This Time Is Different
Carmen M. Reinhart & Kenneth S. Rogoff
FUTURECASTS online magazine
Vol. 11, No. 12, 12/1/09
That old time religion:
Government as well as private
debt - internal as well as
external - can be dangerous. Contradicting decades of Keynesian assurances
to the contrary, that is the main theme of "This Time Is Different: Eight
Centuries of Financial Folly," by Carmen M. Reinhart and Kenneth S. Rogoff.
"Many players in the global financial system often dig a debt hole far larger than they can reasonably expect to escape from, most famously the United States and its financial system in the late 2000s."
"Although private debt certainly plays a key role in many crises, government debt is far more often the unifying problem across the wide range of financial crises we examine."
Debt capital has its uses - indeed, its essential uses - but the temptations for abuse are irresistible and the outcome of abuse ultimately unavoidable.
The authors find general patterns typical of financial crises and economic contractions. These patterns appear in their work and that of other scholars.
Asset inflation can surge above sustainable price levels, providing a false sense of increasing wealth that provides the financial system and its banks with a false appearance of strength and profitability.
"[Large] scale debt buildups pose risks because they make an economy vulnerable to crises of confidence, particularly when debt is short term and needs to be constantly refinanced. Debt-fueled booms all too often provide false affirmation of a government's policies, a financial institution's ability to make outsized profits, or a country's standard of living."
Many classic indicators of financial trouble and economic
contraction are covered by the authors. Surges in housing prices are primary
indicators of a developing banking crisis. They are more reliable than current
account deficits and real stock price fluctuations which produce more false
alarms. However, housing bubbles are less reliable as indicators of currency
crashes. As might be expected, currency crashes are more clearly related to current account
and trade deficits. Currency crashes are also frequently a result of banking crises.
The authors provide a treasure trove of historic data from modern research while candidly acknowledging the lack of precision due to the considerable limitations of the available statistical evidence. This history provides a plethora of precedents that are never quite the same but that are each always in numerous ways analogous to other historic crises and to conditions today.
A highly leveraged financial system is always vulnerable to a
sudden collapse of confidence. The exact triggering factor may be unforeseeable,
so the exact timing of collapse will surprise most participants, but the
ultimate fate of highly leveraged systems is certain and should surprise nobody
(not even Keynesians).
When confidence collapses, depositors can withdraw funds from banks or creditors can refuse to roll over short term debt, forcing financial entities to liquidate illiquid assets at fire-sale prices.
Moral hazard eliminates fear of borrower default and so eliminates market constraints on abuse of credit. The result inevitably is bigger and more dangerous financial bubbles as aggressive participants take ever larger risks.
The history of financial crises is dominated by highly leveraged banks and governments. Other actors, such as government sponsored enterprises,
investment banks and money market mutual funds, are relative recent inventions.
However, they can be just as fragile as banks and just as prone to collapse. When
confidence collapses, depositors can withdraw funds from banks or creditors can
refuse to roll over short term debt, forcing financial entities to liquidate
illiquid assets at fire-sale prices.
Banking, inflation, asset price and exchange rate crises, and sovereign debt default are different types of financial crises that occur repeatedly throughout history. Of course, they often occur together and in clusters.
Prior to the advent of fiat paper currencies, price inflation was predominantly cyclical rather than chronic. Today it is chronic everywhere.
Prior to the advent of fiat paper currencies, price inflation was predominantly cyclical rather than chronic. Today it is chronic everywhere. The authors concede that they do not consider reserve losses or interest rate spikes in their lists of crises. (In the two decades prior to its 1970s devaluations, the U.S. ran through most of its huge hoard of gold reserves as a result of its Keynesian efforts to "obsolete" the business cycle.)
This time is different:
The "This time is different" syndrome is summarized by the authors.
There are numerous examples provided by the authors, including several in the twentieth century that were multinational in extent.
Debt is tempting for politicians as a way of
avoiding hard political decisions about spending and taxing. Some nations and
political systems are more prone to succumbing to this temptation than others.
