COSTS OF FINANCIAL INSTABILITY - GDP, INVESTMENT AND CONSUMPTION E Philip Davis NIESR and Brunel University West London [email protected] .

COSTS OF FINANCIAL INSTABILITY - GDP, INVESTMENT AND

CONSUMPTION

E Philip Davis

NIESR and Brunel University

West London

[email protected]

www.ephilipdavis.com

groups.yahoo.com/group/financial_stability

Course on Financial Instability at the Estonian Central Bank,9-11 December 2009 – Lecture 10

PAPER 1:CORPORATE FINANCIAL

STRUCTURE AND FINANCIAL STABILITYby E Philip Davis and Mark

Stone

Published in Journal of Financial Stability

Structure of paper

• Introduction

• Literature review

• The data and corporate financial structure

• Corporate financial structure and financial stability – descriptive analysis

• Econometric analysis

• Conclusions

Introduction

• Dimensions of corporate financing– Depth – quantity of financing available– Breadth – variety of financing options (loans, bonds,

shares, trade credit – and liquidity)

• Measure implications of structure for fragility via:– Measuring real expenditure responses to banking and

currency crises and variation therein linked to balance sheet indicators

– Examining shifts in size and composition of corporate financing during a crisis, controlling for normal cyclical changes (shifts may link to supply-side, i.e. rationing of finance or demand-side, i.e. corporate balance sheet adjustment)

Literature review

• General determinants of corporate financial structure– Exceptions to Modigliani-Miller – costs of bankruptcy

versus tax benefits to debt– Asymmetric information and higher cost of external

financing (issue how well financial system deals with agency costs)

• Economic and financial development– Overall development of financial services important to

growth - and not bias to bank or market financing?– Stages of development in financial systems – stock and

bond markets come later than banks. Legal aspects important

• Financial effects on the business cycle– Financial accelerator and effect of net worth

(debt/equity ratio) on availability of external finance– Credit channel and specialness of bank finance (bank

debt/total debt ratio)

• Recent theories of financial crises– Introduction of banks and liquidity effects into most

recent models– Collateral effects and micro and macroeconomic

rigidities– Financial breadth (“multiple avenues”) and the impact

of crises on expenditures

• Empirical work on post crisis output contractions• Macroprudential indicators of corporate sector

fragility

Descriptive analysis of crises, growth and investment

• Standard set of banking and currency crises (Eichengreen and Bordo 2002) – 59 events for 29 countries, 17 in EMEs, 18 banking crises and four twin crises

• Methodology is to measure effects as deviation of contribution to GDP from trend effect over 11 years; 5 before crisis year and 5 after

• Larger impact of crises on EMEs GDP – mainly accounted for by domestic demand and corporate investment expenditures and inventory decumulation therein. Consumption relatively robust and external demand expands

• Banking crises greater effect than currency crises

Impact of crises on GDP

Percent GDP

Total domestic demand

Total public

domestic demand

Total private

domestic demand

Private consumpt

ion

Private investme

nt

Change in inventorie

s Foreign balance

Total average -1.5 -2.6 0.1 -2.9 -0.7 -1.7 -0.4 1.1

median -1.0 -1.1 -0.1 -1.8 -0.7 -1.1 0.0 0.5

EME average -3.2 -6.4 -0.4 -5.6* -1.3 -3.2* -1.1 2.7

median -3.3 -4.3 -0.3 -4.1* -2.0 -1.9* -0.1 1.6

OECD average -0.9 -1.5 0.2 -2.0 -0.5 -1.3 -0.1 0.6

median -0.2 -0.7 -0.1 -0.9 -0.5 -0.4 0.1 0.4

Currency average -1.1 -1.9 -0.1 -2.0 -0.5 -1.2 -0.3 1.0

median -0.2 -0.7 -0.1 -0.9 -0.5 -0.4 0.0 0.5

Banking average -3.1 -5.1 0.2 -5.0 -1.4 -3.1 -0.4 1.9

median -2.8 -4.1 0.3 -4.0 -0.8 -3.1 -0.1 1.6

Indicators of financial structure and crises

• Size of balance sheets per se is unrelated to expenditure response…

• …but marked impact of corporate leverage on investment and inventory contraction

