Remittances and Reserves Accumulation in Guatemala 1
Guatemala has experienced a large and unprecedented inflow of remittances over the past ten years, doubling as a percentage of GDP to nearly 19 percent. This structural change in the economic model has contributed to a large improvement in the country’s current account balance (CAB), which has increased from -3.9 percent in 2008 to +1.3 percent in 2022. The CAB improvement has fueled a significant increase in international reserves holdings, amounting to approximately 21 percent of GDP in 2022. This paper studies the role of remittance growth in Guatemala’s economy using a small open economy dynamic stochastic model. We find that remittance growth accounts for higher consumption levels, reserves holdings, and current account balances while exhibiting depressing effects on GDP. Furthermore, this note shows that keeping reserves/imports constant would have had positive welfare effects on Guatemala’s households.
A. External and Monetary Sectors’ Development in Guatemala
1. Guatemala’s flow of remittances doubled from 2008 to 2022, stabilizing at about 19 percent of GDP. Remittances have been stable after the Great Recession until 2015, at 10 percent of GDP. Since 2016, Guatemala experienced a significant and unprecedented surge in remittance inflows, predominantly from the U.S., with peak inflows in 2023 (approximately 19 percent in terms of GDP). In the first months of 2024 remittances growth has continued, though with lower growth rates.
Remittances
(percent of GDP)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: IMF Staff elaborations2. Reserves accumulation doubled as well from 2008 to 2022, having reached 19.6 percent of GDP (Figure 2). International reserves peaked at 22.7 percent of GDP in 2021 to stabilize around 19.6 percent of GDP in 2022 (Figure 2a). In terms of months of imports, the increment in international reserves has been remarkable, from 3.5 months of (current) imports in 2008 to 6.6 in 2022, with a peak of 10.8 months in 2020 (Figure 2b).
Figure 2a. Reserves
(percent of GDP)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: Banguat and IMF Staff elaborationFigure 2b. Reserves
(months of imports)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: Banguat and IMF Staff elaboration3. Guatemala’s CAB displays a significant improvement after 2015. Prior to 2016, Guatemala’s current account position was negative, with an average deficit of approximately 2.7 percent of GDP. From 2016 Guatemala started recording moderate and strong current account surpluses (2 percent of GDP on average), with a peak of 5 percent in 2020. As depicted in Figure 3, the CA improvement has come from the improvement in remittances inflow and lower imports bills (on average, 34.2 percent, and 29.1 percent in terms of GDP before and after 2015).
Contributions to Current Account
(in percent of GDP)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
4. Inflation remains low and the nominal exchange rate has remained relatively stable around 7.7–7.8 quetzales per US$. Following the Great Recession Guatemala has managed to keep the inflation rate under control and within the 4 ±1 percentage point inflation target for almost the entire period, with the recent exemption of 2022-early 2023, which was characterized by high commodity prices (Figure 4a). The nominal exchange rate maintained its value over the period. The quetzal experienced a moderate appreciation from 2009 to 2017, with the exchange rate moving from 8.2 to 7.4 quetzales per US$. From 2018 onwards, the exchange rate has depreciated and converged to the medium-term average of 7.7 quetzales per USD (Figure 4b).
Inflation
(Percent)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: Banguat and IMF 5taff elaborationNominal Exchange Rate
(Quetzales per US Dollar)
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: Banguat and IMF Staff elaoorationB. A Quantitative Model to Study Remittance Growth
5. This paper studies the transition of Guatemala from a low-remittance to a high-remittance country and assesses the suitability of the current reserves accumulation policy. We construct a structural model that captures the change from a low-remittance to a high-remittance regime and its transitional dynamics. We study the dynamics of such transition using the impulse response functions approach, which shows the evolution of GDP, current account, consumption, and inflation, among other outcomes. Furthermore, we construct a counterfactual economy where the monetary authority keeps the reserves/import ratio constant. Finally, we compare the counterfactual economy with the baseline economy to assess whether the counterfactual economy could have been welfare improving.
6. We build a small open economy dynamic stochastic model to reproduce the dynamics of the Guatemalan economy (Box 1). The model is an adaptation of the small open economy New-Keynesian model of Gali and Monacelli (2005) that incorporates remittances as a source of income for the economy. In addition to households, producers, and the rest of the world, we account for an inflation-targeting monetary authority that engages in the market for international reserves to stabilize the relative consumption of domestic goods and imports and smooth consumption over time. Households consume both domestically produced and imported goods, while firms produce domestic goods, which are sold to households and exported to the rest of the world. Remittances constitute a source of foreign currency that can be used to purchase imports. The monetary authority purchases and sells international reserves to smooth the consumption of imports over time and stabilize inflation by affecting the change in relative prices (i.e., the terms of trade).
