## INTRODUCTION (1canea2020001) ## Source details **Canonical URL:** [INTRODUCTION (1canea2020001)](https://www.imf.org/-/media/files/publications/cr/2020/english/1canea2020001.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2020/english/1canea2020001.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2020/english/1canea2020001.pdf.json) --- ### Financial system structure and performance - The financial system’s performance has been strong. - The banking sector: solid profitability and sizeable capital buffers. - The insurance sector: remained financially sound in the low interest rate environment. - Nonbank sectors: pension funds and mutual funds dominate institutional and retail asset management. - Major banks, life insurers and pension funds have expanded abroad. - Canada has strong financial linkages with the United States. - At end-2018: - Total assets of Canadian financial institutions: US$10.2 trillion or 626 percent of GDP. - Outstanding debt securities: US$2.2 trillion, or 133 percent of GDP. - Stock market capitalization: US$1.9 trillion, or 119 percent of GDP. - Composition of total assets (2017 shares): - Deposit-taking institutions: about 45 percent. - Pension funds: about 18 percent. - Mutual funds: about 17 percent. - Insurers: about 13 percent. - Growth since end-2013: - Total assets increased by 31 percent. - Banking sector claims on nonresidents rose to 41 percent of banking sector assets (from 31 percent). - Canada’s international portfolio investment assets rose to 95 percent of GDP (from 60 percent). - Concentration: - Six largest banks and Québec’s major credit cooperative group (D-SIFIs): about 90 percent of deposit-taking sector assets. - Three largest life insurers: about 70 percent of total net premiums. ### Macrofinancial vulnerabilities and channels of risk - Main macrofinancial vulnerabilities: elevated household indebtedness and housing market imbalances. - Drivers of household vulnerability: - Decades-long credit upcycle, low interest rates, low capital charges for mortgage lending, policies promoting housing ownership. - Downside risks to house prices described as sizeable given existing overvaluation. - Canada-specific housing finance characteristics amplify procyclical effects (refinancing pressures; potential lender adoption of risk-based mortgage pricing). - Severe downturn transmission: - Household sector: significant increase in debt belonging to financially weak households. - Corporate sector: would remain more robust. ### Stress testing and resilience findings - Deposit-taking institutions: - In a severe adverse scenario, major deposit-taking institutions would be able to rebuild capital positions to meet regulatory requirements. - High starting capital ratios, robust structural profitability, and dual layered mortgage market provide buffer. - Mortgage insurers: - Would probably need additional capital; face some capital shortfalls in stress scenarios. - Financial stability implications limited because of government backstopping of mortgage insurance contracts. - Liquidity and funding: - Major deposit-taking institutions hold sufficient liquidity buffers to withstand sizeable funding outflows. - Areas for closer monitoring: increased balance sheet complexity, reliance on wholesale and FX funding, extensive use of derivatives. - Life insurers: - Some exposure to financial market stress and lower interest rates. - Solvency of some major life insurers could come under pressure due to widening credit spreads and falling equity prices. - Life insurers hold sizeable amounts of low-rated and unrated bonds. - Solvency would be hit hard in a more sustained low interest rate environment. - Nonbank and market risks: - Increased investor risk-taking, stretched asset valuations, dependence on foreign creditors raise market volatility risk and potential sharp tightening in financial conditions. - Pension funds, institutional investors: increased use of private market strategies and derivatives—adding leverage and liquidity risks. - Fixed income-focused investment funds: rapid growth, increased exposure to credit and duration risk; redemption shock risk. - Dependence on foreign investors for corporate bond market funding has increased significantly; elevated foreign funding dependence raises tightening risk. ### Data, modelling and supervisory capacity gaps - Data and modelling limitations: - Lack of granular data impedes projection of key financial items by significant geographies and full modelling of risk components. - Limited historical downturn episodes constrain assessment of downturn or stressed parameters. - Data gaps for cross-sectoral exposures, unregulated nonbank intermediation, and funding market activities (e.g., securities lending). - Recommended capacity improvements: - Enhance top-down stress testing capacity for banks and insurers; prioritize Bank of Canada’s bank solvency stress testing framework. - Improve comprehensive quality assurance in supervisory or macroprudential stress testing exercises. ### Key policy recommendations (priority and timeframe as provided) - Raise required capital for mortgage exposures at both banks and mortgage insurers to fully account for through-the-cycle credit risk (OSFI, AMF) — Priority: H; Timeframe: NT - Enhance risk monitoring of banks’ funding and use of derivatives, risk-taking by nonbanks, housing finance-related vulnerabilities, and cross-border and intra-system interconnectedness; carry out Canada-wide surveillance in key sectors such as deposit-taking and insurance (BOC lead; HOA, SAC; OSFI, AMF) — Priority: H; Timeframe: NT - Strengthen oversight of large public pension funds, and increase transparency of their financial disclosures (DOF, provincial governments) — Priority: H; Timeframe: NT - Enhance scope and granularity of bank specific data reported regularly to capture data requirements for microprudential and system-wide/macroprudential stress testing (BOC, OSFI, AMF) — Priority: M; Timeframe: MT - Develop coherent framework (data sets, indicators, analytical models) for monitoring, assessment and mitigation of risks linked to household and corporate balance sheets tailored to domestic specificities (BOC lead) — Priority: M; Timeframe: MT - Further develop macroprudential stress testing capacity to fully account for market risk, counterparty risk, and derivatives-related risk (BOC lead) — Priority: M; Timeframe: MT - Closely monitor life insurers’ portfolios of low-rated and unrated debt securities (OSFI, AMF) — Priority: M; Timeframe: MT - Consider interactions of risk margins with regulatory solvency framework for life insurers in preparation for IFRS 17 (OSFI, AMF) — Priority: M; Timeframe: MT - Use proxies from international experience or innovative modelling to complement limited historical downturn episodes in prudential tasks (BOC, OSFI, AMF) — Priority: M; Timeframe: MT - Strengthen capacity to monitor complex nonbank financial sector; strengthen data collection for private fund structures, captive FIs and money lenders (BOC; CSA) — Priority: M; Timeframe: MT - Strengthen risk surveillance for investment funds; enhance bond funds’ liquidity risk management via enhanced disclosures and consider liquidity management tools such as redemption gates, fees and swing pricing (BOC; CSA) — Priority: M; Timeframe: MT *Italic: Content derived from "INTRODUCTION" (1canea2020001) of the Canada FSAP technical note.