• Falls in flows of external finance mainly in bank loans and liquidity falls while bond issues rise. Falls in external finance and bank lending larger for EMEs and in banking crises

Private investment deviation from trend and leverage

Chart 1, Private Fixed Investment Deviation from Trend Growth and Debt-Equity Ratio, Crisis Years

Australia

Australia

AustraliaCanada

Canada

Finland

Finland

FranceFrance India

India

Israel

Italy

#REF!Italy

-11

-9

-7

-5

-3

-1

1

3

0.0

50.0

100.

0

150.

0

200.

0

250.

0

300.

0

350.

0

400.

0

450.

0

Debt-Equity Ratio

Inve

stm

ent C

umul

ativ

e G

row

th

Inventory accumulation and leverage

Chart 2, Private Fixed Inventory Deviation from Trend Growth and Debt-Equity Ratio, Crisis Years

Australia

Australia

Australia Canada

Canada

Finland

Finland France

France India India

Israel Italy #REF!

Italy Italy

-8

-6

-4

-2

0

2

0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0 450.0

Debt-Equity Ratio

Investment Cumulative Growth

Impact of crises on flow of funds/GDP

All crises (27)

External finance

Bank loans

Bond issues

Equity issues

Trade credit Liquidity

Average -0.6 -0.5 0.1 -0.1 -0.1 -0.7

OECD -0.1 0.0 0.1 -0.2 0.5 -0.1

EME -1.4 -1.4 0.0 -0.1 -1.0 -1.6

Banking crises (9)

External finance

Bank loans

Bond issues

Equity issues

Trade credit Liquidity

Average -2.0 -2.2 0.3 0.2 0.1 -1.7

OECD -0.5 -0.6 0.3 -0.3 1.7 -0.1

EME -3.4 -4.3 0.3 0.6 -1.5 -3.6

Currency crises (19)

External finance

Bank loans

Bond issues

Equity issues

Trade credit Liquidity

average -0.7 -0.4 -0.1 -0.2 -0.4 -0.8

OECD 0.1 0.3 0.0 -0.1 0.1 0.0

EME -1.8 -1.4 -0.1 -0.3 -1.2 -1.9

Econometric analysis

• Estimate specifications capturing normal cyclical developments before testing for additional effects of crises

• First estimated corporate expenditure functions with crisis and balance sheet variables

• Then estimated flow of funds/GDP determinants• Cross section weighted GLS panel with fixed

effects and cross section weights (common factors world growth, world trade, share prices, bond yields)

• Results suggest strong crisis effects on both corporate expenditure and corporate financing

• Crisis effects smaller for currency crises than banking crises and for OECD than EME

• Effects on investment aggravated by high debt/equity (financial accelerator) and low bank/debt (credit channel)

• Flow of funds equations suggest fall in bank lending, equity issue (information spillovers from bank lending) but rise in bond issue (multiple avenues of intermediation)

Jorgensen investment functions…

Dependent variable – log difference of private investment All EME OECD DLY 2.4 (0.112)** 3.2 (0.25)** 2.11 (0.13)** DLY(-1) -0.23 (0.14)* 0.26 (0.44) -0.27 (0.16)* DLIP(-1) 0.26 (0.04)** 0.16 (0.095)* 0.23 (0.05)** LIY(-1) -0.144 (0.014)** -0.203 (0.03)** -0.127 (0.015)** DIRD -0.0004 (8E-5)** -0.00027 (7.6E-

5)** 0.0006 (0.0008)

IRD(-1) -0.0006 (0.0002)** -0.0005 (8.3E-5)** -0.0028 (0.0006)** BDUM(-1) -0.02 (0.009)** -0.073 (0.035)** -0.016 (0.0085)* CDUM(-1) -0.025 (0.006)** -0.0074 (0.018) -0.027 (0.0068)** Adjusted R2 0.69 0.77 0.69 SE 0.06 0.08 0.05 Observations 517 105 412 Countries 23 6 17

