Financial Stability Report - bcb.gov.br

Financial Stability ReportVolume 18 ǀ Issue No. 1 ǀ April 2019

ISSN 2179-5398

Financial Stability ReportVolume 18 ǀ Issue No. 1 ǀ April 2019

ISSN 2179-5398CNPJ 00.038.166/0001-05

Financial Stability Report

Semiannual Central Bank of Brazil publication.

Reproduction permitted only if source is stated as follows: Financial Stability Report, Volume 18, n. 1.

Occasional discrepancies between constituent figures and totals as well as percentage changes are due to rounding.

There are no references to sources in tables and graphs originated in the Central Bank of Brazil.

Contents

Preface 5

Executive summary Ï

Unfolding of topics/risks analyzed in previous FSRs ______________________________________10Exposure of the banking sector to subnational entities and potential

impacts to Financial Stability _____________________________________________________10Financial Stability Committee’s decisions on the Countercyclical Capital Buffer ________________10

1 Financial system overview 11

1.1 Liquidity _____________________________________________________________________ 111.2 Credit__________ ______________________________________________________________15 1.2.1 Introduction ______________________________________________________________15 1.2.2 Broad credit and long run trend ______________________________________________16 1.2.3 Companies _______________________________________________________________17 1.2.4 Households ______________________________________________________________19 1.2.5 Domestic banking credit by ownership _________________________________________21 1.2.6 Risks and provisioning _____________________________________________________211.3 Profitability _________________________________________________________________221.4 Solvency______________________________________________________________________251.5 Capital stress tests ______________________________________________________________27 1.5.1 Scenario analysis – Macroeconomic stress tests __________________________________27 1.5.2 Sensitivity analysis ________________________________________________________29 1.5.3 Simulation of direct interbank contagion _______________________________________301.6 Financial Stability Survey ________________________________________________________31 1.6.1 Introduction ______________________________________________________________31 1.6.2 Risks to financial stability ___________________________________________________31 1.6.3 Financial and economic cycles _______________________________________________35 1.6.4 Expectations for the Countercyclical Capital Buffer ______________________________37 1.6.5 Resilience and confidence in the financial system stability _________________________37 1.6.6 Final remarks _____________________________________________________________381.7 Systemically important financial market infrastructures ________________________________39

2 Selected issues 42

2.1 Debentures _________________________________________________________________422.2 Evolution of vehicle financing outstanding __________________________________________43

2.3 Housing credit – Modality profile __________________________________________________452.4 Assessment of changes in coverage eligibility rules of the Credit Guarantee Fund (FGC) ______492.5 Portfolio flows for emerging economies and the non-residents behavior ____________________502.6 National Monetary Council introduces changes to the definition of regulatory capital _________552.7 The Central Bank of Brazil enhances its credit risk capital requirements ___________________562.8 Large exposures limit ___________________________________________________________582.9 Introduction of the Net Stable Funding Ratio (NSFR) in Brazil __________________________58

Appendix 62

Annex 66

Concepts and methodologies ________________________________________________________66Concepts and methodologies – Capital stress ____________________________________________68Working papers about financial stability ________________________________________________72

April 2019 | Financial Stability Report | 5

The Financial Stability Report (FSR) is a semiannual publication issued by the Central Bank of Brazil (BCB) that presents an overview of recent developments and the outlook on financial stability in Brazil, focusing on the main risks and on the domestic financial system resilience, as well as conveys the Financial Stability Committee (Comef) view on the policy and measures to preserve financial stability. The current edition covers the second half of 2018, highlighting more recent events when relevant.

The BCB defines financial stability as the regular operation, over time and in any economic scenario, of the system responsible for the financial intermediation among households, non-financial corporations, and the government.

The report comprises two chapters. Chapter 1 – Financial system overview – presents an analysis of risks related to liquidity, credit, profitability and solvency, of capital stress tests and their effects on the solvency of financial institutions, of the Financial Stability Survey (FSS) results, and of systemically important financial market infrastructures operation. Chapter 2 – Selected issues – discusses relevant but not necessarily recurring topics that may have implications to financial stability in Brazil.

The Statistical annex shows charts and tables underlying data and can be found on the FSR website, <http://www.bcb.gov.br/?fsr>, as well.

Moreover, important time series for financial stability monitoring (e.g. total capital ratio, short term liquidity ratio, delinquency ratio, return on equity) can be downloaded from the Time Series Management System (SGS) in <https://www3.bcb.gov.br/sgspub/localizarseries/localizarSeries.do?method=prepararTelaLocalizarSeries>.

Preface


Executive summary

Economic activity continued to indicate gradual recovery of the Brazilian economy in the second half of 2018. However, the pace of growth is lower than expected, the industrial capacity utilization indexes are still low, and the unemployment rate remains high. The continuity of reforms and adjustments in the domestic economy is essential to a sustainable recovery, and to maintain low inflation in the medium and long term1.

In this scenario of gradual recovery and historically low both inflation and Brazilian Benchmark Interest Rate (Selic), the capital market continues to be attractive to non-financial corporations (NFC). This is evidenced by the fact that the growth in NFC indebtedness was stronger in the capital market than in the banking system in 2018. Nonetheless, corporate bank credit interrupted a two-and-a-half-year period of negative growth rates.

In general, risks to financial stability have reduced, despite the high level of problem-assets in banking loans to large companies. Profitability has resumed to pre-recession level, liquidity is high, and solvency and leverage ratios are well above the regulatory minimums.

The results of the Financial Stability Survey (FSS) show that the market remains highly concerned about the risks related to the approval of necessary measures for the fiscal balance and to the resumption of economic growth. Banking institutions remain confident in the robustness and in the ability of the financial system to absorb shocks.

Capital market emerges as an important source of funding for non-financial corporations. At the same time, non-earmarked credit resumed.

• New issuers account for a significant proportion of the issued securities, which indicates increased access to capital market.

• Banks and investment funds hold 83% of the debentures outstanding in the market. Investment funds absorbed 40% of the debentures issued in 2018. They have been increasing their share in these securities since 2015, aiming at increasing investors’ earnings and at preserving the balance between attractiveness and fees charged to the investors.

• Given the context of low interest rates and high liquidity in the economy, both debentures’ new issues and investment funds’ share in these new issues shall remain growing.

• As for the annual growth of NFC debt, the banking system increased both its securities portfolio and its non-earmarked credit portfolio. The improvement in non-earmarked loans was sufficient to offset the decline in earmarked credit, thereby halting the downward trend in total credit to NFCs that had persisted since mid-2016. The trend for non-earmarked credit is to continue growing, mainly in less risky credit types.

• The credit risk of Small and Medium-sized enterprises (SME) continues to decline. As for loans to large companies, however, the risk remains high: the problem-assets index remained in upward path and the applications for judicial recovery have slightly increased.

1 See Inflation Report, available at https://www.bcb.gov.br/htms/relinf/ing/2018/12/ri201812sei.pdf.


The credit portfolio to households grew at the highest annual rate since 2015, driven by consumption-related loans.

• The improvement in household income, and the historically low inflation and interest rates have contributed to boost consumer confidence.

• All credit types to households grew in 2018. Annual origination of credit card loans and non-payroll deducted personal loans reached the highest records since 2013. The former were more pronounced in “sight” credit card purchases (those with a single debit entry in the card balance). The latter occurred as a result of an increased supply by financial institutions.

• There is no evidence of an increase in credit risk held by financial institutions. Problem-assets either remained stable or decreased in most household credit types.

• Bank credit to households shall continue to grow as long as the economic recovery is confirmed and the improvement in expectations captured by the consumer confidence index materializes.

Banks’ profitability remains growing. Smaller banks that operate in specific markets still face challenges for a more consistent recovery.

• Profitability remained favored by the decrease in loan-loss provisioning, by the drop in funding cost, and by operational efficiency gains in 2018.

• Smaller banks try to diversify their revenues to face the challenges in the markets where they operate. Those who lend to large companies have faced growing risks stemming from their niche of clients, and larger competition with capital market. Those who operate in “Treasury and investment activities” deal with margins under pressure due to the low interest rate environment.

• With the prospect of both stable provisioning expenses and stable funding costs, profitability growth tends to lose strength. In this context, credit portfolio growth and changes in the portfolio mix, with a higher share of more profitable portfolios – household and SME loans –, can contribute to further marginal improvements.

The banking system has robust capital, both in quantity and quality, is fully compliant with Basel III rules and is able to withstand the resumption of demand for credit.

• All capital adequacy and leverage ratios remain well above regulatory requirements. The fully-fledged implementation of Basel III framework regarding the Common Equity, concluded in January 2018, and regarding the minimum capital requirement, concluded in January 2019, was smooth for the Brazilian banks. The same can be said regarding the ongoing transition related to Tier 2 capital and to additional Tier 1 capital.

• As in previous periods, loan-loss provisions remain comfortable, in line with the portfolio’ risk profile, indicating that there is no need for additional provisioning that could undermine the amount of available capital.

• The gradual credit recovery and the balance sheet expansion shall require more capital to face risks. In this scenario, robust capital and banks’ ability to generate profits will contribute to keep the capital adequacy ratios stable and in a comfortable level.


The slight resumption of credit did not imply an increase in liquidity risk. Both short-term and long-term liquidity risk indexes improved further in the second half of 2018.

• The banking system’s Short-term Liquidity Ratio (IL) has reached its highest level since 2012. This increase is due to the combined effects of the growth of inflation-linked sovereign bonds market value – the main source of increase in liquid assets –, with the reduction in stressed cash flow. In addition, all banking conglomerates required to comply with the Liquidity Coverage Ratio (LCR) presented ratios above 100%.

• The Structural Liquidity Ratio (ILE) has been rising for about three years, mitigating long-term liquidity risk. This has been possible thanks to the slow pace of credit origination and to the reduction of the credit portfolio duration, along with a slight increase in equity, in retail funding, and in long-term funding.

• The Net Stable Funding Ratio (NSFR) minimum requirement for the larger banks started in the last quarter of 2018, in accordance with the Basel III rules. All banks subject to comply with it have maintained NSFR ratios above the regulatory minimum, in line with the ILE behavior. Both ratios corroborate the perception of a low liquidity risk in the long run.

The results of capital stress tests continue confirming the resilience of the banking system, which is able to absorb the estimated losses in all the simulated scenarios.

• Tests in scenarios where unemployment, interest and exchange rates, inflation, economic activity and US interest rates are all in stressed levels indicate that the problem-assets as a share of total credit portfolio would reach less than 10% in the worst scenario. The estimated capital needs to comply with the minimum requirements would be less than 1% of regulatory capital.

• The results of sensitivity tests to abrupt shocks in exchange rates indicate negligible impact on regulatory capital. However, the results were more sensitive to sudden shocks in interest rates than in the previous semester. An abrupt increase of 100% in the Selic rate would imply a significant regulatory capital need to comply with minimum requirements.

• The sensitivity to credit risk shocks indicates that, even in the most extreme scenario simulated, with problem-assets to total credit peaking at 18.1%, more than twice the highest ratio ever observed, there would be an additional capital need equivalent to 2.9% of the system’s regulatory capital.

• The simulation of sensitivity to reductions in residential property prices demonstrates that there would not be non-compliance and/or restrictions on dividend distribution in case of a decrease in nominal prices of up to 30%.

• The simulation of direct interbank contagion suggests a reduced need of resources for recapitalizing the banking system in case of some contagion event in each individual institution that could affect the others. In the worst scenario, the capital needs would amount to less than 1% of the system’s regulatory capital.

The systemically important financial market infrastructures worked efficiently throughout the second half of 2018.

• The effective liquidity need of the system was, on average, 1.9% of the available intraday liquidity, being 12.6% the maximum in the period.

• The backtest analyses for the clearing and settlement systems of securities, derivatives, and foreign currency transactions in which there is an entity acting as a central counterparty have presented satisfactory results.


The recent expansion of the vehicle financing does not carry the risks observed throughout the 2009-2012 expansion.

• The recent expansion of the vehicle loans has occurred with adequate risk, maturity and loan-to-value (LTV) parameters. The evolution of delinquency considering both stock and different cohorts per month of credit granted corroborates this perception.

• Problem-assets have declined substantially since the end of the previous expansion and remain stable over the current expansion. Non-performing loans, which accounts for more than 70% of the problem-assets ratio, are at the lowest level since December 2008.

The risks related to real estate loans have, in general, decreased after the rise during the recession. The real estate loans account for one third of household credit and are one of the less risky portfolios.

• The risk materialization observed in the problem-assets ratio declined due to the lower volume of restructuring and to the fall in delinquency. This decline has occurred in a scenario where the growth in real estate loans was low and stable. At the same time, the coverage ratio to face this risk rose about 25%.

• The risk of real estate loans has also declined due to the increased share of loans with LTV ratios lower than 80%. In addition, roughly 97% of borrowers take real estate loans to buy only one property, which minimizes the risk of excessive leverage.

• Property foreclosures related to deteriorated loans remain rising. However, these properties represent only 1.6% of the real estate portfolio and 1.4% of the regulatory capital, thereby not posing risks to financial stability.

• Real estate loans’ interest margins are reaching back the 2013 level. Although they have not represented risk, these margins narrowed substantially in 2014-2015. In that period, the real estate funding costs raised due to increases in both Selic rate and funding through real estate credit bills (LCI), without elevating interest rates of new loans.

The US economic policy decisions had a negative impact on the portfolio investment flow to emerging economies in 2018.

• At the global level, external factors, such as the US monetary policy and the risk appetite for emerging markets’ assets, seem to explain most of capital flows to emerging countries in 2018. Idiosyncratic attraction factors have also contributed significantly to deepen the deterioration of flows, which suggest that other variables pressured the portfolio investment flow.

• At the domestic level, the reduction in flows in 2018 was also influenced mainly by external factors. Idiosyncratic attraction factors of the Brazilian economy were not relevant to explain the recent behavior of portfolio flows to Brazil.

• The more accommodative US monetary policy and the increase in the risk appetite for emerging markets’ assets at the beginning of 2019 suggest a recovery of capital flows to emerging economies.

There was a moral hazard mitigation related to the possibility of multiplying the deposits guaranteed by the Credit Guarantee Fund (FGC).

• This reduction is due to a new BCB rule that limited the FGC coverage to R$1 million per investor for each four-year period. As a result, investment flows above R$1 million per investor decreased by ¼ in 2018.


• This rule discourages splitting the investments in smaller portions without the adequate evaluation of the issuer risk, and maintains the protection to the retail investor.

The BCB and the National Monetary Council (CMN) continue promoting the resilience of the National Financial System (SFN) both to increase its efficiency and safety and to enhance its compliance with international regulatory standards. In this sense, the following measures stand out:

• Improvement in the definition of regulatory capital. This measure addresses issues on Tier 2 capital and aims to ensure that the capital requirement converts into an adequate loss-absorbing capacity.

• Adjustments in the standardized approach to credit risk, mainly in the required capital for exposures related to loans to NFC and to foreign sovereign entities. These changes aim at improving prudential rules and at boosting incentives by optimizing capital allocation based on the level of risk incurred by financial institutions.

• Improvement in the rules applicable to maximum limit of both exposure per customer and concentrated exposures. Such enhancement widens the scope of the transactions considered in the calculation of the limits, changes the calculation basis, and establishes the concept of linked counterparties, who are considered as a single customer for the purpose of applying the limits.

Unfolding of topics/risks assessed in previous FSRs

Exposure of the banking sector to subnational entities and potential impacts to financial stability

As for the exposure of the financial system to subnational entities, impact simulation results remain suggesting low risk to financial stability. The current test outcomes show a risk reduction when compared to simulations reported in the FSR previous edition. Even though debt has slightly grown, the system’s higher capitalization lowers the effect of defaults. The financial system is prepared to withstand an occasional default by subnational entities rated C or D by the National Treasury Secretariat, who are not eligible to be guaranteed by the federal government, as well as by their active and retired civil servants, by their suppliers and suppliers’ employees.

Financial Stability Committee’s decisions on the Countercyclical Capital Buffer

The Financial Stability Committee (Comef) has decided in its quarterly meetings held on November 22th, 2018, and on March 7th, 2019, to keep the Countercyclical Capital Buffer for Brazil (ACCPBrasil) at 0% (zero percent)2. Both decisions were taken by Comef in the exercise of its duties provided by Circular 3,927, of February 11th, 2019, and have followed the objectives and procedures described in Communiqué 30,371, of January 30th, 2017.

2 Communiqués 32,794 (November 22th, 2018) and 33,240 (March 7th, 2019).


1Financial system overview3

1.1 Liquidity3

The liquidity risk remains a low concern for the Brazilian banking system. Due to beneficial mark-to-market effects and a positive net cash-flow4 being used to purchase sovereign bonds, the aggregate banks’ liquidity buffer increased in the second half of 2018. In this context, both regulatory and monitoring liquidity risk ratios followed-up by the BCB improved in the period, indicating that the credit granting in the semester did not lead to increases in the liquidity risk-taking of banks. Summing up, this risk is still at low levels, scenario that tends to continue during 2019.

Funding has grown 2.2% over the second half of 2018 and 4.6% over the full year. Time deposits are still absorbing part of the reduction in the stock of repurchase agreements collateralized by securities issued or endorsed by institutions within the same prudential conglomerate (due to Resolution 4,527 published on September 29, 2016) and savings deposits net flows remained positive in the second half of 2018. Funding via agribusiness credit bills (LCA) has remained stable, whereas that via

3/ Please note that, within sections 1.3 Profitability, 1.4 Solvency and 1.5 Capital stress tests, the granularity of the analysis is in the level of prudential conglomerates, as defined by Resolution no. 4,280, from October 31, 2013, to which the minimal capital requirements are applied, as stated by Resolution no. 4,193, from March 1, 2013. In sections 1.1 Liquidity, the basis of the analysis is the whole banking system, comprised by Commercial Banks (CB), Multiple Banks (MB), FX (foreign exchange) Banks and Investment Banks (IB) and by financial conglomerates including at least one of these types of institutions. In 1.2 Credit, the scope is the whole SFN.

4/ Due to a slight increase in the stock of funding in the second half of 2018, concomitantly with the lower level of credit granting in relation to the cash inflows of principal and interest payments from the credit portfolio in the same period.


real estate credit bills (LCI) has decreased due to a shift by the main issuers to cheaper funding (Chart 1.1.1)5.

Funding profile by type of investor shows the prevalence of natural and legal persons not classed otherwise6, whose stock represented 62% of the outstanding amount at the end of 2018. Brokered funding provided by natural and legal persons, even though still accounting for less than 2% of total funding, has been growing steadily over the past 5 years (22% in the second half of 2018 alone). On the one hand, this brokered funding broadens the customer base to which small and medium-sized financial institutions have access, allows longer maturity issues and reduces their dependence on professional asset managers, on the other hand, it increases some institutions dependence on the brokers themselves (Chart 1.1.2).7

The ratio of external funding over the total amount of borrowing from the Brazilian financial system remained, by the end of 2018, at the same level of the previous semester (charts 1.1.3 and 1.1.4). On December 31st, the outstanding amount of external funding internalized in the domestic market was BRL 464.7 billion, of which BRL 131.5 billion were destined for export/import financing or on-lending operations, and the remaining BRL 333.2 billion were used for free destination credit lines. The internal uncertainties of the electoral period did not interfere with the availability of external credit for the SFN, although the costs grew, aligned with rates practiced in the external markets, which were impacted by the increase of the basic US interest rate. Considering the internal conjuncture of low demand for credit, the external funding is expected to maintain the current level for the next semester, with its cost following the trend of foreign interest rates.

As more than 90% of the external funding is referenced in USD, its average cost showed the same upward trend as the Libor USD rate. The monthly average of the 6-month Libor USD rate ranged from 2.5% p.a. in June 2018 to

5/ Time deposits: certificates of deposit, receipts of deposit, time deposits with special guarantee by the Credit Guarantee Fund (Fundo Garantidor de Crédito – FGC). Subordinated debts: subordinated certificates of deposit, subordinated financial notes and other capital instruments. Other instruments: structured notes, bills of exchange, financial notes, mortgage notes, box spread strategies with options. Repurchase agreements (repo): refers only to repo collateralized by private-issued securities. Repo collateralized by the Brazilian federal government securities were not considered into the funding concept used in this report, as they are operations by which banks exchange liquidity.

6/ Definition comprising not only companies, but other entities (such as non-profit ones) as well.

7/ Funding from the public sector is almost entirely linked to the brokering of government loans or credit lines.

