## wp1830 — 4.1–5.1 excerpts: Exchange Rate Measure, Scaling of Reserves, Data, Baseline EMP, and The EMP and Global Liquidity Drivers ## Source details **Canonical URL:** [wp1830 — 4.1–5.1 excerpts: Exchange Rate Measure, Scaling of Reserves, Data, Baseline EMP, and The EMP and Global Liquidity Drivers](https://www.imf.org/-/media/files/publications/wp/2018/wp1830.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2018/wp1830.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2018/wp1830.pdf.json) --- ### Exchange rate measure and base-currency choices - Baseline EMP is constructed using the bilateral exchange rate vis-`a-vis the main monetary base currency of the country. - Klein and Shambaugh [2008] (KS) classification is used; available quarterly until the first quarter of 2014 with values extrapolated to 2017. - For the US and the euro area (which do not have KS base currencies) the euro and the USD are used respectively. - In practice most countries in the sample have the USD as base currency. - Notable exceptions: - Several European non-euro countries use the euro as main base currency (and the Deutsche mark before the euro). - Singapore uses the Malaysian baht as base currency. - New Zealand uses the Australian dollar as base currency. - Alternative measure (effective exchange rate) not used because it would not allow converting reserves to effective exchange rate units. ### Scaling of reserves — elasticities and baseline choices - Empirical scaling factor Π_{e,t} requires estimates of elasticities ε^{F A}_e and ε^{F L}_e and information on gross assets and liabilities. - Baseline choice: set elasticities ε^{F A}_e and ε^{F L}_e equal to 0.05. - These elasticities of 0.05 are estimated from simple panel regressions of foreign investment positions for the sample countries (details in Appendix D). - Although elasticities are assumed constant, Π_{i,e} varies over time and across countries with the size of gross investment positions. - Appendix (D) illustrates consequences of alternative elasticity choices. - Measurement and interpretation issues: - Causality from exchange rate movements to portfolio shares is difficult to identify; literature often uses fund-level portfolio equity flows for granularity and timing. - Alternative empirical proxies (e.g., correspondence between central bank FX interventions and exchange rate changes) translate into elasticities of gross private foreign investment positions; correspondence varies across countries and depends on trade, GDP, and net FDI stocks as proxies for local currency market turnover. - The level of gross foreign positions (not month-to-month variation) matters for the scaling factor. ### Data sources, coverage, and sample design - Data drawn from: - National central banks - IMF’s International Financial Statistics - International Investment Positions - The Financial Flows Analytics Databases - Lane and Milesi-Ferretti “External Wealth of Nations Dataset” - Quarterly data supplemented with earlier annual values of gross foreign assets and liabilities to construct Π_{e,t}. - All data sources and definitions are provided in Appendix (A) Table 4; descriptive statistics in Appendix (A) Table 5 for the unbalanced sample period extending from January 1990 to October 2017. - Countries without their own currency are excluded (this excludes individual euro area countries; the euro area as a whole is included). - Estonia and Latvia included up until euro entry dates: January of 2011 and 2014 respectively. - For consistency across all 44 countries, the monthly baseline EMP starts in 2000m1. - Sample includes countries for which data on realized net capital flows and the EMP span a minimum of 10 years. ### The Baseline EMP — constructional distinctions and implementation - Major distinctions from earlier EMP measures: - Different scaling of foreign exchange reserve changes and different weighting of components relative to Girton and Roper [1977], Eichengreen et al. [1994], Kaminsky and Reinhart [1999]. - Scaling with reserves as done by Kaminsky and Reinhart [1999] tends to overstate fluctuations in reserves when reserve levels are very small; problem reduced at larger reserve levels. - Applying precision weights to mismeasured reserve changes can amplify measurement problems. - Baseline scaling scheme properties: - Results in largely zero scaled fluctuations in foreign currency reserves in many cases (reducing spurious large relative movements when reserves are small). - Baseline EMP places weight of one on exchange rate changes, and a weight on reserves changes between zero and one; precision weights in other measures sum to one. - Illustrative country examples: - Switzerland: - Foreign currency reserves grew strongly after the Global Financial Crisis (GFC). - Baseline EMP does not amplify early-sample reserve changes when levels were low, but captures large changes at higher reserve levels later. - Late 2008: exchange rate against the euro appreciated while Swiss franc depreciated against the USD; EMP based on USD bilateral rate captures mild capital outflow pressure, while euro-based EMP suggests capital inflow pressure. - Large realized capital inflow in late 2008 and early 2009 linked to large-scale foreign exchange swap operations by the Swiss National Bank—when swaps are added to reserves the EMP suggests a large capital inflow pressure. - Authors did not correct for central bank FX swap operations because data are not consistently available across counterparties. - Australia: - EMP shows large period-by-period directional swings mainly driven by sharp exchange rate moves vis-`a-vis the USD. - Brazil: - EMP driven by both exchange rate and reserve movements; large reserve accumulations before the GFC, then reserve sales and currency depreciation post-crisis. - Base-currency choice matters: - Base-currency choice affects EMP during episodes when exchange rate changes dominate; less effect when reserve changes dominate. - United Kingdom: persistent directional differences between exchange market pressures against the euro versus against the dollar lead to larger differences between EMP constructions. ### EMP formula, regime weight, and scaling expression - EMP baseline formula with valuation adjustment: - EMP_c,t = de_c,t / e_c,t − Ω_c,t * ˆdR_c,t / Π_e,c,t - Π_e,c,t = (FL_c,t−1 / e_c,t) i + F L_e,c * FL_c,t / e_c,t − F A_e,c * FA_c,t - Ω_t empirical weight: - Ω_t = var(ˆdR_t / Π_e,t) / [ var(ˆdR_t / Π_e,t) + var(de_t / e_t) ] - Ω_t ∈ (0,1); Ω_t = 0 indicates fully flexible rate regime; Ω_t = 1 indicates firmly pegged. - Rationale: - Ω_t captures de facto exchange rate regime from observed variation in scaled reserve changes relative to exchange rate changes. - Rolling standard deviations used for Ω require five years of data before first EMP observation; monthly EMP spans 1995m1 through 2017m10 for countries with data back to 1990. ### Elasticity estimation for gross positions (panel regressions) — key results and caveats - Purpose: estimate elasticities F A_e,c, F L_e,c, F A_i,c, F L_i,c via panel regressions using quarterly data 2000Q1–2016Q4 (excluding crisis quarters 2008Q3 and 2009Q2). - Estimated equations: - dlog(FA_c,t) = F A_e,c dlog(e_c,t−1) + F A_i,c d(i_c,t−1) + F A_w,c dlog(GDP) + φ_c + τ_t + ε_c,t - dlog(FL_c,t) = F L_e,c dlog(e_c,t−1) + F L_i,c d(i_c,t−1) + F L_w,c dlog(GDP) + φ_c + τ_t + ε_c,t - Regression results (Table 6 highlights; preserved exactly where reported): - dlog(FA): - dlog(e): −0.052 ** (significant) - d(i): 0.008 *** (significant) - dlog(GDP): −0.041 (not significant in baseline); when allowing differences between advanced and emerging, dlog(GDP) = 0.067 ** for advanced (column II) - R^2: 0.48 (columns I and II) - No. Obs: 1969; No. countries: 49 - dlog(FL): - dlog(e): −0.006 (column III baseline, not significant); when interacting with emerging & developing countries dlog(e) shows significance for emerging and developing countries (column IV) but with the wrong sign according to portfolio rebalancing expectations. - d(i): 0.0003 (not significant) - dlog(GDP): −0.041 (column III), −0.023 (column IV) - R^2: 0.20 (columns III and IV) - Interpretation highlights: - For FA, a 1% exchange rate depreciation against the US dollar is estimated to lead to a 0.05% reduction in foreign asset holdings (domestic retrenchment implying capital inflow) in baseline. - Elasticity estimates likely downward biased due to remaining endogeneity; negative sign for F A_e,c aligns with portfolio rebalancing theory. - Elasticities appear larger for emerging and developing countries in some specifications. - FL regressions show lower fit and weaker or counterintuitive signs. - Caveats: - Excluding FDI mitigates composition effects but does not solve heterogeneity across asset types. - Global factors controlled via time fixed effects rather than explicit global factor variables. ### Sensitivity analysis for elasticity choices (scenarios) - Baseline elasticity used for Π_e,t scaling in EMP construction: 0.05. - Alternative elasticity scenarios illustrated: - Baseline elasticity: 0.05 - Low alternative elasticity: 0.005 - High alternative elasticity: 1 - Sensitivity findings: - Floating regimes: reserves changes weighted very low; exchange rate component dominates and elasticity choice matters little. - Pegged regimes: reserves changes dominate; elasticity choice rescales amplitude which cancels if index normalized. - Intermediate regimes: elasticity choice matters mainly through amplitude changes in specific episodes. - Country examples: - Switzerland: elasticity matters most — in 2012 (lower bound on exchange rate) baseline and low-elasticity EMPs indicate strong interventions; high-elasticity EMP suggests inflow pressure but less exceptional. - United States: EMP not sensitive to elasticity assumption because reserves play no role under freely floating rates. - Japan and euro area: EMP series shown for baseline and alternatives (figures in source). ### The EMP and realized capital flows — regime-dependent relationships - Comparison (Figure (3)) of baseline EMP with quarterly net private capital outflows in percent of GDP for Australia, Brazil, and Switzerland: - Sign and direction constructed to be directly comparable under model assumptions; level and amplitude differences standardized by country. - Degree of correlation depends on exchange rate regime: - Freely floating regimes: exchange rate adjustments moderate realized private capital flows (e.g., Switzerland before 2008). - Highly managed regimes: capital flow pressures materialize as private net capital flows fully accommodated by changes in central bank foreign reserves, stunting exchange rate response (e.g., post-GFC Switzerland when policy tools shifted toward active exchange rate management). - Scatter (Figure (4)) shows: - Realized private capital flows and the EMP tend to be highly correlated in countries with a high average degree of exchange rate management. - Countries with little exchange rate management (US, EU, UK) exhibit low correlation between EMP and realized net capital flows. - Implication: EMP is advantageous as a measure of capital flow pressures comparable across exchange rate regimes; relying exclusively on realized capital flows risks missing important aspects of pressures. ### The EMP and Global Liquidity Drivers — panel specification and sign expectations - Estimating equation (11): - EMP_t,c = β_i di_t,c + β_i* di*_t + β_s ds*_t + κ_c + ε_t,c. - Sign expectations (from equation (7)): - β_i: expected negative (a lower domestic interest rate induces capital outflows and increases EMP_t,c). - β_i*: expected positive (a higher foreign interest rate induces capital outflows and increases EMP_t,c). - β_s: sign depends on elasticities of gross foreign assets and liabilities to increases in risk perception; literature finds β_s positive for emerging markets and negative for safe-haven countries. - Country fixed effects κ_c capture unobserved country-specific factors; time fixed effects used in alternative specifications to capture common global factors. ### Data, samples, and global-factor measurement for regressions - Sample periods: - Full sample: 2000m1 to 2017m10. - Pre-crisis: up to 2007m6. - Crisis: 2007m7 to 2009m6. - Post-crisis: 2009m7 to 2017m10. - Country sample: 44 countries total: - 11 non-“safe haven” advanced economies (Australia, Canada, Denmark, Estonia, Israel, Latvia, New Zealand, Norway, Sweden, United Kingdom, and the euro area). - 3 “safe havens”: U.S., Japan, Switzerland. - 30 emerging markets and developing countries (listed in source). - Foreign interest rate di*: base-currency short-term policy rate; shadow policy rates from Krippner [2016] used for the US, euro area, Japan, and the UK during ZLB periods. - Global financial risk proxy: dlog(VIX). Robustness: time fixed effects as alternative global factor proxy; GRR later allows substitution of VIX with any monthly global risk proxy. ### Key descriptive statistics for the EMP (Table 2 excerpts; preserved exactly) - (a) U.S., Japan and Switzerland (obs): - Total: Mean -0.010, Max 0.181, Min -0.416, Std. Dev. 0.053, Obs 644. - Pre-crisis: Mean -0.007, Max 0.115, Min -0.272, Std. Dev. 0.043, Obs 269. - Crisis Period: Mean -0.010, Max 0.090, Min -0.116, Std. Dev. 0.044, Obs 72. - Post-crisis: Mean -0.012, Max 0.181, Min -0.416, Std. Dev. 0.062, Obs 303. - (b) Advanced Economies excl. U.S., Japan and Switzerland: - Total: Mean -0.007, Max 0.568, Min -0.552, Std. Dev. 0.073, Obs 2,232. - Pre-crisis: Mean -0.009, Max 0.568, Min -0.440, Std. Dev. 0.076, Obs 988. - Crisis Period: Mean -0.004, Max 0.436, Min -0.450, Std. Dev. 0.099, Obs 264. - Post-crisis: Mean -0.005, Max 0.344, Min -0.552, Std. Dev. 0.062, Obs 980. - (c) Emerging Markets: - Total: Mean -0.032, Max 8.722, Min -5.427, Std. Dev. 0.205, Obs 3,619. - Pre-crisis: Mean -0.044, Max 8.722, Min -5.427, Std. Dev. 0.266, Obs 2,665. - Crisis Period: Mean -0.034, Max 1.140, Min -0.799, Std. Dev. 0.193, Obs 720. - Post-crisis: Mean -0.021, Max 0.794, Min -1.275, Std. Dev. 0.131, Obs 934. - (d) Full Sample: - Total: Mean -0.025, Max 8.722, Min -5.427, Std. Dev. 0.175, Obs 9,195. - Pre-crisis: Mean -0.033, Max 8.722, Min -5.427, Std. Dev. 0.223, Obs 3,922. - Crisis Period: Mean -0.025, Max 1.140, Min -0.799, Std. Dev. 0.168, Obs 1,056. - Post-crisis: Mean -0.017, Max 0.794, Min -1.275, Std. Dev. 0.115, Obs 4,217. ### Panel regression results and interpretation (Table 3 highlights) - General findings: - Variation explained by dlog(VIX) and interest rates is generally low, but higher for safe-haven countries and during crisis and post-crisis samples. - dlog(VIX): - Significant and positive for emerging markets (increased risk aversion → capital outflow pressures). - Positive but insignificant for advanced non-safe haven countries since the crisis. - Negative for safe-haven countries (significant in the pre-crisis period). - Interest rate coefficients mostly insignificant (possibly reflecting poor measurement of funding costs during ZLB). - Time fixed effects (period fixed effects): - Including period fixed effects (excluding VIX and di*) substantially increases explained variation. - Share of variation explained by common global factors highest in safe-haven countries, about 35% (Table 3 columns show R2 values such as 0.352 and 0.358). - EMP attributes a slightly stronger role to a global financial cycle than regressions on realized net capital outflows in percent of GDP for pooled groups. - Representative R2 figures (selected): - Safe-haven full-sample with period fixed effects: R2 0.352 and 0.358. - Advanced non-safe-haven panels: R2 include 0.006, 0.059, 0.013, 0.012, 0.137, 0.176, 0.115 across specifications. - Emerging