In addition to the wildly popular topic of the market's record low volatility, coupled with speculation what could break the current spell of "endemic complacency" and what its impact would be on asset prices (yesterday we posted a fascinating take by DB's Aleksandar Kocic on the issue of the markets current "metastability"), another question that has fascinated the economic community is when will the next recession hit.
As we discussed last weekend, according to one of the more popular contextual indicators, commercial bank C&I loan creation, the US economy is poised for a contraction as soon as months, if not weeks from now, as C&I loan growth has tumbled from 7% at the start of the year to just 1.6%, the lowest since 2011, and at the current rate is poised to go negative in the very near future.
Loan creation, however, is just one of several "pre-recessionary" early indicators. Another one comes straight from the market, and specifically the Treasury yield curve which traditionally has always flattened sharply, and in some cases turned negative, going into a US recession.
Currently the implied odds of a recession - according to Deutsche Bank's model - are roughly 33%, well below what loan creation would imply.
In other words, there is a disconnect. So in an attempt to "time" the next contraction, and answer when the next recession will emerge and what might cause it, Deutsche Bank this week looked at the broader universe of factors that could precipitate a US economic contraction.
Before we dive into the findings, however, here is where we stand: the current economic expansion is growing old by historical standards. At 32 quarters, it now ranks as the third longest expansion on record over the history of US modern national income accounts.
By early next year it will be number two, and by mid-2019 number one. The German bank thinks the expansion does have a reasonable chance to set a new record, but it could end not long after it does so. As DB's chief economist Peter Hooper writes, US economic expansions do not die of old age, rather they are brought to an end by one of three types of developments (sometimes more than one of these simultaneously): i) domestic economic imbalances, ii) international shocks and iii) Fed tightening.
A quick look at the first two does not reveal any particularly troubling developments. First, a look at the domestic economy:
Sometimes recessions result from overinvestment. The housing bubble and severe overinvestment in residential structures in the mid-2000s was at the heart of the collapse into the great recession and financial crisis of 2007-09. Previously, the decade-long expansion of and overinvestment in business capital during the 1990s, which culminated in the dot-com tech bubble, was the primary cause of the business-spending-led recession of 2001. During both these episodes, private investment as a share of GDP had risen to nearly 20%.
It had risen to similar heights ahead of the recession of 1980 as well. The current expansion is still “young” in terms of real investment spending. US households and firms are underinvesting in both homes and business capital, as the housing vacancy rates continue to test new lows and as the growth of business capital services remains mired at near historic lows, helping to hold down labor productivity growth. While there may be some stress in commercial real estate and autos, these sectors are not large enough to touch off a more generalized downturn in investment spending from relatively modest levels to begin with
Next, Deutsche looks at international developments, where it similarly finds little to be concerned about. We beg to differ, especially when it comes to Deutsche's sanguine assessment of China's "stability."
Global developments have contributed to downturns in the US in the past, though typically in concert with domestic imbalances or overheating. Oil shocks caused by wars in the Middle East factored importantly into US recessions during the 1980s and early 90s. These events generally exacerbated inflation pressures that led to significant monetary tightening. The development of the tight oil market in the US has reduced the likely severity of such an event in the future. At the same time, as emerging market economies, especially China, have grown in importance in the global economy, the sensitivity of the US economy to events abroad may have increased. However, a sudden collapse of activity in China or elsewhere on a substantial scale would seem to be a relatively low probability event for the foreseeable future.
Which brings us to the topic of this piece: will the Fed tighten into a recession, and it is here that Deutsche is mostly concerned, "noting that significant Fed tightening, with the real fed funds rate rising more than one percentage point above its neutral rate, has been at play ahead of almost all of the previous downturns.
To be sure, it is the Fed's hiking cycle - which as shown in the BofA chart below has always ended with some "financial event" - is what recently prompted David Rosenberg to predict that “the yield curve is flat enough that if the Fed raises rates four more times, that’s all it takes. We probably will have a recession next year.”
