The Loss of Confidence on Bank Money in the Great Depression by pptfiles

VIEWS: 8 PAGES: 1

									he Loss of Confidence in Bank Money in the Great Depression
by

A. Gerali & F. Passacantando
Joint BoE / ECB Conference on Financial Stability Frankfurt, November 13th, 2007

A sharp decline of stock of money associated with fall in output
110 100 90 80 70 60 50 40 Jan-25

Jan-26

Jan-27

Jan-28

Jan-29

Jan-30

Jan-31

Jan-32

Jan-33

Jan-34

Jan-35

Jan-36

Industrial Production (adjusted)

Money M2

Outline

• A glimpse at traditional views • The impact of the disruption of the payment system
• Quantitative evaluation

• Lessons for 2007 crisis

Standard explanations
• Different views on the direction of causality:
 from money to income  from income to money

• The prevailing view: a sharp decline in money supply (due to Fed misperception of the problem, Gold Standard rigidity…) amplified the fall in output (Friedman & Schwartz).

Our approach: focusing on transaction services
• The focus is not on the quantity of money but on transaction services associated with the use of a given quantity of money (money as medium of exchange)

• The capacity of money to serve as a payment medium depends on a confidence factor, because most transactions are based on instruments bearing credit risk
• Credit risk has three components:
 Debtor risk: probability of insolvency of the debtor  Settlement risk: illiquidity/ insolvency of the Bank  Systemic risk: domino and contagion effects

• All three factors played a role in reducing the acceptability of bank money in the Great Depression

The disruption in the payment system due to bank failures
Money took an unprecedented drop, but payments made with bank money declined even more sharply
Banks Deposits
1 20 10 1 500,000 1 00 90 80 300,000 70 60 50 1 00,000 40 30 Jan-25 Jan-26 Jan-27 Jan-28 Jan-29 Jan-30 Jan-31 Jan-32 Jan-33 Jan-34 Jan-35 Jan-36 0 200,000 400,000 600,000

Money M2

Deposits Turnover

Banks Suspended Deposits

The effect on real activity
• Limited ability to hedge against credit risk embedded in payments instruments • Diversification can take place out of bank deposits into currency and various forms of script money
• However the process is long ……….. • …………and costly

• Increase in transaction costs has real effects on the economy

Policy implications: difference from monetarist view
• (F&S) Whatever shock started the depression, its effects were magnified by an unprecedented decline in the quantity of money (mainly on account of a policy mistake by the Fed)

• Banking crises were important only because they led to a reduction in the stock of money • The impact of same contraction of money supply with no bank failures would have been “equally severe and probably even more so” • In our opinion the loss of confidence on bank money and the unwillingness of the public to use it as a payment medium would have limited the impact of a more expansionary monetary policy on economic activity • Structural policies should accompany macro policies

Policy implications complementarity with Bernanke’s view
• Banks failures disrupted the ability of channeling funds from savers into the hands of good borrowers
• Increases in borrowing costs and credit rationing

• … and turned a normal recession (up to October 1930) into a full depression. Thus banking crises amplified the recession.
• Our view complements it by adding payment services to the picture

Relationship between money, transaction services and economic activity
N

Transaction services of money

M P



Costs of credit intermediation (Bernanke)

CCI

Y

Income Velocity
(Friedman & Schwartz)

A model of transaction services
A typical model of the relationship between transaction services (N) and money holdings emphasizes a factor linked to the uncertainty about the future level and variability of prices (Klein 1974): M N  f ( ,) Price risk factor P We focus on bank money and add a factor linked to the risk that the banking payment instrument might not settle (credit risk):

Mb N  f( ,  , ) P

Confidence risk factor

The econometric exercise
We quantify the impact of variations of  on the production of transaction services N. Our proxy for the transaction services N is the ratio between bank debits (the total amount of the financial transactions in bank money) and the amount of bank deposits. We call it deposits turnover (DT).
Our proxy for  is the amount of bank deposits held in suspended or failed banks. The regression we estimate is:
DTt  f ( Dep . Failed Bankst ; X t )

where X is a vector of controls.

The dependent and independent variables
Deposits Turnover 5,0 Velocity
4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0

Suspended Deposits
500.000 450.000 400.000 350.000

Deposits Turnover

300.000 250.000 200.000 150.000

Suspended Deposits

100.000 50.000 -

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

The results
Table 1: Regressions on Log of Deposits Turnover (lnDT)
Equation no. lnDT(-1) lnSVOL lnCYCLE lnPCPR lnM2 lnBSDEP(-1) lnBSDEP(-2) lnBSDEP(-3) lnBSDEP(-4) AR(1) error coeff. Std. error Adj. R2 D.W. Impact from a 100% increase in BSDEP 0.5715 0.1010 0.55 2.24 (*)

(1)

(2)
(*)

(3)
(*)

(4)
(*)

(5) 0.1810
(*) (*) (*) (*) (*)

0.0064 0.6751 0.2943 0.8411
(*) (*)

0.2085 0.6998 0.4341 -0.0183 -0.0328 -0.0175 -0.0063 0.4882 0.1025 0.63 2.25 -7.49%

0.1834 0.4207 0.3216 -0.0212

0.1422 0.3300 0.4121 0.3985 -0.0188 -0.0308 -0.0146 -0.0017

0.2611 0.2898 -3.8807 0.3622 -0.0226 -0.0157 -0.0036 -0.0058

(*)

(*)

(*)

(*) (*)

(*) (*)

(*)

-0.0331 -0.0160 -0.0033

(*)

0.4052 0.0938 0.75 2.24 -7.37%

(*)

0.3877 0.0933 0.77 2.23 -6.59%

(*)

0.0843 0.92 2.17 -4.78%

Notes: Bold numbers indicate a coefficient significant at the 95% confidence level. An asterisk indicates significance above the 99% threshold. In all the equations, we instrument lnSVOL, lnCYCLE, lnPCPR and lnM2 using their lagged values.

