## 1. Market Reaction to Domestic APP Announcements

## Source details

**Canonical URL:** [1. Market Reaction to Domestic APP Announcements](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020292-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020292-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020292-print-pdf.pdf.json)

---

### Introduction and summary of findings
- The COVID-19 pandemic caused a synchronized and deep global recession and, since March 2020, nearly all major central banks responded forcefully including by lowering rates, providing liquidity to the financial system, and adopting or expanding unconventional monetary policy measures.
- Nearly 20 emerging market (EM) central banks resorted to asset purchase programs (APPs).
- For the purpose of this paper, asset purchases refer to the purchase of private and public sector bonds with the intention of providing monetary accommodation, broadly easing financial conditions, reducing market volatility, or possibly engaging in monetary financing via primary market purchases.
- The study examines effects of APP announcements by EM central banks on domestic financial markets across fixed income, foreign exchange (FX), and equity markets using:
  - an event study analysis at various frequencies; and
  - a regression analysis using local projections methodology that controls for global factors (Federal Reserve measures, swings in global risk appetite as proxied by the VIX, and domestic policy rate changes) and captures persistence of effects.

### Key empirical findings
- Impact on long-end bond yields:
  - Event study: median bond yields (and term premia) declined by almost 35 (25) basis points (bps) in the few days (and in some cases, on an intra-day basis) following APP announcements.
  - Local projections: the size of the effect ranges from 20 to 60 bps for up to 6 trading days in the aftermath of the announcements.
  - The Federal Reserve announcement of extensive new measures (the March 23 announcement is considered to be the QE announcement date) also helped lower bond yields in EMs.
  - A rise in the VIX had a negative effect, increasing EM yields.
  - Domestic policy rate cuts did not have a significant effect on bond yields after controlling for APP announcements.
- Impact on currencies:
  - APP announcements did not lead to a depreciation or a significant change in EM currencies against the dollar.
  - Possible explanation: several central banks sterilized purchases, which may have alleviated market concerns about exchange rate depreciation pressures.
  - The QE announcement by the Federal Reserve was followed by an appreciation of EM currencies against the US dollar, whereas the increase in the VIX was followed by a depreciation.
  - Domestic policy rate cuts during the period of COVID-19 generated either short-lived depreciations, or less significant effects on currencies.
- Impact on equities:
  - Domestic APP announcements did not have a strong effect on EM equities, although they somewhat helped stock markets rebound.
  - Global factors had a more significant effect on domestic stock markets: the QE announcement by the Federal Reserve supported EM stock markets, whereas a higher VIX put downward pressures.
  - Domestic policy rates appear to have no significant effect on equities.

### Data and methodology (overview)
- Daily event study sample: Chile, Colombia, Hungary, Malaysia, Indonesia, India, South Africa, Poland, Turkey, and the Philippines (over a total of 16 dates).
- Data sources: Bloomberg for bond yields, exchange rates and equity markets; BNP Paribas for term premia calculations using the methodology in Adrian and others (2013).
- Change metrics: bond yields in basis points; exchange rates and equity markets in percentage terms. All series normalized at the announcement date (t=0).
- Intra-day event study: high-frequency data from Bloomberg at tick level (aggregated to minute level); focused on South Africa, India and Poland.
- Regression analysis dependent variables (three separate specifications):
  - (i) Changes in 10-year government bond yields in basis points.
  - (ii) Changes in the nominal exchange rate vis-à-vis the US dollar.
  - (iii) Changes in equity market indexes in percent.
- Key explanatory variables:
  - Domestic APP announcement dummy: takes 1 on APP announcement dates, 0 otherwise.
  - Domestic policy rate cuts: variable measured as percentage points decrease in policy rates.
  - Global factors: either dummy for the QE announcement of the Federal Reserve (assigned 1 on March 23, and 0 otherwise) or the VIX (from Bloomberg).
- Sample for regressions: 13 EMs between the beginning of January and the mid-May at daily frequency using trading days only.
- Methodology:
  - Intra-day event studies at very high frequency to attribute sharp asset price movements to major policy announcements rather than global conditions.
  - Formal regression analysis using local projections methodology (Jorda, 2005) with the extension introduced by Teulings and Zubanov (2014) to capture size and persistence of impacts and control for lead values of explanatory variables.
  - Regression specification includes four lags for explanatory variables, country fixed effects, robust standard errors, and estimation horizon p = 0,...,6 (6 trading days). 1-standard error confidence intervals reported.