The risk of default rises significantly for debt intolerant states whenever debt ratios rise above as little as 30%.
It can take decades for countries to improve their credit ratings, but they can lose them in a flash.
"Debt intolerance" is what the authors call the loss
of national credit. It can afflict nations that suffer several repeated episodes of
default. This leaves debtor nations vulnerable to
debt crises at levels of debt burden much lower than readily sustained by
financially stronger nations. They suffer a degree of credit fragility.
The authors list 30 nations that have suffered defaults since 1970.
Ecuador has had three, Argentina, Chile and Peru two each.
Institutional failings are the real cause of debt intolerance. Corruption and governance failings are primary culprits.
Lack of access to international credit markets is a consequence
of debt intolerance. Institutional failings are the real cause. Corruption and
governance failings are primary culprits. Moreover, the benefits of access to
international credit markets may be far less than idealized theory suggests, and
such access may be pro-cyclical, increasing instability rather than being a
factor of stability. The tendency of capital flows to collapse during economic
contractions prevents developing countries from using Keynesian monetary and
fiscal policies to mitigate the impact of such contractions. (This risk also
imposes limits on Keynesian policies in advanced nations.)
The authors cover 66 nations, many of which
were not in existence in 1800 when their main sequence begins.
There have been at least 250 sovereign defaults on external debts
since 1800 and 68 defaults on domestic debts. The totals and details are inexact
because the statistics are murky. Governments have not been forthcoming in
reporting their financial weaknesses. The authors describe the efforts made to
overcome the statistical difficulties. The study of these defaults is "more
akin to archeology than economics."
Creditors depend more on the debtor nation's willingness to repay than on its ability to repay. The default decision will be based on the politics of the payment effort.
More than half the sovereign defaults covered by the authors occurred at debt levels below 60% of GDP. However, the view as a percentage of exports or government revenues provides a different picture.
National default decisions are generally more political than
financial. Creditors depend more on the debtor nation's willingness to repay
than on its ability to repay. The default decision will be based on the politics
of the payment effort. Default is typically a strategic decision that repayment
is not worth the necessary sacrifice.
Confidence is supported by much more than just a desire to borrow increasing sums in the future. The latter reason alone would reduce such borrowing to just a Ponzi scheme.
More important is the need to retain credit and access to financial markets to finance international commerce and meet emergency needs. Fourteenth century England borrowed heavily from Italian financiers for its wars, but was heavily dependent on Italian weavers for its wool exports and Italian merchants for its spice imports. Today, foreign trade is heavily dependent on access to short term credit.
Moreover, nations have a variety of national security arrangements and
other important needs and issues that require them to maintain a reputation for
reliability. They may want to attract foreign direct investment to facilitate
their development. These and other complex reasons can solidly support
confidence that nations will repay their debts. Confidence is supported by much
more than just a desire to borrow increasing sums in the future. The latter
reason alone would reduce such borrowing to just a Ponzi scheme.
Even Russia, in the 1990s, ultimately felt obliged to make token payments on its Tsarist debts.
"Compared to corporate debt, country defaults often lead to much larger recoveries, especially when official bailouts are included."
Usually, sovereign defaults are partial. After much haggling and sometimes considerable delay, creditors manage to get sometimes fairly significant sums back. The restoration of normal international relations and access to financial markets is of significant value. Even Russia, in the 1990s, ultimately felt obliged to make token payments on its Tsarist debts. The authors list incidents of default, but fuzzy statistics leave magnitude of each incident uncertain.
Sovereign defaults generally also involve and exacerbate a collapse of
credit and economic depression within the domestic economy. The government and
its private entities will face higher risk premiums when they do return to the
credit markets. New regimes that replace clearly kleptocratic predecessors may
feel no obligation to repay the creditors who enabled the kleptocratic
Governments can default on domestic
debts denominated in local currency not only by refusal to repay but - more
frequently - by inflation that reduces the value of the debt in terms of
purchasing power. The U.S. and many other nations did this in the 1970s.
