Abstract
This paper shows how the macroeconomic instability that affects more financially open emerging economies can be explained in terms of imperfect international financial integration. The analytic tool used is a stochastic dynamic equilibrium model. The model, based on a small open economy, is calibrated to Malaysia. International financial relations are characterized by the presence of a borrowing constraint, that amplifies the volatility of exogenous shocks. Impulse response and simulation analyses are conducted. Main result is that, in presence of international financial frictions, greater access to international liquidity can cause high short-run macroeconomic instability.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See also Paasche (2001). Iacoviello and Minetti (2003) analyze the effects of the financial liberalization on the business cycle of open industrialized countries. Another strand of the literature on the macroeconomics of emerging markets has studied the implications of financial frictions—such as borrowing constraint or risk premia—for the optimality of currency regimes. See among the others Cespedes et al. (2004). Alternatively, it is possible to formalize financial frictions in terms of share of households that do not have access to financial markets, but in each period consume all available income (rule of thumb agents), as in Gali et al. (2004). Rule of thumb agents can be thought as agents that want to borrow but unfortunately cannot, because the loan-to-value ratio in the borrowing constraint is set to zero in each period.
Faia (2008) uses a similar framework to perform a welfare analysis of capital account liberalization.
The first group includes many countries of the South-Est Asia and South America. The second group many countries of Africa.
To take into account for the possible smoothing role played by public expenditure—that could be important especially in developing economies—Kose et al. (2003a) also report total consumption. This variable is generally less volatile than private consumption. However, the general pattern is confirmed: consumption volatility declined for industrial and less financially open developing economies, while it increased for the more financially open developing countries.
See also Kose et al. (2003b).
Real estate services is a rather standard assumption in macroeconomic models of the real estate credit channel. The real estate is formalizes as a real asset, that gives utility to households, in such a way to get a positive demand for it in each period. See Iacoviello (2005) among the others.
The utility based price index P is defined as the minimum expenditure required to buy one unit of the composite good, given the prices of the Home and Foreign goods. See Corsetti and Pesenti (2001) among the others.
The consumption demands are the solution of an intratemporal allocation problem, consisting in allocating a given level of nominal expenditure among domestically produced and imported goods. See Corsetti and Pesenti (2005). A similar problem applies to the investment demand.
With a slight abuse of notation, I use the terms “real estate” and “asset” as interchangeably in the paper.
Hence, \(p_{H}=\frac{P_{H}}{P}\).
This symplifying assumption is rather common in the open economy literature, where the public sector does not issue public debt nor raise distortionary taxation. See for example Corsetti and Pesenti (2001).
This assumption is rather common in the literature. See, for example, Iacoviello (2005). This is not necessarily unrealistic if we interpret housing in a broad way, which also includes land.
See the Appendix for derivation.
I do not consider the years after 1996 to avoid the macroeconomic effects of the East-Asia currency and financial crises.
The liberalization of the capital account, that was accompanied by measures to deregulate the financial system, began in the late 1980s. See Ariyoshi et al. (2000).
For more details, see Kim et al. (2003).
I have estimated the two laws of motion using annual logged and filtered data of Malaysia. I have used the Hodrick–Prescott filter, with smoothing parameter equal to 100. The data go from 1963 to 1996. Data for export are from the International Financial Statics of the International Monetary Fund. Those for technology are calculated using data of the Penn Table.
If the variance of the shocks is very large, the agent might not borrow up to the limit after a long series of positive productivity or export shocks. Instead, she can decide to keep a buffer stock of resources to use in bad times to avoid the possibility of hitting the borrowing constraint. By assuming that the discount rate of the constrained agents, γ, is well below that of the agents in the rest of the world, β, I can minimize the probability that credit constraints become non-binding in some states of the world.
The model without the borrowing constraint is a standard small open economy real bysiness cycle model. The first order conditions with respect to B t and h t are now:
$$ E_{t}\left( -\frac{1}{C_{t}\left( 1+r\right) \Phi (B_{t})}+\gamma \frac{1}{ C_{t+1}}\right) =0 $$$$ \frac{q_{t}}{C_{t}}=j^{\,h}\frac{1}{h_{t}}+E_{t}\left( \gamma \frac{q_{t+1}}{ C_{t+1}}\right) $$I assume the steady state of the model without borrowing constraint is similar to that of the model with m = 0.3. To make the model stationary, I assume the Home agent is as patient as the agents in the rest of the world (hence \(\gamma =\beta =\frac{1}{1+r}\)) and that there is a function Φ(B t ) capturing the cost of undertaking positions in the international asset market. The agent does not take into this extra-term when maximizes her utility. As borrower, the agent will be charged a premium on the interest rate; as lenders, she will receive a remuneration lower than the interest rate. I introduce this additional cost to pin down a well defined steady state. The payment of this cost is rebated to agents belonging to the rest of the world. I adopt the following functional form:
$$ \Phi (B_{t})=\exp \left[ \delta _{b}\left( B_{t}-B\right) \right] $$where − 1 ≤ δ b ≤ 0 and B is the steady state asset position. The parameter δ b controls the speed of convergence to the steady state. I set it as small as possible, compatibly with the stationarity of the model. See Schmitt-Grohé and Uribe (2001) for more details.
