## 1.     Monetary Transmission Through Lending Rates

## Source details

**Canonical URL:** [1.     Monetary Transmission Through Lending Rates](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13251.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13251.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13251.pdf.json)

---

### Introduction
- During the past two years, private banks’ credit seems to have been less responsive to monetary policy changes than in previous loosening cycles.
- In August 2011, the Brazilian central bank started an easing cycle.
- Since then and until the start of the tightening cycle in April 2013, the policy rate was cut by 525 bps.
- Timeframe referenced in figures: Jun-02 through Feb-13.

### Stylized facts
- Credit growth and monetary policy
  - Credit growth by private banks continued to decline despite substantial monetary stimulus; new lending operations were recovering only very gradually.
  - The surge in NPLs since late-2011 likely contributed to more cautious and limited credit supply by private banks.
  - Demand-side constraints: weak consumer and business confidence, elevated household indebtedness, and relatively high debt service ratios.
  - Public banks expanded credit at an annual rate of nearly 25 percent during 2011–12, increasing their share of total credit to 48 percent (compared to 42 percent by end-2010).
- Interest rate pass-through and borrower heterogeneity
  - The pass-through of the policy rate to loan rates was ultimately complete. During the last easing cycle, pass-through was initially delayed but private banks fully passed policy cuts to lending rates for firms and individuals, with tightening of spreads for some loan segments.
  - The negative correlation between Selic changes and credit growth weakened in the recent monetary easing cycle, particularly for bank lending to corporates.
- Credit demand indicators
  - Serasa indicator: consumer credit requests were relatively weak during 2012 and picked up in early 2013; corporate demand for credit weakened since end-2011 and remained soft.
- Bank ownership patterns
  - Time series of credit growth by bank ownership shows differing behavior for public banks and private banks over the sample.

### Methodology and data
- Dataset and sample
  - Panel quarterly data on 37 private banks (19 domestic and 18 foreign), sample period 2005Q1–2012Q4.
  - Dependent variable: growth in new lending by private banks (new loans to sector i by private bank j at time t).
- Empirical specification (as reported)
  - Δln x_ijt = β0 + β1 ΔSelic_t + β2 Δln public_it + β2 k_it + β3 A_ijt + β4 Z_jt + υ_ij + ε_ijt
  - Definitions:
    - x_ijt = new loans to sector i (corporate or individual) by private bank j at time t.
    - Selic = Brazil Central Bank policy rate.
    - public_it = new loans by public banks to sector i.
    - k_it = demand proxies (confidence index, expectations index, Serasa credit requests).
    - A_ijt = credit risk factors (bank- and sector-specific NPL ratio; common factors: VIX, EMBI, Bovespa, household debt service).
    - Z_jt = lending capacity constraints (bank capitalization, liquidity; reserve requirements).
    - υ_ij = bank-specific fixed effect; ε_ijt = error term.
  - Dummy δ_t equals 1 during 2012Q1–2012Q4 to identify whether private banks’ new credit growth was lower during the last easing cycle and to interact with other variables (e.g., ΔSelic) to test changed sensitivities.
- Data sources: Brazil Central Bank dataset on new credit concessions and bank balance-sheet items; Haver; Serasa Experian Brazil.

