## wp1765 — Section 3: Indicators of Monetary Policy

## Source details

**Canonical URL:** [wp1765 — Section 3: Indicators of Monetary Policy](https://www.imf.org/-/media/files/publications/wp/2017/wp1765.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp1765.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp1765.pdf.json)

---

### Introduction: scope and central question
- Sample and period:
  - 994 publicly listed financial institutions in 22 countries (19 advanced economies plus Brazil, Mexico, and South Africa) from 1998 Q1 to 2014 Q4.
- Core research questions:
  - Does prolonged monetary policy easing increase domestic financial-system vulnerability?
  - Does prolonged U.S. monetary policy easing affect vulnerabilities of financial systems abroad?
- Key descriptive findings (from Figure 1):
  - Banks: baseline median asset-to-equity (A/E) = 10.5; after two consecutive years of policy easing A/E = 12.5.
  - Insurance companies: baseline median A/E = 6.5; after two years of easing A/E = 7.4.
  - U.S. policy easing produces effects on non-U.S. firms that are approximately as large — or larger — than own-country easing.

### Measures of prolonged monetary policy easing and financial vulnerability
- Policy easing (duration) definition:
  - Duration D_t is number of consecutive quarters with declines in the 8-quarter moving average of an interest rate series.
  - D_t := { D_{t-1} + 1 if MA_t < MA_{t-1}; 0 otherwise }, where MA_t := (1/8) ∑_{τ=1}^8 i_{t-τ+1}.
- Interest-rate measures used (four):
  1. nominal short-term rate
  2. real short-term rate
  3. nominal two-year sovereign yields (preferred indicator)
  4. nominal 10-year sovereign yields
- Rationale for preferring Y2 yields:
  - Short rates lose informativeness near the zero lower bound; Y10 yields are less sensitive pre-crisis; two-year yields offer a balanced indicator of stance.
- Summary statistics of duration measures (country-level panel 1998 Q1–2014 Q4; Table 1):
  - Short rate: 25th = 3, 50th = 6, 75th = 10, Max = 37, observations = 857
  - Real short rate: 25th = 2, 50th = 5, 75th = 9, Max = 26, observations = 695
  - Y2 yields: 25th = 3, 50th = 6, 75th = 10, Max = 25, observations = 843
  - Y10 yields: 25th = 4, 50th = 8, 75th = 14, Max = 33, observations = 892
- Financial institution vulnerability measures:
  - Primary focus: leverage (assets-to-equity, A/E).
  - Alternatives for robustness: risk-adjusted ROE (ROE divided by its standard deviation over previous two years) and z-score (average ROA + average equity/assets, divided by standard deviation of ROA over previous two years).
- Leverage by industry (Table 2; firm-level medians):
  - Banks: 25th = 6.5, 50th = 10.5, 75th = 19.7, Number of firms = 245
  - Insurance companies: 25th = 3.2, 50th = 6.5, 75th = 12.6, Number of firms = 122
  - Investment banks: 25th = 3.2, 50th = 4.5, 75th = 12.2, Number of firms = 47
  - Asset managers: 25th = 1.2, 50th = 1.5, 75th = 2.1, Number of firms = 124
  - Real estates: 25th = 1.6, 50th = 2.0, 75th = 2.6, Number of firms = 370
  - Other financials: 25th = 1.5, 50th = 2.6, 75th = 5.2, Number of firms = 86

### The impact of own-country monetary policy easing on leverage
- Empirical specification (industry by industry):
  - ln(Y_ikt) = α_0 + α_1 D_kt + β X_{k,t-1} + c_{ki} + ε_ikt
  - Controls: lagged real GDP growth, stock price growth, volatility index, sovereign bond rating; country-industry fixed effects.
- Main marginal impacts (Y2 yields; effect of one additional quarter of domestic easing on A/E; Table 3):
  - Bank: 0.19***
  - Insurance: 0.08***
  - Investment Bank: 0.08***
  - Asset Manager: 0.01*
  - Real Estate: 0.00
  - Other: 0.02
- Interpretation of magnitude:
  - For banks, an additional quarter of easing raises leverage by 0.19; given median A/E = 10.5, this is roughly a 2 percent increase in leverage.
  - Nonbanks (insurance, investment banks) show economically and statistically significant increases in leverage as well.
- Nonlinearity:
  - Leverage increases are non-linear in duration: risk-taking rises most quickly at the onset of policy easing (marginal impact of an additional quarter is larger early in the easing spell).

