Monetary EconomicsMonetary and Fiscal Policy Responses A simple model of banking crisis Monetary Economics Lecture 8 September 2010 2 / 43. ... constant dividend d E = Equity (book

Monetary EconomicsFinancial Crisis in Theory and Practice

Nicola Viegi

September 2010

Monetary Economics () Lecture 8 September 2010 1 / 43

Introductions

Models with Credit Market imperfection are only partially useful tounderstand the crisis

Origin of the Crisis are endogenous to the �nancial system

This lecture:

The Financial Origin of the CrisisThe Economics of the Crisis,Monetary and Fiscal Policy ResponsesA simple model of banking crisis


Some Literature

Markus Brunnermeier "Deciphering the 2007-08 Liquidity and CreditCrunch" A detailed analysis of the �nancial architecture thatgenerated the crisis and the ampli�cation mechanism - a must read(but do not expect to understand everything - nobody could)

Ross Levine "An Autopsy of the U.S. �nancial System" - Focus onCredit Rating agencies and the problem in �nancial regulatoryframework

Diamond and Rajan "The Credit Crisis: Conjectures about Causesand Remedies" Origin in the housing market.

John Taylor "The Financial Crisis and the Policy Responses: AnEmpirical Analysis of What Went Wrong" - Responsability is in toolax monetary policy and wrong policies

Paul De Grauwe "Keynes�Savings Paradox, Fisher�s Debt De�ationand the Banking Crisis" - A macroeconomic analysis that highlightthe need of public intervention to eliminate coordination failure


The �nancial origin of the crisis

Leverage + Financial innovation + Easy Money = Financial Crisis

1 Originate and distribute banking model2 Increased leverage/maturity mismatch (on/o¤ balance sheet)3 Lax lending standards4 "Greenspan put"5 "Savings Glut" in the rest of the world


Global Imbalances


Flow-Stock de�ation spirals

Keynesian Saving Paradox

Fisher�s Debt De�ation

Cost Cutting De�ation

Bank Credit De�ation

Coordination Failures


Keynesian Saving Paradox

When one individual desires to save more, and he is alone todo so, his decision to save more (consume less) will not a¤ectaggregate output. He will succeed to save more, and once he hasachieved his desired level of savings he stops trying to save more.When the desire to save more is the result of a collective lack

of con�dence (animal spirits) the individual tries to build upsavings when all the others do the same. As a result, output andincome decline and the individual fails in his attempt to increasesavings. He will try again, thereby intensifying the decline inoutput, and failing again to build-up savings


Fisher�s Debt De�ation

When one individual tries to reduce his debt, and he is aloneto do so, this attempt will generally succeed. The reason is thathis sales of assets to reduce his debt will not be felt by theothers, and therefore will not a¤ect the solvency of others. Theindividual will succeed in reducing his debt.When the desire to reduce debt is driven by a collective

movement of distrust, the simultaneous action of individuals toreduce their debt is self-defeating. They all sell assets at thesame time, thereby reducing the value of these assets. This leadsto a deterioration of the solvency of everybody else, therebyforcing everybody to increase their attempts at reducing theirdebt by selling assets.


Cost Cutting De�ation

When one individual �rm reduces its costs by reducing wagesand �ring workers in order to improve its pro�ts, and this �rm isalone to do so, it will generally succeed in improving its pro�ts.The reason is that the cost cutting by an individual �rm does nota¤ect the other �rms. The latter will not react by reducing theirwages and �ring their workers.When cost cutting is inspired by a collective movement of

fear about future pro�tability the simultaneous cost cutting willnot restore pro�tability. The reason is that the workers who earnlower wages and the unemployed workers who have less (or no)disposable income will reduce their consumption and thus theoutput of all �rms. This reduces pro�ts of all �rms. They willthen continue to cut costs leading to further reductions of outputand pro�ts.


Bank Credit De�ation

When one individual bank wants to reduce the riskiness of itsloan portfolio it will cut back on loans and accumulate liquidassets. When the bank is alone to do so (and provided it is nottoo big), it will succeed because the strategy of the bank will notbe felt by the other banks, which will not react.When banks are gripped by pessimism and extreme risk

aversion the simultaneous reduction of bank loans by all bankswill not reduce the risk of the banks�loan portfolio for tworeasons. First, banks lend to each other. As a result when banksreduce their lending they reduce the funding of other banks. Thelatter will be induced to reduce their lending, and thus thefunding of other banks. Second, when one bank cuts back itsloans, �rms get into trouble. If �rms get into trouble banks willsee that their loan portfolio has become riskier. They will in turnreduce credit thereby increasing the riskiness of the loan portfolioof other banks.


