Transcript An Historical Perspective on the Current Crisis

An Historical Perspective on
the Current Crisis
Michael D. Bordo
Rutgers University and NBER
Prepared for: The conference “The Global Financial Crisis: Historical
Perspective and Implications for New Zealand”
Reserve Bank of New Zealand
Wellington
Wednesday June 17, 2009
An Historical Perspective on the Current Crisis
Introduction

The current global financial crisis displays many similarities to crises
observed across the world in the past two centuries.

There are also some novel features.

This talk considers how the present crisis fits into the historical pattern. I
discuss what is familiar and what is novel about recent events with
emphasis on financial innovation, policy errors and private sector behavior.

As background I also briefly survey the historical evidence for many
countries on crisis incidence, duration and depth.

Finally I assess how economic policy has addressed the current crisis and
conclude with some lessons for policy from an historical perspective.
2
2. Historical Evidence on Crisis Incidence
2.1 – Bordo, Eichengreen, Kliengebiel and
Martinez-Peria(2001)

Bordo, Eichengreen, Kliengebiel, and
Martinez-Peria(2001), provide evidence for
a panel of 21 countries for 120 years and
56 countries for the 4 recent decades on:
the frequency, duration and severity of
currency, banking and twin crises across 4
policy regimes.
3
2.2 - Counting Crises




We define financial crises as episodes of financial turbulence leading
to distress – significant problems of illiquidity and insolvency—
among major financial market participants and/or to official
intervention to contain those consequences.
We identified currency crisis dates using “exchange market
pressure” measure and, alternatively, survey of expert opinion. We
use the union of these indicators and an EMP cutoff of 1.5 standard
deviations from the mean.
For banking crises, we adopted World Bank dates for post-1971
period, and used similar criteria (bank runs, bank failures, and
suspensions of convertibility, fiscal resolution) for earlier periods.
Twin crises: banking and currency crises in same or consecutive
years. Crises in consecutive years counted as one event.
4
We distinguish four periods:

1880-1914: prior period of financial
liberalization and globalization

1919-1939: period of exceptional currency,
banking and macro instability


1945-1971: Bretton Woods period of tight
regulation of domestic financial systems and of
capital controls
1973-1998: second period of financial
liberalization and globalization
5
2.3 – Frequency of Crises

We divide the number of crises by the number of country year
observations in each sub-period. (See figure l)

Alarmingly, all crises appear to be growing more frequent




Crisis frequency of 12.2% since 1973 exceeds even the unstable
interwar period and is three times as great as the pre 1914 earlier
era of globalization
Results driven by currency crises, which have become much more
frequent in recent period
This challenges the notion that financial globalization creates
instability in foreign exchange markets since pre 1914 was the
earlier era of globalization
May be due to a combination of capital mobility and democratization
6
7




In contrast, incidence of banking crises only slightly
larger than prior to 1914, while twin crises more
frequent in the late twentieth century
Note that interwar period had highest incidence of
banking crises
Bretton Woods period was notable for the absence of
banking crises due to financial repression
A comparison of crisis frequency between emerging and
industrial countries (See figure 2), suggests that with the
exception of the interwar period, the majority of crises
occurred in the emerging countries
8
9
2.4 – Duration of Crises

We define the duration of crises as the average recovery time. The
number of years before the rate of GDP growth returns to its 5-year
trend preceding the crisis. (See Figure 3 and Table 1)

Recovery time today for currency crises is longer than preceding 2
regimes but shorter than pre 1914

Banking crises last not much longer now than in earlier periods

Recent period twin crises produce the longest slump for emerging
markets

The dominant impression from comparison of pre 1914 and today
for all crises is how little has changed

To the extent that crises have been growing longer, we have simply
been going back to the future
10
11
2.5 – Depth of Crises


We calculate the depth of crises by calculating, over the years prior
to full recovery, the difference between pre-crisis trend growth and
actual growth. (See Figure 4 and Table 1 – which shows cumulative
output loss as a percentage of GDP)
We find that output losses from currency crises were even greater
before 1914 than today. The difference is most pronounced for
emerging countries.

