## _wp1262 ## Source details **Canonical URL:** [_wp1262](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1262.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1262.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1262.pdf.json) --- ### I. Introduction and scope - Purpose: studies reasons why Brazil has relatively high real interest rates and provides insights into possible factors that may help reduce them over time. - Focus: determinants of the evolution of the (ex-post) short-term real interest rate (the policy rate set by the Central Bank adjusted for inflation). - Exclusion: leaves for future research the question of why Brazil has high intermediation spreads in the banking system. - Sample and period: panel data set of 15 emerging market inflation targeting (IT) countries over the 1980–2009 period. - Note on coverage: takes into account monetary policy stances adopted in the aftermath of the 2008 financial crisis; does not cover 2010 and 2011. ### Trends in Brazil’s ex-post short-term real interest rate - Historical trajectory (excluding hyperinflation period 1988–1994): - About 40 percent in the 1980s. - About 20 percent in the second half of the 1990s (prior to inflation targeting and floating of the currency in 1999). - Annual average of about 10 percent during 2000–2005. - Below 8 percent over the 2006–2009 period. - Reached the lowest historical level (just below 5 percent) in 2009. - Comparative position: in 2000–2009 Brazil’s real interest rates were about four percentage points above the average of emerging market inflation-targeting regimes. ### Theoretical discussion and stylized facts - Five groups of explanations for historically high interest rates in Brazil: - (A) Fiscal considerations - (B) Low domestic savings - (C) Institutional weaknesses - (D) Previous history of high inflation and inflation volatility - (E) Factors affecting the monetary policy transmission mechanism - A. Fiscal considerations - Two types of fiscal dominance (Coates and Rivera (2004)): - Type I. Monetary Subordination: monetary policy directed at financing fiscal deficit through money creation (relevant in hyperinflation; not an issue in last 15 years). - Type II. Crowding Out Effect in the Credit Market: fiscal deficit financed in domestic local-currency capital markets may bid up interest rates. - Empirical evidence is mixed: - Muinhos and Nankane (2006) find no evidence of a negative relationship between public debt levels and the real interest rate. - Simple bivariate analysis does not show a clear relationship; inclusion of gross public debt in panel regressions does not produce robust results and sometimes yields opposite sign. - Fiscal discipline improved over time: gradual reduction of the overall fiscal deficit due to sustained high primary surpluses and fiscal responsibility law approved in 2000 likely contributed to lower real interest rates. - B. Domestic savings - Brazil has relatively low domestic savings; classical intuition: if investment demand exceeds domestic savings supply, equilibrium real interest rate increases. - Cross-country stylized facts: - Korea, Indonesia, Thailand around 30 percent of GDP have lower real interest rates. - Chile and Mexico: average domestic savings 6–7 percentage points higher than Brazil and considerably lower real interest rates. - Brazil and Turkey appear as outliers: lower domestic savings and higher real interest rates than linear prediction (linear prediction would suggest average interest rates about four percentage points lower than actually observed for Brazil). - C. Institutional weaknesses - Jurisdictional Uncertainty (anticreditor bias) not supported by empirical evidence (Goncalves, Holland, and Spacov (2007)). - Lack of full Central Bank independence theoretically relevant but difficult to test; Nahon and Meuer (2009) find recent Board changes did not undermine Central Bank credibility. - D. History of inflation and inflation volatility - Long history: - 1970s: annual inflation averaging 30 percent - 1980–88: averaging over 200 percent - 1989–1994: hyperinflation averaging 1,400 percent - Correlation: high inflation levels associated with high real interest rates. - Post-1995: largest reduction in inflation volatility among sample IT emerging markets; by 2006 inflation volatility converged to the group average. - Stylized statistic: an increase of one standard deviation in inflation volatility is associated with an increase in the real interest rate of about 1.4 percentage points. For Brazil, reduction