* ### Macrofinancial conditions and outlook - Real GDP growth projections: - Projected 1.5 percent in 2019 and 1.9 percent in 2020. - Medium-term outlook: growth expected to slow to around 1.6 percent by 2024. - Credit and housing indicators: - As of 2019Q1: credit growth moderated to 4.8 percent year-on-year. - Household debt: 96 percent of GDP at end-2018. - Net household wealth: 489 percent of GDP. - Mortgage credit at end-2018: Can$1.8 trillion, or 81 percent of GDP. - Mortgage credit provided by banks: 69 percent; households main borrowers: 81 percent. - Insured mortgages (2018Q3): Can$723 billion; government-guaranteed MBS (NHA MBS): Can$485 billion. - Growth-at-risk (2018Q3): - 5 percent probability that real GDP growth would be -1.7 percent or less over the next year. - 5 percent probability that real GDP growth would be -1.6 percent (annualized) over the next three years. - Selected economic indicators (2015–2020): - Real GDP: 2015: 0.7; 2016: 1.1; 2017: 3.0; 2018: 1.8; 2019: 1.5; 2020: 1.9 - Unemployment rate (average): 2015: 6.9; 2016: 7.0; 2017: 6.3; 2018: 5.8; 2019: 5.9; 2020: 6.0 - CPI inflation (average): 2015: 1.1; 2016: 1.4; 2017: 1.6; 2018: 2.2; 2019: 1.7; 2020: 1.9 - Household credit growth: 2015: 4.9; 2016: 5.5; 2017: 5.5; 2018: 4.0; 2019: 3.0; 2020: 7.2 - Business credit growth: 2015: 9.3; 2016: 5.3; 2017: 8.2; 2018: 6.5; 2019: 3.6; 2020: 4.2 ### FSAP analytical scope and stress-testing modules - FSAP covered: - Macrofinancial vulnerabilities (household and corporate sectors, housing market). - Bank solvency stress tests (top-down for seven largest deposit-taking institutions). - Bank liquidity stress tests (LCR tests and cash-flow analysis). - Life insurance solvency stress tests (five largest life insurers). - Mortgage insurance solvency stress tests (all three mortgage insurers). - Nonbank financial intermediation and market risk analysis (pension funds, investment funds, valuation assessments). - Investment fund liquidity stress tests (fixed income-focused mutual funds sample covering about 20 percent of mutual funds’ AUM and 77 percent of corporate bonds held by mutual funds). ### Macroeconomic scenarios and adverse scenario features - Two macrofinancial scenarios: baseline (October 2018 WEO) and adverse (simulated with Global Macrofinancial Model). - Baseline: economic growth about 1.9 percent (annualized) over 2019–21; interest rates rise with monetary tightening. - Adverse scenario: - Cumulative real GDP growth: -2 percent (annualized) during 2019–21 (equivalent to 3 standard deviations). - Likelihood (growth-at-risk): 3.8 percent. - House prices: expected average cumulative decline of 40 percent by end-2020 in the adverse scenario. - Transmission: trade disruptions, disorderly market adjustments, tightening global financial conditions, amplified by elevated household debt and housing imbalances; significant exchange rate depreciation; monetary policy snapback followed by accommodation. ### Household sector vulnerabilities and stress-test outcomes - Aggregate: - Household indebtedness peaked at 97 percent of GDP in 2016Q4; leveled off over two years. - Net household wealth: 489 percent of GDP as of 2018Q4. - Debt-at-risk: - Financially weak households typically defined DTI>4.5 or TDS>40 percent. - Share of debt held by households with DTI>4.5: more than 30 percent of total household debt at end-2016 (three times 1999 share). - Range of debt-at-risk at end-2016: between 17 and 42 percent depending on definition. - Stress assumptions aligned with adverse scenario: - House prices: decline by 40 percent over next two years (by end-2020). - Household disposable income: decline of 15 percent. - Interest rates: increase up to 230 basis points. - Interest rate adjustments by renewal profile (relative to 2016): 2017: 10 basis points; 2018: 70 basis points; 2019: 220 basis points; 2020: 230 basis points. - Stress outcomes: - Share of debt-at-risk (TDS basis) increases from 17 percent in 2016 to 29 percent in 2020 (adverse). - Share of debt-at-risk held by bottom 40 percent increases from 16.2 percent in 2016 to 31 percent in 2020 (adverse); baseline: 21 percent. - Share of debt-at-risk not covered by real estate assets rises from <0.5 percent in 2016 to about 4.5 percent by 2020 (adverse). ### Housing market valuations and HaR findings - Canada-wide house prices increased by more than 40 percent over the past five years (period covered). - House price-at-risk (HaR): - Three years-ahead HaR across Canada: about -12 percent (annualized) at end-2018 Q3 (5 percent VaR). - City tail risks: Toronto and Vancouver ~ -24 percent and -25 percent respectively (three-year horizon). - Supply-side and capital flows: - Residential investment mitigates forward-looking HaR. - FDI inflows: 1 percentage point increase in FDI inflows associated with reduction of about 0.5–1 percentage point in HaR for certain cities. - Other capital inflows (foreign bank transactions) amplify downside risks in cities such as Toronto and Vancouver. ### Corporate sector vulnerabilities and debt-at-risk - Corporate debt rose to 111 percent of GDP at end-2018. - Two-fifths of corporate borrowing is liabilities to affiliates. - Construction and real estate firms relatively leveraged; oil and gas and mining firms have weak earnings. - Earnings shock increases debt-at-risk by about 3.6 percentage points; interest rate shock increases debt-at-risk by 2.6 percentage points; together they generate an increase of 3.6 percentage points. - Severe shocks could double share of corporate debt-at-risk. ### Banking sector stress testing: methodology and key assumptions - Top-down solvency stress tests for seven largest deposit-taking institutions (cover >90 percent of assets); three-year horizon; cut-off 2018Q3 regulatory data. - Risks captured: credit, market, non-interest income, interest-rate risk. - Capital buffers and requirements: - Capital conservation buffer (CCB): 2.5 percent (fully phased in 2019). - D-SIB surcharge: 1 percent for all 7 banks. - Additional domestic stability buffer (DSB): 1.75 percent for six federally regulated D-SIBs. - Dividend policy: dividends anchored to last fiscal year; distribution restrictions applied when ending ratios fall below regulatory thresholds (detailed CET1 bands provided exactly in source). ### Bank stress test results and sensitivity - Aggregate CET1 under adverse scenario: - Decline by 4.8 percentage points to 7.4 percent in 2020; recover to 9.6 percent in 2021. - Baseline-adverse CET1 difference at end-horizon: 5.6 