Output losses from banking crises also greater in pre 1914 regime
than today

Twin crises show comparable output losses for today and pre-1914
for emerging markets

Key unsurprising fact is the large output losses in the interwar from
both currency and twin crises
12
13
14
2.6 – Reinhart and Rogoff



Carmen Reinhart and Kenneth Rogoff (2008) have
extended the data base of financial crises in Bordo et al
(2001) to include many more countries, to include
episodes back to 1800 and forward to 2008
The incidence of banking crises they show in Figure 5
(the proportion of countries with crises weighted by their
shares of income) presents a pattern for banking crises
which echoes that in Bordo et al(2001), with the highest
incidence in the interwar and a recurrence of crises since
the early 1970’s.
The recent episode promises to be as severe as the
crises of the 90s
15
16
3. Credit Crunches, Asset Busts,
Banking Crises and Recessions



The current recession was preceded by a massive housing boom in the U.S., U.K.,
and a number of other countries. The boom turned into a bust in 2007 and this led to
serious financial stress, then a credit crunch and now a serious recession which has
become global
There is considerable evidence both recent and historical linking the severity of
recessions to asset price busts and credit crunches
Claessens, Kose, and Terrones (2008) study the behavior of financial variables
around recessions, credit crunches, and asset price busts for a panel of 21 OECD
countries, 1960-2007

They find that house price busts (declines in real prices from cyclical peaks to troughs
of at least 30%) significantly deepen and prolong recessions

They also find that credit crunches (declines of at least 20% from cyclical peaks to
troughs) also worsen and prolong recessions

Real equity price busts (declines of at least 50% from cyclical peak to trough) do not
have as serious effects on recessions.
17

See Table 3 which summarizes their results.
18


Reinhart and Rogoff (2009) present historical evidence on systemic
financial crises.
They include the recent “big five” advanced economy crises (Spain
1977, Norway 1987, Finland 1991, Sweden 1991, and Japan 1992), the
Asian crisis of 1992, Colombia 1998, Argentina 2001 and two historical
episodes Norway 1899 and the U.S. 1929.

They calculate peak to trough changes in real housing prices, real
equity prices and real per capita GDP for each banking crisis episode.

Figure 6 shows that real housing price busts associated with banking
crises last an average 6 years. Housing prices fell on average by 35.5%

Figure 7 shows that real equity price busts, associated with banking
crises, although deeper than housing price busts, last much shorter, for
3.4 years. Stock prices fell on average by 55.9%

Finally, they show in Figure 8 that real per capita GDP in these events
declines on average by 9.3 percent and lasts 2 years.

The current U.S. recession which began in December 2007 may very
well fit this pattern.
19
Figure66
Figure
20
Figure76
Figure
21
Figure86
Figure
22
An Historical Perspective on the Crisis of 2007-2008
Introduction




Today’s events have echoes in earlier big international financial
crises which were triggered by events in the US financial system.
e.g. 1857,1893,1907 and 1929-33.
This crisis has many similarities to those of the recent past but also
some important modern twists.
Crisis started with collapse of US subprime mortgage market spring
2007.
Causes include : government initiatives to extend home ownership,
changes in regulation , lax oversight, relaxing of lending standards,
a prolonged period of abnormally low interest rates, and a savings
glut in Asia.
23
Introduction





Default on mortgages following housing bust spread via elaborate
network of derivatives.
The crisis has spread over the real economy through a virulent
credit crunch.
The crisis and recession has spread from the U.S. to the rest of the
world.
Fed and other CBs responded in a classical way by flooding the
financial markets with liquidity.
Fiscal authorities still to deal fully with the decline in solvency in
banking system following template of earlier bailouts like RFC in the
1930s, Sweden 1992 and Japan late 1990’s.
24
Introduction

I provide an historical perspective on the
current crisis, contrasting the old with the
modern and offer some lessons for policy.
25
Some Descriptive Historical
Evidence




Figure 1 upper panel shows monthly Baa-Ten year Treasury
constant maturity bond spread and NBER recessions : 1953 to
January 2009.
The quality spread can reflect asymmetric information and signal a
credit crunch.
Figure 1 also shows important events like banking crises, stock
market crashes and political events. Lower panel shows the Federal
Funds rate
Figure 2 shows Baa, ten year Treasury composite bond spread from
1921 to January 2009. The discount rate is substituted for the FFR.
26
FIGURE 1: FEDERAL FUNDS RATE AND Baa
AND 10-YEAR TCM SPREAD
Percentage points
S and L crisis and
Continental Illinois
Stock market crashes
Tech bust
Y2K
Penn Central
Kennedy assassination
LTCM
Bear
Stearns
Rescue
Sub-prime
crisis Lehma
fails
Spread
September 11
20
16
Federal Funds Rate
12
8
4
0
1953
1963
1973
1983
1993
2003
27
Sources: Federal Reserve Board and NBER
FIGURE 2: DISCOUNT RATE AND Baa AND
COMPOSITE TREASURY OVER 10 YEARS SPREAD
8
Percentage points
Bear Stearns
Rescue
Banking Crises
S and L crisis and Tech bust
Stock market crashes Penn Central Continental Illinois Y2K Subprime
Crisis Lehma
Pearl Harbor
September
fails
Kennedy assassination
11th
6
4
Spread
2
LTCM
0
1921
1931
1941
1951
1961
1971
1981
1991
2001
1991
2001
16
14
12
Discount rate
10
8
6
4
2
0
1921
1931
1941
1951
1961
1971
1981
28
Sources: Federal Reserve Board and NBER
Some Descriptive Historical
Evidence




