Policy Research Working Paper 11132 Forecast Sensitivity to Global Risks A BVAR Analysis Heather Ruberl Remzi Baris Tercioglu Adam Elderfield Economic Policy Global Department May 2025 Policy Research Working Paper 11132 Abstract Developing countries face uncertainties driven by global into the World Bank’s macro-structural model to assess macroeconomic variables over which they have little to no how a range of potential global disturbances could impact control. Key exogenous factors faced by most developing economic outcomes across countries. The methodology countries include interest rates in high-income countries, is applied to 115 countries, using the World Bank’s fall commodity prices, global demand for exports, and remit- 2024 edition of the Macro-Poverty Outlook forecasts as a tance inflows. While these variables are sensitive to common baseline. Although the individual country results are het- global shocks, they also exhibit idiosyncratic fluctuations. erogeneous, the aggregate distribution of gross domestic This paper employs a Bayesian Vector Autoregression model product outcomes across the 115 countries suggests that to capture interdependencies of global variables and sim- global factors influence gross domestic product levels in ulates global risks using the empirical joint distribution of individual developing countries by less than plus or minus global shock as captured by joint Bayesian Vector Autore- 2 percent in most years, but by between 2 and 4 percent in gression errors. The simulated shocks are then integrated about 3 in 10 years. This paper is a product of the Economic Policy Global Department. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at hruberl@worldbank.org or rtercioglu@worldbank.org@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Forecast Sensitivity to Global Risks: A BVAR Analysis∗ Heather Ruberl1, Remzi Baris Tercioglu2, Adam Elderfield 1,2 Economic Policy Global Department, World Bank JEL Classification: C10, C50, E17 Keywords: Economic modeling, forecasting, global risks, macroeconomic shocks, macroeconomic modeling and statistics ∗We thank Andrew Burns, Fernando Giuliano and Charl Jooste who provided valuable feedback and advice on this work. 1 Introduction Macroeconomic forecasting is an exercise undertaken in the face of considerable uncertainty. Indeed, forecasting can be seen as an exercise in reducing uncertainty – in part by exploiting empirical regular- ities in the way economies operate and in part by making artificial assumptions about what the state of the world may be. For developing countries, a key source of uncertainty lies in global macroeconomic conditions. Forecasts typically rest on assumptions about variables such as oil prices, trade policies of major economies, and demand from key trading partners. Each of these reduces the range of possible outcomes, and helps forecasters develop a conditional projection of what economic outcomes might look like if these global assumptions are realized. Typically these projections are presented as point forecasts, what the world will look like assuming these global conditions hold. Of course, in the real world the assumptions are never realized exactly (and sometimes are very far from expectations). This paper proposes a methodology to move beyond these point estimates conditional on specific global assumptions by generating a range of outcomes around a given point estimate by running hundreds of simulations each based on the underlying uncertainty of different external-to-the country global economic variables. These include the world prices of oil and other commodities, monetary policy conditions in high-income countries, the level of import demand in the countries to which a country exports, and the level of remittance inflows to the country (in many developing countries these remittances can be as high as 25% of GDP). A Bayesian Vector Autoregression (BVAR) across external variables allows controlling for the endogeneity between external variables (typically when oil prices are high so are those of other commodities, while oil prices, monetary policy, and imports are sensitive to the economic cycle in high-income countries). By randomly drawing from the distribution of historical errors of the BVAR, an infinite range of potential outcomes can be generated. These can then be fed into a macroeconomic model to generate a distribution of outcomes reflecting the underlying uncertainty surrounding the assumptions that generated a given forecast. Such an exercise is performed for 115 countries using as inputs the point estimates of the World Bank’s forecasts produced for the fall 2024 edition of the Macro-Poverty Outlook (MPO). Drawing from the joint distribution of errors for the above set of exogenous (to the country) global variables, 1,000 simulations are run for each country using the World Bank’s MFMod macro-structural forecasting model. From these simulations, a range of GDP, current account, inflation and fiscal outcomes are derived and used to provide a probabilistic presentation of future outcomes. The paper proceeds as follows: Section 2 outlines the process for constructing a BVAR model of exogenous macroeconomic shocks and incorporating these into MFMod; Section 3 demonstrates how modeling global macroeconomic shocks can provide insight into the economic outlook for a country; and Section 4 points to some of the limitations of this approach and suggested areas for further work. 2 Constructing a BVAR model of exogenous global shocks 2.1 Global cycles and constructing the BVAR Kose et al. (2003) provide evidence of global business cycles, highlighting the role of a common global factor in driving output fluctuations across most countries. Rey (2015) emphasizes the constraints imposed by global financial cycles on the monetary policy independence of developing economies. Miranda-Agrippino and Rey (2020) demonstrate that U.S. monetary policy shocks generate external spillovers through global financial cycles, influencing domestic credit markets and economic activity in developing countries. Similarly, Kalemli-O¨ zcan (2019) argues that the risk channel associated with U.S. 2 monetary policy shocks disproportionately affects emerging market economies compared to advanced economies. The role of commodity prices as both a source and transmission channel of global shocks is analyzed by Juvenal and Petrella (2024). Global macroeconomic variables play a crucial role in forecasting the economic performance of developing countries. Among the most significant global variables are interest rates in high-income countries, commodity prices, global demand for exports, and remittance inflows. These variables are determined on a global scale, leaving policy makers in developing economies with little to no influence over them. During periods of global expansion, growth in high-income countries and commodity prices typically rise.1 If accompanied by low interest rates, capital inflows to developing economies increase, further fueling economic