markets panels: R2 include 0.002, 0.051, 0.017, 0.006, 0.078, 0.145, 0.065 across specifications. - Frequency and obs counts example: - Safe-haven panels show No.Obs 269, 722, 294, 635, 382, 10207 across various specifications and frequencies (monthly "M" and quarterly "Q" as indicated). ### Time fixed effects, cross-group comparisons, and regression fits (Figure 5 excerpts) - Time fixed effects capture common global factors that impact sample countries in the same way; they do not identify individual drivers. - Regression fit examples (from column 5 regressions in Table 3): - EMs vs Non-safe haven AEs (regression fits by period): - Pre − 2007m7: Reg. Line y = 0.0798 x; P-val 0.0243; R2 0.056. - 2007m7 − 2009m6: Reg. Line y = 0.1219 x; P-val 0.1087; R2 0.113. - Post − 2009m7: Reg. Line y = 0.1878 x; P-val 0.0000; R2 0.214. - EMs vs Safe Haven AEs (regression fits by period): - Pre − 2007m7: Reg. Line y = 0.0371 x; P-val 0.3668; R2 0.009. - 2007m7 − 2009m6: Reg. Line y = −0.1270 x; P-val 0.0124; R2 0.252. - Post − 2009m7: Reg. Line y = −0.1649 x; P-val 0.0535; R2 0.042. - Interpretation: - Global factor size for emerging markets is generally about 5 times larger than for other countries (text note). - Global effects pushed more toward capital outflow pressures post-crisis than pre-crisis. - Global factors for emerging markets and advanced economies are significantly positively correlated; safe-haven global factors are significantly negatively correlated with those of emerging markets. ### The Global Risk Response (GRR) index — definition, computation, and interpretation - Definition: - GRR_t,c: country-specific, time-varying empirical correlation of monthly changes in the global factor s* with the monthly country-and-time-specific baseline EMP. - Computation details: - GRR is the partial correlation coefficient between standardized monthly EMP_t,c and the global factor (e.g., dlog(VIX)), controlling for changes in domestic and foreign interest rates. - All variables standardized by sample mean and standard deviation to enable comparability across countries. - GRR values range between -1 and 1. - Interpretation: - |GRR| closer to 1: global risk sentiment is a stronger driver of exchange market pressures relative to other drivers. - GRR positive: capital outflow pressures increase relative to the base currency when global risk sentiment increases. - GRR negative: risk-off events tend to be associated with relative capital inflow pressures. - Bilateral interpretation: GRR is sensitivity relative to the anchor (base currency) country; a positive GRR means more outflow pressure relative to the anchor, not necessarily absolute outflows. - Implementation notes: - Rolling computation uses five-year windows for each country (so for countries with EMP starting 1995m1, GRR can start in 2000m1). - Rolling GRR plotted with one standard error bands (Figure 6 examples for Australia, Brazil, Switzerland). - VIX used in illustrations but can be substituted with other monthly global risk proxies. ### Empirical GRR results and country/time variation (Figure 7 highlights) - Figure 7: GRR_t,c computed for pre-crisis (2000m1–2007m6), crisis (2007m7–2009m6), and post-crisis (2009m7–2017m10); countries ordered by post-crisis GRR. - Key patterns: - Size and frequency of positive GRR values increased strongly from pre-crisis to crisis; positive GRRs are more frequent in crisis and post-crisis samples. - Safe-haven countries (Switzerland, the US, Japan) rank at top of post-crisis list (negative GRR interpreted as relative inflows during risk-off). - Some countries exhibit surprising negative GRRs (e.g., Guatemala, Lithuania), raising questions on bilateral drivers relative to base currency. - Some countries show muted GRRs, suggesting possible roles for capital controls in reducing relative sensitivity. - No clear dichotomy by income/type: both advanced and emerging markets appear at both ends of GRR rankings across periods. - Notable episodes: India, Mexico, and Malaysia show particularly elevated sensitivity during the crisis period. - Statistical significance: - Bars shaded black in Figure 7 indicate correlations significantly different from zero at 90% confidence. ### Conclusions — substantive findings - The EMP index: - Provides a theoretically grounded, balance-of-payments based measure of capital flow pressures that accounts for exchange rate regimes, reserve portfolio composition, gross foreign asset and liability drivers, and both realized flows and incipient pressures manifested in prices. - Offers monthly frequency, broad-country coverage, and computation with only a few months’ lag. - Requires baseline parameter choices, introducing scope for mismeasurement. - The GRR index: - Captures country-specific sensitivity of EMP to global risk sentiment. - Demonstrates substantial cross-country and over-time variation in sensitivities; sensitivities to global risk aversion have increased over recent decades, especially since the global financial crisis. - Shows that some countries exhibit stronger capital inflow pressures during risk aversion increases, while others have stronger outflow pressures—safe-haven status is not immutable. - Overall: - Time fixed effects and the EMP suggest a somewhat stronger role for a global financial cycle than results based solely on realized capital flows. - Important open questions remain on the drivers of cross-country and time variation in responses of capital flows to global risk aversion; the EMP and GRR metrics facilitate further exploration. *Source: wp1830 - 4.1 Exchange Rate Measure; 4.2 Scaling of Reserves; 4.3 Data; 4.4 The Baseline EMP; 5.1 TheEMPand Global Liquidity Drivers (excerpts).* ### 4.1 Exchange Rate Measure ### 4.1 Exchange Rate Measure ### Key definitions and base-currency choices - The baseline EMP is constructed using the bilateral exchange rate vis-`a-vis the main monetary base currency of the country. - The Klein and Shambaugh [2008] (KS) classification is used; it is available quarterly until the first quarter of 2014 with values extrapolated to 2017. - For the US and the euro area, which do not have KS base currencies, the euro and the USD are used respectively. - In practice, most countries in the sample have the USD as base currency. - Exceptions noted: a number of European non-euro countries use the euro as main base currency (and the Deutsche mark before the euro), Singapore uses the Malaysian baht as base currency, and New Zealand uses the Australian dollar as base currency. - An alternative would be to use the effective exchange rate, but an effective exchange rate would not allow converting reserves to effective exchange rate units. ### 4.2 Scaling of Reserves ### Elasticities and baseline choice - Empirical measures of the scaling factor Π_{e,t} require estimates of elasticities ε^{F A}_e and ε^{F L}_e and information on gross assets and liabilities. - The literature typically finds that foreign shares in investors’ portfolios respond significantly negatively to an appreciation shock to the exchange rate of the foreign currency (examples: Hau and Rey [2004], Hau and Rey [2006], Curcuru, Thomas, Warnock, and Wongswan [2014]). - To construct the baseline EMP, and lacking empirical guidance on aggregate gross positions’ responsiveness, the authors set elasticities ε^{F A}_e and ε^{F L}_e equal to 0.05. - These elasticities of 0.05 are estimated from simple panel regressions of foreign investment positions for the sample countries explored (details in Appendix D). - Although elasticities are assumed constant, the resulting scaling factor Π_{i,e} varies over time and across countries with the size of gross investment positions. - The approach to assessing elasticities warrants more research; for application to specific countries, elasticity estimates could be refined based on country-specific