percentage points. - Income and losses (adverse scenario): - Cumulative pre-loss income 2019–20: 2.9 percent of RWAs. - Losses due to credit and market risks: 3.9 and 1.8 percent of RWAs, respectively. - Deteriorating credit quality impact on CET1: 1.2 percentage points via higher RWAs. - Credit loss cumulative default rates (adverse vs baseline, end of horizon): - Uninsured mortgages and HELOCs: 6.4 percent (adverse) versus 0.7 percent (baseline). - Consumer retail exposures: 13.3 percent (adverse) versus 8.3 percent (baseline). - Corporate exposures: 8 percent (adverse) versus 3.8 percent (baseline). - Additional cumulative impairment flow under adverse scenario: approximately Can$90 billion (37 percent corporates; 34 percent consumer credit; 29 percent uninsured mortgages and HELOCs). - Net interest income (NII): - Average yearly NII ~24 percent lower compared with pre-cut-off and baseline in adverse scenario. - Aggregate NII lowest in Year 2: 34 percent below cut-off year level. - Sensitivity analyses: - Stressed undrawn exposures add Can$18.5 billion impairments → additional CET1 impact of 91 basis points by end of Year 3. - Alternative IFRS 9 amortization tenor assumptions increase Stage II mortgages and total impairments by Can$5.2 billion → additional CET1 decline of 26 basis points. ### Mortgage market and risk-based pricing - Risk-based pricing simulation (mortgage structural model): - DSR>0.4: additional spread 100 basis points. - DSR>0.3 band: additional spread 50 basis points. - Prime (DSR<0.15): spread tighter by 50 basis points. - PD PiTs for insured and uninsured segments increase by 40 to 50 percent under risk-based pricing; loss rates increase similarly. - Mortgage amortization vulnerability: - Concentration of remaining amortization tenors between 20 and 30 years noted; cap on spread over personal demand deposit rate of 330 basis points used in modelling. ### Bank liquidity analysis: LCR and cash-flow results - Aggregate LCR starting point: 133.3 percent. - Scenario 1 (Retail): aggregate average LCR 127.6 percent. - Scenario 2 (Wholesale): aggregate average LCR 104.4 percent. - Scenario 3 (Combined): aggregate LCR 90.3 percent. - Currency-specific LCR movements: - U.S. dollar LCR: from 121.5 percent to 118.7 percent (retail) and 98.3 percent (wholesale). - Canadian dollar LCR: from 116 percent to 109.8 percent (retail) and 90.7 percent (wholesale). - Cash-flow analysis (3-month horizon) results: - Aggregate shortfalls in severely adverse scenario (banks with shortfall): Can$91 billion. - Highest individual bank shortfall: Can$26 billion. - Shortfalls considered manageable given Bank of Canada liquidity facilities. ### Life insurance stress tests and results - Five largest life insurers covered (~90 percent of life insurers’ total assets; ~80 percent of net premiums). - Aggregate stress outcomes (adverse scenario, 2019Q3): - Aggregate core capital ratio decline by 34 percentage points to 61 percent by 2019Q3. - Aggregate total capital ratio decline by 32 percentage points to 104 percent by 2019Q3. - Some entities would fall below 50 percent core capital hurdle and 90 percent total capital hurdle. - By 2021Q4: - Aggregate core capital ratio: 90 percent. - Aggregate total capital ratio: 132 percent. - All insurers above regulatory minimums. - Asset composition (five largest life insurers): - Bonds: C$483bn - Equity: C$36bn - Mortgages: C$70bn - Property: C$36bn - Assets by region: Canada C$278bn (39); USA C$182bn (26); UK C$51bn (7); Europe excl UK C$93bn (13); Japan C$34bn (5); Other C$73bn (10) - Credit quality: - Two-thirds of bonds rated A or higher. - 16 percent of bonds unrated or below B. - Policy recommendation: - Monitor life insurers’ low-rated and unrated bond portfolios. - In preparation for IFRS 17, carefully consider how risk margins interact with regulatory solvency framework. ### Mortgage insurers: stress-test results and policy implications - Mortgage insurer stress test (adverse scenario): - Aggregate capital would turn negative. - Mortgage insurers would need additional capital of Can$15 billion to meet MICAT supervisory target (150 percent). - Cumulative insurance claims projected: Can$25 billion (vs Can$1 billion during 2015Q4–2018Q3). - PD assumed 2.8 percent per year average 2019–2021; LGD assumed 43 percent. - Aggregate losses: Can$19.5 billion; capital available falls from Can$18 billion (2018Q3) to negative Can$1 billion by 2021Q4. - Alternative assumptions (updated LTVs): required capital Can$8.5 billion higher; capital shortfall implied ~Can$23 billion under conservative updating. - System-wide capital buffer assessment: - Seven D-SIFIs: capital buffers for insured mortgage exposures equivalent to 0.17 percent of outstanding insured mortgages (insured exposures Can$453 billion as of October 2018). - Accounting for mortgage insurers’ required capital: system-wide buffers = 1.96 percent. - In adverse scenario buffers should rise to 4.32 percent → additional capital need Can$28 billion. ### Nonbank financial intermediation and market risks - Size and growth: - NBFIs excluding insurers: almost two-thirds of Can$13.9 trillion in financial system assets (635 percent of GDP). - Mutual funds: Can$2.3 trillion (105 percent of GDP). - Pension funds: 113 percent of GDP (noted as 114 percent elsewhere). - CFIMLs’ total assets (2017): Can$3.3 trillion. - Collective investment schemes: 73 percent of nonbank intermediation; 18 percent related to credit provision dependent on short-term funding. - Institutional investor risk-taking: - Alternative/private market investments contributed 40 percent of pension fund asset increase since 2013. - Share of Level 3 assets in six large public pension funds rose from 31 to 36 percent. - Reported gross leverage for six large public pensions: about 130 percent. - Estimated effective leverage in publicly traded portfolios: closer to 150 percent. - Gross notional derivatives positions for six large public pensions: nearly 200 percent as of 2017. - Asset valuations: - Fixed income valuations stretched; term premiums at multi-decade lows. - Corporate bond duration-adjusted credit spreads near historic lows. - Canada has among the flattest yield curves. ### Investment fund liquidity stress tests: sample and results - Sample: 238 Canadian domestic-currency corporate bond-focused funds; Can$323 billion in assets (about 25 percent of Canadian dollar-denominated corporate bond market). - Redemption shock calibrations: 1. First-percentile historical fund flows by mandate → aggregate outflows 7.2 percent (sample-wide -7.1 percent). 2. First-percentile per individual fund → aggregate outflows 4.3 percent. 