in inflation volatility over time is associated with a reduction in the real interest rate of over two percentage points. - E. Factors affecting the monetary policy transmission mechanism - Credit market segmentation: subsidized public lending (BNDES, housing, agricultural sectors) may raise equilibrium real interest in the free credit market by reducing ability of policy rate to control aggregate credit; direct transfers to BNDES that increased lending capacity by over 6 percent of GDP began in 2008 and expanded in 2009–10 (cannot be fully tested in sample period). - Minimum remuneration requirements in savings accounts (cadernetas de poupança): remunerated at about 6 percent plus a reference rate and tax-exempt interest income; acts as a floor for the policy rate (Selic). - Inflation target midpoint and bands: midpoint 4.5, with upper and lower bands of 6.5 and 2.5 percent; increases nominal rates for a given equilibrium real rate but target per se not associated with higher real rates. - Other factors: inflation inertia via indexation, weakness of wealth channel due to indexation and short maturity of public debt, relatively low credit-to-GDP ratio compared with other emerging markets. ### Econometric analysis — sample, data, and model - Sample: emerging market IT regimes (Brazil, Chile, Colombia, Czech Republic, Hungary, Indonesia, Korea, Mexico, Peru, Philippines, Poland, Romania, South Africa, Thailand, Turkey). - Excludes advanced IT regimes (Australia, Canada, Iceland, Israel, New Zealand, Norway, Sweden and the U.K.). - Data sources: International Finance Statistics (IFS) and World Economic Outlook (WEO) for gaps. - Time period: 1980–2009 (panel unbalanced). - Model: panel error correction model (ECM). - Dependent variable: ex-post short-term real interest rate, it, - Independent variables vector C includes: - PPPCAP: per-capita GDP in purchasing power parity values - CAB_Y: current account balance in percent of GDP - PRIVSAVINGS_Y: private savings in percent of GDP - GOVBAL_Y: overall fiscal balance of the general government in percent of GDP - FED_RATE: federal funds rate (control for international monetary conditions) - INFVOL: inflation volatility - IT: dummy for adoption of an inflation-targeting regime - Estimation methods: OLS with Huber iterations (robust regression); model also estimated with fixed effects and alternative techniques (panel-corrected standard errors, White-corrected standard errors, Arellano-Bond GMM, GLS). ### Main econometric results and selected coefficient estimates (dependent variable D.REALINT) - Broad findings: - Real interest rate declines when private savings increase. - Real interest rate declines when overall government fiscal and external current account balances improve. - Increases in inflation volatility are associated with increases in the real interest rate. - Adoption of an inflation-targeting regime results in a significant short-term reduction in real interest rates. - International monetary conditions matter (federal funds rate coefficients reported below). - Selected coefficient estimates (Table 1; exact values preserved): - L.REALINT: Coefficient = - .31207*** ; Standard Error = .0122 ; t = -25.48 ; P>|t| = 0.000 - L.PPPCAP: Coefficient = 0.00000 ; Standard Error = .0000 ; t = 0.04 ; P>|t| = 0.971 - D.PPPCAP: Coefficient = 0.00028 ; Standard Error = .0006 ; t = 0.45 ; P>|t| = 0.652 - L.CAB_Y: Coefficient = - .12359** ; Standard Error = .0614 ; t = -2.01 ; P>|t| = 0.045 - D.CAB_Y: Coefficient = - .12691 ; Standard Error = .0856 ; t = -1.48 ; P>|t| = 0.139 - L.PRIVSAVINGS_Y: Coefficient = - .08903* ; Standard Error = .0539 ; t = -1.65 ; P>|t| = 0.100 - D.PRIVSAVINGS_Y: Coefficient = - .06259 ; Standard Error = .0906 ; t = -0.69 ; P>|t| = 0.490 - L.GOVBAL_Y: Coefficient = - .26077*** ; Standard Error = .0729 ; t = -3.58 ; P>|t| = 0.000 - D.GOVBAL_Y: Coefficient = - .37139*** ; Standard Error = .1064 ; t = -3.49 ; P>|t| = 0.001 - L.FED_RATE: Coefficient = .164647** ; Standard Error = .0759 ; t = 2.17 ; P>|t| = 0.031 - D.FED_RATE: Coefficient = .347821*** ; Standard Error = .1181 ; t = 2.94 ; P>|t| = 0.003 - L.IT: Coefficient = - .47424 ; Standard Error = .6177 ; t = -0.77 ; P>|t| = 0.443 - D.IT: Coefficient = -1.6391** ; Standard Error = .9321 ; t = -1.76 ; P>|t| = 0.080 - L.INFVOL: Coefficient = .458668* ; Standard Error = .2721 ; t = 1.69 ; P>|t| = 0.093 - D.INFVOL: Coefficient = 1.31338*** ; Standard Error = .5901 ; t = 2.23 ; P>|t| = 0.027 - Table notes: - (***) Significant at the 99 percent