## Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset

## Source details

**Canonical URL:** [Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026018-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026018-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026018-source-pdf.pdf.json)

---

### Dataset and methodology
- New cross-country high-frequency monetary policy surprises dataset constructed using asset price movements in 30-minute windows around scheduled central bank announcements.
- Sample coverage:
  - 16 countries — 11 advanced economies (AEs) and 5 emerging markets (EMs).
  - Sample end date: May 2025.
  - Total number of monetary policy events in the dataset: 2,528.
  - Total AEs: 1,783; Total EMs: 745.
- Primary instruments and data:
  - OIS rates (1- and 3-month), government bond yields (1-, 2-, and 10-year), equity indices, exchange rates vs USD.
  - Minute-by-minute data sourced from London Stock Exchange Group (LSEG).
  - Instrument substitutions where necessary: interest rate swaps (Brazil, Chile, Mexico), forward rate agreements (Poland, Czech Republic), FX forward-implied rates (Norway).
  - Special handling: Brazil and Chile use many outside-trading-hour announcements (Brazil: 198 out of 216; Chile: 256 out of 257).
- Event window and measurement:
  - Baseline 30-minute window: from 10 minutes before to 20 minutes after scheduled announcement.
  - Pre-announcement rate: median price during the 20-to-10-minute window before the event.
  - Post-announcement rate: median price during the 20-to-30-minute window after the event.
  - Monetary policy surprises: bond surprises = 100 × change in yields over announcement window; stock and exchange-rate surprises = percentage changes over the same window.
- Decomposition:
  - Target surprises: change in short-term interest rate (3-month OIS where available).
  - Path surprises: residual from regression of 2-year yield change on target factor (∆y_{jt+θ}^{2y} = β ∆y_{jt+θ}^{3m} + ε_{jt+θ}^{2y}; path = ε_{jt+θ}^{2y}).
  - Scaling: factors scaled to correspond to a 100-basis-point change in the respective yield.
  - Winsorization at the 1st and 99th percentiles prior to shock construction and regression analysis.
- QE/long-end (risk-premium) factor included in dataset but not analyzed in this paper.

### Validation and stylized facts
- Validation evidence:
  - U.S. and Euro Area estimates correlate with benchmark series near or above 0.9 (Swanson (2021) and Altavilla et al. (2019)).
  - Announcement-window yield volatility markedly higher than on non-announcement days.
    - Volatility of 3-month yields on announcement days is 3.96 times volatility on non-announcement days (average median).
    - Volatility of 2-year yields on announcement days is 4.1 times volatility on non-announcement days (average median).
- Three stylized facts:
  - Target surprises are small on average (means close to zero across countries).
  - Predictability deteriorates during crises/regime shifts; interquartile ranges widen during the Global Financial Crisis, COVID-19 pandemic, and the recent inflation surge—particularly for path shocks.
  - EMs exhibit fatter tails than AEs, especially for path shocks (greater volatility, shallower markets, communication challenges).
- Summary statistics of high-frequency shocks (Table 3 excerpt preserved exactly):
  - Advanced economies
    - Target: N 1584; Mean 0.00; SD 0.05; Min -0.46; Median 0.00; Max 0.32
    - Path: N 1748; Mean 0.00; SD 0.05; Min -0.36; Median 0.00; Max 0.99
  - Emerging markets
    - Target: N 717; Mean 0.00; SD 0.08; Min -0.55; Median 0.00; Max 0.49
    - Path: N 717; Mean 0.00; SD 0.09; Min -0.82; Median 0.00; Max 0.58

### Main empirical findings on transmission
- Baseline country-level regression (30-minute window):
  - ∆y_{jt+30} = α_j + β_j * Target_{jt+30} + γ_j * Path_{jt+30} + ε_{jt+30}, where ∆y is change in yields (medium-term 1–3 years; long-term 5–10 years), stock prices, or bilateral exchange rate vs USD.
  - Positive shocks denote monetary tightening.
- Key quantitative median estimates and significance (30-minute window, scaled to 100-basis-point shocks):
  - Target factor:
    - Median increase in medium-term yields: 43 basis points (for a 100-basis-point target surprise).
    - Median increase in long-term yields: 20 basis points.
    - Median equity price decline: 1.83 percent.
    - Median exchange rate appreciation: 2.45 percent.
    - Statistical significance: medium-term yields significant in 13 out of 16 country-yield pairs; long end significant in 12 out of 16 (averages include only countries with transmission significantly different from 0).
  - Path factor:
    - Median equity price decline: 1.34 percent (for a 100-basis-point path surprise).
    - Median exchange rate appreciation: 5.06 percent.
    - Effects stronger in AEs than EMs; effects significant in the majority of countries (13 out of 16 for equities; 13 out of 15 for exchange rates).
- Country-level representative coefficients (Table A.2 excerpts preserved exactly for illustrative countries):
  - Australia — Stock Price: -2.20*** (0.47); Exchange Rate: 4.27*** (0.63); 171 obs.; R-squared 0.69
  - Euro Area — 10YT: -0.00 (0.07); Path 0.74*** (0.04); Stock Price: -0.42 (1.99); -4.64*** (1.09); 178 obs.; R-squared 0.14 (stock)
  - Japan — 1YT: 0.08*** (0.01); 10YT: 0.09*** (0.03); Stock Price: -2.32* (1.23); Exchange Rate: 2.36*** (0.50); 160 obs.; R-squared 0.76 (1YT)
  - U.S. — Stock Price: -8.38*** (1.08); 10YT: 0.01 (0.05) target; 0.56*** (0.04) path; 162 obs.; R-squared 0.28 (stock)
- Robustness:
  - Panel regressions with country fixed effects confirm country-level findings.
  - Controlling for contemporaneous U.S. monetary policy announcements yields similar estimates (dummy = 1 if a U.S. announcement occurred within 24 hours).
  - Coefficients for Japan scaled down by factor of 10 in cross-country visuals because responses are an order of magnitude larger (likely ELB-related).

### Information effects and the PM restriction
- PM (Poor Man’s) restriction classifies target shocks as “pure” monetary policy shocks only when tightening lowers equity prices:
  - Target_MP = Target if Target > 0 and ∆Stock Price < 0; 0 if Target > 0 and ∆Stock Price > 0.
- Findings:
  - PM restriction has limited impact on 30-minute high-frequency specification; HFI already filters information contamination.
  - In daily specifications, PM restriction improves precision and magnitude of stock-price responses and clarifies identification; it can alter exchange-rate results in daily data.
  - Overall: little evidence of pervasive information effects in high-frequency regressions.