3. BOC calibration: 100 basis points parallel yield increase → larger redemption shock (BOC shock: 18.0 percent; median fund-level redemption: 17.5 percent). - Forced selling (pro rata, calibrations): - Calibration 1: Can$11.9 billion corporate bonds (≈2 percent of domestic-currency corporate bonds). - Calibration 2: Can$6.5 billion corporate bonds (≈1 percent). - Calibration 3: Can$29.7 billion corporate bonds (≈5 percent). - BOC price-impact estimate: - Can$31.7 billion corporate bond sales → 93 basis points increase in excess bond premium. - If long-term investors do not absorb supply → ≈140 basis points increase (GFC observed premium: 165 basis points). - Policy recommendations: - Strengthen risk surveillance and data granularity for investment funds and funds’ bond portfolios. - Enhance liquidity risk management capabilities and disclosures. - Consider liquidity management tools: redemption gates, fees, swing pricing. ### Structural modelling and satellite models (methodology highlights) - Credit PDs: - Bayesian Model Averaging (BMA) used for PD PiT satellite models across corporate and consumer segments. - 16 asset exposure classes × 3 geographies × 2 regulatory approaches → 96 asset-class segments. - Mortgage structural model: - Uses borrower affordability (DSR), LTV, and liquid wealth; Monte Carlo over partitioned DSR-LTV space; produces PD, LGD, EL, and RWA density projections. - RWA density calibration example (selected): - Aggregate pre-stress / post-stress RWA density: 18.3 / 43.4 (percent). - Insured: 24.0 / 53.8; Non-insured: 14.0 / 35.8. - Aggregate three-year cumulative PD under adverse: 6.7 percent (insured 8.5 percent; uninsured 5.3 percent). - Accounting integration: - IFRS 9 treatment: stage transitions projected; excess accounting provisions added back to Tier-2 up to cap 0.6 percent of credit RWAs; regulatory expected losses in excess of accounting provisions deducted from CET1. - Liquidity haircuts, run-off rates and LCR stress-test templates detailed in appendices (explicit run-off rates and haircut tables used in cash-flow/LCR analysis). _Italic: Extracted from INTRODUCTION and associated sections of "1canea2020001" (Canada FSAP technical note). _ ### INTRODUCTION _________________________________________________________________________________ 9 ### INTRODUCTION ### Financial system structure and performance - The financial system’s performance has been strong. - The banking sector has enjoyed solid profitability and sizeable capital buffers. - The insurance sector has remained financially sound even in the low interest rate environment. - Other nonbank sectors have grown considerably, with pension funds and mutual funds dominating the institutional and retail asset management landscape. - Major banks, life insurers and pension funds have expanded their footprints abroad. - Canada has strong financial linkages with the United States. ### Macrofinancial vulnerabilities and channels of risk - Macrofinancial vulnerabilities—particularly elevated household indebtedness and housing market imbalances—remain substantial and pose financial stability concerns. - Drivers of household vulnerability include: decades-long credit upcycle, low interest rates, low capital charges for mortgage lending, and policies promoting housing ownership that fueled borrowing amid rapidly rising house prices. - Downside risks to house prices in the medium term are described as sizeable given existing overvaluation. - Canada-specific housing finance characteristics may amplify procyclical effects of falling house prices due to borrowers’ refinancing pressures and lenders’ sudden adoption of risk-based mortgage pricing. - During severe downturns: - The household sector would be affected, with a significant increase in debt belonging to financially weak households. - The corporate sector would remain more robust. ### Stress testing and resilience findings - Major deposit-taking institutions would be able to manage severe macrofinancial shocks: - In a severe adverse scenario, major deposit-taking institutions would be able to rebuild their capital positions to meet the regulatory requirements. - Relatively high starting point capital ratios, robust structural profitability, and the dual layered mortgage market structure provide a generous buffer for deposit-taking institutions. - Mortgage insurers would probably need additional capital: - Mortgage insurers would face some capital shortfalls in stress scenarios. - Financial stability implications are limited because of the government’s backstopping of mortgage insurance contracts. - Liquidity and funding: - Major deposit-taking institutions hold sufficient liquidity buffers to withstand sizeable funding outflows. - Areas of concern warranting closer monitoring: increased balance sheet complexity, reliance on wholesale and FX funding, and extensive use of derivatives. - Life insurers: - Large life insurers appear somewhat exposed to financial market stress and lower interest rates. - Solvency of some major life insurers could come under pressure during severe financial market stress, largely due to widening credit spreads and falling equity prices. - Life insurers hold a sizeable amount of low-rated and unrated bonds. - Life insurers’ solvency would be hit hard in a more sustained low interest rate environment. - Nonbank and market risks: - Increased investor risk-taking, stretched asset valuations, and dependence on foreign creditors raise the risk of market volatility and a sharp tightening in financial conditions. - Pension funds and other institutional investors: increased use of private market strategies and derivatives—adding leverage and liquidity risks. - Fixed income-focused investment funds: rapid growth and increased exposure to credit and duration risk, raising the risk of a redemption shock. - Fixed income and real estate asset valuations are described as stretched. - Dependence on foreign investors for corporate bond market funding has increased significantly. - In a market stress scenario, valuation and liquidity risks could magnify losses and amplify market volatility, with spillovers to banks and insurers via holdings of securities and derivatives. - Elevated dependence on foreign funding increases the risk that financial conditions would tighten sharply. ### Analytical and supervisory capacity gaps - Data and modelling limitations: - Lack of granular data impedes the ability to project key financial items by significant geographies and to fully model all risk components. - Limited historical episodes of economic downturns constrain assessment of downturn or stressed parameters. - Data gaps particularly noted for cross-sectoral exposures, unregulated nonbank financial intermediation, and funding market activities (e.g., securities lending). - Recommended capacity improvements: - Enhance top-down stress testing capacity for banks and insurers, prioritizing further development of the Bank of Canada’s bank solvency stress testing framework. - Improve comprehensive quality assurance in supervisory or macroprudential stress testing exercises. ### Key policy recommendations (Table 1: Recommendations to Bolster the Financial System’s Resilience) - Raise required capital for mortgage exposures at both banks and mortgage insurers to