data. - Appendix (D) illustrates examples of consequences of alternative choices for the elasticities. ### Measurement and interpretation issues - Difficulty identifying causality from exchange rate movements to portfolio shares leads literature to use fund-level portfolio equity flows for granularity and higher frequency to assess timing. - Alternative empirical approaches (e.g., correspondence between central bank foreign exchange interventions and exchange rate changes) provide proxies for responsiveness but are translated into elasticities of gross private foreign investment positions; correspondence varies across countries and depends on trade, GDP, and net FDI stocks as proxies for local currency market turnover. - The level of gross foreign positions, not month-to-month variation, matters in the scaling factor. ### 4.3 Data ### Data sources and coverage - Data on exchange rates, interest rates, central bank foreign reserves, gross foreign assets, and gross foreign liabilities denominated in foreign currency are drawn from: - National central banks - IMF’s International Financial Statistics - International Investment Positions - The Financial Flows Analytics Databases - Lane and Milesi-Ferretti “External Wealth of Nations Dataset” - To extend coverage back in time, quarterly data are supplemented with earlier annual values of gross foreign assets and liabilities for constructing the Π_{e,t} scaling factor. - All data sources and definitions are provided in Appendix (A) Table 4; descriptive statistics are in Appendix (A) Table 5 for the unbalanced sample period extending from January 1990 to October 2017. - Analyses sorting countries into Advanced Economies versus Emerging and Developing economies utilize the IMF’s country classifications. - Countries without their own currency are excluded (this excludes individual euro area countries, while the euro area as a whole is included). - Estonia and Latvia are included up until their dates of entry into the euro area in January of 2011 and 2014 respectively; countries that joined the euro earlier are not included. - For consistency across all 44 countries, the monthly baseline EMP starts in 2000m1. - The sample includes countries for which data on realized net capital flows and the EMP span a minimum of 10 years. ### 4.4 The Baseline EMP ### Constructional distinctions versus earlier literature - The baseline EMP departs significantly from metrics in Girton and Roper [1977], Eichengreen et al. [1994], and Kaminsky and Reinhart [1999] in both: - the scaling of foreign exchange reserve changes, and - the weighting of components. - Scaling with reserves as done in Kaminsky and Reinhart [1999] tends to overstate fluctuations in reserves when reserve levels are very small; this mismeasurement is reduced at larger reserve levels. - Applying precision weights to mismeasured changes in reserves amplifies the problem. - Differences across measures can be strong in both direction and sign, reflecting: - signal-to-noise ratio of measurement of reserves changes related to valuation effects, and - occasional overweighing of reserves through precision weights. - For countries with floating exchange rates and minimal central bank foreign exchange reserves, even small fluctuations in reserves can be large in relative terms. - In contrast to previous measures, the baseline scaling scheme results in largely zero scaled fluctuations in foreign currency reserves in many cases. - The baseline EMP places a weight of one on exchange rate changes, and a weight on reserves changes between zero and one; precision weights in other measures sum to one. ### Illustrative country examples and base-currency sensitivity - Switzerland: - Absolute value of foreign currency reserves grew strongly after the Global Financial Crisis (GFC). - The baseline EMP does not amplify changes in reserves in the early sample when reserve levels were still low, but captures large changes in reserves at higher levels later in the sample. - In late 2008, the onset of the GFC triggered important capital inflow pressures: exchange rate against the euro appreciated, while Swiss franc depreciated against the USD; the EMP based on USD bilateral rate captures a mild capital outflow pressure, while the baseline euro exchange rate EMP suggests capital inflow pressures. - Large realized capital inflow in late 2008 and early 2009 not reflected in the EMP is related to large scale foreign exchange swap operations by the Swiss National Bank—when swaps are added to reserves the EMP suggests a large capital inflow pressure. - The authors did not correct for central bank foreign exchange swap operations because data are not consistently available across counterparties. - Australia: - EMP shows large period-by-period directional swings mainly driven by sharp exchange rate moves vis-`a-vis the USD (Australia often described as a commodity currency). - Brazil: - EMP driven by both exchange rate and reserve movements; large reserve accumulations occurred before the GFC, with reserve sales and currency depreciation post crisis. - Base-currency choice matters: - Figure (2) shows baseline EMPs for Switzerland and the United Kingdom use the euro as base currency per KS; the alternative EMP uses the USD with US dollar denominated changes in official foreign reserves. - Base currency matters during episodes when exchange rate changes dominate the index, less so when reserves changes dominate. - In the United Kingdom, persistent directional differences between exchange market pressures against the euro versus against the dollar lead to larger differences between the two EMP constructions. ### Relationship with realized capital flows and exchange rate regime - Figure (3) compares baseline EMP with quarterly net private capital outflows in percent of GDP for Australia, Brazil, and Switzerland. - The sign and direction of change in the EMP and realized net private capital flows are constructed to be directly comparable under model assumptions; level and amplitude differences are standardized by country. - Degree of correlation between EMP and actual capital flows depends on the currency regime: - In freely floating regimes, capital flow pressures result in exchange rate adjustments that moderate realized private capital flows (e.g., Switzerland before 2008). - In highly managed exchange rate regimes, capital flow pressures materialize as private net capital flows fully accommodated by changes in central bank foreign reserves, stunting exchange rate response (e.g., post-GFC Switzerland when policy tools shifted toward active exchange rate management). - Figure (4) (scatter) shows country-specific correlation of EMP and actual net private capital flows plotted against a de facto index of exchange rate management: - Realized private capital flows and the EMP tend to be highly correlated in countries with a high average degree of exchange rate management. - Countries with little exchange rate management (US, EU, UK) exhibit low correlation between EMP and realized net capital flows. - Implication: EMP is advantageous as a measure of capital flow pressures comparable across exchange rate regimes; relying exclusively on realized capital flows risks missing important aspects of capital flow pressures. ### 5 Recent Patterns in Capital Flow Pressures (introductory remarks) ### Advantages and analytic applications of the EMP - Advantages of the EMP include: - comprehensiveness (covers all net non-reserve flows), - monthly frequency, and - comparability across countries, currency regimes, and over time. - EMP allows assessment of the link between capital flow pressures and global factors at higher frequency and more consistently across countries than using realized capital flow data based on Balance of Payments Statistics and global liquidity series. - The EMP is applied in regression specifications typically used to study international capital flows and global factors, bearing on the debate over the importance of the global factor over the cycle, at crisis points, and over time. *Source: wp1830 - 4.1 Exchange Rate Measure (excerpts).