level; (**) Significant at the 95 percent level; (*) significant at the 90 percent level. - N = 364; F( 29, 334) = 384.16; Prob. > F = 0.0000. - “L” prefix indicates variable in levels with one lag; “D” prefix indicates first-differenced. - The average fixed effect is about 2, implying that, controlling for model variables, real interest rates in Brazil are about two percentage points higher than in other IT regimes for unidentified reasons. ### Effect of U.S. policy rate - Short-term effect: 0.3 percentage points in the short-term (D.FED_RATE). - Long-term effect: about 0.5 percentage points (captured by L.FED_RATE as reported). ### Domestic savings, counterfactuals, and magnitudes - Model indications: - Increases in private savings and government savings (improvement in overall fiscal balance) are strongly associated with reductions in real interest rates. - Increase in domestic savings via improvement in the fiscal position has an effect about three times higher than an equivalent increase in private savings. - Brazil’s sample average domestic savings: 16.5 percent of GDP. - Mexico average domestic savings: 22.6 percent of GDP. - Counterfactual: if Brazil’s average domestic savings increased from 16.5 percent of GDP to Mexico’s 22.6 percent of GDP, the model predicts real interest rates would decline by over 2 percentage points. - A full reduction of 4–5 percentage points would require increasing domestic savings to the levels of Thailand or Korea (about 30 percent of GDP), levels not observed in Brazilian history. ### Robustness checks - Results robust to different model specifications and estimation techniques, with exceptions: - Effects of private savings and the overall fiscal balance vary with estimation technique; however, the coefficient associated with the overall fiscal balance is always larger than that for private savings. - Collapsing private and government savings into one variable produces an estimate between the two coefficients. - Estimation methods used include: OLS with panel-corrected standard errors, OLS with White-corrected standard errors, Arellano-Bond GMM, and GLS. ### Interpretation and policy-relevant implications - Main drivers of decline in Brazil’s real interest rates: drastic reduction in inflation volatility, improved fiscal position, stronger external position, and credibility effect of the IT regime. - Further reductions from inflation volatility are limited given convergence to peers; the major remaining margin is increasing domestic savings. - Raising domestic savings to Mexico’s level would reduce the average difference with the rest of IT emerging markets by almost 50 percent. - Increasing domestic savings through fiscal improvement likely produces larger effects than equivalent private savings increases. - Structural features that may sustain upward pressure on real rates and blunt monetary transmission include credit market segmentation, minimum savings remuneration, indexation, and short maturity of public debt; subsidized public lending is a potentially important untested factor due to lack of comparative data. - Controlling for model variables, Brazil’s real interest rates are still about 2 percentage points higher than inflation-targeting peers, indicating other country-specific factors may be at play. *Italic: Source — IMF Working Paper _wp1262 (References), PDF chapter content covering 1980–2009.* ### References ............................................................................................................. ### _wp1262 - References ### I. Introduction and scope - The paper studies reasons why Brazil has relatively high real interest rates and provides insights into possible factors that may help reduce them over time. - Focus: determinants of the evolution of the (ex-post) short-term real interest rate (the policy rate set by the Central Bank adjusted for inflation). - Exclusion: the paper leaves for future research the question of why Brazil has high intermediation spreads in the banking system. - Sample and period covered: a panel data set of 15 emerging market inflation targeting (IT) countries over the 1980–2009 period. - Note on coverage: the paper covers the period 1980–2009 and therefore takes into account the monetary policy stance adopted by the countries in the sample in the aftermath of the financial crisis that started in 2008. It does not cover 2010 and 2011 due to data limitations in some cases and ongoing asymmetric monetary policy adjustments in a number of countries responding to the sovereign