### Event study results (selected statistics)
- Daily analysis:
  - Government bond yields: median bond yields declined by almost 35 bps over the next few days after the announcement, compared to an average 50 bps tightening prior to the announcement.
  - ACM term premia: median term premia reduced by almost 20 bps in a week after the announcement, as compared to a 15 bps tightening going into the announcements.
  - Currencies: impact relatively limited; likely reflects that asset purchases were planned to be sterilized.
  - Equity markets: EM equity markets improved a few days post announcements.
- Intra-day analysis (selected EMs):
  - South Africa: yields declined by almost 150 bps intra-day.
  - India and Poland: intra-day yield declines around 10 bps.
  - South African rand appreciated (reversion from a sharp 7 percent depreciation in the last few days); Indian rupee and Polish zloty relatively unchanged.
- Noted wide variation across countries, implying country-specific factors mattered significantly.

### Regression analysis — baseline results and specific coefficient estimates
- Government bond yields (Figure 3(a) baseline):
  - Domestic APP announcements lowered bond yields in EMs, controlling for other factors; effect persistent and significant up to 6 trading days.
  - Point estimate: yield was 35 bps lower than pre-announcement level, with one standard error confidence interval ranging between around 20 to 50 bps even on the 6th day after the announcement.
  - QE announcement by the Federal Reserve: in the 6th day following the announcement, yield was 17 bps lower than pre-announcement level, with the range of the effect 5–30 bps.
  - Domestic policy rate changes did not affect long-end bond yields significantly.
- FX markets (Figure 3(b) baseline):
  - Domestic APP announcements were not followed by significant depreciations in local currencies.
  - QE announcement by the Federal Reserve: positive effect persistent and around 1 percent (appreciation) even in the 6th trading day after the announcement.
  - Domestic policy rate cuts: a 1 percentage point cut led to a depreciation in domestic currencies up to 1.4 percent in the second day after the cut, then statistically insignificant from the fifth day.
- Equity markets (Figure 3(c) baseline):
  - Domestic APP announcements did not have much impact in the very short term but began to show a statistically significant positive effect in later days.
- Appendix Table A.2 — APP dummy coefficient estimates (percentage points for yields; percent for FX and equities):
  - Bond yields (percentage points): Day 0: -0.151*** (0.055); Day 1: -0.221** (0.113); Day 2: -0.235* (0.121); Day 3: -0.157 (0.116); Day 4: -0.064 (0.117); Day 5: -0.339** (0.162); Day 6: -0.340** (0.156).
  - Currencies (percent change vis-à-vis the US dollar): Day 0: -0.297 (0.257); Day 1: -0.244 (0.295); Day 2: -0.399 (0.495); Day 3: -0.224 (0.494); Day 4: -0.044 (0.505); Day 5: 0.546 (0.525); Day 6: 0.681 (0.571).
  - Equities (percent change in stock market indexes): Day 0: -0.337 (0.888); Day 1: -0.139 (1.050); Day 2: 0.394 (1.593); Day 3: 1.118 (1.883); Day 4: 1.748 (1.966); Day 5: 3.257** (1.588); Day 6: 3.067** (1.560).
- Global factors:
  - The Federal Reserve QE announcement had a pronounced positive effect on EM equities; effect persistent and around 7 percent in the 6th day.
  - Controlling for global factors is important: global QE announcements and changes in global risk appetite significantly impacted EM bond yields, currencies, and equities.

### Controlling for the VIX and day fixed effects
- Replacing the Fed QE announcement with the VIX yields broadly similar results:
  - Domestic APP announcements lowered bond yields, did not have much effect on FX markets and stock markets.
  - An increase in the VIX increased bond yields, led to depreciations in EM currencies, and lowered EM equities.
- Including day fixed effects yields similar conclusions: domestic APP announcements had a significant and persistent effect on bond yields, while effects on currencies and equities remain much less clear.

