Optimal Monetary Policy under Balance-Sheet Effects on the Non-tradable Sector in a Small Open Economy

Abstract

The choice of an exchange rate regime is crucial in small open economies (SOEs) with a dollarized financial sector. While the traditional Mundell–Fleming model supports a floating exchange rate, evidence shows that central banks frequently intervene in exchange markets. One of the reasons for these interventions is the consequences of large depreciations that could trigger negative balance-sheet effects. This paper extends the literature about the optimal monetary policy in SOEs, by considering a heterogeneous hedge across tradable and non-tradable sectors. Our findings support a ‘leaning against the wind’ policy as an optimal response to negative external shocks. This result is present even if only one sector of the economy faces credit constraints. We show that the vulnerability of the economy to large negative external shocks depends not only on the overall leverage, but also on the distribution of foreign currency debt across economic sectors.

Keywords:

• Credit constraints
• exchange rate intervention
• monetary policy
• tradable and non-tradable firms

JEL Classifications:

• E5
• F3
• G15

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 See Faia and Monacelli (2008), Sutherland (2005) and De Paoli (2009a, 2009b).

2 As it is reported in Calvo and Reinhart (2002) and Mihaljek (2005).

3 For a theoretical explanation of the effects of currency substitution on the volatility of money demand, see Miles (1978), McKinnon (1982) and Borensztein and Berg (2000).

4 As in Kiyotaki and Moore (1997).

5 Ganesh-Kumar et al. (2001) argue, in response to Ghironi and Melitz (2005), that exporting provides a signal, thus granting easier access to financial sources.

6 Based on the de-dollarization program (Contreras et al., 2019) followed by the Central Bank of Peru.

7 For a discussion, see Korinek and Jeanne (2013), Mendoza and Bianchi (2010), Chen et al. (2013) and Benigno et al. (2013).

8 See Sanchez (2006).

9 We assume the ‘law of one price’ in the tradable sector: $P_t^T=S{P}_t^{T,\ast }$.

10 Further, the LM curve is dependent on $Y_t^T$, $Y_t^N$ and $S_t$. We map $(Y^T,Y^N)$ to $Y_t^H$ with Equation (9(9) $P_t^H{Y}_t^H=P_t^N{Y}_t^N+P_t^T{Y}_t^T.$(9) ).

11 Similar to the LM in the unconstrained case, the LM curve is dependent on $Y_t^T$, $Y_t^N$ and $S_t$. We map $(Y^T,Y^N)$ to $Y_t^H$ with Equation (9(9) $P_t^H{Y}_t^H=P_t^N{Y}_t^N+P_t^T{Y}_t^T.$(9) ).

12 We compute the IS curve with Equations (28(28) $Y_t^T=\underset{\mathrm{Negative}}{\underbrace{\frac{A_t^T}{\kappa W_t^{\omega }(S_t{Q}_t^{\ast }{)}^{1-\omega }}}}[\underset{\mathrm{Positive}}{\underbrace{\frac{(1-\alpha \theta )(1-\alpha )}{\gamma }\tilde{X}{S}_t}}-\underset{\mathrm{Positive}}{\underbrace{\frac{\alpha (1-\theta )}{\gamma }(N_t^N-S_t{D}_t^{N,\ast })}}]$(28) ), (29(29) $Y_t^N=\underset{\mathrm{Negative}}{\underbrace{\frac{A_t^N{\omega }^{\omega }}{W_t^{\omega }(S_t{Q}_t^{\ast }{)}^{1-\omega }}}}{[\underset{\mathrm{Positive}}{\underbrace{\frac{(1-\alpha )\omega \alpha \theta }{\gamma }\tilde{X}{S}_t}}-\underset{\mathrm{Positive}}{\underbrace{\frac{\alpha \theta }{\gamma }(N_t^N-S_t{D}_t^{N,\ast })}}]}^{\omega }{\left[\underset{\mathrm{Negative}}{\underbrace{N_t^N-S_t{D}_t^{N,\ast }}}\right]}^{1-\omega },$(29) ) and (9(9) $P_t^H{Y}_t^H=P_t^N{Y}_t^N+P_t^T{Y}_t^T.$(9) ).

13 Under this assumption, $\tilde{X}$ has an exponential distribution.

14 We depart from Devereux and Poon (2011), as they assume a discrete distribution for $\tilde{X}\in \{X_t(1),\dots ,X_t(Z)\}]$, with a given probability for each shock.

15 Devereux and Poon (2011) prove this result. Under commitment, the monetary policy is unable to consistently exploit the market power that the economy has over home goods.

16 Indeed, this is a pre-shock policy. The normal state is defined as the average state, in which $\tilde{X}=1$.

17 See Appendix 2 for more details on the calibration of the optimal monetary policy.

18 The values assumed for the debt in foreign currency and the asset value in both sectors are detailed in Table A1.

19 In the flexible case, an increase in the foreign demand due to the positive shock increases the prices of the tradable goods in the foreign country. This generates a possibility of arbitrage from the law of one price, thus the exchange rate will appreciate and eliminate the arbitrage condition. However, in the case of a fixed exchange regime, the monetary policy should be expansionary to increase the prices of tradable goods. This expansion will eliminate the arbitrary condition in the law of one price and lead to a new equilibrium. This is how a currency peg is set in the model that makes the monetary policy an exchange rate stabilizer.

20 Nevertheless, it binds in the positive shock realization. An explanation for this result is that to stabilize the exchange rate, the monetary policy should be more expansive that increases the demand for intermediate inputs that thus helps the constraint to hold.

21 Also known as the Ramsey policy.

22 For a given exchange rate, same foreign debt implies same dollarization.

Additional information

Notes on contributors

Marco Ortiz

Marco Ortiz, Assistant Professor of the Academic Department of Economics of the Universidad del Pacífico. Ph.D. in Economics and Master in Research in Economics from the London School of Economics and Political Science.

Gerardo Herrera

Gerardo Herrera, Research Assistant at Universidad del Pacífico Research Center. B.A in Economics at Universidad Nacional Mayor de San Marcos.