594 623 658 642 652 634 661 671 720 744 792

576 565 534 504 535 568 603 710 717 806 875320 326 370 358 397 376 356 248 215 127

80411 435 450 456467 466 475 476 488 483 482264 278

358 403494 402 393 385 388 450 465

215252

277 302335

343 330 310 304333 323

151161

174 180

181189 202 220 202

224 224

95123

159 192

203 206 201 198 183167 155

149152

164 165

179 180 174 173 167150 157

109133

147 178

186 187 180 153 144145 144

180157

172 146

148135 155 138 160

149 165

-

1,000

2,000

3,000

4,000

Dec2013

Jun2014

Dec2014

Jun2015

Dec2015

Jun2016

Dec2016

Jun2017

Dec2017

Jun2018

Dec2018

R$ biChart 1.1.1 – Funding ProfileBanking System

Saving deposits Time deposits Repos (private securities) Domestic borrowing and onlending External funding Other instruments Judicial deposits Real estate credit bills Subordinated debt Agribusiness credit bills Demand deposits Interbank deposits

Sources: BCB, [B]3

1,940 2,005 2,142 2,137 2,264 2,260 2,329 2,321 2,330 2,349 2,432

11 1620 24

28 30 36 41 48 5466

377413

420 435457 455 430 399 375 380

354445

470494 500

511 509 520 521 533 532 531

264278

359 403

494 402 393 385 389 451465

8690

91 89

8987 84 84 88

8586

-

1,000

2,000

3,000

4,000

Dec2013

Jun2014

Dec2014

Jun2015

Dec2015

Jun2016

Dec2016

Jun2017

Dec2017

Jun2018

Dec2018

R$ biChart 1.1.2 – Funding Profile by Type of Investor Banking System

Natural and legal persons Natural and legal persons - Brokered Institutional Public setor Non residents Financial institutions

Sources: BCB, [B]3

0

5

10

15

20

25

Dec2010

Dec2011

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

%

External funding kept abroad versus total fundingExternal funding that entered the country (linked to foreign trade) versus total fundingExternal funding that entered the country (free-purpose) versus total funding

Chart 1.1.3 – Profile of external fundingAs a percentage of total funding

Statistical annex

Statistical annex

Statistical annex

http://www.bcb.gov.br/ingles/estabilidade/2018_10/fsrStatisticalAnnex.xlsx

https://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsx


https://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsxhttps://www.bcb.gov.br/ingles/estabilidade/2019_04/fsrStatisticalAnnex.xlsx




2.89% in December. The average cost of externally-funded export-oriented credit lines, which are relevant to Brazilian exports, presented the highest correlation with the variation in external rates (Chart 1.1.5).

During the second half of 2018, the maturity profile of the domestic funding stayed relatively stable. There was a slight increase in the share of redeemable deposits and sight deposits in the total funding (Chart 1.1.6)8. On the other hand, the long-term instruments (maturing in more than one year), that have a lower liquidity risk, also rose their share, with the reduction occurring mainly in the category of funding maturing between 30 days and one year.

While the Selic rate accrued 3% in the second semester, the stock of liquid assets measured by the BCB nominally rose 11%, due mainly to the increase in the market value of fixed-rate and inflation-linked domestic sovereign bonds, which account for around 60% of the aggregate liquidity buffer. This price movement was a result of the shrinkage of the yield curve in the period, with the agents estimating a more benign scenario for the Brazilian economy. Beyond the carrying effect of the liquid assets, in the second half of 2018, banks also bought more sovereign bonds using their positive net cash flows, which, in turn, were a result of the credit granting pace lower than the inflows of interest and principal payments from the stock of credit in course.

Analyzing the banking system by type of control, in the same period the aggregate stock of liquid assets of public banks remained nearly stable (+2%). Despite the sovereign prices increases, those banks observed a slight decrease in the total funding in the second half. Meanwhile, there was an increment of 18% in the aggregate liquidity buffer of private institutions, which benefited from an expansion of their funding, resources partially applied in liquid assets. Due to this enhancement, the liquidity levels of private banks returned to the ones of the beginning of 2018, compensating therefore the liquidity reduction occurred during the first half.

In the second half of 2018, the stressed cash outflows in the next 21 business days estimated in the stress scenario designed by the BCB decreased 2% in relation to the end of first half; result explained by the reduction in interest rates and US dollar exposures. The funding run-offs

8/ BCB changed the process to generate the time series, resulting in minor differences in relation to the data that had been published until the last Financial Stability Report (October 2018).

0

50

100

150

200

250

Jun2011

Jun2012

Jun2013

Jun2014

Jun2015

Jun2016

Jun2017

Jun2018

USD billion

External funding kept abroad

External funding that entered the country (linked to foreign trade)

External funding that entered the country (free-purpose)

Chart 1.1.4 – Profile of external fundingAbsolute amounts in dollars

0.0

0.6

1.2

1.8

2.4

3.0

3.6

Jun2017

Aug Oct Dec Feb2018

Apr Jun Aug Oct Dec

%

Export credit lines Libor 6 month - monthly average

Chart 1.1.5 – External credit lines for exportInterest rates

Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec0%

20%

40%

60%

80%

100%

Chart 1.1.6 – Domestic fundingResidual maturity profile

Judicial deposits Funding maturing above 1 year Funding maturing between 30 days and 1 year Saving deposits Funding maturing between 1 and 30 daysSight deposits + other redeemable funding

Statistical annex

Statistical annex

Statistical annex





estimated in the BCB scenario remained stable, with the increase in the stock of funding being counterbalanced by the diminution of liabilities maturing in the short-term. Thus, due to the joint effect of liquid assets expansion and stressed cash outflows reduction, the banking system`s aggregate short-term liquidity ratio (IL)9 closed 2018 at 2.42, a positive variation of 29 points in comparison with the first half (Chart 1.1.7), and surpassing the previous peak of 238%, observed in the end of 2017.

Individually, the banking institutions reduced their short-term liquidity risk, a move that happened mainly to the private-owned banks, including the large ones, reversing the risk-raising movement that had begun in the first half of the year on some of these institutions, and closing the year of 2018 at levels similar to those of twelve months ago. In June of 2018, 98% of the assets of the banking system were on balance sheets of banks with enough liquid assets to withstand a liquidity crisis (i.e. IL above 1 as shown in Chart 1.1.8). In comparison with the previous semester, the number of institutions with IL below 1 increased from 18 to 20, all banks with low representativeness in terms of total assets.

The Liquidity Coverage Ratio (LCR)10 corroborates the low liquidity risk already outlined by IL. All banking conglomerates required to comply with this ratio kept their ratios above 100%, regulatory minimum to be required by 2019. The aggregate LCR of those institutions reached 210% by the end of 2018, an increase of 7 p.p. in twelve months and +13 p.p. compared to the first semester of 2018 (Chart 1.1.9). Thus, both the IL and the average LCR of the biggest institutions indicate a high capacity to face stress scenarios in the short term.

By measuring the liquidity risk over a period of more than 30 days, the banks’ funding structures continue

9/ The IL measures whether banks have enough liquid assets to cover their short-term (21 business days) cash-flow needs in a simulated stress scenario, defined and calibrated by the BCB. Such cash outflows arises from the run-off of maturing or redeemable liabilities, losses from market risk exposures, for instance, margin calls and derivative pre-settlements, and other contractual outflows maturing in the next month. Institutions with IL above 1 have enough liquid assets for such scenarios. For further calculation details, please refer to appendix Concepts and Methodologies, item a.

10/ In Brazil, all institutions considered in the S1 segment, under terms of art.2nd of Resolution no. 4,553, of January 30, 2017 and Resolution no. 4,616, of November 30, 2017 must comply with the LCR. It requires that financial institutions maintain a stock of high quality liquid assets to withstand cash outflows for the next 30 days, under a standardized stress scenario set by the BCBS (www.bis.org/publ/bcbs238.htm). In 2018, the minimum requirement was 90% and as of January 2019, it is 100%. Domestic regulation set by Resolution no. 4,401, of February 27, 2015, and Circular n° 3,749, of March 5, 2015.

0

100

200

300

400

500

600

700

800

900

1,000

0.00

0.50

1.00

1.50

2.00

2.50

Jan2012

Jun Dec2012

Jun Dec2013

Jun Dec2014

Jun Dec2015

Jun Dec2016

Jun Dec2017

Jun Dec2018

BRL billion

Chart 1.1.7 – Short-term liquidity ratio (IL)

Liquid assets Net stressed cash outflowsShort-term liquidity ratio (IL)

un.

135

10

14

12

76

18 2

69

21

74

0

18

36

54

72

90

<0,8 0,8-1,0 1,0-1,2 1,2-1,6 1,6-2,0 ≥2,0

Chart 1.1.8 – IL Frequency distribution1/

June 2018 December 2018

Short-term liquidity ratio

1/ The numbers above the bars are the number of financial institutions with IL within the corresponding interval.

% of system assets

0%

100%

200%

300%

400%

500%

600%

Oct2015

Dec Mar2016

Jun Sep Dec Mar2017

Jun Sep Dec Mar2018

Jun Sep Dec

Chart 1.1.9 – LCRHigh, low and aggregate1/

Aggregate LCR IL Maximum LCRMinimum LCR

1/ Until Sep18: aggregate LCR and IL series comprise banks with more than BRL 100 bi in total assets. From oct18 on: encompass banks in the S1 Segment.

Regulatory minimum

Statistical annex

Statistical annex

Statistical annex





to indicate low susceptibility to liquidity crises in the medium and long term. The aforementioned growth in long-term funding (maturity above a year) and in short-term retail funding in the second half of 2018 reinforces the amount of stable funding. Meanwhile, despite the maintenance of the credit recovery movement in the period, credit portfolios presented a shorter duration, which reduces the need for long-term financing. These movements led the Structural Liquidity Ratio (ILE)11 to 1.15 in December of 2018 (Chart 1.1.10), a slight increase (+0.02) in relation to the first semester and to the previous year (+0.03). The ILE is a measure of long-term liquidity risk assessment based on the Net Stable Funding Ratio (NSFR) methodology. The ILE indicator has been showing an upward trend for approximately three years, mainly due to the slow pace of credit granting concomitant to capital and stable funding growth in this period.

The most representative banks in terms of assets (95.9% of banking system total assets in December 2018) maintained balance sheet structures that minimize long-term liquidity risk, i.e. ILE above 1.00 (Chart 1.1.11). Twenty-two banks had an ILE of less than 1.00, one bank more than in the previous semester. Twelve of these banks also present liquidity shortfalls in an eventual short-term crisis (IL < 1), all micro and small-sized institutions.

1.2 Credit

1.2.1. Introduction

Generally speaking, credit and risk indicators improved in the second semester of 2018, with better portfolio quality and reduced write-offs. Despite some degree of oscillation, the macroeconomic environment showed a gradual recovery, with a positive impact on bank financing to families and companies. As pointed out in the previous FSRs, the recovery movement occurs mainly in credit for individuals and granted by private banks, especially in non-earmarked loans. In the case of corporate credit, there is a positive real growth in non-earmarked loans, more than compensating the negative variation on earmarked loans. In addition, for these borrowers, the growth of

11/ ILE aims to measure whether banks have enough stable funding resources (numerator) to finance their long-term activities (denominator). Thus, institutions with ILE equal to or greater than 1 are less susceptible to future liquidity problems. For details, see appendix Concepts and Methodologies, item b.

0

1,000

2,000

3,000

4,000

5,000

0.40

0.55

0.70

0.85

1.00

1.15

Dec2011

Jun Dec2012

Jun Dec2013

Jun Dec2014

Jun Dec2015

Jun Dec2016

Jun Dec2017

Jun Dec2018

BRL billion

Chart 1.1.10 – Structural liquidity ratio (ILE)

Available stable funding

Required stable funding

Structural liquidity ratio (ILE)

un.

10 1 10

24

8

10

6710 6 6

16

13

8

71

0

14

28

42

56

70

< 0,8 0,8-0,9 0,9-1,0 1,0-1,1 1,1-1,2 1,2-1,3 ≥ 1,3

Chart 1.1.11 – ILE Frequency distribution1/


Structural liquidity ratio (ILE)

1/ The numbers above the bars are the number of financial institutions with ILE within the corresponding interval.

% of system assets

Statistical annex

Statistical annex




capital market financing remained the highlight between June and December 2018.

In line with the recent reduction in the basic interest rate of the economy, average interest rates on bank credit lending fell throughout the year and in the second half of 2018 (average interest rate were 25.6% p.y. in December 2017, 24.6% p.y. in June 2018 and 23.2% p.y. in December 2018). This interest rate drop occurred for both individuals and companies loans; in the first case, average interest was 31.7% p.y., 30.9% p.y. and 29.0% p.y. in December 2017, June 2018 and December 2018, respectively; for companies, credit granting average interest rates were 16.9% p.y., 15.6% p.y. and 14.7% p.y., respectively on the same periods. Total loan portfolio increased 4.0% p.y. between June and December 2018 (5.1% p.y. in 2018), with growth levels of 5.6% for credit to families (8.4% p.y. in 2018) and 2.1% p.y. (1.3% p.y. in 2018) for credit to companies.

The resumption of the National Financial System (SFN) credit growth – still below its long-term trend –, the reduction of problem assets portfolio and the stability of the coverage index are factors that corroborate the perception that SFN credit risk is dissipating gradually and continuously.

1.2.2. Broad credit and long run trend

The Basel Committee on Banking Supervision (BCBS) and the international literature12 consider the credit-to-gross domestic product (GDP) gap a good metric to assess whether the growth of the credit outstanding in a country is sound regarding its long-term trend. This gap could signal an excessive increase in credit granted to companies and households, which could result in abrupt adjustments. Therefore, BCBS suggests that countries with the aforementioned gap above 2.0 p.p. should consider adopting measures to reduce credit growth.

Currently the credit-to-GDP gap is negative at 5.6 p.p. of GDP. Albeit the credit granting improvement in the last semester, the gap trend is expected to remain negative in the short run: the main contributions come from bank credit, that, in almost all facilities, continues growing below the long-term, especially the earmarked

12/ Drehmann, M., Borio, C., and K. Tsatsaronis (2011): “Anchoring countercyclical capital buffers: the role of credit aggregates”, BIS Working Papers, no 355. Drehmann, M., and Juselius, M. (2011): “Evaluating early warning indicators of banking crises: Satisfying policy requirements”, BIS Working Papers, no. 421.

-10

-5

0

5

10

15

Dec1997

Dec2000

Dec2003

Dec2006

Dec2009

Dec2012

Dec2015

Dec2018

Chart 1.2.2.1 – Credit/GDP gap without FX variationBy financing type

External market Capital marketCompanies - earmarked credit Companies - non-earmarked creditHousehold credit Total

p.p.of GDP

Source: BCB, Cetip, CVM, Previc and IBGE

Statistical annex



loans to companies. The exceptions—i.e., the positive contributions to the gap—are the capital market and, to a lesser degree, the external market (Chart 1.2.2.1).

1.2.3. Companies

Although the recovery scenario for non-financial corporations in 2018 showed signs in different directions, there was evidence of an improvement in the credit for Brazilian companies, especially in non-earmarked loans and in the capital market. This occurred despite the uncertainties that permeated the Brazilian economy in 2018. Albeit there are elements that suggest the continuity of the companies’ economic recovery, such as the increase in profitability of listed companies and the growth in demand for corporate credit from medium and large companies13, the judicial recovery requirements in 2018, in relation to the observed in 2017, increased for large companies, despite having slightly decreased for all companies (Chart 1.2.3.1).

Regarding domestic bank credit, the recovery was driven by non-earmarked credit, which continued to grow in the second semester of 2018 and offset the reduction in earmarked loans (Chart 1.2.3.2). Overall, at least nominally, the amount of companies’ financing has recovered to the amount of 12 month earlier for both small and medium-sized enterprises (SMEs) and large companies.

In the 12-m period, the outstanding of earmarked loans increased by 8.0% for SMEs and 8.3% for large companies, the latter was also influenced by the appreciation of the dollar in the period (discounting for the FX (foreign exchange) variation, the credit outstanding would have grown by 5.9% in the period).

External credit had a light increase in 2018, when measured in US dollars (Chart 1.2.3.3). This movement demonstrates that, despite the increase in interest rates in developed economies and the appreciation of these countries’ currencies, there is no scarcity of resources to Brazilian non-financial corporations in the international market.

13/ According to Serasa, the Indicator of Credit Demand by Companies in December 2018 reached 72.5 for medium companies and 149 for large companies, compared to 58.7 and 121.4 respectively in June 2018. Available at https://www.serasaexperian.com.br/amplie-seus-conhecimentos/indicadores-economicos.

0

400

800

1,200

1,600

2,000

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Chart 1.2.3.1 – Corporations in judicial recoveryCumulative requests by year

2012 2013 2014 2015 2016 2017 2018

Source: Serasa Experian

-20

-10

0

10

20

30

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

%

Chart 1.2.3.2 – YoY credit growthBy company size

SME - Total SME - Non-earmarked creditSME - Earmarked credit Large corporates - TotalLarge corporates - Non-earmarked credit Large corporates - Earmarked credit

80

110

140

170

200

230

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.3.3 – Corporate indebtednessDec/2013 = 100

Non-earmarked credit Earmarked credit Capital market

External market External market - US$

Statistical annex

Statistical annex

Statistical annex





The capital market continued to be the highlight in 2018, increasing by R$25.4 billion in the second semester and R$64.6 billion in the year (7.3% and 20.8%, respectively).14 Despite the stable outstanding in the last quarter, an expressive amount of issuances continued to occur, but they were offset by a large payment in the period. It is important to note that the change in the financing structure has not been uniform within the capital market, being concentrated in a few companies and specific sectors such as “Health, Sanitation and Education”, “Telecommunications” and “Energy”.

It should be noted that, of the BRL 64.6 billion capital market increase in 2018, BRL 22.9 billion is in the SFN portfolio. In other words, in spite of the slight growth in bank credit (BRL 4.7 billion), the SFN financing to non-financial corporations considerably increased via capital market, disregarding the amount of companies under judicial recovery restructuring plans (approximately 30% of the capital market increase in the banking system in 2018).

For the coming months, it is expected the maintenance of the non-earmarked credit growth, especially of less risky credit modalities. Regarding the capital market, it is expected the continuity of the issuances upward trend, given the scenario of low interest rates, strong liquidity and search for profitability by investment funds and other participants of the market.

Credit risk indicators remained stable in the second half of 2018 for companies, although with distinct directions regarding the companies’ sizes (Chart 1.2.3.4). Among SMEs, to which the main source of financing is domestic banks credit, the share of problem assets decreased by 1.36 p.p. in the semester and 2.40 p.p. in the 12-m period, with problem asset portfolio decrease along with credit outstanding increase, after a long decrease period (Chart 1.2.3.5). On the other hand, among large corporates, the share of problem assets increased in the semester (+0.93 p.p.), mainly due to the increase of “E to H” rated performing loans in “Transportation”, “Telecommunications”, “Construction, Wood and Furniture” and “Media and Leisure” sectors, and, to a

14/ The analysis of the capital market as an alternative financing for the banking system can be found in the “Selected Issues 2.2 - Companies debt prepayment, funding migration and foreign exchange hedge” of the October 2018 FSR and in boxes on the subject in the Inflation Reports published in 2018. In this FSR, for additional information of the debenture market, see “Selected Issues 2.1 - Developments in the Debentures Market”.

0

3

6

9

12

15

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

% Total Credit

Chart 1.2.3.4 – Problem assets

Corporate Small and medium enterprises Large companies

70

90

110

130

150

170

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.3.5 – Problem assetsSmall and medium enterprises (Dec/2014 = 100)

Problem assets Total credit

0

100

200

300

400

500

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.3.6 – Problem assetsLarge companies (Dec/2014 = 100)


Statistical annex

Statistical annex

Statistical annex

https://www.bcb.gov.br/ingles/estabilidade/2018_10/fsrSection2_2.pdf

https://www.bcb.gov.br/ingles/estabilidade/2018_10/fsrSection2_2.pdf





lesser extent, the delinquency of the “Sugar cane” sector and legal entities overseas (Chart 1.2.3.6 and Table 1.2.1).

Summing up, given the recent indicators of economic activity that showed a gradual recovery of the economy, there are signals of an improvement in the credit market for companies, especially for SMEs. The downward trend of credit growth to companies, which begun two and a half years ago, reversed. The capital market should continue to increase due to the improvement of the economic framework and the concession and privatization plans of the new government. There is growing evidence that the risk materialization cycle for the SME loan portfolio is over, but it should still last for large companies.

1.2.4. Households

The gradual economic growth in 2018 contributed to the continuity of the increase in credit to individuals. The slight reduction of unemployment15 and the improvement in household income16, coupled with historically low inflation and interest rates17, have positively contributed to lower levels of household indebtedness18 over the last few years. Consumer confidence reversed the downward trend, from mid-2018 on, returning to the levels of the first quarter of 2014 (Chart 1.2.4.1). All of these factors contributed positively to the highest YoY growth of household credit portfolio since 2015.

The most impacted credit facilities in the second semester of 2018 were those linked to consumption, such as credit card, vehicle financing and non-payroll deducted personal credit, with YoY growth higher than 10% p.y. at the end of 2018 (Chart 1.2.4.2).19 Nonetheless, it should be noted that both vehicle financing20 and non-payroll deducted personal credit have had reasonable periods of negative growth in recent years and have returned to a positive figure as of the second semester of 2017. The expectation is that credit to households endures, given the level of

15/ BCB Time Series no. 24,369 – Unemployment rate – Continuous National Household Sample Survey (PNADC)

16/ BCB Time Series no. 24,382 – Real habitually average earnings of employed people – Continuous PNAD

17/ BCB Time Series no. 433 – National household price index (IPCA) – IBGE e no. 4,390 – Brazilian Benchmark Interest Rate – Selic montly accumulated.

18/ BCB Time Series no. 19,881 – Household debt service ratio – Seasonally adjusted data%.

19/ Non-payroll deducted personal credit accounts for 5.8% of the total credit to households.

