Transcript Economic Theory and the Current Economic Crisis

Economic Theory
and the Current
Economic Crisis
Joseph E. Stiglitz
Manchester
October 2008
Current economic crisis has many
lessons for economists

Probably most serious economic disturbance in U.S.
since Great Depression

Most downturns since have been inventory cycles—


Or a result of Central Bank stepping on brakes too hard


Economy recovers as soon as excess inventories are decumulated
Economy recovers as soon as Central Bank discovers its mistake,
removes its foot from brake
This economic downturn is a result of major financial mistakes


Akin in many ways to frequent financial crises in developing
countries
Worse version of S & L crisis


Which led to 1991 recession
Effects spreading to Europe



Partly because of major financial losses in Europe
Partly because of exchange rate adjustments, impact on exports
Both part of globalization
Pathology teaches lessons



Useful in discriminating among alternative
hypotheses
Great Depression led to new insights—into how
periods of unemployment could persist
Led to conclusion that markets are not selfadjusting
 At
least in the relevant time frame
 Role for government in maintaining economy at full
employment
New Lessons

Insights into macro-economics
 Debate
about source of macro-economic
failures
Nominal wage/price rigidities (in tradition of early
Hicks)
 Real wage rigidities (efficiency wage models)
 Imperfect contracting (Greenwald-Stiglitz/Fischer
debt deflation/Minsky, later Hicks)

 These
events are already drawing attention to
Greenwald/Stiglitz/Fisher/Minsky models
Insights into micro-economics
Are markets really as efficient and
innovative as market advocates claim?
 How do we explain these market failures?

 Imperfections
of information?
 Irrationality?
What does traditional finance theory have
to say?
 What advice does economic theory give
about what should be done now?

Neoclassical synthesis



Belief that, once markets were restored to full
employment, neo-classical principles would
apply—economy would be efficient
Not a theorem, but a belief
Idea was always suspect—why should market
failures only occur in big doses
 Recessions tip of iceberg
 Many “smaller” market failures
 Imperfect information
 Incomplete markets
 Irrational behavior
 But huge inefficiencies—e.g. tax paradoxes
This is a micro-economic
failure leading to a macroeconomic problem
Financial markets are supposed to
allocate capital and manage risk
 Misallocated capital
 Mismanaged risk

Innovation



Financial markets did not create risk products that would
have enabled individuals to manage the risks which they
faced
Innovations—tax, regulatory, and accounting arbitrage
Actually resisted innovations that would have made
markets work better





Inflation indexed bonds
GDP indexed bonds
Danish mortgage
Better mortgages that would have managed risk better
Better auctions of Treasury Bills
Incentives


Incentives matter
But incentives in financial markets were distorted
 Focus
only on short term profits
 Asymmetric rewards—20% of gains, none of losses
 Designed to encourage gambling
 Succeeded
 Reliance on non-transparent stock options
encouraged distorted accounting


Easier to increase reported returns than to provide better
products
Opposed reforms for improved accounting
■
Mismatch between private rewards and
social returns (which may be negative)
 Social
returns may have been negative
 Yet as a sector and as individuals they were
generously compensated
 Some 40% of corporate profits

With predictable consequences—behavior
designed to enhance private returns, not
social benefits
Understanding market failure

General Theorem: whenever information is
imperfect or markets incomplete (that is, always)
markets are not constrained Pareto efficient
 Taking
into account costs of collecting and processing
information or creating markets, there are government
interventions that can make everyone better off
 Pecuniary externalities matter

B. Greenwald and J.E. Stiglitz, “Externalities in Economies with
Imperfect Information and Incomplete Markets,” Quarterly Journal of
Economics, Vol. 101, No. 2, May 1986, pp. 229-264.

R. Arnott, B. Greenwald, and J. E. Stiglitz, “Information and Economic
Efficiency,” Information Economics and Policy, 6(1), March 1994,
pp. 77-88.
Application: Securitization

While it enhances opportunities for
diversification, creates new agency
problems
 Resulting
market equilibrium will not in
general be (constrained) Pareto Efficient
 Originator of mortgages did not have sufficient
incentives to screen and monitor

J. E. Stiglitz “Banks versus Markets as Mechanisms for
Allocating and Coordinating Investment,” in The Economics of
Cooperation: East Asian Development and the Case for ProMarket Intervention, J.A. Roumasset and S. Barr (eds.),
Westview Press, Boulder, 1992, pp. 15-38.
Application: Lending
based on collateral