…with balance sheet variables

All All All BDUM(-1) -0.026 (0.013)** -0.143 (0.102) -0.178 (0.082)** CDUM(-1) -0.036 (0.01)** -0.003 (0.02) -0.026 (0.02) LDER(-1) -0.0092 (0.0032)** -0.015 (0.004)** BDUM*LDER(-1) -0.02 (0.01)** -0.0176 (0.01)* CDUM*LDER(-1) -0.034 (0.013)** -0.038 (0.011)** LBDEBT(-1) 0.115 (0.029)** 0.115 (0.03)** BDUM*LBDEBT (-1)

-0.17 (0.04) -0.214 (0.118)**

CDUM*LBDEBT (-1)

0.05 (0.04) 0.028 (0.031)

Tobin investment function

All DLY 2.3 (0.11)** 2.3 (0.11)** 2.3 (0.11)** DLY(-1) -0.38 (0.16)** -0.41 (0.15)** -0.42 (0.15)** DLIP (-1) 0.26 (0.049)** 0.29 (0.05)** 0.28 (0.05)** LIY(-1) -0.14 (0.02)** -0.16 (0.023)** -0.16 (0.023)** LTOBIN(-1) 0.016 (0.0024)** 0.017 (0.003)** 0.019 (0.003)** BDUM(-1) -0.027 (0.014)* -0.033 (0.015)** -0.18 (0.09)** CDUM(-1) -0.028 (0.009)** -0.033 (0.008)** -0.022 (0.013)* LBDEBT(-1) 0.056 (0.024)** 0.051 (0.024)** BDUM*LBDEBT(-1)

-0.207 (0.12)*

CDUM*LBDEBT(-1)

0.015 (0.03)

Adjusted R2 0.77 0.74 0.74 SE 0.041 0.039 0.039 Observations 258 227 227 Countries 19 18 18

Inventory adjustment function

Dependent variable: change in inventories/GDP All EME OECD DLY 0.06 (0.008)** 0.17 (0.04)** 0.049 (0.007)** IIY(-1) 0.46 (0.13)** 0.38 (0.28) 0.5 (0.108)** DIRD 2.1E-5 (1.2E-5)** 1.9E-5 (1.1E-5)* 0.00011 (3.4E-5)** IRD(-1) 2.4E-6 (1.3E-5)** 8.6E-6 (1.4E-5) -1.2E-7 (2E-5) BDUM 0.0002 (0.0007) -0.02 (0.008)** 0.001 (0.0006)* CDUM -9.7E-6 (0.0003) 0.0006 (0.003) -4.2E-5 (0.0003) BDUM(-1) 0.0008 (0.0004)** 0.004 (0.009) 0.00036 (0.00036) CDUM(-1) -5.9E-6 (0.0003) -0.0027 (0.003) -5E-5 (0.0003) Adjusted R2 0.49 0.49 0.53 SE 0.009 0.018 0.006 Observations 569 108 461 Countries 23 6 17

Bank lending function

Dependent variable: difference of bank lending/GDP All EME OECD DLY 0.21 (0.028)** 0.13 (0.054)** 0.203 (0.032)** DLY(-1) 0.078 (0.024)** 0.045 (0.032) 0.077 (0.04)* BLY(-1) -0.47 (0.046)** -0.71 (0.117)** -0.38 (0.052)** DIRD 0.0014 (0.00016)** 0.0015 (9.8E-5)** 0.0017 (0.0004)** IRD(-1) 0.0008 (0.0003)** 0.0013 (0.00027)** 0.00022 (0.00027) BDUM -0.0084 (0.0036)** -0.019 (0.008)** -0.0016 (0.0032) CDUM -0.0015 (0.0025) -0.011 (0.0049)** 0.002 (0.0032) BDUM(-1) -0.0093 (0.0015)** -0.013 (0.0028)** -0.01 (0.0025)** CDUM(-1) -0.0039 (0.0011)** -0.0046 (0.0028)* -0.0057 (0.002)** Adjusted R2 0.47 0.63 0.44 SE 0.025 0.032 0.019 Observations 362 120 242 Countries 23 6 17