20/ For more details about the vehicle financing, see selected issues 2.2, “Evolution of the vehicle financing to households”, in the Chapter 2 of this Report.

Table 1.2.1 – Problem assets by economic sectorLarge companies

Dec/17 Jun/18 Dec/18

Public administration by economic sector 2.3% 2.5% 2.0%Agriculture 3.0% 3.1% 3.7%Food 5.7% 4.3% 3.9%Financial activities 0.2% 0.2% 0.2%Automotive 2.4% 2.7% 2.6%Beverages and tobacco 2.4% 3.9% 3.4%Construction, wood and furniture 33.0% 33.1% 38.3%Animal production 1.8% 2.5% 2.2%Eletric and electronic components 2.3% 1.4% 1.1%Energy 2.6% 1.9% 2.2%Machinery and equipment 9.2% 9.4% 7.6%Media and leisure 2.1% 2.3% 6.0%Others 15.1% 16.9% 20.9%Pulp and paper 1.3% 1.2% 0.8%Petrochemical 2.3% 1.7% 1.5%Legal entities located overseas 14.1% 15.9% 27.4%Chemicals, pharmaceuticals and hygiene 0.2% 0.2% 0.1%Health, sanitation and education 3.1% 2.1% 3.4%Services 8.4% 7.6% 8.5%Metallurgy 3.8% 2.9% 3.1%Sugar cane 16.1% 17.1% 20.1%Telecommunications 24.1% 20.5% 26.7%Textiles and leather 6.1% 10.1% 9.9%Transport 16.6% 15.3% 17.7%Retail 2.0% 2.2% 1.4%

% total credit outstanding

80

92

104

116

128

140

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

%

Chart 1.2.4.1 – Consumer confidence index

Source: Fecomércio

-30

-15

0

15

30

45

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Real estate financing Payroll deductedCredit card Non-payroll deducted personal creditVehicles financing Total

%

Chart 1.2.4.2 – YoY credit growthHousehold credit facilities

Statistical annex

Statistical annex

Statistical annex





economic recovery and the possible confirmation of the expectation captured in consumer confidence indexes.

Regarding credit cards, there has been an even more pronounced increase in the credit card purchases facility since September 2018, in addition to the maintenance of growth of payroll backed card balance and installments from revolving credit facilities21 (Chart 1.2.4.3)22. If, on the one hand, credit card purchases followed the resumption of household consumption, on the other, the evolution of payroll backed card balance is the result of more intense offering of this product by some financial institutions. The installments from revolving credit are directly related to credit card purchases and, therefore, accompanied the growth of this facility23.

The increase of average invoice payment (Chart 1.2.4.4)24 continued to occur during the second semester of 2018, at higher levels than the ones before the prohibition of outstanding revolving credit for a period exceeding thirty days. The growth, verified during 2018 and even higher in the last months, shows the effectiveness of the regulation adopted in 2017.

Regarding the risk analysis, it is possible to observe the maintenance or decrease of problem assets in the majority of credit modalities throughout 2018 (Chart 1.2.4.5). Indeed, the only modality with a recent increase is the non-payroll deducted personal credit, but in amounts well below the peaks of 2015 and 2016. It is important to mention that, despite the portfolio risk reduction since the end of 2016, the decrease in the problem assets ratio during 2018 was due to the increase in the credit outstanding, and not to the reduction of the problem assets portfolio (Chart 1.2.4.6).

21/ The credit facility “installments from revolving credit” shows the outstanding of the revolving credit operations that were divided in installments in better conditions to the debtor.

22/ The information presented in this section regarding credit cards may differ from other BCB publications, once the credit card facilities are classified differently.

23/ According to Resolution no. 4,549, of January, 26, 2017. This resolution made it possible for the invoice financed through a revolving operation to be the object of an advantageous offer to the debtor after the expiration of the subsequent invoice.

24/ Data from the Credit Bureau of this Central Bank was used to calculate a proxy of the credit card invoice. Firstly, it is estimated the invoice to be paid in the subsequent month, per individual: this is done through the Credit Bureau information of the payment due within 30 days, as well as revolving and overdue credits (which enter in full on the monthly invoice). Then it is calculated how much of the invoice was not paid, measured from the increased amount in the revolving balances or invoice installments. The difference between these two figures is considered as the amount that was actually paid from the monthly invoice.

0.0

3.6

7.2

10.8

14.4

18.0

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.4.5 – Problem assets Household credit facilities

Real estate financing Payroll deductedCredit card Non-payroll deducted personal creditVehicles financing Total

%

50

60

70

80

90

100

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

%

Chart 1.2.4.4 – Credit card invoice – Average payment

0

40

80

120

160

200

0

8

16

24

32

40

Jan2014

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.4.3 – Credit card by modalities

Revolving credit up to one month Installments from revolving creditRevolving credit past one month Payroll backed credit card balanceInstallments financed purchases Credit card purchases (L)

BRL bnBRL bn

Statistical annex

Statistical annex

Statistical annex





Summing up, the credit portfolio to households shows an improvement in risk levels and resumption of loan granting, especially in more consumer-related modalities. Credit to individuals is expected to maintain its upward trend throughout 2019, given the continuity of the economic environment improvement.

1.2.5. Domestic banking credit by ownership

The domestic banking credit has maintained the recovery trend for the last six months for banks of all type of ownership. The nominal credit growth reached 5.5% in 2018 (Chart 1.2.5.1). The recovery continues to happen with greater strength in private banks whose portfolio registered rates higher than in 2012. The public banks have also shown signs of recovery, although the public development banks still present annual negative variation.

The recovery of the domestic credit in private banks is also demonstrated by the increase in their credit lending (Chart 1.2.5.2). In the second half of 2018, private banks’ credit lending returned to the levels observed in 2014. For the next semester, it is still expected the credit growth concentrated in the private banks.

1.2.6. Risks and provisioning

Considering the gradual recovery of the macroeconomic environment, the credit risk indicators continue to show improvements in the portfolio quality. Problem assets decreased by 0.08 pp in the last semester, recording a decrease of 0.49 pp compared to December 2017 (Chart 1.2.6.1). The improvement in the problem asset portfolio has been mainly impacted by the decreased of the delinquency. The 90 days past due indicator has maintained its downward trend which began in the first half of 2017 in almost every bank segment.

The problem assets index has presented different trends by type of ownership (Chart 1.2.6.2). The private banks have been keeping the downward trend of problem assets since 2017, as a result of their better quality credit lending, especially in the household portfolio. Regarding the public development banks, the problem asset has still presented growth due to the increase of the delinquency of the large companies’ loans. Problem assets of public commercial banks have remained nearly stable throughout the last semester. This occurs because

100

106

112

118

124

130

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 1.2.4.6 – Problem assetsHousehold (Dec/2014 = 100)


-20

-10

0

10

20

30

40

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

%

Chart 1.2.5.1 – Year over year credit growth By ownership

Private banks Public commercial banks

Public development banks Total

0

50

100

150

200

250

300

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

Chart 1.2.5.2 – Credit lendingDeflated and seasonally adjusted – By ownership

Private banks Public commercial banks

Public development banks

BRL billion

Statistical annex

Statistical annex

Statistical annex





even with a decrease in the intensity of loss recognition, the discrete growth of the portfolio prevented significant reductions in the indicator (Chart 1.2.6.3).

The delinquency decrease was mainly caused by the reduction on the inflow from performing loans observed in the last quarter (chart 1.2.6.4). The current course of the delinquency flow strengthens the perception of the end of a cycle of loss materialization in the credit portfolio and a probable new cycle of credit in the economy.

Although the trend of improvement in the credit portfolio should continue, the downward trend of the problem assets is expected to lose strength in the next semesters, as it was already mentioned in the last FSR. This perspective is based on the assessment that the household credit portfolio may have completed their losses materialization related to the crisis period, and the loans to corporate will still tend to present a significant volume of renegotiation.

The coverage index (CI) of problem assets stayed above 80% in the last six months (Chart 1.2.6.5). The maintenance of CI level, even in the most severe risk materialization period (2015/2016), reinforces the perception of the sufficiency of this coverage level as well as the capability of National Financial System to maintain provisions to support losses compatible to their assumed risks.

1.3 Profitability

The banking system continued to increase its profitability in the second half of 2018, reaching pre-crisis levels. In the last two years, public banks25 profit have grown in a faster pace, approaching private banks profitability levels. The reduction in loan-loss provisions (LLP) and funding costs and operational efficiency gains are the main reason for the improvement in the banking system profitability, despite the reduction in treasury incomes and the stagnation of corporate loans portfolios26.

The change in the portfolio mix and the recovery of credit growth, especially to households and small and medium-sized enterprises27, should be positive for banks profits in the coming years. Another important issue that can benefit the profitability is the improvement in operational

25/ Public banks ROE was affected by an extraordinary LLP related to the real estate loan portfolio at the end of 2018.

26/ For more details, see chapter 1, item 1.2 of this Report.27/ For more details, see chapter 1, item 1.2 of this Report.

0

2

4

6

8

10

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

%

Chart 1.2.6.1 – Problem assetsBy component

90 days past due loans90 days past due or restructured loans90 days past due or restructured or "E to H" rated loans

0

2

4

6

8

10

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

Private banks Public commercial banksPublic development banks Total

%

Chart 1.2.6.2 – Problem assetsBy ownership

0

10

20

30

40

50

1stSem2014

2ndSem

1stSem2015

2ndSem

1stSem2016

2ndSem

1stSem2017

2ndSem

1stSem2018

2ndSem

Chart 1.2.6.3 – Restructured loan flow accumulated in the semester

Private banks Public commercial banksPublic development banks

BRL billion

Statistical annex

Statistical annex

Statistical annex





efficiency, as a result of efforts to restrain administrative expenses and to improve services income related to digital banking solutions. However, it is expected a slowdown in the LLP reduction process, which, together with the lower margins in the investment securities portfolio, tend to stabilize the net interest margin (NIM).

Smaller banks operating with “Corporate Credit”28 were deeply affected by the economic crisis and by the unfavorable environment for large companies in Brazil. The ones operating with “Treasury and investment activities”29 were pressured by the lower Selic level. These two groups of banks were, in general, the most affected by the recession cycle. In this context, some adjustments in business models may be necessary to improve their profitability.

In general, a slowdown in the profitability growth path is expected, but without increasing risks to the financial stability of the banking system.

The banking system Return on Equity (ROE)30 reached 14.8% in December, an elevation of 0.4 pp compared to June 2018 (Chart 1.3.1). This improvement was driven by the factors discussed in the first paragraph of this topic. The comparison between ROE and the risk-free rate proxy also confirms the recovery cycle31. The profitability risk premium improved due to the higher ROE but also due to the reduction in the risk-free rate proxy, as it incorporates the effects of the Selic decline.

In the context of the Selic lowest historical level, the net interest margin32 remained stable throughout the second half of the year. The net credit margin slightly increased33, influenced by the growth in household loans, as well as in Small and Medium-sized Enterprises (SMEs). These groups usually have higher interest spread.

Risk-adjusted interest margin maintained an upward movement throughout the semester, mainly due to LLP

28/ “Corporate Credit” segment is comprised by banks that focus on corporate consumers (corporate = balance contracts > R$1 million).

29/ “Treasury and investment activities” segment is composed of banking conglomerates that greatly rely on treasury and business operations (bonds, repurchase agreements - repo - and investments) as a source of income.

30/ The adjustment intents to restrict the analysis to recurring values.31/ The risk-free rate proxy in this report is defined by the average Selic

annual rate over the last 36 months multiplied by 0.85 to consider tax effects. The loan portfolio’s weighted average maturity, the main source of banking revenue, based the choice for the 36-month period.

32/ In this report, net interest margin represents the difference between the return on the loan and securities portfolio and the cost of funding.

33/ Banks funding interest rate in Brazil is predominantly post-fixed while loans portfolio interest rate is predominantly pre-fixed.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

-0.9

-0.6

-0.3

0

0.3

0.6

0.9

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

Chart 1.2.6.4 – Delinquency flow

Inflow from performing loansWrite-offsOutflow to performing loansOther outflows (liquidated, transfered or restructured loans)90 days past due loans (left)

%%

0.5

0.6

0.7

0.8

0.9

1.0

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

Chart 1.2.6.5 – Coverage index of problem assets

0

22

44

66

88

110

Jun2012

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec0

5

10

15

20

25

Net income (left axis)1/ROE 1/Profitability risk premium: ROE (-) risk-free rate proxyPublic banks ROE 1/Private banks ROE 1/

BRL billion %

1/ Identified non-recurrent adjustments.

Chart 1.3.1 – Return on equity (ROE)1/

Trailing twelve months

Statistical annex

Statistical annex

Statistical annex





reduction. This is a key factor for the net income increase over the last year (Chart 1.3.2). Nevertheless, LLP has shown a flat trend in recent months and, together with the NIM stabilization, has contributed to the net income stabilization (Chart 1.3.3).

Services income has kept growing but in a slower pace when compared to the previous semester. Increases in bank account fees and payment card fees have still been significant. Investment fund service fees have presented a slower growth. Social program funds34, which are usually managed by public banks, were the main driver of the increase in the semester. The lower raise in assets under management, which has been the most important driver in the recent years, mainly explained the fund fees behavior. In addition, the current level of the Selic rate pressured the management fee in fixed-rate funds. Under such a scenario, incomes from investment fund services tend to drop down in the coming semesters unless banks change their strategy and the product profile.

The coverage ratio (services income versus administrative expenses) has kept its positive path in the second half of the year. Administrative expenses (trailing twelve months) increased 0.3% between June and December 2018, compared to a 3.1% growth in services income (Charts 1.3.4). The expectation35 of operational efficiency gains over 2018 was confirmed due to the positive effects of cost restraint policies and organizational restructuring36 in recent years as well as increasing incomes due to digital banking solutions. The outlook is still the improvement of operational efficiency in the next semester.

The asset share and number of banks with profitability above the risk-free rate proxy continued growing in the last half-year. The number of banks with net loss decreased from 22 to 21 between June and December 2018, maintaining the reduction trend over the last eighteen months. However, the recent stabilization of the frequency distribution indicates that banks may be reaching their profitability equilibrium levels, while some of them still face challenges to realign its market position after the recession period (Chart 1.3.5).

34/ These include funds for workers and students such as: Fund for period labored (FGTS), salary variation compensation fund (FCVS), Social Integration Program (PIS), Student Financing Fund (FIES), Federal Lotteries, among others.

35/ According to the fourth paragraph of item 1.3 of the Financial Stability Report (Volume 17 | Number 2 | October 2018).

36/ Voluntary Dismissal Programs and/or shutdown of branches plans announced in the public media.

.0

3.20

6.40

9.60

12.80

16.0

Dec2012

Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

%

Net interest margin (Credit + Securities)Risk-adjusted net interest margin (Credit + Securities)Funding costInterest income rate (Credit + Securities)

Chart 1.3.2 – Net interest marginTrailing twelve months

0

1

2

3

4

5

0

30

60

90

120

150

Jun2012

Dec Jun2013

Dec Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

BRL billion

Net provision expenses (trailing 12 months)Net provision expenses (trailing 3 months)Net provision expenses / Credit portfolio (left axis)

Chart 1.3.3 – Provision expenses

%

39

116

40

119

42

121

44

121

47

123

48

12264

101

65

101

63

99

60

102

62

105

65

106

25 25 26 28 29 3013 13 14 14 14 14

64.8 65.4 65.7 65.3 66.868.7

0

14

28

42

56

70

0

50

100

150

200

250

Jun2016

Dec Jun2017

Dec Jun2018

Dec

BRL billion

Other administrative expensesPersonnel expensesFees: debt collection and guaranteesRevenues: funds and capital marketRevenues: cards and othersService feesServices income versus administrative expenses (left axis)

Chart 1.3.4 – Main components of administrative expenses and services incomeTrailing twelve months%

Statistical annex

Statistical annex

Statistical annex





Although smaller banks have already shown signs of recovery due to a better economic environment, some groups most affected by the recession period still need to overcome certain challenges to achieve a robust recovery in profitability. Banks operating with “Corporate Credit” face adversities such as: a) higher risks in the large companies market when compared to the pre-crisis period and b) companies with better credit rating looking for alternative sources of funds, outside of the banking system (capital markets and debentures37). At the same time, the profitability of the banks in the “Treasury and investment activities” segment has been strongly influenced by the Selic rate lowest historical level, which directly affects their net interest margin. Among strategies to raise profitability identified in some of these banks is the diversification of income sources, such as: a) digital banking solutions to expand into new business niches; b) diversification of financial services; and c) increment of private securities portfolio.

1.4 Solvency

The banking system solvency reported a positive trend during the second half of 2018, not constituting a concern for financial stability. Capital and leverage ratios remain in robust levels, substantially higher than regulatory requirements (Chart 1.4.1), highlighting that the system is prepared for the fully-fledged Basel III timetable conclusion38.

The banking system´s Common Equity Tier 1 ratio (CET1) reached 13.3%, reflecting the increase in profits – mostly due to the reduction in loan loss provisions and the operational efficiency improvement – as well as the prudential adjustments decrease, particularly related to deduct Deferred Tax Assets (DTA) arising from tax losses, as a result of the Resolution CMN no. 4.680/1839

37/ For more details on this trend, see Section 2.1 of this Report.38/ From January 1st, 2019 on, the F factor timetable schedule will be fully

implemented, reducing from 8.625% to 8.0% according to the art. 4 of the Resolution CMN no 4.193, of March 1st, 2013, which will affect RWA for market and operational risk, as well as the banking portfolio (RBAN). Also from January 1st, 2019 on, the phase-out factor for debt capital instruments issued before Basel III took place will evolve from 40% to 30%, according to the art. 28 of the Resolution CMN no 4.192, of March 1st, 2013. The Basel III timetable for common equity tier 1 capital is already finished.

39/ The Resolution CMN no. 4.680, of July 31st, 2018, enabled financial institutions not to deduct DTA arising from tax losses generated from hedging foreign currency investments, originated between January 1st, 2018 and December 31st, 2019. The referred Resolution established a deduction timetable, according to the following: (i) at least 50%, up to June 30st, 2020; (ii) 100%, by December 31st, 2020.

22 42

66

2137

73

0

20

40

60

80

100

ROE < 0 0 < ROE < Proxy²/ ROE > Proxy²/


1/ Values above the bars represent the number of financial institutions in the correspondent ROE range.2/ Risk-free rate proxy.

% Assets

Chart 1.3.5 – Return on equity (ROE) frequency distribution1/

12.613.3

13.8 14.5

17.2 17.9

7.2 7.6

0

4

8

12

16

20

Jun2016

Dec Jun2017

Dec Jun2018

Dec

%

CET1 Ratio T1 Ratio Capital Ratio Leverage Ratio

Chart 1.4.1 – Capital Ratios and Regulatory Requirements1

1/ The arrows represent the regulatory requirements for different capital levels, including Conservation Buffer( 6.375% for CET1 Capital, 7.875% for T1 Capital, 10.5% for Total Capital and 3.0% for Leverage ratio). The leverage ratio data considers only Commercial, Multiple, Foreigh Exchange, Investment and Saving banks classified as S1 or S2, according the Resolution CMN no 4.615/17.

0.47 0.52 0.27

0.63 0.55 0.57 0.59

-0.38 -0.62

-0.27 -0.78 -0.93

0.35

-0.28

0.54

0.49

0.34

-0.35

-0.23

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Dec2015

Jun Dec2016

Jun Dec2017

Jun Dec2018

Chart 1.4.2 – CET1 Variation DecompositionHalf-year variation (p.p.)

Share capital and Profit Reserves Prudential AdjustmentsOther CET1 components RWA for credit riskRWA for market risk RWA for operational riskCET1 change

Percentagepoints

Statistical annex

Statistical annex

Statistical annex





(Chart 1.4.2). The system´s risk weight assets, in turn, presented a slight growth (2.1%), especially in the credit portfolio.

Relevant new issuances of debt capital instruments in the second half, particularly in private banks, strengthen the replacement process of Basel III non-compliant instruments to the compliant ones, preserving their share in capital structure even considering the common equity tier 1 (CET1) growth in the period (Chart 1.4.3).

Public banks maintained higher earnings retention rates, preserving the strategy of strengthening their capital base after a period of strong credit growth and deterioration of credit quality, which resulted in capital restrictions for these entities. Private banks remain with higher profit distribution, albeit with an increasing retained earnings trend, reflecting the beginning of a recovery cycle (Chart 1.4.4).

The banking system´s risk weight assets (RWA) increased by 2.1%, mainly due to the RWA´s expansion in private banks (2.9%) and, to a lesser extent, in public banks (0.5%). The RWA for credit risk reported the same dynamics, particularly due to repo and retail credit operations growth.

Reflecting the positive path of the banking system solvency, the fully-fledged CET1 ratio histogram shows that 129 institutions, accountable for 99.9% of the system´s total assets, reported ratios above 7.0%, which is the minimum requirement for 2019 on (Chart 1.4.5). Moreover, the projected capital shortfall is just BRL 0.6 billion in December 2018, representing 0.11% of the current system´s total capital.

Considering the fully-fledged Basel III framework, capital and leverage ratios presented a steady trend (Chart 1.4.6), with immaterial projected capital shortfall. Prospectively, the gradual recovery of credit growth may materialize in a greater risk weight assets increase in the forthcoming periods. However, the alignment and complete adaptation to the Basel III framework, combined with a strong earnings-generating capacity, may contribute for the stability of the system´s capital ratios, keeping capital surplus in comfortable levels.