### The impact of U.S. monetary policy easing on foreign financial firms (spillovers)
- Empirical specification for spillovers:
  - ln(Y_ikt) = α_0 + α_1 D_kt + α_US D_t^{US} + β X_{k,t-1} + c_{ki} + ε_ikt, estimated on non-U.S. sample.
- Main marginal impacts of U.S. duration (Y2 yields; one additional quarter of U.S. easing; Table 4):
  - Bank: 0.36***
  - Insurance: 0.10***
  - Investment Bank: 0.10**
  - Asset Manager: 0.01***
  - Real Estate: 0.02***
  - Other: 0.04***
- Median (A/E) for non-U.S. sample (reported):
  - Banks 15.9; Insurance 8.2; Investment Bank 5.8; Asset Manager 1.4; Real Estate 2.1; Other 2.4.
- Comparison with domestic effects:
  - For Y2 yields, U.S. easing effects are equal to domestic effects for some industries (investment banks, asset managers), larger for banks, and substantially larger for some other firms.
- Suggested transmission channels for U.S. spillovers:
  - Capital outflows from the U.S. (“global liquidity”) ease funding constraints abroad.
  - Large share of global trade and finance denominated in U.S. dollars; lower U.S. rates reduce global funding costs directly.

### Robustness checks and extensions
- Symmetry (tightening) results (Table 5; Y2 yield increases):
  - Domestic tightening example coefficients (opposite sign, roughly same magnitude):
    - Bank: -0.27***
    - Insurance: -0.14***
    - Investment bank: -0.09**
    - Asset manager: -0.01**
    - Real estate: -0.03***
    - Other: -0.05***
  - U.S. tightening (non-U.S. sample) example:
    - Bank: -0.55***; Insurance: -0.15***; Investment bank: -0.14**; Asset manager: -0.01***; Real estate: -0.05***; Other: -0.06***
- Macroprudential policies:
  - Controlling for Cerutti et al. (2015) MPI (0–12 index) produces virtually no change in baseline coefficients.
  - Example when controlling for MPI (domestic Y2 declines): Bank 0.22***; Insurance 0.09***; Investment bank 0.07***.
- Endogeneity (system GMM following Arellano-Bover and Blundell-Bond):
  - Results nearly unchanged; example (domestic Y2 declines, GMM): Bank 0.27***; Insurance 0.10***; Investment bank 0.07**.
- Alternative vulnerability measures and nonlinearity (Tables 6 and 7):
  - Risk-adjusted ROE (linear model, domestic):
    - Bank: -0.13***; Insurance: -0.08**; Investment bank: -0.07***; Asset manager: -0.05*; Real estate: -0.03*; Other: 0.02
  - Z-score (linear model, domestic):
    - Bank: -2.8***; Insurance: -0.7**; Investment bank: -0.1; Asset manager: -0.3; Real estate: -0.2; Other: 0.0
  - Nonlinear models: marginal impact of an additional quarter is larger at the onset (e.g., risk-adjusted ROE and z-score impacts much larger for 0 quarter current duration than for 8 quarters).
  - Results for alternative measures and nonlinearities are similar for domestic and U.S. duration specifications.

### Conclusions and policy implications
- Main findings:
  - Prolonged monetary policy easing increases leverage and other measures of vulnerability across banks and nonbanks.
  - Prolonged U.S. easing raises vulnerabilities in foreign financial firms by amounts comparable to domestic easing.
  - The relationship is nonlinear: the largest marginal increase in risk-taking occurs at the onset of easing.
  - Results robust to controls for macroprudential policy, endogeneity, and alternative vulnerability measures.
- Policy implications highlighted by authors:
  - Countries should closely monitor financial-sector risks during periods of monetary accommodation at home and in the U.S.
  - Develop solid prudential and regulatory frameworks to preserve monetary policy space for inflation and output objectives.
  - Frameworks should apply to both banks and nonbanks.
- Directions for further research:
  - Establish link between leverage increases and the probability or severity of crises to enable cost-benefit analysis of using monetary policy for financial stability.
  - Develop benchmarks for risk-taking behavior to judge whether observed increases in risk are excessive.

*Source: wp1765 — IMF Working Paper (Section 3: Indicators of Monetary Policy), authors’ calculations based on Bloomberg, Datastream, Haver, Worldscope, and WEO data (1998 Q1–2014 Q4).*

### References .............................................................................................................

