## _090313a ## Source details **Canonical URL:** [_090313a](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2013/_090313a.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2013/_090313a.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2013/_090313a.pdf.json) --- ### Key empirical findings on bond yields and tapering events - Changes in yields are computed in the day following each announcement, then cumulated. - Tapering announcements occurred on May 22 and June 19, 2013. - Z-scores for capital flows represent the difference of average capital flows (equity and bond) in the month after first U.S. tapering announcement (May 22, 2013) relative to flows since 2009, normalized by the standard deviation of flows. - Example contemporaneous responses to a recent 25 bps jump in U.S. yields: - Australia: 35 bps jump in Australian long-term rates. - Brazil: 24 bps jump in Brazilian long-term rates. - Canada: 23 bps jump in Canadian long-term rates. - China: 6.8 bps jump in Chinese long-term rates. - India: 12 bps jump in Indian long-term rates. - Indonesia: 36 bps jump in Indonesian long-term rates. - Korea: 21 bps jump in Korean long-term rates. - Mexico: 58 bps jump in Mexican long-term rates. - Poland: 35 bps jump in Polish long-term rates. - Russia: 46 bps jump in Russian long-term rates. - South Africa: 23 bps jump in South African long-term rates. - Thailand: 20 bps jump in Thai long-term rates. - Turkey: 95 bps jump in Turkish long-term rates. ### Measures of exposure and resilience (overview) - Exposure — market volatility and sensitivity: - Sensitivity to a change in U.S. long-term bond yield: regression of non-UMP country 10-year yields on U.S. 10-year yields (changes over two-day intervals after FOMC announcements, Jan 1, 2003–May 20, 2013). - Change in long-term bond yields following U.S. tapering announcements: cumulative change in two days after Fed tapering on May 22 and June 19, 2013. - Exposure — capital outflows: - Net inflows following U.S. tapering announcements: average EPFR Global bond and equity fund flows in the month following May 22, 2013 relative to average monthly flows since 2009 (Z-score). Z-scores of -2 or lower indicate net inflows fell by more than 2 average standard deviations. - Sovereign credit rating: lower of Moody’s and S&P, expressed in terms of S&P ratings. - Resilience — domestic market conditions: - Market capitalization: percent of GDP. - Turnover ratio: total value of shares traded divided by average market capitalization. - Size of domestic institutional investors: market size of pension, mutual fund and insurance companies, in percent of external debt. - Resilience — dependence on foreign funding: - Foreign equity liabilities: foreign ownership of domestic equities in percent of GDP. - Total external debt: gross external debt in percent of GDP (all sectors). - Resilience — policy room: - Fiscal policy room: primary balance adjustment (percent of GDP) necessary to achieve debt target in 2030. - Monetary policy room: inflation gap from its target and output gap. - Room for exchange rate adjustment: difference in REER relative to level consistent with medium-term fundamentals (as of May 2013). - Trade balance: percent of GDP. - FX reserves: months of import cover. - Bank capital ratio: bank capital and reserves to total assets. - Non-performing loans (NPLs): NPLs to total gross loans. ### Cross-indicator interpretation - Countries most responsive to U.S. Fed bond-purchase announcements did not necessarily show the largest response to tapering announcements; changing investor positioning and market liquidity may explain differences. - Main vulnerability in several cases is the large increase in foreign-held sovereign bonds; banking and corporate sectors often appear better positioned. ### Case studies — Australia and Brazil (highlights) - Australia - Growth slowed to 2.6 percent in Q1 2013. - Inflation within RBA target range of 2–3 percent. - Policy rate 2.5 percent after cuts from 4.75 percent since 2011. - Net external liabilities about 60 percent of GDP. - Gross inflows US$60 billion in 3Q 2012 (4Q rolling sum). - Nonresidents hold about 75 percent of Australian debt stock (increase of 15 percent since end-2009). - Exposure: historical sensitivity to U.S. monetary shocks; a recent 25 bps U.S. jump led to a contemporaneous 35 bps jump in Australian long-term rates. - Inflows fell significantly after tapering (z-score of -4.04). - Resilience: deep, liquid markets; nonresident holdings up, banks’ short-term wholesale external funding reliance declined and largely hedged; exchange rate has space to depreciate; fiscal space sufficient. - Note: strong credit rating may prevent larger capital outflows. - Brazil - Growth slowed to 0.9 percent in 2012. - Headline inflation close to upper band of 4½ ±2 percent target range. - BCB increased policy rate by 175 bps since April 2013 (as of August 2013). - Gross public debt 68 percent of GDP in 2012. - External debt about 14 percent of GDP. - 2010 gross inflows US$162 billion (7.4 percent of GDP); recent year gross inflows US$104 billion. - Net portfolio inflows declined most; net FDI stable and sufficient to cover current account deficit. - Exposure: high sensitivity to U.S. shocks; a recent 25 bps U.S. jump led to a contemporaneous 24 bps jump in Brazilian long-term rates. - Inflows fell after tapering (z-score of -2.15). - Resilience: banking sector stable with ample capital buffers; corporate leverage and external debt increased but within historical ranges. ### Selected country summaries (key statistics preserved) - Canada - Growth slowed to 1.7 percent in 2012; Q1 2013 rebound. - Current account deficit 3.4 percent of GDP in 2012. - Inflation in H1 2013 lower than 1 percent (yoy). - External gross debt low relative to other advanced economies. - Net portfolio inflows ~5 percent of GDP per year in 2009–12. - Exposure: 25 bps U.S. jump → 23 bps jump in Canadian long-term rates. - Inflows did not significantly fall after tapering (z-score of -0.46). - China - Growth 7.8 percent in 2012; inflation 2.5 percent in 2012. - Central government debt 14.4 percent of GDP in 2012. - Capital account subject to broad capital controls. - FDI inward 3.8 percent of GDP in 2011; net capital outflows ~US$20 billion in 2012; net inflows resumed since Q1 2013. - Exposure: 25 bps U.S. jump → 6.8 bps jump in Chinese long-term rates. - Inflows fell after tapering (z-score of -1.97). - Resilience: low dependence on foreign funding; FX reserves ample; on-budget fiscal stimulus recommended if growth falls well below 7½ percent target. - India - Growth 5.0 percent in FY 2012–13; potential 6.5 percent. - Consumer price inflation around 10 percent. - External debt ~21 percent of GDP. - Public deficit and government debt 67 percent of GDP limit fiscal space. - Net foreign inflows ~US$100 billion in 2012–13. - Exposure: 25 bps U.S. jump → 12 bps jump in Indian long-term rates. - Inflows fell after tapering (z-score of -1.67). - Resilience constraints: shallow markets; high current account deficit and dependence on portfolio debt and equity flows. - Indonesia - Growth 6.2 percent in 2012; inflation 4.3 percent end-2012; rose to 8.6 percent in July 2013 after fuel price increases. - Policy and deposit rates raised by 125 bps since June 2013 to 7 percent and 5.25 percent respectively (as of end-August 2013). - Public debt under 25 percent of GDP. - About one-third of rupiah government debt held by nonresidents. - Exposure: 25 bps U.S. jump → 36 bps jump in Indonesian long-term rates. - Inflows fell after tapering (z-score of –2.05). - Resilience: banking sector well capitalized (CAR 16 percent); government debt exposure to foreign funding high; domestic markets shallow. - Korea - Growth 2.3 percent yoy in 2Q 2013; inflation 1.4 percent in July 2013. - Public debt 34 percent of GDP. - External debt 34 percent of GDP (late 2012). - Capital inflows net US$-44 billion in 1Q 2013 (4Q rolling sum). - Exposure: 25 bps U.S. jump → 21 bps jump in Korean long-term rates. - Inflows: modest outflows from equity post-tapering (z-score of -1.42); bond inflows continued. - Mexico - Average 4 percent annual growth since 2010. - Core inflation 2½ percent in July. - Public debt ~43 percent of GDP. - Gross inflows US$86 billion in 2012. - Exposure: 25 bps U.S. jump → 58 bps jump in Mexican long-term rates. - Inflows fell moderately after tapering (z-score of -1.65). (Additional country entries in the source include Poland, Russia, South Africa, Thailand, Turkey with country-specific numeric indicators and z-scores retained in the source.) ### Macrofinancial model: nonlinear amplification, design, and simulation scenarios - Purpose: - Endogenize macrofinancial amplification mechanisms and capture global nonlinear feedback in large distress scenarios; patterns/mechanisms are primary outputs, not precise numerical predictions. - Novel features: - Aggregate, non-diversifiable credit risk on banks’ loan books endogenously derived from macro developments. - Optimal bank balance-sheet and capital-buffer choices under uncertainty constrained by risk-bearing capacity of bank capital. - Endogenous interaction of aggregate credit risk, loan portfolio value, bank capital buffers, and relative costs of internal and external equity. - Model structure: - Real sector: small open economy with production for local consumption and exports; distinguishes short-run and long-run elasticity of substitution and allows import substitution. - Financial sector: banks extend non-traded loans and create matching deposits; banks hold capital buffers above regulatory minima that vary endogenously. - Risk and feedback: loan defaults driven by income/wealth evolution; loan risks share a common systemic component; adverse shocks raise impaired loans → write-downs → asset liquidation → higher lending costs → fire sales and credit crunches. - Calibration and limitations: - Real sector calibrated to characteristics of emerging Central and Eastern European countries. - Parametrization of macro-financial linkages limited by sparse data on nonlinear distress episodes; quantitative predictions should not be taken literally. - Simulation design (stress scenarios): - Phase I: three years of cheap foreign finance; Phase II: sudden reversal and gradual return to normal. - Phase I shock: cost of foreign financing falls by 200 bps. - Phase II shock: cost of foreign financing initially increases by 300 bps (overshooting normal by 100 bps), then converges back to normal within ~five years. - Behavioral assumption: banks and borrowers are myopic in Phase I and do not internalize reversal risk. - Two economies: “resilient” with 5 percent FX loans share; “vulnerable” with 50 percent FX loans share. - For each economy: run (i) linearized model and (ii) global nonlinear solution; policy passive (no countercyclical macroprudential policies). - Phase I dynamics: - Exchange rate appreciation, rising asset prices, faster credit growth, stronger import demand and net foreign liabilities increase. - FX lending magnifies upturn via valuation improvements that reduce loan-to-value ratios. - Capital adequacy ratios decline as capital buffers thin. - Linearized models over-predict the upturn relative to nonlinear model. - Phase II reversal dynamics: - Capital flow reversal and higher foreign funding costs reduce sustainable debt; current account adjusts via lower consumption/investment and improved exports from depreciation. - Access to bank credit deteriorates; currency depreciation lowers real household income; wage pass-through sluggish. - Vulnerable economy (high FX loans): valuation effects magnify