* ### 5.1 TheEMPand Global Liquidity Drivers ### 5.1 TheEMPand Global Liquidity Drivers ### Panel specification, sign expectations, and estimation strategy - Estimating equation (11): EMP_t,c = β_i di_t,c + β_i* di*_t + β_s ds*_t + κ_c + ε_t,c. - Expectations based on equation (7): - β_i: expected negative (a lower domestic interest rate induces capital outflows and increases EMP_t,c). - β_i*: expected positive (a higher foreign interest rate induces capital outflows and increases EMP_t,c). - β_s: sign depends on elasticities of gross foreign assets and liabilities to increases in risk perception; literature finds β_s positive for emerging markets and negative for safe-haven countries. - Country fixed effects κ_c capture country-specific unobserved factors. Time fixed effects are used in alternative specifications to capture common global factors. ### Data, samples, and global factor measurement - Sample periods: - Full sample: 2000m1 to 2017m10. - Pre-crisis: up to 2007m6. - Crisis: 2007m7 to 2009m6. - Post-crisis: 2009m7 to 2017m10. - Country sample: 44 countries total: - 11 non-“safe haven” advanced economies (Australia, Canada, Denmark, Estonia, Israel, Latvia, New Zealand, Norway, Sweden, United Kingdom, and the euro area). - 3 “safe havens”: U.S., Japan, Switzerland. - 30 emerging markets and developing countries (listed in source). - Foreign interest rate di*: base-currency short-term policy rate; shadow policy rates from Krippner [2016] used for the US, euro area, Japan, and the UK during ZLB periods. - Global financial risk sentiment proxy: dlog(VIX). Robustness: time fixed effects as an alternative global factor proxy; GRR later allows substitution of VIX with any monthly global risk proxy. ### Key descriptive statistics for the EMP (Table 2; samples as defined above) - (a) U.S., Japan and Switzerland (obs): - Total: Mean -0.010, Max 0.181, Min -0.416, Std. Dev. 0.053, Obs 644. - Pre-crisis: Mean -0.007, Max 0.115, Min -0.272, Std. Dev. 0.043, Obs 269. - Crisis Period: Mean -0.010, Max 0.090, Min -0.116, Std. Dev. 0.044, Obs 72. - Post-crisis: Mean -0.012, Max 0.181, Min -0.416, Std. Dev. 0.062, Obs 303. - (b) Advanced Economies excl. U.S., Japan and Switzerland: - Total: Mean -0.007, Max 0.568, Min -0.552, Std. Dev. 0.073, Obs 2,232. - Pre-crisis: Mean -0.009, Max 0.568, Min -0.440, Std. Dev. 0.076, Obs 988. - Crisis Period: Mean -0.004, Max 0.436, Min -0.450, Std. Dev. 0.099, Obs 264. - Post-crisis: Mean -0.005, Max 0.344, Min -0.552, Std. Dev. 0.062, Obs 980. - (c) Emerging Markets: - Total: Mean -0.032, Max 8.722, Min -5.427, Std. Dev. 0.205, Obs 3,619. - Pre-crisis: Mean -0.044, Max 8.722, Min -5.427, Std. Dev. 0.266, Obs 2,665. - Crisis Period: Mean -0.034, Max 1.140, Min -0.799, Std. Dev. 0.193, Obs 720. - Post-crisis: Mean -0.021, Max 0.794, Min -1.275, Std. Dev. 0.131, Obs 934. - (d) Full Sample: - Total: Mean -0.025, Max 8.722, Min -5.427, Std. Dev. 0.175, Obs 9,195. - Pre-crisis: Mean -0.033, Max 8.722, Min -5.427, Std. Dev. 0.223, Obs 3,922. - Crisis Period: Mean -0.025, Max 1.140, Min -0.799, Std. Dev. 0.168, Obs 1,056. - Post-crisis: Mean -0.017, Max 0.794, Min -1.275, Std. Dev. 0.115, Obs 4,217. ### Panel regression results and interpretation (Table 3 highlights) - General findings: - Variation explained by dlog(VIX) and interest rates is generally low, but higher for safe-haven countries and during crisis and post-crisis samples. - dlog(VIX): - Significant and positive for emerging markets (indicating increased risk aversion → capital outflow pressures). - Positive but insignificant for advanced non-safe haven countries since the crisis. - Negative for safe-haven countries (significant in the pre-crisis period). - Interest rate coefficients are mostly insignificant, possibly reflecting poor measurement of funding costs during ZLB periods. - Time fixed effects (period fixed effects) as a proxy for common global factors: - Including period fixed effects (and excluding VIX and di*) substantially increases explained variation. - Share of variation explained by common global factors: - Highest in safe-haven countries, about 35% (noted as "about 35%" and Table 3 columns show R2 values such as 0.352 and 0.358 in safe-haven panels). - EMP appears to attribute a slightly stronger role to a global financial cycle than regressions on realized net capital outflows in percent of GDP for pooled groups (columns 6 and 7 comparisons in Table 3). - Representative R2 figures from Table 3 (selected): - Safe-haven full-sample specification with period fixed effects: R2 0.352 and 0.358 in related columns. - Advanced non-safe-haven panels: R2 values reported include 0.006, 0.059, 0.013, 0.012, 0.137, 0.176, 0.115 across specifications. - Emerging markets panels: R2 values reported include 0.002, 0.051, 0.017, 0.006, 0.078, 0.145, 0.065 across specifications. - Frequency and observation counts noted in Table 3: - Example: Safe-haven panels show No.Obs 269, 722, 294, 635, 382, 10207 across various specifications and frequencies (monthly "M" and quarterly "Q" as indicated). ### Time fixed effects, global factors, and cross-group comparison (Figure 5) - Time fixed effects capture common global factors that impact the sample countries in the same way; they cannot identify individual drivers. - Scatter plots and regression fits (derived from column 5 regressions in Table 3): - EMs vs Non-safe haven AEs (regression fits by period): - Pre − 2007m7: Reg. Line y = 0.0798 x; P-val 0.0243; R2 0.056. - 2007m7 − 2009m6: Reg. Line y = 0.1219 x; P-val 0.1087; R2 0.113. - Post − 2009m7: Reg. Line y = 0.1878 x; P-val 0.0000; R2 0.214. - EMs vs Safe Haven AEs (regression fits by period): - Pre − 2007m7: Reg. Line y = 0.0371 x; P-val 0.3668; R2 0.009. - 2007m7 − 2009m6: Reg. Line y = −0.1270 x; P-val 0.0124; R2 0.252. - Post − 2009m7: Reg. Line y = −0.1649 x; P-val 0.0535; R2 0.042. - Interpretation: - Global factor size for emerging markets is generally about 5 times larger than for other countries (as noted in text). - Global effects pushed more toward capital outflow pressures post-crisis than pre-crisis. - Global factors for emerging markets and advanced economies are significantly positively correlated; safe-haven global factors are significantly negatively correlated with those of emerging markets. ### The Global Risk Response index (GRR): definition and properties - GRR_t,c: country-specific, time-varying empirical correlation of monthly changes in the global factor s* with the monthly country-and-time-specific baseline EMP. - Computation details: - GRR is the partial correlation coefficient between standardized monthly EMP_t,c and the global factor (e.g., dlog(VIX)), controlling for changes in domestic and foreign interest rates. - All variables standardized by sample mean and standard deviation to enable comparability across countries. - GRR values range between -1 and 1. - Interpretation: - |GRR| closer to 1: global risk sentiment is a stronger driver of exchange market pressures relative to other drivers. - GRR positive: capital outflow pressures increase relative to the base currency when global risk sentiment increases. - GRR negative: risk-off events tend to be associated with relative capital inflow pressures. - Bilateral interpretation: GRR is sensitivity relative to the anchor (base currency) country; a positive GRR means more outflow pressure relative to the anchor, not necessarily absolute outflows. - Implementation notes: - Rolling computation uses five-year windows for each country (so for countries with EMP starting 1995m1, GRR can start in 2000m1). - Rolling GRR plotted