### References

### Tables
- 1. Summary Statistics of the Duration Measures of Monetary ..................................................9
- 2. Leverage Ratio by Financial Industry ..................................................................................10
- 3. Marginal Impacts of the Duration of Domestic Monetary Easing on Leverage ..................12
- 4. Marginal Impacts of the Duration of U.S. and Domestic Monetary Easing ........................15
- 5. Robustness ...........................................................................................................................18
- 6. Alternative Measures of Financial Institution Vulnerability and Nonlinearities:Own 
  Country Duration ...............................................................................................................20
- 7. Alternative Measures of Financial Institution Vulnerability and Nonlinearities: U.S. 
  Duration .............................................................................................................................21

### Figures
- 1. Change in the Asset-to-Equity Ratio .....................................................................................5
- 2. Periods of Monetary Policy Easing for the U.S. ....................................................................8
- 3. Estimated Effect of the Duration of Domestic Monetary Policy Easing on Leverage ........13
- 4. Estimated Effect of the Duration of U.S. Monetary Policy Easing on Leverage ................16

### Appendix
- I. Data and Definitions of Variables ........................................................................................28

### Appendix Tables
- 1. Definitions and Sources .......................................................................................................29
- 2. Main Stock Indicators ..........................................................................................................30

*Source: wp1765 - References (IMF Working Paper PDF).*

### 3. Indicators of Monetary Policy .......................................................................................

### 3. Indicators of Monetary Policy

### Introduction: scope and central question
- Sample and period:
  - 994 publicly listed financial institutions in 22 countries (19 advanced economies plus Brazil, Mexico, and South Africa) from 1998 Q1 to 2014 Q4.
- Core research questions:
  - Does prolonged monetary policy easing increase domestic financial-system vulnerability?
  - Does prolonged U.S. monetary policy easing affect vulnerabilities of financial systems abroad?
- Key descriptive findings (from Figure 1):
  - For banks, baseline median asset-to-equity (A/E) = 10.5; after two consecutive years of policy easing A/E = 12.5.
  - For insurance companies, baseline median A/E = 6.5; after two years of easing A/E = 7.4.
  - U.S. policy easing produces effects on non-U.S. firms that are approximately as large — or larger — than own-country easing.

### Measures of prolonged monetary policy easing and financial vulnerability
- Policy easing (duration) definition:
  - Duration D_t is number of consecutive quarters with declines in the 8-quarter moving average of an interest rate series.
  - Four interest-rate measures used:
    1. nominal short-term rate
    2. real short-term rate
    3. nominal two-year sovereign yields (preferred indicator)
    4. nominal 10-year sovereign yields
  - Example formula: D_t := { D_{t-1} + 1 if MA_t < MA_{t-1}; 0 otherwise }, where MA_t := (1/8) ∑_{τ=1}^8 i_{t-τ+1}.
- Rationale for preferring Y2 yields:
  - Short rates lose informativeness near the zero lower bound; Y10 yields are less sensitive pre-crisis; two-year yields offer a balanced indicator of stance.
- Summary statistics of duration measures (Table 1; country-level panel 1998 Q1–2014 Q4):
  - Short rate: 25th = 3, 50th = 6, 75th = 10, Max = 37, observations = 857
  - Real short rate: 25th = 2, 50th = 5, 75th = 9, Max = 26, observations = 695
  - Y2 yields: 25th = 3, 50th = 6, 75th = 10, Max = 25, observations = 843
  - Y10 yields: 25th = 4, 50th = 8, 75th = 14, Max = 33, observations = 892
- Financial institution vulnerability measures:
  - Primary focus: leverage (assets-to-equity, A/E).
  - Alternatives used in robustness: risk-adjusted ROE (ROE divided by its standard deviation over previous two years) and z-score (average ROA + average equity/assets, divided by standard deviation of ROA over previous two years).
- Leverage by industry (Table 2; firm-level medians):
  - Banks: 25th = 6.5, 50th = 10.5, 75th = 19.7, Number of firms = 245
  - Insurance companies: 25th = 3.2, 50th = 6.5, 75th = 12.6, Number of firms = 122
  - Investment banks: 25th = 3.2, 50th = 4.5, 75th = 12.2, Number of firms = 47
  - Asset managers: 25th = 1.2, 50th = 1.5, 75th = 2.1, Number of firms = 124
  - Real estates: 25th = 1.6, 50th = 2.0, 75th = 2.6, Number of firms = 370
  - Other financials: 25th = 1.5, 50th = 2.6, 75th = 5.2, Number of firms = 86