### Advantages of high-frequency identification (HFI) vs lower-frequency measures
- Narrow 30-minute windows yield more precise and statistically stronger estimates than wider windows.
  - As shock and response windows widen (30-minute → 2 hours → 4 hours → full trading day), point estimates similar but fewer countries show statistically significant responses.
  - 30-minute shocks produce stronger and more consistent relationships with daily asset price changes than 2-hour, 4-hour, or daily shocks.
  - Stock-price effects of target surprises fade with longer response windows and often become statistically insignificant at daily frequency; path surprises remain more persistent.
- Implication: daily measures are more likely to be confounded by concurrent macroeconomic developments; HFI better isolates exogenous policy surprises.

### Post-COVID-19 assessment
- PostCovid dummy: PostCovid = 1 for observations after May 2021; 0 otherwise.
- Findings:
  - No systematic change in transmission to bond yields or equity prices after May 2021 across most countries and asset classes.
  - Notable exception: exchange rate responses in AEs weaken significantly after May 2021 (interaction terms negative and significant, particularly for path factor).
  - Interpretation: globally synchronized tightening and compressed interest rate differentials plausibly muted exchange-rate channel.

### Contributions, policy implications, and avenues for future research
- Contributions:
  - First cross-country high-frequency monetary policy surprise database enabling unified analysis of target and path factors across 16 countries.
  - Demonstrates global relevance of target and path factors for asset-price movements.
- Policy and research uses:
  - Inform studies on international monetary spillovers and exchange rate dynamics.
  - Explore effects of unconventional policies (QE/long-end factor) across countries.
  - Examine how institutional characteristics, central bank credibility, and market depth influence transmission.
  - Investigate interactions between monetary and fiscal policies during systemic shocks.
- Suggested further work:
  - Cross-country analysis of QE/risk-premium (long-end) factor.
  - Deeper study of institutional and market-structure drivers of cross-country heterogeneity in transmission.

*Source: IMF Working Paper — Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset (wpiea2026018-source-pdf).*

### Executive Summary ......................................................................................................

### Executive Summary

### Dataset and methodology
- Introduces a new cross-country dataset of high-frequency monetary policy surprises constructed using asset price movements in 30-minute windows around central bank announcements.
- Monetary policy surprises are decomposed into target and path factors, following established methodologies.
- Dataset uses consistent financial instruments across a broad sample of advanced and emerging market (EM) economies.
- Applies the “Poor Man’s” (PM) restriction (Jarociński and Karadi, 2021) to assess information effects.

### Validation and stylized facts
- Validates the shock dataset by showing high correlations with benchmark series for the United States (U.S.) and Euro Area.
- Documents significantly higher yield volatility during policy announcements than on non-announcement days.
- Three stylized facts:
  - Target surprises are small on average.
  - Predictability deteriorates during crises, especially for path shocks.
  - EMs exhibit fatter tails, reflecting greater volatility, shallower markets, and communication challenges.

### Monetary policy transmission: main findings
- Target surprises (unexpected changes to the current policy rate) lead to:
  - Statistically significant increases in medium- and long-term bond yields.
  - Declines in equity prices.
  - Appreciation of the domestic currency.
- Path surprises (changes in the expected future policy path) affect long-term bond yields and exchange rates.
- Transmission to long-term yields is stronger for advanced economies (AEs) than for EMs, consistent with deeper financial markets and greater use of forward guidance in AEs and interest rates not being at the effective lower bound (ELB) for an extended period in EMs.
- Evidence suggests limited directional influence of central bank information effects, as indicated by strongly negative stock market reactions to tightening shocks.

### Identification and advantages of high-frequency identification (HFI)
- High-frequency identification (HFI) achieves more precise identification than daily or lower-frequency measures.
- The PM restriction improves identification in daily data but has limited additional value in the high-frequency setting.
- As shock and response windows widen, estimated effects become less precise and statistically weaker, with fewer countries showing significant responses.
- Daily measures are more likely to be confounded by other macroeconomic developments.

### Post-COVID-19 assessment
- No systematic change found in transmission of monetary policy shocks to bond yields and equity prices after the COVID-19 period.
- Exchange rate responses to monetary policy shocks weaken notably after May 2021, likely due to high degree of monetary policy synchronization across countries during the pandemic and subsequent tightening cycles, compressing interest rate differentials and muting relative currency movements.

### Implications and avenues for further research
- Demonstrates global relevance of target and path factors in driving asset prices.
- The dataset provides a foundation to explore:
  - International monetary spillovers.
  - Impact of unconventional policies on long-term yields.
  - Exchange rate dynamics and financial integration.
  - How institutional frameworks, central bank credibility, and market depth influence transmission.
  - Interactions between monetary and fiscal policies during periods of systemic stress.

*IMF Working Paper — Executive Summary: Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset.*

### Introduction

### Introduction

### Context and motivation
- Renewed interest in assessing the transmission and effectiveness of monetary policy following the largest runup in global inflation in decades.
- Accurate identification of monetary policy shocks is central to this discussion.
- High-frequency identification (HFI) strategies are emphasized to isolate exogenous monetary policy shocks from movements driven by broader macroeconomic news or expectations.
- HFI techniques rely on asset price changes in narrowly defined windows around monetary policy announcements, minimizing contamination from other news and avoiding strong parametric assumptions.
- Alternative methods (Taylor-rule residuals, VAR innovations, survey-based surprises) often suffer from limited time resolution, endogeneity, or difficulties capturing unconventional policy actions.
- HFI has become the benchmark approach, particularly during periods of heightened policy uncertainty or market volatility.

### Dataset and methodology
- Constructed a comprehensive cross-country database of high-frequency monetary policy shocks covering 16 AEs and EMs.
- Dataset includes intra-day surprise measures derived from interest rate instruments in a 30-minute window around monetary policy announcements.
- Standardized set of financial instruments: short-term Overnight Indexed Swaps (OIS), government bonds, and long-term yields, adjusted where necessary for market-structure differences.
- Decomposed monetary policy shocks into distinct components:
  - Target surprises: unexpected changes in the current policy rate.
  - Path surprises: revisions to the expected trajectory of future rates.
- Measurement approach:
  - Target factor measured using changes in the OIS-implied short-term rate.
  - Path factor measured as changes in short- to medium-term bond yields, orthogonalized with respect to the target factor, following Rogers et al. (2018) and Braun et al. (2025).
- Note on QE factor (long end/risk premium): dataset includes this component, but the present paper focuses on target and path factors; cross-country analysis of the QE factor is left for future research due to heterogeneity in unconventional measures across countries.