The exercises consist in simulating the model in correspondence of different values of ρ. Remaining parameters are kept constant to their baseline values (see Table 2). When ρ is equal to 2 (1.1 in the benchmark), the volatilities of consumption, investment and output are respectively equal to 2.2 (3.54 in the benchmark), 7.9 (14.27), 3.1 (3.15).
In steady-state, the amount of export is equal to:
$$ \frac{X}{Y}=1-a_{H}\left( \frac{C}{Y}+\frac{I}{Y}+\frac{G}{Y}\right) $$See the Appendix for more details.
When a H is equal to 0.75 (0.44 in the benchmark), the volatilities of consumption, investment and output are respectively equal to 3.43 ( 3.54 in the benchmark), 13.9 (14.27), 3.18 (3.15).
When the persistence of both shocks is equal to 0.92, the volatilities of consumption, investment and output are respectively equal to 4.93 (3.54 in the benchmark), 14.98 (14.27), 3.13(3.15).
References
Aghion P, Bacchetta P, Banerjiee A (2004) Financial development and the instability of open economies. J Monet Econ 51(6):1077–1106
Aguyar M, Gopinath G (2007) Emerging market business cycles: the cycle is the trend. J Polit Econ 115:69–102
Ariyoshi A, Habermeier K, Laurens B, Otker-Robe I, Canales-Kriljenko JI, Kirilenko A (2000) Capital controls: country experiences with their use and liberalization. International monetary fund occasional paper, 190
Caballero RJ, Krishnamurthy A (2001) International and domestic collateral constraints in a model of emerging market crises. J Monet Econ 48:513–548
Cespedes LF, Chiang R, Velasco A (2004) Balance sheets and exchange rate policy. Am Econ Rev 94(4):1183–1193
Collyns C, Senhadji A (2002) Lending booms real estate bubbles and the Asian crisis. IMF working paper, WP/02/20
Corsetti G (1998) Interpreting the Asian financial crisis: open issues in theory and policy. Asian Dev Rev 16(2):1–44
Corsetti G, Pesenti P (2001) Welfare and macroeconomic interdependence. Q J Econ 116(2):421–445
Corsetti G, Pesenti P (2005) International dimensions of optimal monetary policy. J Monet Econ 52(2):281–305
Corsetti G, Pesenti P, Roubini N (1998) What caused the Asian currency and financial crisis. Japan World Econ 11:305–73
Faia E (2008) Macroeconomic and welfare implications of capital account liberalization, mimeo
Gali J, Lopez-Salido JD, Vallés J (2004) Rule of thumb consumers and the design of interest rate rules. J Money, Credit Bank 36(4):739–764
Iacoviello M (2005) House prices, borrowing constraints and monetary policy in the business cycle. Am Econ Rev 95(3):739–764
Iacoviello M, Minetti R (2003) Financial liberalization and the sensitivity of house prices to monetary policy: theory and evidence. Manch Sch 71(1):20–34
Jansen WJ (2003) What do capital inflows do? dissecting the transmission mechanism for Thailand, 1980–1996. J Macroecon 25:457–480
Kiyotaki N, Moore J (1997) Credit cycles. J Polit Econ 105(2):211–48
Kim SH, Kose A, Plummer M (2003) Dynamics of business cycles in Asia. Rev Dev Econ 7(3):462-477
Kose MA, Prasad ES, Terrones ME (2003a) Financial integration and macroeconomic volatility. IMF working paper, WP/03/50
Kose MA, Eswar SP, Terrones ME (2003b) Volatility and comovement in a globalized world economy: an empirical exploration. IMF working paper, WP/03/246
Krugman P (1999) The return of depression economics. W.W. Norton, New York
Mera K, Renaud B (eds) (2000) Asia’s financial crisis and the role of real estate. M.E. Sharpe, Armonk
Paasche B (2001) Credit constraints and international financial crises. J Monet Econ 48:623–650
Radelet S, Sachs J (1998) The onset of the East Asian financial crisis. NBER working paper no. 6680
Schmitt-Grohé S, Uribe M (2001) Closing small open economy models. J Int Econ 61:163–185
Schneider M, Tornell A (2004) Balance sheet effects, bailout guarantees and financial crises. Rev Econ Stud 71(3):883–913
Tornell A, Westermann F (2002) The credit channel in midlle income countries. NBER working paper no. 9355
Tornell A, Westermann F (2003) Credit market imperfections in middle income countries. NBER working paper no. 9737
Tornell A, Westermann F, Martinez L (2004) The positive link between financial liberalization growth and crises. NBER working paper no. 10293
Acknowledgements
This paper is an extract from one chapter of my Phd thesis at LSE, defended on January 2007. I owe thanks to my supervisor, G. Benigno. I also thanks for useful comments K. Aoki, G. Corsetti, B. De Paoli, M. Eichenbaum, N. Kiyotaki, M. Iacoviello, R. Minetti, C. Pissarides, A. Sarichev, and seminar participants at the CEP-FMG International Financial Stability Seminar, ZEI Summer School 2003 on ‘Monetary Theory and Policy’, CFS Summer School 2003 on ‘Empirical Methods for Macroeconomic Models’, Universitat Pompeu Fabra-EDP Jamboree 2003, Bank of Italy Research Department (2004), University of Rome III (2005). All errors are mine. Usual disclaimers hold.