local-currency outstanding bank credit; nonlinear feedbacks sharply increase non-performing loans; capital buffers drop by as much as 1 percentage point (negligible in resilient economy). - Banks cut lending, raise spreads to recapitalize; higher spreads depress demand and asset prices, depreciate currency further, and increase NPLs—feedback loop. - Linearized models underestimate downside severity; nonlinear simulations show much larger adverse effects for vulnerable economies. - Simulation outcomes report variables including: - Cost of foreign funds (percent PA), Real GDP, Domestic demand, Lending spread (percent PA), Lending rates (percent PA), Policy rate (percent PA), Inflation (percent PA), CAR, Real bank loans, Trade balance to GDP, Labour income, Nominal exchange rate. ### Empirical analysis of UMP effects on flows and asset prices - Data and methods: - EPFR Global weekly net portfolio flows from >14,000 equity funds and >7,000 bond funds (US$8 trillion AUM); study covers 25 AEs and 17 EMEs from Jan 1, 2007 to Aug 26, 2013. - Adjusted EPFR flows for increasing coverage; adjustment almost halves observed bond and equity flows by end of sample. - Analytical tools: Bayesian dynamic latent factor model; regressions linking weekly flows to central bank announcements and purchases; regressions of flows on asset prices; event studies on daily asset prices for forward guidance. - Baseline weekly regression: Flow_it = α_i + β_i Announce_t + γ_i Purchase_t + η_i conditioning variables_t + ε_it. - Conditioning variables include VIX, oil prices, and non-oil prices. - EPFR coverage and validation: - EPFR coverage represents <20 percent of market capitalization in equity and bonds for most countries (lower for bonds). - Correlations between adjusted EPFR quarterly flows and BoP flows (2008–12): EMEs bond/equity correlations ~0.5 in aggregate; country-level variation (examples: China bond correlation 0.10, equity 0.73; India bond 0.15, equity 0.79; Russia bond 0.78, equity 0.70). - EPFR flows are a reasonable proxy for gross international bond flows to EMEs and equity flows to most economies, with caveats. - Global factor, synchronization, and May 22, 2013 episode: - Common global factor explains synchronized outflows after Fed tapering (May 22, 2013). - Equity flows: global factor explained 30–80 percent early in crisis; after U.S. QE1, global factor has explained ~80 percent of EMEs bond flow variation since then. - Since May 22, the role of global factors in explaining portfolio flow variability jumped; Table 2 indicates the share of actual flows due to global factor is virtually 100 percent for bond flows over May 22–Aug 26, 2013. - Quantified flows around May 22 to Aug 26, 2013: - EMEs bond funds: estimated outflows US$10 billion (0.2 percent of EMEs bond market size). - AEs bond funds: estimated outflows US$11 billion (0.1 percent of AE bond market size); US$7.6 billion from U.S. bond funds alone. - Equity funds: EMEs outflows ~US$20 billion, of which US$16 billion flowed into U.S. equity funds and remainder into Japanese and European equity funds. - Note: mutual fund flows represent <20 percent of overall portfolio flows; EPFR may understate total portfolio rebalancing. - Regression and program-specific findings: - Mutual fund flows respond more to actual UMP bond purchases than to announcements alone. - Different UMP programs operate via different mechanisms; impact varies by program and market conditions. - Key program findings: - U.S. LSAP1A, U.S. tapering, Japan LSAP: both announcements and purchases had significant coefficients for bond flows. - Purchases drove U.S. LSAP1B, U.S. LSAP2, U.S. LSAP3, U.K. FLS, euro-area LTRO programs. - U.S. LSAP3 (in more stable conditions) initially produced synchronized inflows into EM bond and non-U.S. equity funds; Fed tapering (May 22) largely reversed these inflows. - BoJ policies supported domestic bond and equity markets; Japan LSAP associated with major inflows into Asian bond funds. - ECB and BoE programs generally had limited global bond-flow effects, with exceptions (e.g., U.K. LSAP1). ### Impact of fund flows on asset prices and forward guidance - Mutual fund flows → asset prices: - Regression ∆Asset price_it = α_i + β_i Flow_t + γ_i ∆Asset price_it-1 + η_i conditioning variables_t + ε_it. - Findings: - Greater inflows associated with higher asset prices. - Equity coefficients universally correctly signed and generally highly significant. - Bond results: ~1/3 of coefficients significant at 10 percent; overwhelmingly correctly signed. - Only 5 of 33 coefficients were incorrectly signed (probability <0.01 percent by chance). - UMP-related flows and asset prices: - UMP-related flows behave similarly to overall EPFR flows; when used alone, UMP-related flows support asset prices with few incorrectly signed coefficients (three for equities and nine for bonds). - Bilateral exchange rates: - Both overall and UMP-related fund flows show detectable links with bilateral dollar exchange-rate movements (13 emerging markets tested). - Forward guidance event study (daily two-day returns, Jan 1, 2003–Jun 26, 2013): - Monetary surprises measured by changes in 1-year-ahead futures 3-month LIBOR rate. - Findings: - Forward guidance primarily increases global equity prices and appreciates local currencies, with little impact on bond yields. - Asset purchases (quantitative easing) particularly in the U.S. lowered global bond yields and associated with some fall in equity prices and local currency depreciation. - Example: U.S. LSAP1 surprise lowered long-term bond rates in Brazil and Mexico by over 9 bps. - Fed tapering announcements led to a major increase in the expected 3-month Libor rate in one year (market “surprise”), contributing to repricing of risk, rising domestic interest rates and weakening equity prices in most EMs. - Three high-level conclusions on UMP: 1. UMP-linked capital flows play a significant role in overall bond flows. 2. Flows are often dominated by actual purchases, highlighting importance of both announcement and purchase effects. 3. Impacts vary by program; aggregate UMP summary measures can be misleading. ### Dominance of a single global factor in EM bond flows (key quantitative evidence) - Key finding: "Most bond flows to emerging markets continue to be driven by a single global factor." - Selected Table 2 entries (May 22–Aug 26, 2013): - Emerging Markets — Bonds: Actual Flows 9873; Estimated flows due to global factor 9818; As percent of bond market size 0.19. - Emerging Markets — Equities: Actual Flows 20389; Estimated flows due to global factor 20361; As percent of equity market size 0.21. - Asia — Bonds: Actual Flows 2870; Estimated flows due to global factor 2862; As percent of bond market size 0.11. - China — Bonds: Actual Flows -105; Estimated due to global factor -104. - India — Bonds: Actual Flows -35; Estimated due to global factor -34. - Latin America — Bonds: Actual Flows 4343; Estimated due to global factor 4337; As percent of bond market size 0.26. - o/w BRICS — Bonds: 3129; Estimated due to global factor 3118. - Total Outflows (Table 2): bonds 21020 actual – 20962 estimated (approx. 0.05 as percent metric shown in the source). - Table 3 (EPFR vs BoP correlations and EPFR/BoP ratios) — selected entries: - Emerging Markets: Bond correlation 0.44; EPFR/BoP 0.08. Equity correlation 0.38; EPFR/BoP 0.22. - BRICS: Bond correlation 0.45; EPFR/BoP 0.08. Equity correlation 0.67; EPFR/BoP 0.40. - China: Bond correlation 0.10; EPFR/BoP 0.02. Equity correlation 0.73; EPFR/BoP 0.46. - India: Bond correlation 0.15; EPFR/BoP 0.01. Equity correlation 0.79; EPFR/BoP 0.27. - Russia: Bond correlation 0.78; EPFR/BoP 0.16. Equity correlation 0.70; EPFR/BoP 0.54. ### Policy-relevant implications drawn in the source - Bond flows to EMs, being largely driven by a single global factor, can concentrate and transmit global market risk into EMs; this makes EM bond markets an important potential source of global risk. - Different UMP instruments have distinct spillover channels: forward guidance primarily affects equities and currencies via expectations; asset purchases affect bond markets via flow and portfolio-balance channels. - Assessment of spillovers should account for program-specific features, market conditions, and the evolving importance of purchases versus announcement effects. - Macroeconomic policy and macroprudential frameworks should consider nonlinear amplification mechanisms (e.g., high shares of FX lending) and prepare for stress scenarios where passive policy responses can magnify downturns. - Stress-scenario simulations indicate policy space to: - Accommodate higher interest rates where necessary, - Use reserves to limit FX overshooting, - Consider macroprudential tools to build buffers during inflow phases to reduce vulnerability to abrupt reversals. *Source: GLOBAL IMPACT AND CHALLENGES OF UNCONVENTIONAL MONETARY POLICIES—BACKGROUND PAPER, INTERNATIONAL MONETARY FUND, September 12, 2013.* ### 1. Total Cumulated Changes in 10-year Bond Yields Following U.S. LSAP ### 1. Total Cumulated Changes in 10-year Bond Yields Following U.S. LSAP Announcements and Tapering Announcements ### Key empirical findings (figures and measures) - Changes in yields are computed in the day following each announcement, then cumulated. - Tapering announcements occurred on May 22 and June 19, 2013. - Source: Fund staff estimates. - Z-scores for capital flows represent the difference of average capital flows (equity and bond) in the month after first U.S. tapering announcement (May 22, 2013) relative to flows since 2009, normalized by the standard deviation of flows. ### Measures of exposure and resilience (Table 1 summary) - Exposure: market volatility - Sensitivity to a change in U.S. long-term bond yield - Based on regressions of non-UMP country 10-year bond yields on U.S. 10-year bond yields (regression run in changes over two-day intervals immediately following FOMC announcements between January 1, 2003 and May 20, 2013). - Larger correlations indicate higher sensitivity to U.S. monetary shocks. Insignificant coefficients suggest no historical sensitivity to U.S. monetary shocks. - Change in long-term bond yields following U.S. tapering announcements - Cumulative change in a country’s 10-year bond yields in two days following the tapering announcements by the U.S. Federal Reserve Bank (Fed) on May 22 and June 19, 2013. Larger changes suggest high sensitivity to the tapering announcements. - Exposure: capital outflows - Net inflows following U.S. tapering announcements - Average EPFR Global bond and equity fund flows in the month following the Fed tapering announcement of May 22, 2013, relative to average monthly flows since 2009, normalized by the standard deviation of flows (Z-score). - Z-scores of -2 or lower indicate net inflows fell by more than 2 average standard deviations. - Sovereign credit rating - Lower of Moody’s and S&P ratings, expressed in terms of S&P ratings. Better rating suggests lower risk of outflows. - Resilience: domestic market conditions - Market capitalization: Market value of listed companies expressed in percent of GDP. - Turnover ratio: Total value of shares traded divided by average market capitalization. - Size of domestic institutional investors: Market size of pension, mutual fund and insurance companies, in percent of external debt. - Resilience: dependence on foreign funding - Foreign equity liabilities: Foreign ownership of domestic equities in percent of GDP. - Total external debt: Gross external debt in percent of GDP (all sectors). - Resilience: policy room - Fiscal policy room: Primary balance adjustment (in