489 490 494 490 519 506 545

38 32 38 38 37 485298 93 94 83 82 757254 40 40 56 60 6267

0

200

400

600

800

1,000

Dec2015

Jun2016

Dec2016

Jun2017

Dec2017

Jun2018

Dec2018

BRL bi

Chart 1.4.3 – Capital Structure

CET1 AT1 T2 (Basel III non-compliant) T2 (Basel III compliant)

59.9%

56.4% 50.7%

32.2% 28.0% 36.8%

67.6% 69.2%

55.4%

0

20

40

60

80

100

0

20

40

60

80

100

2011 2012 2013 2014 2015 2016 2017 2018

BRL bi

Chart 1.4.4 – Dividends and Interest on own capital

Accounting net profitProfit distribution (Dividends plus Interest on own capital)(%) Profit distribution Banking System (left axis)(%) Profit distribution Public banks (left axis)(%) Profit distribution Private banks (left axis)

(%)

1 1 0

7

121

2 0 09

120

0

20

40

60

80

100

0 - 4.5 4.5 - 5.125 5.125 - 7 7 - 10.5 >10.5

Simulation in June 2018 Simulation in December 2018

% system assets

Chart 1.4.5 – Fully-fledged Basel III – Common Equity Tier I (CET1)¹

Histogram

1/ Inside the bars are the # of financial institutions in the correspondent CET1 ratio range

Statistical annex

Statistical annex

Statistical annex





1.5 Capital stress tests

Capital stress tests are financial stability tools that assess the resilience of the banking system related to its ability to absorb losses in adverse macroeconomic scenarios. The tests simulate effects on the banking system’s capital adequacy ratios, stemming from extreme shocks in the main economic-financial variables. In addition, simulations of sensitivity analysis to the main risk factors undertaken individually and contagion among financial institutions40 complement the analysis.

Stress tests results indicate that the banking system maintains its loss absorbing capacity against all simulated shocks, with no relevant capital shortfalls due to noncompliance41 or insolvency events. The results are the consequence of the appropriate capitalization cushion, as well as the resilience of banks´ profitability even under extreme scenarios.

Sensitivity analysis indicates a low impact on banks` capital assuming severe shocks on FX (foreign exchange) rates, once mostly banks exposures to exchange rates are hedged. Nevertheless, shocks on interest rates point to some risks arising from abrupt increases in rates. With regard to credit risk, the system showed a small increase in capital shortfalls compared to the previous simulation (June, 2018). The sensitivity to residential real estate prices also demonstrated slightly higher capital shortfalls with respect to the June 2018 simulation, albeit not reflecting any relevant risks from exposures to mortgages on banks՚ balance sheet.

1.5.1 Scenario analysis – Macroeconomic stress tests42

Table 1.5.1.1 displays the economic variables for the different stress test scenarios: Baseline, Structural Break and Worst Historical. Each scenario brings different stressed values in each quarter of the test horizon, and only the values of the last quarter are presented.

As of December 2018, the time horizon for each scenario was extended to twelve quarters, against six until the

40/ The scope of the contagion simulation reach all authorized institutions to operate by the BCB, except consortiums. The scope of macroeconomic stress tests comprehends only banks.

41/ A financial institution is considered as non-compliant if it does not comply with at least one of the capital requirement ratios: total capital ratio, tier 1 and common equity tier 1.

42/ The stress test assumptions are in accordance with the Resolution no 4,680.

12.5 12.4 13.1

13.3 13.514.3

15.5 15.6 16.1

7.0 7.1 7.6

0.0

3.5

7.0

10.5

14.0

17.5

Dec-17 Jun-18 Dec-18 Dec-17 Jun-18 Dec-18

Capital Ratio Leverage Ratio

(%)

Chart 1.4.6 – Fully-fledged Basel IIICapital and Leverage Ratios

CET1 Ratio T1 Ratio Total Capital Ratio Leverage Ratio* Fully-fledged Basel III simulation considers the phase-out factor of debt capital instruments, according the Resolution CMN no 4.192/13 and no 4.679/18. The leverage ratio data considers only Commercial, Multiple, Foreign Exchange, Investment and Saving banks classified as S1 or S2, according the Resolution CMN no 4.615/17.

Statistical annex



last FSR. The Structural Break scenario is obtained by applying the observed changes of economic variables in previous periods on the current levels by making use of a quarterly rolling window. The financial system´s most unfavourable historic path within an eight quarters horizon is chosen, for each variable independently. The Worst Historical scenario simulates the historical behaviour of each variable by choosing the patterns observed in a twelve-quarter rolling window since July 2003, that would result in the banking system´s lowest earnings before taxes.

The Chart 1.5.1.1 features the proportion of problem assets to total credit portfolio, for all scenarios. Under the worst-case scenario, the problem assets do not reach 10% of the total loan portfolio at the end of the simulation, even with the three-year scenario.

The estimated additional43 capital to avoid both noncompliance as well as dividends distribution limitations amounts to 0.9% of the current regulatory total capital as shown in Chart 1.5.1.2. The capital shortfall of

43/ The concept of capital shortfall encompasses the amount necessary to avoid both minimum capital non-compliances as well as limitations on profits distributions imposed by Resolution no. 4,193, from March 1st, 2013, in which systemically important financial institutions are subject to the systemic buffer requirement.

Table 1.5.1.1 – Macroeconomic Stressed Scenarios (december/2021)

Structural Break Worst Historical

Output(IBC-Br) 1.2% 2.9% -4.5% -1.6%

Domestic Interest Rates (Selic) 6.5% 8.0% 11.9% 7.6%

Exchange Rate (BRL/USD) 3.81 3.80 5.48 5.49

Inflation(annual IPCA) 3.8% 3.7% 4.1% 0.8%

Unemployment(PNAD-C IBGE) 11.6% 11.6% 16.5% 22.2%

Country Risk(Brazil EMBI+)2/ 269 269 796 610

Foreign Int. Rates (US G. Bonds Yield 10yr)4/ 3.0% 1.9% 4.1% 2.6%

Scenarios

Variables Dec 2018 Base Scenario1/

4/ For the Baseline scenario, the trajectory of the US G. Bonds Yield 10yr was extracted from the Federal Reserve (FED) Adverse Scenario in Dodd-Frank Act Stress Testing (DFAST) 2018 (https://www.federalreserve.gov/newsevents/pressreleases/files/bcreg20180201a1.pdf). For the Structural Break and Worst Historical scenarios, the forecasts are based on historical behavior of the variable.

1/ GDP preview, Selic, FX and inflation rates were collected from the Focus survey published in December 31th, 2018. Both unemploymentand country risk remain constant.2/ The table shows the maximum values for the EMBI+Brazil in each scenario. For Structural Break scenario, the EMBI+Brazil peak of 796 isreached in June 2020. For the Worst Historical scenario, the maximum level is reached in September 2020.

3/ The method employed for building each scenario can be found in the annex Concepts and methodologies - Capital stress.

0.0

2.4

4.8

7.2

9.6

12.0

Dec2018

Mar2019

Jun Sep Dec Mar2020

Jun Sep Dec Mar2021

Jun Sep Dec

Base scenario Structural BreakWorst Historical Provisions (Dec 2018)

Chart 1.5.1.1 – Macroeconomic stress testProblem assets forecast (% total credit)

%

Statistical annex

Statistical annex




the financial system shows stability when compared to the results of the three previous semesters. (Chart 1.5.1.3).

The dispersion analysis of the total capital adequacy ratio demonstrates that most of the institutions would continue to show ratios above the minimum regulatory requirements (10.5%). This group represents about 90% of the total assets of the banking system (Chart 1.5.1.4).

1.5.2 Sensitivity analysis

Sensitivity analysis assesses the impact on the banking system´s capital arising from incremental changes in interest rates, foreign exchange rates, problem assets and residential real estate prices, all of them independently. In the case of the problem assets, only positive changes are considered; for mortgage prices, only negative changes are analysed. With regard to interest rates and FX (foreign exchange), changes in either way (positive or negative) are allowed since they can denote gains or losses for banks. The shocks alter both interest rates as well as FX (foreign exchange), individually and in steps of 10%, over a range of values with lower and upper bounds corresponding to 10% and 200% of the actual values, respectively.

The FX (foreign exchange) sensitivity analysis showed no capital shortfall within this interval. In the case of interest rate risk, if rates increased by 40% over the actual values, the additional capital needed in order to avoid non-compliance with capital requirements would amount to 2.8% of the total regulatory capital. If rates rose 100%, the capital shortfall would increase to 30.2% of total regulatory capital. It is worth noting that these are parallel shocks.

The results of the sensitivity test to incremental credit risk shocks (Chart 1.5.2.1) indicate that the problem assets would have to reach 12.0% of the total loan portfolio in order to the system require additional capital of 0.1% of total regulatory capital. This level is above the historical peak of 8.3% observed in December 200044. Under extreme scenarios, if the number of problem assets reached 18.1% of the total loans, there would be a capital shortfall equivalent to 2.9% of the total regulatory capital of the system, stemming from institutions that represent 31.1% of the total assets. This amount is lower than the one observed in June 2018.

44/ Before January 2012, the proportion of E-H rated loans to total credit portfolio is used for comparison.

15

17

19

21

23

25

0.0

0.4

0.8

1.2

1.6

2.0

Dec2018

Mar2019

Jun Sep Dec Mar2020

Jun Sep Dec Mar2021

Jun Sep Dec

CET1 (right) AT1 (right) Tier 2 (right) Capital ratio

Chart 1.5.1.2 – Capital requirement gap and Capital ratioStructural Break

Regulatory capital (%)Capital ratio (%)

0.0

1.4

2.8

4.2

5.6

7.0

Dec2018

Mar2019

Jun Sep Dec Mar2020

Jun Sep Dec Mar2021

Jun Sep Dec

Quarters Dec 2018 Jun 2018 Dec 2017 Jun 2017

Regulatory capital (%)

Chart 1.5.1.3 – Capital requirement gapStructural Break – Evolution

1 - 6 18

19

87

19 3 8 7 7

87

0

20

40

60

80

100

<= 8.5 8.5 - 10.5 10.5 - 12.5 12.5 - 14.5 14.5 - 16.5 >= 16.5

December 2018 (realized data) December 2021 (stressed scenario)

Chart 1.5.1.4 – Macroeconomic Stress TestFrequency distribution of system assets by Capital ratio band – Structural Break1/

System assets (%)

1/ The value above each bar represents the number of institutions per band.

Statistical annex

Statistical annex

Statistical annex





The assessment of changes in residential property prices indicates that there is no regulatory breaching or dividend distribution limitation for nominal price drops of up to 35%. Only a price slump of 40% or more would lead to insolvency, represented by negative regulatory capital (Chart 1.5.2.2).

In December 2018, the average loan-to-value (LTV) on mortgages outstanding balance was 62.2%, when revaluing collateral prices by the Residential Mortgage Collateral Value Index (IVG-R)45. Credit granted with low LTV, in addition to the utilization of the Constant Amortization System that reduces LTV during the lifetime of the operation, are healthy features for the mortgage loans and contribute to improving the loss absorbing capacity of the system under extreme scenarios.

Therefore, sensitivity analysis confirms that the Brazilian banking system presents sound loss absorbing capacity since relevant capital shortfalls would only happen under extremely adverse situations. The results of the stress tests simulations suggest that the banking system has an adequate capital cushion to withstand severe shocks stemming from the hypothetical worsening of economic fundamentals.

1.5.3 Simulation of direct interbank contagion

In addition to the macroeconomic and sensitivity stress tests, direct inter-financial contagion simulations take place, comprising all financial entities authorized by the Central Bank, except for consortiums. In this exercise, all direct national inter-financial exposures are taken into account, although second-order effects such as fire sales or liquidity are not considered.

In the assessment of direct inter-financial contagion, the individual failure of each financial institution is simulated, one at a time, and the impact on its counterparties is evaluated. If the failure of one institution leads to a breach on its counterparties, additional rounds are ran until a new equilibrium is found (domino effect). The impacts stem from the write-off of exposures to different instruments, such as interbank deposits, granting of guarantees, OTC (over the counter) derivatives, and any other entailing credit risk, in which there are neither third-party guarantees nor collateral. With the

45/ The IVG-R is calculated and disclosed by the BCB based on the values of real estate financing collateral.

1.81.10.6

0.0 0.2

18.0 14.0

0.1 19.9

36.4 35.1 31.1

-

2 .0

4 .0

6 .0

8 .0

1 0 .0

1 2 .0

1 4 .0

1 6 .0

1 8 .0

2 0 .0

0

4

8

1 2

1 6

2 0

2 4

2 8

3 2

3 6

4 0

4 4

4 8

5 2

5 6

6 0

6 4

6 8

7 2

7 6

8 0

8 4

8 8

9 2

9 6

1 00

1 04

1 08

1 12

1 16

1 20

1 24

1 28

1 32

1 36

1 40

1 44

1 48

1 52

1 56

1 60

1 64

1 68

1 72

1 76

1 80

1 84

1 88

7.2 8.3 9.4 10.5 11.6 12.7 13.8 14.9 15.9 17.0 18.1

Chart 1.5.2.1 – Sensitivity analysisCredit risk

Problem assets (%)

Banks in insolvency (% of Total Assets)

Noncompliant banks (% of Total Assets)

Capital ratio (%)

Capital requirement gap (% of Total Capital):

CET 1Additional Tier 1Tier 2

Maximum observedproblem assets

(since Dec/2000)

23.1

0.9 0.1

16.3

32.3

0.0

16.3

16.77.5

0 10 20 30 40 50 60 70 80 90

Capital requirement gap (% of Total Capital):

CET 1

Additional Tier 1

Tier II

Banks in insolvency (% of Total Assets)

Noncompliant banks (% of Total Assets)

Capital ratio (%)

Fall in housing prices (%)

Acumulated fall ofS&P Case-Shiller 10 (USA)

during subprime crisis -(Apr/06 to May/09)

Chart 1.5.2.2 – Sensitivity analysisHousing prices risk

Statistical annex

Statistical annex




bankruptcy simulation, the exposures identified cause losses to creditors, and the effects are evaluated from the perspective of the amount of capital required to prevent contagion to spread.

The results show a low capital shortfall in case of default of each institution separately. In the worst scenario, the figure is less than 1% of the regulatory capital of the entire system. Two points help explain this result. Firstly, there is a regulatory cap of 25% on exposures to any single counterparty, as a proportion of the creditor institution’s capital. Secondly, the great majority of inter-financial transactions occurs through repurchase agreements collateralized by federal bonds, which are assumed riskless. The remaining operations, although small in aggregate volume in the financial system, may be relevant in some particular cases, which explains the situation above described.

1.6 Financial Stability Survey

1.6.1 Introduction

This section presents the latest results from the Financial Stability Survey (FSS). The survey is conducted quarterly with selected financial institutions and aims to identify sources of risk to financial stability, as perceived by regulated institutions.

The FSS sample comprises 55 financial institutions, covering 94.4% of the Brazilian National Financial System (SFN) in terms of assets,46 including public banks, development banks, foreign banks, and Brazilian private banks with and without foreign shareholders.

Since the latest edition of the Financial Stability Report (FSR), two rounds of the FSS were run, from October 29th to November 5th, 2018 and February 5th to 14th, 2019, with response rates of 100% and 98%, respectively. This section compares results from these FSS rounds with those rounds occurred from August 2nd to 16th, published in the October 2018 FSR.

The BCB asked survey respondents about their perception of the main risks to the financial stability in the next three years, considering their occurrence probability and

46/ The share of respondents’ total assets relatively to the Brazilian National Financial System is evaluated with data from December 2018.


impact on the SFN.47 Each institution can freely describe up to three sources of risk, which are then classified by the BCB into different risk categories (Table 1.6.2.1).48

The citation frequency of risks related to the foreign scenario has increased in the last four quarters. The average citation frequency rose from 0.76 per institution in August 2018 to 1.13 per institution in the last survey. This increase is explained by the number of ways risks from the foreign scenario could affect the balance sheet of financial institutions. In addition to concerns about the effects of the withdrawal of monetary stimulus in the United States and the trade tensions between the United States and China, financial institutions believe that the level of world economic activity will possibly be lower than expected. Also, political instability in Europe could lead to unfavorable financial market conditions, hampering international liquidity conditions.

The concern with non-fiscal political risks fell relative to the August 2018 FSS, with a reduction in the average citation frequency from 0.22 to 0.19 per institution. The main descriptions of non-fiscal political risks focus on governability conditions, including those in adverse scenarios.

The average citation frequency of fiscal-political risks fell from 1.11 per institution in November 2018 to 0.96 in February 2019, reflecting confidence in the new government’s economic team. However, financial institutions consider that Brazil fiscal situation is still

47/ Question: “In the next three years, what are the risks to the financial stability that your institution considers most relevant considering probability and impact on the SFN? Describe the three risks in order of importance (the most important first, considering the combination of probability of occurrence of the event and the magnitude of the impact in terms of losses measured as a fraction of the total assets of the SFN).”

48/ Since the same institution can describe two or more risks that could later be classified into the same risk category (for example monetary policy in the US and trade war), the reported frequency does not necessarily correspond to the number of institutions that quoted a given risk category.

Note: the risk of “delinquency and activity” corresponds to the previous “delinquency and recession” risk. The new terminology amplifies the previously adopted concept by including recent concerns highlighted by respondents, such as the low level of economic activity. Additionally, the previous political risks and fiscal risks have been reclassified to represent, respectively, political risks not related to fiscal risks, and fiscal-political risks, which, in addition to the previous fiscal risks, include political risks associated with fiscal risks.

Table 1.6.2.1 – FSS – The most cited risk factorsProbability Impact

Aug 2018 Nov 2018 Feb 2019

Foreign Scenario 0.76 1.02 1.13 Mid-High Medium

Fiscal-political Risks 0.89 1.11 0.96 Mid-High High

Delinquency and Activity 0.55 0.44 0.37 Mid-Low Medium

Non-fiscal Political Risks 0.22 0.09 0.19 Mid-High High

RiskAverage frequency (citations/financial institution)

Feb 2019

Statistical annex



at worrying levels and that, if structural reforms are not carried out, the economic environment will remain unfavorable to business. Respondents consider important that the government succeed in approving reforms capable of stabilizing the fiscal balance of the public sector and the necessary measures for the economic growth recovery in the country.

The average citation frequency of risks related to delinquency and activity fell to 0.37 per institution in the last survey, in comparison with 0.55 in August 2018, reflecting the gradual recovery of the Brazilian economy together with an improvement in the credit quality.

Fiscal-political risks are the most important risk out of the three reported by respondents, being cited by 65% of institutions in February 2019 in comparison with 53% in August 2018. Such risks are reported with mid-high probability and high impact on the financial system (Table 1.6.2.2). The risks associated with the foreign scenario are stronger, reaching 20% in February 2019 against 9% in August 2018.

The comparison between the results of the current survey and those of August 2018 that respondents' perceptions of the probabilities of materialization of fiscal-political and foreign scenario risks have decreased49 (Chart 1.6.2.1).50

According to respondents, fiscal-political risks were those that showed the greatest decrease in the probability of materialization, without significant changes in the assessment of the impact that they would produce on the

49/ Question: “For each of the three mentioned risks, indicate the probability and the impact, considering the following classes: i) probability: low (<1%); medium-low (1% -10%); medium-high (10% -30%); high (>30%); ii) impact (volume of SFN assets): very low (<0.1%); low (0.1% -1%); medium (1% -5%); high (5% -10%); very high (> 10%)”.

50/ The circle size represents the risk frequency. The x and y coordinates represent, respectively, the midpoint of the probability and impact.

Table 1.6.2.2 – FSS – Citation frequency of the most important riskProbability Impact

Aug 2018 Nov 2018 Feb 2019

Fiscal-political Risks 53 67 65 Mid-High High

Foreign Scenario 9 16 20 Mid-High High

Delinquency and Activity 18 9 7 Mid-Low Medium

Non-fiscal Political Risks 15 4 4 Mid-High High

RiskFrequency (%)

Feb 2019

Statistical annex



financial system in case of materialization. A possible explanation for this result is that the uncertainties associated with the electoral process were dissipated, leaving concerns about governability issues and the scope of the economic reforms to be adopted by the new government.

Financial institutions also consider fiscal-political risks (with 59% of citations) and risks from the foreign scenario (with 48% of citations) as the most difficult to mitigate without the assistance of measures of the BCB and/or the Federal Government.51

The most relevant channels of transmission of high impact events perceived by respondents remained the same as those reported in the survey of August 2018. The most relevant transmission channels were a sharp drop in asset prices; an increase in risk aversion and uncertainty; capital flight or currency depreciation; and credit downgrade (Table 1.6.2.3).52

51/ Question: “Which of the risks listed above does your institution consider to be more difficult to mitigate with the adoption of internal risk management strategies by financial institutions without the assistance of measures of the BCB and/or the Federal Government?” The response may involve more than one risk, so that the sum of frequencies may exceed 100%.

52/ Question: “In the case of the occurrence of the most relevant event of high impact, what is the probability that this shock will be transmitted by the following channels:” The reported numbers represent the median of the answers. The last column shows the distribution of responses from the last survey.