The peaks in the credit cycle proxied by the spread line
up with upper turning points of the business cycle.
Many events like banking crises and stock market
crashes occur close to the peaks.
Policy rates peak very close to or before the peaks of the
credit cycle.
In the recent crisis the spreads are higher than the
recession of 2001 and the recession of the early 1980’s.
They are close to those of the early 1930’s.
29
Historical Parallels and Modern
Twists




Many of the financial institutions and instruments caught up in the
crisis are part of the centuries old phenomenon of financial
innovation.
The rise and fall of financial institutions and instruments occurs as
part of the lending booms and busts cycle financed by credit. Credit
cycle connected to the business cycle.
Irving Fisher ( 1933) and others tell the story of a business cycle
upswing driven by a displacement leading to an investment boom
financed by bank credit and new credit instruments.
The boom leads to a state of euphoria and possibly bubble.
30
Historical Parallels and Modern
Twists





A state of overindebtedness leads to a crisis.
A key dynamic in the crisis stressed is information asymmetry
manifest in the spread between risky and safe securities.
(Mishkin, 1997)
The bust would lead to bank failures and possibly panics. This
led to the case for a LLR following Bagehot’s rule.
Countercyclical monetary policy is also an integral part of the
boom-bust credit cycle.
Stock market booms occur in environments of low inflation,
rising real GDP growth and low policy interest rates. As the boom
progresses and inflationary pressure builds up, central banks
inevitably tighten policy helping to trigger the ensuing crash.
(Bordo and Wheelock, 2005)
31
Historical Parallels and Modern
Twists





Similar story in housing (Leamer 2007)
Stock market crashes can have serious real
consequences via wealth effects and liquidity crises.
Housing busts can destabilize the banking system and
depress the real economy.
The recent housing boom was likely triggered by a long
period of abnormally low interest rates reflecting loose
monetary policy from 2001 to 2004 and a global savings
glut.
The bust was likely induced by a rise in rates in reaction
to inflationary pressure.
32
The Non Bank Financial Sector, Financial
Innovation and Financial Crises




The traditional view sees a financial crisis as coming from the liability
side as depositors rush to convert deposits into currency. In recent
decades, since advent of deposit insurance, pressure has come from
the asset side.
Examples include the commercial paper market Penn Central 1970 ;
Emerging market debt in 1982 ; hedge funds LTCM in 1998.
Historical example of 1763 crisis in market for bills of
exchange.(Schnabel and Shinn 2001)
Financial innovation which increased leverage is part of the story with
Penn Central (commercial paper) ; savings and loan crisis (junk
bonds) ; LTCM (derivatives and hedge funds) ; today securitization of
subprime mortgages
33
Modern Twists





Recently risk has been shifted from the originating bank into
mortgage backed securities which bundles shaky risk with the
more creditworthy
Asset backed securities absorbed by hedge funds, offshore
banks and commercial paper.
Shifting of risk didn’t reduce systematic risk and may have
increased risk of a more widespread meltdown.
Another key modern twist is growth of the unregulated shadow
banking system
Repeal of Glass Steagall in 1999 encouraged investment banks
to increase leverage. So did the investment banks going public.
34
Modern Twists


When the crisis hit they were forced to delever,
leading to a fire sale of assets in a declining market
which lowered the value of their assets and those
of other financial firms.
A similar feedback loop in Great Depression
(Friedman and Schwartz, 1963)
35
Prospects for the Emerging
Markets




Financial crises have always had an international dimension.
Contagion spreads through asset markets ,international banking
and international standard.
Baring crisis of 1890 classic historical example of contagion,
central bank tightening led to sudden stops, currency crises and
debt defaults in the emergers similar to 1997-98.
Current crisis has spread from the US to the advanced countries
via holding of opaque subprime mortgage derivatives in diverse
banks in Europe and elsewhere.
36
Prospects for the Emerging
Markets



Current crisis initially was contained to the advanced countries,
which were exposed to subprime mortgages.
Pressure has subsequently spread to emerging markets,
especially those indebted in hard currency to advanced
countries, e.g. Iceland, Hungary, Latvia and Ukraine.
Many Asian and Latin countries initially avoided the crisis
because of defensive measures taken in reaction to the 1990s
meltdown.