percent of GDP) necessary to achieve debt target in 2030 (from the IMF Fiscal Monitor, April 2013 (IMF 2013e)). - Monetary policy room: Inflation gap from its target and output gap. - Room for exchange rate adjustment: Differences in Real Effective Exchange Rate (REER) relative to its level consistent with medium term fundamentals and desirable policy (as of May 2013, taken from the 2013 External Stability Report, IMF 2013d). - Trade balance: Trade balance for goods and services in percent of GDP. - FX reserves: FX reserves in months of import cover. - Bank capital ratio: Ratio of bank capital and reserves to total assets. - Non-performing loans (NPLs): NPLs to total gross loans. ### Cross-indicator evidence and interpretation - Evidence from the second and third indicators of “exposure”: - Countries showing the largest response to U.S. Federal Reserve announcements of bond purchases did not necessarily show the largest response to tapering announcements. - This may be due to the changing nature of investor positioning and market liquidity. ### Case study highlights — Australia - Background macro indicators and policy settings - Growth has slowed to 2.6 percent in the first quarter of 2013. - Inflation has remained contained within the Reserve Bank of Australia’s (RBA) target range of 2–3 percent. - The monetary policy rate is at 2.5 percent after a series of cuts from 4.75 percent since 2011. - External debt is relatively low; net external liabilities are high (about 60 percent of GDP). - Capital account is open with limited intervention in the Australian dollar since 1983. - Capital inflows (up to the May 2013 tapering announcements) - Gross inflows stand at US$60 billion in 3Q 2012 (4Q rolling sum), similar to their 2000–07 average. - Nonresidents hold about 75 percent of Australian debt stock (an increase of 15 percent since end-2009). - Foreign investors reportedly repatriated funds following RBA rate cuts; the Australian dollar depreciated by nearly 15 percent between April and mid-August 2013. - Coping with inflows — potential vulnerabilities and resilience - Policy response to capital inflows was passive, mainly allowing the exchange rate to appreciate. - REER appreciated by almost 40 percent between late 2008 and April 2013. - The cyclically-adjusted current account deficit was assessed to be around 1–2 percent of GDP weaker than implied by medium-term fundamentals and desirable policy settings. - Asset prices buoyed on the back of rising house prices and capital inflows. - Banking sector is mostly stable with satisfactory capital buffers and/or appropriate hedges. - Exposure: Australian long-term rates are historically sensitive to U.S. monetary policy shocks (regression results). The recent jump in U.S. yields of 25 bps immediately following U.S. tapering announcements led to a contemporaneous 35 bps jump in Australian long-term rates. - Inflows fell significantly in the month following the recent U.S. tapering announcements (z-score of -4.04). - Resilience: Nonresident holdings of government debt have increased, while banks’ reliance on short-term wholesale external funding has declined and foreign liability positions are largely hedged. Australia’s deep and liquid financial markets should be better suited to deal with shocks. - The exchange rate has some space to depreciate. Fiscal policy space is sufficient. - Notes - Australia’s strong credit rating may prevent larger capital outflows. ### Case study highlights — Brazil - Background macro indicators and policy settings - Growth slowed to 0.9 percent in 2012—the lowest rate since 2009. - Headline inflation is close to the upper band of the Central Bank of Brazil’s (BCB) 4½ ±2 percent target range. - BCB has recently embarked on a tightening cycle and increased the policy rate by 175 bps since April 2013, as of August 2013. - Gross public debt stood at 68 percent of GDP in 2012, about 3–4 percentage points of GDP higher than pre-crisis levels. - External debt is about 14 percent of GDP, with mostly long-term maturities. - Brazil has made wide use of capital flow management measures (CFMs). - Capital inflows (up to the May 2013 tapering announcements) - In 2010 gross inflows amounted to US$162 billion (7.4 percent of GDP)—exceeding the pre-crisis peak at US$125 billion. - In the past year gross inflows were moderate at US$104 billion. - Net portfolio inflows—both debt and equity—have declined the most, while net FDI inflows have remained mostly stable and sufficient to cover the current account deficit. - Coping with inflows — potential vulnerabilities and resilience - Policy response included exchange rate adjustments, interventions to mitigate volatility, prudential measures (reserve requirements) and ample use of CFMs such as the financial transaction tax (IOF). Since late 2011, authorities started relaxing some of these measures. - The cyclically-adjusted current account deficit was assessed to be moderately larger than implied by medium-term fundamentals and desirable policies of about -1¾ to -2¾ percent of GDP. The REER was deemed moderately overvalued (IMF 2013d), but depreciated recently. - Asset prices: 10Y government local currency bond yields decreased to less than 10 percent by mid-May 2013 but have risen since then. Equity prices have been trending lower. Real-estate prices in some cities rose sharply while housing credit expanded strongly from a low base. - Banking sector is stable with ample capital buffers. Corporate leverage and external debt have increased since 2010, while remaining within historical ranges. Liquidity has improved as maturities have lengthened. - Exposure: Brazilian long-term rates are historically sensitive to U.S. monetary policy shocks (regression results). Brazil’s financial markets are very sensitive to global financial conditions (higher beta than most EMs). The recent jump in U.S. yields of 25 bps immediately following U.S. tapering announcements led to a contemporaneous 24 bps jump in Brazilian long-term rates. - Capital inflows as measured using high frequency mutual fund data fell significantly in the month following the recent U.S. tapering announcements (z-score of -2.15). *Source: GLOBAL IMPACT AND CHALLENGES OF UNCONVENTIONAL MONETARY POLICIES—BACKGROUND PAPER, INTERNATIONAL MONETARY FUND, September 12, 2013.* ### Appendix Table 1 for details on the ### _090313a - Appendix Table 1 for details on the ### General findings and resilience (selected excerpts) - Balance of payments coverage is much wider and suggests that flows have in fact picked up somewhat. - Resilience: - Neither the banking nor corporate sectors are heavily dependent on foreign funding, but capital flow reversals could complicate funding of current account deficits. - Large accumulated reserves (US$374 billion) can be used to prevent FX overshooting or disorderly adjustments. - The BCB recently announced a substantial program of preannounced interventions to limit such risks. - The BCB can lower reserve requirements from current high levels to ease liquidity pressures in the banking system. ### Canada - Background: - Growth slowed to 1.7 percent in 2012, below medium-term potential at around 2 percent, but rebounded in Q1 2013 driven by stronger external demand. - The current account deficit widened to 3.4 percent of GDP in 2012 but is projected to narrow somewhat in 2013. - Inflation in H1 2013 has been lower than 1 percent (yoy), well below the 2 percent inflation target. - The current accommodative monetary policy stance remains appropriate. - External gross debt is low relative to other advanced economies but is expected to gradually increase over time. - Capital account is fully open with no unilateral intervention on the Canadian dollar since 1998. - Capital inflows: - Average net portfolio outflows from 2000–07 turned into inflows of close to 5 percent of GDP per year on average in 2009–12. - The share of Canadian debt securities in the U.S. investors’ bond portfolio increased from 11 percent in 2007 to 16 percent in 2011. - About one-third of non-financial private sector credit is financed by external sources. - Potential vulnerabilities and exposure: - No specific policy response to capital inflows per se; currency appreciated in line with flexible exchange rate policy. - Cyclically-adjusted current account deficit of around 2½ percent of GDP in 2012 is 1–3 percentage points larger than the value implied by medium-term fundamentals and desirable policies. - REER was deemed moderately overvalued (IMF 2013d). - Housing prices above levels consistent with fundamentals; household debt at a historical high relative to disposable income. - Banking sector is stable with satisfactory capital buffers in line with Basel III. - Exposure: Canadian long-term interest rates are historically sensitive to U.S. monetary policy shocks: a recent 25 bps jump in U.S. yields led to a contemporaneous 23 bps jump in Canadian long-term rates. - Inflows did not significantly fall in the month following U.S. tapering announcements (z-score of -0.46). - Resilience and risks: - Domestic capital market is relatively deep; net external liabilities relatively modest; sizeable fraction of external liabilities denominated in Canadian dollars. - If long-term interest rates overshoot following a U.S. exit, risks could increase of a disorderly rebalancing of housing prices and pressures on household balance sheets. ### China - Background: - Growth moderated to 7.8 percent in 2012. - Inflation slowed to 2.5 percent in 2012. - Monetary policy prudent; monetary aggregate growth exceeded annual target of 14 percent so far this year. - Central government debt declined to 14.4 percent of GDP in 2012 (expanding government definition would yield a higher figure). - Capital account remains subject to broad capital controls on both inflows and outflows covering the bulk of non-FDI portfolio flows and external borrowing. - Capital inflows: - FDI equal to net FDI was US$230 billion; inward FDI was equal to 3.8 percent of GDP in 2011. - In 2012, China experienced net capital outflows of around US$20 billion; since Q1 2013 net capital inflows seem to have resumed, along with appreciation pressures. - Potential vulnerabilities and exposure: - Policy tools used: reserve requirements, higher interest rates, tighter prudential measures, direct administrative limits, and currency intervention. - Exports slowed to 8 percent growth in 2012. - REER appreciated by 19 percent from end-2009 through May 2013 and 35 percent since the mid-2005 exchange rate reform. - Staff assessment: external position appears moderately stronger and the currency moderately undervalued compared with the level consistent with medium-term fundamentals and desirable policies. - Real estate market susceptible to large cyclical swings; credit stock (including off-balance sheet and non-bank lending) increased sharply by 60 percent of GDP in just 4 years since 2009. - Stock price index nearly doubled in 2008–09, but decreased by 30 percent from 2010 to May 2013. - Bank balance sheets appear healthy based on reported data, but non-bank activity has grown significantly; rapid financial innovation raises risks. - Exposure: Chinese long-term rates historically not significantly related to U.S. monetary policy shocks; a recent 25 bps jump in U.S. yields led to a contemporaneous 6.8 bps jump in Chinese long-term rates. - Inflows fell significantly in the month following U.S. tapering announcements (z-score of -1.97). - Resilience and policy space: - Low dependence on foreign funding provides resilience. - Ample FX reserves should help cushion outflows. - If growth slowed too much below the