Author information
Authors and Affiliations
Corresponding author
Appendix: The solution of the model
Appendix: The solution of the model
The equilibrium dynamics is characterized by solving a first-order log-linear approximation to the equilibrium conditions around the non-stochastic steady state. In what follows, the deterministic steady state and the log linearized equations are shown.
1.1 The steady state equilibrium
A steady-state equilibrium is considered in which all the shocks are zero, there is no net capital accumulation (K t = K t − 1 = K), no debt change ( B t = B t − 1 = B) and in which all price are constant. The following price normalization is used: P H = P F = P = 1. Hence, the relative prices are equal to one.
The steady state value of r satisfies:
where β is the discount factor of the rest of the world. Using the steady state version of the consumer Euler Eqs. 13 and 14, the following value for λ, the Lagrange multiplier, can be obtained:
By assumption, 0 < γ < β < 1; so the Lagrange multiplier is strictly greater than zero in the steady state (and in a small neighborhood of it). As a consequence, the borrowing constraint is binding.
I normalize the steady-state value of the total productivity factor, A, to 1. From the firms’ first order conditions, Eqs. 19 and 20, the following two equations are respectively obtained:
From Eqs. 15, 16, 34 the capital-to-output ratio is:
From the capital accumulation law Eq. 12, the investment-to-output ratio can be derived:
The borrowing constraint Eq. 11 becomes:
From the budget constraint Eq. 10, using Eqs. 34 and 35, the consumption-to-output ratio can be written as:
Combining Eqs. 38, 39 and the steady state version of Eq. 14 the following equation for the real estate value-to-output ratio is obtained:
Given that I/Y is given by Eq. 37 and G/Y (as well as T/Y) is exogenous, the ratio is function of the structural parameters of the model. Substituting back into Eqs. 38 and 39, also the ratios B/Y and C/Y become function of the structural parameters.
From the intratemporal first order conditions of Home agents—Eqs. 6, 7, 8—the following steady state allocations are obtained:
The demand of the rest of the world for the Home good in steady state is (see Eq. 24):
The Home demand for the Foreign good is (see Eq. 25 and the above steady state equations of C F , I F , G F ):
Finally, the budget constraint of the public sector is:
1.2 The Loglinearized equilibrium
Local dynamics is studied by linearizing the equilibrium conditions around the steady state.
Let variables with a time subscript t, t + 1 or t − 1 denote log-deviations from steady-state and let those without a time subscript denote steady state values. The log-linearized equilibrium of the model is system of 14 equations in 14 variables: \(\hat{Y},\) \(\hat{Y}_{F},\) \(\hat{C},\) \(\hat{B},\) \(\hat{K},\hat{I},\) \(\hat{L},\) \(\hat{q},\hat{h},\) \(\hat{p}_{H},\) \( \hat{p}_{F},\) , \(\hat{T},\hat{A},\hat{X}\). They are log-deviations from steady state of, respectively, output, amount of imports, consumption, net foreign asset position, stock of capital, investment, labor supply, real estate price index, real estate holdings, price of the Home good, price of the Foreign good, terms of trade, technology, amount of export (for example, \( \hat{Y}_{t}\equiv \ln (Y_{t}/Y)\)).
To save on notation, I drop the expectation operator between variables dated t + 1. However the variables must be intended as in expected value conditional on the information available at time t.
The equations are the following ones:
-
1.