Serial defaults of international sovereign debt:
Since 1800, there have been four periods of more than a decade in length during which 40% or more of the independent nations in the authors' study were in a state of default or restructuring.
Each period of relative calm is inevitably followed by a new surge of widespread defaults as reasons are found for the belief that "this time is different."
Waves of sovereign
defaults can periodically sweep the world.
The quarter century prior to WW-I, "the halcyon days of the gold standard," was a marvelous period of worldwide financial stability, with nations in default accounting for only about 2% of world income.
The two world wars with the trade war and Great Depression in between caused over four decades of financial disaster.
When viewed in terms of share of world income, this picture changes sharply.
Until WW-I, defaults predominantly involved smaller nations
accounting for less than 15% of world income during five major spikes and less
than 5% during other times. These peaks of default appear of much
shorter duration than the periods of widespread numerical default. (A nation's
share of world income tends to decline sharply after the outset of default,
accounting for the brevity of these spikes.) The quarter century prior to WW-I,
"the halcyon days of the gold standard," was a marvelous period of
worldwide financial stability, with nations in default accounting for only about
2% of world income.
Inflation, is a form of partial default on ordinary government liabilities. With the advent of fiat currencies, inflation crises unsurprisingly increased in frequency.
Commodity price declines and sovereign defaults are unsurprisingly closely correlated. The former generally precede and act as a precipitating factor for the latter. The global capital flow cycle also correlates well with sovereign default rates.
There is also a close correlation between sovereign debt defaults and inflation crises. Political leaders try to inflate their way out of their problems - kicking the can down the road but invariably only making matters much worse. Inflation, the authors point out, is a form of partial default on ordinary government liabilities. With the advent of fiat currencies, inflation crises unsurprisingly increased in frequency.
Serial defaulters have poor rehabilitation records. Efforts to renegotiate debts are frequently followed by renewed borrowing to levels that expose fragile nations to subsequent default crises. For example, of the 17 nations whose external obligations were restructured in the late 1980s pursuant to the "Brady Plan," debt levels were again rising in 10 of them within three years. Four have since defaulted.
French monarchs were in the habit of executing their major creditors - an extreme form of restructuring.
Between 1300 and 1800, England defaulted perhaps three times on
external debts, France eight times and Spain six times, with Austria, Prussia
and Portugal once each. Spain defaulted another seven times in the 19th century.
French monarchs were in the habit of executing their major creditors - an
extreme form of restructuring. It graduated from the ranks of serial defaulters,
with no defaults on external sovereign debt since 1800. However, it has not been
free from periods of hyperinflation.
Domestic public debt:
The limits of available data constrain
examination of the history of defaults of domestic public debt. There is
little data on consolidated public debt. The authors estimate the ratio of
domestic public debt to total domestic debt and the ratio of long term to short
term domestic debt, and compare market rates to instances of administered rates.
Various methods have been used to direct domestic savings into public debt
Contrary to the beliefs of many modern economists, the authors find that the levels of domestic debt play a significant role in determining the sustainability of foreign sovereign debt.
Domestic debt instruments denominated in dollars or other
foreign hard currencies or otherwise linked to hard currencies are a special
category. In the 1990s, Mexico and Brazil successfully floated dollar-linked
debt as they strove unsuccessfully to avert financial crisis. A century earlier,
Argentina floated pound denominated bonds in London. In the 1960s, Thailand
issued dollar denominated bonds that it successfully serviced.
"[In] the 250 episodes of default on external debt in our database, it is clear that domestic debt loomed large across the majority of them."
The usual practice is to inflate domestic debts away. However,
inflation has many costs. It distorts both banking and financial systems. It is
less effective in diminishing indexed or short term debt. In the 1930s, the U.S.
had to default on its gold clause commitments prior to inflating away much of
its debts. Unsurprisingly, the greatest surge in domestic sovereign debt
defaults coincided with the Great Depression and WW-II. The quarter century
after 1980 saw a conspicuous increase in defaults.