Bond issuance function

All EME OECD DLY 0.007 (0.006) -0.0004 (0.017) 0.02 (0.01)** DLY(-1) 0.032 (0.011)** 0.03 (0.035) 0.044 (0.008)** BOY(-1) -0.48 (0.08)** -0.34 (0.14)** -0.55 (0.096)** DIRD -0.00014 (6.3E-

5)** 0.00022 (0.00029) -0.00029 (6.6E-

5)** IRD(-1) 8.4E-5 (3.8E-5)** 0.00021 (0.00016) 0.00012 (5.1E-5) BDUM 0.0014 (0.0007)** 0.0009 (0.0015) 0.0023 (0.0009)** CDUM -0.0009 (0.0005)* -0.0037 (0.002)* -0.00023 (0.0006) BDUM(-1) 0.001 (0.0015) -0.0032 (0.0057) 0.0021 (0.001)** CDUM(-1) -0.0015 (0.0005)** -0.0042 (0.004) -0.0008 (0.0002)** Adjusted R2 0.22 0.14 0.27 SE 0.009 0.011 0.008 Observations 346 104 242 Countries 22 5 17

Dependent variable: difference of bond issuance/GDP

Summary table of crisis dummiesEquation All EME OECD Tobin -0.033 BDUM(-1) n.a. -0.036 BDUM(-1) -0.033 CDUM(-1) n.a. -0.036 CDUM (-1) Jorgensen -0.02 BDUM(-1) -0.073 BDUM(-1) -0.016 BDUM (-1) -0.025 CDUM(-1) -0.027 CDUM (-1) Inventories 0.001 BDUM -0.02 CDUM 0.0008 BDUM(-1) Bank lending -0.0084 BDUM -0.019 BDUM -0.011 CDUM -0.0093 BDUM(-1) -0.013 BDUM(-1) -0.01 BDUM(-1) -0.0039 CDUM(-1) -0.0046 CDUM(-1) -0.0057 CDUM(-1) Bond issuance 0.0014 BDUM 0.0023 BDUM -0.0009 CDUM -0.0037 CDUM 0.0021 BDUM(-1) -0.0015 CDUM(-1) -0.0008 CDUM(-1) Equity issuance -0.0042 BDUM(-1) -0.001 CDUM(-1) Trade credit -0.00075 BDUM(-

1)

-0.00048 CDUM(-1)

External financing -0.019 BDUM -0.019 CDUM -0.012 BDUM(-1) -0.013 BDUM(-1) -0.018 BDUM(-1) -0.004 CDUM (-1) -0.008 CDUM(-1) Liquidity -0.0017 CDUM(-1)

Conclusion

• Probed impact of crises on corporate financing and expenditures in OECD and EME countries

• Marked differences in balance sheet size and composition as well as flows between OECD and EME countries

• Investment and inventory contractions key components of GDP decline in wake of crises, notably in EMEs, underlining importance of better understanding of dynamics of corporate behaviour during crises

• Role of debt equity ratio in determining impact as well as bank lending/debt

• Role of bond market as shock absorber in OECD

• Evidence strengthens the case for financial sector

reforms, notably in EMEs, given role of external finance.

• Bank lending and trade credit warrant particular focus in EMEs

• Development of bond markets also warranted• IMF surveillance needs to focus on corporate sector

performance in macroprudential surveillance and not merely banking sector – and also in Article IV in assessing monetary policy transmission during a crisis

• Encourage countries to produce flows of funds including back data

• Further research in this overall area desirable

PAPER 2: COSTS OF FINANCIAL

INSTABILITY, HOUSEHOLD-SECTOR

BALANCE SHEETS AND CONSUMPTION

Ray Barrell, E Philip Davis and Olga Pomerantz

 Published in Journal of Financial Stability

Introduction

• Literature on costs of financial instability tends to focus on fiscal costs and the impact on GDP of banking crises.