authorities’ 7½ percent target, on-budget fiscal stimulus should be used to support activity, emphasizing rebalancing and protecting vulnerable households. ### India - Background: - Growth slowed to 5.0 percent in fiscal year 2012–13, below potential of 6.5 percent. - Consumer price inflation remained around 10 percent. - Monetary policy accommodative with policy rate cuts since April 2012; more recently tightened liquidity and interest rate defense measures as the rupee came under pressure. - External debt low at around 21 percent of GDP. - Public deficit and government debt of 67 percent of GDP limit fiscal policy space. - Current account deficit reached a record 4.8 percent of GDP in 2012–13. - Capital account relatively closed but gradually liberalized to attract FDI and portfolio investment. - Capital inflows: - Net foreign inflows around US$100 billion in 2012–13 (below pre-crisis peak of about US$110 billion). - FDI inflows robust; debt flows (especially short-term) increased notably. - Potential vulnerabilities and exposure: - Policy supported capital inflows via relaxation of external commercial borrowing and FDI rules. - Floating rupee has been an important shock absorber; authorities intervened sporadically. - Cyclically adjusted current account deficit and REER deemed broadly in line with fundamentals (IMF 2013d), but the rupee is vulnerable to flow fluctuations. - Asset prices evolved moderately since 2008; equity price index recovered to pre-crisis level though some moderation since May 2013. - Banking sector mostly stable with satisfactory capital buffers, though asset quality deteriorated considerably. - Corporate debt increased since 2008; foreign ownership rose as regulations eased. - Exposure: Long-term bond yields historically not sensitive to U.S. monetary policy shocks; recent 25 bps U.S. jump led to a contemporaneous 12 bps jump in Indian long-term rates. - Inflows fell in the month following the recent tapering announcements (z-score of -1.67), especially on the debt side. - Resilience and policy constraints: - Some markets are shallow and illiquid. - Combination of high current account deficit and dependence on external portfolio debt and equity flows is a key vulnerability. - Policy room limited: higher interest rates could widen the output gap; international reserves could dampen exchange rate volatility; fiscal space limited due to high public deficits. - As of mid-August 2013 the rupee was down by 12 percent since the Fed’s tapering announcement; Reserve Bank of India tightened money market liquidity drastically and implemented measures aimed at reducing gold imports. ### Indonesia - Background: - Growth eased to 6.2 percent in 2012. - Year-on-year inflation 4.3 percent at end-2012, within central bank’s target range of 4.5±1 percent, but rose to 8.6 percent in July 2013 following a 33 percent (average) increase in subsidized fuel prices. - Policy and deposit rates raised by 125 bps since June 2013 to 7 percent and 5.25 percent respectively, as of end-August 2013. - Public debt fell to under 25 percent of GDP from around 76 percent in 2001. - Capital account relatively open, but most banking flows are subject to capital controls. - Capital inflows: - Since 2009 Indonesia remained an attractive destination; in 2012 FDI hit a new high and net debt issuance rebounded. - Going forward FDI expected around 1½-2 percent of GDP, with net debt issuance amounting to 1-1½ percent of GDP. - About one-third of rupiah government debt is held by nonresidents. - Potential vulnerabilities and exposure: - Active policy response: FX interventions, lowering bottom of interest rate corridor, tightened CFMs including minimum holding period for central bank bills, prudential limits on consumer and mortgage lending. - Cyclically-adjusted current account deficit appears ½–2½ percentage points of GDP weaker than implied by fundamentals and desirable policies. - REER deemed approximately fairly valued (IMF 2013d). - Asset prices rose sharply since 2008: 10Y government bond yields decreased by 600 bps; stock price index quadrupled between end-2008 and May 2013. - Banking sector profitable and well capitalized; capital to risk weighted asset ratio at 16 percent. - Exposure: Indonesian long-term rates historically highly sensitive to U.S. monetary policy shocks; a recent 25 bps jump in U.S. yields led to a contemporaneous 36 bps jump in Indonesian long-term rates. - Net inflows fell significantly in the month following U.S. tapering announcements (z-score of –2.05). - Resilience and constraints: - Government debt exposure to foreign funding is high; other sectors’ exposure limited but rising. - Domestic capital markets are shallow. - Fiscal policy space seems limited in the event of significant outflows. ### Korea - Background: - Growth rebounded to 2.3 percent yoy in 2Q 2013 after slowdown in 2012; output gap remains negative. - Inflation fell to 1.4 percent in July 2013, below the central bank’s 2.5-3.5 percent target band. - Public debt at 34 percent of GDP provides ample fiscal space; subdued inflation provides monetary policy space. - External debt fell to 34 percent of GDP as of late 2012, majority long-term. - Capital account relatively open. - Capital inflows: - Remained volatile—particularly equity and other flows—since 2008. - In net terms, capital inflows stood at US$-44 billion in 1Q 2013 (4Q rolling sum). - Potential vulnerabilities and exposure: - Policy response included macroprudential measures (since 2010), withholding tax on foreign purchases of treasury and monetary stabilization bonds, levy on banks’ non-deposit FX liabilities, FX interventions to rebuild reserves. - Current account surplus widened to 3.8 percent of GDP in 2012, 1–4 percentage points above value implied by fundamentals and desirable policies. - REER assessed moderately undervalued (IMF 2013d). - Asset prices not subject to sustained upward pressure since 2008; 10Y government bond yields decreased to around 3.7 percent from 5.8 percent. - Banking sector strong with large capital buffers; corporate distress indicators have risen to a fairly high level. - Exposure: Korean long-term rates historically somewhat sensitive to U.S. monetary policy shocks; a recent 25 bps U.S. jump led to a contemporaneous 21 bps jump in Korean long-term rates. - Modest outflows from equity market occurred in the month following tapering announcements (z-score of -1.42), though bond inflows continued. - Resilience and policy room: - Korea’s open and integrated financial markets increase exposure but vulnerability reduced relative to 2008. - Banking sector still dependent on foreign funding, though much less than in 2008. - Policy room exists: monetary policy could be deployed and fiscal policy has space to provide support if needed. ### Mexico - Background: - Economy averaged 4 percent annual growth since 2010. - Headline inflation has come down close to the 3 percent inflation target; core inflation fell to a record low 2½ percent in July. - Monetary policy accommodative since 2009; reference rate cut by 50 bps early this year. - Public debt around 43 percent of GDP. - Capital account open with deep and liquid financial markets. - Capital inflows: - Gross inflows stood at US$86 billion in 2012, exceeding pre-crisis peak of US$60 billion. - Foreign portfolio debt exposures increased significantly; share of foreign-held sovereign debt increased to 34 percent since mid-2009. - Potential vulnerabilities and exposure: - Policy response consistent with framework: monetary and fiscal policy, exchange rate flexibility, rule-based FX interventions, absence of CFMs. - Cyclically-adjusted current account deficit stable around 1 percent of GDP. - REER deemed fairly valued (IMF 2013d). - Asset prices increased significantly since 2008: equities increased by 74 percent between January 2009 and May 2013. - Banking sector mostly stable with solid capital buffers and limited FX loan exposure. - Corporate leverage around 49 percent; short-term liabilities account for 48 percent of total liabilities. - Exposure: Mexican long-term rates historically somewhat sensitive to U.S. monetary policy shocks; a recent 25 bps U.S. jump led to a contemporaneous 58 bps jump in Mexican long-term rates. - Inflows fell moderately in the month following U.S. tapering announcements (z-score of -1.65). *Source: _090313a - Appendix Table 1 for details on the (excerpted content).* ### Appendix Table 1 for details on the ### _090313a - Appendix Table 1 for details on the ### Overarching resilience note - Main vulnerability relates to the large increase in foreign-held sovereign bonds. - Banking and corporate sectors appear well positioned in case of a sudden stop. - In the event of significant outflows: - The economy can accommodate higher interest rates. - Reserves can be used to limit currency overshooting. - The Flexible Credit Line (FCL) arrangement has been an effective complement to international reserves as insurance against global tail risks. - Fiscal policy has little room to expand. ### Poland - Background: - Growth slowed to 1.9 percent in 2012, and is below potential. - Inflation fell sharply below the lower band of the target range (1.5–3.5 percent) in 2013. - Monetary policy began easing in November 2012; cumulative rate cuts since then amount to 225 bps bringing the reference rate to 2½ percent. - Public debt dropped to 55.6 percent of GDP in 2012. - Poland granted a two-year extension to exit the EU excessive deficit procedure. - Capital account is mostly open. - Capital inflows (up to the May 2013 tapering announcements): - Net inflows (4Q rolling average) stood at US$5.1 billion in Q4 2012, compared with a peak of US$12.8 billion in Q2 2008. - Coping with inflows — Potential vulnerabilities and indicators: - Policy response mostly passive; monetary policy focused on inflation; sporadic FX intervention; prudential tightening on FX mortgages. - Exports closely linked to core EU countries (particularly Germany). - Cyclically-adjusted current account improved in 2012 and broadly consistent with fundamentals. - REER deemed fairly valued (IMF 2013d). - Asset prices: EMBI spreads dropped from 350 bps (end 2011) to around 100 bps (May 2013); sovereign CDS spreads dropped by about 200 bps to about 80 bps (between May 2012 and May 2013). - Credit growth declined in 2012 vs 2011; real-estate prices declining. - Banking sector: CAR of 15¼ percent (core tier 1 ratio of 13¾ percent) at end-March 2013; FX-mortgage overhang remains a risk. - Corporate leverage (debt-to-equity) rose from 42 percent in 2007 to 49 percent in 2012 with pick-up in bond financing in 2012. - Exposure: - Historically Polish long-term bond rates not significantly related to U.S. monetary policy shocks; however, a recent U.S. 25 bps jump led to a contemporaneous 35 bps jump in Polish long-term rates. - Inflows fell significantly in the month following U.S. tapering announcements (z-score of -1.95). - Resilience: - Market used to express regional views but relatively small and vulnerable to emerging market investor sentiment. - Foreign ownership share of government bonds reached 37 percent in April 2013. - Mitigants: stable investor base, prudent public debt management, strong banking-sector liquidity. - Stress scenario: - If significant outflows occur, fiscal space is limited and higher interest rates would weigh on inflation and output (both below target/potential); exchange rate would bear the brunt. ### Russia - Background: - Growth above potential since early-2010 but slowed to below 2 percent in 1Q 2013. - Inflation projections close to target range; monetary policy bias towards