Problems with domestic debt can force nations to default on relatively low levels of external debt. Clearly, domestic debt levels have to be considered when evaluating external debt offerings. Domestic sovereign debt levels are, of course, also a factor in driving inflation rates. The authors convincingly demonstrate the often overlooked role that high long-term debt levels play in inducing major inflation episodes. For this purpose, they examine high inflation as well as hyperinflation episodes.
Indeed, except in Latin America, external debt typically accounted for
less than half of total sovereign debt during an external debt default. Taking
domestic debt into account provides one answer why emerging market states
default at such relatively low ratios of external debt to GDP.
Frequently, emerging market defaults follow closely after a period of prosperity.
Overall economic performance is generally far worse in the three years prior to domestic defaults than to external defaults, although often the two come together. Both the rates of inflation and the decline in GDP are significantly worse for domestic default crises. The difference is even greater in the three years after default. It appears that overt domestic default tends to occur only in times of severe macroeconomic distress.
Of course, there is wide variety in the mix of debts and responses in
The lack of transparency in government accounting is notorious. The U.S. government, for example, has massive indefinite entitlement obligations and even more indefinite off-balance sheet guarantees. Smoke and mirrors accounting practices are widespread. Why global investors do not insist on better historic and current financial data from sovereign borrowers is a major puzzle for the authors.
Periodic banking crises seem to be inherent in the
capitalist system. "Countries can outgrow a history of repeated bouts
with high inflation, but no country yet has graduated from [the tendency to
suffer repeated bouts of] banking crises."
The S&L crisis and the Credit Crunch about two decades
later have shown that the U.S. is still subject to serial banking crises, the
authors point out. Indeed, banking crises show remarkable similarities whether
in wealthy nations or developing nations, whether in large nations or small
nations. Real estate price booms and busts generally correlate with these banking crises.
"In modern financial systems, it is not only banks that are subject to runs but also other types of financial institutions that have highly leveraged portfolios financed by short-term borrowing."
The nature of the banking business, which involves the accumulation of
short term deposits to
channel into longer term loans, is inherently unstable and prone to bank runs.
Whenever confidence in banks is lost, depositors rush to try and withdraw their
funds at a pace that is too fast for the liquidation of sufficient assets to
meet their demands. "In modern financial systems, it is not only banks that
are subject to runs but also other types of financial institutions that have
highly leveraged portfolios financed by short-term borrowing."
The development of financial markets is relatively recent in
developing countries, and the incidence of banking crises has increased with
financial market development. Argentina, Brazil and Mexico have had 9, 11 and 7
respectively. Smaller nations have simply used foreign creditors as their
bankers for much of the last two centuries.
Often, liberalization proceeds initially without adequate regulation and supervision.
The authors set forth some of the characteristics of banking crises.
The authors emphasize that banking crises are generally a phase of,
and amplification of, some larger economic or financial crises and generally not
their fundamental cause.
Real estate cycles appear far less resilient than equity price
cycles. The contraction in housing prices in these cycles typically are long
lasting - for from 4-to-6 years. The magnitude of the contractions is similar in
both advanced and developing nations. Real equity prices, on the other hand,
generally start declining about a year prior to a crisis and regain prior peaks
within three years of a crisis. Japan is a notable exception. There are, of
course, cyclical expansions and contractions in financial sectors related to
real estate and equity price cycles.
Clearly, rich countries remain just as prone to financial crises as emerging market countries, and the characteristics are remarkable similar.
The history of monetary inflation is traced
by the authors back to Dionysius of Syracuse in the fourth century B.C.
"Inflation has long been the weapon of choice in sovereign default."
During their developing stage, modern advanced nations were prone to the same financial problems as modern developing nations.
