There was a VERY significant private inventory buildup from 2011.Q2 to 2011.Q4 (see chart below); over this period private domestic investment ("PDInv") accounted for 77 percent of real GDP change while a more typical proportion is closer to 40 percent (see third chart in this post). Some of this increase in PDInv may have been to rebuild inventories from prior quarters during the business cycle.

Click on Image to Enlarge

As can be seen in the chart above, the quarterly average for personal consumption expenditures ("PCE") between 2009.Q2 to 2011.Q1 was $51.6 billion, 75 percent of the net change in the quarterly average real GDP change during that period (51.6 + 36.5 – 13.5 – 5.7, or $68.9 billion). That’s running pretty hot and no doubt some inventories were thin and some restocking was in order.

Further, some inventory building may have occurred purposely in anticipation of future sales increasing. However, with the benefit of 20/20 hindsight, any business betting on improving future sales was probably disappointed given how events unfolded. As can be seen on the first chart, real disposable personal income (“DPInc”) was actually contracting from 2011.Q2 to 2011.Q4 (negative real change). As a result, as the chart below shows, PCE fell from a quarterly average of $51.6 to $34.1 billion during the period 2011.Q2 to 2011.Q4.

Which leads to the third reason business inventories were increasing during this period: business inventory was piling up because consumers weren’t buying, creating an overhang in inventory. Thus, even though real GDP lurched higher over the last half of 2011, giving off signs that it had pivoted away from contraction, in actuality the economy was choking.

Click Image to Enlarge

By early 2012 the inventory overhang became clear and so domestic investment sharply contracted during 2012.Q1 to 2012.Q3; as can be seen in the chart above, PDInv on an average quarter basis plummeted from $65.6 billion to $20.2 billion.



So, to wrap up learning from what happened in late 2011, increases in PDInv were the principal forces pushing real GDP changes higher. Those increases were likely exacerbated by contractions in real DPInc.



Our observation is that despite the correction in PDInv over the past three quarters, the degree of real GDP growth attributable to PDInv is still high for the entire cycle when compared to the two most recent cycles’ first 13 quarters of growth. As can be seen in the chart below, PDInv has accounted for about 40 percent of real GDP change while, to date in the current business cycle, it accounts for 54 percent.

This suggests to us there will need to be yet more reductions in private investment over the coming quarters or a significant ramp-up in PCE to bring the rates back into balance with each other compared to other business cycles for the economy to truly move into a sustainable growth path.



The best outcome for the economy would be an increase in DPInc that translates into an increase in PCE. However, as the first graph of this post shows, during 2012.Q2 and 2012.Q3 the pace of real DPInc growth was slowing, not expanding. This makes a significant increase in PCE unlikely.

In fact, we would say the chance of this improving soon recently took another hit; the day after the second estimate for 2012.Q3 GDP release the BEA reported the latest statistics on DPInc in the November 30th Personal Income and Outlays release. The new data indicates that in real terms October’s DPInc fell by 0.1 percent (a drop of $12.2 billion) from September’s level while real PCE decreased by 0.3 percent (a drop of $29.5 billion).

Just to keep things in perspective, the average QUARTERLY GAIN in real PCE during the current recovery has been $45.2 billion (see second graph in this post) while the initial estimate for the MONTH of October was a DROP of $29.5 billion). Not exactly a great way to start 2012.Q4 if what is needed is a significant ramp-up in PCE.

In addition, buried down in the November 30 Personal Income and Outlays release, the BEA noted it had revised DPInc back through April 2012 and PCE back through July 2012. The BEA release provided revision details only for the months of August and September, not for all the months revised. For August and September, DPInc was revised $4.9 billion HIGHER in real dollars (-28.6 to -28.2 in August and -2.3 to +2.2 in September) and real PCE was revised $14.4 billion LOWER (+12.8 to -3.2 in August and +38.9 to +40.5 in September).

These revisions were incorporated into the second 2012.Q3 GDP estimate released on November 29.  The revisions explain the downward revision in consumer spending between the advance and second estimate that sent newswires buzzing with the seemingly contradictory messages of upwardly revised third quarter 2012 GDP growth but weaker fundamentals when internal details were examined. This, by way of example, from Reuters:

“It was the fastest growth since late 2011 and much quicker than the 2 percent rate the government estimated last month…Growth in consumer spending, which accounts for about 70 percent of U.S. economic activity, was cut by more than half a percentage point….”