• We analyse the effect of a banking or currency crisis on consumption.

• We show consumption plays an important role in the macroeconomic adjustment following a financial crisis

• Effect of a crisis is aggravated by high leverage, also greater in a small open economy

• Falling house prices shown to be part of the transmission process of financial instability

• High nominal interest rates are an indicator of sharp declines in consumption

• Results imply that a banking crisis taking place now could have a greater incidence than in the past, especially if macroeconomic policy is unable to respond, as for a small country in EMU.

• Also highly relevant as Basel II encourages consumer lending via lower risk weights

1 Measuring costs of banking crises

• More studies on prediction than costs• Components of costs

– Losses by stakeholders in failing banks– Losses by borrowers– General recession hitting consumption and

investment (but crises also occur in recessions)– Fiscal resolution costs, which also affect others

• Issue of timing of crisis, linked to definition

• Fiscal costs ca. 12% of GDP (larger in EMEs), depending on role of banks, scope of liquidity support

• Output losses usually growth or (Hoggarth-Sapporta) output level difference from trend or recession in comparable countries

• Stylised facts crises last longer in OECD countries and output costs can be greater

• Also twin crises costlier than banking crises• Davis-Stone components of GDP analysis –

investment bears the brunt (-3.1% of GDP) compared to consumption (-1.4% of GDP)

• Econometric approach – dummies in investment function – that we follow here

2 The impact of banking and currency crises on consumption

• Probe the nature and determinants of reaction of consumption to financial instability over and above impact of crisis on real personal disposable income and net financial wealth

• Test for dummies in standard consumption function, also with focus on financial liberalisation and resulting rise in leverage

• Standard Caprio-Klingebiel definitions of crises:– Currency crises entail forced change in parity, abandonment of a

pegged exchange rate, or an international rescue. – Banking crisis involve bank runs, widespread bank failures and the

suspension of convertibility of deposits into currency, or significant banking sector problems that result in the erosion of most or all of banking system capital.

List of crises Banking crises Currency crises US 1984 1971, 1985 UK 1974 1964-7, 1976, 1982, 1992 Germany 1977 Japan 1992 1979 France 1994 1992 Sweden 1991 1971, 1992 Italy 1990 1976, 1992, 1995 Netherlands 1971 Canada 1983 1962, 1981, 1986 Belgium 1971, 1982 Spain 1977 1971, 1976, 1982, 1992,

1995 Portugal 1976, 1978, 1983 Finland 1991 1986, 1991, 1993 Denmark 1987 1971, 1976, 1992, 1993 Ireland 1976, 1986, 1992 Austria Australia 1989 1976, 1983, 1985 S Korea 1998 1980, 1998 Norway 1987 1986 Total 14 42 1960s 0 2 1970s 3 15 1980s 5 13 1990s 6 13 G-7 7 14 SOEs 7 28

Consumption specification• Function of human and non human wealth,

where human proxied by its return (RPDI)