### Robustness checks and identification
- Country-dropping exercise: re-running regressions dropping one of the 13 EMs at a time keeps results similar; findings are not driven by any single country or date.
- Consecutive announcement dates and identification:
  - Applied the Teulings and Zubanov (2014) extension to control for forward announcement values.
  - Using only the first announcement date for countries with multiple dates yields very similar results.
  - A large set of alternative combinations/identifications of announcement dates produces results that stay almost the same.
- Additional checks:
  - Including a linear common day trend in regressions leaves results almost unchanged.

### Examples of sterilization and country practices
- The Polish Central Bank sterilized a significant portion of its asset purchases through the sale of 7-day NBP bills.
- The Chilean central bank provided a transparent breakdown of liquidity drains it performed to counterbalance liquidity injections due to bond purchases and other exceptional policy measures.
- The South African central bank governor noted that persistent and unsterilized increases in bank reserves due to asset purchases would create downward pressure on interest rates which would be inflationary in his view.

### Trade-offs and risks of APPs in emerging markets
- Identified risks:
  - weakening of institutional credibility;
  - intensification of capital outflow pressure;
  - distortion of market dynamics and impaired price discovery;
  - concerns of fiscal dominance and increased balance sheet exposure to long-term debt.
- Implications:
  - Large-scale or open-ended APPs could weaken credibility; central banks with higher institutional quality tended to see greater reductions in local market stress.
  - In economies with weak monetary and fiscal frameworks, APPs can invite concerns about fiscal dominance, raise risk premiums and steepen yield curves.
  - Persistent central bank presence as buyers may impair primary market price discovery, affect collateral availability and policy rate transmission, and risk asset overvaluation.

### Conclusion and directions for further research
- Summary:
  - In the post-March 2020 selloff period, APP announcements by EM central banks lowered bond yields, did not lead to a depreciation of domestic currencies, and did not have much effect on equities.
  - Global factors, including the Federal Reserve QE announcement and increases in global risk aversion, had significant impacts on EM assets.
  - No evidence so far that APP announcements led to heightened investor concerns about fiscal dominance; likely due to the small scale and time-bound nature of programs during the pandemic.
- Policy implications:
  - APPs could be part of the monetary toolkit for central banks constrained by the effective lower bound, with steady inflation expectations, low concerns over capital outflows and FX depreciation, or limited domestic absorption capacity of new bond supply.
  - Policymakers should weigh benefits against potential significant costs, especially for large-scale and open-ended programs.
- Research gaps and suggested analyses:
  - Investigate effectiveness of APPs beyond announcement effects and into implementation periods, controlling for interactions with local currency government net issuance and duration risk.
  - Compare time-bound versus open-ended APPs, secondary versus primary market purchases, and the role of sterilization in driving financial market outcomes.
  - Evaluate trade-offs between stabilizing effects of APPs and risks to central bank credibility, fiscal dominance, monetary anchor loss, and market distortion.

### Appendix — key data elements (excerpt)
- Sample of 13 EMs and APP announcement dates (excerpt):
  - Chile: March 16, April 8
  - Colombia: March 23
  - Hungary: March 24, April 7, April 28
  - India: March 18, March 20, April 23
  - Indonesia: April 1
  - Korea: March 19, March 25, April 9
  - Mexico: April 21
  - Philippines: March 24, April 10
  - Poland: March 17, April 8
  - Romania: March 20
  - South Africa: March 25
  - Thailand: March 19, March 23, April 7
  - Turkey: March 31, April 17

*Source: IMF working paper section "1. Market Reaction to Domestic APP Announcements" from wpiea2020292-print-pdf*

### 1. Market Reaction to Domestic APP Announcements .............................................................9

### 1. Market Reaction to Domestic APP Announcements

### Introduction and summary of findings
- The COVID-19 pandemic caused a synchronized and deep global recession and, since March 2020, nearly all major central banks responded forcefully including by lowering rates, providing liquidity to the financial system, and adopting or expanding unconventional monetary policy measures.
- Nearly 20 emerging market (EM) central banks resorted to asset purchase programs (APPs).
- For the purpose of this paper, asset purchases refer to the purchase of private and public sector bonds with the intention of providing monetary accommodation, broadly easing financial conditions, reducing market volatility, or possibly engaging in monetary financing via primary market purchases.
- The study examines effects of APP announcements by EM central banks on domestic financial markets across multiple asset classes (fixed income, foreign exchange (FX), and equity markets) using:
  - an event study analysis at various frequencies; and
  - a regression analysis using local projections methodology that controls for global factors (Federal Reserve measures, swings in global risk appetite as proxied by the VIX, and domestic policy rate changes) and captures persistence of effects.