fully account for through-the-cycle credit risk (OSFI, AMF) — Priority: H; Timeframe: NT - Enhance risk monitoring of banks’ funding and use of derivatives, risk-taking by nonbanks, housing finance-related vulnerabilities, and cross-border and intra-system interconnectedness; carry out Canada-wide surveillance in key sectors such as deposit-taking and insurance (BOC lead; HOA, SAC; OSFI, AMF) — Priority: H; Timeframe: NT - Strengthen oversight of large public pension funds, and increase transparency of their financial disclosures (DOF, provincial governments) — Priority: H; Timeframe: NT - Enhance the scope and level of granularity of bank specific data that is reported on a regular basis to fully capture the data requirements for fully-fledged microprudential and system-wide/macroprudential stress testing (BOC, OSFI, AMF) — Priority: M; Timeframe: MT - Develop a coherent framework (including appropriate data sets, indicators and analytical models) for the monitoring, assessment and mitigation measures of risks linked to household and corporate balance sheets. The framework should be tailored to capture domestic market specificities (renewal and roll-over risks, affordability and indebtedness constraints, lack of historical data covering full economic cycles, risk-based pricing, and product related vulnerabilities) (BOC lead) — Priority: M; Timeframe: MT - Further develop macroprudential stress testing capacity to fully account for some risk types that are currently partially covered (market risk, counterparty risk, and derivatives-related risk) (BOC lead) — Priority: M; Timeframe: MT - Closely monitor life insurers’ portfolios of low-rated and unrated debt securities (OSFI, AMF) — Priority: M; Timeframe: MT - Carefully consider how risk margins interact with the regulatory solvency framework for life insurers in the context of preparation for the new accounting standards (IFRS 17) (OSFI, AMF) — Priority: M; Timeframe: MT - Use proxies from the international experience or innovative modelling to complement the limited historical episodes of economic downturns in assessing downturn or stressed parameters in their prudential or supervisory tasks (BOC, OSFI, AMF) — Priority: M; Timeframe: MT - Strengthen capacity to monitor risks in the complex and fast-evolving nonbank financial sector, including strengthening data collection for private fund structures, captive financial institutions and money lenders, and other nonbank financial institutions, to improve their ability to track their growth, financial system linkages, and risk profile (BOC; CSA) — Priority: M; Timeframe: MT - Strengthen risk surveillance for investment funds; enhance bond funds’ liquidity risk management capabilities by requiring enhanced disclosures to supervisors or investors on relevant policies and assessments; consider liquidity management tools such as redemption gates, fees and swing pricing (BOC; CSA) — Priority: M; Timeframe: MT _Italic: Content derived from "INTRODUCTION" (1canea2020001) of the Canada FSAP technical note._ ### INTRODUCTION ### INTRODUCTION ### A. Financial System Structure - Canada has one of the largest and most developed financial systems in the world. - As of end-2018: - Total assets of Canadian financial institutions: US$10.2 trillion or 626 percent of GDP. - Outstanding debt securities: US$2.2 trillion, or 133 percent of GDP. - Stock market capitalization: US$1.9 trillion, or 119 percent of GDP. - Composition of financial institutions’ total assets (2017 shares by type): - Deposit-taking institutions: about 45 percent. - Pension funds: about 18 percent. - Mutual funds: about 17 percent. - Insurers: about 13 percent. - Each segment—deposit-taking, insurance, pension, asset management, and capital markets—is among the largest in the world in nominal terms. - Growth and internationalization since end-2013: - Total assets of financial institutions increased by 31 percent. - Banking sector claims on nonresidents rose to 41 percent of banking sector assets (from 31 percent). - Canada’s international portfolio investment assets rose to 95 percent of GDP (from 60 percent). - Concentration: - The six largest banks and Québec’s major credit cooperative group (D-SIFIs) account for about 90 percent of deposit-taking sector assets. - The three largest life insurers account for about 70 percent of total net premiums. - Business lines and activity: - Major banks’ main business lines: retail and wholesale banking, wealth management, and capital markets. - Banks finance about two-thirds of private sector credit domestically; bond issuance and nonbanks are important alternative funding sources. - Public and housing finance: - Bond markets expanded by about 39 percent since end-2013. - Share of foreign-currency debt securities increased from 26 percent to 34 percent. - Government roles: mortgage insurance through CMHC, backstops private insurers’ mortgage insurance (subject to 10 percent deductibles of original mortgage principal), and timely payment guarantee for securitization of qualified insured mortgages. - As of 2018Q3: insured mortgages = Can$723 billion; government-guaranteed MBS (NHA MBS) = Can$485 billion. - Exposure to housing market: - Total residential and nonresidential mortgage credit at end-2018: Can$1.8 trillion, or 81 percent of GDP. - Mortgage credit provided by banks: 69 percent; households are main borrowers: 81 percent. - Financial institutions hold around Can$180 billion in NHA MBS. - Government’s central role in housing finance fortifies the financial-sovereign nexus. ### B. Macrofinancial Conditions - Economic performance and outlook: - Real GDP growth: projected 1.5 percent in 2019 and 1.9 percent in 2020. - Medium-term outlook: growth expected to slow to around 1.6 percent by 2024. - Past dynamics: slowdown in 2015 due to low oil prices; recovery during 2016–17; weak performance in recent quarters (first three quarters of 2018 performed well). - Structural constraints: low labor productivity growth, population aging, deteriorating international competitiveness. - Financial conditions and monetary policy: - Bank of Canada initiated a tightening cycle in mid-2017 with five rounds of rate hikes; more recently signaled accommodative stance. - Long-term bond yields declined after rising during the tightening phase. - Overall pricing of risk remains near historical lows despite recent market volatility. - Credit and housing market developments: - As of 2019Q1: credit growth moderated to 4.8 percent year-on-year. - Policy measures reduced insured mortgage lending and improved credit quality; banks’ new lending to highly indebted borrowers fell sharply. - House prices broadly stable over the past couple years; housing activity moderated (construction, inventory and sales, mortgage lending). - HELOCs have grown rapidly, some with interest-only payment features; HELOC growth concentrated in the amortizing mortgage component with total outstanding balance recently contracting (regulatory findings referenced). - Market-based systemic stress: - Market analysis of 19 large financial institutions as of December 2018 indicates the probability that several financial institutions experience distress simultaneously was near historical lows. - Systemic stress measure broadly stable; potential contagion effects appear to have risen over the past decade reflecting interconnectedness and/or growing common exposures to the housing market. - Macrofinancial vulnerabilities (still substantial despite recent declines): - High household indebtedness: - Household debt reached 96 percent of GDP at end-2018. - Households’ net wealth: 489 percent of GDP. - Pockets of financially weak households exist with excessive indebtedness or weak debt servicing capacity. - Persistent housing market imbalances: - Overvalued house prices relative to fundamentals; imbalances persist in major cities such as Toronto and Vancouver. - Growing corporate debt: - Corporate debt rose to 111 percent of GDP at end-2018. - Growth driven by debt issuance (including in foreign currency) and non-mortgage borrowing. - Weak earnings persist in oil and gas and mining sectors; rapid debt increase in real estate sector with weak income growth. - Growth-at-risk results (as of 2018Q3): - 5 percent probability that real GDP growth would be -1.7 percent or less over the next year. - 5 percent probability that real GDP growth would be -1.6 percent (annualized) over the next three years. - Downside risk to growth has declined over the past year due to some reductions in housing market imbalances and credit-to-GDP gap. ### C. Scope of the Financial Stability Analysis in the FSAP - FSAP approach: comprehensive assessment of financial stability risks including macrofinancial vulnerabilities, resilience of deposit-taking institutions and insurers, risk-taking in nonbank sectors and markets, intra-system and cross-border interconnectedness, systemic liquidity, and housing finance. - Modules and analyses undertaken: - Analysis of macrofinancial vulnerabilities: - Focus on corporate and household sectors and housing market. - Employed the debt-at-risk approach and house price-at-risk analysis. - Bank solvency stress tests: - Top-down balance sheet approach for seven largest deposit-taking institutions (covering more than 90 percent of deposit-taking institutions’ total assets) using regulatory and supervisory data. - Bank of Canada carried out a parallel top-down exercise using its own methodology. - All tests used same baseline (October 2018 World Economic Outlook projections) and the adverse scenario; sensitivity tests also performed. - Bank liquidity stress tests: - Top-down exercise for the seven largest deposit-taking institutions using regulatory data. - Included Liquidity Coverage Ratio (LCR) tests and cash-flow analysis, including foreign-currency funding outflows. - Life insurance solvency stress tests: - Top-down exercise for five largest life insurers (cover about 80 percent of life insurers’ total net premiums) using regulatory and supervisory data. - Assessed instantaneous impact of macrofinancial shocks consistent with the adverse scenario. - Mortgage insurance solvency stress tests: - Top-down exercise for all three mortgage insurers using regulatory data. - Conducted consistently with bank solvency stress tests; considered baseline and adverse scenarios to assess system-wide capital buffers for mortgage lending. - Analysis of risk-taking in nonbanks and financial markets: - Examined landscape of nonbank financial institutions and market participants. - Analyzed risk-taking among institutional investors—particularly pension funds—and assessed asset price valuations. - Investment fund liquidity stress tests: - Top-down exercise for fixed income-focused mutual funds accounting for about 20 of mutual funds’ assets under management (AUM) using publicly available data. - Focused on potential effects of large redemption shocks on the corporate bond market; sample captured 77 percent of corporate bonds held by mutual funds. *Source: INTRODUCTION, 1canea2020001* ### 14. The remainder of this technical note is structured as follows. Section II presents key ### 14. The remainder of this technical note is structured as follows. ### Structure of the Note - Section II presents key macrofinancial risks and the scenarios underpinning the financial stability analysis. - Section III discusses macrofinancial vulnerabilities—particularly, in the corporate and household sectors and related to the housing market. - Sections IV, V, VI, and VII cover: - bank solvency stress tests, - bank liquidity stress tests, - life insurance solvency stress tests, and - mortgage insurance solvency stress tests, respectively. - Section VIII analyzes risk-taking in nonbank financial institutions and financial market activities. - Section IX covers investment fund liquidity stress tests. - Each section concludes with relevant policy issues necessary to manage identified risks. ### Financial System Structure — Key findings - Canada’s financial system is well developed based on the combined metrics of depth, access and efficiency. - The financial system is large, with banks, investment funds and pension funds dominating the landscape. - The bond and stock markets are deep, serving as important venues for governments, financial institutions, and nonfinancial entities to raise funding. - Nonbank financial intermediation is relatively large, but mainly comprises collective investment schemes susceptible to run. - Canadian banks have significant exposures to the United States due to their overseas operations via subsidiaries and branches. - The money market continues to function well, though asset-backed commercial paper activity has not fully recovered to pre-2008 levels. ### Macrofinancial Developments — selected observations - The Canadian economy has been very resilient since the global financial crisis, including its adjustment to lower oil prices. - Resources-rich provinces were disproportionately affected; recovery partly benefited from internal labor mobility. - A tightening cycle of monetary policy started, with inflation within the target band. - Long-term bond yields had increased during monetary tightening but have declined more recently, along with greater market volatility. - Financial conditions remain loose due to still favorable risk pricing conditions. - Credit growth has moderated due to monetary tightening and prudential measures that curb mortgage lending, contributing to a smaller credit-to-GDP gap. - The slowdown in residential mortgage lending is largely led by the decline in insured mortgages. - Credit quality of mortgage lending has improved recently, with a lower share of highly indebted borrowers. - House prices became more stabilized in the past two years; immigration appears to be an important driver of rapidly rising housing prices in some regions. - Construction activity: - A construction boom is evident in