The results
Table 1: Regressions on Log of Deposits Turnover (lnDT)
Equation no. lnDT(-1) lnSVOL lnCYCLE lnPCPR lnM2 lnBSDEP(-1) lnBSDEP(-2) lnBSDEP(-3) lnBSDEP(-4) AR(1) error coeff. Std. error Adj. R2 D.W. Impact from a 100% increase in BSDEP 0.5715 0.1010 0.55 2.24 (*)

(1)

(2)
(*)

(3)
(*)

(4)
(*)

(5) 0.1810
(*) (*) (*) (*) (*)

0.0064 0.6751 0.2943 0.8411
(*) (*)

0.2085 0.6998 0.4341 -0.0183 -0.0328 -0.0175 -0.0063 0.4882 0.1025 0.63 2.25 -7.49%

0.1834 0.4207 0.3216 -0.0212

0.1422 0.3300 0.4121 0.3985 -0.0188 -0.0308 -0.0146 -0.0017

0.2611 0.2898 -3.8807 0.3622 -0.0226 -0.0157 -0.0036 -0.0058

(*)

(*)

(*)

(*) (*)

(*) (*)

(*)

-0.0331 -0.0160 -0.0033

(*)

0.4052 0.0938 0.75 2.24 -7.37%

(*)

0.3877 0.0933 0.77 2.23 -6.59%

(*)

0.0843 0.92 2.17 -4.78%

Notes: Bold numbers indicate a coefficient significant at the 95% confidence level. An asterisk indicates significance above the 99% threshold. In all the equations, we instrument lnSVOL, lnCYCLE, lnPCPR and lnM2 using their lagged values.

Relationship between money, transaction services and economic activity
Our regression
Transaction services of money

N

anecdotal evidence

M


P

Costs of credit intermediation (Bernanke)

CCI

Y

Income Velocity
(Friedman & Schwartz)

• We showed that transaction services embedded in bank money likely declined. • But did this had any sizeable effect on real activity?

Some remarks from those times …
“This factor of confidence is an intangible one and greatly underestimated. Gradually we have developed to a point where 90 per cent of our transactions are conducted on credit… (and) only one out of ten of our daily business transactions is a cash transaction. The nine others depend upon our state of mind. If doubt enters and one out of the nine deals falls through, there is a slowdown of ten per cent, and much less than a 10 per cent slowdown changes the nation’s figures from black to red”
(Merle Thorpe “Cheerful Facts about 1930” World’s Work, May 1931)

" Farmers do not like to take it because the stamp or sales tax is relatively heavy in view of the low prices they get for their products. The same is true in the case of merchants selling products on which the profit margin is small."
(Business Week, January 11 1933)

" The use of it [scrip money] is one form of inflation, which may be a very good thing at a time when prices are too low. But in view of the contrast which must always exist between scrip and "real money" and the constant danger of a loss of confidence in the former, it is not likely to produce any general rise in the level of prices as measured in terms of legal tender."
(The Christian Century, February 1933)

Some remarks from those times …
“This factor of confidence is an intangible one and greatly underestimated. Gradually we have developed to a point where 90 per cent of our transactions are conducted on credit… (and) only one out of ten of our daily business transactions is a cash transaction. The nine others depend upon our state of mind. If doubt enters and one out of the nine deals falls through, there is a slowdown of ten per cent, and much less than a 10 per cent slowdown changes the nation’s figures from black to red”
(Merle Thorpe “Cheerful Facts about 1930” World’s Work, May 1931)

" Farmers do not like to take it because the stamp or sales tax is relatively heavy in view of the low prices they get for their products. The same is true in the case of merchants selling products on which the profit margin is small."
(Business Week, January 11 1933)

" The use of it [scrip money] is one form of inflation, which may be a very good thing at a time when prices are too low. But in view of the contrast which must always exist between scrip and "real money" and the constant danger of a loss of confidence in the former, it is not likely to produce any general rise in the level of prices as measured in terms of legal tender."
(The Christian Century, February 1933)

Inside the “liquidity shock” of Christiano, Motto, Rostagno
• Recently, an influential paper on the Great Depression goes as far as saying that it is all that it takes to explain the recession phase of the depression. • CMR estimated a dynamic stochastic general equilibrium macromodel with detailed real and financial sectors (and various shocks: technology, preference, financial, monetary policy, etc … ) • According to their estimation, the depression was started by a liquidity preference shock that greatly resembles our “loss of confidence” factor • In essence, for some unexplained reason households shifted away from demand deposits and into currency! • Bank failures are part of this story (at least implicitly, but they were not used in the estimation procedure)

Conclusions & lessons for the crisis of 2007
• Great Depression was not a unique event. Disruption in payment system in the crises of Russia and Argentina
• In the current crisis the payment system has performed well

• However interbank market suffers from confidence crisis
• Also in 1930s interbank market froze:

The introduction of bankers and trade acceptances will “liquify the great volume of frozen floating commercial debt and unfreeze a great mass of book credit and accounts receivables by which businesses is carried on normally and by which it is being retarded at present” (Business Week, June 22, 1932)

Policy actions to restore confidence

• Liquidity injections and macro policies are obviously important

• But in 1933, deposit insurance, emergency lending powers to Fed were essential to restore confidence • Today?


								
To top