### Validation and stylized facts
- Validation: U.S. and Euro Area estimates compared to Swanson (2021) and Altavilla et al. (2019), finding correlation coefficients near or above 0.9.
- Announcement-window volatility: yield volatility during policy announcements is significantly higher than during the same window on non-announcement days in nearly all countries—suggesting surprises capture policy-induced movements rather than routine market noise.
- Three key stylized facts from the dataset:
  - Target surprises are generally small across both advanced and emerging economies.
  - Time-series evolution: predictability deteriorates sharply during crises or regime shifts; the interquartile range widens during the Great Financial Crisis, the pandemic, and the recent inflation surge, reflecting repeated market misjudgments of policy persistence, particularly in path factors.
  - Cross-sectional distribution: both AEs and EMs centered near zero, but EMs display fatter tails—especially for path shocks—suggesting less predictable policy paths amid greater macroeconomic volatility, shallower markets, and communication challenges.

### Empirical findings on transmission to financial variables
- Baseline empirical framework: regress 30-minute changes in financial variables (bond yields, equity prices, exchange rates) on identified target and path surprises to capture immediate market responses.
- Main results:
  - Target surprises associated with significant increases in short- and medium-term yields, reductions in equity prices, and appreciations of exchange rates.
  - Path surprises significantly affect yields and asset prices, particularly in AEs—potentially reflecting deeper financial markets and greater reliance on forward guidance, and the fact that interest rates in EMs were not at the ELB for an extended period.
  - Effects are robust, statistically significant, and directionally consistent with standard asset pricing models.
- Information effects:
  - Potential information effects (markets interpreting policy actions as signals about central-bank private information) appear modest and do not alter qualitative responses.
  - Little evidence of information effects in high-frequency data.
  - Controlling for information effects in daily data strengthens estimated transmission, particularly for equity prices—suggesting HFI narrow windows already achieve more precise identification.

### Robustness and comparison with daily data
- Varying the width of shock and response windows shows broader windows reduce precision and statistical significance of estimated transmission effects.
- Broader-window estimates may understate asset-price sensitivity to policy shocks or misattribute movements to monetary policy when driven by other factors.
- Caution urged when interpreting monetary-policy transmission estimates based solely on daily data, commonly used in EM contexts.
- Temporal changes in transmission:
  - No systematic change in transmission of monetary shocks to bond yields or equity prices.
  - Exchange rate responses in AEs weaken significantly after May 2021, coinciding with the globally synchronized tightening phase.
  - Attenuation in exchange-rate responses likely due to smaller interest rate differentials across countries and increased policy coordination.

### Contributions, implications, and avenues for future research
- Contributions:
  - Provides new cross-country evidence on structure and stability of monetary transmission using a standardized HFI approach across 16 AEs and EMs.
  - Demonstrates global relevance of target and path factors in influencing asset prices.
  - Offers a foundation for future research on international spillovers, unconventional measures targeting the long end of the yield curve, exchange rate dynamics, and financial market integration.
- Potential further research enabled by the dataset:
  - Cross-country analysis of QE/risk-premium (long-end) factor.
  - How institutional characteristics, central bank credibility, and market depth influence the strength and speed of transmission.
  - Interactions between monetary and fiscal policies in large systemic shocks.

### Organization of the paper
- Section II: recent literature.
- Section III: underlying data, methodology, and resulting shocks database.
- Section IV: analysis of transmission of shocks to select financial variables.
- Section V: conclusion.

*wpiea2026018-source-pdf - Introduction*

### references therein for Australia, Nakamura et al. (2021) and Kubuta and Shintani (2022) for Japan, Almerud et al. (2024)

### wpiea2026018-source-pdf - references therein for Australia, Nakamura et al. (2021) and Kubuta and Shintani (2022) for Japan, Almerud et al. (2024)

### Study scope and contribution
- Title in document: Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset.
- The paper constructs a comprehensive cross-country HFI dataset for both AEs and EMs to identify monetary policy shocks using intra-day financial market data.
- Positions itself relative to existing work:
  - Ferrari et al. (2021) and Kearns et al. (2023): identify monetary policy shocks through high-frequency methods for multiple countries but limited to six and seven AE central banks respectively.
  - Choi et al. (2024): compile a cross-country dataset using a mix of intra-day and daily data.
  - Bolhuis et al. (2024): apply a consistent methodology across countries but rely on daily data and capture only one dimension of monetary policy.
- The dataset and methodology fill the gap by offering broad cross-country coverage with consistent intra-day identification.

### Sample coverage and events
- Sample comprises 16 countries—11 AEs and 5 EMs.
- Sample start dates vary depending on data availability, generally beginning in the early 2000s; exception: India begins in 2016 following the adoption of the inflation targeting framework.
- Sample end date: May 2025.
- The study focuses on scheduled monetary policy announcements. Announcement dates and times are sourced from Bloomberg and cross-verified with central bank websites.
- The study excludes announcements coinciding with:
  - the Global Financial Crisis (September 2008–March 2009)
  - the COVID-19 outbreak (February–April 2020)
- Unscheduled announcements are excluded due to unavailability of precise timing.
- Total number of monetary policy events in the dataset: 2,528.

- Table 1. Sample Coverage (selected entries and totals preserved exactly)
  - Australia: Data start dates 12/2/2008; Data end dates 5/20/2025; Number of events 178
  - Canada: 7/10/2007 — 4/16/2025 — 146
  - Czech Republic: 3/25/2004 — 5/7/2025 — 186
  - Euro Area: 5/4/2006 — 4/17/2025 — 188
  - Japan: 12/1/2009 — 5/1/2025 — 163
  - New Zealand: 1/24/2001 — 5/28/2025 — 187
  - Norway: 1/22/2003 — 5/8/2025 — 174
  - Sweden: 7/6/2001 — 5/8/2025 — 149
  - Switzerland: 12/16/2010 — 3/20/2025 — 61
  - UK: 9/6/2007 — 5/8/2025 — 180
  - U.S.: 6/30/2004 — 5/7/2025 — 171
  - Total AEs: 1783
  - Brazil: 10/21/2009 — 5/7/2025 — 118
  - Chile: 5/10/2007 — 4/29/2025 — 149
  - India: 2/2/2016 — 4/9/2025 — 57
  - Mexico: 10/14/2005 — 5/15/2025 — 179
  - Poland: 1/21/2004 — 5/7/2025 — 242
  - Total EMs: 745
  - Total (all countries): 2528
- Note: The final column shows the total number of events with underlying series available for shock construction.