### Key empirical findings
- Impact on long-end bond yields:
  - Event study: median bond yields (and term premia) declined by almost 35 (25) basis points (bps) in the few days (and in some cases, on an intra-day basis) following APP announcements.
  - Local projections: the size of the effect ranges from 20 to 60 bps for up to 6 trading days in the aftermath of the announcements.
  - The Federal Reserve announcement of extensive new measures (the March 23 announcement is considered to be the QE announcement date) also helped lower bond yields in EMs.
  - A rise in the VIX had a negative effect, increasing EM yields.
  - Domestic policy rate cuts did not have a significant effect on bond yields after controlling for APP announcements.
- Impact on currencies:
  - APP announcements did not lead to a depreciation or a significant change in EM currencies against the dollar.
  - Possible explanation: several central banks sterilized purchases, which may have alleviated market concerns about exchange rate depreciation pressures.
  - The QE announcement by the Federal Reserve was followed by an appreciation of EM currencies against the US dollar, whereas the increase in the VIX was followed by a depreciation.
  - Domestic policy rate cuts during the period of COVID-19 generated either short-lived depreciations, or less significant effects on currencies.
- Impact on equities:
  - Domestic APP announcements did not have a strong effect on EM equities, although they somewhat helped stock markets rebound.
  - Global factors had a more significant effect on domestic stock markets: the QE announcement by the Federal Reserve supported EM stock markets, whereas a higher VIX put downward pressures.
  - Domestic policy rates appear to have no significant effect on equities.

### Policy implications and risks
- APPs could be an additional tool to achieve a central bank’s monetary policy objectives, diffuse financial market stress and promote stability.
- This is particularly important for central banks with relatively limited conventional policy space, such as when policy rates are close to the effective lower bound (Cavallino and Sandri, 2020), or where domestic bond markets are being severely disrupted.
- Inherent risks of APPs include:
  - weakening of institutional credibility;
  - intensification of capital outflow pressure;
  - distortion of market dynamics; and
  - concerns of fiscal dominance.
- These risks underscore the need for ongoing evaluation of APPs and whether they should be part of the regular policy toolkit.

### Examples of sterilization and country practices (as noted in the source)
- The Polish Central Bank sterilized a significant portion of its asset purchases through the sale of 7-day NBP bills.
- The Chilean central bank has a transparent breakdown of the liquidity drains it performed to counterbalance the liquidity injections due to bond purchases and other exceptional policy measures.
- The South African central bank governor was explicit that persistent and unsterilized increases in bank reserves due to asset purchases would create downward pressure on interest rates which would be inflationary in his view.

### Literature contributions and positioning
- This study is among the first to analytically explore the impact of EM asset purchase programs on domestic financial markets.
- It complements Arslan and others (2020) by:
  - exploring effects of domestic APP announcements and global factors conditional on each other;
  - addressing potential downward bias from consecutive APP announcements in EMs; and
  - extending analysis to stock market implications in the COVID-19 period.
- It contributes to literature on spillovers from US monetary policy and global financial conditions to EMs, documenting that the QE announcement by the Federal Reserve helped EMs improve bond yields, currencies and equities, whereas weaker global risk appetite (proxied by the VIX) had the opposite impact.

### Data and methodology (overview)
- Event study analysis at various frequencies to capture immediate impacts over different windows (including intra-day).
- Regression analysis using local projections methodology to control for global factors (Federal Reserve measures, VIX) and domestic policy rate changes, and to capture persistence of effects.