A list of currency debasement episodes in Europe from 1258
through 1799 prior to the advent of paper currency is provided by the authors. A
second list covers the 19th century. They demonstrate the remarkably steady
decline totaling 50% of the silver content of ten major currencies from 1400 to
the reign of Louis XVI with a major further decline thereafter running through
the Napoleonic Wars. Monetary inflation is clearly not a phenomenon just of the fiat
money era, but the advent of fiat paper currencies has greatly accelerated the
process. The authors emphasize that, during their developing stage, modern
advanced nations were prone to the same financial problems as modern developing
Annual rates of price inflation above 20% were
experienced by every nation studied for significant periods. In the three
centuries from 1500, such rates were experienced at least 4% of the time in
those nations where sufficient data is available for this study.
Since the 2001 global recession, instances of 20% price inflation have been much reduced. The sophistication of central banks has improved, so many are now applying "this time is different" rationalizations to price inflation, too. The authors advise skepticism.
This record continued for emerging nations after 1800. The
Great Inflation decade of the 1970s pushed inflation in Singapore and
Taiwan above 20%. The WW-II and recovery period pushed Hong Kong and
Malaysia briefly above 20%. African states have been wracked by price inflation.
Latin American states are even worse, having suffered price inflation rates
above 20% for more than 10% of their history. Bolivia, Ecuador, Mexico,
Nicaragua and Paraguay have been above the 20% inflation level at least 30% of
| Exchange rate crises go together with price inflation crises in
almost all cases. Again, major wars coincide with or immediately precede
major episodes. However, the last three decades of the 20th century have in some
respects been the worst period going back to 1800 for currency exchange rate crashes.
These three decades equal or exceed even the great war periods. The pain of high
rates of currency devaluation and price inflation forces most monetary
miscreants back into fixed exchange rates or dollarization of their financial
system. It generally takes a long time for confidence to be restored in a
free floating domestic currency, so restoration of fixed exchange rates becomes
a necessary alternative.
The Credit Crunch:
The Credit Crunch is no different than many preceding periods of crisis.
The Credit Crunch is thus not some novel, unprecedented financial
calamity. It is no different than many preceding periods of financial crisis.
The U.S. economy which is at the epicenter of the crisis showed many of the
historic signs of impending deep financial crisis during the preceding years. "Outsized
borrowing from abroad" manifested in huge deficits in international trade
and payments accounts were hardly the only warning signals.
The IMF as late as April 2007 provided assurances that worldwide financial risks were extremely low.
"What could in retrospect be recognized as huge regulatory mistakes, including deregulation of the subprime mortgage market and the2004 decision by the Securities and Exchange Commission to allow investment banks to triple their leverage ratios - - - appeared benign at the time."
Many other nations, including Great Britain, Spain and Ireland, showed similar signs. It is the global character of the Credit Crunch that constitutes the primary similarity with the Great Depression.
There was no shortage of warnings. The authors mention Nouriel Roubini and Brad Setzer in 2004 and Paul Krugman in 2007. (FUTURECASTS warnings began towards the end of 2002.) However, the authoritative voices of government and the economics profession countered that this time was different and advised calm. They offered the following reasoning.
Among such authoritative voices noted by the authors were Federal Reserve officials Allen Greenspan and Ben Bernanke and prominent economists Michael Dooley, David Folkers-Landau, Peter Garber, Richard Cooper, Ricardo Housmann, Federio Sturzenegger, Ellen McGrattan, Ed Prescott, and Marc Gertler. The IMF as late as April 2007 provided assurances that worldwide financial risks were extremely low.
The worst of 18 bank centered financial crises
in advanced nations since WW-II was the 1992 crisis in Japan which lasted a decade. However, by
2009, the U.S. Credit Crunch was already well beyond average in terms of
duration, the decline in real housing prices and equity prices, the
international payments current account deficit, GDP decline, and the increase in
The bust was preceded by all the standard signs, including housing and equity price booms, international payments deficits, a surge in GDP, and rising public and private indebtedness. However, not to worry! These red flags were to be determinedly ignored. Authoritative voices were providing reasons why this time was different.