And the full-tale may not yet be told. The blogsite Zerohedge compared October’s revised real DPInc series published on the St. Louis Federal Reserve site to the same data set they had downloaded the prior month and noted the BEA’s revisions resulted in reducing cumulative DPInc between April and September by $40 billion.

Even though these revisions were included as part of 2012.Q3’s second GDP estimate, BEA indicated the DPInc extend back into the second quarter as well. This suggests that 2012.Q2 GDP reported growth could be reduced in subsequent revisions (probably next year). But even though the revised statistics may not be known until next year, the reality they reflect is felt now.

Focusing in on 2012.Q3’s “final” estimate of real GDP change, while the estimate was revised higher, the increase relative to the advance/first estimate was mostly due to more spending on PDInv while PCE was revised lower than originally reported in the advance estimate (see our MacroPulse post on the second estimate for details).  The third 2012.Q3 estimate returned some of what was originally trimmed in PCE between the first and second estimates but only marginally while PDInv essentially unchanged between the second and third estimates (see our MacroPulse post on the third/final estimate for details).

While the increase in PDInv was not as out of balance with PCE as during the latter half of 2011, recall that through the course of this business cycle PDInv is higher than would be considered healthy compared to PCE. Thus, based on the final GDP estimate during 2012.Q3 no material progress was being made on bringing those back into balance.

Further, as was stated earlier, October’s PCE has turned down in real terms, potentially making the situation worse, not better. This is certainly not a good way to start 2012.Q4. While some of the reduction in PCE may be due to the impact of Hurricane Sandy, it is unclear to what extent the storm is affecting the numbers.

Some argue rebuilding and restoration activity in the aftermath of Hurricane Sandy will spur economic activity. However, to make such an argument means the deployment of economic resources to redress the hurricane’s damage is better than the deployment that would have occurred with those same resources if the economy hadn’t experienced hurricane damage. Our view is that, at best, such expenditures are simply redirected and a “push” in terms of the national economy, but not a source of economic stimulus resulting in increased economic activity.

Net exports were revised higher in both the second and third revisions of 2012.Q3’s GDP. Net exports contributed 0.38 percentage points of the 3.1 percent of real GDP growth in the latest 2012.Q3 GDP estimate, an increase from the 0.23 percentage point contribution in 2012.Q2.

Despite the net increase in the contribution between 2012.Q2 and 2012.Q3 we believe the underlying details point to a less positive fundamental: slowing global economic growth. First, U.S. exports fell from a contribution of 0.72 percent in 2012.Q2 to 0.27 percent in 2012.Q3 as trading partners’ economies slowed. However, the reduction in this case was mitigated by declining crude oil prices that caused the value of U.S. imports to fall. The cheaper price of crude oil imports was also due to shrinking global economic growth, the common denominator in this case.

Government spending, which has been mostly retreating throughout this business cycle, was revised higher in the third estimate from the advance estimate. However, the fact that government spending provided a positive contribution to 2012.Q3 real growth is significant when analyzing what lies ahead; a significant share of the government spending during 2012.Q3 was in defense spending.

We suspect the increase in government spending came about from three factors:

1. The end of the third quarter also marked the end of the Federal, and many state and local governments’, fiscal year, when frequently the “use it or lose it” mentality kicks into high gear.
2. With the fiscal cliff looming the potential for significant budget cuts are a real possibility for federal departments, particularly the Defense Department, spurring expenditures in advance of the calendar year deadline.
3. 2012.Q3 was also the homestretch of a presidential election, so any restraint that may have otherwise been imposed on federal spending was probably softened because government spending would serve to lift reported GDP, which could only help the incumbent.

So collecting the pieces together, in terms of government spending, none of these factors are replicable going forward in the near term and so government spending will likely subtract rather than add to GDP moving forward. Slowing global economic activity places any further near-term positive contribution to U.S. GDP in doubt as well. Finally, at some point the economy cannot continue to increase PDInv without consumers buying goods and services. And, consumers will have difficulty ramping up purchases if – as seems to be happening – DPInc growth stalls.

Our conclusion is that 2012.Q3’s rate of increase in GDP change relative to 2012.Q2’s rate of change is not the first step towards a repeat of what unfolded over the last half of 2011 when GDP churned higher. Instead, it is more a matter of levitation, the unlikely convergence of either unrepeatable or unsustainable events, than improvement based on economic fundamentals.