*Source: IMF working paper section "1. Market Reaction to Domestic APP Announcements" from wpiea2020292-print-pdf*

### Section 5 documents the results from the regression analysis. Section 6 discusses the trade-

### wpiea2020292-print-pdf - Section 5 documents the results from the regression analysis. Section 6 discusses the trade-

### Data
- Daily event study sample: Chile, Colombia, Hungary, Malaysia, Indonesia, India, South Africa, Poland, Turkey, and the Philippines (over a total of 16 dates).
- Data sources: Bloomberg for bond yields, exchange rates and equity markets; BNP Paribas for term premia calculations using the methodology in Adrian and others (2013).
- Change metrics: bond yields in basis points; exchange rates and equity markets in percentage terms. All series normalized at the announcement date (t=0).
- Intra-day event study: high-frequency data from Bloomberg at tick level (aggregated to minute level); focused on South Africa, India and Poland.
- Regression analysis dependent variables (three separate specifications):
  - (i) Changes in 10-year government bond yields in basis points.
  - (ii) Changes in the nominal exchange rate vis-à-vis the US dollar.
  - (iii) Changes in equity market indexes in percent.
- Key explanatory variables:
  - Domestic APP announcement dummy: takes 1 on APP announcement dates, 0 otherwise. Dates assigned from EM central banks’ websites; set of dates illustrated in Table A.1 in the Appendix. APP announcement dates largely consistent with Arslan and others (2020) and Hartley and Rebucci (2020).
  - Domestic policy rate cuts: policy rates obtained from Bloomberg; variable measured as percentage points decrease in policy rates.
  - Global factors: either dummy for the QE announcement of the Federal Reserve (assigned 1 on March 23, and 0 otherwise) or the VIX (from Bloomberg).
- Sample for regressions: 13 EMs between the beginning of January and the mid-May at daily frequency using trading days only. Robustness: results similar if start from February or end at beginning of May; ensured no single country drives results.

### Methodology
- Two complementary approaches:
  1. Intra-day event studies at very high frequency to attribute sharp asset price movements to major policy announcements rather than global conditions.
  2. Formal regression analysis using local projections methodology (Jorda, 2005) with the extension introduced by Teulings and Zubanov (2014) to capture size and persistence of impacts and control for lead values of explanatory variables.
- Regression specification (local projections with controls and lags):
  - Dependent variable: cumulative change in the dependent variable from day t-1 to day t+p (훥훥훥훥
푐푐,푡푡−1→푡푡+푝푝).
  - Explanatory variables: contemporaneous and lagged values (four lags) of APP announcement dummy (푨푨푨푨푨푨 풂풂...), global factor (QE dummy or VIX), and policy rate changes (훥훥퐴퐴푎푎퐺퐺 훥훥푎푎훥훥 푟푟푎푎푡푡푎푎
푐푐). Four lags included for all explanatory variables; results robust to lower/higher number of lags.
  - Country fixed effects (μ
푐푐
) to absorb unobserved country-specific features (e.g., pre-COVID-19 market conditions, institutional capacity, policy credibility, central bank independence).
  - Standard errors are robust; results virtually the same if clustering standard errors at country level.
- Estimation horizon: coefficients plotted for p = 0,...,6 (6 trading days following each action). 1-standard error confidence intervals reported. Appendix Table A.2 provides coefficient estimates, standard errors, and statistical significance at the 10, 5 and 1 percent levels.
- Caveat noted: APP dummy approach ignores size, duration or specific nature of announcements; difficult to apply size-based analysis because features of APP announcements often not clearly defined by EM central banks.

### Results: Event Study Analysis
- Daily analysis (Figure 1):
  - Government bond yields (Figure 1.a): median bond yields declined by almost 35 bps over the next few days after the announcement, compared to an average 50 bps tightening prior to the announcement.
  - ACM term premia (Figure 1.b): median term premia reduced by almost 20 bps in a week after the announcement, as compared to a 15 bps tightening going into the announcements.
  - Currencies (Figure 1.c): impact relatively limited; likely reflects that asset purchases were planned to be sterilized.
  - Equity markets (Figure 1.d): EM equity markets improved a few days post announcements.
- Intra-day analysis (selected EMs: South Africa, India, Poland; Figure 2):
  - South Africa: yields declined by almost 150 bps intra-day.
  - India and Poland: intra-day yield declines relatively modest at around 10 bps.
  - Currencies: mixed and relatively limited impact. South African rand appreciated (potentially reflecting increased investor confidence and a reversion from a sharp 7 percent depreciation in the last few days); Indian rupee and Polish zloty relatively unchanged.
- Noted wide variation across countries, implying country-specific factors (vulnerabilities, buffers, program details) mattered significantly.