Macroeconomic Consequences

Credit Crunch

Collapse in Demand

Collapse in production and employment

Collapse of international trade


GDP Growth

Global GDP Growth

-10

-5

0

5

10

15

2006

Q1

2006

Q2

2006

Q3

2006

Q4

2007

Q1

2007

Q2

2007

Q3

2007

Q4

2008

Q1

2008

Q2

2008

Q3

2008

Q4

2009

Q1

2009

Q2

2009

Q3

2009

Q4

2010

Q1

2010

Q2

2010

Q3

2010

Q4

2011

Q1

2011

Q2

2011

Q3

2011

Q4

World Advanced economies Emerging and developing economies


Industrial Production

Industrial Production

80

85

90

95

100

105

110

115

120

125

130

2007

m1

2007

m2

2007

m3

2007

m4

2007

m5

2007

m6

2007

m7

2007

m8

2007

m9

2007

m10

2007

m11

2007

m12

2008

m1

2008

m2

2008

m3

2008

m4

2008

m5

2008

m6

2008

m7

2008

m8

2008

m9

2008

m10

2008

m11

2008

m12

2009

m1

2009

m2

2009

m3

2009

m4

2009

m5

2009

m6

2009

m7

2009

m8

2009

m9

2009

m10

2009

m11

2009

m12

2010

m1

2010

m2

2010

m3

2010

m4

World Advanced economies Emerging economies


Merchandise Exports

80

90

100

110

120

130

140

15020

07m

1

2007

m2

2007

m3

2007

m4

2007

m5

2007

m6

2007

m7

2007

m8

2007

m9

2007

m10

2007

m11

2007

m12

2008

m1

2008

m2

2008

m3

2008

m4

2008

m5

2008

m6

2008

m7

2008

m8

2008

m9

2008

m10

2008

m11

2008

m12

2009

m1

2009

m2

2009

m3

2009

m4

2009

m5

2009

m6

2009

m7

2009

m8

2009

m9

2009

m10

2009

m11

2009

m12

2010

m1

2010

m2

2010

m3

2010

m4



Employment

Advanced Economies: Emplyment Growth

-5

-4

-3

-2

-1

0

1

2

3

2007

m1

2007

m2

2007

m3

2007

m4

2007

m5

2007

m6

2007

m7

2007

m8

2007

m9

2007

m10

2007

m11

2007

m12

2008

m1

2008

m2

2008

m3

2008

m4

2008

m5

2008

m6

2008

m7

2008

m8

2008

m9

2008

m10

2008

m11

2008

m12

2009

m1

2009

m2

2009

m3

2009

m4

2009

m5

2009

m6

2009

m7

2009

m8

2009

m9

2009

m10

2009

m11

2009

m12

2010

m1

2010

m2

2010

m3

2010

m4


Financial Markets

90

92

94

96

98

100

102

104

106

108

110

01/01

/2010

01/07

/2010

1/13/2

010

1/19/2

010

1/25/2

010

1/29/2

010

02/04

/2010

02/10

/2010

2/16/2

010

2/22/2

010

2/26/2

010

03/04

/2010

03/10

/2010

3/16/2

010

3/22/2

010

3/26/2

010

04/01

/2010

04/07

/2010

4/13/2

010

4/19/2

010

4/23/2

010

4/29/2

010

05/05

/2010

05/11

/2010

5/17/2

010

5/21/2

010

5/27/2

010

06/02

/2010

06/08

/2010

6/14/2

010

6/18/2

010

6/24/2

010

6/30/2

010

S&P 500 MSCI EM


In�ation

Inflation

-2

0

2

4

6

8

10

2002

m1

2002

m4

2002

m7

2002

m10

2003

m1

2003

m4

2003

m7

2003

m10

2004

m1

2004

m4

2004

m7

2004

m10

2005

m1

2005

m4

2005

m7

2005

m10

2006

m1

2006

m4

2006

m7

2006

m10

2007

m1

2007

m4

2007

m7

2007

m10

2008

m1

2008

m4

2008

m7

2008

m10

2009

m1

2009

m4

2009

m7

2009

m10

2010

m1

2010

m4



How Does it compared to 1929?

World Industrial Production now vs. then



World Trade now vs. then



Equity Markets now vs. then


Why the Di¤erence?

Monetary Policy


Why the Di¤erence?