• Log transform to allow for growth

• Comparable work Ludvigsen-Steindel, Davis-Palumbo, Barrell- Davis

• Panel GLS estimation, error correction

tttt bWaYC 1

tttt WYcC 10 lnlnln

ktijti

tttt

YW

WYCC

ln*ln*

ln*ln*ln*ln 1211110

Regression data and ADFs Data

sample LC LRPDI LRNW DLC DLRDPI DLRNW

US 61q2-03q2 -3.14 -3.55 -1.87 -6.47 -8.68 -8.4 UK 64q2-02q4 -1.47 -2.3 -2.1 -7.8 -9.0 -8.5 Germany 71q1-02q4 -1.4 -1.5 -3.9 -9.0 -8.5 -10.1 Japan 71q1-02q4 0.5 0.7 -1.2 -11.0 -13.9 -8.4 France 70q1-01q4 -2.6 -1.8 -2.9 -5.9 -11.2 -6.5 Sweden 61q2-01q4 -2.6 -2.0 -1.75 -5.4 -6.6 -7.1 Italy 72q1-01q4 -2.2 -1.4 -0.4 -4.9 -9.8 -7.8 Netherlands 70q2-01q4 -1.0 -3.13 -3.8 -6.6 -9.5 -8.5 Canada 61q2-02q4 -2.1 -1.6 -2.3 -7.6 -8.8 -9.2 Belgium 70q2-01q4 -3.4 -2.2 -3.6 -5.7 -8.5 -7.1 Spain 70q2-01q4 -1.5 -1.5 -2.4 -4.0 -10.0 -5.2 Portugal 80q3-01q4 -2.0 -3.0 -2.1 -3.9 -4.6 -5.6 Finland 79q2-02q4 -1.6 -2.8 -1.3 -4.3 -5.5 -4.6 Denmark 71q2-01q4 -2.1 -4.3 -2.0 -7.0 -7.1 -6.8 Ireland 77q2-97q2 -1.0 -1.1 -3.0 -2.5 -5.9 -5.5 Austria 70q2-01q4 -4.4 -2.8 -2.5 -11.2 -8.0 -5.8 Australia 88q3-03q3 -2.3 -2.6 -1.2 -4.6 -6.8 -4.9 S Korea 76q1-03q3 -1.2 -1.4 -0.9 -5.9 -3.0 -6.8 Norway 81q1-03q3 -1.1 -1.7 -3.8 -5.3 -3.8 -9.3 Panel Unit Root

-1.9 -2.1 -2.3 -6.3 -7.9 -7.2

Basic panel results Full

sample G-7 SOEs Scandinavia 1961-89 1990-2003

LC(-1) -0.036 (7.7)

-0.037 (5.5)

-0.0378 (5.0)

-0.0426 (3.5)

-0.0447 (6.3)

-0.0468 (4.4)

LRPDI(-1) 0.0277 (5.8)

0.0278 (4.6)

0.0291 (3.8)

0.033 (2.7)

0.0356 (5.1)

0.0354 (2.9)

LRNW(-1) 0.00396 (8.0)

0.00487 (3.6)

0.00388 (6.7)

0.00416 (5.8)

0.0042 (6.7)

0.0093 (6.2)

DLRPDI 0.229 (13.6)

0.262 (11.5)

0.205 (8.6)

0.184 (4.1)

0.217 (10.7)

0.253 (7.5)

DLRNW 0.0037 (3.7)

0.021 (3.3)

0.0033 (3.0)

0.0030 (2.9)

0.0033 (3.2)

0.0024 (4.3)

F (pooling) (90,2204) 3.9 [8.6]

(30,957) 6.1 [7.3]

(55,1247) 3.1 [7.9]

(15,448) 2.2 [6.5]

(90,1278) 3.5 [8.5]

(90.812) 3.3 [8.5]

F (fixed effects)

(18,2292) 7.7 [7.9]

(6,987) 5.6 [6.9]

(11,1302) 6.9 [7.3]

(3,463) 2.0 [6.1]

(18,1378) 5.5 [7.5]

(18,902) 3.4 [7.1]

R-bar sq 0.15 0.17 0.15 0.12 0.17 0.14 LM het 0.14 [0.91] 0.006 [0.94] 0.005 [0.982] 6.6 [0.01] 0.32 [0.574] 0.15 [0.696] DW 1.98

[0.31,0.39] 2.1 [0.941,0.970]

1.9 [0.063,0.106]

1.89 [0.077,0.17]

2.05 [0.805,0.871]

1.88 [0.022,0.046]

Implicit long run:- income effect 0.77 0.75 0.77 0.77 0.80 0.76 wealth effect 0.11 0.13 0.10 0.10 0.09 0.20

Dummies for crises Full

sample G-7 SOEs Scandinavia 1961-89 1990-

2003 BANK -0.0038

(2.7) -0.0004 (0.2)