0

2

4

6

8

10

0 4 8 12 16 20 24

Impa

ct (l

oss/

syst

em a

sset

s)

Probability

Foreign Scenario

Feb/2019 Aug/2018

0

2

4

6

8

10

0 4 8 12 16 20 24

Impa

ct (l

oss/

syst

em a

sset

s)

Probability

Fiscal-political risks

Feb/2019 Aug/2018

Chart 1.6.2.1 – FSS – Cited risks: probability, impact and

(%)

(%)

Table 1.6.2.3 – FSS – Transmission channels of high impact eventsAug 2018(median)

Nov 2018(median)

Feb 2019(median)

Distribution (last survey)

3 3 3

3 3 3

4 4 4

4 4 4

3 3 3

4 4 4

4 4 4

Very low 1 2 3 4 5 6 Very highProbability

Widespread credit rating downgrade, including sovereign ratings

Increase in risk aversion and uncertainty, affecting consumption and investment decisions

Decline in depositors confidence, including flight-to-safety

Transmission channel

Contagion between markets and domestic institutions

Liquidity squeeze, including interbank markets and foreign credit

Sharp decline in domestic financial asset prices, including collateral prices

Capital flight or strong currency depreciation

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

Chart 1.6.2.1 – FSS – Cited risks: probability, impact and frequency

Statistical annex

Statistical annex




1.6.3 Financial and economic cycles

The perception of the economic activity recovery remains relatively stable. Most respondents (91%) believe that the economy is in the recovery phase (Chart 1.6.3.1).

Respondents perceive that the credit to GDP gap increased marginally, with most of them classifying it as low (74% of respondents, considering the three corresponding categories), but with an upward trend (56% of respondents).

According to the survey, the trend of the increasing willingness of financial institutions to take risk is growing. The participation of those who consider that the risk appetite is low, but with an upward trend, went from 53% in August 2018 to 65% in February 2019, with a significant decrease in the share of those who consider that the risk appetite is low and stable (from 36% to 19%). This optimism about risk-taking is consistent with the prospects of improvement in the domestic economic environment.

The households and firms leverage is still considered at high levels by financial institutions. The leverage level of households is deemed to be high by 72% of the respondents, compared to 76% in August 2018, with 52% of them pointing out that the cycle phase is of a downward trend.

Regarding firms, 72% of the financial institutions classify their leverage as high compared to 85% in August 2018, with 56% of them pointing to a downward trend.

In general, there is a high degree of agreement among financial institutions that access to funding and liquidity remain high, with 72% of answers concentrated in the high classes in February 2019, compared to 76% in the survey of August 2018. In the same period, indications that the cycle is in the bullish phase rose from 18% to 33%, suggesting some optimism regarding market liquidity.

Regarding asset prices and economic fundamentals, most respondents consider that the cycle is in its high phase (54% in February 2019 against 38% in August 2018), with an increase in the indications of an upward trend. Additionally, the percentage of citations of the two categories with an upward trend (“high with upward trend” and “low with upward trend”) rose from 40% in August 2018 to 50% in February 2019.

0

20

40

60

80

100

Expansion Boom(peak)

Contraction Recession Depression(valley)

Recovery

%Economic Cycle

Feb 2019 Aug 2018

0

20

40

60

80

100

High withupward trend

High andstable

High withdownward

trend

Low withdownward

trend

Low andstable

Low withupward trend

%Credit to GDP Gap

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Risk Appetite

Feb 2019 Aug 2018

Chart 1.6.3.1 – FSS – Economic and financial cycles


0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Asset prices with respect to the economy fundamentals

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Funding and liquidity

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Firms leverage

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Households leverage

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Asset prices with respect to the economy fundamentals

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Funding and liquidity

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Firms leverage

Feb 2019 Aug 2018

0

20

40

60

80

100


High andstable

High withdownward

trend

Low withdownward

trend

Low andstable


%Households leverage

Feb 2019 Aug 2018

Statistical annex



1.6.4 Expectations for the Countercyclical Capital Buffer

Following the recommendations of the Basel Committee for Banking Supervision (Basel III), the Financial Stability Committee (Comef) is the body responsible for defining and reporting the value of the Countercyclical Capital Buffer for Brazil (ACCPBrazil). In the February survey, most financial institutions expected (94% of answers) and recommended (93% of answers) keeping the value of ACCPBrazil at zero percent (Chart 1.6.4.1). The decision of the Comef meeting on March 7th, 2019, was to maintain the value at zero percent.

1.6.5 Resilience and confidence in the financial system stability

The perception of the resilience conditions of the SFN53 remains positive (Table 1.6.5.1). Results show a high level of agreement among institutions on the suitability and adequacy of the available instruments to deal with a severe financial crisis if materialized.

The financial stability confidence index54 has remained high and stable, confirming the high values trajectory initiated in 2016 (Chart 1.6.5.1), with no negative citations (classes: “no confidence” and “low confidence”) recorded in the last ten quarters.

Therefore, despite uncertainties regarding the country’s fiscal balance, in addition to increased risks associated with the foreign scenario, the institutions surveyed rely on the resilience and stability of the financial system.

1.6.6 Final remarks

While participants of the survey are more concerned about risks stemming from the foreign scenario and the ongoing domestic process related to economic reforms, their probabilities of materialization and impact slightly decreased. As for risks from the foreign scenario, there were concerns about a possible decrease in the level of

53/ Question: “How does your institution evaluate the responsiveness of the financial system to the event described in field 1.1? (Scale the degree of satisfaction from 1 to 6, 1 being very satisfactory and 6 being very unsatisfactory)”

54/ Question: “What is the degree of confidence in the stability of the SFN in the next three years?” The confidence index is calculated by weighing the responses according to the following weights (multiplied by 100): full confidence (1); high confidence (0.75); mid confidence (0.5); low confidence (0.25), and lack of confidence (0).

93%

2%2% 2% 2%

Suggestions for the countercyclical additional capital

Keep as is Increase 0.05 p.p. Increase 0.5 p.p. Increase 1 p.p. No answer

94%

2%2% 2%

Expectations for the countercyclical additional capital

Keep as is Increase 0.05 p.p. Increase 0.5 p.p. No answer

Chart 1.6.4.1 – FSS – Expectations for the countercyclical Chart 1.6.4.1 – FSS – Expectations for the countercyclical additional capital

Statistical annex



the world economic activity. Issues related to political tensions in Europe and the “trade war” between the United States and China are perceived as risk factors with significant impact on the Brazilian financial system and are considered difficult to mitigate without the assistance of measures of the BCB or the Federal Government.

The perception of the economic and financial cycle is more positive. The risk appetite increased, although financial institutions remain cautious about the outcome of the domestic economic measures to improve the activity level. The valuation of asset prices improved, with a considerable number of respondents believing in a rising trend. According to most respondents, the leverage of firms and households is still high, but with a downward trend, and depend on the level of economic activity and the scope of the reforms proposed by the new government.

Most respondents expect and recommend maintaining the value of the Countercyclical Capital Buffer at zero percent, which actually occurred, suggesting the alignment of expectations concerning the capital buffer required to ensure the stability of the Brazilian financial system.

Confidence in financial stability is high, following an upward trend since 2016. The ability of the financial system to respond to relevant events is considered satisfactory.

40

48

56

64

72

80

Feb2012

Feb2013

Feb2014

Feb2015

Feb2016

Feb2017

Feb2018

Feb2019

%

Chart 1.6.5.1 – FSS – Confidence in the stability of the Financial System

0

20

40

60

80

100

No confidence Lowconfidence

Mid confidence Highconfidence

Full confidence

%Relative Distribution of the Confidence

Perceptions

Feb 2019 Aug 2018

Table 1.6.5.1 – FSS – Financial system capacity of reacting to high impact events

Aug 2018(median)

Nov 2018(median)

Feb 2019(median)

Distribution (last survey)

2 2 2

2 2 2

2 2 2

2 2 2

2 2 2

Satisfactory 1 2 3 4 5 6 Unsatisfactory

Median of the distribution of reaction capacities

Instrumental availbability for risk prevention and mitigation by the BCB

Financial system resilience factors

Financial system capital adequacy

Financial system liquidity adequacy

Financial institutions monitoring and attention

Government and Regulatory Agencies monitoring and attention

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

Statistical annex

Statistical annex




1.7 Systemically important financial market infrastructures

In the second half of 2018, the systemically important financial market infrastructures (FMIs) observed a safe and efficient operation. In the Reserves Transfer System (STR), the sole systemically important funds transfer system, the aggregate balance of funds available for payments and interbank transfers – named intraday liquidity – remained above the effective needs of participating financial institutions, assuring the smooth functioning of settlement operations. During the semester, on average, the need for funds – effective liquidity needs – of the system was 1.9% of the available liquidity, with a peak of 12,6% in the period.

Federal public securities (TPF) held by financial institutions in their portfolios and reserve requirements held at BCB contribute to the system´s high liquidity level (Chart 1.7.1). Reserve requirements balances can be transferred to the reserves accounts and TPFs can be converted into central bank money by way of intraday repo operations, both with no intraday financial cost to the financial institutions. A high and stable level of intraday liquidity allows an uninterrupted flow of payments, removing incentives for liquidity retention and the risk of insufficient resources for the settlement of obligations throughout the day.

The BCB performs monthly backtesting analyses for securities clearing and settlement systems for transactions with securities, derivatives and foreign currency, in which there is an entity acting as central counterparty (CCP). The aim of these analyses is to establish (i) the adequacy of the amount of additional55 collaterals and safeguards to cover default of the two participants with the most critical financial exposure (credit risk) and (ii) the existence of liquid resources that guarantees the timely settlement of obligations assumed by the two participants with the highest financial obligations (liquidity risk) on each day of the period.

A participant’s Net Financial Risk (RFL) is a metric used to assess its credit risk. It consists of the comparison between the financial result arising from the simulation of the closing of the positions56 and the guarantees of a defaulting participant. Both systems

55/ Each system has a fund available for the CCP to deal with the credit risk that exceeds the value of the investors՚ guarantees.

56/ Calculated by the CCP based on the strategy for closure and on the actual variation of asset prices, assessed on the subsequent days.

0

200

400

600

800

1,000

1,200

1,400

1,600

Jul 2017 Sep Nov Jan2018

Mar May Jul2018

Sep Nov

R$ billion

Federal public securities Reserve requirements

Reserves in the beginning of day Liquidity effective need

Chart 1.7.1 – Liquidity effective need

Statistical annex



evaluated - BM&FBovespa FX (foreign exchange) and BM&FBovespa Clearinghouse, operated by B3 S.A. - presented satisfactory results during the second semester of 2018. In both systems, the sum of the RFL of the two participants with the largest financial exposures did not exceed the value of the assets that make up the additional safeguards of that system.

In the BM&FBovespa Clearinghouse, the RFL of the two participants with the largest exposures corresponded to 8.9% of the additional safeguards available on the day it reached its maximum value (Chart 1.7.2). In the foreign exchange clearinghouse, the RFL was null every day.

Still regarding the BM&FBovespa Clearinghouse, Table 1.7.1 shows that in the second half of 2018, two-day accumulated changes in main Primitive Risk Factors’ value57 remained within the limits established in their stress scenarios. The table shows the highest observed percentage in the period for the ratio between two-day accumulated return and the respective high or low scenario.

The accuracy estimate58 of the risk model used by BM&FBovespa Clearinghouse remained above 99%, considered as a reference value59 for central counterparties (Chart 1.7.3).

Clearinghouses and other clearing and settlement providers must maintain liquid resources in order to guarantee at least the timely settlement of the participant’s obligations with the highest debtor position60. In this respect, both BM&FBovespa-FX (foreign exchange) Clearinghouse61 and BM&FBovespa Clearinghouse complied with the rules and followed international recommendations. Besides, although not required by regulation and by international principles, BM&FBovespa-FX (foreign exchange) Clearinghouse

57/ The PRF associated with a derivative contract is the denomination given to the financial variables relevant to the contract’s price formation.

58/ The accuracy level is defined as the correctness ratio of a risk management model within a given period.

59/ According to the Principles for Financial Market Infrastructures (PFMI – CPSS/IOSCO 2012), a CCP should use a 99% confidence level of the estimated distribution for future exposure when calculating the guarantees required for a participant or a portfolio. The accuracy estimate presented in Chart 1.7.3 is calculated in aggregated form for all individual portfolios, therefore being indirectly related to the PFMI recommendation.

60/ Resolution n° 2.882, of 8/30/2001.61/ The international principles adopted by BCB (PFMI – CPSS/IOSCO 2012)

recommend that a CCP, such as the Foreign Exchange Chamber, keep liquid resources to ensure the settlement of the largest debtor position. In turn, CCPs that are considered systemically important in more than one jurisdiction or that have a complex risk profile should maintain sufficient liquid resources to ensure settlement of the two largest debtor positions.

Table 1.7.1 – BM&FBOVESPA ClearinghousePrimitive Risk Factors (PRF)

Discrimination Low1/ High1/

Ibovespa spot 25% 39%

USD spot 36% 28%

Fixed rate 42 7% 7%

Fixed rate 126 32% 14%

Fixed rate 252 41% 23%

Fxed rate 756 43% 23%

DDI4/ 180 16% 11%

DDI 360 22% 13%

DDI 1080 23% 16%

Sources: BM&FBOVESPA and BCB

1/ Second semester of 2018.2/ Foreign exchange coupon.

99.0

99.2

99.4

99.6

99.8

100.0

Jul2018

Aug Sep Oct Nov Dec

%

Chart 1.7.3 – BM&FBOVESPA ClearinghouseCredit risk*

* Accuracy estimates from the model for individual margin evaluation with a rolling window of the last three months (63 days).

0

80

160

240

320

400

Jul2018

Aug Sep Oct Nov Dec

R$ millions

Sources: BM&FBOVESPA and BCB

Chart 1.7.2 – BM&FBOVESPA ClearinghouseNet Financial Risk

Statistical annex

Statistical annex

Statistical annex





maintained sufficient liquid funds to guarantee the timely settlement of the two greatest debt positions, except for nine days in the period for the settlement in reais (Chart 1.7.4) and thirteen days for the settlement in dollars (Chart 1.7.5).

0

120

240

360

480

600

Jul2018

Ago Set Out Nov Dez

Default of one member Default of two members

Chart 1.7.4 – BM&FBovespa-FX ClearinghouseLiquidity shortage R$

R$ millions

0

60

120

180

240

300

Jul2018

Ago Set Out Nov Dez

Default of one member Default of two members

Chart 1.7.5 – BM&FBovespa-FX ClearinghouseLiquidity shortage US$

US$ millions

Statistical annex

Statistical annex




2Selected issues

2.1 Debentures

Since mid-2017, there has been substantial growth in the issuance of debentures, the determinants of which were addressed in the last two issues of the Financial Stability Report (REF) and the Banking Report (REB). The fall in the Selic rate – which, since March 2018, remains at the level of 6.5% p.a. – the creation of the Long-Term Rate (TLP) and the government review of the business model of the National Bank of Economic and Social Development (BNDES) were identified as the main drivers of this surge. In this issue, we will present an analysis of the issuers, the allocation of resources and buyers in the debentures market.

The participation of new issuers, i.e. companies that by 2017 had not accessed the market of debentures – corresponding to 43% (R$64.4 billion) of the accumulated emissions in 2018 – indicates better access of companies to the debenture market. Another highlight was the issuance of debentures with tax benefits, which reached 16% of the total accumulated in twelve months (R$24.1 billion, of which R$13.8 billion of new issuers – Graph 2.1.1).

According to the Brazilian Association of Financial and Capital Market Entities (Anbima), the share of the volume offered to the market in 2018 for refinancing (including roll-over of previous issues of debentures) and working capital formation decreased, while the share allocated to investments in infrastructure is 17%, which represents a higher level than in 2015 (Graph 2.1.2). The general picture of the allocation of funds is more and more similar to that observed prior to the recession that began in the second half 2014.

-

36

72

108

144

180

Dec2015

Mar2016

Jun Sep Dec Mar2017

Jun Sep Dec Mar2018

Jun Sep Dec

R$ biChart 2.1.1 – Debentures IssuanceCumulative value of rolling 12 monthsIssuers before and after Dec/2016

Issuers until Dec/16 - w/o tax benefits Issuers until Dec/16 - with tax benefitsIssuers after Dec/16 - with tax benefits Issuers after Dec/16 - w/o tax benefits

In order to highlight the role played by new issuers in the acceleration of emissions from mid-2017, it was decided to separate them into two groups: those that had issued debenturesuntil December 2016 and those that only did so after that date.

Sources: BCB, [B]3

37% 33% 33% 35%

16%16%

44%

27%

15%10%

9%

17%16%

28%

9%

9%

12% 9%3%

11%1%

2% 1%

0%

20%

40%

60%

80%

100%

2015 2016 2017 2018

Chart 2.1.2 – DebenturesDestination of funds at issue date

Refinancing Working capitalInfrastructure Debentures roll-over or buy-backsInvestment in shares Not available / Not informedInvestment in fixed asset

Source: Anbima (Capital Markets Bulletin)

Statistical annex

Statistical annex




Mutual funds and other investors62 jointly absorbed 50% of total debentures offered on the primary market in 2018.63 This share has been increasing in recent years and tends to continue upwards since debentures are a higher yield alternative in the context of a Selic rate reduction cycle (Chart 2.1.3).

There is significant growth in the outstanding stock held by mutual funds since mid-2017 (Chart 2.1.4). Standing at the same level as that of banks in December 2018, this increased mutual fund representation constitute an alternative source of funds outside the financial system and some degree of financial disintermediation (from a relative point of view) in the acceleration of the issuance of debentures in the recent past.

The expansion of the capital market allows companies to obtain financing without the need to financial intermediation by banks, directly accessing other agents of the market. Investment funds are now presenting themselves as relevant buyers in the debentures market. This extends the financing options of companies, allowing the reduction of interest expenses.

From the point of view of financial stability, this movement, on the one hand, reduces the risk in the banking system and shifts profitability from credit to tariffs and services. On the other hand, it increases the maturity transformation and consequently the need for elaborated controls and risk management outside the financial system.

2.2 Evolution of vehicle financing to households

After a contraction period started in 2012, the vehicle financing portfolio for individuals has been recovering since the end of 2017, with a 13.5% YoY growth at the end of 2018 (Chart 2.2.1). Such portfolio does not currently present the risks observed in the past, with growth and credit conditions less vigorous than those in 2010. However, it is necessary to constantly monitor the various credit facilities, in order to verify whether the pattern of loan lending remains reasonable and with adequate risk balance.

62/ Non-banking financial institutions, institutional investors, non-residents, public sector, natural and legal persons.

63/ The BCB considers the issue year to be the date of effective transfer of the debenture to the acquirer in the systems of [B]3, instead of the date of registration of the distribution in the CVM or the date of commencement of the transaction - criteria used by Anbima.

41

34

76

17

51

60

6

7

15

- 30 60 90 120 150 180

2016

2017

2018

R$ biChart 2.1.3 – Issuance of DebenturesBuyers at the primary market

Banks Funds Others

Source: [B]3

0

84

168

252

336

420

Dec2015

Mar2016

Jun Sep Dec Mar2017

Jun Sep Dec Mar2018

Jun Sep Dec

R$ biChart 2.1.4 – DebenturesAmount outstanding by nature of holder

Banks Funds Others

Source: [B]3

0

44

88

132

176

220

0

20

40

60

80

100

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Chart 2.2.1 – Vehicle financing to individualsby vehicle type

cars (new) cars (up to 3 yrs) cars (above 3 yrs)motorcycles heavy vehicles Total (L)

BRL bi

Source: BCB (SCR, SGS), SNG

BRL bi

Statistical annex

Statistical annex

Statistical annex





Currently, the largest financed amounts in the vehicle portfolio are related to new cars and to those older than three years64 (Chart 2.2.1). The recent loan granting followed the resumption of sales in this market, reaching in 2018 the largest figure since 201565.

In 2010, loans with loan-to-value (LTV) between 80% and 100% and maturities between 4 to 5 years (Chart 2.2.2 and 2.2.3) were increasing, as the BCB increased the risk weight for vehicle loans granting with long maturities and high LTVs.66 Despite the recent growth in the vehicle financing outstanding credit (which is associated with the gradual economic recovery in the country), current LTV and maturities standards in higher brackets are still far from those observed in 2010. In addition, even with the recovery of loan granting, the current level is reasonably lower than those when the aforementioned BCB regulation was put in place67 (Chart 2.2.4).

Regarding the portfolio risk level, a decrease in the problem assets ratio was observed since 2017, with the lowest figure occurring in December 2018 (Chart 2.2.5)68. The problem assets portfolio have declined gradually since 2012 and has been relatively stable since the last half of 2017 (Chart 2.2.6). Recently, the vehicle financing outstanding increase has caused the problem assets ratio to decrease. Given the current trend of stability of the problem assets portfolio, the outstanding credit increase is expected to continue driving the problem assets ratio of household vehicle financing to even lower levels throughout 2019, not necessarily meaning a risk reduction.