As the credit crunch continued and recession spread in US
and Europe, emergers also were affected.

They are facing sudden stops and some may end up
defaulting on sovereign debt.
37
Policy Lessons

Crisis has implications for monetary
policy on key issues of : liquidity,
solvency, stability of real economy.
38
Liquidity





Central banks reacted quickly in the Bagehot manner to unfreeze
interbank market in August 2007.
Run on Northern Rock September 14, 2007 reflected inadequacies in
UK deposit insurance and separation of financial supervison and
regulation from the central bank.
Bear Stearns crisis in March 2008 led Fed to develop new programs
for access to discount window lending.
The Fed changed its tactics away from general liquidity provision via
OMO and leaving distribution of liquidity to individual firms to the
market.
It has developed a large number of credit allocation facilities. Much
of the credit extended was sterilized. However since September 2008,
the monetary base has expanded.
39
Liquidity


Targeted lending exposes the Fed to politicize its
selection of recipients of credit. This is a throwback
to policies followed earlier in the twentieth century.
Also the Fed has greatly reduced its holdings of
Treasury securities. How will it be able to eventually
tighten with its large holdings of unmarketable
mortgage backed securities?
40
Solvency




The Fed and US monetary authorities bailed out firms too
systematically connected to fail : Bear Stearns March 2008,
GSEs July 2008, AIG September 2008 and three times since,
Citigroup (three times) and Bank of America.
Lehman Brothers was allowed to fail in September on the
grounds it was basically insolvent and not as systematically
important as the others.
Had Bear Stearns been allowed to fail, could the more severe
crisis in September/October 2008 have been avoided?
Had Bear Stearns been closed and liquidated it is likely that
more demand for Fed credit would have come forward than
actually occurred.
41
Solvency



When Drexel Burnham Lambert was shut down in
1990, there were no spillover effects.
Assume a crisis in March 2008 like the one that
followed Lehman’ failure in September had
occurred as the Fed feared at the time.
It would have not been as bad as what actually
occurred in September because the Bear Stearns
rescue led investment banks and other market
players to follow riskier strategies than otherwise
on the assumption that they also would be bailed
out.
42
Solvency




This surely made the financial system more fragile than
otherwise so that when the MA let Lehman fail the shock that
occurred and the damage to confidence was much worse.
The deepest problem facing the banking system is solvency.
The problem stems from the difficulty of pricing securities
backed by a pool of assets because the quality of individual
components of the pool varies
The credit market is plagued by the inability to determine
which firms are solvent and which firms are not.
Lenders are unwilling to extend loans when they cannot be
sure that a borrower is credit worthy
43
Solvency




This is a serious shortcoming of the securitization process
that is responsible for the paralysis of the credit markets
The Fed was slow to recognize the solvency problem.
The US Treasury's TARP plan of October 13 2008 based on
the UK plan to inject capital into the banking system only
went part way to help solve this problem. However the
credit crunch continues and bad assets keep piling up as
house prices have continued to decline.
There is precedence to the Treasury's plan with the RFC in
the 1930s, Sweden in 1992 and Japan in the late 1990s.
44
Solvency



The recent Financial Stability Plan to set up a Public Private
Investment Fund whereby private institutions will be
subsidized to acquire troubled assets may work but the
banks may be reluctant to sell them. They fear that this
would further erode their capital base.
The Treasury’s plan to provide capital to the leading U.S.
banks as warranted by a stress test combined with the
(PPIP) may only help to overcome the banking crisis when
insolvent banks have been taken over, their management
replaced and they are recapitalized.
Adopting the Good Bank Bad Bank solution of the Swedes
may still be required.
45
Real Economy




Given the Fed’s dual mandate, attenuating the
recession consequent upon the credit crunch is
now the primary goal of monetary policy.
The recent shift to Quantitative Easing will most
likely attenuate the recession.
The fiscal stimulus will not likely be as potent as
monetary ease. The record of the 1930s in the U.S.
and Japan in the 1990s makes the case.
Once recovery is in sight and inflationary
expectations pick up it will behoove the Fed to
return to its (implicit) inflation target.
46
Conclusion

The monetary authorities in the US and Europe acted
quickly to resolve the liquidity aspects of the crisis.

This is in stark contrast to the Great Depression
when the Fed did virtually nothing.

Expansionary monetary policy and less likely fiscal
stimulus packages may attenuate the downturn.

The solvency aspect still remains.

Without a resolution to the banking crisis as occurred
in the U.S. in the 1930’s, 1990’s, Sweden in 1992 and
elsewhere, and until the solvency problem is solved,
recovery will be limited.
47