Global economy: Back to the future?
This post is a guest contribution by Spyros Andreopoulos of Morgan Stanley.
Mr. Strickland: I noticed your band is on the roster for the dance auditions after school today. Why even bother, McFly? You don’t have a chance; you’re too much like your old man. No McFly ever amounted to anything in the history of Hill Valley!
Marty McFly: Yeah, well, history is gonna change.
With an eye on the current situation, we go back in economic history to investigate slowdowns and double-dips. Does a US slowdown to ‘stall speed’ – as occurred in the first two quarters of this year – presage recession? Is a young US expansion that slows down to ‘stall speed’ likely to end up in recession – more generally, does the US economy double-dip? Are balance sheet recessions particularly prone to double-dips? And what are the catalysts for double-dips?
Our historical, and international, evidence suggests that double-dips do not occur unless brought about by monetary or fiscal policy tightening. This provides a sobering historical perspective on the DM recession risks we flagged recently (see Global Economics: Dangerously Close to Recession, August 17, 2011). While DM monetary policy should remain supportive of the recovery going forward – indeed, become more accommodative – past ECB tightening may have done some damage already. The prospect of meaningful fiscal tightening in several DM economies – as well as the harmful effects of uncertainty surrounding the fiscal stance – mean that the second catalyst could also be falling into place. Related to this, the eurozone debt crisis remains a potential catalyst on both sides of the Atlantic. So, while we are not as optimistic as Marty McFly that history is going to change, we hope that he will be proven right in the end.
‘Stall Speed’ and Double-Dips: Post-1950 US Expansions
From post-1950 US cycles, we conclude that double-dips have occurred only upon Fed tightening. Specifically:
• Slowdowns – expansions arriving at ‘stall speed’ – have ended in recessions 70% of the time (unconditional probability of a slowdown being followed by a recession within a few quarters): in nine out of 13 historical cases.
• Young expansions that ‘stall’ (enter a slowdown phase) are followed by recessions 50% of the time (probability of a slowdown being followed by a recession within a few quarters conditional on the expansion being young): in two out of four historical cases, young expansions that stalled subsequently double-dipped.
• Both of these double-dips, however, have been due to the Fed tightening monetary policy. Hence, no expansion has ever died young but for the Fed inducing the recession.
Double-Dips and Balance Sheet Recessions: The Great Depression and Japan
From expansions following balance sheet recessions, we again conclude that abortive recoveries and double-dips have been the result of policy tightening on either the fiscal or monetary side.
• Although the economy did not double-dip following exit from the Great Depression in 1933 – the post-Depression expansion was long by the standards of the time – it was terminated by premature monetary tightening in 1936/37 as the Fed increased reserve requirements and the Treasury sterilised gold inflows.
• In 1997-99, Japan experienced a recession relatively soon into the recovery from the long post-bubble recession of the early 1990s. Fiscal policy tightening was the main catalyst.
I. ‘Stall Speed’ and Double-Dips: Post-1950 US Expansions
We ask: have ‘slowdowns’ – the economy reaching ‘stall speed’ as it did in the previous two quarters of this year – always presaged a recession? What if the expansion was only a few quarters old, as is the case now? More to the point, does the US economy have a proclivity to double-dips? To answer these questions, we look at post-1950 US expansions.
Post-1950 expansions: two convenient definitions: There have been 10 expansions in the US economy post-1950 (the current one is the 11th). The average length of an expansion (trough to peak) has been almost exactly five years (20.1 quarters). For the purposes of our analysis, we make the following definitions:
• ‘Slowdown’: Two successive quarters with an average growth rate of at most 1%Q SAAR – i.e., the economy reaching ‘stall speed’.
• ‘Young’ expansion: An expansion that is eight quarters old or less (note that, given the average duration of post-1950 US expansions, our definition of ‘young’ is reasonable as it is less than half of the average duration).
Any resemblance to current circumstances is purely intentional: growth in both previous two quarters, 1Q11 and 2Q11, was less than 1%Q SAAR (0.7% on average), while the expansion as of 2Q11 was eight quarters old – ‘young’ on our definition.
Slowdowns to ‘stall speed’ are common – and come in three different guises: Of the 10 post-1950 expansions, all but one have had at least one slowdown (as defined above); and some had more than one, as there have been 13 such slowdowns in total. A few quick observations:
• Four of these slowdowns (4/13) have been ‘early/mid-cycle’ in the sense that they occurred in the first half or around the middle of the respective expansion (i.e., we consider early/mid-cycle slowdowns as not having been followed by a recession, although of course they eventually all were).
• Another four have been ‘late cycle’ (4/13), which we define as having occurred during the second half of the expansion but were not immediately followed by a recession; although they were followed by a recession after – usually a few – more quarters of expansion.