expansion. Conversely, during contraction phases, global demand and commodity prices tend to decline, leading to declining export revenues. This contractionary effect is particularly pronounced in commodity-exporting developing economies. U.S. monetary tightening tends to weaken capital flows to developing countries, tightens monetary conditions in countries with fixed exchange rates and can influence interest rates for developing country borrowers. Figure 1: HP filtered cyclical components of annual US GDP, EU GDP, China GDP (on the left axis), US 10 year rates and oil prices (on the right axis). Figure 1 presents the cyclical components of GDP for three major economies, alongside US interest rates and oil prices. While expansion and contraction phases are clearly identifiable, each series exhibits distinct dynamics. This is particularly evident in commodity prices, which are influenced by both global cycles and idiosyncratic shocks. For instance, geopolitical conflicts may drive oil prices higher, whereas favorable weather conditions in key agricultural regions can lower global commodity prices, regardless of the phase of global cycles.2 The literature employs several approaches to integrating global risks into macroeconomic forecasts. Carabenciov et al. (2008) develop a macro-structural model that incorporates oil prices and simulate it using historical variations in oil prices. In their framework, historical oil shocks are estimated with the cyclical component modeled as an AR(1) process. Dynamic Stochastic General Equilibrium (DSGE) models frequently include simplified representations of the rest of the world, where global variables are 1 Commodity prices also show idiosyncratic fluctuations which may or may not be correlated with global cycles. 2 The impact of geopolitical events depends on whether a country is a price maker. For example, geopolitical tensions in a small oil-exporting economy may not significantly influence global oil prices. 3 modeled as AR(1) processes (Gali and Monacelli, 2005; Lubik and Schorfheide, 2007). However, some studies estimate the external sector using a VAR framework instead, allowing for global interactions (Christoffel et al., 2008; Fasolo et al., 2024). Notably, Giuliano et al. (2024) employ a VAR model of global economic variables to integrate global risks into the World Bank’s Uruguay macrostructural model.3 This paper builds on the methodology of Giuliano et al. (2024), extending the analysis to 115 coun- tries in the World Bank’s Macro Fiscal Modeling system (MFMod). To effectively capture interde- pendencies between global variables, we employ an annual BVAR model for each of the 115 developing countries spanning from 2005 to 2023. These models include US long-term interest rates, Eurozone long-term interest rates, United States GDP, Eurozone GDP, China GDP, India GDP (for countries in the South Asia region), global oil prices, country-specific export market demand, a weighted average of commodity prices, and remittances inflows. We include GDP of three major economies separately in the BVAR, rather than using a weighted average series, as business cycles do not fully align, as shown in Figure 1 and discussed in the literature (Kose et al., 2012). Similarly, oil prices are included separately from the weighted average of import and export prices to prevent variations in commodity prices driven by oil price changes from being absorbed into the error terms. Our analysis is confined to the most recent two decades for two primary reasons: 1) the rapid evolution of the global economy post-2000s implies substantial structural change and means that the transmission channels in the most recent period differ significantly from those that were at play in earlier periods and 2) many countries lack data for the earlier period to support the methodology. Despite the shorter period used to estimate the BVAR, it is sufficiently long to have comprised two major economic crises, notably the global financial crisis of 2008 and the recent COVID-19 pandemic.4 VAR models are often used in econometrics to capture the interdependencies among multiple time series. Unlike univariate autoregressive models, which predict the future values of a single variable based on its own past values, VAR models consider the joint dynamics of a set of variables, treating each variable as potentially influenced by its own past values and the past values of all other variables in the system.5 This makes VAR models particularly powerful for analyzing economic systems where feedback effects and simultaneous interactions between variables are significant. More important for our purpose in this paper, VAR models tend to perform well in forecasting (Sims, 1980; Stock and Watson, 2001; Del Negro and Schorfheide, 2004). This is mainly because VAR models are data-driven and they do not impose strong theoretical restrictions on data. We use a BVAR instead of a classical VAR primarily due to data limitations, which lead to degrees of freedom constraints in a classical VAR. By imposing prior restrictions on parameters, BVARs help mitigate the risk of overfitting. The literature highlights the advantages of BVARs in data-limited environments and their superior performance over classical VARs, particularly in large systems (Litterman, 1986; De Mol et al., 2008; Ban ´ bura et al., 2010). The construction of the BVAR model begins by extracting the cyclical component of the global variables. This is done by estimating a trend for each variable using the Hodrick-Prescott (HP) filter (with a smoothing parameter of λ = 100).6 The cyclical component is obtained by subtracting the 3 An alternative approach involves using forecast errors from global forecasters. For instance, to capture risks for the U.S., Adams et al. (2021) utilize forecast errors from the Survey of Professional Forecasters to construct predictive distributions for domestic variables. 4 One can include more variables in the mix such as a financial risk indicator VIX. However, with the short time span, adding more variables will make identifying additional parameters difficult. 5 Using a VAR instead of multiple AR(1) models is crucial for our purposes here because it allows us to exploit the interdependencies among the error terms — something that would not be possible with isolated AR(1) processes. 