The main component of the problem assets portfolio is the 90 days past due loans (Chart 2.2.5). Currently, this component represents almost 74% of the total of problem assets of household vehicle financing. Analyzing the evolution of the 90 days past due loans rate, one can compare the current level of risk materialization with

64/ Data by vehicle type is only public since December of 2014.65/ BCB Time Series no. 7,384 - Sales of factory authorized vehicle outlets -

Passenger cars sales, whose primary source is the Associação Nacional dos Fabricantes de Veículos Automotores (Anfavea), shows that the monthly average car sales in 2018 was above 175 thousand vehicles, the largest since 2015.

66/ Circular BCB no. 3,515, of December 3rd of 2010.67/ BCB Time Series no. 3,996 was used prior to March of 2011 and BCB

Time Series no. 20,673 after that date. Even so, there is no significant discontinuity in the real granting loans seasonally adjusted.

68/ Data presented here may differ from those of other BCB publications, including the BCB Time Series.

0

10

20

30

40

50

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Chart 2.2.2 – Distribution of lending LTVVehicle financing to individuals

up to 60% 60% to 70% 70% to 80% 80% to 100%

%

0

10

20

30

40

50

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Chart 2.2.3 – Distribution of lending due ratesVehicle financing to individuals

up to 1 yr 1 to 2 yrs 2 to 3 yrs3 to 4 yrs 4 to 5 yrs greater than 5 yrs

%

0

4

8

12

16

20

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Chart 2.2.4 – Real and seasonally adjusted credit lendingVehicle financing to individuals BRL bi

Statistical annex

Statistical annex

Statistical annex





those observed in the last credit cycle69 (Chart 2.2.7).70 The conclusion is identical to the one made from the observation of problem assets portfolio: currently, 90 days past due loans, whose data collection begins in December 2008, is at a historical low. Moreover, new loans are being granted at a lower level of 90 days past due ratios than the observed in 2010 and 2011 and without significant growth in 90 days past due loans per cohort71 (Chart 2.2.8).

2.3 Housing credit – Modality profile

Real estate loan has grown significantly since 2008, which justifies a more detailed analysis about its progress over this period. Taking into account this type of lending is long-term oriented, it is necessary to evaluate its capacity to generate and maintain positive interest margins in different economic contexts.

Real estate loan went through a significant expansion between 2008 and 2015. After this period, it maintained a moderate pace, in line with nominal GDP growth (Chart 2.3.1). Currently, it represents approximately one third of household outstanding credit. In the recent economic downturn, credit risk and problem assets portfolio increased. However, since 2017, problem assets have been falling, while coverage index for LLP has been increasing. Foreclosed assets (BNDU), i.e. residential mortgages collaterals taken by banks, despite their upward trend, have reduced when compared to both real estate portfolio and regulatory capital of the banking system. It is also worth highlighting the reduction in LTV related to new mortgages. Finally, the housing portfolio capacity to generate positive interest margins remained sound, despite being pressured by a greater shortage of earmarked saving deposits over a specific

69/ The restructured loans are only available from 2012 on, so the problem assets do not have total comparability between the current numbers and those verified prior to 2012.

70/ The data presented here may differ from other BCB publications or BC Time Series.

71/ Delinquency cohort in this report is the percentage of the total credit granted for household vehicle financing in a particular date that was 90 days past due after 6 months. For delinquency cohort of household vehicle financing, see https://www.bcb.gov.br/estabilidadefinanceira/inadimplencia_coorte. A cohort consists of all credit sent to the Credit Information System (SCR) in a granular basis, with month and year in the “Loan granting date” field, according to the SCR document 3040 (www.bcb.gov.br/?doc3040), correspond to the month and year of the reference date (https://www.bcb.gov.br/content/estabilidadefinanceira/Documents/scr/inadimplencia_coorte/Notas_Metodologicas.pdf).

0

2

4

6

8

10

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.2.5 – Problem assetsVehicle financing to individuals

90 days past due loans90 days past due loans or restructured debt90 days past due loans or restructured debt or "E to H" rated loans

%

40

52

64

76

88

100

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.2.6 – Problem assetsVehicle financing to individuals (Dec/2012 = 100)


0.0

1.6

3.2

4.8

6.4

8.0

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Chart 2.2.7 – 90 days past due loansVehicle financing to individuals %

Statistical annex

Statistical annex

Statistical annex

https://www.bcb.gov.br/estabilidadefinanceira/inadimplencia_coorte

https://www.bcb.gov.br/estabilidadefinanceira/inadimplencia_coorte





period. Currently, the real estate interest margin benefits from the SELIC rate reduction.

Given that, housing credit presents a favorable outlook, not entailing risks to the financial stability.

During the economic downturn, even though the housing credit is a portfolio characterized by low risk level, the problem assets portfolio increased, especially as of September 2015. Although there has not occurred a significant increase in 90 days past due loans, an important increase in restructured loans arose between September 2015 and December 2016. The risk mitigation trend began in the second half of 2017, with the decrease of problem assets portfolio (Chart 2.3.2), primarily due to the decrease in restructured loans (either by the exit by curing72 of the previously restructured loans or by the reduction of new restructured loans) and, subsequently, by the 90 days past due loans decrease (Chart 2.3.3).73 Likewise, in an analysis of granting loans by cohort,74 there is also a reduction in the problem assets portfolio after 12-m of the loan lending date (Chart 2.3.4).

Within the period of portfolio risk increase, the coverage ratio of problem assets to provisions in the housing credit decreased. The indicator recorded the lowest level in July 2017 (40%). After this period, in line with the gradual improvement of the risk scenario from the second half of 2017, the coverage ratio reached 50% in December 2018, close to the level of the last quarter of 2015 (Chart 2.3.5).

The period of credit risk increase of the modality also resulted in an increase in properties recovery from 2015 onwards. Simultaneously, given the downward trend of economic environment, the real estate market did not provide a compatible pace of sale of these assets, which contributed to the increase in the Residential mortgages collaterals taken by banks (Chart 2.3.6). It should be noted, however, that the housing properties registered as BNDU do not provide a significant risk to financial stability, since in December 2018, the outstanding

72/ After 12-m of performing (without new restructuring), the loans are considered “cured”, that is, the restructuring would have been successful and the operations would no longer be considered as problem assets.

73/ Data presented here may differ from those of other BCB publications, including the BCB Time Series.

74/ See footnote 71, of the selected issues “Evolution of vehicle financing outstanding” for further explanations on the cohort analysis. In the case of housing credit, a window of 12-m was used: for a respective cohort, it was observed 12-m ahead the problem assets portfolio in relation to the total housing credit of that cohort.

0.0

0.8

1.6

2.4

3.2

4.0

Jun2009

Jun2010

Jun2011

Jun2012

Jun2013

Jun2014

Jun2015

Jun2016

Jun2017

Jun2018

Chart 2.2.8 – 6-m Cohort 90 days past due loansVehicle financing to individuals %

0

120

240

360

480

600

0

2

4

6

8

10

Dec2008

Dec2010

Dec2012

Dec2014

Dec2016

Dec2018

Milh

ares

%

Chart 2.3.1 – Housing credit

housing credit/GDP housing credit outstanding (L)

BRL bi

96

108

120

132

144

156

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.3.2 – Problem assets Housing credit (Dec/2014 = 100)


Statistical annex

Statistical annex

Statistical annex





volume represented only 1.62% of the housing credit portfolio and 1.4% of the institutions regulatory capital.

Currently, there is an improvement trend in the quality of the real estate loan portfolio. This perspective is further reinforced by the analysis of the evolution of the granting profile of the last three years, with a significant increase in loans with LTV of up to 80% (Chart 2.3.7).

The real estate loans are long-term oriented, with higher average ticket and lower interest rates when compared to other household loans. Likewise, it is necessary to ensure its economic-financial sustainability and viability for the banking system. The aforementioned attributes of real estate loans are strongly related to its funding profile, which is composed mostly by earmarked resources. Saving accounts (Brazilian system of savings deposits and loans – SBPE - and SFH – Housing financial system - funding model) and Fund for period labored (FGTS)75, which usually present reduced costs76 and indexed rates compatible with its destined credit, account for about 90% of real estate funding.

Resolution CMN 4,676 of 2018, as of January 2019, allowed financial institutions to use different interest rate indexes rather than the “Reference Rate” (TR). However, it should be emphasized that the application of different interest rate indexes is at institution’s sole discretion, which should consider their credit policy, their expectations about price dynamics and the level of risk; including capital requirements to cover risks arising from adverse movements in interest rates. From January 2019 on, financial institutions are given more flexibility to manage its real estate portfolio, moreover, allowing them to use asset securitization structures and loans as collateral of new funding. However, it is expected that the TR indexed loans will continue to predominate, at least initially. Other floating rates will probably be used in shorter-term agreements with corporate and private clients.

In this context, the risk-adjusted net credit margin77 pattern indicates housing portfolio is viable in economic-financial terms over time. The risk-adjusted net credit margin remained positive over the last five years, including the recent economic recession period. Despite

75/ Resolution CMN 4,676 of 2018 determines that 65% of the saving resources must be directed to real estate loans. In the case of FGTS funding, Housing Financial System (SFH) sets out specific agreement conditions.

76/ Mainly due to the tax exemption of saving income and FGTS funding.77/ Net credit margin minus LLP expenses (risk adjusted net credit margin).

0%

1%

2%

4%

5%

6%

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.3.3 – Problem assetsHousing credit

90 days past due loans or restructured debt or "E to H" rated loans90 days past due loans or restructured debt90 days past due loans

0.0

0.6

1.2

1.8

2.4

3.0

Dec2014

Dec2015

Dec2016

Dec2017

Chart 2.3.4 – 12-m Cohort problem assetsHousing credit

0.3

0.4

0.5

0.6

0.7

0.8

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.3.5 – Coverage index of problem assetsHousing credit

Statistical annex

Statistical annex

Statistical annex





that, in September 2015, net interest margin reached its lowest level. Between 2015 and 2017, margins experienced significant volatility. Nevertheless, the funding profile and SFH requirements were a safeguard to real estate portfolio revenue and funding cost. It is also worth highlighting the small difference between net credit margin and risk-adjusted net credit margin, thus indicating the low LLP expenses and risk profile of the housing lending portfolio, reinforced by its collateral and real guarantees structures (Chart 2.3.8).

When Selic rate increases above 8.5% a year, TR also tends to increase, even to a lesser extent78, acting as a natural hedge since it is used for the interest rate index for both funding and housing loans contracts. However, when Selic rate falls below 8.5% a year, saving accounts assume a floating rate, based on Selic (70% of Selic). Thus, while funding costs tend to follow the Selic rate fluctuation path, the assets side continue to generate incomes based on its pre-fixed interest rates, with positive effects on net interest margin.

In order to have a better view of these effects, a theoretical portfolio with stable outstanding amount79 was simulated from December 2005 to December 2018. On top of this, changes in funding cost and interest rate parameters were applied. Further, effects in the funding profile were considered when Real Estate Credit Bills (LCI)80 constituted part of it. The funding profile available over time can be seen in Chart 2.3.9.

The simulation demonstrates the ability of risk-adjusted margin to remain positive during higher Selic cycle, even applying conservative parameters throughout the simulated time range, such as: i) TR + 7.5% a year as interest rate (unchanged) and ii) stop accrual81 and LLP expenses worst historical levels. On the other hand, interest margin rises in periods when Selic rate falls below 8.5% a year (Chart 2.3.10).

78/ TR rate is based on the Base Financial Rate (TBF) parameters. TBF represents the average interest rate paid by term deposits, limited to the thirty largest banks in the national financial system. However, a deduction rule is applied on TBF parameters. As bank term deposit rates are influenced by Selic, TR is also indirectly affected, but in a lesser extent due to the deduction rule.

79/ It was assumed interest and principal were paid and amortization payments equal new loans amounts.

80/ In periods when the available saving deposits did not cover the real estate loan portfolio of the banking system.

81/ Past due 60 days operations must cease interest accrual.

0.0%

0.4%

0.8%

1.2%

1.6%

2.0%

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.3.6 – Foreclosed AssetsHousing properties

in relation to the housing credit in relation to the institutions RP

0%

20%

40%

60%

80%

100%

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Chart 2.3.7 – LTV distribution of lending creditHousing credit

up to 50% 50% to 70% 70% to 80% greater than 80%

0

3

6

9

12

15

Dec2013

Jun2014

Dec Jun2015

Dec Jun2016

Dec Jun2017

Dec Jun2018

Dec

Referential Rate (TR) (accrual effect - 12 months)Interest Income (Housing) 1/Funding cost (Housing)Net interest margin (Housing) 1/Risk-adjusted net interest margin (Housing) 1/Selic Rate (accrual effect - 12 months)

Chart 2.3.8 – Net interest margin – Housing lending portfolio Margin components trailing twelve months

%

1\ Note: Comprises equalization incomes from the Treasury to cover costs and to ensure the return onthe investment for MCMV contracts

.

Statistical annex

Statistical annex

Statistical annex





Nevertheless, the simulation reinforces the significance of funding sufficiency and the balance between loan interest income and funding cost to support positive margins of the housing portfolio. The choice of other types of funding and interest rate indexes must be made in appropriate manner in order not to put the sustainability of real estate lending at risk.

Finally, as reported in section 1.5.2, the specific sensitivity analysis assessed for housing portfolio indicates that only decreases of more than 40% in real estate prices would require additional capital.

2.4 Assessment of changes in coverage eligibility rules of the Credit Guarantee Fund (FGC)

Deposit insurance schemes are cornerstones of financial safety nets as they uphold public confidence in the banking system by protecting depositors from loss, up to a specified limit, in case of bank failure; therefore bank runs are averted and financial stability is preserved in the short run. In Brazil, in addition to demand and savings deposits, the Fundo Garantidor de Crédito (FGC) insures other types of financial institutions’ liabilities, thereby providing additional protection to investors.

However, in order to maintain financial stability in the long run, balance is needed between the development of financial safety nets and limiting moral hazard, since investors may increase their exposure to risk when insured. In fact, this kind of behavior was identified in a small group of investors by the BCB, which prompted the National Monetary Council (Conselho Monetário Nacional – CMN) to issue Resolution 4,620 (published on December 21, 2017). While keeping a BRL 250,000 coverage limit per financial institution or conglomerate (thus protecting low-ticket investors), Resolution 4,620 implemented a global coverage limit of BRL 1 million per investor for each four-year period, in order to curb excessive risk taking by a limited number of investors.

In light of the lower flow of new investments82 in a sample83 of financial institutions over 2018 (Chart 2.4.1) and the lower growth rates of the outstanding amount of investments made by individuals with investment

82/ Comprising investments in time deposits, bills of exchange, agribusiness, real estate and mortgage credit bills.

83/ Sample of 55 banks and 39 retail financial companies for which FGC-guaranteed funding is relevant.

0

20

40

60

80

100

Dec2005

Dec2006

Dec2007

Dec2008

Dec2009

Dec2010

Dec2011

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Savings account (earmarked) Funding (FGTS)Savings account (non-earmarked) Funding LCI (used) 1/

Chart 2.3.9 – Funding profile of the real estate loan portfolio%

1/ Included as funding only when traditional resources were insufficient to cover the credit portfolio insome institutions.

0

3

6

9

12

15

Dec2006

Dec2007

Dec2008

Dec2009

Dec2010

Dec2011

Dec2012

Dec2013

Dec2014

Dec2015

Dec2016

Dec2017

Dec2018

Risk-adjusted net interest margin (housing)LLP exepenses 1/Funding costFunding cost (usage of LCI)Referential Rate (TR) (accrual effect - 12 months)Selic Rate (accrual effect - 12 months)

Chart 2.3.10 – Simulation of housing loan portfolioBehavior of interest income components

%

1\ Conservative parameters: TR + 7.5% a year as interest rate (unchanged) and stop accrual and LLP expenses worst historical levels. Further, effects in the funding profile were considered when LCIconstituted part of it.

2,3%

0,35%

1,8%

usage of LCI

1.9 1.9 2.0 1.5

5.8 7.7 8.6 11.4

7.6

11.5 14.8

18.6 12.8

9.6

10.3

12.3

0

10

20

30

40

50

2015 2016 2017 2018

R$ billionChart 2.4.1 - Investments in selected institutionsFlow by type of investor

Depositors – investments above R$ 250 K by FI

Depositors – investments below R$ 250 K

Depositors – investments up to R$ 250 K by FI and total between R$ 250 K and R$ 1 MM

Depositors – investments up to R$ 250 K by FI and total above R$ 1 MM

Sources: BCB, [B]3

Statistical annex

Statistical annex

Statistical annex





portfolios worth more than BRL 1 million in the same sample (Chart 2.4.2), Resolution 4,620 is effectively curbing moral hazard, allowing the FGC to concentrate on its mission of protecting low-ticket investors.

2.5 Portfolio flows for emerging economies and the non-residents behavior

In this section, we analyze how United States political economy decisions84 resulted in monetary and financial global tightening and had negative impacts on the portfolio investment flow for emerging market economies and, particularly, for Brazil. The applied methodology estimated autoregressive models with structural identification of shocks, in order to identify variables cross-effects and construct a historical decomposition of US monetary policy, risk appetite for emerging economies assets index, portfolio investment flow for emerging economies and portfolio flows for Brazil.

Capital flows are an important feature in the analysis of the vulnerability of financial systems in emerging market economies. In long periods of excess liquidity in international markets, when global capital flows increase, portfolio flows, domestic assets’ prices and the volume of credit grant may react to more favorable financial conditions, stimulating firm and consumer leverage. Increase in asset prices and credit, in turn, may contribute to the growth of the economic activity and reinforce new capital inflows. A tightening in monetary and financial global conditions can modify this benign scenario in emerging market economies, with increase in asset prices volatility. Economies that are dependent on short-run flows and where credit expansion are not in line with fundamentals can endure stress periods and reversion of the credit cycle. Economic fundamentals, introduction of buffers during expansion and policy adjustments are determinant for the resilience of the financial system to the external shocks.

Non-resident capital flows may be for direct investment or for portfolio investment. The nature of these flows are different: direct investment flows are less volatile and associate to macroeconomic fundamentals while portfolio flows are more volatile, reacting to global and local short-run financial conditions. Non-resident portfolio flows are

84/ As detailed below, the monetary conditions index not only reflects central banks decisions on monetary policy, but also the recent trade policy constraints that affected the foreign exchange and interest rate markets.

2.4 4.2 5.7 6.66.5

11.015.2

21.48.7

13.9

19.7

26.1

18.9

19.2

18.1

20.3

0

20

40

60

80

2015 2016 2017 2018

R$ billionChart 2.4.2 – Investments in selected institutionsAmount outstanding by type of investor

Depositors – investments above R$ 250 K by FIDepositors – investments below R$ 250 KDepositors – investments up to R$ 250 K by FI and total between R$ 250 K and R$ 1 MMDepositors – investments up to R$ 250 K by FI and total above R$ 1 MM

Sources: BCB, [B]3Statistical annex



classified according to their portfolio allocations: debt securities or equities.

Except for the abrupt recovery of flows observed soon after the financial crisis of 2008-2009, probably induced by the expansionist reactions of central economies’ monetary policies, a historical decline trend between 2011 and middle-2016 is noted (Chart 2.5.1). A short cycle of flow growth recovery started in the middle of 2016, ending in the beginning of 2018.

However, non-resident capital flows for direct investment were less volatile than portfolio capital flows in the same period (Chart 2.5.2). Although the volume of direct investment capital also shows a decelerated pace since 2011, growth is still positive and significant in the whole period. Actually, the portfolio flows present the larger risk of reversal and therefore pose greater challenges to ensure financial stability in emerging countries in the face of adverse external shocks.

Brazil was one of the countries that received less capital inflows as a percentage of GDP, considering a selection of emerging market economies in the last four years (Chart 2.5.3). However, Brazil received consistent and significant amount of direct investment inflows (Chart 2.5.4).

The determinants of capital flows fluctuations are essential to define strategies for risk monitoring and managing vulnerabilities. Literature shows that the drivers of the dynamics of the capital flows to the emerging economies can be internal or external and that the relative importance of these factors depends on the type of the flow, on the global liquidity scenario and on the characteristics of each country85. Other recent studies point risk appetite for emerging economies assets as an important factor in the determination of capital flows, in line with the discontinuity observed in emerging markets inflows during the 2008 global financial crisis86.

The behavior of the risk appetite for emerging economies assets is a difficult to measure non-observable variable. In this respect, we produced our own index, using qualitative analysis of orthogonal principal components from a set

85/ IMF, 2017. “The Drivers of Capital Flows in Emerging Markets Post Global Financial Crisis”, IMF Board Paper, International Monetary Fund.

86/ Jerome H Powell, 2017. “Prospects for Emerging Market Economies in a Normalizing Global Economy”, Prospects for Emerging Market Economies in a Normalizing Global Economy.

-1.0%

0.0%

1.0%

2.0%

3.0%

4Q2008

3Q2009

2Q2010

1Q2011

4Q2011

3Q2012

2Q2013

1Q2014

4Q2014

3Q2015

2Q2016

1Q2017

4Q2017

3Q2018

% GDP

Chart 2.5.1 – Portfolio investment (12-month accumulated) for emerging economies

Equity Debt Portfolio, totalSource: IMF

-5.0%

-2.0%

1.0%

4.0%

7.0%

Sou

th A

frica

Col

ombi

a

Chi

le

Mex

ico

Per

u

Indo

nesi

a

Bul

garia

Turk

ey

Indi

a

Pol

and

Bra

zil

Phi

lippi

nes

Rus

sia

Thai

land

% GDP

2014 2015 2016 2017 2018*

Chart 2.5.3 – Portfolio flow (12-month accum.)