### The impact of own-country monetary policy easing on leverage
- Empirical specification (industry by industry):
  - ln(Y_ikt) = α_0 + α_1 D_kt + β X_{k,t-1} + c_{ki} + ε_ikt
  - Controls: lagged real GDP growth, stock price growth, volatility index, sovereign bond rating; country-industry fixed effects.
- Main marginal impacts (Y2 yields; effect of one additional quarter of domestic easing on A/E; Table 3):
  - Bank: 0.19***
  - Insurance: 0.08***
  - Investment Bank: 0.08***
  - Asset Manager: 0.01*
  - Real Estate: 0.00
  - Other: 0.02
  - Median (A/E) reference levels used to compute ∂Y/∂D.
- Interpretation:
  - For banks, an additional quarter of easing raises leverage by 0.19; given median A/E = 10.5, this is roughly a 2 percent increase in leverage.
  - Nonbanks (insurance, investment banks) show economically and statistically significant increases in leverage as well.
- Nonlinearity:
  - Leverage increases are non-linear in duration: risk-taking rises most quickly at the onset of policy easing (illustrated in nonlinear specifications).

### The impact of U.S. monetary policy easing on foreign financial firms (spillovers)
- Empirical specification:
  - ln(Y_ikt) = α_0 + α_1 D_kt + α_US D_t^{US} + β X_{k,t-1} + c_{ki} + ε_ikt, estimated on non-U.S. sample.
- Main marginal impacts of U.S. duration (Y2 yields; one additional quarter of U.S. easing; Table 4):
  - Bank: 0.36***
  - Insurance: 0.10***
  - Investment Bank: 0.10**
  - Asset Manager: 0.01***
  - Real Estate: 0.02***
  - Other: 0.04***
  - Median (A/E) for non-U.S. sample reported: Banks 15.9; Insurance 8.2; Investment Bank 5.8; Asset Manager 1.4; Real Estate 2.1; Other 2.4.
- Comparison with domestic effects:
  - For Y2 yields, U.S. easing effects are equal to domestic effects for some industries (investment banks, asset managers), larger for banks, and substantially larger for some other firms.
- Suggested transmission channels for U.S. spillovers:
  - Capital outflows from the U.S. (“global liquidity”) ease funding constraints abroad.
  - Large share of global trade and finance denominated in U.S. dollars; lower U.S. rates reduce global funding costs directly.

### Robustness checks and extensions
- Symmetry (tightening) results (Table 5):
  - Y2 yield increases (i.e., tightening): coefficients roughly same magnitude, opposite sign. Example (domestic tightening):
    - Bank: -0.27***
    - Insurance: -0.14***
    - Investment bank: -0.09**
    - Asset manager: -0.01**
    - Real estate: -0.03***
    - Other: -0.05***
  - U.S. tightening (non-U.S. sample) example:
    - Bank: -0.55***; Insurance: -0.15***; Investment bank: -0.14**; Asset manager: -0.01***; Real estate: -0.05***; Other: -0.06***
- Macroprudential policies:
  - Controlling for Cerutti et al. (2015) MPI (0–12 index) produces virtually no change in baseline coefficients.
  - Example when controlling for MPI (domestic Y2 declines): Bank 0.22***; Insurance 0.09***; Investment bank 0.07***.
- Endogeneity (system GMM following Arellano-Bover and Blundell-Bond):
  - Results nearly unchanged; example (domestic Y2 declines, GMM): Bank 0.27***; Insurance 0.10***; Investment bank 0.07**.
- Alternative vulnerability measures and nonlinearity (Tables 6 and 7):
  - Risk-adjusted ROE (linear model, domestic):
    - Bank: -0.13***; Insurance: -0.08**; Investment bank: -0.07***; Asset manager: -0.05*; Real estate: -0.03*; Other: 0.02
  - Z-score (linear model, domestic):
    - Bank: -2.8***; Insurance: -0.7**; Investment bank: -0.1; Asset manager: -0.3; Real estate: -0.2; Other: 0.0
  - Nonlinear models show the marginal impact of an additional quarter is larger at the onset (e.g., risk-adjusted ROE and z-score impacts are much larger for 0 quarter current duration than for 8 quarters).
  - Results for alternative measures and nonlinearities are similar for domestic and U.S. duration specifications.