tightening. - Non-oil fiscal deficit remains above level needed to save oil revenue; public and external debt low at 12 percent and 28 percent, respectively. - Capital account relatively open since financial liberalization. - Capital inflows (up to the May 2013 tapering announcements): - Net outflows: 2012 net outflows US$30 billion (1.5 percent of GDP), compared to US$21 billion in 2010. - Coping with inflows — Potential vulnerabilities and indicators: - No policy response to capital inflows. - Exports decreased in 2012 after increases in 2010–12. - REER appreciated 16 percent between January 2009 and May 2013 due to higher inflation than trading partners. - Cyclically-adjusted current account surplus declined from 6 percent of GDP in 2008 to 2.9 percent of GDP in 1Q 2013. - Current account broadly in line with fundamentals; REER deemed moderately undervalued (IMF 2013d). - Asset prices: 10Y government bonds fell to just below 7 percent by May 2013 (from average 10 percent in 2009); equities up 120 percent between January 2009 and May 2013. - Concerns on asset quality amid rapid credit growth and volatile oil prices. - Banking sector improved 2009–2012 but trend reversed slightly; external debt increasing; corporate leverage mostly unchanged since 2008. - Exposure: - Historically Russian long-term rates not significantly related to U.S. monetary policy shocks; recent U.S. 25 bps jump led to a contemporaneous 46 bps jump in Russian long-term rates. - Inflows fell moderately in the month following U.S. tapering announcements (z-score of -1.69). - Russia may be exposed if commodity prices are affected by U.S. tapering. - Resilience and policy space: - Financial sector depth and liquidity relatively low. - Not very dependent on foreign funding (2.3 percent of GDP). - Both fiscal and monetary authorities have space to operate if significant outflows materialize. ### South Africa - Background: - Growth slowed to 0.9 percent in 1Q 2013, below potential of 3-3½ percent. - Current account deficit widened to 6.3 percent of GDP in 2012; narrowed moderately in 1Q 2013. - Inflation at top of 3-6 percent target range. - Policy rate at 5 percent with no cuts since July 2012; real rate slightly negative. - Rand vulnerable to foreign investor flows and commodity cycle. - Public debt over 40 percent of GDP. - Fiscal policy appropriately accommodative short run; achieving government debt target may require additional measures. - Capital account mostly open; central bank committed to fully flexible exchange rate. - Capital inflows (up to the May 2013 tapering announcements): - Strong inflows largely led by portfolio flows to local debt market; inclusion in WGBI in 2012. - Portfolio inflows recorded US$14.4 billion in 2010. - Nonresidents hold about 40 percent of government bonds and 30 percent of equities. - Sovereign ratings downgraded by all three major agencies; two maintain negative outlook. - Coping with inflows — Potential vulnerabilities and indicators: - Policy response eclectic: allowing appreciation, accommodative monetary stance, liberalizing outflow controls, building reserves; CFMs considered. - Exports mostly robust until 2011; current account deficit of 6 percent high. - Cyclically-adjusted current account 2–4 percentage points of GDP weaker than implied by fundamentals; REER moderately overvalued (IMF 2013d). - Asset prices rose since 2008 with foreign buying of local government bonds; 10Y yields 7 percent as of May 2013; some reversal since May 2013. - Banking sector mostly stable with satisfactory capital buffers; risks from rapid increase in unsecured lending. - Corporate debt increased but manageable. - Exposure: - South African long-term rates historically somewhat sensitive to U.S. monetary policy shocks. - Recent U.S. 25 bps jump led to contemporaneous 23 bps jump in South African long-term rates. - Inflows fell significantly in the month following U.S. tapering announcements (z-score of -1.87). - Resilience and risks: - Large domestic investors could substitute for withdrawing foreign investors. - Significant outflows could make financing the current account deficit difficult; higher interest rates would squeeze an economy below potential; fiscal space diminished. - Lower global commodity prices also pose a risk. ### Thailand - Background: - Growth rebounded to 6.4 percent in 2012 after 2011 floods, then slowed in H1 2013. - Headline and core inflation declined over 2013. - Policy rate history: lowered to 1¼ percent in April 2009; raised cumulatively 2¼ percent since July 2010; then cut cumulatively 1 percent since November 2011. - Public debt 45.4 percent of GDP at end-2012, but expected to rise. - Fiscal policy needs tightening once recovery entrenched. - Most outward capital transactions and many inward transactions subject to BOT approval ceilings. - Capital inflows (up to the May 2013 tapering announcements): - Gross capital flows to Thailand reached US$40 billion in 2012. - Capital and financial account surplus of US$10 billion in 2012. - Net portfolio inflows 1.2 percent of GDP to bonds; other investments 2.9 percent of GDP. - Gross FDI largest inflow form; net FDI negative since 2011 due to outward FDI. - Coping with inflows — Potential vulnerabilities and indicators: - Policy response included allowing appreciation with interventions, lowering policy rate, relaxing outflow controls; macroprudential measures and CFMs being assessed. - Exports gained market share over last decade despite REER appreciation. - Cyclically-adjusted current account about 0-2 percentage points of GDP stronger than value consistent with fundamentals; REER fairly valued (IMF 2013d). - Current account surplus fell from 8¼ percent of GDP in 2009 to 0.7 percent in 2012 (2 percent cyclically adjusted). - Asset prices: Thai stock index increased by 275 percent between end-2008 and May 2013 (annual average growth of 35 percent); 10Y government bond yield around 3.5 percent; housing prices increased 3–4 percent nationally in 2012, but 50 percent in selected Bangkok areas over last 12 months. - Banking sector sound: declining NPL ratio, improved capital adequacy, strong profitability; early signs of rising systemic risk. - Exposure: - Thai long-term rates historically sensitive to hikes in U.S. long-term rates. - Recent U.S. 25 bps jump led to contemporaneous 20 bps jump in Thai long-term rates. - Inflows fell significantly in the month following U.S. tapering announcements (z-score of -2.37). - Resilience: - Foreign equity holdings high, but government debt mostly domestically financed. - High domestic market capitalization and relatively high turnover support resilience. - In event of significant outflows, higher interest rates can be accommodated due to a relatively small output gap; ample FX reserves can limit exchange rate volatility; banking sector appears sound. ### Turkey - Background: - Growth increased to 3 percent as of 1Q 2013, below potential around 4 percent. - Inflation jumped to 8.8 percent in July 2013, at risk of missing target range (3 to 7 percent). - Monetary policy framework changes caused transmission distortions and higher inflation expectations. - Fiscal policy loose and pro-cyclical; public debt 35 percent of GDP. - External debt increasing and vulnerable to large exchange rate shock. - Capital account mostly open. - Capital inflows (up to the May 2013 tapering announcements): - In 2012 gross inflows US$70 billion (9 percent of GDP), similar to pre-crisis peak. - Inflows financed current account deficit and increased reserves by more than US$30 billion in 2012. - Inflows toward FDI declined; inflows toward debt increased; short-term debt remains predominant financing instrument. - Coping with inflows — Potential vulnerabilities and indicators: - Policy response: monetary policy adopted financial-stability objective and unorthodox tools (reserve requirements, repo facilities, FX interventions, macroprudential measures) to contain lending and avoid exchange rate volatility. - Exports increased somewhat since 2008, but imports rose; REER appreciated by 4 percent on average in 2012. - Cyclically-adjusted current account is 1.5-3 percent of GDP weaker than implied by fundamentals; REER moderately overvalued (IMF 2013d). - Asset prices: 10Y government bonds around 6 percent by May 2013; equity prices increased 16 percent annually since 2009 with strong correction after May 2013. - Credit growth large swings since 2008; picked up over 2013 as of August 2013. - Banking and corporate sectors mostly stable with satisfactory capital buffers; external funding dominated by short-term debt. - Corporate leverage risen but manageable. - Exposure: - Turkish long-term rates historically highly sensitive to U.S. monetary policy shocks. - Recent U.S. 25 bps jump led to contemporaneous 95 bps jump in Turkish long-term rates. - Inflows fell moderately in the month following U.S. tapering announcements (z-score of -1.48). - Resilience and contingency: - Large external financing needs and relatively small markets increase vulnerability to external shocks. - In event of significant outflows, higher interest rates could weigh on an economy growing below potential; exchange rate has room to adjust; FX reserves could support interventions to prevent overshooting; some fiscal policy space remains. *Source: Appendix Table 1 (text provided).* ### 6. The impact of capital flows and their reversals on the financial sector and the real ### _090313a - 6. The impact of capital flows and their reversals on the financial sector and the real ### Nonlinearity and limits of standard models - Capital flows and their reversals can produce nonlinear and asymmetric effects that force the economy out of a corridor of stability into regions where global nonlinearities dominate. - Standard macroeconomic models: - Overlook endogenous aggregate (non-diversifiable) balance-sheet risks of financial institutions and borrowers. - Are routinely solved by local approximation methods (linearization or higher-order approximation) and thus cannot capture global nonlinearities arising from balance-sheet risks. - Are therefore only capable of explaining behavior within corridors of stability; extrapolating them to large distress yields overly benign outcomes. ### New IMF macrofinancial model — purpose and novelty - Purpose: - Provide a framework that endogenizes macrofinancial amplification mechanisms and captures global nonlinear feedback in large distress scenarios. - Illustrative output: patterns and mechanisms are the primary output rather than precise numerical predictions. - Novel features integrated: - Aggregate, non-diversifiable credit risk on banks' loan books, derived endogenously from macroeconomic developments. - Optimal bank balance-sheet and capital-buffer choices under uncertainty, constrained by the risk-bearing capacity of bank capital. - Endogenous interaction of aggregate credit risk, loan portfolio value, bank capital buffers, and relative costs of internal and external equity flows. ### Brief description of model structure - Macroeconomic side: - Real sector: small open economy with production for local consumption and exports. - Distinguishes short-run and long-run elasticity of substitution in expenditure switching; allows for varying degrees of import substitution; includes weights on permanent and current income in aggregate demand. - Financial sector: - Consists of banks extending non-traded bank loans and creating matching liabilities (bank deposits) at loan origination. - Banks are required to hold capital; model distinguishes saving and financing. - Deposits finance consumption, investment, imports, or nonresidents’ purchases of local exports. - Risk and capital: - Individual loans and loan portfolios are risky; debtors can default depending on income and wealth evolution. - Loan risks share a common systemic component driven by aggregate macroeconomic developments; diversification cannot eliminate aggregate risk. - Banks hold capital buffers above regulatory minima because of non-diversifiable risk; buffers vary endogenously and pro-cyclically with loan riskiness. - Feedback and asymmetry: - Adverse shocks raise impaired loans and cause write-downs; thinner capital buffers force asset liquidation and higher bank lending costs, triggering fire sales and credit crunches. - Feedback is nonlinear and asymmetric: marginal upside of bank finance is limited in very good times (banks as sellers of call options) while downside in bad times is practically unlimited, leading to rapidly growing losses. ### Calibration, limitations, and interpretation - Real sector calibration: broadly matches characteristics of emerging Central and Eastern European countries. - Parametrization of macro-financial linkages is challenging: - Global nonlinear models require vastly more simulations and parameter combinations to achieve confidence comparable to linear business cycle models. - Limited data on periods of nonlinear financial distress constrains precise calibration. - Quantitative predictions from nonlinear models should not be taken literally; focus is on coherent analysis of mechanisms in tail-risk events and guidance for robust policy design. ### Design of simulation experiments - Shock structure: - Phase I: initial period of three years of cheap foreign (nonresident) finance and capital inflows. - Phase II: sudden and unexpected reversal of capital flows, followed by gradual return to normal. - Simulations are stress scenarios: low-probability but plausible with large adverse impact; not baseline forecasts. - Specific parameter shocks: - Phase I: cost of foreign financing falls by 200 bps. - Phase II: cost of foreign financing increases initially by 300 bps (thus overshooting the normal level by 100 bps), then converges back to normal within about five years. - Behavioral assumption: - During Phase I, banks and borrowers behave myopically and do not internalize reversal risk, creating an externality and underpriced bank lending risk. - Scenarios: - Two economies: “resilient” with 5 percent share of foreign currency (FX) loans, and “vulnerable” with 50 percent share of FX loans. - For each economy, two simulations: (i) linearized model; (ii) global nonlinear solution. - Throughout simulations, policy remains passive; no pro-active countercyclical macroprudential policies (e.g., capital surcharges) are applied. - Additional model features: - Share of FX loans is parametric; currency choice is not endogenous. - Reasons for FX lending (demand- and supply-side) are noted but not endogenized. ### Phase I: dynamics under capital inflows - Effects observed during three-year low-cost foreign finance period: - Exchange rate appreciation and rising asset prices lead to faster credit growth. - Households and firms increase demand for local goods, imports, and assets (housing, productive capital, stocks) — real activity booms. - Import demand is particularly strong, current account deteriorates, net foreign liabilities increase. - FX lending magnifies the upturn: - Appreciation improves borrowing capacity by reducing loan-to-value or debt-to-income ratios. - Risk of future defaults given a large depreciation grows but is not fully internalized. - Bank capital adequacy: - Capital adequacy ratios decline over time as banks hold thinner capital buffers perceived as safer lending. - This creates pro-cyclical capital requirement problems. - Comparison of model solutions: - Linearized simulations over-predict the upturn in both economies relative to the nonlinear model. - Capital adequacy ratios are not reported for linearized simulations since credit risk lacks first-order dynamics in that formulation. ### Phase II: reversal and amplified downturn - Immediate macro adjustments: - Unanticipated rise in foreign financing costs and capital flow reversal reduce sustainable debt; current account adjusts via lower consumption and investment and improved exports from depreciation. - Credit, income, and demand effects: - Access to bank credit deteriorates rapidly due to currency depreciation and asset price falls. - Currency depreciation lowers real household income; wage pass-through is assumed sluggish, undermining aggregate demand. - Vulnerability of high-FX economy: - Valuation effects magnify impact: outstanding bank credit measured in local currency rises as depreciation valuation effects dominate actual lending volume decline. - Nonlinear real-financial feedbacks sharply increase non-performing loans. - In the vulnerable (high FX) economy, capital buffers drop by as much as 1 percentage point; effect negligible in resilient (low FX) economy. - Banking responses and vicious cycle: - Banks write off losses, capital buffers deteriorate, and banks cut back lending to recapitalize. - Banks raise lending spreads and tighten lending conditions, effectively rationing credit; raising spreads is key to rebuild capital via increased profit margins. - Reduced credit depresses demand, depresses asset prices, depreciates currency further, increases non-performing loans, and fuels the feedback loop. - Linear vs nonlinear outcomes: - Conventional linearized models overestimate upside during booms and greatly underestimate downside during reversals; nonlinear simulations reveal much larger adverse effects in vulnerable economies. ### Simulation outcomes and figures - The chapter reports model-based simulation experiments comparing linearized and nonlinear outcomes for low (5 percent) and high (50 percent) FX loan shares across variables such as: - Cost of foreign funds (percent PA) - Real GDP - Domestic demand - Lending spread (percent PA) - Lending rates (percent PA) - Policy rate (percent PA) - Inflation (percent PA) - CAR (capital adequacy ratio) - Real bank loans - Trade balance to GDP - Labour income - Nominal exchange rate - Legend clarifies line styles for: - Linearized simulation, low share (5 percent) of foreign-currency loans. - Nonlinear simulation, low share (5 percent) of foreign-currency loans. - Linearized simulation, high share (50 percent) of foreign-currency loans. - Nonlinear simulation, high share (50 percent) of foreign-currency loans. ### Effects of UMP on bond and equity flows and prices (overview) - Empirical approach uses multiple techniques to assess UMP effects on mutual fund flows and asset prices: - Factor analysis of global, regional, and country factors driving weekly mutual fund flows. - Regressions linking weekly flows to purchases by FED, BoE, ECB, BoJ and to UMP announcements. - Regressions of how weekly portfolio flows affect asset prices. - Event studies on daily asset prices to assess forward guidance effects independent of asset purchase announcements. - Key empirical findings: - Mutual fund flows respond more to actual UMP bond purchases than to UMP announcements; announcements alone provide only part of the impact. - Patterns of flows vary by UMP program and market conditions: - When the Fed used MBS and Treasury purchases to stabilize markets, money initially flowed out from global markets and then into advanced markets. - U.S. LSAP3, undertaken in more stable market conditions, initially resulted in synchronized flows into emerging markets bonds and non-U.S. equity funds. - The Fed’s May 22 tapering announcement is associated with generalized repricing of risk and a notable increase in the level of correlation of flows—especially EM bond flows—as U.S. LSAP3 and Japanese LSAP flows were partly reversed. - Mutual fund flows significantly affect asset prices, especially equities. - Forward guidance announcements have a separate and strong surprise effect on stock prices and foreign currencies, distinct from asset purchase announcements. - Three general conclusions: - Different forms of UMP operate through different mechanisms; asset purchases matter for flows and asset prices in addition to announcement effects. - Impact of UMP varied with market conditions; the boost to domestic markets per dollar of asset purchases likely falls as leakage to the rest of the world rises. - The large global component in EM bond flows implies this market may be an important potential source of global risk. ### Empirical methods — factor analysis brief - A Bayesian dynamic latent factor model was used on a 42-country sample divided into nine regions to estimate: - A global dynamic factor common to all aggregates, regions, and countries. - Nine regional dynamic factors common across aggregates within each region. - 42 country factors to capture comovements in net flows of bonds and equities within each country. - Asset-market-specific components capturing idiosyncratic dynamics. - By design, these dynamic factors capture intertemporal cross-correlations among observable variables and assess the role of global risk-on/risk-off movements and time-varying share of volatility attributable to global factors. *Prepared by Jaromir Benes (RES); and by Tamim Bayoumi, Qianying Chen, Cristina Costantinescu, and Silvia Sgherri (SPR) as indicated in the chapter.* ### 34. The study relies on the EPFR Global dataset. The EPFR Global database contains weekly ### _090313a - 34. The study relies on the EPFR Global dataset. The EPFR Global database contains weekly ### EPFR dataset and sample - EPFR Global contains weekly portfolio investment (net) flows by more than 14,000 (mutual and ETF) equity funds and more than 7,000 (mutual and ETF) bond funds, with US$8 trillion of capital under management. - EPFR coverage represents only less than 20 percent of the market capitalization in equity and in bonds for most countries, generally with a lower proportion for bonds compared to equities. - The study uses data for 25 AEs and 17 EMEs, over the period from January 1, 2007 to August 26, 2013. - EPFR data by country were adjusted for the increasing coverage of funds over time; in aggregate this adjustment almost halves the amount observed bond and equity flows by the end of the sample compared to the raw data, with very similar rate of shrinkage for advanced and emerging markets. - A key strength: high (weekly) frequency and broad geographic coverage for both AEs and EMEs. ### Global factor, synchronization, and major episodes - Common dynamics across advanced and emerging markets point to synchronized outflows following the Fed’s tapering announcement on May 22, 2013. - The global component in equity and bond flows reflects global risk-on/risk-off movements and global liquidity conditions with notable episodes: - Persistent and synchronized outflows from the beginning of the crisis (August 2007) until mid-2009. - Significant synchronization of outflows between August 2011 to February 2012 (peak of the euro area debt crisis). - A record large global outflow following the Fed’s tapering announcement on May 22, 2013. - Equity markets showed more coherence than bond markets in the early crisis: - Global factor explained 30–80 percent of equity flows versus 10–20 percent of bond flows early in the crisis. - After U.S. QE1 (November 2008), pattern switched: global factor has explained a surprisingly consistent 80 percent of the variation in EMEs bond flows since then. - Proportion for advanced economies is lower and more variable (e.g., rose in summer 2011). ### Variance explained and shifts after May 22, 2013 - Since May 22, the role of global factors in explaining variations of portfolio flows appears to have jumped, indicating a generalized repricing of risk. - The increase in market volatility following Fed tapering announcements is primarily driven by push factors common to EMEs and non-UMP AEs rather than idiosyncratic EME weakness. - Table 2 (referenced) indicates the share of actual flows due to the estimated global factor is virtually 100 percent for bond flows over May 22 to August 26, 2013. ### Quantified flow movements around May 22, 2013 - Following the tapering announcement (May 22 to August 26, 2013): - EMEs bond funds: estimated outflows of US$10 billion (approximately 0.2 percent of EMEs bond market size). - AEs bond funds: estimated outflows of US$11 billion (approx. 