Friday, March 10, 2017

Secular Stagnation vs. Financial Cycle Drag

With rates of unemployment and inflation at low levels by historical standards, the central issue for the US economy is slow growth in productivity and the overall economy.  Claudio Borio looks at two of the main hypotheses for the slowdown in "Secular stagnation or financial cycle drag?" which was given as a lecture last week at the National Association for Business Economics annual meeting held in Washington, DC.

Borio is head of the Monetary and Economic Department at the Bank of International Settlements. In turn, BIS was set up in 1930 and its membership is made up of 60 central banks from all over the world. BIS acts as a bank for central banks in certain international financial transactions, and also holds meetings and does research to encourage communication between central banks. Under Borio's leadership, the BIS has been a strong voice expressing concerns over financial cycle and their deleterious effects, so no one who knows the BIS research output will be surprised that he finds financial cycle drag, rather than secular stagnation, the more plausible explanation for slow growth. Here's how he lays out the argument (for readability, citations, footnotes, and references to graphs are omitted):

The secular stagnation hypothesis can be summarised in three propositions. First, the world has been haunted for a very long time, well before the crisis, by a structural aggregate demand deficiency that is likely to persist well into the future and keep growth sluggish. Many factors are typically mentioned in this context, including ageing populations, growing income and wealth inequality, and falling tangible investment owing to technological change. Second, the pre-crisis financial boom (or “bubble”) was the only reason why output reached potential, ie full employment. Third, and more technically, the natural (or equilibrium) real interest rate has been falling steadily and has been negative for some time. Now, the natural or equilibrium interest rate is typically defined as the rate that would prevail if output was at its potential level and hence inflation was stable. So, in plainer language, given the major structural demand deficiency, real (inflation-adjusted) interest rates must be negative in order to ensure that the economy operates at full employment and to avoid a costly deflationary spiral. Such a spiral would arise because, with nominal interest rates stuck at the zero lower bound, falling prices would raise real interest rates, which would cut spending further, which, in turn, would depress output and employment and hence prices, and so on. 
The financial cycle drag hypothesis can also be summarised in three propositions – largely the mirror image of the previous ones. First, the world has been haunted by the inability to restrain financial booms that, once they turn to bust, cause huge and long-lasting economic damage – deep and protracted recessions, weak and drawn-out recoveries, and persistently slower productivity growth. Such outsize financial cycles are best characterised by the joint fluctuations in credit and asset prices, especially property prices, as risk-taking ebbs and flows. And they tend to be much longer than “traditional” business cycles (say, 15–20 years rather than 8–10). Second, the pre-crisis boom actually pushed output above potential and undermined productivity. In other words, it was not even required to achieve full employment. Third, the natural or equilibrium real interest rate is positive and considerably higher than the secular stagnation hypothesis would suggest. There are two related reasons for this. Defining and measuring an equilibrium rate without explicitly considering the build-up of financial imbalances is too narrow an approach. In addition, the global demand deficiency has been overestimated while the role of primarily positive, and benign, secular supply side global factors in driving inflation has been underestimated. ...
The [secular stagnation] hypothesis is quite compelling in some respects, but even a cursory look at the facts raises some questions. The hypothesis was originally developed for the Unites States, a country that posted a large current account deficit even pre-crisis – hardly a symptom of domestic demand deficiency. True, US growth pre-crisis was not spectacular, but it was not weak either – recall how people hailed the Great Moderation, an era of outstanding performance. Likewise, the world as a whole saw record growth rates and low unemployment rates – again, hardly a symptom of global demand deficiency. Finally, recent declines in unemployment rates to historical averages – and, in some cases, such as the United States, close to estimates of full employment – point to supply, rather than demand, constraints on growth. 
At the same time, a number of specific pieces of evidence support the financial cycle drag hypothesis. First, there is plenty of evidence that banking crises, which occur during financial busts, cause very long-lasting damage to the economy. They result in permanent output losses, so that output may regain its pre-crisis long-term growth trend but evolves along a lower path. There is also evidence that recoveries are slower and more protracted. And in some cases, growth itself may also be seriously damaged for a long time. If so, given the GFC’s almost unprecedented depth and breadth, the subsequent evolution of output is not that surprising – although it would have been so for forecasters that did not adjust their “models” to take such patterns into account. 
Second, BIS research has found evidence that financial (credit) booms tend to undermine productivity growth, further helping to explain the post-crisis weakness ... Drawing on a sample of over 40 countries and over 40 years, the data suggest that this happens mainly as a result of a misallocation of resources towards lower-productivity growth sectors, notably construction, and that the impact of the misallocations that occur during the boom is twice as large in the wake of a subsequent banking crisis. The reasons are unclear, but may reflect, at least in part, the fact that overindebtedness and a broken banking system make it harder to reallocate resources away from bloated sectors during the bust. ...  The findings could help explain the faster pace of the long-term decline in productivity growth seen in recent years. 
Third, measures of output gaps used in policymaking now show that output was indeed above potential pre-crisis. ... The reason is simple: the symptom of unsustainable expansion was not rising inflation, which stayed low and stable, but the buildup of financial imbalances, in the form of unusually strong and persistent credit growth and property price increases. 
Or course, one need not make a totally black-or-white choice between the secular stagnation and the financial cycle drag hypotheses. For example, one could believe in secular stagnation, and still feel that it's pretty important to find ways to prevent financial cycles from blowing up into bubbles and crises.