### Results — main findings
- Overall lending dynamics
  - Lending growth was weaker since end-2011.
  - Changes in the Selic have a negative and statistically significant impact on credit growth.
  - The dummy δ_t coefficient is negative and significant, indicating lower lending growth during the past cycle.
- Monetary transmission and interactions
  - Interaction δ_t · ΔSelic: positive and statistically significant in Column II (suggesting weaker monetary transmission during the recent cycle), but coefficient sign reverts and is significant in Column III when δ_t is included, implying reductions in Selic did have a positive impact on loan disbursements after controlling for other factors.
  - No statistically significant difference between foreign and domestic private banks in transmission (δ_t · Foreign bank dummy not significant).
  - Evidence indicates the lending channel was not weakened; after controlling for demand and supply shifts, the lending channel was more effective than during the rest of the sample (negative sign on Selic, interaction results in Table 4.c).
  - Similar results when restricting sample to individuals or corporates: interaction terms not statistically significant, suggesting similar transmission across borrower types.
- Demand and supply determinants (Table 4)
  - Demand proxies (confidence index, expectations index, Serasa credit requests) each have positive and significant impacts on credit growth.
  - Market risk measures (EMBI, VIX, or declines in Bovespa) exert negative impacts on credit growth.
  - Deterioration in credit portfolio (increase in NPL or decrease in ROA) reduces credit supply.
  - Reserve requirements coefficient positive (indicative of endogeneity: central bank raises reserve requirements in response to excess credit growth).
  - Bank capitalization and liquidity show expected positive signs and are statistically significant.
- Bank caution and sensitivity (Table 5)
  - Banks became more sensitive in lending to changes in monetary policy during the past easing cycle (δ_t · ΔSelic negative and significant in some specifications).
  - Banks also showed heightened sensitivity to macro/global environment (δ_t · EMBI negative and significant).
  - Interaction terms with NPL and other variables generally have expected signs but are often not statistically significant, supporting increased caution.
- Public banks’ impact (Table 6)
  - When lending to individuals, public banks’ lending moves in tandem with private banks’ lending (dlnpublic positive and significant), suggesting different products or borrower types.
  - Competition (substitution) exists between private and public banks for corporate lending: dlnpublic negative effect on private lending to corporates in some specifications.
  - Interaction δ_t · dlnpublic generally negative but not statistically significant, providing no strong evidence that competition intensified during 2012Q1–2012Q4 beyond existing patterns.
  - Product-level results: public banks’ lending increases associate positively with private lending in individual consumption and negatively or null in some corporate or working capital categories.
- Representative estimated coefficients (preserved from tables)
  - dlnx_t-1 coefficients range: -0.13 to -0.30 across regressions (often significant at *** p<0.01).
  - ΔSelic coefficients typically negative and significant (examples: -0.06 (3.85)***; -0.04 (2.78)***).
  - Δnpl coefficients often negative and significant (example: -0.04 (3.57)***).
  - dlnpublic t-1 when lending to individuals: 0.36 (2.71)***; when lending to corporates: -0.13 (2.57)**.

### Conclusions and policy-relevant interpretations
- Effectiveness of monetary transmission
  - Monetary policy transmission worked efficiently during the last monetary easing cycle: neither lending rates nor volumes transmission was impaired.
  - Private bank lending weakened since end-2011 despite substantial policy cuts; however, after accounting for demand and supply factors, Selic reductions positively affected disbursements and the lending channel remained operative.
- Drivers of weak lending outcomes
  - Demand-side: shifts in demand for credit (weaker confidence and lower credit requests).
  - Supply-side: tighter bank profitability and capitalization, surge in non-performing loans, and heightened market risk prompting greater caution and selectivity by banks.
  - Public banks’ rapid expansion of credit likely contributed to weaker private bank lending; competition/substitution is evident in some segments (notably corporates), but there is no strong evidence that competition intensified during 2012Q1–2012Q4 beyond existing patterns.
- Implications for monitoring and policy
  - Importance of disentangling demand versus supply influences when assessing the impact of policy rate changes on lending volumes.
  - Monitor NPLs, bank capitalization, liquidity, and market risk indicators (EMBI, VIX, Bovespa) to assess banks’ willingness to extend credit.
  - Track public banks’ targeted lending to understand potential displacement effects across borrower types and product categories.

*Source: IMF Fund staff calculations and analysis in _wp13251 - 1.     Monetary Transmission Through Lending Rates (PDF chapter).*

### 1.     Monetary Transmission Through Lending Rates ..............................................................5

### 1.     Monetary Transmission Through Lending Rates

### Introduction
- During the past two years, private banks’ credit seems to have been less responsive to monetary policy changes than in previous loosening cycles.
- In August 2011, the Brazilian central bank started an easing cycle.
- Since then and until the start of the tightening cycle in April 2013, the policy rate was cut by 525 bps.