-0.0069 (3.2)

-0.0022 (0.8)

0.0007 (0.4)

-0.0081 (3.9)

CURR -0.0032 (3.6)

-0.0013 (1.1)

-0.0043 (3.4)

-0.0021 (1.2)

-0.0014 (1.2)

-0.0047 (3.1)

BANK -0.0037

(2.7) -0.0004 (0.3)

-0.0072 (3.3)

-0.0034 (1.2)

0.0009 (0.4)

-0.0084 (4.0)

CURR -0.0031 (3.6)

-0.0013 (1.0)

-0.0042 (3.4)

-0.0014 (0.9)

-0.0013 (1.1)

-0.005 (3.3)

BANK(-4) -0.0012 (0.8)

0.0003 (0.5)

-0.004 (1.8)

-0.01 (3.5)

-0.0008 (0.4)

-0.00035 (0.2)

CURR(-4) -0.0009 (1.1)

-0.0019 (1.6)

-0.0002 (0.2)

0.0002 (0.1)

-0.0009 (1.2)

-0.0019 (1.3)

BANK -0.0038

(2.78) -0.0006 (0.3)

-0.0074 (3.4)

-0.0038 (1.4)

0.0009 (0.5)

-0.009 (4.2)

CURR -0.0031 (3.5)

-0.0014 (1.0)

-0.004 (3.2)

-0.0011 (0.6)

-0.0014 (1.2)

-0.0044 (2.9)

BANK(-4) -0.0014 (1.0)

0.0009 (0.5)

-0.0046 (2.1)

-0.011 (3.7)

-0.0011 (0.5)

-0.006 (0.3)

CURR(-4) -0.0008 (0.9)

-0.0019 (1.5)

-0.00002 (0.0)

0.0006 (0.4)

-0.0009 (0.8)

-0.0017 (1.1)

BANK(-8) -0.0018 (1.3)

-0.0011 (0.6)

-0.003 (1.4)

-0.0048 (1.7)

-0.00003 (0.0)

-0.0028 (1.5)

CURR (-8) 0.0011 (1.3)

0.0014 (1.1)

0.001 (0.9)

0.001 (0.6)

0.0013 (1.1)

0.0019 (1.4)

Memo items for dummies at level and (-4) Implicit long run income effect

0.77 0.75 0.79 0.81 0.80 0.90

Implicit long run wealth effect

0.10 0.12 0.09 0.07 0.09 0.12

Chart 1: Response of consumption to financial crises (quarters)

-0.016

-0.014

-0.012

-0.01

-0.008

-0.006

-0.004

-0.002

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

banking currency

Chart 2: Response of consumption to financial crises in the 1990s (quarters)

-0.035

-0.03

-0.025

-0.02

-0.015

-0.01

-0.005

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

banking currency

Subsets of crisis eventsDummy variables for systemic banking crises Full

sample G-7 SOEs Scandinavia 1961-89 1990-

2003 BANKSYS -0.0089

(4.1) -0.002 (0.4)

-0.01 (4.0)

-0.004 (1.3)

-0.0022 (0.6)

-0.01 (4.2)

BANKSYS -0.0086

(4.0) -0.002 (0.4)

-0.01 (4.0)

-0.005 (1.7)

-0.0007 (0.2)

-0.01 (4.2)

BANKSYS (-4)

-0.0005 (0.2)

0.0013 (0.3)

-0.0016 (0.6)

-0.009 (2.6)

0.0002 (0.1)

0.0019 (0.7)

BANKSYS -0.0087

(4.0) -0.002 (0.4)

-0.011 (4.0)

-0.006 (1.9)

-0.0007 (0.2)

-0.011 (4.2)

BANKSYS (-4)

-0.0006 (0.3)

0.0013 (0.3)

-0.0018 (0.7)

-0.01 (2.9)

0.0005 (0.1)

0.0016 (0.6)

BANKSYS (-8)

0.0007 (0.2)

0.0008 (0.2)

0.0017 (0.4)