British Columbia. - Alberta saw a boom-bust cycle driven by oil prices. ### External and Fiscal Vulnerabilities — highlights - The favorable macroeconomic outcome benefits from relatively strong public finance and an international net creditor status. - The current account deficit indicates reliance on foreign funding. - Provincial governments have substantial financial liabilities, more than the federal government. - Provincial governments’ financial strength varies, potentially complicating the ability to backstop financial institutions in their respective jurisdictions. ### FSAP Analytical Framework - The FSAP approach integrates: - analysis of macrofinancial vulnerabilities (household sector vulnerabilities, corporate sector vulnerabilities, housing market imbalances), - bank and insurance stress testing (bank solvency and liquidity, mortgage insurance, life insurance), - analysis of nonbank and market activities (interconnectedness, risk-taking in asset management, asset price valuation, nonbank credit intermediation), - investment fund stress tests, - intra-system linkages and cross-border linkages, - exposures to housing markets and the financial-sovereign nexus, - systemic liquidity and resilience of core funding markets, including external and FX funding and mortgage funding market structure. ### Selected financial system metrics and structure (from tables and figures) - Financial institutions (total assets, in billion Canadian dollars and percent of GDP): - Memo: Nominal GDP (2018): C$ 2,217 billion (US$ 1,711 billion) - GDP per capita (2018): US$ 46,243 - Population (2018): 37.0 million - Quota: SDR 11,023.9 million - Nonbank financial intermediation: - In 2017, CFIMLs’ total assets amounted to Can$3.3 trillion. - In 2017, 73 percent of nonbank financial intermediation was related to collective investment schemes with features that make them susceptible to runs, and 18 percent was related to credit provision that is dependent on short-term funding. - Banking sector overseas exposures (2018Q4): - The figures show sizable exposures to the U.S. (displayed as percent of total claims on an ultimate risk basis in the source figures). - Money market instruments (selected, 2007-18): - Series illustrated include bankers' acceptance, other commercial paper, asset-backed commercial paper, corporate commercial paper, repo trading (in trillions of Canadian dollar; right scale). ### Selected economic indicators (2015–2020, percentages or levels as shown) - Real GDP: - 2015: 0.7 - 2016: 1.1 - 2017: 3.0 - 2018: 1.8 - 2019: 1.5 - 2020: 1.9 - Unemployment rate (average, in percent): - 2015: 6.9 - 2016: 7.0 - 2017: 6.3 - 2018: 5.8 - 2019: 5.9 - 2020: 6.0 - CPI inflation (average): - 2015: 1.1 - 2016: 1.4 - 2017: 1.6 - 2018: 2.2 - 2019: 1.7 - 2020: 1.9 - Household credit growth (average, in percent): - 2015: 4.9 - 2016: 5.5 - 2017: 5.5 - 2018: 4.0 - 2019: 3.0 - 2020: 7.2 - Business credit growth (average, in percent): - 2015: 9.3 - 2016: 5.3 - 2017: 8.2 - 2018: 6.5 - 2019: 3.6 - 2020: 4.2 - Three-months treasury bill: - 2015: 0.5 - 2016: 0.5 - 2017: 0.7 - 2018: 1.4 - 2019: 1.7 - 2020: 1.9 - Ten-years government bond yield: - 2015: 1.5 - 2016: 1.3 - 2017: 1.8 - 2018: 2.3 - 2019: 2.3 - 2020: 2.5 - Current account balance (in percent of GDP): - 2015: -3.5 - 2016: -3.2 - 2017: -2.8 - 2018: -2.6 - 2019: -3.1 - 2020: -2.9 - General government fiscal indicators (in percent of GDP): - Revenue (2015–2020): 40.0, 40.1, 39.9, 40.1, 39.8, 39.9 - Expenditures (2015–2020): 40.0, 40.6, 40.3, 40.6, 40.7, (2019–2020 values shown in table) - Overall balance (2015–2020): -0.1, -0.4, -0.3, -0.4, -0.8, -0.8 - Gross debt (2015–2020): 91.3, 91.8, 90.1, 89.7, 87.5, 84.9 - Net debt 1/ (2015–2020): 28.5, 28.8, 27.6, 26.8, 26.7, 25.9 ### Macrofinancial Risks — FSAP Risk Assessment Matrix (RAM) — key risks, likelihood and expected impacts - Sharp tightening of global financial conditions, with bouts of large market volatility: - Likelihood: Low/Medium - Expected impact: Medium - Potential channels and effects: - Decompression of risk premiums prompting worldwide decline in asset prices. - Tightening of liquidity conditions and increase in funding costs given Canada’s strong financial linkages with the U.S. and the global financial system. - Moderating economic activity in Canada and the U.S., contributing to deterioration in banks’ asset quality. - Losses on investment portfolios, weakening solvency of some financial institutions, and triggering portfolio reallocations that affect asset prices. - Weaker-than-expected global growth (including a significant China slowdown and rising protectionism): - Likelihood: High - Expected impact: High - Potential channels and effects: - Impaired debt-servicing capacity of corporates and households, raising banks’ nonperforming assets. - Financial difficulty faced by firms with cross-border operations, with spillovers to the financial system. - Depressed commodity prices, leading to growing vulnerabilities in the commodity-related sector, reduction in investment, jobs and real income, especially in resources-rich provinces, and spillovers to regional housing markets that heighten financial stability concerns and weaken provincial public finance. - Sharp house price correction on the back of rising unemployment and higher funding costs: - Likelihood: Medium - Expected impact: High - Potential channels and effects: - Deteriorating asset quality, particularly related to mortgage lending and real estate financing. - Rippled effects in financial markets that fund mortgage lending. - Increase in the government’s contingent liabilities through claims on mortgage insurance. - Cyber-attacks on the interconnected financial system: - Likelihood: Medium - Expected impact: Medium - Potential channels and effects: - Successful cyber-attacks on a key financial institution could spread widely; the situation appears manageable given robust financial market infrastructures (FMIs). - RAM methodological note: - Relative likelihood is staff’s subjective assessment: “low” indicates a probability below 10 percent, “medium” between 10 and 30 percent, and “high” between 30 and 50 percent. - RAM reflects staff views as of the time of discussions with the authorities. Non-mutually exclusive risks may interact and materialize jointly. *Source: Haver Analytics; IMF staff estimates; IMF staff.