When money matters: liquidity shocks with real effects

John Driffill and Marcus Miller

Birkbeck and University of Warwick

Abstract• In their ‘workhorse model of money and liquidity’,

Kiyotaki and Moore (2008) show how tightening credit constraints can cut current investment and future aggregate supply.

• Aggregate demand matches current supply, thanks to a flex-price ‘Pigou effect’

• Switching from a flex-price to a fix-price framework implies that demand failures can emerge after a liquidity shock.

• Quantitative estimates by FRBNY using such a framework produce dramatic results: what about the analytics?

Diagram 1. Effect of a stochastic liquidity shock in US that lasts 10 quarters, Del Negro et al. (2009)

Macro Paradigm: Woodford’s Synthesis Interest and Prices (2003)

• marked decisive shift in monetary economics from

looking at the quantity of money to the cost of

borrowing (i.e. from Friedman back to Wicksell

• inspired by an over-arching vision: to create a new

synthesis reconciling mainline macroeconomics with

dynamic General Equilibrium (GE), as practised by

RBC theorists in particular.

Hammond’s view of GE without credit constraints?

The Arrow Debreu paradigm at risk?

• In the absence of collateral or other credible

enforcement , Peter Hammond (1979) argued that, the

‘core’ of the inter-temporal GE model is not sub-game

perfect. Further discussion tomorrow?

• If this is true for Arrow-Debreu paradigm of GE, it is also

true for the DSGE specialisation developed for

macroeconomics. Does it matter?

Great Moderation has succumbed to credit crunch

• US unemployment rate has doubled from 4.8 per cent to

9.8 percent

• “the world is currently undergoing an economic shock

every bit as big as the Great Depression shock of 1929-

302” Eichengreen and O’Rourke (2009).

• “The good news is that the policy response is very

different” (zero interest rate, deficit spending, QE)

• The bad news is that it lies outside the reach of DSGE

• Can Kiyotaki and Moore (2008) model help?

KM(2008) framework

• addresses the Hammond critique of DGE: firms cannot borrow at will - with real consequences for composition of output.

• heterogeneous investors facing liquidity constraints

• want to hold money as a precaution against a lack of finance for investment opportunity

• need to add sticky wages/prices for liquidity shocks to have significant real effects;

• otherwise the Pigou effect acts as automatic stabiliser!

Woodford’s synthesis: the New Paradigm

Supply conditions

Goods market

Money market

RBC (representative agent with RE)

Productive efficiency (Flex-price, FE)

Inter-temporal optimisation (Euler equation)

Efficient contracts (money an epiphenomenon)

Keynesian Orthodoxy

Fix-price,UE (Phillips curve)

Agg. Demand (IS)

Liquidity Preference (LM)

Kiyotaki and Moore(2008), but with sticky wages/prices as in FRBNY and Driffill/Miller

Supply conditions

Goods market

Money market

RBC (representative agent with RE)

Productive efficiency (Flex-price, FE)

Inter-temporal optimisation subject to

Efficient contracts (money an epiphenomenon)

Keynesian Orthodoxy

Fix-price,UE (Phillips curve)

(credit constraints) Agg. Demand (IS)

Liquidity Preference (LM)

Fix price macro

• If prices are inflexible downward, there will be no Pigou

effect to stabilise aggregate demand in the face of a fall

of investment

• A fall in demand will contract employment if the real

wage is determined by bargaining, as argued for the UK

in Layard and Nickell, Alan Manning.

• Graphical representation follows of how liquidity

contraction can cut income conditional on K and q.

45°

L

wage bill (w*L)

X

‘workers spend what they earn;

entrepreneurs earn what they

spend’

Marginal Product of Labour

Aggregate Demand

Bargaining Wage

w*

X

Xf

real wage rate

L

D(X;q,K,)

μ

E

E*

D

D*

Net Output (X = r(Y)K)

Net Output (X=r(Y)K)

Figure 3. Short-run determination of X and Y

Kiyotaki and Moore (2008): “Liquidity, Business Cycles, and Monetary Policy”

• Assets involved:

• Money and equity

• Money is liquid

• Equity is not (completely) liquid – only a fraction of holdings can be sold each period– only a fraction of newly produced capital goods can

be financed by issuing new equity

Flex price to Fix price

Planes of stationarity

IR, Investment for Replacement I=(1-λ)K

AM, Asset Market equilibrium

GM, Goods Market equilibrium

Other variables

Dynamics of adjustment

Otherwise, Capital adjusts

out of equil, Share prices adjust

out of equil, Prices move

n/a

Variable determined

K q p M, φ

Nature State variable Jump Jump exogenous Notes Two unstable and one stable eigenvectors In 3 dimensional p,q,K space, with one state variable,K.

Planes of stationarity

IR, Investment for Replacement I=(1-λ)K

AM, Asset Market equilibrium

Change of status (Goods Market not in equil)

Other variables as before

Dynamics of adjustment

Otherwise, Capital adjusts

out of equil, Share prices adjust

n/a n/a

Variable determined K q p M, φ Nature State variable Jump exogenous exogenous Note One unstable and one stable eigenvectors In 2 dimensional q,K space, with one state variable,K.

Workers – not the focus of attention

• Spend what they get

• Rational and forward-looking, but impatient and credit

constrained.