6 To reduce the end-point bias, we smooth over 2005-2026, including the forecasts of the coming three years in addition to the historical data. 4 estimated trend from the actual value. We use an unstructured BVAR of order 1 where each variable is modeled as a linear function of lags of itself and lags of the other variables in the system. By doing this, the BVAR captures the linear interdependencies between variables and system-wide feedback effects. We do not impose any structural assumptions, as our focus is on forecast uncertainty which is related to the error variance- covariance matrix of the BVAR, rather than on impulse-response functions or the causal relationships between variables. The system of equations for the BVAR(1) model can be compactly expressed in matrix form as xt = Φxt−1 + ϵt, where xt is the vector of variables at time t, Φ is the matrix of autoregressive coefficients, and ϵt is the vector of shocks (or random errors). The estimated coefficients capture the relationships among the variables, while the residual variance- covariance matrix, Σϵ, quantifies the uncertainty (random shocks) associated with each variable. The interaction of shocks with this matrix determines the dynamic adjustment of the system. The error terms ϵt represent the shocks to the economic variables. These shocks are assumed to be random and can be interpreted as unexpected changes in each variable due to factors not captured by its own past values and contemporaneous and past changes of the other variables. The correlation between the errors provides insights into how unexpected shocks to one variable are related to shocks in other variables. For instance, if ϵt,1 (shock to GDP) is positively correlated with ϵt,2 (shock to inflation), it suggests that unexpected increases in GDP are often associated with unexpected increases in inflation. Formally, this can be analyzed using the covariance matrix of residuals, Σϵ, which captures the covariances and correlations:   Var(ϵt,1) Cov(ϵt,1, ϵt,2)Cov(ϵt,1, ϵt,3)   Σϵ = Cov(ϵt,2, ϵt,1) Var(ϵt,2) Cov(ϵt,2, ϵt,3)  . Cov(ϵt,3, ϵt,1) Cov(ϵt,3, ϵt,2) Var(ϵt,3) By standardizing these covariances, we obtain the correlation matrix (shown in Section 3 for sample countries), which provides a measure of the inter-temporal relationships between the shocks to the variables. We estimate the BVAR(1) model using the prior specification of Sims and Zha (1998). The prior mean of all coefficients is set to zero. This implies that our prior pushes all cyclical variation in global variables to shocks. We set the overall tightness hyperparameter, λ0, which controls the degree of shrinkage on all coefficients, to 0.5. The relative cross-variable tightness hyperparameter, λ1, which governs the shrinkage on coefficients of other variables’ lags, is set to 0.1. This prior specification imposes relatively loose shrinkage, allowing the data to speak while yielding impulse response functions consistent with empirical regularities. To assess the robustness of our results, we conduct a sensitivity analysis by setting λ0 = λ1 = 0.01, increasing prior shrinkage towards zero, and by setting λ0 = λ1 = 1, relaxing prior shrinkage. The main results remain robust across these alternative specifications, with no material changes in the distribution of forecast errors. The results of this sensitivity analysis are reported in Appendix A. After estimating the BVAR, we conduct conditional forecasting to reproduce the mean global forecasts of the 2024 Annual Meeting (AM24) version of the MPO, with uncertainty defined by the BVAR model’s joint error distribution. We interpret this uncertainty as global shocks – or global risks. For instance, not all oil price increases qualify as exogenous shocks. When global demand is high, import demand rises, driving up oil prices. Similarly, because oil is a key input in the production of many commodities, and these commodities are also influenced by global demand conditions, not all commodity price increases are exogenous shocks. The BVAR model captures these demand-driven 5 price movements in response to global business cycles. Only the portion unexplained by other global variables and past prices qualifies as a shock. Our BVAR model, much like policy makers and World Bank forecasters, can anticipate the impacts of global economic cycles but cannot predict unforeseen events or ‘shocks’ such as the U.S. mortgage crisis, the COVID-19 pandemic, or a major drought. To simulate future uncertainty around global variables, random disturbances are drawn from a multivariate normal distribution with mean zero and a variance-covariance matrix estimated from the residuals of the BVAR model. For each time step t, future values are generated iteratively by applying the estimated coefficients from the BVAR to the past values of the variables and adding the randomly drawn shocks. This process is repeated 1,000 times, resulting in a set of stochastic pathways for each exogenous variable. 2.2 Exogenous macroeconomic variables Global variables included in the BVAR model exogenously enter into the macro-structural MFMod country models. Some of these series are the same for all countries (for example, the US interest rate) and some of them have been specifically constructed or sourced specifically for the country being modeled (for example, export demand). Most country-specific MFMod models already include these variables as exogenous inputs into the model, and contain historical time-series data for each variable. It is therefore possible to deploy this methodology with limited changes to pre-existing models, and so it can be incorporated fairly easily into most analytic work and forecasting . Table 1 outlines the global variables included in the BVAR model and their “hook” into MFMod. For a full technical description of the standard MFMod framework, see Burns et al. (2019). 2.2.1 Key Transmission Channels from Global Shocks to the Domestic Economy Demand-Side Global shocks affect the domestic economy through several demand-side channels. First, all eco- nomic activity is price-elastic, and variations in import and oil prices directly influence domestic price deflators, particularly in sectors where imports constitute a significant share of expenditure (e.g., busi- ness investment). In addition, household consumption is sensitive to changes in import prices, while remittances also affect household incomes. Export volumes are strongly influenced by the economic performance of major trading partners, and economic downturns or geopolitical events can substan- tially reduce export demand. Global commodity prices and shipping costs, which are linked to oil prices, further influence export demand. Monetary Policy Central banks closely monitor import and export prices due to their effect on domestic inflation. They also track broader economic indicators such as overall economic growth and the output gap to assess the economy’s performance relative to its potential. Shocks to global prices and global economic activity will therefore influence monetary policy decisions through the impact on domestic inflation and domestic economic growth. Exchange Rates In MFMod, economies operate under various exchange rate regimes: pegged (where the domestic currency tracks another currency), floating (determined by market forces), and mixed. In countries with floating exchange rates, the equilibrium exchange rate follows the uncovered interest parity con- dition. Under this framework, capital inflows occur when the domestic interest rate exceeds the foreign risk-free rate (typically the US or EU rate), leading to currency appreciation as monetary policy tight- 6 Exogenous Vari- Data Source Linkages in MFMod able Import Price Country-specific-trade-weighted Key exogenous