-0.0027 (0.4)

0.002 (0.5)

0.0014 (0.4)

Dummy variables for twin crises Full

sample G-7 SOEs Scandinavia 1961-89 1990-

2003 TWIN -0.0096

(4.6) -0.0018 (0.7)

-0.018 (5.6)

-0.0073 (2.0)

-0.0004 (0.1)

-0.014 (5.3)

TWIN -0.0098

(4.6) -0.0018 (0.7)

-0.018 (5.6)

-0.008 (2.2)

-0.0004 (0.1)

-0.014 (5.2)

TWIN(-4) -0.001 (0.5)

0.0011 (0.4)

-0.0033 (1.0)

-0.013 (3.6)

-0.0001 (0.0)

0.0003 (0.1)

TWIN -0.01

(4.7) -0.0019 (0.7)

-0.019 (5.7)

-0.0086 (2.3)

-0.0005 (0.2)

-0.014 (5.5)

TWIN(-4) -0.001 (0.6)

0.001 (0.4)

-0.0037 (1.1)

-0.014 (3.8)

-0.0002 (0.1)

-0.0006 (0.2)

TWIN(-8) -0.0045 (2.1)

-0.004 (1.5)

-0.0055 (1.7)

-0.0082 (2.2)

-0.003 (0.8)

-0.0048 (1.9)

Dummies leveraged by household debt/income

Full sample

G-7 SOEs Scandinavia 1961-89 1990-2003

BANK -0.0083 (3.2)

-0.0018 (0.4)

-0.015 (3.9)

-0.0086 (1.5)

-0.0006 (0.2)

-0.028 (5.2)

CURR -0.0066 (3.7)

0.0015 (0.4)

-0.01 (4.0)

-0.005 (1.2)

-0.0038 (1.4)

-0.0092 (3.1)

BANK(-4) 0.0057 (2.2)

0.0045 (0.1)

0.004 (1.1)

-0.0053 (0.9)

0.006 (1.6)

0.0086 (2.0)

CURR(-4) -0.0021 (0.1)

-0.0033 (1.0)

0.0016 (0.6)

0.0039 (0.8)

0.0013 (0.5)

-0.0065 (4.2)

BANKDY 0.0059 (2.1)

0.0019 (0.3)

0.0088 (2.6)

0.0043 (1.1)

0.0016 (0.5)

0.026 (4.2)

CURRDY 0.0054 (2.3)

-0.0047 (0.9)

0.0083 (2.9)

0.0041 (1.1)

0.0037 (1.0)

-0.0072 (2.3)

BANKDY(-4) -0.0091 (3.3)

0.0008 (0.1)

-0.0098 (3.0)

-0.0042 (1.1)

-0.0082 (2.4)

-0.01 (2.1)

CURRDY(-4) -0.0016 (0.7)

0.0023 (0.4)

-0.0032 (1.1)

-0.0034 (0.8)

-0.0038 (1.0)

0.0045 (1.5)

Points to note• Crises take time to impact and effects build• Effects largest in small open economies and since

1990• Role for leverage in augmenting eventual effects• Higher gearing, higher effect of liquidity

constraints…but high gearing in liberalised systems with less liquidity constraints

• Mean D/Y 0.65 to 1989, 0.84 from 1990• In first year crisis effect ameliorated by high D/Y,

but in second makes substantially worse

Conclusions

• Literature focuses largely on fiscal costs and impact of crises on GDP

• Seen that consumption key element of macroeconomic adjustment

• Effect is worsened by high leverage, despite the benefit of liberalisation

• Also more serious in Small Open Economies – especially if macro policy cannot adjust to offset

• Need for macroprudential focus on household leverage

References

• Barrell, R., Davis, E.P., and Pomerantz, O., (2006) "Costs of Financial Instability, Household-Sector Balance Sheets and Consumption", Journal of Financial Stability, 2, 194-216

• Davis E P and Stone M (2004), "Corporate Financial Structure and Financial Stability", Journal of Financial Stability, 1, 65-81