Increased price of houses gives rise to increased lending
Leading to increased demand
Leading to increased prices
Socially excessive lending
Bubbles
Similar problems arise in amount of foreign borrowing
(endogenous exchange rates—Anton Korinek)

J.E Stiglitz and M. Miller, “Bankruptcy protection against
macroeconomic shocks: the case for a ‘super chapter 11’,” World
Bank Conference on Capital Flows, Financial Crises, and Policies,
April 15, 1999.
But this does not fully
explain what went wrong

Hard to reconcile behavior with rationality
 Or
even rational herding behavior
 Many borrowed beyond their ability to repay
 Should have been obvious to both borrower and
lender



But those in financial market were supposed to be financially
sophisticated
Borrowing based on pyramid scheme—belief that prices
would always go up
But how could low income individuals continue to pay more
and more as their real incomes declined?
Models used by banks and rating
agencies flawed and obviously so
Underestimated correlations
 Underestimated systemic risks
 Once in a lifetime events happened every
ten years

 Should
have used fat tailed distributions
rather than lognormal distributions
 But there already were several instances of
failures from using these models—financial
markets didn’t learn

Zero (or negative) non-recourse mortgages are
an option
 Issuing
such options is equivalent to giving away
money
 Giving away money is hard to reconcile with profit
maximizing behavior



Unless there is an underlying belief in the irrationality of
borrower (won’t exercise options)
Or of those to whom one will sell the mortgage
Or part of a scheme of fraud



Design was an invitation to fraud
Conflicts of interest made these more likely
But market participants seemed to ignore this
Standard models and policy
prescriptions used by Central Bank
did not anticipate problem



Indeed, they made it worse
Denied existence of bubble (a little froth)
Encouraged people to take out variable rate
mortgages when interest rates were at record
lows
 With
individuals borrowing to capacity
 And likelihood that interest rates would go up
 Especially with negative amortization and balloon
mortgages, high likelihood of system blowing up

Change in interest rates would lead to defaults, difficulty
refinancing

Denied any ability to ascertain that there
was a bubble
 Econometric
Models to predict economic
vulnerability

J.E. Stiglitz and J. Furman, “Economic Crises:
Evidence and Insights from East Asia,” Brookings Papers on
Economic Activity, 1998(2), pp. 1-114.
 Shiller
 Basic
economics—how could prices keep going up
when real incomes of most Americans were declining

Believed in self-regulation—oxymoron
 And
can’t take into account interactions arising from
banks’ simultaneously following similar policies


Believed that if there was a problem, it would be
easy to fix
Argued that interest rate was too blunt of an
instrument
 If
tried to control asset price bubble, would interfere
with focus on current markets
 But refused to use instruments at its disposal


Regulatory instruments rejected
Even though one Fed governor tried to get them to act

Central banks were focused on models
centered on second order problems—
micro-misallocations that occur when
relative prices get misaligned as a result of
inflation
 Economics

professor shares blame
First order problem was integrity of the
financial system
Why is this a problem?

Standard model (representative agent models)
without institutions says this is no problem
 Misallocations
couldn’t have happened
 Were acting on best information available
 Simply a negative shock
 Some redistributions
 But redistributions don’t matter
 Economy simply goes on with new capital stock as if
nothing had happene
Redistributions and
institutions do matter

Loss in bank equity will not be readily replaced


Heavy dilution demanded
Consistent with theories of asymmetric information


With loss of bank capital, there will be reduced lending


Greenwald and Stiglitz, New Paradigm of Monetary Economics
What matters is not just interest rates but credit availability




Asquith and Mullins; Greenwald, Stiglitz, and Weiss “Informational Imperfections in the
Capital Markets and Macroeconomic Fluctuations,” American Economic Review, 74(2),
May 1984, pp. 194-199.
Credit availability affected also regulations (capital adequacy
requirements) and risk perceptions
As important as open market operations and interest rates
Spread between T-bill rate and lending rate an endogenous
variable
With reduced lending, reduced level of economic activity

Problems exacerbated by reduction in interbank
lending
 Tightening
credit constraints and leading to higher
lending interest rates
 Banks know that they don’t know own balance sheet
 And so can’t know balance sheet of others
 But there are still high levels of information
asymmetries
 Market breakdown