variable in MFMod average of 32 global commodity - Influences overall inflation through input prices costs, consumer prices, government prices, and investment prices. - Affects import volumes and prices, govern- ment revenues from import taxes, current ac- count position and exchange rate pressures and through these domestic monetary policy conditions. Export Price Country specific trade weighted Key exogenous variable in MFMod average of 32 global commodity - Affects export volumes and prices, govern- prices ment revenues from export taxes, current ac- count position and exchange rate pressures and through these domestic monetary policy conditions. Oil Price West Texas Intermediate (WTI) For non-oil producing countries: Affects oil price country-specific Import and Export prices. For oil producing countries: Key exogenous variable in the model which impacts on oil ex- port prices, current account and government revenues associated with oil. Export Market Country specific trade weighted Key exogenous variable in MFMod Demand average of major trading partner - Affects export volumes, government revenues import volumes from export taxes, current account position and exchange rate pressures and through these domestic monetary policy conditions.. US Interest Rate US Federal Reserve (10-year Key exogenous variable in MFMod Treasury bond rate) - Affects global borrowing costs and govern- ment interest payments on external debt. - Short-term rate enters into uncovered inter- est parity and impacts on exchange rates. EU Interest Rate European Central Bank policy For countries with a currency pegged to the rate Euro this is the global interest rate variable in the model. Remittances in- Country specific data from bal- Key exogenous variable in MFMod flows ance of payments dataset (de- - Impacts household disposable income and nominated in USD) private consumption. - Affects current account balance. Table 1: Summary of BVAR variables data sources and how they enter MFMod models ens in response to global shocks. In contrast, economies with fixed exchange rate regimes maintain a direct link to the base currency, making their exchange rates less responsive to monetary policy changes. However, global shocks still affect reserves and overall external sustainability. Government fiscal balance Higher prices or lower GDP growth stemming from global shocks can also affect the government’s fiscal balance by altering the cost of government expenditures and impacting the government’s tax base. The overall effect on fiscal balances depends on whether the response in prices or demand is stronger, leading to either increased costs or shifts in revenue streams. Domestic and international interest rates also significantly influence the cost of financing government debt. Elevated rates, whether domestic or global, tend to increase borrowing costs and the overall debt servicing burden, thereby 7 exerting pressure on fiscal balances. Current Account Balance Export/import prices, volumes and exchange rates determine trade balances of countries, all of which are subject to global shocks as explained above. Remittances, which are determined by incomes in high-income economies, enter net factor services and transfers. 3 Applications For each of the 115 countries modeled, a three-step simulation process is followed. First, the BVAR is shocked 1,000 times by drawing from the distribution of joint errors in the BVAR system. This generates 1,000 outcomes for the global variables included in the BVAR, with the mean of these outcomes tracking the point forecasts for the country. In step 2, the shocked values of the global variables are introduced into each macroeconomic model, generating country specific responses to the changed global conditions. These 1,000 simulation results are then collated into a distribution of results for all of the endogenous variables of each model and these distributions are presented for a selection of main economic indicators. We compute the median outcome for each year. To capture the dispersion of outcomes, we determine the 20, 50 and 80 percent intervals of distributions. This approach allows us to quantify the range of outcomes with specified likelihood, providing insights into the potential variability and uncertainty surrounding the impact of shocks on key exogenous variables. Figure 2 shows the aggregate distribution of GDP responses for 115 countries to combined global shocks. Fifty percent of the simulations fall within ± 2 percent of the baseline real GDP level, while 80 percent lie within ± 4 percent. The median GDP growth across countries is approximately 3 percent, with the 10th percentile around -4 percent and the 90th percentile around 10 percent. The distribution of growth rates is wider than that of GDP levels. GDP levels reflect the cumulative impact of shocks, with positive and negative shocks partially offsetting each other over time, thereby constraining deviations within a narrower range. 8 Figure 2: Aggregate forecast intervals from combined exogenous macroeconomic shocks to 115 countries. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an addi- tional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. We show detailed results for 4 sample economies: An emerging market economy with high trade openness (Poland), an oil producing economy (Angola), an economy with a significant share of remit- tance inflows (The Gambia), and a relatively diversified economy with limited exposure to external shocks (the Philippines). Appendix B presents results for the remaining 111 economies. 3.1 Poland Figure 3 presents distribution of global shocks to Polish economy as estimated from the BVAR model. Interest rate shock ranges within ± 150-basis points. Oil price can deviate up to ± 50% from the baseline. For export and import prices, BVAR model estimates ranges around ± 10% - 15% and export market fluctuates around ± 7.5% of the baseline. Remittances inflows range within ± 20% of its baseline.7 7 While external shocks are the same across countries, their economic impact depends on each country’s economic structure. For instance, in an oil-importing economy, a positive oil price shock raises import costs, worsens the current account balance, and fuels inflation. In contrast, for an oil-exporting economy, higher oil prices boost export prices, increase export tax revenues, and improve the current account balance. Similarly, the transmission of U.S. interest rate shocks depends on a country’s exchange rate regime. Under a fixed exchange rate, U.S. monetary policy directly influences domestic interest rates which enter consumption, investment decisions, and debt servicing costs. In a floating exchange rate regime, by contrast, the exchange rate would adjust—depreciating or appreciating—to absorb part of the shock, with implications for export and import prices. 9 Figure 3: Fan charts of 1000 shocks to exogenous macroeconomic variables for Poland. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. The correlation matrix of shocks is given in Figure 4. As expected, export and import prices are highly correlated with each other and each is strongly correlated with oil prices. US interest rates are correlated with export market growth as they signal expansions/contractions in advanced economies. Figure 4: Poland Correlation matrix. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, REM is remittances and PXW is world average export price. Figure 5 shows how these shocks affect the level of Real GDP (in local currency, millions), GDP growth rate, Inflation rate, Fiscal balance (% of GDP), Public Debt (% of GDP) and Current Account Balance (% of GDP) of Poland. The GDP level distribution shows annualized cumulative growth rate ranging from -1% to 7.4% while the growth rate dispersion measure has no memory and for any given year the growth rate could be either much more or less positive, ranging from -3.6% to 10.3%. 10 Figure 5: Poland forecast intervals from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 6 decomposes the impact of all global risks occurring simultaneously into individual shock components. For Poland, shocks to export market demand are the biggest global source of risk for real GDP, followed by shocks to import prices. Another noteworthy result is that the joint macro shock is very close to summation of individual shocks which is not necessarily the case as the Angola application below shows. Figure 6: Individual and joint shock impacts for Poland. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 3.2 Angola The range of global shocks hitting Angola is shown in Figure 7 and the correlation matrix is given in Figure 8. US interest rate, oil prices and export market uncertainties are comparable to the BVAR results for Poland. This is an expected result as we include key global variables in all BVAR specifica- tions. On the other hand, export prices excluding oil and remittance inflows show a wider uncertainty. 11 Figure 7: Fan charts of 1000 shocks to exogenous macroeconomic variables for Angola. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 8: Angola correlation matrix. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PXEX is export price excluding oil, PM is import price, and REM is remittances. Figures 9 and 10 show how these shocks are transmitted to the macro-economy of Angola. There is a bigger risk than in Poland around the median forecast for fiscal balance, public debt and current account balance. This is due to the uncertainty of oil export revenues linked to oil price shocks. Export market growth and export prices excluding oil do not have any significant impact because non-oil exports are less than 10% of total exports. Another interesting result is that distributions are not symmetric. In fact, they are assigning more probability to upward risks than downward risks. Finally, aggregate macro impacts are less than the summation of individual impacts because of the high correlation between oil prices and other commodity prices. 12 Figure 9: Angola forecast confidence intervals from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 10: Individual and joint shock impacts for Angola. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 3.3 Philippines The range of global shocks for the Philippines is shown in Figure 11 and the correlation matrix is given in Figure 12. Global risks are comparable to Poland (lower in remittances). 13 Figure 11: Fan charts of 1000 shocks to exogenous macroeconomic variables for the Philippines. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 12: The Philippines correlation matrix. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. Figures 13 and 14 show the response of the Philippines economy to global shocks. With a relatively stable growth history (aside from the COVID-19 shock), the Philippines is less exposed to global risks, and this is reflected in the size of the uncertainty bands around median forecasts. Combined shocks are expected to move GDP ± 3% with respect to its baseline where the biggest risk is coming from export market growth which is followed by export and import prices. 14 Figure 13: Philippines forecast confidence intervals from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 14: Individual and joint shock impacts for the Phillipines. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 3.4 The Gambia Figure 15 shows the distribution of global shocks to The Gambia and Figure 16 gives the correlation matrix. Shock distributions are similar to the other three cases with slightly higher uncertainty around export prices and lower uncertainty around import prices. 15 Figure 15: Fan charts of 1000 shocks to exogenous macroeconomic variables for The Gambia. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 16: The Gambia correlation matrix. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. Figures 17 and 18 present the macroeconomic results for The Gambia. The most significant finding is the crucial role of remittances in shaping domestic economic outcomes, as they account for the majority of global risks impacting the Gambian economy. 16 Figure 17: The Gambia forecast confidence intervals from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure 18: Individual and joint shock impacts for The Gambia. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 17 4 Limitations of this analysis While the methodology for simulating stochastic shocks to exogenous macroeconomic variables using a VAR model offers valuable insights into potential future outcomes, it is not without its limitations. Understanding these constraints is crucial for interpreting the results and acknowledging the boundaries of the analysis. Firstly, the VAR model’s predictive power is inherently limited by its reliance on historical data. The VAR framework generates future paths based on historical shocks and their estimated relation- ships, which constrains its ability to forecast rare or unprecedented events, such as long-tail risks or “black swan” occurrences. These extreme or outlier events are infrequently observed in historical data and thus may not be adequately captured within the model’s stochastic simulations. Consequently, the risks associated with rare but significant economic disturbances are not fully captured by this analysis. Secondly, the VAR model’s utility is confined to the exogenous shocks that are explicitly included in the analysis. The model’s scope is limited to the variables selected for incorporation and their historical dynamics. It does not account for potential shocks outside the specified set, which could also influence the macroeconomic environment. Finally, the impacts of these shocks are contingent upon the structure and constraints of the MFMod used in conjunction with the VAR model. The functional form of equations within the MFMod, along with any imposed coefficient restrictions and identity equation specifications, shapes the response of the model to the shocks. As a result, the analysis is bounded by the assumptions and parameterization of the MFMod, which may not fully capture all the complex interactions and non-linearities in the real economy. 