7. The model simulates the observed increase in remittances from 10 to 20 percent of GDP and replicates the patterns of international reserves (Figure 5a). We simulate the model to replicate the observed growth in remittances observed in the mid-2010s in Guatemala, and we calibrate the model to generate the observed pattern of reserves accumulation. We model the growth of remittances as a regime change in a two-stage Markov process: before 2015, remittances followed a “low” regime accumulation process, while after 2015, the process switched to a “high” accumulation regime. At the onset of the shock, the drift term of the remittance process changes permanently. The autoregressive component of remittances makes the transition slow, reflecting the observed data pattern.
Impulse Response Functions
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: IMF Staff elaboration.8. The model predicts a 1.5 percent drop in output, a 3.5 percent increase in consumption, and a current account deterioration in the long run. The increase in remittances constitutes an additional source of foreign currency for the households that can be spent for purchasing additional foreign goods. Moreover, the increased demand for imports generates, on the one hand, an additional demand for domestic consumption because of complementarity reasons and, on the other hand, a deterioration of the terms of trade, as the relative price of foreign goods is lower. Initially, the increased demand for domestic consumption is satisfied by lower export volumes, caused by a deterioration of the terms of trade, which in turn reduces production needs and causes an output drop. The long-run combined effect of higher imports and lower output deteriorates the trade balance, which is not sufficiently compensated by the higher remittance support, causing a deterioration of the current account as well.
9. The impulse response functions to a regime change shock replicate the observed dynamics of the transition from a low-remittance to a high-remittance economy (Figure 5a and Figure 5). We construct the impulse response function to assess the dynamics of the variables during the transition from a low to a high-remittance economy. Interestingly, the regime change produces an initial positive effect on GDP caused by an increased demand for domestic goods. The increased demand cannot be accommodated immediately through lower exports, forcing producers to increase production in the short term, generating inflationary pressures and depreciating the domestic currency. In the long run, reducing export volumes allows the economy to sustain lower production levels, lowering the GDP. Part of the resources arising from remittances are saved, allowing a strong accumulation of international reserves and improving the current account in the short run. In the long run, once the reserves accumulation process has stopped, the return generated by the high level of reserves allows the economy to sustain a larger current account deficit, which is partially mitigated by less favorable terms of trade.
C. Counterfactual Exercise and Welfare Analysis
10. The counterfactual exercise shows the dynamics of the regime transition under a constant reserves’ accumulation rule (Figure 6a and Figure 6). We extend the analysis of the model by considering the dynamic impact of remittance growth under an alternative reserves accumulation policy. Under such a policy, the monetary authority keeps the reserves/imports ratio constant at 4.3 months of imports, the steady state value of reserves in the “low” remittance regime (i.e., average value pre-2016 with the switch to the “high” remittance regime).
Counterfactual Exercise—Part I
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: IMF Staff elaboration.11. Under the counterfactual policy rule, households consume a higher share of remittances, and the trade balance deteriorates more. In the model, the lower reserves accumulation policy implemented by the monetary authority frees resources for consumption. Households increase the demand for imports, contributing to a deterioration of the trade balance. The complementarity of the two consumption goods leads to a stronger demand for domestic goods, which, paired with stronger import demand, leads to a more intense deterioration of terms of trade. Consequently, exports decline more, amplifying the trade balance deficit.
12. The output contraction is faster, with lower inflationary pressure and a less pronounced current account improvement. Given the faster export deterioration, GDP falls faster than in the baseline economy, alleviating inflationary pressures on domestic producers. Nominal exchange rate depreciation is more negligible, with faster appreciation dynamics after that. Finally, the current account improvement is more limited since the combination of higher imports and lower exports worsens the trade balance. In the long run, the lower levels of reserves accumulated generate less resources for the economy, which does not allow to afford the larger current account deficit sustained in the baseline scenario.
13. Holding reserves constant would have led to a welfare gain of 0.02 percent units of consumption for the households. We compare the two models, the baseline and the counterfactual economy. Given the paths of consumption, labor supplied, and inflation of both models, the welfare gain is defined as the additional units of consumption that households should receive every period, such that the two economies are indifferent to the households. According to the simulations, households should receive 0.02 percent units of consumption as compensation for living in an economy with the baseline reserves accumulation rule rather than living in a world with a constant reserves rule. Figure 5a and Figure 5b provide some intuition for the welfare gain: the lower reserve accumulation allows households to consume more, work fewer hours, and enjoy a lower level of inflation in the short-run, at the onset of the remittance growth period, which provides a higher level of utility. In the long run, the inflation rate and the hours households work are roughly the same for both economies. At the same time, consumption levels are higher in the baseline economy, owing to the initially higher saving rate. However, since households are impatient, the short-run gains more than compensate for the higher long-run benefits of higher consumption, explaining the welfare gain of the counterfactual economy.