• No borrowing

• They can hold money and equity if they choose

• Save nothing

• Consumption equals wages

Investment

Entrepreneurs can only finance investment using

money, selling existing equity claims to others,

raising equity on new capital, and spending out of

current income

Entrepreneurs – play central role, manage production and invest and hold assets

• May (prob π) or may not (prob 1-π) have an idea for a profitable investment

• Those with no ideas (no investment)– Consume – Save in form of money and equity holdings

• Those with an idea (Investors)– Buy new capital goods– Issue equity against them– Use money, other equity holdings, and current income

to finance investment

Liquidity constraints – on investment

• Entrepreneurs can raise equity against up to a fraction θ of new investment.

• They can sell off a fraction φt of pre-existing equity (theirs and others) nt

• Money is perfectly liquid

1 (1 ) (1 )t t t tn i n

1 0tm

Entrepreneur’s budget constraint

• Budget:

• p – price of money; q – equity price

• λ – 1-depreciation rate

• n equity held by entrepreneur

• Objective - max exp U:

1 1( ) ( )t i t t t t t t t t tc i q n i n p m m r n

log( )s tt s

s t

E c

Production

• CRS / C-D production function, capital and labour

• KM: wage clears labour market

• DM: fix money wage and price level – entrepreneurs keep the surplus

1t t t ty A k l

t t t t ty w l r k

Investment and Net Demand

(1 )1 1

t t t t t t

t t Rt t t

r q K p Mq I

q K

( ) 1

( ) 11

t t t

t t t t t tRt t

r x qr x K I K p M

q

Investment demand

Entrepreneurs’ income equals their demand (GM equilibrium)

1

1R t

t

qq

.

Entrepreneur’s Portfolio Balance (AM)

1 1 1

1 1 1 1

1 1 1 1 1 1

1 1 1 1 1 1 1

/ /(1 )

/ (1 ) /

(1 )

t t t t tt s

t t t t

Rt t t t t t t t

t R st t t t t t t

r q q p pE

r q N p M

p p r q q qE

r q q N p M

1 (1 )st t t t tN I K K

q

Equity Price

K

AM

E

Capital Stock

RI'GM

K*

Zero net investment

ΔK/Δt = 0

Asset price stationary

Δq/Δt = 0

GM'

AM'

RI

Δp/Δt = 0

Note that, at E’, the price level is lower than at E; 3D dynamics

E'

KM (2008): liquidity driven expansion, ϕincreases, equity more liquid

q

Equity Price

K

AM

E'

Capital Stock

RI'

K*

Zero net investment

ΔK/Δt = 0

Asset price stationary

Δq/Δt = 0

GM

AM'

RI

Fixed price, 2D dynamics with respect to AM and RI, GM not applicable: output is demand determined

E

Y

DM (2010): liquidity driven contraction

Saddle Path Dynamics in fix price case: driven by Asset Market and Investment disequilibrium

q

Equity Price

K

AM

E

Capital Stock

RI

K*

Zero net investment

ΔK/Δt = 0

Asset price stationary

Δq/Δt = 0

Note: Goods Markets has excess supply

Using AM and RI to get phase diagram

q

Equity Price

K

AM

E

Capital Stock

RI

SU

K*

K Zero net investment

ΔK/Δt = 0

Asset price stationary

Δq/Δt = 0 U

S

liquidity shock shifts E to E': with stock market fall leading to recession – or recovery if shock is to be

reversed

q

K

E

K**

U'

D

K*

E'

U'

L

A

A

I

P

 Figure 6. Numerical Results from DM simulation using FRBNY parameters

Calibration using FRBNY parameters (qtly)

• φ = 0.13 (fraction of existing assets an entrepreneur can sell);

• discount factor β = 0.99;

• fraction of new capital against which an entrepreneur can raise equity, θ = 0.13;

• probability of an entrepreneur having an idea for an investment, π = 0.075;

• the quarterly survival rate of the capital stock λ = 0.975

• [ our base case steady state: q = 1.12, r = 0.0374,

• Mp/K =0.1171, K = 152.5, y =17.26]

Temporary and permanent liquidity shock

t

Y

Table 2. Impact effects of a 20% cut in ϕ for different lengths of time

Short (2 years) Long (8 years) Permanent

q -1.25% -2.86% -3.57%

r -10.90% -12.23% -12.50%

X -10.27% -11.48% -11.73%

y -18.65% -20.54% -20.92%

Figure 8. Tobin’s q and the capital stock between the wars

Figure 9. Bubble collapse preceding liquidity shock: like 1929

q

K

U

E

K**

U

E'

D

K*

U'

U'B

Credit crunch

• With firms who want to invest more credit constrained -

and workers income constrained - no Pigou effect to

stimulate entrepreneurial consumption, a ‘credit crunch’

causes recession.

• The antidote discussed by KM should work here too:

Quantitative Easing as the government supplies

liquidity in exchange for corporate securities.

Conclusion• Switching from a flex-price to a fix-price framework

means that demand failures can emerge after a liquidity shock.

• AM and RI offer simple analytical treatment of impact and dynamic effects.

• Adding bubble might help explain the origin of the shock- it’s when the bubble bursts

• Need to add financial intermediaries to get to the heart of the matter

Current UK recession (blue) relative to earlier recessions(brown: 1930s, green and yellow: oil shocks)

Titanic sinking

Titanic

Queen Hermione imprisoned for sixteen years: then came reconciliation

Time for a change in Macro?

• discredited and discarded in the stagflation that followed

the oil price shocks of the 70s and 80s, the Keynesian

paradigm of macroeconomic stabilisation has suffered

in silence for many years.

• but the new DSGE paradigm failed to predict the ‘credit

crunch’ - or explain its effects.

• Let’s briefly review recent fashions in macro