While the 2011–2012 euro sovereign debt crisis, the 2008 global financial crisis and their many smaller predecessors have receded in the rearview mirror
for many market participants, financial crises remain an important subject for academic researchers, and rightly so. In fact, the last several decades have
been characterized by a series of asset bubbles and subsequent localized or global financial crises, and these crises appear to have become ever more
severe over time.
Famous German soccer coach Sepp Herberger once said, “After the match is before the match.” The same can be said for financial markets: After the crisis is
before the crisis. The complication, of course, is that while soccer players usually know exactly when the next match will kick off, the timing of the next
crisis is always uncertain for financial players. All we know is that, eventually, there will be another one.
Last month, I was pleased to join a panel with UC Berkeley Professors Christina and David Romer and IMF Chief Economist Maurice Obstfeld on “The
(Un)Changing Nature of Financial Crises and Their Aftermath” at the Clausen Center Conference on Global Economic Issues at the University of California, Berkeley. In
his presentation at the conference, San Francisco Fed President John Williams focused on the natural (or neutral, or equilibrium) real rate of interest, or
r*. This rate has been a hot topic ever since the minutes of the October FOMC (Federal Open Market Committee) meeting revealed that the Committee had
extensive discussions about models estimating the current level of the time-varying r*, its likely future evolution and the implications for monetary
Virtually all of these models, including the seminal one by John Williams and Thomas Laubach, suggest that r* has declined over time, became negative
during and after the 2008 financial crisis, and has picked up only slightly to around zero more recently. PIMCO’s own enhanced version of the
Laubach-Williams model (run by Josh Davis in our quantitative portfolio management group) confirms this result. In fact, as our colleague Richard Clarida
emphasized in the November 2015 Global Central Bank Focus, “ r*=New Neutral,” a neutral real interest rate of around
zero has actually been PIMCO’s core thesis ever since our May 2014 Secular Forum.
So what does r* have to do with financial crises? Well, in my contribution to the conference, I argued that there is actually a strong connection between
the trend decline in r* and the ever-escalating bubbles and financial crises over the past few decades. My thesis: The decline in r* and the serial bubbles
and crises have a common cause – the global savings glut.
That the decline in r* can be attributed to an excess of desired saving over desired investment isn’t exactly new news. Ben Bernanke and Larry Summers have
been making this point for a long time, and both John Williams in Berkeley and Janet Yellen during her testimony last week to the Joint Economic Committee
of the U.S. Congress referred to global excess savings as an explanation for low (natural) interest rates.
Our September Macro Perspectives discussed
the explanations for how a global excess of desired saving over desired investment has depressed r* over time. The reasons for high desired saving include
demographics, inequality and voluntary or forced savings surpluses in emerging market economies. On the other hand, technology and the progression to a
service economy have reduced desired investment and thus the demand for savings. And with the (planned) supply of savings exceeding (planned) investment,
the interest rate that equilibrates the two at high levels of employment has had to decline.
Inflating asset bubbles
What’s perhaps less obvious is that the global savings glut also helps to explain the occurrence of financial bubbles and subsequent crises of the past few
decades. While this is not the place to recap the entire history of financial crises, it’s worth considering a few facts about the financial crises of the
past quarter century first.
An illuminating starting point is Christina and David Romer’s joint paper, “New Evidence on the Impact of Financial Crises in Advanced Countries,”
presented at the Berkeley conference. The Romers present a new, improved indicator of financial stress and a precise chronology of postwar financial crises
in the advanced (OECD, or Organisation for Economic Co-operation and Development) economies. Their paper finds that there were essentially no episodes of
financial distress, and certainly nothing that would count as a significant crisis, in the 1970s and much of the 1980s. This changed in the late 1980s and
especially the 1990s, a period of extensive financial stress. Examples in the advanced economies include the U.S. savings and loan crisis, Japan’s
financial woes, and banking crises in Sweden, Finland and Norway. And in emerging markets (not covered in the Romers’ paper) there were of course the Asian
crises and the Russian crisis in the second half of the 1990s.
While the crises of the 1990s (with the notable exception of the Asian crisis) were relatively contained with little or no spillover to other countries,
the crises of the past 15 years were different: The 2008 crisis in the U.S. and Europe had strong global repercussions and was the worst in almost a
century. And the 2011–2012 euro area sovereign debt and banking crisis also had strong global spillovers and was so severe that it threatened the existence
of the euro. In short, the evidence suggests a clear trend over the past quarter century toward ever more severe and contagious financial crises with
increasingly persistent and protracted effects on growth and inflation.
As I see it, the global savings glut plays an important role in explaining this evidence. How? Excess savings not only pushed down r* and actual interest
rates but also drove up asset prices and caused serial asset bubbles in equities, emerging markets, housing, credit, eurozone peripheral bonds and
commodities. Whenever a bubble burst, it sparked financial distress and crisis.
In addition, there is a feedback loop between financial crises and the savings glut. This is because a financial crisis and the related destruction of
wealth leads to even higher desired saving (or deleveraging), and because the depressing impact on growth reduces investment and thus the demand for
Moreover, as the savings glut drove down r* and as central banks consequently had to lower official rates further and further in order to prevent an
unwanted tightening of monetary conditions, they got closer and closer to the lower bound for interest rates, which in turn makes it more difficult to
react to financial crises with conventional policy tools. And now that exhaustion has set in almost everywhere for many unconventional policy tools, such
as quantitative easing, there is a significant risk that central banks may not be able to deal effectively with the next crisis.
Long-term risks of a long-term savings glut
If my basic thesis that both the decline in r* and the serial asset bubbles and financial crises can be attributed in large part to the global savings glut
is right, the implications could be quite dire. Despite attempts to make the financial system safer and more crisis-proof, avoiding bubbles and crises will
be very difficult as long as desired saving significantly exceeds desired investment. And overcoming the savings glut in the foreseeable future will be
very difficult, too. This is because the main factors fueling desired saving, such as demographics, inequality, demand in emerging markets for perceived
safe assets, cannot be influenced easily. It’s likely the only viable way out would be a joint effort by the major countries to raise public spending on
infrastructure, education, and more in order to absorb excess savings and raise r*. Alas, the political obstacles to such a policy appear to be
So, in short and thinking long-term, get used to the savings-glut-fueled New Neutral and brace yourself for the next bubble and crisis – even if we can’t
yet foresee when or where it will begin.