• And five (5/13) have been ‘pre-recession’, i.e., were immediately followed by a recession (with no intervening quarters of expansion prior to the recession).
Note that the previous two expansions (2Q91 to 1Q01 and 1Q02 to 3Q07) both had an ‘early/mid-cycle’ slowdown; that is, although the economy reached ‘stall speed’, it did not double-dip.
While nine out of 13 (70%) slowdowns ended in recession soon after… Tallying the above would suggest that nine out of 13 slowdowns, as defined above, were followed, within a few quarters (i.e., either immediately or following a few more quarters of expansion), by a recession. That is, 9/13, or 70%, is the unconditional probability of a slowdown to ‘stall speed’ being followed by a recession (immediately or within a few quarters).
…only two out of four (50%) young expansions that reached stall speed double-dipped… Yet, the more relevant question in current circumstances is the probability of a slowdown ending in a recession when the expansion is ‘young’ in the sense defined above (that is, what matters is the probability of a slowdown being followed by a recession, conditional on the expansion being young) – essentially, this is the question ‘does stall speed presage a double-dip?’ The historical record suggests a total of four slowdowns during young expansions, of which:
• Two were ‘early/mid-cycle’, i.e., they were not followed by a recession (1955/56, 2002/03).
• One was ‘late-cycle’, i.e., was followed by a recession within a few quarters (1959 – the recession commences three quarters later).
• One was ‘pre-recession’ (1981), i.e., was immediately followed by a recession.
Hence, two out of four occasions of the economy reaching ‘stall speed’ in a young expansion were followed by a recession within a few quarters. So, young expansions that slowed to ‘stall speed’ double-dipped 50% of the time. But what were the catalysts?
…both because of monetary tightening: It turns out that both these recessions were precipitated by monetary policy. The 1981 recession was – deliberately – induced by the Fed in order to squeeze inflation out of the system (the recession essentially marked the beginning of the ‘Volcker disinflation’). And even the 1960/61 recession is thought by economic historians to have been caused by “the drastic tightening of money that occurred in 1959/60”.
Conclusion: double-dips have only occurred upon Fed tightening: Whenever in post-war US history expansions have died young, the catalyst has been monetary policy tightening. Put differently: double-dips have occurred only when induced by the Fed.
Caveat: stall speed and vulnerability: The economy slowing down to stall speed may mean that it is unusually vulnerable to shocks. In addition to monetary tightening, therefore, the looming US fiscal tightening is a worry (after all, fiscal tightening derailed the Japanese recovery in the late 1990s – see the next section). Moreover, given the weakness of sentiment in the US, a serious deterioration of the European crisis could just be sufficient to cause a double-dip.
‘Stall speed’ models: a technical note: A fascinating recent Fed research paper that employs the ‘stall speed’ concept to forecast recessions has received much attention in the market (Jeremy Nalewaik, Forecasting Recessions Using Stall Speeds). As shown above, given the frequency with which stall speeds precede recessions, quantitative models can be set up that utilise this property to attempt to forecast recessions. At the same time, the existence of early/mid-cycle slowdowns (which have not been followed by recessions) makes such models generate ‘false positives’, i.e., wrong predictions of recession. As the author notes: “the false positive recessions signals…are in the sluggish, jobless recoveries immediately following the 1990/91 and 2001 recessions, suggesting an important caveat to our results. The dynamics at play in the early part of the 1990s and 2000s expansions may have been different than the dynamics at play in the more mature part of those and other expansions, and our stall speed models may have omitted an additional phase of the business cycle that has appeared in recent decades, namely the sluggish, jobless recovery phase. If so, the applicability of these stall speed models may be somewhat limited at certain times, such as in the middle of 2010 when the economy evidently slowed while still in the early stages of recovery”. Precisely because of such limitations of these models, our historical analysis suggests that looking out for catalysts, particularly on the policy side, would be helpful.
II. Double-Dips and Balance Sheet Recessions: The Great Depression and Japan
However, the post-war US historical experience may only provide limited insight into the current situation – indeed, we have argued as much in the past. An economy that has just exited a ‘balance sheet’ recession, with some sectors in deleveraging mode, may exhibit different dynamics to an economy coming out of a garden-variety recession. Hence, it is important to study periods that bear closer resemblance to current circumstances: we look at the expansion following the Great Depression in the US of the 1930s; and at Japan’s experience of the 1990s.
We ask: are expansions following ‘deleveraging’ or ‘balance sheet’ recessions prone to falling back into recession? Put differently, do post-balance sheet recessions have a tendency to double-dip? If so, why?