5 Conclusion and next steps This paper has documented a simple framework for modeling shocks to global macroeconomic variables, and outlined the method for iteratively incorporating these shocks into a macro structural model to model the range of possible impacts on key economic indicators such as GDP growth, inflation, fiscal balance, public debt, and current account balance. The results from our example countries highlight the range of potential outcomes for these variables, illustrating the inherent uncertainties facing policy makers. By understanding these potential outcomes, policy makers can better prepare for economic volatility, ensure fiscal sustainability, and enhance the resilience of their economic strategies. 18 References Adams, P. A., Adrian, T., Boyarchenko, N. and Giannone, D. (2021). Forecasting macroeconomic risks, International Journal of Forecasting 37(3): 1173–1191. ´ bura, M., Giannone, D. and Reichlin, L. (2010). Large bayesian vector auto regressions, Journal Ban of applied Econometrics 25(1): 71–92. Burns, A., Campagne, B. P. M., Jooste, C., Stephan, D. A. and Bui, T. T. (2019). The world bank macro- fiscal model technical description, Policy Research Working Paper WPS 8965, World Bank Group, Washington, D.C. URL: http://documents.worldbank.org/curated/en/294311565103938951/The-World-Bank-Macro- Fiscal-Model-Technical-Description Carabenciov, I., Ermolaev, I., Freedman, C., Juillard, M., Kamenik, O., Korshunov, D., Laxton, D. and Laxton, J. (2008). A small quarterly multi-country projection model with financial-real linkages and oil prices. Christoffel, K. P., Coenen, G. and Warne, A. (2008). The new area-wide model of the euro area: a micro-founded open-economy model for forecasting and policy analysis. De Mol, C., Giannone, D. and Reichlin, L. (2008). Forecasting using a large number of predic- tors: Is bayesian shrinkage a valid alternative to principal components?, Journal of Econometrics 146(2): 318–328. Del Negro, M. and Schorfheide, F. (2004). Priors from general equilibrium models for vars, Interna- tional Economic Review 45(2): 643–673. Fasolo, A. M., Arau ´ jo, E., Jorge, M. V., Kornelius, A. and Marinho, L. S. G. (2024). Brazilian macroe- conomic dynamics redux: Shocks, frictions, and unemployment in samba model, Latin American Journal of Central Banking 5(2): 100110. Gali, J. and Monacelli, T. (2005). Monetary policy and exchange rate volatility in a small open economy, The Review of Economic Studies 72(3): 707–734. Giuliano, F., Navia, D. and Ruberl, H. (2024). The macroeconomic impact of climate shocks in uruguay, World Bank. Juvenal, L. and Petrella, I. (2024). Reprint of “unveiling the dance of commodity prices and the global financial cycle”, Journal of International Economics 149: 103941. ¨ zcan, 2019). Us monetary policy and international risk spillovers, Technical report, National Kalemli-O Bureau of Economic Research. Kose, M. A., Otrok, C. and Prasad, E. (2012). Global business cycles: convergence or decoupling?, International economic review 53(2): 511–538. Kose, M. A., Otrok, C. and Whiteman, C. H. (2003). International business cycles: World, region, and country-specific factors, american economic review 93(4): 1216–1239. Litterman, R. B. (1986). Forecasting with bayesian vector autoregressions—five years of experience, Journal of Business & Economic Statistics 4(1): 25–38. 19 Lubik, T. A. and Schorfheide, F. (2007). Do central banks respond to exchange rate movements? a structural investigation, Journal of Monetary Economics 54(4): 1069–1087. Miranda-Agrippino, S. and Rey, H. (2020). Us monetary policy and the global financial cycle, The Review of Economic Studies 87(6): 2754–2776. Rey, H. (2015). Dilemma not trilemma: the global financial cycle and monetary policy independence, Technical report, National Bureau of Economic Research. Sims, C. A. (1980). Macroeconomics and reality, Econometrica 48(1): 1–48. Sims, C. A. and Zha, T. (1998). Bayesian methods for dynamic multivariate models, International Economic Review pp. 949–968. Stock, J. H. and Watson, M. W. (2001). Vector autoregressions, Journal of Economic perspectives 15(4): 101–115. 20 A Appendix A: Prior Sensitivity We conduct a sensitivity analysis by relaxing and tightening the prior restrictions on the variance- covariance matrix. We present the results for Poland under these alternative prior settings. Figure A.1 reports the shocks obtained under looser restrictions, with λ0 = λ1 = 1, while Figure A.2 shows the shocks obtained under tighter restrictions, with λ0 = λ1 = 0.01. The distribution of shocks does not change much across these specifications. Under looser priors, we observe a slight reduction in the bandwidth of shocks, as less restrictive coefficients absorb some of the variation that would otherwise load onto the residuals. Figure A.1: Fan charts of 1000 shocks to exogenous macroeconomic variables for Poland under loose priors. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. 21 Figure A.2: Fan charts of 1000 shocks to exogenous macroeconomic variables for Poland under tight priors. Note: The dark blue segment reflects the results from 20 percent of simulations, the lighter blue distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. 22 B Appendix B: Results for 115 countries in AM 24 MPO Figure A.3: ALB results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.4: ALB Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 23 Figure A.5: ARE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.6: ARE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 24 Figure A.7: ARG results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.8: ARG Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 25 Figure A.9: ARM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.10: ARM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 26 Figure A.11: AZE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.12: AZE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 27 Figure A.13: BDI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.14: BDI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 28 Figure A.15: BEN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.16: BEN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 29 Figure A.17: BFA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.18: BFA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 30 Figure A.19: BGD results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.20: BGD Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 31 Figure A.21: BGR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.22: BGR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 32 Figure A.23: BHR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.24: BHR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 33 Figure A.25: BIH results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.26: BIH Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 34 Figure A.27: BLR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.28: BLR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 35 Figure A.29: BLZ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.30: BLZ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 36 Figure A.31: BOL results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.32: BOL Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 37 Figure A.33: BRA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.34: BRA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 38 Figure A.35: CHL results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.36: CHL Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 39 Figure A.37: CHN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.38: CHN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 40 Figure A.39: CIV results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.40: CIV Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 41 Figure A.41: CMR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.42: CMR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 42 Figure A.43: COD results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.44: COD Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 43 Figure A.45: COG results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.46: COG Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 44 Figure A.47: COL results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.48: COL Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 45 Figure A.49: COM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.50: COM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 46 Figure A.51: CPV results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.52: CPV Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 47 Figure A.53: CRI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.54: CRI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 48 Figure A.55: DOM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.56: DOM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 49 Figure A.57: DZA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.58: DZA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 50 Figure A.59: ECU results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.60: ECU Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 51 Figure A.61: EGY results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.62: EGY Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 52 Figure A.63: ERI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.64: ERI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 53 Figure A.65: ETH results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.66: ETH Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 54 Figure A.67: GAB results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.68: GAB Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 55 Figure A.69: GEO results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.70: GEO Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 56 Figure A.71: GHA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.72: GHA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 57 Figure A.73: GIN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.74: GIN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 58 Figure A.75: GNB results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.76: GNB Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 59 Figure A.77: GNQ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.78: GNQ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 60 Figure A.79: GTM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.80: GTM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 61 Figure A.81: HND results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.82: HND Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 62 Figure A.83: HRV results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.84: HRV Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 63 Figure A.85: HTI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.86: HTI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 64 Figure A.87: IDN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.88: IDN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 65 Figure A.89: IND results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.90: IND Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 66 Figure A.91: IRN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.92: IRN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 67 Figure A.93: IRQ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.94: IRQ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 68 Figure A.95: JAM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.96: JAM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 69 Figure A.97: JOR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.98: JOR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 70 Figure A.99: KAZ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.100: KAZ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 71 Figure A.101: KEN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.102: KEN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 72 Figure A.103: KGZ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.104: KGZ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 73 Figure A.105: KHM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.106: KHM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 74 Figure A.107: KWT results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.108: KWT Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 75 Figure A.109: LBN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.110: LBN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 76 Figure A.111: LBR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.112: LBR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 77 Figure A.113: LBY results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.114: LBY Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 78 Figure A.115: LKA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.116: LKA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 79 Figure A.117: MAR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.118: MAR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 80 Figure A.119: MDA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.120: MDA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 81 Figure A.121: MDG results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.122: MDG Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 82 Figure A.123: MDV results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.124: MDV Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 83 Figure A.125: MEX results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.126: MEX Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 84 Figure A.127: MLI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.128: MLI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 85 Figure A.129: MNE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.130: MNE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 