Going back to Deutsche, and its question "When will the Fed tighten enough to cause the next recession", the bank says that while current market expectations suggest the next recession (or at least the next Fed tightening that would be forceful enough to cause a recession) could still be many years away, the bank "thinks this is too optimistic."
Here's why, courtesy of Peter Hooper:
As we have written recently, the US labor market is on an impressive improving trend with momentum behind it. It is now at full employment and on the verge of moving well beyond full employment. Simple arithmetic says that the unemployment rate could continue to fall into the mid to low 3s or even lower over the next couple years. This would happen if real GDP growth continues at its moderate 2% pace of recent years and the labor force participation rate levels out this year and resumes its secular downtrend next year thanks to the ongoing retirement of the baby boom generation. How far and how fast the unemployment rate moves in that direction would depend on how rapidly labor productivity growth came off the floor it has been on for the past six years. Figure 9 shows a couple scenarios. One has productivity growth recovering a full percentage point from its recent 6-year average, which is roughly consistent with our baseline House forecast, and the second has it recovering only 1/2 of a percentage point. To get a sustained full percentage point recovery in productivity growth may well require a good deal more private capex or supportive government policy than is currently envisioned.
The last time the unemployment rate fell into the mid-3s – in the 1960s – inflation took off. It is worth taking a closer look at this episode. GDP grew at above potential rates through the first half of the 1960s, and unemployment fell fairly steadily from around 7% to below 4%. As it moved down through the 4s, core CPI inflation, which had been very low for the previous seven years, remained below 2% and even declined some in the neighborhood of 1.5%.
Once the unemployment rate fell below 4%, however, inflation began to rise, rapidly. It jumped by 2 percentage points over the next year and another two percentage points to a level of 6% in ensuing years as the unemployment rate remained below 4%. Only after the Fed tightened and unemployment jumped in the recession of 1970 did inflation ease again.
There are of course limits to such historical comparisons. Inflation expectations may now be anchored more firmly by well understood central bank inflation objectives, and technological advances like global sourcing and the dissemination of pricing information via the internet may help to quell inflationary pressures. But we expect that the laws of supply and demand have not been rescinded totally in the labor market, and as the supply of available workers continues to shrink relative to demand with the declining unemployment rate, we would still expect to see at some point a substantial increase in wage and price pressures.
With this thought in mind, a quite plausible scenario is that as the unemployment rate moves down well into the 3s, over the next 18 months, inflation could very well begin to rise significantly at some point by early 2019. Finding itself behind the curve, the Fed would tighten aggressively to catch up. This tightening would lead to the recession of late 2019 or 2020.
How does this scenario translate into the real fed funds rate gap measure of Fed policy stance that we outlined earlier? Our current baseline forecast has the Fed raising rates four times in 2018 and then three more times in 2019 as the real neutral fed funds rate rises to about 0.8% by end-2019. We also expect the Fed's balance sheet to unwind over time, adding to the monetary restraint. Based on Fed Board staff estimates, the 10-year Treasury term premium could rise by about 15bp per year in 2018 and 2019 as the Fed unwinds its balance sheet. Using simulations from the Fed's model of the US economy, this rise in long-term yields is consistent with about two rates hikes per year given historical relationships. We consider an alternative scenario where a significant overshoot of core inflation causes the Fed to tighten more aggressively in 2019, raising rates eight times for a total of 200 bps, in line with the "measured pace" episode of the mid-2000s.
Under our baseline scenario, the Fed's policy stance, accounting for a rundown in the balance sheet, approaches levels of restrictiveness that have historically signaled a recession – our measure of the Fed's policy stance is around 0.8% around end-2019, while historical recessions were typically preceded by an increase to 1% or above. The more hawkish scenario would clearly move the Fed's policy stance to a level that would make a recession likely by late-2019 or 2020.