with one standard error bands (Figure 6 examples for Australia, Brazil, Switzerland). - VIX used in illustrations but can be substituted with other monthly global risk proxies. ### Empirical GRR results and country/time variation (Figure 7 and discussion) - Figure 7: GRR_t,c computed for pre-crisis (2000m1–2007m6), crisis (2007m7–2009m6), and post-crisis (2009m7–2017m10) samples; countries ordered by post-crisis GRR. - Key patterns: - Size and frequency of positive GRR values increased strongly from pre-crisis to crisis; positive GRRs are more frequent in crisis and post-crisis samples. - Safe-haven countries (Switzerland, the US, Japan) rank at top of post-crisis list (negative GRR interpreting as relative inflows during risk-off). - Some countries exhibit surprising negative GRRs (e.g., Guatemala, Lithuania) raising questions on bilateral drivers relative to base currency. - Some countries show muted GRRs, suggesting possible roles for capital controls in reducing relative sensitivity. - No clear dichotomy by income/type: both advanced and emerging markets appear at both ends of the GRR rankings across periods. - Notable episodes: India, Mexico, and Malaysia show particularly elevated sensitivity during the crisis period. - Statistical significance: Bars shaded black in Figure 7 indicate correlations significantly different from zero at 90% confidence. ### Conclusions (summary of substantive findings) - The EMP index: - Provides a theoretically grounded, balance-of-payments based measure of capital flow pressures that accounts for exchange rate regimes, reserve portfolio composition, gross foreign asset and liability drivers, and both realized flows and incipient pressures manifested in prices. - Offers monthly frequency, broad-country coverage, and computation with only a few months’ lag. - Requires baseline parameter choices, introducing scope for mismeasurement. - The GRR index: - Captures country-specific sensitivity of EMP to global risk sentiment. - Demonstrates substantial cross-country and over-time variation in sensitivities; sensitivities to global risk aversion have increased over recent decades, especially since the global financial crisis. - Shows that some countries exhibit stronger capital inflow pressures during risk aversion increases, while others have stronger outflow pressures—safe-haven status is not immutable. - Overall: - Time fixed effects and the EMP suggest a somewhat stronger role for a global financial cycle than results based solely on realized capital flows. - Important open questions remain on the drivers of cross-country and time variation in responses of capital flows to global risk aversion; the EMP and GRR metrics facilitate further exploration. *Source: 5.1 TheEMPand Global Liquidity Drivers (wp1830)* ### References ### References ### Bibliographic entries - Joshua Aizenman, Sebastian Edwards, and Daniel Riera-Crichton. Adjustment patterns to commodity terms of trade shocks: The role of exchange rate and international reserves policies. Journal of International Money and Finance, 31(8):1990–2016, 2012. - Joshua Aizenman, Mahir Binici, and Michael M. Hutchison. The Transmission of Federal Reserve Tapering News to Emerging Financial Markets. NBER Working Papers 19980, National Bureau of Economic Research, Inc, Mar 2014a. - Joshua Aizenman, Yin-Wong Cheung, and Hiro Ito. International Reserves Before and After the Global Crisis: Is There No End to Hoarding? Working Paper 20386, National Bureau of Economic Research, August 2014b. - Stefan Avdjiev, Leonardo Gambacorta, Linda S. Goldberg, and Stefano Schiaffi. The Shifting Drivers of Global Liquidity. BIS Working Papers 17/tbd, Bank for International Settlements, 2017. - Olivier Blanchard, Francesco Giavazzi, and Filipa Sa. International investors, the u.s. current account, and the dollar. Brookings Papers on Economic Activity, 36(1):1–66, 2005. - Olivier J Blanchard, Gustavo Adler, and Irineu E de Carvalho Filho. Can Foreign Exchange Intervention Stem Exchange Rate Pressures from Global Capital Flow Shocks? IMF Working Papers 15/159, International Monetary Fund, July 2015. - Luigi Bonatti and Andrea Fracasso. Hoarding of international reserves in China: Mercantilism, domestic consumption and US monetary policy. Journal of International Money and Finance, 32(C):1044–1078, 2013. - Dennis P. J. Botman, Irineu E. Carvalho Filho, and Raphael W. Lam. The Curious Case of the Yen as a Safe Haven Currency: A Forensic Analysis. IMF Working Papers 13/228, International Monetary Fund, November 2013. - William H. Branson and Dale W. Henderson. The specification and influence of asset markets. In R. W. Jones and P. B. Kenen, editors, Handbook of International Economics, volume 2 of Handbook of International Economics, chapter 15, pages 749–805. Elsevier, January 1985. - Robin Brooks and Michael Cahill. Fear of Floating: Foreign Exchange Intervention in Asia. Global Economics Paper 229, Goldman Sachs, February 2016. - Bundesbank. Exchange rates and financial stress. Monthly report, Bundesbank, Jul 2014. - Ricardo J. Caballero, Emmanuel Farhi, and Pierre-Olivier Gourinchas. Safe Asset Scarcity and Aggregate Demand. American Economic Review, 106(5):513–518, May 2016. - Eugenio Cerutti, Stijn Claessens, and Damien Puy. Push Factors and Capital Flows to Emerging Markets: Why Knowing Your Lending Matters More than Fundamentals. Working Paper 127, International Monetary Fund, 2015. - Eugenio Cerutti, Stijn Claessens, and Andrew K Rose. How Important is the Global Financial Cycle? Evidence from Capital Flows. CEPR Discussion Papers 12075, C.E.P.R. Discussion Papers, June 2017. - Anusha Chari, Karlye Dilts Stedman, and Christian Lundblad. Taper Trantrums: QE, its aftermath and Emerging Market Capital Flows. Working Papers 17/tbd, UNC-Chapel Hill, 2017. - Stephanie E. Curcuru, Charles P. Thomas, Francis E. Warnock, and Jon Wongswan. Uncovered Equity Parity and rebalancing in international portfolios. Journal of International Money and Finance, 47(C):86–99, 2014. - Irineu de Carvalho Filho. Risk-off Episodes and Swiss Franc Appreciation: the Role of Capital Flows. Working Papers 13.07, Swiss National Bank, Study Center Gerzensee, July 2013. - Michael P. Dooley, David Folkerts-Landau, and Peter Garber. The revived Bretton Woods system. International Journal of Finance & Economics, 9(4):307–313, 2004. - Barry Eichengreen and Poonam Gupta. Tapering Talk: The Impact of Expectations of Reduced Federal Reserve Security Purchases on Emerging Markets. MPRA Paper 53040, University Library of Munich, Germany, Jan 2014. - Barry Eichengreen, Andrew K. Rose, and Charles Wyplosz. Speculative Attacks on Pegged Exchange Rates: An Empirical Exploration with Special Reference to the European Monetary System. NBER Working Papers 4898, National Bureau of Economic Research, Inc, October 1994. - Barry Eichengreen, Andrew Rose, and Charles Wyplosz. Contagious Currency Crises: First Tests. Scandinavian Journal of Economics, 98(4):463–484, December 1996. - Barry Eichengreen, Livia Chiandtcedilu, and Arnaud Mehl. Stability or Upheaval?; The Currency Composition of International Reserves in the Long Run. IMF Economic Review, 64(2):354–380, June 2016. - Rasmus Fatum and Yohei Yamamoto. Intra-safe haven currency behavior during the global financial crisis. Globalization and Monetary Policy Institute Working Paper 199, Federal Reserve Bank of Dallas, September 2014. - Kristin Forbes and Francis Warnock. Capital flow waves: Surges, stops, flight, and retrenchment. Journal of International Economics, 88:235 – 251, 2012. - Marcel Fratzscher. Capital flows, push versus pull factors and the global financial crisis. Journal of International Economics, 88(2):341–356, 2012. - Atish R. Ghosh, Mahvash S. Qureshi, Jun Il Kim, and Juan Zalduendo. Surges. Journal of International Economics, 92(2):266–285, 2014. - Lance