Stiglitz and Weiss, “Credit Rationing in Markets with Imperfect
Information,” American Economic Review, 71(3), June 1981,
pp. 393-410
Akerlof, Lemons
Credit interlinkages



As important as interlinkages emphasized in standard
general equilibrium model
Not fully mediated through price system
Bankruptcy in one firm can lead to bankruptcy in others
(bankruptcy cascades)




Collapse of economic system
Worry underlies bail-outs (1998 LTCM, 2008 Bear Stearns)
Agent based models more likely to bring insights
No hope from representative agent models

S. Battiston, D. Delli Gatti, B. Greenwald and J.E. Stiglitz ,“Credit
Chains and Bankruptcy Propagation in Production Networks,” Journal
of Economic Dynamics and Control, Volume 31, Issue 6, June 2007, pp.
2061-2084.

It will take time to restore bank capital, and
therefore for full restoration of economy


B. Greenwald and J. E. Stiglitz, “Financial Market Imperfections and
Business Cycles,” Quarterly Journal of Economics, 108(1), February 1993,
pp. 77-114.
Pace will be affected by magnitude of
fiscal stimulation
 Money
to those who are credit constrained
(unemployed)
 Would not work if Ricardian equivalence held
or if redistributions didn’t matter

Pace will also be affected by government
sponsored capital injections
 Hidden in bail-outs, huge wealth transfers
 Many banks focusing on selling “bad assets”
 By itself, doesn’t solve capitalization problem, only reduces
uncertainty
 They seem to be paying a high price
 American
bail-outs particularly non-transparent
 With credit and interest rate options embedded
 Access to Fed window by investment banks
 Discriminatory patterns?
Paulson plan badly flawed





Based on trickle down economics—throwing enough
money at Wall Street will trickle down to rest of economy
Like mass transfusion—while patient is dying from
internal bleeding
Does nothing to stop hemorrhaging
Buying hundreds of thousands of toxic mortgages and
derivatives based on them is complex—and because of
lemons problem taxpayer will overpay
If we don’t overpay, won’t repair hole in balance sheet
What should have been done

Equity injection
 Preferred shares with warrants
 Downside protection, upside potential

Helping people stay in their homes
 Already
3 million foreclosures, 2 million more
expected in next year
 Converting tax deduction to tax credit
 Bankruptcy reform—homeowners’ chapter 11
 Direct lending to homeowners at government’s lower
cost of capital and better enforcement mechanisms


Combined with conversion to recourse loans
And major haircut for banks—reducing loan amount to 90%
of house value
Stimulus

Even with program, economy is headed for recession




What is needed





Credit contraction
Worsening of balance sheets
Cutbacks in state and local spending
Expanded unemployment benefits
Aid to states and localities
More investment
Given high national debt, important to have large bang
for buck
America does not need to stimulate consumption


Problem has been too much consumption
Simply postpones day of reckoning
What was going on? Macro

At macro-level—insufficient aggregate demand
induced Fed to flood economy with liquidity and
have lax regulations to keep economy going
 Created
new bubble to replace dot.com bubble
 Lower interest rates major effect on mortgage equity
withdrawals, much of which was consumed

Decline in net worth, unlike case where investment is
stimulated

High level of demand for U.S. dollars to put in
reserves
 Massive reserve accumulation
 Partly in response to IMF/US treasury
response to
1997/1998 crisis
 But exporting T-bills rather than automobiles does not
create jobs

High oil prices
 Massive
redistribution to oil exporters
 If redistributions don’t matter, wouldn’t have any
consequences
 But redistributions do matter
 Part of global imbalances
 But real side of imbalances—inadequate global
aggregate demand


Myopic, short sighted response
Akin to how Latin America avoided negative impact of oil
price shock—borrowing for consumption




Paid a high price—lost decade
Housing bubble fueled consumption boom that offset
higher expenditures on oil, large trade deficit—for a while
Not sustainable
There were alternatives—none of this was inevitable

See J. E. Stiglitz and Linda Bilmes, The Three Trillion Dollar War,
2008
What was going on? Micro



Regulatory arbitrage—financial alchemy
converting F-rated toxic mortgages into financial
products that could be held by fiduciaries had a
private (but not necessarily social) pay-off
Accounting arbitrage—bonuses based on
reported profits, incentive to book profits (e.g.
from repackaging), leaving unsold (risky) pieces
“off balance sheet”
Distorted incentive systems
Hard to explain