### Conclusions and policy implications
- Main findings:
  - Prolonged monetary policy easing increases leverage and other measures of vulnerability across banks and nonbanks.
  - Prolonged U.S. easing raises vulnerabilities in foreign financial firms by amounts comparable to domestic easing.
  - The relationship is nonlinear: the largest marginal increase in risk-taking occurs at the onset of easing.
  - Results robust to controls for macroprudential policy, endogeneity, and alternative vulnerability measures.
- Policy implications highlighted by authors:
  - Countries should closely monitor financial-sector risks during periods of monetary accommodation at home and in the U.S.
  - Develop solid prudential and regulatory frameworks to preserve monetary policy space for inflation and output objectives.
  - Frameworks should apply to both banks and nonbanks.
- Directions for further research (authors’ suggestions):
  - Establish link between leverage increases and the probability or severity of crises to enable cost-benefit analysis of using monetary policy for financial stability.
  - Develop benchmarks for risk-taking behavior to judge whether observed increases in risk are excessive.

*Source: IMF Working Paper (Section 3: Indicators of Monetary Policy), authors’ calculations based on Bloomberg, Datastream, Haver, Worldscope, and WEO data (1998 Q1–2014 Q4).*

### REFERENCES

### wp1765 - REFERENCES

### Key literature cited
- Extensive bibliography on monetary policy, financial stability, and the risk-taking channel, including but not limited to contributions by:
  - Acharya, Viral V.; Pedersen, Lasse H.; Philippon, Thomas; Richardson, Matthew (2010) — “Measuring Systemic Risk.”
  - Adrian, Tobias; Liang, Nellie (2014) — “Monetary Policy, Financial Conditions, and Financial Stability,” Staff Report, No. 690.
  - Adrian, Tobias; Shin, Hyun Song (2009, 2010) — “Money, Liquidity and Monetary Policy” and “Liquidity and Leverage.”
  - Borio, Claudio; Zhu, Haibin (2012) — “Capital Regulation, Risk-Taking and Monetary Policy.”
  - Cerutti, Eugenio; Claessens, Stijn; Laeven, Luc (2015) — “The Use and Effectiveness of Macroprudential Policies: New Evidence,” IMF Working Paper WP/15/61.
  - Dell’Ariccia, Giovanni; Laeven, Luc; Suarez, Gustavo (2013) — “Bank Leverage and Monetary Policy’s Risk-Taking Channel,” IMF Working Paper WP/13/143.
  - Jiménez, Gabriel; Ongena, Steven; Peydró, José-Luis; Saurina, Jesús (2014) — “Hazardous Times for Monetary Policy,” Econometrica, Vol. 82, No. 2, pp. 463–505.
  - Numerous working papers and journal articles covering topics such as large-scale asset purchases, zero lower bound effects, credit cycles, real interest rates, leverage, and bank risk-taking.

### Thematic emphases in cited works
- Monetary policy transmission and its interactions with financial stability and leverage.
- Risk-taking channel of monetary policy across banks and nonbank financial institutions.
- Effects of unconventional monetary policy, including large-scale asset purchases (LSAPs), on market yields and corporate credit risk.
- Macroprudential policy effectiveness and interaction with monetary policy.
- Panel data methods and estimation techniques for dynamic panels and spatial dependence.

### Appendix I. Data and definitions of variables — dataset overview
- Panel data set on publicly listed financial firms in 22 countries covering 1998 Q1 to 2014 Q4.
- Coverage:
  - Six industries based on Global Industry Classification Standard: banks, insurance companies, investment banks, asset managers, and other financials.
  - Sample countries: 19 advanced economies (Austria, Australia, Belgium, Canada, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, Portugal, Republic of Korea, Spain, Sweden, Switzerland, the United Kingdom, and the U.S.) and three emerging market economies (Brazil, Mexico, and South Africa).
- Data sources and construction:
  - Firm-level financial data from Worldscope (Thomson Reuters); market capitalization from Datastream.
  - Total of 1,039 listed financial firms available in Worldscope.
  - Sample used: 38,883 firm-quarter observations based on 994 firms whose indicators of financial vulnerabilities are available.
  - Firm-level variables are merged with country-level macroeconomic indicators.
- Novelty and purpose:
  - Dataset covers both banking and nonbank financial sectors, enabling analysis of heterogeneous effects of monetary policy across industries.
  - Sufficient cross-firm and cross-country variation to identify effects of the duration of monetary policy easing on financial vulnerabilities.