### Structure and Major Sections (as presented)
- 2.a. Descriptive Statistics
- 2.b. Definition of Variables
- 3. Monetary Transmission and Changes during the 2012 Easing Cycle
- 4. Impact of Selected Factors on Lending Growth
- 5. Testing for Changes in the Sensitivity of Lending Growth to Factors
- 6. Impact of Public Banks’ Lending

### Figures and Data Highlights (as presented)
- Figure 1. Brazil. Credit Growth and Monetary Policy (In percent)
  - Series shown (per figure caption and axis labels):
    - Monetary policy rate (RHS)
    - Outstanding private sector credit by private banks, yoy growth
    - New referenced credit operations in past 12 months, yoy growth
  - Sources: Central Bank of Brazil; and Fund staff calculations.
- Figure listing includes:
  - 1. Credit Growth and Monetary Policy (In percent)
  - 2. By Borrower, Change in Selic and Lending Growth
  - 3. Serasa Indicator of Credit Requests (Growth, yoy, in percent)
  - 4. Credit Growth by Bank Ownership (yoy, in percent)
  - 5. Description of the Variables

### Key Observations (from provided text and figure captions)
- The material focuses on monetary transmission through lending rates in Brazil, examining credit growth, monetary policy rates, and new credit operations.
- There is an explicit emphasis on differences in responsiveness by bank ownership (private banks vs. public banks) and by borrower type (implied by figure titles).
- The analysis includes descriptive statistics and explicit definitions of variables used for empirical work.
- The timeframe explicitly referenced includes Jun-02 through Feb-13 on the figure x-axis, and policy actions between August 2011 and April 2013.

*Source: _wp13251 - 1.     Monetary Transmission Through Lending Rates ..............................................................5*

### 7.25 percent, a decade’s low. Despite the

### _wp13251 - 7.25 percent, a decade’s low. Despite the

### II. Stylized facts
- Credit growth by private banks continued to decline despite substantial monetary stimulus; new lending operations were recovering only very gradually.
- The surge in NPLs since late-2011 likely contributed to more cautious and limited credit supply by private banks.
- Demand-side factors holding back credit demand: weak consumer and business confidence, elevated household indebtedness, and relatively high debt service ratios.
- Expansion of credit by public banks may have competed with private banks; public banks’ credit expanded at an annual rate of nearly 25 percent during 2011–12, increasing their share of total credit to 48 percent (compared to 42 percent by end-2010).
- Figure 2 (described): The negative correlation between Selic changes and credit growth weakened in the recent monetary easing cycle, particularly for bank lending to corporates.
- Interest rate pass-through: Pass-through of the policy rate to loan rates was ultimately complete. During the last easing cycle, pass-through was initially delayed but private banks fully passed policy cuts to lending rates for firms and individuals, with tightening of spreads for some loan segments.
- Serasa indicator: Consumer credit requests were relatively weak during 2012 and picked up in early 2013; corporate demand for credit weakened since end-2011 and continued soft.
- Figure 4 (described): Time series of credit growth by bank ownership shows public banks and private banks differing behavior over sample.

### III. Methodology and data
- Panel dataset: quarterly data on 37 private banks (19 domestic and 18 foreign), sample period 2005Q1–2012Q4.
- Dependent variable: growth in new lending by private banks (new loans to sector i by private bank j at time t).
- Empirical specification:
  - Δln x_ijt = β0 + β1 ΔSelic_t + β2 Δln public_it + β2 k_it + β3 A_ijt + β4 Z_jt + υ_ij + ε_ijt
  - x_ijt = new loans to sector i (corporate or individual) by private bank j at time t.
  - Selic = Brazil Central Bank policy rate.
  - public_it = new loans by public banks to sector i.
  - k_it = demand proxies (confidence index, expectations index, Serasa credit requests).
  - A_ijt = credit risk factors (bank- and sector-specific NPL ratio; common factors: VIX, EMBI, Bovespa, household debt service).
  - Z_jt = lending capacity constraints (bank capitalization, liquidity; reserve requirements).
  - υ_ij = bank-specific fixed effect; ε_ijt = error term.
- Dummy δ_t equals 1 during 2012Q1–2012Q4 to identify if private banks’ new credit growth was lower during the last easing cycle and to interact with other variables (e.g., ΔSelic) to test changed sensitivities.
- Data sources: Brazil Central Bank dataset on new credit concessions and bank balance-sheet items; Haver; Serasa Experian Brazil.