### Results: Regression Analysis — Baseline Results
- Baseline regressions control for QE announcement by the Federal Reserve in equation (1). Plots in Figure 3 examine effects on (a) government bond yields, (b) FX markets, (c) stock indexes, for three factors: domestic APP announcements, QE announcement by the Federal Reserve, and domestic policy rate cuts.
- Government bond yields (Figure 3(a)):
  - Domestic APP announcements lowered bond yields in EMs, controlling for other factors.
  - Effect persistent and significant upwards of 6 trading days.
  - Point estimate: yield was 35 bps lower than pre-announcement level, with one standard error confidence interval ranging between around 20 to 50 bps even on the 6th day after the announcement.
  - QE announcement by the Federal Reserve also had a long-lasting effect: in the 6th day following the announcement, yield was 17 bps lower than pre-announcement level, with the range of the effect 5–30 bps.
  - Domestic policy rate changes did not affect long-end bond yields significantly in these regressions.
- FX markets (Figure 3(b)):
  - Domestic APP announcements were not followed by significant depreciations in local currencies.
  - QE announcement by the Federal Reserve helped EM currencies: positive effect persistent and around 1 percent (appreciation) even in the 6th trading day after the announcement.
  - Domestic policy rate cuts: a 1 percentage point cut led to a depreciation in domestic currencies, which reaches up to 1.4 percent in the second day after the cut, but then becomes statistically insignificant from the fifth day.
- Equity markets (Figure 3(c)):
  - Domestic APP announcements did not have much impact in the very short term but started to have somewhat a statistically significant positive effect in later days.
- Robustness checks:
  - Alternative dependent variable using change in yield in percent (difference between two days normalized by initial yield) yields similar persistence and shape for charts (a), (b) and (c). Quantitatively, both APP announcements and the QE announcement by the Federal Reserve are found to lower the yield around 6 percent of its initial level even on the 6th trading day following the events.
  - Using changes vis-à-vis the euro for European EM currencies yields similar FX results.
- Notes:
  - Term premia calculations used Adrian, Crump, and Moench (2013) methodology; effectiveness analysis for advanced economies often focuses on term premia, but this study uses bond yields for EMs for stated reasons in the text.
  - Appendix materials referenced for dates (Table A.1) and coefficient tables (Table A.2).

*Source: wpiea2020292-print-pdf - Section 5 documents the results from the regression analysis. Section 6 discusses the trade- (IMF PDF content provided).*

### 3.3 percent in the 5

### wpiea2020292-print-pdf - 3.3 percent in the 5

### Effects of APP Announcements on EM Financial Assets
- Bond yields:
  - Domestic APP announcements lowered bond yields; effects are significant and persistent.
  - Appendix Table A.2 reports coefficient estimates (percentage points) for the APP dummy:
    - Day 0: -0.151*** (0.055)
    - Day 1: -0.221** (0.113)
    - Day 2: -0.235* (0.121)
    - Day 3: -0.157 (0.116)
    - Day 4: -0.064 (0.117)
    - Day 5: -0.339** (0.162)
    - Day 6: -0.340** (0.156)
- Currencies:
  - Domestic APP announcements did not have much effect on FX markets.
  - Appendix Table A.2 reports coefficient estimates (percent change in value vis-à-vis the US dollar):
    - Day 0: -0.297 (0.257)
    - Day 1: -0.244 (0.295)
    - Day 2: -0.399 (0.495)
    - Day 3: -0.224 (0.494)
    - Day 4: -0.044 (0.505)
    - Day 5: 0.546 (0.525)
    - Day 6: 0.681 (0.571)
- Equities:
  - Domestic APP announcements did not have much effect on equities overall.
  - Appendix Table A.2 reports coefficient estimates (percent change in stock market indexes):
    - Day 0: -0.337 (0.888)
    - Day 1: -0.139 (1.050)
    - Day 2: 0.394 (1.593)
    - Day 3: 1.118 (1.883)
    - Day 4: 1.748 (1.966)
    - Day 5: 3.257** (1.588)
    - Day 6: 3.067** (1.560)
- Global factors:
  - The Federal Reserve QE announcement had a pronounced positive effect on EM equities; the effect was persistent and around 7 percent in the 6th day of the event.
  - Controlling for global factors is important: global QE announcements and changes in global risk appetite significantly impacted EM bond yields, currencies, and equities.
- Domestic policy rate cuts:
  - Domestic policy rate cuts appear to have no effect on equities, and generally did not appear to have much effect when controlling for other variables.