### Appendix Table 1 — Firm-level and country-level variable definitions and sources
- Firm-level variables:
  - Leverage ratio: The "market" asset-to-equity ratio calculated as the sum of total liabilities and market capitalization divided by market capitalization. — Computed
  - Risk-adjusted return on equity: The returns on equity divided by its standard deviation over the past 8 quarters (t-7 to t). — Computed
  - Total liability: All short and long term obligations expected to be satisfied by the company (Field 03351). — Worldscope
  - Market capitalization: The share price multiplied by the number of ordinary shares in issue (Field MV). — Datastream
  - Returns on equity: Net income divided by total equity (Field 08301). — Worldscope
- Country-level macroeconomic variables:
  - Real GDP growth: The year-on-year percent change of the real GDP. — WEO
  - Growth of stock price index: The year-on-year percent change of the natural log of the main stock indicator. — Datastream
  - Volatility index: Estimated time-varying volatility of the main stock indicator using a GARCH(1,1). — Datastream
  - Sovereign bond rating: Moody's Local Currency Long-Term Debt Rating. Values from 1 to 22 are assigned so that a higher value indicates a better rating (e.g., 22 indicates "AAA"). — Bloomberg
  - Three-month yields: The three-month government bond yields. For countries where three-month government bonds are not available, three-month interbank rate is used. — Bloomberg, Datastream, Haver
  - Inflation: The year-on-year percent change of the Consumer Price Index (CPI). — Haver
  - Output gap: The output gap estimates provided by Oxford economics. — Datastream
  - Macroprudential policy index: The number of macroprudential measures used out of the total of 12 macroprudential measures. — Cerutti et al. (2015)

### Appendix Table 2 — Main stock indicators (country to indicator mapping)
- Austria: ATX—Austrian Traded Index (Field ATXINDX) — Datastream
- Australia: Standard and Poor's / Australian Stock Exchange 300 (Field ASX300I) — Datastream
- Belgium: Belgium 20 (Field BGBEL20) — Datastream
- Brazil: Brazil Bovespa (Field BRBOVES) — Datastream
- Canada: Standard and Poor's / Toronto Stock Exchange 60 Index (Field TTOSP60) — Datastream
- Switzerland: Swiss Market (SMI) (Field SWISSMI) — Datastream
- Germany: DAX 30 (Field DAXINDX) — Datastream
- Spain: IBEX 35 (Field IBEX35I) — Datastream
- Finland: OMX Helsinki 25 (OMXH25) (Field HEX25IN) — Datastream
- France: France CAC 40 (Field FRCAC40) — Datastream
- United Kingdom: FTSE 100 (Field FTSE100) — Datastream
- Greece: FTSE / Athex Large Capital (Field FTASE20) — Datastream
- Ireland: Iseq 20 (Field ISECP20) — Datastream
- Italy: FTSE MIB Index (Field FTSEMIB) — Datastream
- Japan: TOPIX (Field TOKYOSE) — Datastream
- Korea, Republic of: Korea Stock Exchange Composite (KOSPI) (Field KORCOMP) — Datastream
- Mexico: Mexico IPC (Bolsa) (Field MXIPC35) — Datastream
- Netherlands: AEX Index (AEX) (Field AMSTEOE) — Datastream
- Portugal: PSI-20 (Field POPSI20) — Datastream
- Sweden: OMX Stockholm 30 (OMXS30) (Field SWEDOMX) — Datastream
- U.S.: Standard and Poor's 500 Composite (Field S&PCOMP) — Datastream
- South Africa: FTSE / JSE Top 40 (Field JSEAL40) — Datastream

### Measures of monetary policy stance and construction
- Duration of monetary policy easing and tightening:
  - Defined as the number of consecutive quarters with a decline in the eight quarter moving average of interest rates (see Section 2.1).
  - Use of the eight quarter moving average is intended to avoid falsely associating easing with temporary fluctuations in interest rates.
  - Results are robust to the use of moving averages over the past one year and the past three years.

### Appendix Table 3 — Indicators of monetary policy (definitions and sources)
- Interest rates:
  - Nominal short rate: The overnight interbank interest rate. — Haver
  - Real short rate: The ex-post real short rate, defined as the difference between the nominal short rate and the realized CPI inflation rate. — Computed
  - Two-year yields: The two-year government bond yields. — Bloomberg, Datastream
  - 10-year yields: The 10-year government bond yields. — Bloomberg, Datastream

*Source: wp1765 - REFERENCES (IMF Working Paper).*>

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp1765.pdf_