14. Relaxing the domestic goods-imports elasticity improves GDP in the long-run (Figure 7). To assess the robustness of our results, we ease the assumption concerning the elasticity of domestically produced good and imports, and we compare the results with the baseline model. We recalibrate the model to match the observe patterns of reserves accumulation of the two models. Relaxing the elasticity of substitution assumption reduces the negative long-run effects on GDP. As in the baseline calibration, higher remittances inflow increases the demand for domestically produced goods, which is partially satisfied by lower exports. However, the increased marginal utility of consumption does not impact labor supply in the long run, thus eliminating the negative effects on production of domestic goods. The dynamic response of the current account balance and the welfare experiment are not affected by this robustness check.
Counterfactual Exercise–Part II
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: IMF Staff elaboration.Counterfactual Exercise with Robustness-Check Calibration
Citation: IMF Staff Country Reports 2024, 267; 10.5089/9798400287459.002.A004
Source: IMF Staff elaboration.Note: baseline (solid); counterfactual exercises (dotted).D. Conclusion
15. Transitioning from a low-remittance to a high-remittance country leads to structural changes in the economic model. Guatemala has seen remittances doubling as a percentage of GDP from 2008 to 2022. This structural change has modified the composition, sign, and magnitude of the current account. The large remittance inflow has improved the sustainability of trade deficits, with higher current account balances fueled by strong reserve accumulation.
16. Higher remittance inflows have positive effects on consumption yet negative effects on production. The large inflow of remittances allows households to sustain higher consumption levels, both imports and domestic goods, thus having a positive welfare effect. However, higher remittances lead to a permanently lower GDP in the long run. This decline is rationalized by a deterioration of the terms of trade, which makes exports less favorable, thus reducing the need to produce domestic goods despite the higher levels of domestic consumption.
17. Keeping the reserves/imports ratio constant would have improved welfare along the transition from a low remittance to a high remittance country. As previously discussed in the counterfactual exercise and the welfare experiment, a different reserves accumulation rule that keeps the reserves/imports ratio constant at 4.3 months of imports would have led to a welfare gain of 0.02 units of consumption. Such gain would have arisen from additional volumes of imports and additional domestic consumption in the initial phase of the transition, which would have more than compensated for the (smaller) additional consumption in the long run.
A Small Open Economy New-Keynesian DSGE Model for Guatemala
Model structure. The model is a small open economy which follows Gali and Monacelli (2005), also followed by Arellano et al (2020), with two types of goods and four agents, i.e. households, domestic producers, monetary authority and rest of the world.
The types of goods are:
Domestic goods: produced by domestic firms and consumed by households and the rest of the world (exports)
Foreign goods: produced by the rest of the world, imported and consumed by households.
The types of agents are:
Households: consume the two types of goods, supply labor to domestic firms in exchange for a wage, receive remittances from the rest of the world, profits from domestic firms and a (net) transfers from the monetary authority. The budget constraint of households, in units of the domestic goods can be represented as:
where CD,p,CF,w,h,r,a,a’,qa,ZD represent, respectively: domestic consumption, terms of trade, imports, real wages, hours worked, remittances, initial reserves, final reserves, price of buying one unit of reserves and profits from domestic producers.
Households aggregate domestic consumption and imports through a CES function, which yields a consumption basket C:
With γ,μ being, respectively, the share of domestic goods in the consumption basket and the elasticity of substitution between domestic goods and imports.
Subject to the budget constraint, households choose the optimal values CD,CF,h that solve the maximization problem:
The instantaneous utility function displays risk aversion in consumption and a quadratic disutility from inflation. This functional form (see Ottonello and Perez (2019)) is a reduced form version a model with cash-in-advance constraint or money in the utility function and it is intended to capture the welfare cost of inflation, as shown in Lucas (2000).