### Underlying financial market data and instruments
- Primary instruments used to identify monetary policy shocks:
  - OIS rates (1- and 3-month)
  - Government bond yields (1-, 2-, and 10-year)
  - Equity indices
  - Exchange rates against the United States dollar (USD)
- Data frequency and source:
  - Minute-by-minute data (as of close of the minute; last traded price in the minute) sourced from London Stock Exchange Group (LSEG).
- Instrument substitutions where OIS markets are thin or undeveloped:
  - interest rate swaps (Brazil, Chile, and Mexico)
  - forward rate agreements (Poland and Czech Republic)
  - foreign exchange (FX) forward-implied rates (Norway)
- Special cases and adjustments:
  - For Brazil and Chile, where most announcements occur outside trading hours, daily opening and closing prices are used.
  - Following discontinuation of EONIA on January 3, 2022, Euro Short-Term Rate (€STR)-based OIS contracts are used for Euro Area countries.
  - For Mexico, the 3-year bond yield is used in place of the 2-year yield.
  - For OIS and interest rate swaps, prices are quoted as the fixed leg rate; forward rate agreements are also quoted as annualized interest rates and used directly.
  - For Norway, where the OIS market is underdeveloped, a synthetic 1-month Norwegian interest rate is constructed from the USD/NOK 1-month FX forward swap using covered interest rate parity (CIP), following Brubakk et al. (2017). The closing price of the U.S. 1-month OIS is used as the corresponding US interest rate on the same date.
- Footnotes and data specifics:
  - The OIS rate definition and properties as in footnote: "The OIS rate is the fixed rate of a contract where counterparties agree to exchange the fixed rate for a floating reference rate, usually the prevailing money market rate (e.g. federal funds rate in the U.S., Sterling Overnight Index Average (SONIA), Euro Overnight Index Average (EONIA) or other interbank rate in other jurisdictions). The fixed rate incorporates market expectations of future short-term interest rates and has in practice been shown to be a reliable measure of market expectations up to the 2-year tenor (Llyod, 2021)."
  - Monetary policy surprises for bond instruments are calculated as 100 times the change in yields over the announcement window; for stock prices and exchange rates, surprises are measured as percentage changes over the same window.
  - In Brazil, 198 out of 216 announcements are outside of trading hours and in Chile, 256 out of 257 announcements are outside of trading hours.

### Identification of monetary policy surprises and event window
- Definition: monetary policy surprises are changes in interest rates within a narrow window around scheduled monetary policy announcements.
- Baseline event window specification (following Braun et al. (2025)):
  - 30-minute window: from 10 minutes before to 20 minutes after the scheduled announcement.
  - Pre-announcement rate: median price during the 20-to-10-minute window before the event.
  - Post-announcement rate: median price during the 20-to-30-minute window after the event.
- Visualization: Figure 1 illustrates the measurement windows; pre- and post-announcement medians defined as above.
- Handling announcements outside trading hours:
  - Pre- and post-announcement windows are defined using the last and first 10 minutes of trading on the relevant days.
  - Where the open-close method yields excessive zero changes (Chile), close-to-close price differences are used instead.
- Decomposition of surprises (following Rogers et al. (2018)):
  - Two components:
    1. Target factor: Change in short-term interest rate (y_jt^3m, 3-month OIS or equivalent) during the announcement window:
       - Target_j,t+θ = Δ y_j,t+θ^3m. (Equation (1))
  - The study focuses on the 3-month maturity where available, as volumes in this market were generally higher than in the 1-month OIS market in the sample (Table 2 referenced).

### Methodological notes and cross-country comparability
- The paper details steps taken to ensure cross-country comparability in data construction and identification methods.
- Uses consistent intra-day measurement principles across heterogeneous market structures, substituting instruments where necessary while preserving the principle of high-frequency identification.

*Source: IMF Working Paper — Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset (extract provided).*

### 2. Path factor: Residual from a regression of the 2-year bond yield change on the target factor:

### 2. Path factor: Residual from a regression of the 2-year bond yield change on the target factor:

### Definition and construction
- Path factor is defined as the residual 휀푗푡+휃2푦 from the regression:
  - ∆푦푗푡+휃2푦 = 훽∆푦푗푡+휃3푚 + 휀푗푡+휃2푦, (2)
  - where 푦푗푡2푦 is the 2-year Treasury yield. If the 2-year yield is unavailable, a 1-year or 3-year yield is used.
- Scaling: factors are scaled so that the target and path components correspond to a 100-basis point change in the respective yield (short-term, 2-year, and 10-year) for each country.
- Winsorization: All series are winsorized at the first and ninety-ninth percentiles prior to shock construction and regression analysis.

### Trading-volume evidence motivating instrument choices (Table 2)
- Total Trading Volume in May 2025, Billions of Local Currency Units (LCU):
  - AUD: 1-month OIS 0.1; 3-month OIS 85
  - BRL: 1-month OIS 6.3; 3-month OIS -
  - CAD: 1-month OIS 10.1; 3-month OIS 20.5
  - CHF: 1-month OIS 0.2; 3-month OIS -
  - EUR: 1-month OIS 32; 3-month OIS 66.4
  - GBP: 1-month OIS 46.1; 3-month OIS 165
  - JPY: 1-month OIS 49.6; 3-month OIS 765
  - MXN: 1-month OIS 3.9; 3-month OIS 631.5
  - NZD: 1-month OIS -; 3-month OIS 13.1
  - USD: 1-month OIS 136.9; 3-month OIS 518
- Note: Table summarizes May 2025 OIS trading volumes (approximate 30-day and 90-day maturities). Evidence shows 3-month OIS trading volume is generally higher than 1-month, motivating use of the 3-month OIS rate for all countries.

### Validation and comparison with existing shock series
- Method choice: Rogers et al. (2018) method chosen for feasibility and robustness; alternative methods (e.g., PCA across many yields) often infeasible in EMs.
- Validation: US and Euro Area shocks compared with Swanson (2021) and Altavilla et al. (2019) respectively, finding correlation coefficients near or above 0.9.
- Swanson (2021) evidence: path shocks derived from the 2-year bond yield using the Rogers method are highly correlated with PCA-based shocks.
- Window choices: path shocks compared to Altavilla et al. (2019) use a 2-hour window for consistency; other figures computed using a 30-minute window.

### Volatility in event vs. non-event windows
- Approach: compare standard deviation of yields and asset prices within the 30-minute monetary policy window to the same window on the following trading day.
- Key findings:
  - Volatility of 3-month yields on announcement days is 3.96 times volatility on non-announcement days (average median).
  - Volatility of 2-year yields on announcement days is 4.1 times volatility on non-announcement days (average median).
- Figures for 10-year yields, exchange rates, and stock indices show volatility markedly higher on announcement days; majority of correlations are statistically significantly different.
- Event-window significance assessed at the 90 percent significance level.