Aggressive Fiscal Policy


Modelling Financial Crisis

Blanchard (2009) - triggers of the �nancial crisis:

intransparency of assets and balance sheets

reliance on short-term funding

forced deleveraging

coordination problems

Plana simple model of banking crisisStarting point: the balance sheet of an investment bank


Balance Sheet of an Investment Bank

Assets Liabilities and EquityA E- L

A = Assets (book value) - claim on investment project - generate aconstant dividend dE = Equity (book value) -L = Liabilities (book value) - short term debt which has to re�nancedevery period at an interest rate equal to rt

Ri�nancing always possible if

(A� E ) rt � d


Introducing Uncertainty (1)

Investment bank borrows from intermediaries on the money market -market for short-term debt instrumentsThe intermediaries have access to money from outsiede investors whodemand interest rate RWith Probability P they will be repaid, with Probability (1� P) theywil loose both the principal and the interest

What Interest will the intermediary charge?

The intermediary (for each dollar borrowed and lended to the investmentbank) will make a pro�t if:

P (1+ rt ) + (1� P) (0) � 1+ Rsolving for rt gives

rt �1+ R � P

PMonetary Economics () Lecture 8 September 2010 25 / 43

Introducing Uncertainty (2)

Competitive Market

rt =1+ R � P

P

P = 1, ....rt = R

P ! 0, ...rt �! ∞

Higher the risk of the investment, higher the cost


Introducing Intransparency (1)

Assume that there is two types of assets: good and bad (or "toxic")assets

the good asset pays d (dividends) in perpetuity (forever)the bad assets disappear in each period with probability q (subprimemortages, for example)

Pricing of the loan with perfect knowledge

For banks with good assets

rgt = R

For Banks with bad assets

rbt =R + q1� q


Introducing Intransparency (2)

What if the intermediaries in the money market cannot disinguish betweengood assets and bas assets? What if they know only the proportion (p) ofbad assets in the market but not their location? What is the expectedpro�t for lending to investment banks?

Probability of incuring a loss : p � q (the probability of lending to abank with bad assets times the probability that those assets disappear)

Expected Pro�ts on each dollar lended if

(1� p � q) (1+ rt ) + (p � q) (0) � 1+ ROptimal Price of the Loan

rt =R + p � q1� p � q

ImplicationsDue to intransparency, �nancing of bad assets becomes more attractive,

�nancing of good assets becomes less attractive.


Bad Assets Dominance

Notice that:

rt =R + p � q1� p � q < r

bt =

R + q1� q

and

rt =R + p � q1� p � q < r

gt = R

Good assets pay part of the cost of having bad assets around - the price ofa loan increase for everybody, independently of the quality of the assets(unknown).


Intransparency and Interest Rate (1)

rt =R + p � q1� p � q

Increased uncertainty about the proportion of good and bad assetsdrives up the interest rate.

δrtδp=

q1� p � q +

(R + p � q) q(1� p � q)2

=1+ R

(1� p � q)2> 0

Re�nancing loans becomes more di¢ cult, also for banks with onlygood assets.



Example - the collapse of Lehman Brothers



If (A� E ) rt > d both, the investment banks with good assets andthe investment banks with bad assets will run intro trouble.

They may have to sell their assets if they cannot re�nance the loan.

But, a bank with good assets would have been able to re�nance if themarket was transparent.

Note that the high interest rate and the re�nancing problem do notarise because of a lack of liquidity.

Problems created by intransparency and re�nancing will not be solvedby the central bank injecting liquidity (at least as long as theinvestment banks do not have the required collateral and restore their�nancing - reducing the perceived risk for the lender).


Intransparency and market price of assets (1)

AssumeG = Price of good assets (A normal mortage)B = Price of bad assets (Subprime mortage)B < G

What would be the price that the bank can extract from the market if it isnot possible to distinguish good from bad?

(1� p)G + pBIncrease uncertainty about the proportion of good and bad assets drivesdown the price. A change in evaluation of the amount of "toxic" assetspresent in the market will chnge the prices of all assets in proportion to theincreas in risk


Intransparency and market price of assets (2)

Implications:

If markets are illiquid (i.e. when a too large supply meets a too smalldemand), prices may go down even further because of the bargainingpower of buyers, i.e. both B and G decline.