### IV. Results
- Table 3 (summary):
  - Lending growth was weaker since end-2011.
  - Changes in the Selic have a negative and statistically significant impact on credit growth.
  - Dummy δ_t coefficient negative and significant, indicating lower lending growth during the past cycle.
  - Interaction δ_t · ΔSelic: positive and statistically significant in Column II (suggesting weaker monetary transmission during recent cycle), but coefficient sign reverts and is significant in Column III when δ_t is included, implying reductions in Selic did have a positive impact on loan disbursements after controlling for other factors.
  - No statistically significant difference between foreign and domestic private banks in transmission (δ_t · Foreign bank dummy not significant).
- Demand and supply factors (Table 4):
  - Demand proxies (confidence index, expectations index, Serasa credit requests) each have positive and significant impacts on credit growth.
  - Market risk measures (EMBI, VIX, or declines in Bovespa) exert negative impacts on credit growth.
  - Deterioration in credit portfolio (increase in NPL or decrease in ROA) reduces credit supply.
  - Reserve requirements coefficient positive (indicative of endogeneity: central bank raises reserve requirements in response to excess credit growth).
  - Bank capitalization and liquidity show expected positive signs and are statistically significant.
- Lending channel during last easing cycle:
  - Evidence indicates the lending channel was not weakened; after controlling for demand and supply shifts, the lending channel was more effective than during the rest of the sample (negative sign on Selic, interaction results in Table 4.c).
  - Similar results when restricting sample to individuals or corporates: interaction terms not statistically significant, suggesting similar transmission across borrower types.
- Increased caution by banks (Table 5):
  - Banks became more sensitive in lending to changes in monetary policy during the past easing cycle (δ_t · ΔSelic negative and significant in some specifications).
  - Banks also showed heightened sensitivity to macro/global environment (δ_t · EMBI negative and significant).
  - Interaction terms with NPL and other variables have expected signs but are often not statistically significant, supporting increased caution.
- Public banks’ impact (Table 6):
  - When lending to individuals, public banks’ lending moves in tandem with private banks’ lending (dlnpublic positive and significant), suggesting different products or borrower types.
  - Competition (substitution) exists between private and public banks for corporate lending: dlnpublic negative effect on private lending to corporates in some specifications.
  - Interaction δ_t · dlnpublic generally negative but not statistically significant, providing no strong evidence that competition intensified during the recent period.
  - Product-level results: public banks’ lending increases associate positively with private lending in individual consumption and negatively or null in some corporate or working capital categories (see Table 6.b specifics).

- Key estimated coefficients and statistical significance preserved in tables (examples):
  - dlnx_t-1 coefficients range: -0.13 to -0.30 across regressions (often significant at *** p<0.01).
  - ΔSelic coefficients typically negative and significant (examples: -0.06 (3.85)***; -0.04 (2.78)***).
  - Δnpl coefficients often negative and significant (example: -0.04 (3.57)***).
  - dlnpublic t-1 when lending to individuals: 0.36 (2.71)***; when lending to corporates: -0.13 (2.57)**.

### V. Conclusions
- Monetary policy transmission worked efficiently during the last monetary easing cycle: neither lending rates nor volumes transmission was impaired.
- Private bank lending weakened since end-2011 despite substantial policy cuts; however, after accounting for demand and supply factors, Selic reductions positively affected disbursements and the lending channel remained operative.
- The diminished lending outcome is attributable to:
  - Shifts in demand for credit (weaker confidence and lower credit requests).
  - Supply-side pressures: tighter bank profitability and capitalization, surge in non-performing loans, and heightened market risk prompting greater caution and selectivity by banks.
- Public banks’ rapid expansion of credit likely contributed to weaker private bank lending; while competition/substitution is evident in some segments (notably corporates), no strong evidence that competition intensified during 2012Q1–2012Q4 beyond existing patterns.

*Source: IMF Fund staff calculations and analysis in _wp13251 - 7.25 percent, a decade’s low. Despite the (PDF chapter).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13251.pdf_