### Stylized facts of monetary policy shocks
- Three notable stylized facts:
  - First: Target surprises are on average small across AEs and EMs, with means close to zero.
  - Second: Predictability deteriorates sharply during crises or regime change; interquartile ranges widen during global financial crisis, COVID-19 pandemic, and recent high-inflation episode. Path factor surprises show greater swings.
  - Third: Cross-sectional distribution differs between AEs and EMs; EMs display fatter tails, particularly for path shocks (less predictable future policy path).
- Table 3 — Summary statistics of high-frequency monetary policy shocks:
  - Advanced economies
    - Target: N 1584; Mean 0.00; SD 0.05; Min -0.46; Median 0.00; Max 0.32
    - Path: N 1748; Mean 0.00; SD 0.05; Min -0.36; Median 0.00; Max 0.99
  - Emerging markets
    - Target: N 717; Mean 0.00; SD 0.08; Min -0.55; Median 0.00; Max 0.49
    - Path: N 717; Mean 0.00; SD 0.09; Min -0.82; Median 0.00; Max 0.58

### Baseline results: transmission to yields, equities, and exchange rates
- Baseline regression (country level, 30-minute window):
  - ∆푦푗푡+30 = 훼푗 + 훽푗 * Target푗푡+30 + 훾푗 * Path푗푡+30 + 휀푗푡+30
  - ∆푦푗푡+30 is change in yields (medium-term: 1-3 years; long-term: 5-10 years), stock prices, or bilateral exchange rate vs USD.
  - Positive shocks indicate monetary tightening.
- Scaling: target surprise → 100-basis-point increase in the 3-month OIS; path surprise → 100-basis-point increase in the 2-year yield.
- Key quantitative findings:
  - A 100-basis-point surprise in the target factor leads to a median increase of 43 basis points in medium-term yields.
  - A 100-basis-point target surprise leads to a median increase of 20 basis points in long-term yields.
  - Statistical significance: medium-term horizon 13 out of 16 country-yield pairs significant; long end 12 out of 16 significant (averages include only countries with transmission significantly different from 0).
  - Equity and exchange rate responses:
    - 100-basis-point target shock → median equity price decline of 1.83 percent and a 2.45-percent appreciation of the exchange rate.
    - 100-basis-point path shock → median equity price decline of 1.34 percent and a 5.06-percent appreciation of the exchange rate.
    - Statistical significance: effects significant in majority of countries (13 out of 16 for equities; 13 out of 15 for exchange rates).
- Robustness:
  - Panel regression with country fixed effects confirms country-level findings.
  - Controlling for contemporaneous U.S. monetary policy announcements yields similar estimates.
- Note on Japan:
  - Coefficients for Japan scaled down by a factor of 10 for cross-country visual comparability due to responses being an order of magnitude larger (likely related to prolonged ELB).

### Assessing information effects (PM restriction)
- PM restriction classification for the target factor:
  - Target Factor_MP = Target Factor if Target Factor > 0 and ∆Stock Price < 0; 0 if Target Factor > 0 and ∆Stock Price > 0.
- Intuition: tightening that raises equity prices is interpreted as information shock (positive private information), whereas tightening that lowers equity prices is a “pure” monetary policy shock.
- Findings:
  - Applying PM restriction has limited impact on 30-minute high-frequency specification (baseline already captures pure shocks).
  - In daily–daily specification, PM restriction improves clarity and precision of stock price responses (larger magnitudes, more countries significant).
  - PM restriction can reduce significance or magnitude for exchange rate responses in daily specification (possibly due to complex drivers of exchange rates).
  - For path factor, PM restriction similarly leaves 30-minute results largely unchanged and improves stock-price identification in daily specification; it does not diminish exchange rate responses.
- Overall: little evidence of pervasive information effects in high-frequency regressions; HFI naturally filters much information contamination.

### Contribution of high-frequency data (HFI) vs lower-frequency approaches
- Three empirical dimensions explored:
  - Vary identification (shock) window: 30-minute → 2 hours → 4 hours → full trading day (response window equal to shock window).
    - Result: point estimates similar, but number of countries with statistically significant responses declines as windows widen.
  - Vary shock window size while keeping response window at daily frequency:
    - 30-minute shocks yield stronger and more consistent relationships with daily asset price changes than 2-hour, 4-hour, or daily shocks.
    - Target factor often not significantly associated with daily stock price changes in majority of countries; path factor shows larger and more statistically significant effects when identified with 30-minute window.
  - Fix shock window at 30 minutes and vary response window from 30 minutes to end of trading day:
    - Stock-price effect of target surprises fades with longer windows and becomes statistically insignificant for most countries at daily frequency.
    - Path surprises remain significant as response window increases, indicating more persistent effect.
    - Exchange rate responses robust across horizons; number of countries with significant exchange-rate responses higher for 30-minute shocks.
- Implication: HFI improves precision and significance of estimated monetary policy effects and better isolates exogenous policy surprises; daily data can understate sensitivity or misattribute effects.

### Did transmission change post-COVID-19?
- Specification adds interaction terms with PostCovid dummy (PostCovid = 1 for observations after May 2021; 0 otherwise).
  - Modified regression includes 휃푗T * Target * PostCovid and 휃푗P * Path * PostCovid.
- Findings:
  - Across most countries and asset classes, interaction terms are statistically insignificant — transmission largely unchanged after May 2021.
  - Notable exception: exchange rate response in AEs — interaction terms (particularly for path factor) are negative and significant, indicating weaker transmission to exchange rates post-2020.
  - Interpretation: global synchronization of monetary policy and compressed interest rate differentials plausibly muted exchange rate channel.
- Conclusion: bond yields and equity prices remained responsive to policy surprises; exchange rate channel in AEs shows meaningful attenuation post-May 2021.

### Conclusions and implications
- Dataset: first cross-country high-frequency monetary policy surprise database enabling unified analysis of target and path factor shocks across many countries.
- Summary quantitative takeaways:
  - 100-basis-point target surprise → median 43 basis-point increase in medium-term yields; 20 basis points at long end.
  - 100-basis-point target surprise → median equity price decline 1.83 percent; exchange rate appreciation 2.45 percent.
  - 100-basis-point path surprise → median equity price decline 1.34 percent; exchange rate appreciation 5.06 percent.
- HFI advantages:
  - Isolates exogenous policy shocks more cleanly than daily data.
  - Yields more precise and significant estimates; narrow (30-minute) windows minimize contamination by concurrent news.
- Cross-country heterogeneity:
  - Path factor (forward guidance) plays larger role in AEs; less salient in EMs.
  - Exchange rate channel sensitive to global policy synchronization (attenuated in AEs post-May 2021).
- Research and policy uses:
  - Dataset can inform studies on international spillovers, unconventional tools, exchange rate dynamics, and market integration.
  - Future work could examine institutional characteristics, central bank credibility, market depth, and interaction with fiscal policy.

*Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026018-source-pdf.pdf*

### Annex I. Individual Country Results

### Annex I. Individual Country Results

### Summary statistics of underlying series by country (Table A.1)
- The table reports summary statistics (N, Daily Mean, SD, Min, P25, P75, Max) for policy instruments (e.g., 3M OIS, 1M FX Swap, 1X4 FAR, 6X9 FAR, 21D IRS, 3M IRS), bond yields (1-, 2/3-, 10-year), stock price indices, and spot exchange rates (vs. USD) for each country.
- Selected country-level daily statistics (examples reported exactly as in source):
  - Australia (N = 178 for most series)
    - 3M OIS: Mean 0.03, SD 0.07, Min -0.27, P25 0.01, P75 0.03, Max 0.47
    - 2Y: Mean -0.01, SD 0.05, Min -0.21, P25 -0.03, P75 0.01, Max 0.20
    - 10Y: Mean -0.01, SD 0.04, Min -0.15, P25 -0.03, P75 0.01, Max 0.28
    - Stock: Mean 0.09, SD 0.34, Min -1.86, P25 -0.06, P75 0.24, Max 1.35
    - FX: Mean 0.04, SD 0.88, Min -2.38, P25 -0.52, P75 0.62, Max 3.36
  - Euro Area (N = 188 for most series)
    - 3M OIS: Mean 0.01, SD 0.04, Min -0.28, P25 0.00, P75 0.03, Max 0.20
    - 2Y: Mean -0.01, SD 0.07, Min -0.27, P25 -0.05, P75 0.01, Max 0.28
    - 10Y: Mean -0.01, SD 0.06, Min -0.20, P25 -0.04, P75 0.03, Max 0.18
    - Stock: Mean 0.05, SD 0.54, Min -1.72, P25 -0.27, P75 0.37, Max 1.77
    - FX: Mean -0.08, SD 0.79, Min -2.10, P25 -0.54, P75 0.43, Max 3.08
  - United States (sample shown)
    - 3M OIS (N = 168): Mean 0.01, SD 0.04, Min -0.29, P25 0.01, P75 0.02, Max 0.20
    - 2Y (N = 171): Mean -0.02, SD 0.07, Min -0.30, P25 -0.05, P75 0.02, Max 0.26
    - 10Y (N = 171): Mean -0.02, SD 0.08, Min -0.48, P25 -0.05, P75 0.03, Max 0.20
    - Stock (N = 171): Mean 0.15, SD 0.70, Min -1.56, P25 -0.25, P75 0.46, Max 4.06
- Notes: Table A.1 presents summary statistics for series used to construct monetary policy shocks and related asset prices. (Full country-by-country detailed table entries are in the source.)

### Baseline regression results by country (Table A.2)
- Table A.2 reports regression estimates of asset price changes on monetary policy surprises (Target and Path factors), with robust standard errors in parentheses and significance markers *, **, *** for 10 percent, 5 percent, and 1 percent levels.
- Representative country-level coefficient estimates (Target (SE); Path (SE); N Obs.; R-squared) — values preserved exactly:
  - Australia
    - 1YT: 0.73*** (0.03); 0.85*** (0.04); 171; 0.92
    - 10YT: 0.35*** (0.04); 0.59*** (0.03); 171; 0.80
    - Stock Price: -2.20*** (0.47); -3.82*** (0.51); 171; 0.42
    - Exchange Rate: 4.27*** (0.63); 6.85*** (0.64); 171; 0.69
  - Brazil
    - 1YT: 0.41 (0.27); 0.36*** (0.08); 121; 0.27
    - 3YT: -0.05 (0.19); 0.53*** (0.09); 119; 0.34
    - Stock Price: -0.02* (0.01); -0.01 (0.01); 121; 0.02
    - Exchange Rate: -0.05 (0.08); -0.01 (0.08); 120; 0.00
  - Canada
    - 1YT: 0.62*** (0.09); 0.53*** (0.06); 136; 0.69
    - 10YT: 0.46*** (0.04); 0.52*** (0.04); 136; 0.75
    - Stock Price: -4.28*** (0.89); -0.21 (0.94); 136; 0.16
    - Exchange Rate: 5.62*** (0.95); 6.57*** (0.76); 136; 0.60
  - Euro Area
    - 1YT: 0.47*** (0.08); 0.39*** (0.06); 178; 0.69
    - 10YT: -0.00 (0.07); 0.74*** (0.04); 178; 0.69
    - Stock Price: -0.42 (1.99); -4.64*** (1.09); 178; 0.14
    - Exchange Rate: 2.34** (0.91); 5.06*** (0.80); 178; 0.30
  - Japan
    - 1YT: 0.08*** (0.01); 0.08*** (0.02); 160; 0.76
    - 10YT: 0.09*** (0.03); 0.28*** (0.04); 160; 0.69
    - Stock Price: -2.32* (1.23); -4.94*** (1.12); 160; 0.2
    - Exchange Rate: 2.36*** (0.50); 3.39*** (1.08); 160; 0.21
  - Mexico
    - 1YT: 0.26*** (0.07); 0.04 (0.04); 167; 0.21
    - 10YT: 0.23*** (0.05); 0.06** (0.03); 170; 0.17
    - Stock Price: -1.89*** (0.55); -0.63*** (0.23); 170; 0.10
    - Exchange Rate: 2.33*** (0.47); -0.32 (0.25); 170; 0.23
  - New Zealand
    - 1YT: 0.46*** (0.09); 0.56*** (0.11); 95; 0.66
    - 10YT: 0.47*** (0.04); 0.60*** (0.05); 103; 0.84
    - Stock Price: -2.50*** (0.47); -1.59*** (0.37); 103; 0.31
    - Exchange Rate: 5.57*** (1.19); 6.81*** (1.23); 103; 0.48
  - UK
    - 1YT: 0.44*** (0.03); 0.94*** (0.05); 170; 0.76
    - 10YT: 0.22*** (0.05); 0.99*** (0.06); 170; 0.79
    - Stock Price: -2.10*** (0.53); -3.48*** (0.72); 170; 0.21
    - Exchange Rate: 3.78*** (0.71); 6.80*** (0.80); 170; 0.53
  - U.S.
    - 1YT: 0.72*** (0.11); 0.57*** (0.05); 95; 0.86
    - 10YT: 0.01 (0.05); 0.56*** (0.04); 162; 0.50
    - Stock Price: -8.38*** (1.08); -3.16* (1.74); 162; 0.28
- Additional country entries (Norway, Poland, Sweden, Switzerland, Chile, Brazil, India, Czech Republic, Poland, Sweden, Switzerland) are reported in the table with coefficients, standard errors, N Obs., and R-squared exactly as presented in the source.
- Notes: Coefficients for medium-term yields used to construct surprises are not shown in the table because X and Y are nearly the same for the 30-minute window.