0.1 percent of AE bond market size), with US$7.6 billion representing U.S. bond funds alone. - Equity funds: outflows from EMEs of circa US$20 billion, of which US$16 billion flowed into U.S. equity funds and the remainder into Japanese and European equity funds. - Note: mutual fund flows represent only less than 20 percent of overall portfolio flows into the economy; EPFR covers a smaller proportion for bond markets, suggesting portfolio switches implied by EPFR may understate total portfolio rebalancing. ### Empirical approach: announcements, purchases, and conditioning variables - Weekly flows (ending Wednesdays) into bond and equity mutual funds from EPFR were adjusted for coverage and used alongside central bank asset purchases from S4 balance sheets. - Sample divided into distinct central banks’ operations: Fed (LSAP1A, LSAP1B, LSAP2, Operation Twist, LSAP3, tapering speech on May 22, 2013), BoE (LSAP1, LSAP2, FLS), ECB (OMT and “whatever it takes” speech, conventional rate cut on May 2, 2013; LTRO and SMP purchases), BoJ (LSAP, QQME, pre-LSAP asset purchases). - Baseline regression specification: - Flow_it = α_i + β_i Announce_t + γ_i Purchase_t + η_i conditioning variables_t + ε_it - Announce_t are 0/1 dummies for weeks with major UMP announcements. - Purchase_t is the amount of securities purchased by a central bank in week t. - Conditioning variables include VIX, oil prices, and non-oil prices. - Two robustness variations to test simultaneity bias: - Lagged RHS variables to guard against purchases being affected by contemporaneous flows (produced extremely similar results). - Dummies for active program periods (replacing weekly purchase amounts); results similar in coefficient significance though sizes varied. ### Data validation and limitations - Correlations between EPFR quarterly-aggregated flows and balance of payments flows since 2008Q1: - For emerging markets, correlations for both bond and equity markets are generally around 0.5. - Correlations are generally fairly high for advanced market equity flows, but low for some AE bond flows and for money centers (U.S., U.K., Switzerland). - Interpretive caveats: - Many AEs have deeper financial markets and domestic bonds used as collateral, which can distort BOP bond flow data. - EPFR data reflect flows to funds regardless of investor domicile and can capture domestic investor behavior (e.g., U.S. EPFR flows representing 64 percent of foreign equity inflows and 19 percent of foreign bond inflows). - Conclusion: adjusted EPFR flows are a reasonable proxy for gross international bond flows to EMEs and equity flows to most economies, and reflect useful information about market behavior. ### Regression results and program-specific findings - Results summarize the significance and direction of announcement and purchase effects across UMP programs: - For bond flows, U.S. LSAP1A, U.S. tapering, and Japan LSAP: both announcements and purchases had many significant coefficients. - Purchases were the driving force for U.S. LSAP1B, U.S. LSAP2, U.S. LSAP3, U.K. FLS and euro area LTRO liquidity programs. - Number of significant coefficients generally higher for bond flows than equity flows. - Overall impacts were computed as the sum of cumulative impacts from announcements and purchases significant at the ten percent level; EPFR flows converted to BOP-equivalent dollars using EPFR/BOP ratios from Table 2, and impacts also reported as ratios to 2012 GDP. ### Key high-level conclusions on UMP effects - Three emphasized conclusions where UMP had statistically significant effects: 1. UMP-linked capital flows play a significant role in overall bond flows—often representing a good proportion of flows into U.S. and other country mutual funds. 2. Flows are often dominated by the impact of actual purchases, indicating the importance of accounting for both announcement effects and the impact of purchases. 3. Analysis should consider individual programs rather than UMP as a whole, as coefficients vary significantly by program and can change sign. - Program-specific summary observations: - U.S. LSAP1 (MBS purchases and initial Treasury purchases): net impact led to outflows from emerging market bond funds into advanced country funds in two phases—generalized bond outflows during LSAP1A, followed by bond inflows mainly to advanced markets during LSAP1B. Global equity impact smaller but similarly involved outflows from EMEs to advanced economies. - U.S. LSAP2: impact mainly within advanced economies, with outflows from U.S. bond funds to U.S. equities and other advanced bond and equity markets. - U.S. LSAP3: led to inflows into bonds of emerging markets and some riskier advanced economies as well as into non-U.S. equities. - Impact on domestic markets per dollar of purchases appears to have fallen over time: LSAP1B boosted private sector inflows to U.S. bond markets, while LSAP2 and LSAP3 were associated with outflows to the rest of the world, diminishing per-dollar domestic impact. - Fed tapering announcement (May 22) induced large generalized outflows from both emerging and advanced bond and equity funds, reversing earlier inflows linked to U.S. LSAP3. - BoJ policies supported domestic bond (LSAP) and equity (QQME) markets; Japan LSAP associated with major inflows into Asian bond funds. - ECB and BoE programs generally had limited effects on global bond flows, with the exception of U.K. LSAP1 which supported U.K. bond and global bond and equity flows. - Net impact of recent policies on emerging markets remains uncertain: inflows from U.S. LSAP3 were largely reversed by tapering, but inflows associated with Japan LSAP do not appear to have been reversed. *Source: Excerpt from GLOBAL IMPACT AND CHALLENGES OF UNCONVENTIONAL MONETARY POLICIES—BACKGROUND PAPER (pages 29–35) — INTERNATIONAL MONETARY FUND* ### 52. This section investigates whether these flows affect asset prices? To answer this ### _090313a - 52. This section investigates whether these flows affect asset prices? To answer this ### Impact of mutual fund flows on asset prices - Regression framework: ∆Asset price_it = α_i + β_i Flow_t + γ_i ∆Asset price_it-1 + η_i conditioning variables_t + ε_it. Asset Prices reflect either local bond yields or equity yields and flows the corresponding EPFR flows. - Findings: - Greater inflows into funds are associated with higher asset prices. - Equity results: coefficients universally correctly signed and generally highly significant; evidence clearest for equities where retail investors play a more important role. - Bond results: about a third of coefficients are significant at the 10 percent level; coefficients overwhelmingly correctly signed. - Only five of the 33 coefficients were incorrectly signed — an event which would occur randomly with a probability of less than current 0.01 percent. - Conclusion: strong evidence that flows into bond and equity funds are associated with changes in asset prices, though this link does not establish causality. ### UMP-related fund flows and asset prices - Experiments: - Added weekly asset flows associated with UMP policies (defined as weekly series summing the impact of program announcements and purchases with significant coefficients) to the base regression. - Also ran regressions using only UMP-related flows. - Findings: - UMP-related flows behaved similarly to overall EPFR flows; coefficients almost always insignificant and show little pattern when added to base regression. - When using only UMP-related flows, results mirror EPFR data: small number of incorrectly signed coefficients (three for equities and nine for bonds) — these would occur randomly less than one percent of the time. - Interpretation: fund flows associated with UMP policies support asset prices and are typical relative to broader fund flows. ### Effects on bilateral exchange rates - Focus: dollar bilateral rate for 13 emerging markets. - Findings: - Both overall and UMP-related fund flows show a link with movements in bilateral dollar exchange rates. - Relatively few incorrectly signed coefficients in these regressions, suggesting a detectable relationship between fund flows and bilateral exchange rate movements. ### Testing forward guidance via event study on daily asset prices - Data and scope: - Two-daily returns from January 1, 2003 until June 26, 2013. - Three asset markets: bond, equity and foreign exchange. - Countries: 23 advanced economies (AEs) and 11 emerging markets (EMs); selected subgroups analyzed. - Monetary surprises measured by changes in 1-year-ahead futures 3-month LIBOR rate (change in futures on a policy announcement day reflects the surprise). - Methodology: - Decomposed impact of UMP announcements into the surprise element and the impact of the surprise on asset prices. - Tested forward guidance in isolation from asset purchase announcements to compare effects per unit of monetary surprise. - Two-stage approach to account for international and domestic financial linkages and differing time zones: first, simultaneous transmission across bond yields, equity prices, and exchange rates within and between four major financial markets; second, use underlying shocks as inputs into similar systems for each small open economy in turn. - Two-day event window used to accommodate differing time zones. - Example result: - U.S. LSAP1 (purchase of MBS and Agency bonds during the first phase) surprise lowered long-term bond rates in Brazil and Mexico by over 9 bps. - Findings on instrument-specific effects: - Forward guidance: - Effective in boosting the economy. - Primary impact: increase global equity prices and appreciate local currencies, with little impact on bond yields. - In the U.S., forward guidance announcements associated with a larger reduction in expectations of future interest rates than bond purchase announcements (see Table 15 referenced). - In Japan, forward guidance appears to have boosted equity prices more than the QQME asset purchase program. - Asset purchases (quantitative easing): - Particularly in the U.S., worked by lowering global bond yields (Table 12). - Associated with some fall in equity prices (Table 13). - Associated with depreciation in local currencies against the dollar (Table 14, where a positive sign implies a depreciation). - ECB case: - Major spillover effects appear to have come from Draghi’s “whatever it takes” speech (pre-announcement of OMT) rather than from later set-up of OMT. - Tapering announcements: - Fed announcements on tapering led to a major increase in the expected level of the 3-month Libor rate in one year (market “surprise”), helping explain repricing of risks, rising domestic interest rates and weakening equity prices in most emerging markets. ### Conclusions and policy-relevant implications - Different forms of UMP operate through different mechanisms: - Forward guidance works conventionally by immediately affecting equity markets via trader revisions to expected future short rates. - Asset purchases operate through capital flows and portfolio balance effects, significantly affecting bond markets. - Measurement caveats: - Impact of UMP depends on market conditions and the specific instrument used; policies in times of market turmoil have different effects than in tranquil times. - The May 22 tapering announcement produced notable surprise and market effects. - It is difficult to summarize UMP by a single metric (e.g., size of purchases or length of guidance). - U.S.