Adverse Selection (Akerlof 1970 "lemon problem"): If the price(1� p)G + pB is lower than what banks with good assets wouldvoluntarily want to sell them for, we can expect that on the market anover-proportion of bad assets will be sold (say ep > p), and thus themarket price for the asset drops even furtherto(1� ep)G + epB < (1� p)G + pB


Ampli�cation (1) : Mark to Market Pricing

Consider two banks who have the same amount of assets at market valueA but di¤erent structures of liabilities:

Bank 1 - Low leverage Bank 2 - High LeverageAssets Liabilities and EquityA E1- L1

Assets Liabilities and EquityA E2- L2

with E1 > E2, and L1 < L2Regulators requires banks to maintain a minimal capital ratio

EA� 4%

evaluated at market prices of assetsUnder these conditions a forced sale of assets by the more leveraged bankmay trigger a forced sale of assets by the less leveraged bank



Assume an exogenous increase in p (Lehman Brother Fails, for examlple) -rt will go up to r ct . How does the increse of rt a¤ect the solvency of thetwo banks?Before the increse in p both banks are solvent

L1rt < L2rt � dafter only bank one is

L1r ct � d < L2r ctBank 2 needs to sell a part of its assets to be able to meet its short termliabilities.As a consequence of this bank de-leveraging, the market price of the assetsdrops to A� ∆.Bank 1 would still be able re�nance L1 = (A� E1) because L1r ct � d .How does regulation a¤ects this solution?



However, the regulator requires a minimum capital ratio and since it usesmark-to-market pricing, the drop in value from A to A� ∆ can forceeven bank 1 to sell parts of its assets.The drop in value makes the minimal capital ratios binding for Bank 1 iffor instance

E1A� 4% > E1

A� ∆In this case, bank 1 has to sell assets only due to regulatory reasons.Implications:

If, due to �re sale prices, the bank is forced to sell assets below whatthey are worth, the bank is in a worse condition after the sale thanbefore.

If the sale of assets by bank 1 leads to further price drops, otherbanks with even lowerleverage ratios might also get into trouble. Adownward spiral occurs.


Ampli�cation (2) : Coordination Problem

Liquidity might disappear from the market even for the safe bank becauseall investor have incentive to follow the same strategy and �ee to safety.(buy Treasury bonds, i.e. lend to the state instead than individualinvestment banks)ScenarioEach lender only lends a fraction L/k of L.So, each investment bank has to �nd k lenders in each period.If the bank does not �nd k lenders, it fails and nobody is paid.What is the interest rate that each lender should ask? It depends on theprobability that the bank will �nd k � 1 other lenders (call it z)Expected Pro�ts on each dollar lended if

z [(1� pq) (1+ rt ) + (pq) (0)] + (1� z) 0 � 1+ R

z (1� pq) (1+ rt ) � 1+ R

Optimal Price of the Loan

rmt =1+ R � (z � zpq)

(z � zpq)Decrease in z increase the interest rate charged by the lender



If rmt increases su¢ ciently, then the inequality (A� E ) rmt > dimplies that the bank fails.In this case, no rational lender would lendto the bank, thus the probability z must be low.

On the other hand if rmt is low, then the inequality (A� E ) rmt > dimplies that the bank does not fail. In this case, a rational lenderwould lend to the bank, thus the probability z must be high.

Circularity between rmt and z . If few people lend interest rate goes upand even less people will lend. Two possible solutions - everybodylends or nobody does! (very simpli�ed)



De�ne Strategies and Pay-o¤s

Many Lend Few LendsLend 0 with zero pro�t condition �1� R (loose principal)No lend Fed rate (�1) Fed Rate (�1)

Two equlibriums - Lend if Many lend and No lend if few lend


Evidence

Uncertainty on the risk evaluation of assets


Evidence

Money Market Dries up


Conclusion

The simple model gives us a framework to interpret the �nancial crisis

intransparency of assets and balance sheets: the complexity of�nancial innovation and the inadequacy of regulators and credit ratingagency made the system vulnarable to spreeding of panic

reliance on short-term funding: the credit boom was relying on shortterm �nancing, making the system very sensitive to external shocks

forced deleveraging: when the crisis arrive, mark-to market rulesforced every �nancial institution to try reducing their debt position

coordination problems: liquidity in the market dried up becauseeverybody was uncertain about what every other market partecipantposition was going to be, preferring to park liquidity in governmenttreasury bonds.

What is the role of monetary and �scal policy in this context? Howdo policies work at the zero bound of Interest rate?


Documents

Monetary EconomicsMonetary and Fiscal Policy Responses A simple model of banking crisis Monetary Economics Lecture 8 September 2010 2 / 43. ... constant dividend d E = Equity (book