### List of tickers used (Table A.3)
- Table A.3 maps instruments and asset classes to tickers used in data sourcing (tickers preserved exactly as shown for each country/instrument).
- Example tickers (preserved exactly):
  - U.S.: USD1MOIS=, USD3MOIS=, US1YT=RR, US2YT=RR, US10YT=RR, SPY, .
  - Euro Area: EUREST1M=, EUREST3M=, EU1YT=RR, EU2YT=RR, EU10YT=RR, .GDAXI EUR=
  - UK: GBP1MOIS=, GBP3MOIS=, GB1YT=RR, GB2YT=RR, GB10YT=RR, .FTSE GBP=
  - Japan: JPY1MOIS=, JPY3MOIS=, JP1YT=RR, JP2YT=RR, JP10YT=RR, 1321.T JPY=
  - Canada: CAD1MOIS=, CAD3MOIS=, CA1YT=RR, CA2YT=RR, CA10YT=RR, .GSPTSE* CAD=
  - Brazil, Chile, Korea, India, Mexico, Norway, Poland, Israel, and other country tickers are listed in the table (tickers preserved as in source).
- Source of tickers: London Stock Exchange Group.
- Note: For India, tickers are used from 2016 onwards, when India adopted inflation targeting.

### Panel regression specification and robustness: controlling for contemporaneous U.S. announcements (Annex II / Figure A.1)
- Panel fixed-effects specification (preserved equation):
  - ∆y_{jt+30} = α_j + β * Target_{jt+30} + γ * Path_{jt+30} + δ_t + ε_{jt+30}.
- Definitions and estimation details (preserved language):
  - ∆y_{jt+30} denotes change in yields (medium-term: 1-3 years; long-term: 5-10 years), stock prices, or bilateral exchange rate against the USD over a 30-minute window around a monetary policy announcement in country j.
  - Target_{jt+30} and Path_{jt+30} capture unexpected changes in the current policy rate and future policy path identified through high-frequency financial market data.
  - α_j is a country-specific fixed effect; δ_t is a fixed effect for the year-month of the surprise.
  - Standard errors are clustered at the country level.
- Robustness check: controlling for contemporaneous U.S. monetary policy announcements
  - A dummy variable equals 1 if a U.S. monetary policy announcement occurred within 24 hours before the domestic announcement.
  - Findings preserved from the source:
    - Controlling for contemporaneous U.S. announcements, estimated effects of target and path factor surprises on domestic bond yields, stock prices, and exchange rates remain pronounced.
    - For a target factor tightening shock (Panel a): medium- and long-term yields in both advanced economies (AEs) and emerging markets (EMs) rise; stock prices fall; exchange rates depreciate notably—especially in AEs.
    - For the path factor (Panel b): bond yields increase; stock prices decline; exchange rates appreciate, with stronger sensitivity in AEs.
    - Panel estimates (black dots) corroborate these patterns, indicating international transmission of U.S. monetary policy surprises reflects a robust spillover effect rather than merely coincident U.S. announcements.
- Notes on Figure A.1:
  - Regressions include a dummy for U.S. events occurring within 24 hours following a US monetary policy announcement.
  - Monetary policy surprises and changes in bond yields and asset prices computed over a 30-minute window.
  - Exchange rates are defined as USD per local currency (an increase denotes an appreciation of the local currency).
  - Solid triangles indicate a significant effect at 90 percent level; empty triangles indicate a non-significant effect; red denotes EMs, blue denotes AEs.
  - Medium-term yields coefficients plot separate regressions for 1-, 2-, and 3-year yields where available; long-term yields similarly plot 5- and 10-year yields.

*Source: Authors’ calculations.*

### References

### wpiea2026018-source-pdf - References

### High-frequency identification and monetary policy surprises
- Acosta, M. (2022). The perceived causes of monetary policy surprises. Published Manuscript.
- Ahn J., Kim, J., & Lee, B. (2021). Identifying Monetary Policy Shocks Using High Frequency Data: Evidence from Korea -Focusing on the Central Bank Information Effect Contained in Monetary Policy. The Korean Journal of Economic Studies, vol.69, no.4, pp. 45–116.
- Altavilla, C., Brugnolini, L., Gürkaynak, R. S., Motto, R., & Ragusa, G. (2019). Measuring euro area monetary policy. Journal of Monetary Economics, 108, 162–179.
- Andrade, P., & Ferroni, F. (2021). Delphic and odyssean monetary policy shocks: Evidence from the euro area. Journal of Monetary Economics, 117, 816–32.
- Bagliano, F. C., & Favero, C. A. (1999). Information from financial markets and VAR measures of monetary policy. European economic review, 43(4-6), 825–37.
- Bauer, M. D., & Swanson, E. T. (2023a). “An Alternative Explanation for the ‘Fed Information Effect’, American Economic Review 113, March 2023, 664–700.
- Bauer, M. D., & Swanson, E. T. (2023b). A reassessment of monetary policy surprises and high-frequency identification. NBER Macroeconomics Annual, 37(1), 87–155.
- Bernanke, B. S., & Kuttner, K. N. (2005). What explains the stock market's reaction to Federal Reserve policy?. The Journal of finance, 60(3), 1221–57.
- Cochrane, J. H., & Piazzesi, M. (2002). The fed and interest rates—a high-frequency identification. American economic review, 92(2), 90–95.
- Faust, J., Rogers, J. H., Swanson, E., & Wright, J. H. (2003). Identifying the effects of monetary policy shocks on exchange rates using high frequency data. Journal of the European Economic association, 1(5), 1031–57.
- Faust, J., Swanson, E. T., & Wright, J. H. (2004). Identifying VARs based on high frequency futures data. Journal of Monetary Economics, 51(6), 1107–31.
- Kuttner, K. N. (2001). Monetary policy surprises and interest rates: Evidence from the Fed funds futures market. Journal of monetary economics, 47(3), 523–44.
- Nakamura, F., Sudo, N., & Sugisaki, Y. (2021). A Quest for Monetary Policy Shocks in Japan by High Frequency Identification (No. 21-E-02). Institute for Monetary and Economic Studies, Bank of Japan.
- Plagborg‑Møller, M., & Wolf, C. K. (2021). Local projections and VARs estimate the same impulse responses. Econometrica, 89(2), 955–80.
- Rigobon, Roberto, and Brian Sack. "Measuring the reaction of monetary policy to the stock market." The Quarterly Journal of Economics 118, no. 2 (2003): 639–69.
- Swanson, E. T. (2021). Measuring the effects of federal reserve forward guidance and asset purchases on financial markets. Journal of Monetary Economics, 118, 32–53.