Typically, instead of acting to take steam out of the overheated financial engine, officials added more fire to the boiler to keep it going longer and more briskly. Mortgage lending standards were reduced and SEC regulatory safeguards were eased for investment banks.
What happens next? The authors turn to crisis history for insights. Mathematical economics, as for so much else, is incompetent for this purpose.
More often than not, since WW-II, the aftermath of severe financial
crises involve deep and prolonged asset market collapses. Housing market
declines average 35% and last for 6 years. Banking crisis aftermaths involve substantial increases in unemployment and
declines in GDP. Average unemployment increases are 7 percentage points during a
four year downward phase. Output decline from peak to trough averages more than
9% during a 2 year contraction.
Equity price declines are far greater, averaging about 56%, but
are of much shorter duration. Equity prices are far more resilient than housing
prices. Stock market recovery tends to precede economic recovery. Nevertheless,
the average peak to trough averages 3.4 years.
Government budget impacts of banking crises are severe. Government debt increases on average 86% in the three
years after a crisis from the combination primarily of reduced revenues and, to a
much lesser extent, the expenses generated by the crisis.
"V-shaped recoveries in equity prices are far more common than V-shaped recoveries in real housing prices or employment."
The Credit Crunch trajectory since 2007 tracked remarkably closely with other severe post WW-II crises.
The collapse of financial systems in September, 2008, made the Credit
Crunch instantly global. Economic activity everywhere was impacted by a lack
of financing. Housing bubbles and equity price booms had developed in many
countries. Also widespread were substantial capital inflows and major increases
in public and private leverage which were subject to Credit Crunch deleveraging.
The collapse of credit markets affected nations worldwide.
There has almost never been a year without a crisis in some Latin American country, with the last three decades of the 20th century especially volatile.
The authors offer some interesting approaches for evaluating
and comparing the severity of different crises. They demonstrate that the Great Moderation low volatility period of
the two decades from 1983 was primarily a phenomenon of the advanced economies.
Developing and undeveloped states of Africa, Asia, Latin America, Eastern
Europe and the former Soviet Union suffered severe crises during that period.
While the Credit Crunch falls well short of the severity of the Great
Depression, it is already the worst global crisis since WW-II. "The
business cycle has evidently not been tamed."
Barriers to effective policy:
reporting leaves much to be desired.
However, the repeated tendency of policy makers to remain determinedly in denial as the crisis rushes upon them is a much more damaging characteristic. "This time is different," they repeatedly insist.
The authors recommend that international institutions push for better
statistics on sovereign and local government debt, off balance sheet
and explicit and implicit guarantee exposure. The Federal Reserve, in
particular, has taken trillions of dollars in opaque assets on its books and has
immeasurable guarantee exposure. More transparency on bank balance sheets should be required for bank
participants in international finance.
Debt levels must be evaluated under the premise that there will be future economic contractions. Assumptions of long periods of prosperity permitting nations to grow there way out of heavy debt burdens are unrealistic.
Price inflation itself is a form of debt default and, at more than de minimus levels, undermines economic stability and growth in many ways.
The authors reemphasize that the domestic debts of governments be considered along with their external indebtedness in evaluating the sustainability of a nation's financial status. Debt levels must be evaluated under the premise that there will be future economic contractions. Assumptions of long periods of prosperity permitting nations to grow there way out of heavy debt burdens are unrealistic.
The persistent temptation to inflate away indebtedness must be
recognized and publicly evaluated. Price inflation itself is a form of debt
default and, at more than de minimus levels, undermines economic stability and
growth in many ways. (See, Understanding
Banking crises are generally of long duration due to the disruption of the financial system. The authors question whether fiscal stimulus can be effective if the financial system is not working in a satisfactory fashion, or whether such stimulus just leaves a government with a much heavier debt burden for the future.
Please return to our Homepage and e-mail your name and comments.
Copyright © 2009 Dan Blatt