* 12-month accum. until 2018 Q3Source: IMF

1.5%

2.0%

2.5%

3.0%

3.5%

4Q2008

3Q2009

2Q2010

1Q2011

4Q2011

3Q2012

2Q2013

1Q2014

4Q2014

3Q2015

2Q2016

1Q2017

4Q2017

3Q2018

% of GDP

Chart 2.5.2 – Direct investment (12-month accumulated) for emerging economies

Source: IMF

Statistical annex

Statistical annex

Statistical annex





Global upswing

US elections

VixUS dollar aprecition vs

EMEsFed

dovish

-2.0

-1.5

-1.0

-0.5

0.0

Sep2016

Nov Jan2017

Mar Jun Aug Oct Jan2018

Mar May Aug Oct Dec Feb2019

Standard deviation of the

mean

Chart 2.5.5 – Risk appetite index-emerging economies

Expectations revision - Fed

of several daily market price indices of a selection of emerging economies87.

Risk appetite index88 (Chart 2.5.5) shows, in a consistent way, the markets’ reaction observed in the main events occurred in the last months: 1) US elections; 2) global upswing; 3) revision of the US monetary policy expectations; 4) VIX sudden high; 5) US dollar appreciation and idiosyncratic events in the emerging economies, and 6) changes in the Fomc (Federal Open Market Committee) communication towards a more dovish direction in 2019. The monthly variations in the BCB risk appetite index are highly correlated with the monthly portfolio flow for a selection of emerging countries89 (Chart 2.5.6).

The twelve-month accumulated portfolio flow for a selection of countries were negative of US$17.7 billions in January 2019, showing a flow reversal, in line with the recent behavior of the risk appetite index (Chart 2.5.7). In fact, the risk appetite index began a more persistent path of decline from February 2018 to December 2018, showing partial recovery only in January 2019.

Several global and idiosyncratic factors contributed to the reduction of risk appetite for emerging economies’ assets throughout 2018, including a contagion effect between emerging economies. Increased trade tensions, US monetary policy tightening, electoral cycles in Mexico, Turkey and Brazil and the exchange rate depreciation in Argentina are examples of factors that contributed for the worsening in the sentiment towards emerging markets. In time, US economic policy, including non-conventional monetary policies adopted in the post-crisis period, has been a major external element on capital flows90, 91.

In order to understand the effects of risk appetite and United States monetary conditions on the portfolio flow for the emerging economies and for Brazil along

87/ South Africa, Brazil, India, Indonesia, Mexico, Russia and Turkey.88/ A positive variation in the index means an increase in the investors’ risk

appetite for assets of emerging economies.89/ For this higher frequency analysis, we used data of portfolio flows from

Bloomberg database for a more restricted set of emerging economies that contain more up-to-date data. This allowed us to produce estimates until December 2018. We selected countries whose Bloomberg monthly data were consistent with quarterly data reported by the IMF. The countries selected are South Africa, Brazil, Bulgaria, Korea, India, Indonesia, Poland, Czech Republic and Turkey.

90/ Koepke, 2015, “What Drives Capital Flows to Emerging Markets? A Survey of the Empirical Literature”.

91/ João Barata R. B. Barroso, Luiz A. Pereira da Silva e Adriana Soares Sales, “Quantitative Easing and Related Capital Flows into Brazil: measuring its effects and transmission channels through a rigorous counterfactual evaluation”, BCB Working Paper Series no. 313.

0.0%

2.5%

5.0%

7.5%

10.0%

Sou

th A

frica

Col

ombi

a

Chi

le

Mex

ico

Per

u

Indo

nesi

a

Bul

garia

Turk

ey

Indi

a

Pol

and

Bra

zil

Phi

lippi

nes

Rus

sia

Thai

land

% GDP

2014 2015 2016 2017 2018*

Chart 2.5.4 – Direct foreign investment (12-month accumulated)

* 12-month accum. until 2018 Q3Source: IMF

Global upswing

US electionsVix US dollar aprecition vs

EMEsFed

dovish

-2.0

-1.5

-1.0

-0.5

0.0

Sept2016

Nov Jan2017

Mar Jun Aug Oct Jan2018

Mar May Aug Oct Dez Feb2019

Standard deviation of the mean

Chart 2.5.5 – Risk appetite index –Emerging economies

Expectations revision - Fed

-0.6

-0.3

0

0.3

0.6

-30

-15

0

15

30

Jan2011

Jul Jan2012

Jul Jan2013

Jul Jan2014

Jul Jan2015

Jul Jan2016

Jul Jan2017

Jul Jan2018

Jul Jan2019

s.d.US$ bi

Chart 2.5.6 – Portfolio flow and risk appetite index

Portfolio flow for EM (monthly)Risk appetite index, monthly var. (Right)

Source: IMF and BCB

Statistical annex

Statistical annex

Statistical annex





2018, we used an autoregressive vector (VAR) model, estimated with an indicator of US monetary conditions92, the emerging market risk appetite index developed by the BCB and the 12-month accumulated portfolio flow for emerging economies. Sample comprises monthly data between January 2007 and January 2019. The results are robust to other model specifications, either with first differences, or with longer differences and fewer lags.

Historical decomposition of the shocks93 shows the contribution of each variable to the risk appetite for emerging economies assets index (Chart 2.5.8) and for the portfolio capital movements in emerging economies (Chart 2.5.9), in the whole sample period and, especially, along 2018. The inflection in the movement of the risk appetite index that occurred in February 2018 is mainly explained by the changes in the US monetary conditions, which intensified at that time. By mid-2018, the deterioration of risk appetite was amplified by specific shocks in the sentiment index itself. The US monetary policy is again relevant to explain the oscillations in investor’s sentiment regarding emerging countries between October 2018 and January 2019, considering the 12-month accumulated variation.

After February 2018, as in risk appetite, it is evident the negative impact of the tightening of US monetary conditions on the capital flows for the emerging economies (Chart 2.5.9). However, especially after May 2018, pull idiosyncratic factors contribute significantly to deepen the deterioration of flows, suggesting a relevant role for other variables besides global monetary conditions and risk aversion that negatively pressured the portfolio flow along the year.

In a second VAR specification, we split monthly flow in two measures of flow: flow for emerging economies excluding Brazil and portfolio flow for Brazil. The

92/ For monetary conditions of the United States, we use an index based on the principal components approach extracted from daily data of interest rate market, stock market, inflation and currencies, the latter only considering developed countries.

93/ Structural identification of shocks in a VAR model allows calculating the effective contribution of each variable to the observed dynamics of capital flows in the portfolio. The identification strategy of this exercise is based on hypotheses for the contemporaneous correlation between the variables. By hypothesis, we assume that the US monetary policy does not react contemporaneously to the emerging capital flows or the variations in assets that compounds the risk appetite for emerging markets assets index. The risk appetite index, another hypothesis, does not respond contemporaneously to changes in capital flows, which are compiled and disclosed with a certain lag. In turn, the behavior of portfolio flows, a priori, can be affected contemporaneously by prices: either by changes in US monetary conditions or by changes in risk appetite for assets of emerging economies.

-2.0

-0.8

0.4

1.6

2.8-2.8

-1.6

-0.4

0.8

2.0

Jan2011

Jul Jan2012

Jul Jan2013

Jul Jan2014

Jul Jan2015

Jul Jan2016

Jul Jan2017

Jul Jan2018

Jul Jan2019

s.d.s.d

Chart 2.5.7 – Capital flow, risk appetite and US monetary conditions

Risk appetite (Right) Flow in 12-monthsUS monetary conditions (Right inv.)

Source: IMF and BCB

-1.8

-0.9

0.0

0.9

1.8

Jan2016

Apr Jul Oct Jan2017

Apr Jul Oct Jan2018

Apr Jul Oct Jan2019

Chart 2.5.8 – Historical decomposition –Risk appetite emerging econ.(12-month variation)

Baseline US monetary policyRisk appetite EM Pull factorSum of the schocks on flows to EM

Statistical annex

Statistical annex




historical decomposition in this specification helps understanding the effects of 2018 external conditions on the flows for Brazil (Chart 2.5.10). The message is similar to the one in the previous specification: capital flows to Brazil have been more sensitive to the US monetary conditions throughout 2018, with some contagion effect after May, led by the reduction of capital flows for the other emerging economies in the sample. After a history of negative contributions over 2016 and 2017, idiosyncratic pull factors do not seem to have contributed significantly to the dynamics of flows throughout 2018, although, in the period, the effect of the idiosyncratic pull factor was the opposite of the one observed in other emerging economies.

In short, the results of the exercises suggest that US monetary conditions and the contagion effect among emerging economies are important triggers for explaining portfolio capital flows. Indeed, the US monetary conditions tightening throughout 2018 was an important factor in reducing the risk appetite of non-resident investors for riskier assets and contributed to the retraction of capital flows to emerging economies. Brazilian economy idiosyncratic pull factors were non-dominant in explaining the recent behavior of portfolio flows in the country, showing the contribution of the external factors for the considerable reduction observed in 2018.

Considering the impacts for financial stability of a sudden reversion of portfolio flows, permanent tracking of macroeconomic risks and mapping of potential vulnerabilities of the national financial systems are valuable and a priority for national macro prudential authorities. In a prospective way, the more accommodative conditions in US monetary policy and the increase in risk appetite for assets of emerging economies in early 2019 suggest a recovery scenario for capital flows for emerging economies.

2.6 National Monetary Council introduces changes to the definition of regulatory capital

The National Monetary Council (CMN) has adjusted the definition of regulatory capital to ensure that the capital requirements entail adequate loss-absorbing capacity and, consequently, threats to financial stability are minimized.

-140.0

-70.0

0.0

70.0

140.0

Jan2016

Apr Jul Oct Jan2017

Apr Jul Oct Jan2018

Apr Jul Oct Jan2019

Chart 2.5.9 – Historical decomposition -capital flows EM (12-month variation)

Baseline US political economyRisk appetite - EM Pull factorSum of shocks on flows - EM

-75.0

-50.0

-25.0

0.0

25.0

Jan2016

Apr Jul Oct Jan2017

Apr Jul Oct Jan2018

Apr Jul Oct Jan2019

Chart 2.5.10 – Historical decomposition-capital flows Brazil (12-month variation)

Baseline US monetary policy Risk appetite - EMPull factor - EM Pull factor - Brazil Flow - Brazil

Statistical annex

Statistical annex




For this purpose, the changes addressed the eligibility of Tier II capital, the treatment of tax credits arising from transactions hedging investments abroad and the deduction of investments made in financial instruments issued by systematically important banks with the intention of absorbing losses in case of their resolution.

Resolution CMN 4,679, of July 31, 2018, aligns the eligibility of Tier II capital between banks controlled by the federal government and the other banks, enhancing the quality of regulatory capital. Regulatory capital (PR in the Portuguese Acronym) comprises Tier I and Tier II capital.

Tier II capital, recently adjusted by Resolution CMN 4,679, is composed of subordinated debt instruments, which can absorb losses in circumstances of peril to the continuity of the issuing entity. Thereafter, all the instruments aiming eligibility to Tier II capital must foresee, in the critical situations established by the regulation, the extinction of the debt or its conversion into shares of the issuing institution.

This measure, by eliminating a differentiation of prudential treatment based on the type of control, follows the principles established by the BC + Agenda, which stipulates that financial institutions should be segmented for prudential regulation in proportion to their size, risk profile and level of international activity. For a smooth implementation, a phase-out period starting in 2020 has been established. By the end of the phase-out period, the Tier II of federal banks will be composed exclusively of financial instruments that meet all the Basel III eligibility criteria.

Measures were also adopted to maintain the equilibrium of regulatory capital in face of the oscillation of tax credits, through Resolution CMN 4,680, of July 31, 2018. Such Resolution complements Draft Law (PL in the Portuguese Acronym) CMN 10,638, of 2018, which aims to improve the tax legislation governing investments abroad and the operations hedging them and gives a higher quality to any tax credits arising from such transactions. The aforementioned Project equalizes the tax treatment of investments abroad and the operations hedging them, eliminating the need for the value of the hedge to exceed that of the investment and the constitution of the resulting tax credit.

Resolution CMN 4,680, of 2018, covers the interval until the expected term of the aforementioned Draft Law,


postponing the deduction of tax credits arising from the operations hedging investments abroad from January 1, 2018 to December 31, 2019. Such tax credits will be deducted in full until the end of 2020.

Also with regard to deductions from regulatory capital, the CMN has determined that investments made in financial instruments issued by systematically important banks with the purpose of absorbing losses during their resolution must be deducted from Tier II capital, pursuant to Resolution CMN 4,703, of 19 of December, 2018. These instruments are associated with the internationally known Total Loss-absorbing Capacity (TLAC) requirement. This measure mitigates contagion risk, since the regulatory capital of the investing bank is preserved in the face of deterioration in the credit quality of the issuing institution.

These measures aim to strengthen the loss-absorbing capacity of banks and their ability to continue providing services without any discontinuity. Therefore, the quality of the capital of the institutions that constitute the financial system is a central aspect of its resilience.

2.7 The Central Bank of Brazil enhances its credit risk capital requirements

The Central Bank of Brazil (BCB) enhanced its credit risk capital requirements under the standardized approach, improving, in particular, the framework´s risk sensitivity. Such enhancements were promoted by means of amendments to Circular BCB 3,644, of March 4, 2013. In particular, the rules governing the capital requirements applied to loans to non-financial legal entities, commonly referred to as corporates, and exposures to foreign sovereigns, were amended.

Circular BCB 3,921, of December 5, 2018, revised the criteria exposures to non-financial legal entities must meet to be eligible for an 85% risk weight. This regulatory adjustment aims to differentiate higher credit quality corporates from riskier ones. Before this amendment, the corporate exposures eligible for an 85% risk weight were comprised of the transactions with non-financial entities whose total debt with the financial system as a whole surpassed BRL 100 million. With the new regulatory rule, which came into force in March, 2019, eligibility for the 85% risk weight is conditioned upon meeting more strict requirements, directly linked to the counterparty´s


credit quality. In particular, the amendment excludes from the preferential treatment defaulted exposures, a concept that incorporates qualitative and quantitative aspects and general criteria concerning obligations that already are or are likely to become past due. Furthermore, the new regulatory norm incorporates a new metric, registered in the Credit Information System (SCR), that analyzes the historical performance of obligations, including the number of days past due, if any, during the last six months.

The second adjustment to Circular BCB 3,644, of 2013, refers to exposures to foreign central governments and their central banks, commonly referred to as sovereigns. Given the recent internationalization of Brazilian banks, the regulatory rule that previously prevailed did not consider the varying degrees of risk posed by the various jurisdictions. In order to change that and more closely align Brazil´s prudential regulation to the Basel standards, the BCB increased the granularity of the risk weights applied to such exposures. This innovation in Brazil´s prudential regulation allows a more precise differentiation of the risk posed by sovereign entities, in accordance with the ratings attributed by the credit rating agencies.

The aforementioned amendments to the credit risk capital requirements under the standardized approach aim to enhance incentives and optimize capital allocation based on the level of risks incurred by the financial institutions. The BCB is convinced that these enhancements will contribute to ensure the safety and soundness of the Brazilian financial system and, concurrently, promote a more efficient allocation of credit in the economy.

2.8 Large exposures limit

Resolution CMN 4,677 of July 31, 2018, enhanced the Brazilian framework of maximum limits for client exposure and maximum limit for the amount of large exposures, in line with the most recent recommendations from the Basel Committee on Banking Supervision.

The new framework expands the list of financial operations that are subject to the limits, so that all exposures considered for capital requirement are also considered for exposure limits, including off-balance operations. The new rules also considers a more restrictive basis for the calculation of limits, using the Tier 1 instead of the total capital. Additionally, the criteria


for the measurement of exposures were also improved, incorporating the recognition of credit risk mitigators.

The individual limit for total exposure to a single client is set at 25% of Tier 1, while the amount of concentrated exposures is limited to 600% of Tier 1. Concentrated exposures are those that have a value equal to or higher than 10% of Tier 1.

The new ruling establishes the concept of connected counterparties which are considered as a single client for the purposes of exposure limitation. Connection of counterparties derives from a controlling relationship between them and, for counterparties with an exposure deemed significant, from the existence of an economic dependence relationship between them identified according to regulatory criteria. New ruling also establishes the reporting to the Central Bank of Brazil on compliance with exposure limitation.

Considering the segmentation of the universe of regulated institutions introduced by Resolution CMN 4,533 of January 30, 2017, less complex criteria apply to the calculation of exposure limits of institutions allocated to Segment 5 (S5), which have a simplified risk profile.

2.9 Introduction of the Net Stable Funding Ratio (NSFR) in Brazil

In the last quarter of 2018 took effect in Brazil the Net Stable Funding Ratio (NSFR)94, the second regulatory liquidity ratio introduced by the Basel III reforms in the afterwards of the last international financial crisis, that highlighted shortcomings in the prudential regulation framework that was in place.95

Acting in a complementary way, the two regulatory liquidity ratios, LCR and NSFR, aim to limit the excessive liquidity risk taking. While the LCR measures the short-term liquidity risk (30 days), the objective of the NSFR is limit the liquidity risk in a longer horizon (1 year), requiring that banks fund their activities with stable sources of funding, i.e., resources with low probability of withdrawal.

94/ As published in section 2.6 of the FSR of April, 2018, the NSFR was introduced in Brazil by means of Resolution CMN no. 4,616, of November 30th , 2017, with the methodology defined by the Circular no. 3,869, of December 19, 2017, and took effect in October 1st, 2018.

95/ The international standard for the NSFR calculation, and the full set of regulatory actions in the context of Basel III reforms may be accessed on: https://www.bis.org/bcbs/basel3.htm.

https://www.bis.org/bcbs/basel3.htm


In Brazil, the compliance with the NSFR is mandatory for institutions in the S1 Segment, according to art. 2o of Resolution no. 4,553, of January 30, 2017. This segment comprises systemically/domestically important financial institutions, due to their size or relevant international activity, and, therefore, are subject to the full compliance with the international standards stated by the Basel Committee on Banking Supervision (BCBS), of which the BCB is a member.

The Brazilian regulatory framework for the NSFR has already been assessed by the BCBS, graded with the best grade, Compliant, that means “totally in compliance with the standard”. The assessment is part of the Regulatory Consistency Assessment Program (RCAP), which verifies the degree of alignment of the prudential regulation of each jurisdiction, in relation to the standards agreed by the BCBS. The details of the Brazilian NSFR assessment are available on https://www.bis.org/bcbs/publ/d458.htm.

The NSFR is the ratio between the amount of Available Stable Funding (ASF) and the amount of Required Stable Funding (RSF). In order to reduce the risk of future liquidity crises, financial institutions must maintain a ratio above 100% on an ongoing basis.

In the fourth quarter of 2018, six banks96 were classified in the S1 Segment and had a weighted average NSFR of 121% in the period (Chart 2.9.1), metrics that remained relatively stable during the period, in line with the aggregate ILE for this group of banks.97 The dispersion indicates that individually all the six banks of the S1 complied with the regulatory minimum in the quarter, endorsing the analysis that, currently, the liquidity risk is not a major concern for those institutions.

In an international comparison, the Brazilian banking system has a long-term liquidity risk, in average, lower than the internationally active banks, including the systemically important banks (G-SIBs), that presented an aggregate NSFR of 116% and 117%, respectively, as of June, 2018.98

96/ In order of total assets: Banco do Brasil, Itau, Caixa, Bradesco, Santander and BTG Pactual.

97/ Although the ILE is based on the NSFR methodology, BCB has adapted some parameters and definitions in order to adjust to the monitoring data available.

98/ See chart no. 77 (page 85) and table C.74 (page 155) in the last Basel III Monitoring Report published by the BCBS, available at: https://www.bis.org/bcbs/publ/d461.pdf.

90%

100%

110%

120%

130%

140%

Oct Nov Dec2018

Chart 2.9.1 – NSFRHigh, low and aggregate1/

Aggregate NSFR ILE Maximum NSFR

Minimum NSFR

1/ NSFR and ILE series comprise FIs in the S1 Segment (nowadays: 6 banks).

Regulatory minimum

Statistical annex

https://www.bis.org/bcbs/publ/d458.htm

https://www.bis.org/bcbs/publ/d461.pdf

https://www.bis.org/bcbs/publ/d461.pdf



The sources of stable funding that comprise the ASF amount are those that are not expected to be redeemed in the next year, due to contractual impediments (for instance, perpetual capital instruments) or, even though redeemable, have low probability of withdrawal in that period (for instance, sight deposits from retail customers that have strong relationship with the institution).

The ASF amount is obtained after the application of ASF weighting factors (0% to 100%) over the carrying value of liabilities and capital items on the banks’ balance sheet, in which, the greater the degree of stability, the higher the ASF factor. The deposits arising from retail customers99 and the own capital constitute the main sources of stable funding for the Brazilian banks, representing 61% of the total ASF amount, as of December, 2018 (Chart 2.9.2).

It’s possible to notice by Chart 2.9.2 that the wholesale funding,100 despite significant in the balance sheet, are considered by the NSFR methodology as sources of non-stable funding in the long run, in the sense that those investors usually reacts in a faster way in a scenario of deterioration of a certain institution, redeeming their resources in relevant amounts. After applying the ASF factors, only 40% of this source of funding is considered to be stable over a year.