How markets used models that were so bad
 Underestimated
systemic risk
 Underestimated obvious correlations
 Underestimated fat tail distributions
 Overestimated value of insurance (undercapitalized
insurance companies)
 Underestimated potential consequences of conflicts
of interest, moral hazard problems, perverse
incentives and scope for fraud


Appraisers owned by originating companies
Rating agencies paid by those producing products

Intellectual incoherence
 Argued
that they had created new products
that transformed financial markets

Justified high compensation
 Yet
based risk assessments on data from
before the creation of the new products
 Argued that financial markets were efficient
 Based pricing on spanning theorems
 Yet also argued that they were creating new
products that transformed financial markets
Hard to Explain
It was individually rational for those in
finance to take advantage of flawed
incentive structure—but not good for the
system
 Even if those originating mortgages had
flawed incentives, why didn’t investors
buying mortgages exercise better
oversight?
 Repeated failures

Hard to Explain
Markets still have not made available
mortgages that would have helped
individuals manage the risks which they
face
 There are alternatives that do a better job

 Danish
mortgages
 Variable rate, fixed payment, variable maturity
Regulatory Failure





Using wrong models
Focusing on wrong thing
Ideological—appointed partly because of
commitment to non-regulation
Political—when appointment was made,
implications for campaign contributions played
key role in appointment
Political (special interest) role in design of Basel
II regulations—not “just” technocratic
Beyond regulatory capture
Regulatory capture model provides too
simplistic model of what happened
 There was a party going on, and no one
wanted to be a party pooper

 But
Fed not only failed to dampen party but
also kept it going
 It had alternatives
Going forward

Actions by market participants generated externalities








Costs borne by taxpayers
Those who are losing their jobs
Social problems—millions of Americans losing homes
Whenever there is an externality, grounds for government
intervention
Those in the financial sector would like us just to build better
hospitals, but do nothing about prevention and contagion
Can we design interventions that encourage “good” innovation
(questionable value of much of recent financial innovation)?
Can we avoid “political economy” problems that have marked
past regulation?
Regulatory systems have to recognize asymmetries of
information and asymmetries of salaries
Regulation

Incentives
 Conflicts of interest
 Longer term
 Asymmetries give rise
 Stock options

to excessive risk taking
Behaviors
 Speed bumps
 Retaining some
responsibility for financial products
created

Accounting
 Reducing
scope for off balance sheet activity
Structures

Financial product safety commission
 With
representation of those who are likely to be hurt
by “unsafe” products
 Skills required to certify “safety” and “effectiveness”
different from those entailed in financial market
dealings

Financial market stability commission
 Need
separate market regulators because complexity
of each market requires specialized regulators
 But need oversight, to understand interactions among
pieces (systemic leveraging, regulatory arbitrage)
Financial market regulation is too
important to leave to those in the financial
sector alone
 Some aspects need to be approached on
a global level

 IMF
and Basel failed to provide adequate
regulatory framework
 Notion underlying Basel II that banks could be
relied upon to assess their own risk seems, at
this juncture, absurd
Rich research agenda ahead

Exploring financial interlinkages
 Bankruptcy cascades
 Optimal network design


(preventing contagion)
Designing financial instruments that better reflect
information imperfections and systematic
irrationalities
Designing appropriate mix of financial
institutions
 Taking into account local information
 Need for renegotiation
 Asymmetries of information created by
securitization
Rich research agenda ahead

Macro-economic models that take into
account complexity of financial system
 Including
financial linkages
 Recognizing role of banks
 And the consequences of redistributions
 Information imperfections, bubbles (rational
herding and irrational)
Research and Policy Agenda

Unfettered financial markets do not work
 But

Markets are not self-adjusting
 At

regulation and regulatory institutions failed
least in the relevant time frame
Darwinian natural selection may not work
 Like
Gresham’s law—bad money drives out good
 Reckless firms forced more conservative firms to
follow investment strategies
 More prudent firms might have done better in long
run—but couldn’t survive to take advantage of that
long run

Design of better regulations
 Not
only designed to discourage destructive
behaviors
 But to encourage financial system to fulfill its core
mission
 May require more extensive intervention in markets

Design of better regulatory institutions
 Based
on a theory of regulation that is better than
simplistic “capture” theory
 Which itself should be an important subject of study


Our financial system failed in its core missions—
allocating capital and managing risk
With disastrous economic and social
consequences
 Huge


disparity between potential and actual GDP
We must do better
And a successful research agenda will help us to
do that