-specific observation: - Inflows into domestic markets per dollar of central bank asset purchases appear to have fallen over time: U.S. LSAP1B associated with private sector inflows into U.S. bond markets, while U.S. LSAP2 and LSAP3 were in particular associated with outflows to the rest of the world. *GLOBAL IMPACT AND CHALLENGES OF UNCONVENTIONAL MONETARY POLICIES—BACKGROUND PAPER* ### 65. Finally, the results also suggest that most bond flows to emerging markets continue to ### _090313a - 65. Finally, the results also suggest that most bond flows to emerging markets continue to ### Key finding - "Most bond flows to emerging markets continue to be driven by a single global factor." - This pattern "seems to be less true of equity flows or of bond market flows to advanced economies." - Implication stated in the source: "This suggests that bond flows to emerging markets may be a particularly important source for market risk." ### Quantitative evidence (selected entries from the source tables) - Table 2 (Impact of the Global Factor on Bond and Equity Funds Flows over May 22 to August 26): - Emerging Markets — Bonds: Actual Flows 9873; Estimated flows due to global factor 9818; As percent of bond market size 0.19. - Emerging Markets — Equities: Actual Flows 20389; Estimated flows due to global factor 20361; As percent of equity market size 0.21. - Asia (selected countries) — Bonds: - Asia: Actual Flows 2870; Estimated flows due to global factor 2862; As percent of bond market size 0.11. - China: Actual Flows -105; Estimated flows due to global factor -104; (equity column shows 6092 and 6092). - India: Actual Flows -35; Estimated flows due to global factor -34; (equity column shows 1565 and 1565). - Latin America — Bonds: Actual Flows 4343; Estimated flows due to global factor 4337; As percent of bond market size 0.26. - o/w BRICS — Bonds: 3129; Estimated flows due to global factor 3118; Equities: 13831; Estimated flows due to global factor 13831. - Advanced Economies — Bonds: Actual Flows 10737; Estimated flows due to global factor 10770 (table entries show "10737- 10770" with adjacent strings for other columns). - Total Outflows (Table 2): 21020- 20962-0.05 (bonds) — 22643- 22616-0.05 (equities). - Net (Table 2): -20610- 20588-0.05... (table shows "Ne t- 20610- 20588-0.05850267550.02"). - Table 3 (Correlation and Relative Size of Adjusted EPFR Data and BoP Data (2008–12)) — selected correlations and EPFR/BoP ratios: - Emerging Markets: Correlation 0.44; EPFR as a ratio of BoP Data 0.08 (bond flows column) — Correlation 0.38; EPFR as a ratio of BoP Data 0.22 (equity flows column). - BRICS (o/w BRICS): Correlation 0.45; EPFR as a ratio of BoP Data 0.08 (bond flows) — Correlation 0.67; EPFR as a ratio of BoP Data 0.40 (equity flows). - Selected country examples: - China: Bond correlation 0.10; EPFR/BoP 0.02; Equity correlation 0.73; EPFR/BoP 0.46. - India: Bond correlation 0.15; EPFR/BoP 0.01; Equity correlation 0.79; EPFR/BoP 0.27. - Russia: Bond correlation 0.78; EPFR/BoP 0.16; Equity correlation 0.70; EPFR/BoP 0.54. - United States: Bond correlation (reported as) 1/0.09; EPFR as ratio 0.19 (equity column shows -0.03 and 0.64 in adjacent entries). - Table 6 and Table 7 (Impact of UMP on Bond and Equity Funds Flows) — headline net impacts reported: - Table 6 (Impact of UMP on Bond Flows) — Net impact (bottom line): -8127263355986868235594214397- 342390 0521268- 192547611274681- 22003516795628502056903893631814448072.15 (table shows a long composite net line). - Table 7 (Impact of UMP on Equity Funds Flows) — Net impact (bottom line): -1658681270742795533433266702- 62076 0239187- 198755891-82505- 103830 -10009001343147565631.21 (table shows a long composite net line). - Table 8 and Table 9 (Fund flows → asset prices): - Table 8 (Impact of Bond Fund Flows on Bond Yields) reports country R2, number of observations, and coefficients for Oil prices, Non-energy commodity prices, Bond funds flows, Lagged dependent variable, and VIX. Example entries: - South Africa (ZAF): R2 0.018; Number of observations 0.126; coefficient entries include 0.0020.001#-0.0060.06285 (table formatting preserved as in source). - China (CHN): R2 0.078; Number of observations 0.249; coefficient entries -0.0010.004#-0.0010.16285. - Table 9 (Impact of Equity Funds Flows on Equity Prices) reports R2 and coefficient patterns. Example entries: - South Africa (ZAF): R2 0.603#; coefficients include -0.118-0.0860.148#0.1900.47285. - India (IND): R2 0.738#; coefficients -0.141-0.0770.115#0.0370.30285. (Note: tables in the source include extensive country-by-country regression coefficients, significance indications, and event-response panels. The excerpted tables preserve the precise numeric strings as presented in the source.) ### Analysis and policy implications (as presented in the source) - Because bond flows to emerging markets are largely driven by a single global factor, they can concentrate and transmit global market risk into emerging markets. - Equity flows and bond flows to advanced economies appear less dominated by the same single global factor, implying different vulnerability profiles across asset classes and country groups. - Understanding the global-factor dominance in EM bond flows is important for assessing spillovers from unconventional monetary policies (UMP) in advanced economies to emerging markets. - The source presents detailed empirical evidence (EPFR-based flows, comparisons with BoP data, regressions, and event-study responses) to quantify how UMP announcements and related flows affect bond yields, equity prices, FX rates, and forward LIBOR responses across country groups. *Source: EPFR database; IMF 2013 April GFSR; and staff calculations.* ### Appendix Table 1. ### Appendix Table 1. ### Fiscal policy room (table structure and examples) - Table lists selected indicators to assess fiscal policy room, including: - Total (percent of GDP) - Short Term Debt (percent of total) - Primary Balance (percent of GDP) adjustment in 2013-20 to achieve debt target in 2030 9/ - Credit rating (WEO projection of 2013 CPI inflation - inflation target 10/) - Output gap of 2013 (percent of GDP) 11/ - Example country entries (values reproduced exactly as in the source): - Australia: 0.54; 35.4; - 4.04; AAA; 84.60; 84.65; ...; 23.85; ...; ...; 0.7; - 0.88; - 0.05; 5 ~ 15; - 1.09; 1.39; 5.6; 1.8 - Brazil: 0.53; 24.1; - 2.15; BBB; 54.60; 67.88; 101.32; 15.86; 14.06; 10.94; - 4.2; - 0.04; 0.00; 10 ~ 15; - 0.96; 14.29; 10.4; 3.6 - Canada: 0.56; 23.3; - 0.46; AAA; 110.69; 61.58; ...; 23.44; ...; ...; 2.3; - 1.73; - 1.24; 5 ~ 15; - 1.99; 1.14; 4.7; 0.6 - China: -0.01; 6.8; - 1.97; AA-; 44.94; 164.44; 144.98; 3.19; 8.96; 72.79; 0.0; ...; - 5.16; -10 ~ -5; 2.82; 19.73; 6.3; 1.0 - India: -0.03; 12.2; - 1.67; BBB-; 68.60; 54.63; ...; 7.00; 20.90; 27.13; 6.7; ...; - 0.91; - 5 ~ 5; - 7.63; 5.42; 6.9; 3.1 ### Resilience and sensitivity to external shocks - Table columns capture resilience and sensitivity indicators, including: - Sensitivity to a change in US long term bond yield 1/ - Change in EM bond yields on US tapering 2/ - EM inflows after US tapering (z-score) 3/ - Credit ratings 4/ - Market capitalization (percent of GDP) 5/ - Turn-over ratios (percent market capitalization) 5/ - Foreign portfolio equity liabilities (percent of GDP) 7/ - Size of pension, mutual funds and insurance companies (percent of external debt) 6/ - Example country entries (values reproduced exactly as in the source): - Indonesia: 1.10; 36.4; - 2.05; BB+; 45.19; 23.30; 5.53; 10.55; 28.59; 17.25; 1.7; 1.41; - 0.11; - 5 ~ 5; - 0.24; 5.92; 11.9; 2.0 - Korea: 0.26; 21.0; - 1.42; AA-; 104.50; 139.22; 85.02; 32.17; 36.60; 26.16; - 4.6; - 2.30; - 1.40; - 8 ~ - 2; 3.63; 6.11; 8.8; 0.7 - Mexico: 0.30; 57.8; - 1.65; BBB+; 44.60; 25.31; 37.73; 14.59; 29.34; 16.92; 1.2; - 0.14; - 0.44; - 5 ~ 5; - 1.21; 4.90; 10.1; 2.4 - Poland: 0.09; 42.8; - 1.95; A; 36.29; 42.56; 19.39; 7.57; 74.35; 16.99; 2.4; - 1.92; - 1.73; - 5 ~ 5; - 0.14; 5.09; 8.6; 5.0 - Russia: -0.26; 45.9; - 1.69; BBB+; 43.41; 87.64; ...; 9.58; 28.61; 9.10; - 0.5; 0.87; - 0.03; - 5 ~ 5; 7.25; 13.13; 12.4; 6.4 - South Africa: 0.19; 23.2; - 1.87; BBB+; 159.33; 54.93; 267.33; 25.85; 35.78; 16.54; 2.3; - 0.15; - 1.17; 10 ~ 20; - 3.05; 4.39; 7.3; 4.4 - Thailand: 0.62; 20.1; - 2.37; BBB+; 104.77; 70.44; 31.77; 20.12; 45.59; 20.57; 1.2; - 0.60; - 0.09; - 5 ~ 5; 1.40; 7.58; 10.5; 2.7 - Turkey: 0.66; 95.1; - 1.48; BBB-; 39.12; 136.51; 4.93; 8.95; 42.70; 36.89; - 1.4; - 0.36; - 0.43; 10 ~ 20; - 5.38; 6.62; 12.0; 2.7 ### Domestic market conditions and dependence on foreign funding (notes and measures) - Domestic market condition indicators in the table include: - Market capitalization (percent of GDP) 5/ - Turn-over ratios (percent market capitalization) 5/ - FX reserves (months of imports) 14/ - Bank capital ratio (share of total assets) 5/ - NPL to Total Gross Loans 15/ - Dependence on foreign funding indicators include: - Foreign portfolio equity liabilities (percent of GDP) 7/ - Size of domestic institutional investors that could provide funding (percent of external debt) 6/ - Footnotes and data details: - 7/ Foreign ownership of domestic portfolio equities as a percent of GDP. Based on 2012 data, except for Indonesia, South Africa and Thailand, which use data from 2011. - 11/ Source: WEO July 2013. - 12/ Differences between REER and those consistent with medium-term fundamentals and desirable policies. Assessments prepared in May 2013. Source: 2013 External Stability Report, IMF 2013d. - 13/ Based on 2012 data. Source: WEO July 2013. - 14/ Number of months of imports covered by FX reserves. Based on 2012 data. Source: Staff estimates and IFS. - 15/ Based on 2012 data from the Financial Soundness Indicators (FSI), except for Thailand, which is based on 2012 data from the Bank of Thailand. - 6/ Market size of domestic investors that could provide funding in case of a sudden stop. Data dates vary by country (see source table notes). ### Monetary policy room - Monetary policy room indicators included in the table: - Room for exchange rate adjustment (percent) 12/ - Trade balance (percent of GDP) 13/ - External Debt (2012) 8/ - FX reserves (months of imports) 14/ - Bank capital ratio (share of total assets) 5/ - NPL to Total Gross Loans 15/ - Measurement notes: - 10/ Measure calculated as the differences between WEO projected CPI inflation in 2013 and inflation target (upper bound used if central bank targets at a range). For South Africa, core inflation target is used. Source: WEO July 2013. - 1/ Based on regression of non-UMP country 10-year bond yields on U.S. 10-year bond yields. The regression was run in changes over two-day intervals immediately following U.S. FOMC announcements between January 1, 2003 and May 20, 2013. - 3/ Z-scores of average flows (equity and bond) in the month after tapering announcement on 5/22/2013 relative to flows since 2009. A z-score represents the deviation from the long-term average expressed in the number of standard deviations. Source: EPFR data and staff calculation. *Source: WEO July 2013 and IFS.* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2013/_090313a.pdf_