debt crisis in Europe. ### Trends in Brazil’s ex-post short-term real interest rate - Excluding the period of hyperinflation (1988–1994), the ex-post real interest rate trajectory for Brazil: - About 40 percent in the 1980s. - About 20 percent in the second half of the 1990s (prior to introduction of inflation targeting and floating of the currency in 1999). - Annual average of about 10 percent during 2000–2005. - Below 8 percent over the 2006–2009 period. - Reached the lowest historical level (just below 5 percent) in 2009. - Relative position: despite the historical decline, Brazil’s real interest rates in 2000–2009 were still about four percentage points above the average of emerging market inflation-targeting regimes. ### Comparative context and analytical approach - Comparative imperative: answering why Brazil has relatively high real interest rates requires analysis beyond Brazilian data; the paper therefore uses the cross-country panel to identify relationships between the real interest rate and key economic and institutional fundamentals. - Importance: Brazil’s high real interest rates have been cited as a significant constraint to economic development; understanding the drivers of the downward trend and potential convergence to other emerging markets is an important exercise. ### Figures and empirical components (as listed) - Figures referenced in the paper include: - Ex-Post Short-Term Real Interest Rate, Average 2000–2009. - Real Interest Rates in Brazil and the Rest of the Emerging Market IT Regimes, 1996–2009. - Brazil: Gross Public Debt in Percent of GDP, 1996–2009. - Public Sector Overall Fiscal Deficit (In percent of GDP), 1996–2009. - Domestic Savings and Real Interest Rates in IT Emerging Markets, Av. 2000–09. - Inflation Volatility in Brazil,* 1996–2009. - Actual and Predicted Interest Rates in IT Regimes, 1996–2009. - Real and Model-predicted Real Interest Rates in Brazil, 1996–2009. - Reduction in Real Interest Rates if Brazil’s Domestic Savings Increased to the Level of Other IT Emerging Market Economies. - Actual and Predicted Differences between the Real Interest Rate in Brazil and the Rest of Emerging Market IT Regimes, 1996–2009. *Italic source attribution: IMF Working Paper _wp1262 (References), PDF covering 1980–2009.* ### section II provides a selected review of the existing literature on interest rates in Brazil and ### _wp1262 - section II provides a selected review of the existing literature on interest rates in Brazil and ### Theoretical discussion and stylized facts (overview) - Literature groups five types of reasons for historically high interest rates in Brazil: - (A) Fiscal considerations - (B) Low domestic savings - (C) Institutional weaknesses - (D) Previous history of high inflation and inflation volatility - (E) Factors affecting the monetary policy transmission mechanism - Section purpose: review these arguments and present stylized facts to inform subsequent econometric analysis. ### A. Fiscal considerations - Two types of fiscal dominance (following Coates and Rivera (2004)): - Type I. Monetary Subordination: monetary policy directed at financing fiscal deficit through money creation (problem during hyperinflation episodes; not an issue in the last 15 years). - Type II. Crowding Out Effect in the Credit Market: fiscal deficit financed in domestic local-currency capital markets may compete with private sector for funding and bid interest rates up. - Empirical evidence on effect of public debt on real interest rates in Brazil is mixed: - Muinhos and Nankane (2006) find no evidence of a negative relationship between public debt levels and the real interest rate. - Simple bivariate examination (Figure 3) does not show a clear relationship; inclusion of gross public debt in panel regressions does not produce robust results and sometimes yields the opposite sign. - Fiscal discipline improved substantially over time (Figure 4): - Gradual reduction of the overall fiscal deficit due to sustained high primary surpluses and the fiscal responsibility law approved in 2000 likely contributed to lower real interest rates by reducing the public sector borrowing requirement. ### B. Domestic savings - Brazil has relatively low domestic savings; arguments (Hausmann 2008; Fraga 2005) posit low domestic savings as a binding constraint to growth and a reason for high real interest rates. - Classical intuition: if investment demand exceeds domestic savings supply, equilibrium real interest rate increases. - Cross-country stylized fact (Figure 5): - Countries with high domestic savings (e.g., Korea, Indonesia, Thailand around 30 percent of GDP) have lower real interest rates. - Chile and Mexico: average domestic savings 6–7 percentage points higher than Brazil and considerably lower real interest rates. - Brazil and Turkey appear as outliers: lower domestic savings and higher real interest rates than linear prediction (linear prediction would suggest average interest rates about four percentage points lower than actually observed for Brazil). ### C. Institutional weaknesses - Two institutional arguments discussed: - Jurisdictional Uncertainty: anticreditor bias, risk of contract value changes or unfavorable judicial interpretation; empirical evidence does not support this hypothesis (Goncalves, Holland, and Spacov (2007)). - Lack of Full Central Bank Independence: theoretically appealing but difficult to test; Nahon and Meuer (2009) find recent Board changes did not undermine Central Bank credibility. ### D. History of inflation and inflation volatility - Long history of high and volatile inflation in Brazil: - 1970s: annual inflation averaging 30 percent - 1980–88: averaging over 200 percent - 1989–1994: hyperinflation averaging 1,400 percent - Strong correlation between high inflation levels and high real interest rates (need to raise real rates to bring inflation down). - Post-1995 reduction in inflation and inflation volatility: - Introduction of inflation-targeting regime associated with the largest reduction in inflation volatility among sample IT emerging markets. - Mid-90s normalized inflation volatility was 2 standard deviations higher than the group average; by 2006 inflation volatility had fully converged to the group average. - Reduction in inflation volatility contributed substantially to the downward path of real interest rates (taming of inflation expectations per Bevilaqua et al, 2007). - Stylized statistic: an increase of one standard deviation in inflation volatility is associated with an increase in the real interest rate of about 1.4 percentage points (see econometric results below). For Brazil, reduction in inflation volatility over time is associated with a reduction in the real interest rate of over two percentage points. ### E. Factors affecting the monetary policy transmission mechanism - Documented particularities that could exert upward pressure on real interest rates: - Credit market segmentation: subsidized public lending (BNDES, housing, agricultural sectors) may push up equilibrium real interest in the free credit market by reducing the Central Bank’s ability to control aggregate credit via the policy rate. Effect size difficult to measure; direct transfers to BNDES that increased lending capacity by over 6 percent of GDP began in 2008 and expanded in 2009–10 and thus cannot be fully tested in the sample period. - Minimum remuneration requirements in savings accounts (cadernetas de poupança): remunerated at about 6 percent plus a reference rate and tax-exempt interest income; acts as a floor for the policy rate (Selic) and limits how low policy rates can fall. - Inflation target midpoint and bands: midpoint of the inflation target in Brazil is 4.5, with upper and lower bands of 6.5 and 2.5 percent; higher nominal rates for a given equilibrium real rate but the target per se not associated with higher real rates. - Other factors (Barbosa 2008): inflation inertia via indexation, weakness of wealth channel due to indexation and short maturity of public debt, relatively low credit-to-GDP ratio compared with other emerging markets. ### Econometric analysis — sample and data - Sample: all emerging market inflation-targeting regimes (Brazil, Chile, Colombia, Czech Republic, Hungary, Indonesia, Korea, Mexico, Peru, Philippines, Poland, Romania, South Africa, Thailand, Turkey). - Excludes advanced IT regimes (Australia, Canada, Iceland, Israel, New Zealand, Norway, Sweden and the U.K.). - Data sources: International Finance Statistics (IFS) and World Economic Outlook (WEO) for gaps. - Time period: 1980–2009 (panel is unbalanced; some countries only have data from early 1990s). ### Econometric model (panel ECM specification and variables) - Estimation based on a panel error correction model (ECM). Core specification (notation in source): - Dependent variable: ex-post short-term real interest rate, it, - Independent variables vector C includes: - PPPCAP: per-capita GDP in purchasing power parity values - CAB_Y: current account balance in percent of GDP - PRIVSAVINGS_Y: private savings in percent of GDP - GOVBAL_Y: overall fiscal balance of the general government in percent of GDP - FED_RATE: federal funds rate (control for international monetary conditions) - INFVOL: inflation volatility - IT: dummy for adoption of an inflation-targeting regime - Baseline estimation: OLS with Huber iterations (robust regression). Model also estimated with fixed effects. ### Main econometric results (summary and key coefficients) - Broad findings: - Real interest rate declines when private savings increase. - Real interest rate declines when overall government fiscal and external current account balances improve. - Increases in inflation volatility are associated with increases in the real interest rate. - Adoption of an inflation-targeting regime results in a significant short-term reduction in real interest rates. - International monetary conditions matter (federal funds rate coefficients reported below). - Specific reported relationships: - Inflation volatility: - A one standard deviation increase in inflation volatility associated with an increase in the real interest rate of about 1.4 percentage points. - For Brazil, reduction in inflation volatility associated with a reduction in the real interest rate of over two percentage points. - Inflation-targeting regime: - Adoption associated with an average short-term reduction in real interest rates of almost 1.6 percentage points. The level (long-term) effect is not statistically significant. ### Selected coefficient estimates (Table 1 highlights; dependent variable D.REALINT) - L.REALINT: Coefficient = - .31207*** ; Standard Error = .0122 ; t = -25.48 ; P>|t| = 0.000 - L.PPPCAP: Coefficient = 0.00000 ; Standard Error = .0000 ; t = 0.04 ; P>|t| = 0.971 - D.PPPCAP: Coefficient = 0.00028 ; Standard Error = .0006 ; t = 0.45 ; P>|t| = 0.652 - L.CAB_Y: Coefficient = - .12359** ; Standard Error = .0614 ; t = -2.01 ; P>|t| = 0.045 - D.CAB_Y: Coefficient = - .12691 ; Standard Error = .0856 ; t = -1.48 ; P>|t| = 0.139 - L.PRIVSAVINGS_Y: Coefficient = - .08903* ; Standard Error = .0539 ; t = -1.65 ; P>|t| = 0.100 - D.PRIVSAVINGS_Y: Coefficient = - .06259 ; Standard Error = .0906 ; t = -0.69 ; P>|t| = 0.490 - L.GOVBAL_Y: Coefficient = - .26077*** ; Standard Error = .0729 ; t = -3.58 ; P>|t| = 0.000 - D.GOVBAL_Y: Coefficient = - .37139*** ; Standard Error = .1064 ; t = -3.49 ; P>|t| = 0.001 - L.FED_RATE: Coefficient = .164647** ; Standard Error = .0759 ; t = 2.17 ; P>|t| = 0.031 - D.FED_RATE: Coefficient = .347821*** ; Standard Error = .1181 ; t = 2.94 ; P>|t| = 0.003 - L.IT: Coefficient = - .47424 ; Standard Error = .6177 ; t = -0.77 ; P>|t| = 0.443 - D.IT: Coefficient = -1.6391** ; Standard Error = .9321 ; t = -1.76 ; P>|t| = 0.080 - L.INFVOL: Coefficient = .458668* ; Standard Error = .2721 ; t = 1.69 ; P>|t| = 0.093 - D.INFVOL: Coefficient = 1.31338*** ; Standard Error = .5901 ; t = 2.23 ; P>|t| = 0.027 - Notes from Table 1: - (***) Significant at the 99 percent level; (**) Significant at the 95 percent level; (*) significant at the 90 percent level. - N = 364; F( 29, 334) = 384.16; Prob. > F = 0.0000. - “L” prefix indicates variable in levels with one lag; “D” prefix indicates first-differenced. - The average fixed effect is about 2, implying that, controlling for model variables, real interest rates in Brazil are about two percentage points higher than in other IT regimes for unidentified reasons. ### Interpretation and policy-relevant implications (from findings) - Reductions in inflation volatility have been a major driver of lower real interest rates in Brazil; further large reductions are unlikely given convergence to peers. - Adoption of inflation-targeting produced a significant short-term reduction in real interest rates (about 1.6 percentage points on average), though long-term level effects are not robustly significant. - Improvements in fiscal balances (GOVBAL_Y) have statistically significant negative effects on real interest rates, both in levels and changes. - Private savings and current account improvements are associated with lower real rates; policies that raise private savings or improve external balances could reduce real interest rates. - International monetary conditions (federal funds rate) matter for domestic real interest rates; global tightening raises domestic real rates (coefficients reported above). - Structural features (credit market segmentation, minimum savings remuneration, indexation, short maturity of public debt) may blunt monetary transmission and sustain upward pressure on real rates; some features are hard to quantify but are relevant for policy design. *Italic: Source — _wp1262 - section II provides a selected review of the existing literature on interest rates in Brazil and (PDF chapter content).