The RSF amount is mainly originated by the long-term and illiquid assets on the balance sheet of financial institutions, such as long-term loans and fixed assets. On the other hand, the stock of liquid assets and illiquid assets maturing in less than one year (for instance, payments of interest and principal of loans in the next twelve months) require a lower proportion of stable funding. The methodology also considers prudential to require that a fraction of certain off-balance sheet exposures, such as credit lines granted but not yet contracted, must be supported by stable funding.

The RSF amount is obtained through the application of RSF weighting factors (0% to 100%) over the carrying value of all assets and some off-balance sheet exposures, in which, the lower the liquidity, the higher the RSF factor. As of December, 2018, the NSFR considered that 69% of the aggregate credit portfolio of S1 banks

99/ It’s considered retail deposits those arising from natural person and non-financial small business.

100/ Wholesale funding are those arising from corporates, government and public sector entities, banks and other financial institutions entities, including investment funds, pension funds and insurance companies.

oct-18 oct-18 nov-18 nov-18 dec-18 dec-18

Accountingvalue

ASF amount(post-

factors)

Accountingvalue

ASF amount(post-

factors)

Accountingvalue

ASF amount(post-

factors)

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

Chart 2.9.2 – Composition of available stable funding (ASF)Aggregate of FIs in S1 Segment

Capital and capital instruments Wholesale fundingRetail deposits DerivativesOther liabilities

BRL billion

Statistical annex



must be financed by stable funding and represented, thus, a bit more than half of the total RSF amount of those banks (Chart 2.9.3). Comparatively, 36% of the securities on those banks’ balance sheet were considered long-term, representing a share of 13% of the total RSF amount, suggesting that the majority of these securities are liquid (e.g. sovereigns). In this category, most of the stable funding requirement arises from corporate bonds acquired by banks.

The aggregate NSFR result indicates a stable funding surplus101 of BRL 616 billion by the end of 2018. In the aggregate LCR, in turn, the high quality liquid assets are BRL 439 billion greater than the stressed net cash flows in 30 days. Together, both measures indicates that, maintaining the current risk profile, the institutions of S1 Segment have a space to expand their investments in less liquid assets, such as the credit for the real economy, without implying in a high vulnerability to liquidity crises.

101/ Difference between the ASF and RSF amounts, i.e., the stable funding excess that could be used to reduce the NSFR up to 100%, the regulatory minimum.

oct-18 oct-18 nov-18 nov-18 dec-18 dec-18

Accountingvalue

RSFamount

(post-factor)

Accountingvalue

RSFamount

(post-factor)

Accountingvalue

RSFamount

(post-factor)

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

Chart 2.9.3 – Composition of required stable funding (RSF)Aggregate of FIs in S1 Segment

Cash and central bank reserves Loans to FI and central banksHigh quality liquid assets (HQLA) Loans to customersSecurities non-HQLA DerivativesOther assets Off-balance sheet exposures

BRL billion

Statistical annex



Appendix

Central Bank of Brazil Management

Acronyms


Central Bank of Brazil Management

Board of Governors

Roberto de Oliveira Campos Neto Governor

Bruno Serra FernandesDeputy Governor

Carlos Viana de CarvalhoDeputy Governor

Carolina de Assis BarrosDeputy Governor

João Manoel Pinho de MelloDeputy Governor

Maurício Costa de MouraDeputy Governor

Otávio Ribeiro DamasoDeputy Governor

Paulo Sérgio Neves de SouzaDeputy Governor

Tiago Couto BerrielDeputy Governor

http://www.bcb.gov.br/pre/quemequem/port/carlosviana.asp?idpai=diretoria


Acronyms

ACCPBrasil Countercyclical Capital Buffer for BrazilAT1 Additional Tier 1 capitalASF Available Stable FundingBCB Central Bank of BrazilBCBS Basel Committee on Banking SupervisionBNDU Foreclosed assetsCET1 Common Equity Tier 1 ratioCCP central counterparty CI coverage indexCMN National Monetary CouncilComef Financial Stability CommitteeDTA Deferred Tax AssetsFGC Credit Guarantee FundFGTS Fund for period laboredFSR Financial Stability ReportFMIs financial market infrastructuresFSS Financial Stability SurveyFX Foreign ExchangeGDP Gross Domestic ProductIL Short-term Liquidity RatioILE Structural Liquidity RatioIPCA National household price indexIVG-R Residential Mortgage Collateral Value IndexLCI real estate credit billsLCR Liquidity Coverage RatioLLP loan-loss provisionsLR Leverage RatioLTV Loan-to-valueNFC non-financial corporationsNIM net interest marginNSFR Net Stable Funding RatioOTC Over the CounterPFMI Principles for Financial Market InfrastructuresPNADC Continuous National Household Sample SurveyRBAN Banking portfolioRCAP Regulatory Consistency Assessment ProgramRFL Net Financial RiskRoE Return on Equity


RSF Required Stable FundingRWA Risk-weighted assetsSBPE Brazilian system of savings deposits and loansSCR Credit Information SystemSelic Brazilian Benchmark Interest RateSGS Series Management SystemSFH Housing Financial SystemSFN National Financial SystemSME Small and Medium-sized enterprisesSTR Reserves Transfer System TBF Base Financial RateTC Total CapitalTH Reference RateTLAC Total Loss-absorbing CapacityTPF Federal public securitiesT2 Tier 2 capitalVAR autoregressive vector


Annex

Concepts and methodologies

a) Short-term Liquidity Ratio (IL) – conceptually similar to the Liquidity Coverage Ratio (LCR), it is the ratio between the stock of liquid assets held by the institution and the net stressed cash flows (estimated disbursements in the next 21 business days under a stress scenario). Therefore, institutions with IL above one (100%) have enough liquid assets to withstand this stress scenario.

i. Liquid assets – liquid resources available for each conglomerate/institution to honor its stressed cash flows for the next 21 business days. It is the sum of highly liquid assets, release of required reserves (due to deposits run-off) and supplemental resources.

a. Highly liquid assets - These include: i) unencumbered Brazilian sovereign bonds held by the institution or received as a collateral in reverse repurchase agreement operations (reverse repos); ii) stocks listed in Ibovespa index; iii) liquid quotas of investment funds; iv) cash; and (v) free central bank reserves.

b. Release of required reserves – amount of the required reserves that would be released to the institution due to the deposit run-off estimated in the stressed cash flows calculation.

c. Supplemental resources – other options for monetization in the scenario’s time-horizon, such as: Bank Deposit Certificate (CDB), Bank Deposit Receipt (RDB), Interbank Deposit (DI), long positions in box strategies (options), reverse repurchase agreements (reverse repos) backed by private securities.

ii. Stressed cash flows – an estimate of the amount of cash that the institution needs within the scenario’s timeframe (21 business days) under a stress scenario. The analyses take into account retail deposits run-off, wholesale funding run-off, market stress and net contractual cash flows.

a. Retail deposits run-off – estimate of the necessary amount to cover the retail-customers withdrawals in demand deposits, time deposits, savings accounts, box strategies, securities issued by the bank, and repurchase agreements (repos) backed by private securities.

b. Wholesale funding run-off – estimate of the necessary amount to cover the possibility of early redemption of the liability positions from the three largest market counterparties.


c. Market stress – estimate of the necessary amount to cover losses arising from market movements affecting the liquid assets or others positions that may cause a cash outflow of the institutions in the stress scenario. The losses comprise: i) margin calls; ii) pre-settlements of derivatives contracts; iii) losses on the marked-to-market values of the liquid assets.

d. Net contractual cash flow – payments due in derivatives positions and in contractual cash flows (assets and liabilities positions) with market agents, maturing within the horizon of the scenario.

b) Structural Liquidity Ratio (ILE) - it is the ratio between the available stable funding (part of the equity and liabilities on which the institution can rely for a one-year horizon) and the required stable funding (part of the assets, including off-balance-sheet assets, which must be financed by stable funding because they have long maturities and/or low liquidity). Institutions with ILE equal or above one (100%) are less susceptible to future liquidity problems. The calculation methodology is based on the final version of the Net Stable Funding Ratio (NSFR), which was introduced as a minimum mandatory compliance in October, 2018.

i. Available stable funding – the funding that shall remain in the institution for at least a year. The main sources of banks’ stable funding are the capital; non-redeemable liabilities with residual maturities above one year regardless of counterparty; and funding with no maturity or with a maturity of less than a year coming from retail customers.

ii. Required stable funding – the amount of stable funding needed to finance the long-term activities of financial institutions. Therefore, it takes into account the liquidity and the maturity of the assets of the institution. The long-term assets are mainly the credit portfolio maturing in over a year; nonperforming assets; less liquid or encumbered securities (i.e. margin requirement in clearings); fixed assets; and the items deducted from the regulatory capital.

c) Total Capital Ratio – Basel Committee on Banking Supervision international concept, consisting of the system regulatory capital (RC) divided by the system RWA. In Brazil, until September 2013, the minimum required ratio was the factor “F”, according to Resolution CMN 3,490, of 29 August 2007, and Circular BCB 3,360, of September 12, 2007. Until October 2013, financial institutions and other institutions authorized to operate should observe the 11% limit established by the BCB, except for individual credit unions not affiliated to central units. From October 2013 on, the minimum required ratio has been disciplined by the Resolution 4,193, of March 1, 2013, which defines a convergent calendar, requiring 11% of RWA from October 2013 to December 2015; 9.875% in 2016; 9.25% in 2017; 8.625% in 2018; and 8% from 2019 on. On top of this requirement must be added a capital buffer, as mentioned in the Common Equity Tier 1 (CET1) Ratio topic.

d) Tier 1 Capital Ratio – According the Resolution 4,193, of 2013, a Tier 1 Capital requirement became effective from October 2013 on, corresponding to 5.5% of RWA, from October 2013 to December 014, and 6% from January 2015 on. On top of this requirement must be added a capital buffer, as mentioned in the Common Equity Tier 1 (CET1) Ratio topic.

e) Common Equity Tier I Ratio (CET1) – According the Resolution 4,193, 2013, a CET1 capital requirement became effective from October 2013 on, corresponding to 4.5% of RWA. In addition to this requirement, the Resolution established a capital buffer, composed by the following items: conservation, countercyclical and systemic. The conservation buffer requirement corresponds to the following RWA percentages: zero, until December 31, 2015; 0.625%, from January to December 2016; 1.25%, from January to December 2017; 1.875%, from January to December 2018; and 2.5% from January 2019 on. The countercyclical buffer requirement is limited to the following maximum RWA percentage: zero, until December 31, 2015;


0.625%, from January to December 2016; 1.25%, from January to December 2017; 1.875%, from January to December 2018; and 2.5% from January 2019 on. The systemic buffer requirement is limited to the maximum RWA percentage: zero until December 31, 2016; 0.5%, from January to December 2017; 1.0%, from January to December 2018; and 2.0% from January 2019 on.

f) Leverage ratio – Basel Committee on Banking Supervision international concept, consisting of Tier I Capital to Total Exposure ratio. In Brazil, the BCB Circular 3.748, of February 27, 2015, established the leverage ratio (LR) methodology. This index intends to complement the current prudential requirements, through a simple, transparent and non-sensitive risk metric. The leverage ratio minimum requirement of 3.0% was established by the Resolution CMN no 4,615, of November 30th, 2017, which is effective from January 2018 on, applicable for institutions classified as S1 or S2, accordingly to the Resolution CMN no 4,553, of January 1st, 2017.

Concepts and methodologies – Capital stress

1.1 Stress test – Introduction

The stress tests executed in BCB comprise a macroeconomic stress test as well as sensitivity analysis to relevant risk factors. These exercises are simulations executed by the BCB in order to estimate potential losses and capital shortfalls in the banking system stemming from extreme adverse, but plausible, scenarios. It also provides assessment of the resiliency of either an individual institution or the banking system as a whole. Hence, it is possible to determine the impact on the capital of institutions taking into consideration unexpected, and thus, not provisioned losses caused by changes in macroeconomic variables.

For each stressed scenario new capital ratios (Basel Ratio, Tier 1 and CET1) are calculated. A financial institution is considered as non-compliant whether any of its capital ratios is below the minimum required and classified as insolvent in the case of total depletion of the CET1. The relevance of non-compliant and/or technically insolvent institutions is assessed and the additional capital required in order that no other bank could get non-compliant is calculated. The relevance of and individual entity is determined based on the representativeness of its Adjusted Assets with respect to the assets of the whole banking system.

The positive effects of the activation of the triggers related to Tier 2 and Additional Tier 1 capitals, in which values are converted into CET1 capital, are classified as income. Furthermore the requirement of additional capital buffers, according to the Resolution no. 4,193 with the redaction given by the Resolution no. 4,443 from Oct. 29th 2015, is taken into account in the calculation of capital shortfalls. And finally, the framework also considers the potential changes of registration and uses of deferred taxes and its implications on regulatory capital calculations, according to the Resolution no. 4,192, from Mar 1st 2013, and posterior modifications.

1.2 Macroeconomic Stress Test

The macroeconomic stress test framework is an exercise that consists of the application of adverse macroeconomic scenarios and the simulation of how the balance sheet of each financial institution individually would behave under such scenarios. With those information in hands, capital shortfall of the whole system is calculated.


1.2.1 Scenarios design

Three macroeconomic scenarios are designed, all of them with time horizon of twelve quarters, based on market information, having the following macroeconomic variables: 1) economic activity (Economic Activity Index measured by the BCB – IBC-Br); 2) exchange rate (Brazilian Real vs US Dollar parity); 3) Brazilian Benchmark Interest Rate (measured by the Selic rate); 4) inflation rate (measured by the National Index of Price to the Ample Consumer – IPCA – accumulated in twelve months); 5) Brazil´s country risk premium (EMBI+Br spread, calculated by J.P. Morgan Chase); 6) the 10-yr US Treasury Yield; 7) unemployment rate (calculated by the IBGE based upon the Brazilian National Household Sample Survey – PNADC); and 8) commodities index (CRB index, calculated by Thomson Reuters/CoreCommodity). All variables are measured as a 3-month average.

The baseline scenario is built using the median of the market expectations (Focus report) for the following variables: economic activity, interest rates, FX (foreign exchange) rates and inflation. The GDP – Focus expectation – and the IBC-Br (VAR variable) are perfectly correlated. The Brazil´s country risk premium, unemployment rate and commodity index are kept constant over the forecast horizon. On the other hand, the path of the 10-yr US Treasury Yield is defined according to the adverse scenario published by the Board of Governors of the Federal Reserve System in the report “2018 Supervisory Scenarios for Annual Stress Tests Required under the Dodd-Frank Act Stress Testing Rules and the Capital Plan Rule”.

The Structural Break scenario is obtained by verifying the historic periods in which each variable showed the greatest change (either positive or negative) through an eight-month interval. In each identified period, it is added the subsequent four quarters in order to form the total projection horizon (three years). Then, the changes between each quarter are calculated and applied onto the observed values of the variables in the reference date.

In the Worst Historical scenario, repetition of the macroeconomic variables behavior is simulated, through a six-quarters rolling window since July 2003. Each window is plugged into dynamic panel data models and the historical scenario is the one with the lowest earnings before taxes.

1.2.2 Stress simulation

The stress simulation is done by projecting six basic groups of the income statement, trying to represent the operational performance of banks presented in the last income statement (net non-operational income are not considered in the test):

1. Net interest income: comprises net credit income, accrued income from bonds and securities and funding costs;

2. Non-interest income: mark-to-market effects, hedges and exchange rates variations;

3. Fees & commissions;

4. Non-consolidated companies;

5. Administrative expenses and;

6. Provisions expenses.


In the “net interest income”, credit and bonds/securities income as well as funding costs are modeled based on the Selic rate. The total funding is adjusted according to their credit portfolio volume, in the proportion of 1:1. Provision expenses are estimated based on the problem assets evolution, resulting from the macroeconomic scenario.

The non-interest group is modeled by applying a shock on market risk sensible positions observed in the starting date of the test. The stressed market risk factors are obtained out of the macroeconomic scenario and positions are then recalculated. The result is the difference between the stressed and the initial values. This amount is applied on the first quarter of projection and incorporated into the final result.

The BCB changed the methodology used in order to capture the interest risk exposures. Hence, from the second semester of 2018 onwards this method will be different. Until recently the shocks were applied only on the trading book positions, all of them informed by banks, according to the Circular No. 3,354, from June, 25th of 2007. However this criteria is no longer in place and now the framework will encompass all the liquid positions, notably both government and corporate bonds as well as derivatives. The effect of this change is that the number of exposures subjected to these shocks have increased, which make the “non-interest” group more significant in the stress test.

The “Fees & Commissions”, “Non-consolidated companies” and “Administrative Expenses” groups are modeled by making use of dynamic panel data models, obtained with the same macroeconomic variables employed in the scenarios.

Besides the performance simulation, verified through the income statement, the Central Bank of Brazil has incorporated the inter-financial contagion into the macroeconomic stress test framework from the first semester of 2019 onwards. In each quarter of the stress test time horizon, there is a verification whether any institution falls below the minimum threshold of 4% of the Core Tier 1 capital ratio. If this is the case, the inter-financial contagion is estimated. The uncollateralized interbank exposures issued by that institution are assumed as losses in the creditors´ balance sheet, and then capital is recalculated. If any financial firm also falls below that threshold, the process is repeated iteratively until there is no more institution below the threshold. The stress test continues with new affected capital levels and the process is repeated in all quarters of the projection, until the end of the time horizon.

1.3 Sensitivity Analysis - Introduction

Sensitivity analysis complements the macroeconomic stress test framework. Its objective is to assess the individual effects of credit or market risk factors that might affect the regulatory capital of institutions, causing eventual capital shortfalls. Those analyses are conducted by applying incremental variations in such risk factors, keeping the other factors fixed.

1.3.1 Sensitivity Analysis – Changes in market risk factors

The exposures subjected to interest rate changes (e.g. fixed rates, currency coupons, price indexes and interest rates) listed in the trading book are stressed. The positions at all vertices (from 21 to 2,520 days) are recalculated after the application of shocks as well as the financial impact on banks’ capital positions. Stressed exposures also affect risk weighted assets (RWA) components. In the case of fixed rates, new regulatory parameters of capital requirements are recalculated based on every new yield curve generated by a shock.


Exposures in foreign currency, gold and other instruments subject to changes in the exchange rates are also stressed, and their impacts on capital and RWA estimated. Here we assume that all exposures are revalued following the percentage points projected for the stressed USD/BRL exchange rate.

We apply shocks individually in each factor, the interest rate and the exchange rate, starting at their current values, in steps of 10% in both directions, until it reaches 200% and 10% of its current value, on the upside and on the downside, respectively. After recalculating capital ratios, we evaluate both the regulatory capital adequacy ratios and the solvency of banks.

The calculation of interest rate shocks follows the same methodology as for the “non-interest” items of the macroeconomic stress test. For the other risk factors all the balance sheet positions are considered.

1.3.2 Sensitivity Analysis – Increases in problem assets

This analysis tries to measure the effect of problem assets increases over the regulatory capital of institutions. We increase problem assets up to 150% of its current level and compute the additional provision required. These additional provisions affect both banks’ capital positions and the RWA component of the required capital. After recalculating capital ratios, regulatory capital adequacy and the solvency of banks are evaluated.

1.3.3 Sensitivity Analysis – Fall in housing prices

The objective of this exercise is to estimate the impacts of fall in housing prices over the capital of financial institutions with outstanding mortgages. Prior to the simulations we proxy housing prices with the value of the updated collateral provided for the loan using the IVG-R index, adding the variations measured by the index since the date that the loan was generated until the date of simulation.

The analysis consists of reducing house prices, simulating a sequence of decreases in steps of 5 p.p. In each step collaterals that become lower than 90% of the remaining loan are considered delinquent.

The loss of each delinquent loan is equal to the difference between the outstanding balance and the present value of the amount recovered from the foreclosure process. In order to calculate the recovered amount, we calculate new housing prices after shocks, net of taxes, maintenance fees and costs related to the foreclosure process. In addition, we consider that the sale in the foreclosure process is done with a discount proportional to the reduction of price due to the shock. The present value is obtained by discounting that sale amount by the 1-year future rate negotiated in the BM&FBovespa. New regulatory capital ratios of each institution are calculated considering the estimated losses to the related decline in housing price.


Working papers about financial stability

470. Inflation Targeting and Financial Stability: does the quality of institutions matter?Dimas Mateus Fazio, Thiago Christiano Silva, Benjamin Miranda Tabak and Daniel Oliveira CajueiroJanuary 2018 Abstract Full text

475. Short-Term Drivers of Sovereign CDS SpreadsMarcelo Yoshio TakamiApril 2018 Abstract Full text

487. Dynamic Interbank Network Analysis Using Latent Space ModelsFernando Linardi, Cees Diks, Marco van der Leij and Iuri LazierNovember 2018 Abstract Full text

https://www.bcb.gov.br/pec/wps/port/wp478.asp?idiom=P

https://www.bcb.gov.br/pec/wps/ingl/wps478.pdf





Financial Stability Report - bcb.gov.br

Documents

Transcript of Financial Stability Report - bcb.gov.br