Aggregate demand
$$ \hat{Y}_{t}=-\rho \left( \hat{p}_{H}\right) +a_{H}\frac{C}{Y}\hat{C}_{t}+a_{H}\frac{I}{Y}\hat{I} _{t}+\frac{X}{Y}\hat{X}_{t} $$(47)$$ \hat{Y}_{F,t}=-\rho \left( \hat{p}_{F}\right) +\left(1-a_{H}\right)\frac{C}{Y}\hat{C}_{t}+\left(1-a_{H}\right)\frac{I }{Y}\hat{I}_{t} $$(48)$$\begin{array}{rll} \hat{q}_{t} &=&\left( 1-m\beta \right) \hat{C}_{t} \nonumber \\ [2pt] &&+\left( \gamma +m\left( \beta -\gamma \right) \right) \left( \hat{q} _{t+1}\right) \nonumber \\ [2pt] &&-(1-\gamma -m\beta +m\gamma )\hat{h}_{t} \nonumber\\ &&-\gamma \left( 1-m\right) \hat{C}_{t+1} \end{array}$$(49)$$\begin{array}{rll} \left( \hat{I}_{t}-\hat{K}_{t-1}\right) &=&\gamma \left( \hat{I}_{t+1}-\hat{K }_{t}\right) +\frac{1}{\psi }\left( \hat{C}_{t}-\hat{C}_{t+1}\right) \nonumber \\ && +[1-\gamma \left( 1-\delta \right) ]\left[ \hat{P}_{H,t}+\hat{Y}_{t+1}- \hat{K}_{t}\right] \end{array}$$(50) -
2.
Borrowing constraint
$$ \hat{B}_{t}=\hat{q}_{t+1}+\hat{h}_{t} $$(51) -
3.
Aggregate supply
$$ \hat{Y}_{t}=\alpha \hat{K}_{t-1}+\left( 1-\alpha \right) \hat{L}_{t} $$(52)$$ \hat{p}_{H,t}+\hat{Y}_{t}=+\hat{C}_{t}+\tau \hat{L}_{t} $$(53) -
4.
Flows of funds/ Other variables
$$ \hat{K}_{t}=\left( 1-\delta \right) \hat{K}_{t-1}+\delta \hat{I}_{t} $$(54)$$ \beta \frac{B}{Y}\hat{B}_{t}=\frac{B}{Y}\hat{B}_{t-1}+\hat{p}_{H,t}+\hat{Y} _{t}-\frac{C}{Y}\hat{C}_{t}-\frac{I}{Y}\hat{I}_{t} $$(55)$$ a_{H}\hat{p}_{H,t}+\left( 1-a_{H}\right) \hat{p}_{F,t}=0 $$(56)$$ \hat{h}_{t}=0 $$(57)$$ \hat{T}_{t}=\hat{p}_{F,t}-\hat{p}_{H,t} $$(58) -
5.
Shock process
$$ \hat{A}_{t}=\rho _{A}\hat{A}_{t-1}+\varepsilon _{A,t} $$(59)$$ \hat{X}_{t}=\rho _{X}\hat{X}_{t-1}+\varepsilon _{X,t} $$(60)
The equations are divided in 5 blocks. The first block contains the demand side of the economy.
The first equation is the demand for the Home produced good (it is the loglinearized version of the Eq. 24).
The second equation is the demand for the imported good (it is obtained by loglinearizing the Eq. 25).
The third equation is the derived from the agent Euler consumption Eq. 13 and real estate demand Eq. 14.
The fourth equation is obtained by combining the investment and capital first order conditions, respectively Eqs. 15 and 16.
The second block is the borrowing constraint. It derives from Eq. 11.
The third block is the supply side of the economy.
The first equation derives from the technology constraint Eq. 18.
The second equation derives from the labor first order conditions Eqs. 17, 20 and from the labor market clearing condition Eq. 22.
The fourth block contains equations describing flows of funds and remaining variables.
The first equation is the log-linearized version of the capital accumulation law Eq. 12.
The second equation derives from the budget constraint Eq. 10.
The third equation defines the consumption as the numeraire (see Eq. 5, with P = 1).
The fourth equation derives from the real estate market clearing condition Eq. 23. The real estate is in a fixed amount.
The fifth equation is the Home terms of trade. It is the loglinearized version of Eq. 9.
Finally, block number five contains the exogenous processes: the law of motion of technology and that of foreign demand.
Rights and permissions
About this article
Cite this article
Pisani, M. Financial Openness and Macroeconomic Instability in Emerging Market Economies. Open Econ Rev 22, 501–532 (2011). https://doi.org/10.1007/s11079-009-9140-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11079-009-9140-x