* ### 0.3 percentage points in the short-term (D.FED_RATE) and has a long-term effect of ### _wp1262 - 0.3 percentage points in the short-term (D.FED_RATE) and has a long-term effect of ### Effect of U.S. Policy Rate - Short-term effect: 0.3 percentage points in the short-term (D.FED_RATE) - Long-term effect: about 0.5 percentate points (captured by the variable L.FED_RATE) . ### Domestic Savings and Real Interest Rates in Brazil - Increases in private savings and government savings (as measured by the improvement in the overall fiscal balance) are strongly associated with reductions in real interest rates. - The model shows the increase in domestic savings via an improvement in the fiscal position has an effect about three times higher than the one that would arise from an equivalent increase in private savings. - Brazil’s margin to increase domestic savings is substantial. - Illustrative counterfactuals and model predictions: - If average domestic savings increased in Brazil from its sample average of 16.5 percent of GDP to the average of Mexico (22.6 percent of GDP), the model would predict that real interest rates would decline by over 2 percentage points. - A full reduction of 4–5 percentage points would require increasing domestic savings to the levels of Thailand or Korea (about 30 percent of GDP), which has never been observed in Brazilian history. - Reaching the level of domestic savings of Mexico should be more feasible for Brazil. ### Robustness Checks - The results are robust to different model specifications and estimation techniques. - Alternative estimations yield similar results for main variables, with exception of the effects of private savings and the overall fiscal balance (proxy for net government savings). - The relative magnitude of the two coefficients varies by estimation technique, but the coefficient associated with the overall fiscal balance is always larger, confirming that improvements in the government’s fiscal position have a stronger long-term impact in reducing real interest rates than similar increases via private savings. - Collapsing private savings and government savings into one variable produces an estimate between the two coefficients. - Estimation methods used include: OLS with panel-corrected standard errors, OLS with White-corrected standard errors, Arellano-Bond GMM, and Generalized Least Squares Model (GLS). ### Conclusions and Policy Implications - Brazil still has considerably higher real interest rates than other emerging market inflation targeting countries, but the difference has been reduced over time thanks to better economic fundamentals (drastic reduction in inflation volatility, an improved fiscal position, a stronger external position, and the credibility effect of the IT regime). - Going forward, increasing domestic savings would seem to be the single most important factor to reduce real interest rates in Brazil over time because Brazil still has a low level of domestic savings and substantial room to expand. - Raising domestic savings to the level of Mexico would reduce the average difference with the rest of the IT emerging markets by almost 50 percent. - Increasing domestic savings through an improvement in the fiscal position is likely to produce larger effects, suggested by the larger magnitude of the regression coefficient associated with this variable. - Fully eliminating the difference through increases in domestic savings alone would require saving rates similar to Korea or Indonesia (30 percent of GDP), which does not seem realistic for Brazil in the near term. - Controlling for variables in the model, Brazil’s real interest rates are still about 2 percentage points higher than those of its inflation targeting peers (captured by the fixed effects), indicating other country-specific factors may be at play. - A potentially important untested factor (due to lack of comparative data) is the effect of public lending at subsidized rates; while effective countercyclically, its use in normal times may reduce monetary policy transmission and contribute to a higher market-determined equilibrium interest rate. *Source: _wp1262 - 0.3 percentage points in the short-term (D.FED_RATE) and has a long-term effect of* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1262.pdf_