To Fix, to Float or to Lay Somew here in Betw een? Eduardo J. J. - - PowerPoint PPT Presentation

To Fix, to Float or to Lay Somew here in Betw een?

Eduardo J. J. Ganapolsky May 2002

Outline

❚ Motivation ❚ What does the literature say? ❚ A different channel ❚ The model ❚ Exercise ❚ Results ❚ Conclusion and extensions

Motivation

❚ In response to some external shocks, countries could

react in different ways regarding to the foreign exchange market:

❙

Full intervention, keeping the ER fixed

❙

Not intervene at all, leaving the ER to fully depreciate

❙

Moderate intervention, allowing the ER to depreciate but avoiding sharp depreciations

❚ What are the factors behind those choices?

❙

Costly intervention

❙

Costly depreciations

What does the literature say?

❚ Fear of floating

❙

Empirical:

❘

Low variability of the nominal ER, even in the presence of real or nominal shocks

❘

Low variability stems from deliberate policy actions

• High variability of international reserves
• High variability of interest rates
• High variability of the domestic prices of commodities

❘

Fear of floating is pervasive in emerging markets

• Credibility problems (Calvo-Reinhart (2000a))
• Liability dollarization (Calvo-Reinhart (2000b))
• Higher degree of pass through and lower the ability to borrow in own

currency (Hausmann-Panizza-Stein (2001))

What does the literature say?

❚ Theoretical:

❙

Lahiri-Végh (2002):

❘

Central banks respond to pressures on their currency with intervention or higher interest rate

• Source of fear of floating: ER variability leads to output cost in the

presence of nominal wage rigidities

• Exogenous intervention cost

❘

Find non-monotonic relationship between nominal ER and the size

• f the monetary shock
• Developing countries are subject to bigger shocks, therefore they are

more reluctant to float than developed countries

❘

Either fix or float, do not find dirty floating

What does the literature say?

❚ Theoretical (cont.):

❙

Cavallero-Krishnamurthy (2001):

❘

Inelastic supply of funds during a crisis

• Monetary policy has no real effects
• ER is very sensitive to monetary policy
• Avoid overshooting because of inflationary consequences

❙

Parrado-Velasco (2002):

❘

Short run price rigidity and imperfect competition

• Optimal exchange rate policy implies a dirty float

A different channel

❚

Intervention is costly because it makes the government to cut valuable spending

❙

The government finances some expenditure with revenues coming from the return on reserves

❘

Hausmann et al. (2001) find that emerging markets hold high stock of reserves

❙

Tax revenues cannot be increased

❚

Exogenous depreciation cost

❙

Currency mismatch between assets and liabilities

❘

Hausmann et al. (2001): “The original sin”

• Find a strong negative link between ER flexibility and liabilities dollarization

❘

Calvo-Reinhart (2000b)

❘

Burnside-Eichenbaum-Rebelo (1999)

• Firms and banks borrow extensively from abroad but do not completely hedge

exchange rate risk

The model

❚ Small open economy ❚ Perfect capital mobility ❚ One tradable good ❚ LOOP holds ❚ Agents:

❙

Household-cum-firm

❙

Bank

❙

Government

❚ Additional ingredients:

❙

Fixed depreciation cost

❙

Private costs in the banking sector

The model

❚ Household

❙

Utility function:

❙

Financial wealth:

❙

Flow constraint:

❙

Produce goods according to:

❙

Money reduce transaction costs:

) ( ] ) ( [

t t t t t l t t t t

m v m i c l r r F y a r a

t

− − Ω + + − − + − + =

• τ

∫

−

= dt e c W

t t β

) log(

t t t t

l b m a − + =

d m m m v

t t t

+ − =

2

1 ) ( α

1 1 < < = η η

η t t

l y

The model

❚ Household

❙ FOCs:

r r l m v i c

l t t t t t

− = = − =

−1

) ( ' 1

η

λ

η

α λ

− −

− = − = =

1 1

) ( 2 1 1 r r l i m c

l t t t d t

d

The model

❚ Bank

❙ Profits:

where; (1-δ)q is a private cost, 0<δ≤1

❙ Balance-sheet: ❙ FOC: ❙ Zero-Profits:

θ

^

] [ E sl ql rb l r F

t t b t t l t t

− + − − + = Ω s q r r l

t

− = −

b t t

b l = θ

^

E F =

The model

❚ Government

❙ Flow constraint: ❙ Intertemporal constraint: ❙ Central Bank balance-sheet:

t t t t t

ql sl m m h r h

t

) 1 ( δ τ ε − + − − + + =

• dt

e ql sl dt e m m h r

rt t t rt t t − −

• ∫

∫

− − + = + + ) ) 1 ( ( ) ( δ τ ε

s t

m E D h

t t

= +

The model

❚ Initial steady state

❙ Assume:

❘ ε = µ = 0 ❘

h0 = 0

❙ Given that, if the government maximize the

household’s welfare, then:

❘

s = (1-δ)q

❙

and:

❘

❘

d r q ra c − − − + − + =

− − 2 1

) 2 1 ( ) )( 1 ( α τ δ η η

η η

q r r l

t

δ = −

Exercise

❚ Unexpected shock to the money demand(dα< 0)

❙ mt

d = ht + D/Et

• r

❚ What to do? ❚ Choose ∆ht and ∆Et such that they maximize the

post-shock welfare; or in other words, they minimize the deviation from the previous

• ptimum

Exercise

❚ If fix dh= dm ❚ If float dh= ds= 0 ❚ Trade-off

d r d ds q ra c fix − − + − + − − + =

− − 2 1

) 2 1 )( ( ) )( 1 ( ' α α τ δ η η

η η 2 1

/ ) 2 1 )( ( ) )( 1 ( ' E D dm d r d q ra c float θ α α τ δ η η

η η

+ − − + − + − + =

− − 1 1

/ ] ) ( ) )[( 1 ( E D dm q ds q c c

float fix

θ δ δ η η

η η η η

− − − − = −

− − − −

Exercise

❚ The problem is:

^ 1 1 ^ 2 1

/ ) ( ] ) [( ) 2 1 )( ( ) )( 1 ( ' E D dm dh E ds ds q l q s d rdh E d r d ds q ra c − = − = − = − + − − + + − − + =

− − − − η η η

δ δ θ τ α α δ η η

∫

−

= dt e c W Max

t t β

) log(

Exercise

❚ Intuition

❙ The marginal disutility of reducing the subsidy

(reduces output) should be compensated for the marginal utility coming from intervention (reduces depreciation costs)

r W f W ds rdh dh ds f W L

dh f

= − − + = ' ' )] ( [ ] ), ( [ φ φ ξ

Results

❚ Case 1:

❙ δ= 1; no private costs ❙ θ= 0; no depreciation costs

either fix or float

❚ Case 2:

❙ δ> 0 ❙ θ= 0

fully depreciate

Results

❚ Case 3:

❙ δ> 0 ❙ θ> 0

❘ For θ small enough: fully depreciate (dh= 0) ❘ For bigger θ: intervene (dh< 0)

❙ The interior solution for dh, gives:

❘ dh = g(θ, η, m0, δ) < 0 ❘ gθ< 0; gη> 0; gm0> 0; gδ< 0

Results

❚

gθ< 0

❙

The higher the depreciation cost, the higher the intervention

❚

gη> 0

❙

The lower the bank loans productivity, the higher the intervention

❚

gm0> 0

❙

The lower the initial real money stock, the higher the intervention

❚

gδ< 0

❙

The lower the distortion, the higher the use of the “distortionary tax” to finance the intervention

❚ How about emerging markets?

Conclusion

❚

Introduces a new trade-off between an “output effect” and a “depreciation cost” generated by a financial need

❚

Finds “partial” depreciations

❚

Emerging markets intervene more than developed countries

❚

It is key the absence of non-distortionary taxes: the government can raise resources only through “distortionary” taxation

❙

Focused on a particular case: distortion in the financial sector

❙

Trade-off between helping the financial sector and keeping the value of the currency

Extensions

❚ Introduces inflation tax as an alternative source of funds ❚ Model explicitly the depreciation cost coming from the

currency mismatch

❚ Incorporate some appreciation cost (traditional

competitiveness story) to obtain intervention on both ups and downs in the exchange rate

❚ Generalize the channel as a “fiscal” explanation of the

fear of floating phenomenon