### Monetary policy transmission, forward guidance, and term structure effects
- Almerud, J., Krygier, D., Lundvall, H., & Njie, M. (2024). Measuring Riksbank monetary policy: Shocks and macroeconomic transmission (No. 445). Sveriges Riksbank Working Paper Series.
- Altavilla, C, Gürkaynak, R., Kind, T.  & Laeven, L. (2025). Monetary Transmission with Frequent Policy Events. CEPR Discussion Paper No. 20196. CEPR Press, Paris & London. https://cepr.org/publications/dp20196
- Brubakk, L., Ter Ellen, S., & Xu, H. (2017). Forward guidance through interest rate projections: does it work?
- Campbell, J. R., Evans, C. L., Fisher, J. D., Justiniano, A., Calomiris, C. W., & Woodford, M. (2012). Macroeconomic effects of federal reserve forward guidance [with comments and discussion]. Brookings papers on economic activity, 1–80.
- Hubert, P., & Labondance, F. (2018). The effect of ECB forward guidance on the term structure of interest rates. International Journal of Central Banking, 14(5), 193–222.
- Ferrarri, M., Kearns, J., & Schrimpf, A. (2021). Monetary policy’s rising FX impact in the era of ultra-low rates. Journal of Banking & Finance, 129, 106142.
- Gertler, M., & Karadi, P. (2015). Monetary policy surprises, credit costs, and economic activity. American Economic Journal: Macroeconomics, 7(1), 44–76.
- Gilchrist, S., López-Salido, D., & Zakrajšek, E. (2015). Monetary policy and real borrowing costs at the zero lower bound. American Economic Journal: Macroeconomics, 7(1), 77–109.
- Jarociński, M., & Karadi, P. (2020). Deconstructing monetary policy surprises—the role of information shocks. American Economic Journal: Macroeconomics, 12(2), 1-43.
- Miranda-Agrippino, S., & Ricco, G. (2021). The transmission of monetary policy shocks. American Economic Journal: Macroeconomics, 13(3), 74–107.
- Ricco, G., Savini, E., & Tuteja, A. (2024). Monetary Policy, Information and Country Risk Shocks in the Euro Area.
- Solís, P. (2023). Does the exchange rate respond to monetary policy in Mexico? Solving an exchange rate puzzle in emerging markets. Journal of Money, Credit and Banking, 55(8), 2093–2113.
- Soosalu, M. (2024). Monetary Policy Transmission in Canada–A High Frequency Identification Approach. The BE Journal of Macroeconomics, 24(2), 781–811.

### Cross-country, emerging markets, and spillovers
- Brandao-Marques, M. L., Gelos, M. R., Harjes, M. T., Sahay, M. R., & Xue, Y. (2020). Monetary policy transmission in emerging markets and developing economies. International Monetary Fund.
- Deb, M. P., Estefania-Flores, J., Firat, M., Furceri, D., & Kothari, S. (2023). Monetary policy transmission heterogeneity: cross-country evidence. International Monetary Fund.
- Kearns, J., Schrimpf, A., & Xia, F. D. (2023). Explaining monetary spillovers: The matrix reloaded. Journal of Money, Credit and Banking, 55(6), 1535–68.
- Ferrari, M., Kearns, J., & Schrimpf, A. (2021). Monetary policy’s rising FX impact in the era of ultra-low rates. Journal of Banking & Finance, 129, 106142.
- Rogers, J. H., Scotti, C., & Wright, J. H. (2018).  . Journal of Money, Credit and Banking, 50(8), 1827–50.
- Choi, S., Willems, T., & Yoo, S.Y. (2024). Revisiting the monetary transmission mechanism through an industry-level differential approach, Journal of Monetary Economics, Volume 145, 2024, 103556, ISSN 0304-3932, https://doi.org/10.1016/j.jmoneco.2024.103556.

### Datasets, methodology, and measurement issues
- Bolhuis, M., Das, S., & Yao, B. (2024). A New Dataset of High-Frequency Monetary Policy Shocks, IMF Working Paper.
- Brennan, C. M., Jacobson, M. M., Matthes, C., & Walker, T. B. (2024). Monetary Policy Shocks: Data or Methods?
- Hambur, J., & Haque, Q. (2023). Can we use high-frequency yield data to better understand the effects of monetary policy and its communication? Yes and no!.
- Cieslak, A., & Schrimpf, A. (2019). Non-monetary news in central bank communication. Journal of International Economics, 118, 293-315.
- Cui, J., In, F., & Maharaj, E. A. (2016). What drives the Libor–ois spread? evidence from five major currency libor–ois spreads. International Review of Economics & Finance, 45, 358–75.
- Schrimpf, A., & Sushko, V. (2019). Beyond LIBOR: a primer on the new reference rates.
- Kubota, H., & Shintani, M. (2022). High-frequency identification of monetary policy shocks in Japan. The Japanese Economic Review, 73(3), 483–513.
- Kutai, A. (2020). Measuring the effect of forward guidance in small open economies: the case of Israel. Available at SSRN 3882421.
- Mertens, K., & Ravn, M. O. (2013). The dynamic effects of personal and corporate income tax changes in the United States. American economic review, 103(4), 1212–47.
- Ramey, V. A. (2016). Macroeconomic shocks and their propagation. Handbook of macroeconomics, 2, 71–162.
- Stock, J. H., & Watson, M. W. (2012). Disentangling the Channels of the 2007-2009 Recession (No. w18094). National Bureau of Economic Research.

### IMF and policy-oriented contributions, remarks, and working papers
- Barrett, M. P., & Platzer, J. (2024). Has the Transmission of US Monetary Policy Changed Since 2022?. International Monetary Fund.
- Borner, D. (2025). Central bank information and pure monetary policy surprises in Switzerland.
- Carrière-Swallow, Y., Kindberg-Hanlon, G., & Smirnov, D. (forthcoming). “Macroeconomic Effects and Spillovers from Bank of Japan Unconventional Monetary Policy,” IMF Working Paper. International Monetary Fund: Washington, DC.
- D’Amico, S., & King, T. B. (2023). Past and future effects of the recent monetary policy tightening. Chicago Fed Letter, 483.
- Deb, M. P., Estefania-Flores, J., Firat, M., Furceri, D., & Kothari, S. (2023). Monetary policy transmission heterogeneity: cross-country evidence. International Monetary Fund.
- Gopinath, G. (2023), Three Uncomfortable Truths For Monetary Policy, Remarks by IMF First Deputy Managing Director Gita Gopinath for the European Central Bank Forum on Central Banking 2023.
- Bolhuis, M., Das, S., & Yao, B. (2024). A New Dataset of High-Frequency Monetary Policy Shocks, IMF Working Paper.

*References list excerpt from "Still Packing a Punch: Monetary Policy Transmission in a New Cross-Country High-Frequency Dataset" (References section).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026018-source-pdf.pdf_