Girton and Dale W. Henderson. Financial capital movements and central bank behavior in a two-country, short-run portfolio balance model. Journal of Monetary Economics, 2(1):33 – 61, 1976. - Lance Girton and Don Roper. A monetary model of exchange market pressure applied to the postwar Canadian experience. Americal Economic Review, 67:537 – 548, 1977. - Linda S. Goldberg, Cindy E. Hull, and Sarah Stein. Do industrialized countries hold the right foreign exchange reserves? Current Issues in Economics and Finance, 19(April), 2013. - Pierre-Olivier Gourinchas and H ́el`ene Rey. Real interest rates, imbalances and the curse of regional safe asset providers at the zero lower bound. Working Paper 22618, National Bureau of Economic Research, September 2016. - Maurizio M. Habib and Livio Stracca. Getting beyond carry trade: What makes a safe haven currency? Journal of International Economics, 87(1):50–64, 2012. - Harald Hau and H ́el`ene Rey. Can Portfolio Rebalancing Explain the Dynamics of Equity Returns, Equity Flows, and Exchange Rates? American Economic Review, 94(2):126–133, May 2004. - Harald Hau and H ́el`ene Rey. Exchange Rates, Equity Prices, and Capital Flows. Review of Financial Studies, 19(1):273–317, 2006. - Dale W. Henderson and Kenneth Rogoff. Negative net foreign asset positions and stability in a world portfolio balance model. Journal of International Economics, 13(1-2):85–104, August 1982. - Ethan Ilzetzki, Carmen M. Reinhart, and Kenneth S. Rogoff. The Country Chronologies to Exchange Rate Arrangements into the 21st Century: Will the Anchor Currency Hold? NBER Working Papers 23135, National Bureau of Economic Research, Inc, February 2017. - Gabriela Kaminsky and Carmen Reinhart. The twin crises: The Causes of Banking and Balance-of-Payments Problems. American Economic Review, 89(3):473 – 500, 1999. - Franc Klaassen. Identifying the Weights in Exchange Market Pressure. Tinbergen Institute Discussion Papers 11-030/2, Tinbergen Institute, February 2011. - Franc Klaassen and Henk Jager. Definition-consistent measurement of exchange market pressure. Journal of International Money and Finance, 30(1):74–95, February 2011. - Michael W. Klein and Jay C. Shambaugh. The dynamics of exchange rate regimes: Fixes, floats, and flips. Journal of International Economics, 75(1):70–92, May 2008. - Pentti J.K. Kouri. Balance of payments and the foreign exchange market: A dynamic partial equilibrium model. Working Paper 644, National Bureau of Economic Research, March 1981. - Leo Krippner. Documentation for measures of monetary policy. July 2016. - Signe Krogstrup and C ́edric Tille. Foreign currency bank funding and global factors. Manuscript, Graduate Institute and International Monetary Fund, 2017. - Jie Li, R. Rajan, and Thomas Willett. Measuring Currency Crises Using Exchange Market Pressure Indices: The Imprecision of Precision Weights). Unpublished manuscript, 2006. - Lukas Menkhoff. Foreign Exchange Intervention in Emerging Markets: A Survey of Empirical Studies. The World Economy, 36(9):1187–1208, 09 2013. - Gian Maria Milesi-Ferretti and Philip R. Lane. International financial integration in the aftermath of the global financial crisis. Working Paper 115, International Monetary Fund, May 2017. - Gian-Maria Milesi-Ferretti and C ́edric Tille. The great retrenchment: International capital flows during the global financial crisis. Economic Policy, 66:285–330, 2011. - Silvia Miranda-Agrippino and H ́el`ene Rey. World Asset Markets and the Global Financial Cycle. NBER Working Papers 21722, National Bureau of Economic Research, Inc, November 2015. - Prachi Mishra, Kenji Moriyama, Papa M’B. P. N’Diaye, and Lam Nguyen. Impact of Fed Tapering Announcements on Emerging Markets. IMF Working Papers 14/109, International Monetary Fund, Jun 2014. - Christopher J. Neely. Are changes in foreign exchange reserves well correlated with official intervention? Review, Federal Reserve Bank of St. Louis, pages 17–32, Sep 2000. - Maurice Obstfeld, Jonathan D. Ostry, and Mahvash Saeed Qureshi. A tie that binds: Revisiting the trilemma in emerging market economies. CEPR Discussion Papers 12093, C.E.P.R. Discussion Papers, June 2017. - Ila Patnaik, Joshua Felman, and Ajay Shah. An exchange market pressure measure for cross country analysis. Journal of International Money and Finance, 73, Part A:62 – 77, 2017. - Angelo Ranaldo and Paul Soederlind. Safe haven currencies. Review of Finance, 14(3):385–407, 2010. - H ́el`ene Rey. Dilemma not Trilemma: The Global Financial Cycle and Monetary Policy Independence. Working Paper 21162, National Bureau of Economic Research, May 2015a. - H ́el`ene Rey. International Channels of Transmission of Monetary Policy and the Mundellian Trilemma. forthcoming, Mundell Fleming Lecture, IMF Economic Review, 2015b. - Ahmed Shaghil, Brahima Coulibaly, and Andrei Zlate. International Financial Spillovers to Emerging Market Economies: How Important Are Economic Fundamentals? International Finance Discussion Papers 1135, Board of Governors of the Federal Reserve System (U.S.), Apr 2015. - Jay C. Shambaugh. The Effect of Fixed Exchange Rates on Monetary Policy. The Quarterly Journal of Economics, 119(1):301–352, 2004. - Hyun Song Shin. The bank/capital markets nexus goes global. Technical report, Bank for International Settlements, Nov 2016. - Evan Tanner. Exchange Market Pressure and Monetary Policy: Asia and Latin America in the 1990s. IMF Working Papers 99/114, International Monetary Fund, August 1999. - Evan Tanner. Exchange Market Pressure, Currency Crises, and Monetary Policy: Additional Evidence From Emerging Markets. IMF Working Papers 02/14, International Monetary Fund, January 2002. - Diana N. Weymark. Estimating exchange market pressure and the degree of exchange market intervention for Canada. Journal of International Economics, 39(3-4):273–295, November 1995. - Diana N Weymark. A General Approach to Measuring Exchange Market Pressure. Oxford Economic Papers, 50(1):106–121, January 1998. - Alfred Wong and Tom Fong. Gauging the Safehavenness of Currencies. Working Papers 132013, Hong Kong Institute for Monetary Research, Sep 2013. ### Data Sources and Definitions (Appendix A) - Variable: e — Baseline bilateral exchange rate. - Definition: End-of-month mid-point between bid and ask, domestic per unit of foreign. - Source and Description: National central banks, the Federal Reserve and IMF International Financial Statistics. - Variable: reer — Real effective exchange rate - Definition: Monthly frequency. Defined such that an increase denotes a depreciation of the domestic currency. - Source and Description: IMF International Financial Statistics. - Variable: R — Official foreign exchange reserves (total reserves minus gold) - Definition: In billions of base currency units, end of period, monthly. - Source and Description: Dollar value from IMF International Financial Statistics. - Note: Observations from 2004m5 to 2004m8 for Sri Lanka are excluded due to huge outlier values. - Variable: i, i* — Monetary policy or short-term rate - Definition: In percentage points, end of period, monthly. - Source and Description: IMF International Financial Statistics or national Central Banks. Constructed as IFS policy rate line 60 if available, else policy rate from national central bank if available, else 3-month money market interest rate from IFS (line 60b) if available, else short-term treasury bond rate (IFS line 60c) if available, else deposit rates from IFS (only needed for parts of the sample period for Nicaragua, Panama, China and Argentina). For countries that have introduced negative policy interest rates, the relevant policy rate prior to the introduction of a negative rate is merged with the relevant rate post *Italic source attribution line.