86 Figure A.131: MNG results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.132: MNG Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 87 Figure A.133: MOZ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.134: MOZ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 88 Figure A.135: MRT results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.136: MRT Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 89 Figure A.137: MUS results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.138: MUS Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 90 Figure A.139: MWI results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.140: MWI Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 91 Figure A.141: MYS results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.142: MYS Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 92 Figure A.143: NAM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.144: NAM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 93 Figure A.145: NER results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.146: NER Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 94 Figure A.147: NGA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.148: NGA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 95 Figure A.149: NIC results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.150: NIC Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 96 Figure A.151: NPL results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.152: NPL Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 97 Figure A.153: OMN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.154: OMN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 98 Figure A.155: PAK results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.156: PAK Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 99 Figure A.157: PAN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.158: PAN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 100 Figure A.159: PRY results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.160: PRY Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 101 Figure A.161: PSE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.162: PSE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 102 Figure A.163: ROU results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.164: ROU Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 103 Figure A.165: RUS results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.166: RUS Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 104 Figure A.167: RWA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.168: RWA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 105 Figure A.169: SAU results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.170: SAU Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 106 Figure A.171: SDN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.172: SDN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 107 Figure A.173: SEN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.174: SEN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 108 Figure A.175: SLE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.176: SLE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 109 Figure A.177: SLV results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.178: SLV Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 110 Figure A.179: SOM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.180: SOM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 111 Figure A.181: SRB results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.182: SRB Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 112 Figure A.183: SWZ results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.184: SWZ Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 113 Figure A.185: SYC results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.186: SYC Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 114 Figure A.187: SYR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.188: SYR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 115 Figure A.189: TCD results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.190: TCD Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 116 Figure A.191: TGO results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.192: TGO Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 117 Figure A.193: THA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.194: THA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 118 Figure A.195: TJK results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.196: TJK Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 119 Figure A.197: TLS results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.198: TLS Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 120 Figure A.199: TUN results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.200: TUN Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 121 Figure A.201: TUR results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.202: TUR Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 122 Figure A.203: TZA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.204: TZA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 123 Figure A.205: UGA results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.206: UGA Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 124 Figure A.207: URY results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.208: URY Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 125 Figure A.209: VNM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.210: VNM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 126 Figure A.211: VUT results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.212: VUT Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 127 Figure A.213: WSM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.214: WSM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 128 Figure A.215: XKX results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.216: XKX Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 129 Figure A.217: YEM results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.218: YEM Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 130 Figure A.219: ZAF results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.220: ZAF Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 131 Figure A.221: ZMB results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.222: ZMB Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 132 Figure A.223: ZWE results from combined exogenous macroeconomic shocks. Note: The dark purple segment reflects the results from 20 percent of simulations, the lighter purple distribution reflects an additional 30 percent of simulations, while the outer line reflects results from 80 percent of simulations. Figure A.224: ZWE Table results for individual and joint global shocks. US10YR is US long-term interest rate, POIL is oil price, XMKT is export market size, PX is export price, PM is import price, and REM is remittances. 133