### Controlling for the VIX and Day Fixed Effects
- Replacing the Fed QE announcement with the VIX:
  - Results are broadly similar to those when using the Fed QE announcement as the global factor.
  - Domestic APP announcements lowered bond yields, did not have much effect on FX markets and stock markets.
  - An increase in the VIX increased bond yields, led to depreciations in EM currencies, and lowered EM equities.
- Day fixed effects:
  - Including day fixed effects (accounting for daily shocks common across EMs) yields similar results: domestic APP announcements had a significant and persistent effect on bond yields, while effects on currencies and equities remain much less clear.

### Robustness and Further Checks
- Country-dropping exercise:
  - Re-running regressions dropping one of the 13 EMs at a time keeps results similar; findings are not driven by any single country or date.
- Consecutive announcement dates and identification:
  - Potential bias from closely spaced announcement dates is addressed by applying the Teulings and Zubanov (2014) extension that controls for forward announcement values.
  - Using only the first announcement date for countries with multiple dates yields very similar results.
  - A large set of alternative combinations/identifications of announcement dates produces results that stay almost the same.
- Additional checks:
  - Including a linear common day trend in regressions leaves results almost unchanged.

### Trade-offs and Risks of APPs in Emerging Markets
- Institutional and central bank credibility:
  - Large-scale or open-ended APPs could weaken credibility; central banks with higher institutional quality tended to see greater reductions in local market stress.
  - Increased balance sheet exposure to long-term debt may raise concerns about the ability to raise interest rates when warranted.
- Fiscal dominance:
  - Large-scale/open-ended APPs in economies with weak monetary and fiscal frameworks can invite concerns about fiscal dominance, raising risk premiums and steepening government bond yield curves.
- Capital outflow pressure:
  - Expectations of large-scale APPs may put downward pressure on long-term yields and exchange rates, risking capital outflows, especially during risk-off periods and in countries with weaker fundamentals.
- Market distortion and price discovery:
  - Persistent central bank presence as buyers in local currency bond markets may impair price discovery, especially in primary markets, affect collateral availability and policy rate transmission, and risk overvaluation of assets.

### Conclusion and Directions for Further Research
- Summary of findings:
  - In the post-March 2020 selloff period, APP announcements by EM central banks lowered bond yields, did not lead to a depreciation of domestic currencies, and did not have much effect on equities.
  - Global factors, including the Federal Reserve QE announcement and increases in global risk aversion, had significant impacts on EM assets.
  - There is no evidence so far that APP announcements have led to heightened investor concerns about fiscal dominance; this is likely due to the small scale and time-bound nature of the programs during the pandemic.
- Policy implications:
  - APPs could be part of the monetary toolkit for central banks constrained by the effective lower bound, with steady inflation expectations, low concerns over capital outflows and FX depreciation, or limited domestic absorption capacity of new bond supply.
  - Policymakers should weigh benefits against potential significant costs, especially for large-scale and open-ended programs.
- Research gaps and suggested analyses:
  - Investigate effectiveness of APPs beyond announcement effects and into implementation periods, controlling for interactions with local currency government net issuance and duration risk.
  - Compare time-bound versus open-ended APPs, secondary versus primary market purchases, and the role of sterilization in driving financial market outcomes.
  - Evaluate trade-offs between stabilizing effects of APPs and risks to central bank credibility, fiscal dominance, monetary anchor loss, and market distortion.

### Appendix — Key Data Elements
- Sample of 13 EMs and APP announcement dates (excerpt):
  - Chile: March 16, April 8
  - Colombia: March 23
  - Hungary: March 24, April 7, April 28
  - India: March 18, March 20, April 23
  - Indonesia: April 1
  - Korea: March 19, March 25, April 9
  - Mexico: April 21
  - Philippines: March 24, April 10
  - Poland: March 17, April 8
  - Romania: March 20
  - South Africa: March 25
  - Thailand: March 19, March 23, April 7
  - Turkey: March 31, April 17
- Appendix Table A.2. (Effect of APP announcements) coefficient estimates and standard errors are reported above in the "Effects" subsection.

*Italic: Sources: Bloomberg Financial L.P.; National Authorities; and IMF staff estimates.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020292-print-pdf.pdf_