Domestic firms: produce domestic goods in a monopolistically competitive environment with price-adjustment costs (a la Rotemberg) and sell the goods to households and the rest of the world. There is a continuum of firms in the unit interval, each producing a different variety j of the domestic good. The demand schedule faced by each firm is:
where the numerator is the price set by the individual firm, the denominator is the aggregate price of domestically produced goods and ε is the elasticity of substitution of the individual variety of the differentiated good. The production function of the differentiated good is represented as:
Yt(j) = ztht(j)
where z is the aggregate productivity. Subject to the demand schedule and the production function, and the production function each firm solves the maximization problem:
where Λ is the stochastic discount factor from the households’ problem and ϕ is the price-adjustment cost in the Rotemberg setup. Each firm faces the following demand for exports:
where we denote with ρ the elasticity of exports with respect to the terms of trade and
Monetary authority: sets the inflation rate (CPI) of the economy and engages in the market for international reserves, transferring the (net) profits of the operations to households. CPI Inflation is characterized by the following equation:
where π,π*, Δe represent CPI Inflation, foreign inflation and nominal exchange rate change. The monetary authority accumulates reserves following the empirical rule given by:
Rest of the world: buy domestic goods from domestic firms (exports), sell foreign goods to households (imports), sell international reserves to the monetary authority and transfer resources to households in the form of remittances.
The (domestic) goods market clearing condition and the Balance of Payment (BOP) condition of this economy can be represented as:
The stochastic process for the aggregate productivity follows a log-normal AR(1) process, while the stochastic process for remittances follows and AR(1) process with drift.
The model is solved using non-linear solution techniques. The Euler equation method is employed for solving the model, a solution technique that delivers accurate results both close to and far away from the steady state. In particular, the method used is the policy function iteration (Coleman, 1990) and time iteration (Judd, 1998). Expectations are computed using Gaussian quadrature (in particular, the Gauss-Hermite polynomials), using linear interpolation over the exogenous state variables z,r and cubic spline interpolation over the endogenous state variables p, a.
The model is calibrated to Guatemala’s data from 2008 to 2022. The calibration process truncates the data in two subsets, before and after 2015. We assume that before 2015 the remittances process follows an AR(1) with a “low” drift process, while after 2015 the drift regime unexpectedly shifts to “high”. The reserves accumulation rule is calibrated to match the average reserves/imports ratios before and after the regime change, while all the other parameters are calibrated from the literature, the data and internally.
The impulse response functions simulate an unexpected rise in remittances. We simulate the unexpected change in the regime process for remittances and we compute the response of the variables following the shock. Prior to the shock, remittances follow a process with a “low” drift. When the shock occurs, remittances start following a process with a “high” drift for the remainder of the simulation period.
The counterfactual exercise simulates a constant reserves/imports accumulation rule. In our simulation we compare the baseline model with another model where the monetary authority follows a different reserves accumulation rule. More specifically, we assume that the monetary authority keeps the reserves/imports ratio constant to the steady state value of the “low” remittance process. We then repeat the calibration exercise and the impulse response functions with this alternative policy rule and compare the results of the two models in terms of economic performance.
The welfare analysis assesses the performance of the baseline model in terms of the constant reserves/imports model.
We compare the baseline and the alternative model to assess the welfare gains arising from having a different reserves accumulation rule. The value functions of the baseline and the alternative models are represented, respectively, as:
The welfare gain from the alternative model is characterized by λ, which denotes the extra units of consumption that households should receive in every period under the baseline scenario such that the value function of the baseline economy matches the value function of the alternative economy. Following Samano (2022) we use the same realizations of the exogenous state variables, productivity and remittances, to construct the time series of consumption, hours worked and inflation for both models. Both patterns simulate the immediate switch from “low” to “high” remittances.
Source: IMF Staff based on SPR tool adapted for GuatemalaReferences
Arellano, C., Bai, Y. & Mihalache, G. (2020), Monetary Policy and Sovereign Risk in Emerging Economies (Nk-Default). NBER Working Paper No. w26671
Coleman, J. (1990), “Solving the stochastic growth model by policy-function iteration.” Journal of Business & Economic Statistics, 8, 27–29.
Galí, J. & Monacelli, T. (2005) Monetary Policy and Exchange Rate Volatility in a Small Open Economy. Review of Economic Studies, Oxford University Press, Vol. 72(3), pages 707–734.
Judd, K. (1992), “Projection methods for solving aggregate growth models.” Journal of Economic Theory, 58, 410–452.
Lucas, R.E. Jr. (2000). Inflation and Welfare. Econometrica 68 (2): 247–74.
Ottonello, P. & Perez, D.J. (2019). The Currency Composition of Sovereign Debt. American Economic Journal: Macroeconomics, 11 (3): 174–208.
Samano, A. (2022). International reserves and central bank independence. Journal of International Economics, Elsevier, Vol. 139©.
Prepared by Andrea Paloschi.