* ### introduction for Denmark, Japan and EU. ### introduction for Denmark, Japan and EU. ### Data and Definitions - Data frequency: monthly; span January 1990 to October 2017. - Sample: 44 countries included; 14 advanced economies (Australia, Canada, Denmark, Estonia, Israel, Japan, Latvia, New Zealand, Norway, Sweden, Switzerland, United Kingdom, United States and the euro area), and 31 emerging markets and developing countries (listed in source). - Key series and units: - SSR: Shadow policy rate in the US, EU, Japan and UK (in percentage points, end of period, monthly). - GDP: Gross domestic product (In USD, quarterly). - M1: Narrow monetary aggregate (In billions of USD, end of period, monthly). - M0: Monetary base (In billions of USD, end of period, monthly). - V IX: CBOE Volatility Index (End of period, monthly). - FA: Total Foreign Assets less FDI (In billions of base currency units, end of period, yearly interpolated to monthly). - FL: Total Foreign Liabilities less FDI (In billions of base currency units, end of period, yearly interpolated to monthly). - NCAP: Net private capital outflows in percent of GDP (In percent, quarterly). - Summary descriptive statistics (selected, preserved exactly): - Advanced (observations and stats aggregated across advanced group): - dR: Mean 0.5389, Max 5.687, Min -32.520, Std. Dev 4.751, Obs 3851 - GDP: 2016.961 19250.00040.9723616.3843628 (preserved as in source) - M1: 952.6628580.7830.0001652.8583699 (preserved as in source) - Emerging (selected): - dR: Mean 0.7061, Max 12.596, Min -107.922, Std. Dev 6.470, Obs 10084 - GDP: 453.95711882.5603.2481073.7528517 (preserved as in source) - M1: 1132.2597411.4010.191579.2358090 (preserved as in source) - Notes: - Gross foreign positions interpolated from yearly frequency; GDP interpolated from quarterly frequency. - Base currency references follow Klein and Shambaugh [2008]; data available yearly until 2014Q1, interpolated to monthly and extrapolated to 2017. ### EMP Construction and Exchange-Regime Weighting - EMP baseline formula with valuation adjustment: - EMP_c,t = de_c,t / e_c,t − Ω_c,t * ˆdR_c,t / Π_e,c,t - Π_e,c,t = (FL_c,t−1 / e_c,t) i + F L_e,c * FL_c,t / e_c,t − F A_e,c * FA_c,t - Ω_t empirical weight: - Ω_t = var(ˆdR_t / Π_e,t) / [ var(ˆdR_t / Π_e,t) + var(de_t / e_t) ] - Ω_t ∈ (0,1); Ω_t = 0 indicates fully flexible rate regime; Ω_t = 1 indicates firmly pegged. - Rationale: - Ω_t captures de facto exchange rate regime from observed variation in scaled reserve changes relative to exchange rate changes. - In hard pegs, exchange rate variation is absent and reserve variation reflects pressure on the peg; in flexible regimes, exchange rate variation dominates. - Implementation detail: - Rolling standard deviations used for Ω require five years of data before first EMP observation; monthly EMP spans 1995m1 through 2017m10 for countries with data back to 1990. ### Exchange-Rate Elasticities and Panel Regression Results - Purpose: estimate elasticities of foreign assets (FA) and foreign liabilities (FL) to exchange rates and interest rates (F A_e,c, F L_e,c, F A_i,c, F L_i,c) using panel regressions of gross positions. - Estimation setup: - Regressions use quarterly data from 2000Q1 to 2016Q4, excluding crisis quarters 2008Q3 and 2009Q2. - Country fixed effects (φ_c) and time fixed effects (τ_t) included. - Lagged explanatory variables by one quarter to reduce endogeneity. - Regressions exclude FDI from gross positions. - Panel of 49 countries; No. Obs: 1969 (for columns reported). - Estimated equations: - dlog(FA_c,t) = F A_e,c dlog(e_c,t−1) + F A_i,c d(i_c,t−1) + F A_w,c dlog(GDP) + φ_c + τ_t + ε_c,t - dlog(FL_c,t) = F L_e,c dlog(e_c,t−1) + F L_i,c d(i_c,t−1) + F L_w,c dlog(GDP) + φ_c + τ_t + ε_c,t - Regression results (Table 6, preserved exact coefficients where reported): - dlog(FA): - dlog(e): −0.052 ** (significant) - d(i): 0.008 *** (significant) - dlog(GDP): −0.041 (not significant in baseline); when allowing differences between advanced and emerging, dlog(GDP) = 0.067 ** for advanced (column II) - R^2: 0.48 (columns I and II) - No. Obs: 1969; No. countries: 49 - dlog(FL): - dlog(e): −0.006 (column III baseline, not significant); when interacting with emerging & developing countries dlog(e) shows significance for emerging and developing countries (column IV) but with the wrong sign according to portfolio rebalancing expectations. - d(i): 0.0003 (not significant) - dlog(GDP): −0.041 (column III), −0.023 (column IV) (mostly not significant) - R^2: 0.20 (columns III and IV) - Interpretation highlights: - For FA, a 1% exchange rate depreciation against the US dollar is estimated to lead to a 0.05% reduction in foreign asset holdings (i.e., domestic retrenchment implying capital inflow) in baseline. - Elasticity estimates are likely downward biased due to remaining endogeneity; negative sign for F A_e,c aligns with portfolio rebalancing theory. - Elasticities appear larger for emerging and developing countries in some specifications. - For FL regressions, fit is lower and signs/significance are weaker or counterintuitive. - Caveats: - Excluding FDI mitigates composition effects but does not fully solve heterogeneity across asset types. - Global factors (US monetary policy, risk aversion, global liquidity) controlled via time fixed effects rather than explicit global factor variables. ### Sensitivity Analysis and Country Examples (Denmark, Japan, euro area) - Baseline elasticity used for Π_e,t scaling in EMP construction: 0.05. - Alternative elasticity scenarios shown in Figure 8: - Baseline elasticity: 0.05 - Low alternative elasticity: 0.005 - High alternative elasticity: 1 - EMPs displayed as quarterly averages of monthly values. - Sensitivity findings: - EMP sensitivity to elasticity depends on exchange rate regime: - Floating regimes: reserves changes weighted very low via valuation adjustment; exchange rate component dominates and elasticity choice matters little. - Pegged regimes: reserves changes dominate; elasticity choice only rescales amplitude which cancels if index normalized. - Intermediate regimes: elasticity choice matters mainly through amplitude changes in specific episodes. - Country examples: - Switzerland: elasticity matters most. In 2012 (lower bound on exchange rate), baseline and low-elasticity EMPs indicate strong interventions (exceptional capital inflow pressures). High-elasticity EMP suggests inflow pressure but less exceptional. - United States: EMP not sensitive to elasticity assumption because reserves play no role in exchange market pressures under freely floating rate. - Japan and euro area: EMP series shown in figures (quarterly averages) alongside alternative elasticity assumptions (baseline, Alt High Elasticity, Alt Low Elasticity). (Figures included in source; preserved conceptually here.) - Individual country charts: - Baseline Goldberg Krogstrup EMP series plotted against realized net private capital outflows in %GDP for many countries including Denmark (panel l), Japan (panel j), and euro area (panel b). - Series are demeaned and normalized by full sample standard deviation. ### Global Risk Response (GRR) and Charts - GRR constructed using the Goldberg Krogstrup baseline EMP and V IX as a measure of global risk sentiment. - Computation notes: - Rolling GRR_base series with 90% confidence bands plotted for each country (examples include Denmark, Japan, euro area among many others). - Global financial crisis period structural break assumed. - Visual findings (from figures): - Rolling GRR_base: mean and one standard error band shown by date for each country (figures cover 2000m1 through 2018m1 in source charts). - GRR series presented for Denmark (panel t in Figure 10), Japan (panel j in Figure 10), and euro area (panel b in Figure 10) among other countries. *Source: wp1830 - introduction for Denmark, Japan and EU.* --- _Source: https://www.imf.org/-/media/files/publications/wp/2018/wp1830.pdf_