1. The Great Depression
The economy did not double-dip after the Great Depression… First off, the US economy did not double-dip following its exit from the Great Depression (GD). The GD is dated by the NBER to have begun in 3Q29 (the stock market began crashing in October 1929) and lasted until 1Q33. The expansion that began in 2Q33 lasted until 1Q37. That is, the expansion lasted a total of 16 quarters or four years – much longer than was common for the economy back then: the average length of expansions between 1875 and 1929 was a little more than eight quarters. In addition, the unemployment rate had more than halved by 1937, from a peak of 25.6% to 11.9%, so progress from the depths of the depression had been meaningful.
…though the expansion was cut short… Still, economic historians generally agree that the expansion was cut short – after all, the unemployment statistic mentioned above suggests that the economy was still some way from anything resembling full utilisation of resources. Therefore, it is instructive to see why the economy returned to recession.
…by monetary action: Between July 1936 and May 1937, the Fed doubled reserve requirements for banks in order to counteract the increase of excess reserves in the banking system. In response, banks tightened lending in order to rebuild their desired levels of excess reserves (see “Reversing Excessive Excess Reserves”, The Global Monetary Analyst, October 28, 2009). In addition, around the same time the Treasury began sterilising substantial gold inflows from Europe; as a result, money supply growth stalled. As a consequence of this monetary tightening, soon the economy fell back into recession.
2. Japan’s ‘Lost Decade’ (Once Again)
Japan double-dipped in the 1990s… Japan suffered a long recession between 2Q91 and 4Q93 after its asset price bubble burst; and it did double-dip (indeed triple-dip) subsequently: in 3Q97, after 14 quarters of expansion, the economy fell into recession again (which lasted until 1Q99).
…because of fiscal tightening: According to Bank of England MPC member Adam Posen – a long-time scholar of Japan’s economy – this recovery “could have been sustainable but [was] cut off by macroeconomic policy mistakes”. Specifically, significant fiscal tightening in the form of both spending cuts and tax hikes was the main factor behind the recession that commenced in 1997 (though the Asian Financial Crisis and a worsening credit crunch due to ongoing banking sector problems undoubtedly contributed): measured by potential GDP, the tightening was worth around two percentage points within a few quarters (OECD data).
It is likely that structural issues – such as the failure for a long time to deal with NPLs and an undercapitalised financial sector – contributed to both the fragility of the economy and the length of its post-bubble downturns (our Japan team has written about these issues extensively; see in particular Deflation: Will America and Europe Follow Japan? September 6, 2010, and Japan Economics: How Japan Got Financial Reform Right, November 27, 2008). However, the most important catalyst for the late 1990s recession has been tighter fiscal policy.
III. Conclusion: Back to the Future?
We’ve shown above that economies don’t just double-dip. Something happens to them, or rather, something is done to them: monetary or fiscal tightening tends to be the factor behind aborted recoveries.
Monetary policy is likely to be more stimulative going forward, at least in some DM economies – indeed, even concerted easing by DM central banks is a possibility (see The Global Monetary Analyst: Concerted Easing? September 7, 2011). But fiscal policy is clearly the source of anxiety at present.
In the US, the Fed will not only remain supportive but is expected to loosen monetary policy further (see US Economics: US Forecast Update: More Fiscal Gridlock Ahead? September 6, 2011). Hence, all eyes are on fiscal policy. Uncertainty around the future path of fiscal policy in the context of President Obama’s recent proposal to extend the Social Security payroll tax cut and extended unemployment benefit will likely persist until December. The outcome here is binary, with adoption of the president’s proposals bringing about 0.8% of GDP of net new stimulus; a rejection by Congress would mean expiration of these measures and bring about an automatic fiscal tightening of 1.2% of GDP. And of course the eurozone debt crisis – a fiscal problem – could yet prove a catalyst for a potentially vulnerable US economy.
In the eurozone, while the ECB is poised to reverse course and lower policy rates in the near future (see ECB Watch: Inching Towards a Rate Cut? September 8, 2011), the damage may have already been done; past rate hikes are part of the reason for our team’s downward revision of growth forecasts (see European Economics: Growth Coming to a Standstill, August 17, 2011). On the fiscal front, crisis-induced austerity measures mean tightening in France, Italy and Spain in addition to Portugal and Greece. And all the while, market-induced tightening in the eurozone periphery is doing additional damage as real yields climb not just for governments, but also for households and corporates: on our estimates, Italy, Spain, Portugal and Greece will all be in recession next year.
The shadow of history is long. Let’s hope Marty McFly is right and history is going to change.
Source: Spyros Andreopoulos, Morgan Stanley, September 16, 2011.
More on this topic (What's this?)
U.S. Economy: This Number Suggests the United States is on the Verge of Economic Collapse (Jutia Group, 7/11/15)
Why Inflation More Likely Than Deflation in the U.S. Economy (Jutia Group, 6/8/15)
U.S. Economy Losing Steam in 3 Charts (EconMatters, 4/3/15)