* ### 16. The financial stability analysis was conducted broadly based on two macrofinancial ### 16. The financial stability analysis was conducted broadly based on two macrofinancial scenarios (Figure 5, Table 5) ### Macroeconomic Scenarios and Stress-Testing Framework - Two macrofinancial scenarios underpin the analysis: the baseline (based on the October 2018 World Economic Outlook projections) and an adverse scenario simulated using the Global Macrofinancial Model. - Baseline: envisages a continued, stable macrofinancial environment, with economic growth of about 1.9 percent (annualized) over the three-year stress testing horizon (i.e., 2019–21). Interest rates are expected to rise in line with monetary policy tightening. - Adverse scenario: designed around key macrofinancial risks from the RAM and simulated with the Global Macrofinancial Model (a structural macroeconometric model). The scenario is described as severe but plausible, with cumulative real GDP growth of -2 percent (annualized) during 2019–21, equivalent to 3 standard deviations; based on growth-at-risk analysis, the likelihood of such a severe growth outcome is 3.8 percent. ### Adverse Scenario: Transmission and Features - Initial trigger and transmission: - Disruptions in international trade and global production chains. - Disorderly financial market adjustments and heightened policy uncertainty. - Tightening of global financial conditions, global housing market and credit cycle downturns. - Amplifying domestic vulnerabilities: - Elevated household debt and housing market imbalances amplify negative external shocks, producing a sharp housing market correction and deterioration in bank asset quality. - Significant financial stress and large exchange rate depreciation in Canada. - Monetary policy dynamics: - Snapback of interest rates in Canada initially (monetary tightening to stabilize inflation expectations after rising inflation caused by global supply shocks and Canadian dollar depreciation). - Accommodative monetary policy in later years due to recession-induced significant deflationary effects. - Scenario outcomes and magnitudes: - Cumulative real GDP growth: -2 percent (annualized) during 2019–21. - Likelihood of such growth outcome (growth-at-risk): 3.8 percent. - House prices: expected average cumulative decline of 40 percent by end-2020 in the adverse scenario. ### Overview of Macrofinancial Vulnerabilities - Main concern: elevated household indebtedness and housing market imbalances. - Drivers of household debt accumulation: low interest rates, low capital charges for mortgage lending, policies promoting housing affordability, and risk mispricing. - FSAP vulnerability analysis methods: - Debt-at-risk approach to quantify debt owed by financially weak corporates and households. - Sensitivity tests on corporate and household balance sheets to macrofinancial shocks. - House price-at-risk analysis to gauge downside risk to house prices. ### Household Sector Vulnerabilities: Key Findings - Aggregate measures: - Household indebtedness peaked at 97 percent of GDP in 2016Q4 and has leveled off over the past two years. - Net household wealth: 489 percent of GDP as of 2018Q4. - Comparative position: - Canada’s household debt-to-GDP and aggregated debt servicing-to-income ratios exceeded most advanced economy peers; only Australia and the Netherlands have higher household debt-to-GDP ratios and aggregated debt servicing-to-income ratios. - Debt-at-risk magnitudes and definitions: - Financially weak households defined typically as DTI above 4.5 or TDS above 40 percent. - Share of debt held by households with DTI above 4.5: more than 30 percent of total household debt at end-2016 (three times larger than the share in 1999). - Range of debt-at-risk at end-2016: between 17 and 42 percent depending on definition (e.g., 17 percent when only borrowers with TDS above 40 percent; 42 percent when households with DTI above 350 percent were considered). - Distributional vulnerabilities: - Low-income borrowers (below the 20th percentile): total debt service as share of income about 55 percent in 2016, compared to 25 percent for the average borrower. - Debt accumulation among poorer borrowers rose from about 3 times annual income in 1999 to about 8 times in 2016. - Stress-test assumptions for households (aligned with adverse scenario): - House prices: decline by 40 percent over the next two years (by end-2020). - Household disposable income: decline of 15 percent. - Interest rates: increase up to 230 basis points. - Interest rate adjustments relative to 2016 level by renewal profile: in 2017 by 10 basis points, in 2018 by 70 basis points, in 2019 by 220 basis points, and in 2020 by 230 basis points. - Household stress outcomes: - Share of debt-at-risk (based on TDS) increases from 17 percent in 2016 to 29 percent in 2020 in the adverse scenario. - Share of debt-at-risk held by the least wealthy borrowers (bottom 40 percent) increases from 16.2 percent in 2016 to 31 percent in 2020 in the adverse scenario (compared with 21 percent in the baseline). - Regional differences: households in British Columbia and Ontario are more financially vulnerable than those in Quebec and Alberta. - Coverage by real estate collateral: - Share of debt-at-risk not covered by real estate assets rises from less than 0.5 percent in 2016 to about 4.5 percent by 2020 in the adverse scenario. - Magnitude of increase is largest in Ontario and British Columbia (share rises by more than eight times). - Robustness checks: total share of debt-at-risk not covered by real estate assets of about 2 to 7 percent when 25 percent and 40 percent average declines in house prices are considered, respectively. ### Housing Market Imbalances - Aggregate price movements: - Canada-wide house prices increased by more than 40 percent over the past five years (up to the period covered). - After a boom between 2011 and 2017, the housing market cooled in 2018 due to tighter monetary policy, stricter lending standards, and other housing market measures. - Recent activity slowdown: - Construction activity, sales, and mortgage lending slowed substantially over the past year. - Slowdown partially reflects macroprudential measures and other policies to curb speculative activity and improve affordability. - Regional heterogeneity in the adverse scenario (house price declines by 2020): - Ontario, British Columbia, Quebec, Alberta, and the rest of Canada: declines range between 35 and 45 percent (Ontario and British Columbia most vulnerable). ### Key Macrofinancial Variables: Selected Canada Figures (from Table 5) - Canada: Real GDP growth - 2017: 3.0 - 2018: 2.1 - 2019 (baseline): 2.0 - 2020 (baseline): 1.8 - 2021 (baseline): 1.8 - 2019 (adverse): -3.1 - 2020 (adverse): -4.2 - 2021 (adverse): 1.3 - Inflation rate (CPI) - 2017: 1.6 - 2018: 2.6 - 2019 (baseline): 2.2 - 2020 (baseline): 2.1 - 2021 (baseline): 2.1 - 2019 (adverse): 3.6 - 2020 (adverse): 3.2 - 2021 (adverse): 1.7 - Unemployment rate - 2017: 6.3 - 2018: 6.1 - 2019 (baseline): 6.2 - 2020 (baseline): 6.2 - 2021 (baseline): 6.3 - 2019 (adverse): 7.2 - 2020 (adverse): 10.1 - 2021 (adverse): 12.1 - Exchange rate (CAD per USD) - 2017: 1.30 - 2018: 1.29 - 2019 (baseline): 1.28 - 2020 (baseline): 1.25 - 2021 (baseline): 1.22 - 2019 (adverse): 1.52 - 2020 (adverse): 1.55 - 2021 (adverse): 1.47 - Equity price (2017=100) - 2017: 100 - 2018: 103 - 2019 (baseline): 107 - 2020 (baseline): 111 - 2021 (baseline): 115 - 2019 (adverse): 61 - 2020 (adverse): 70 - 2021 (adverse): 93 - House price (2017=100) - 2017: 100 - 2018: 104 - 2019 (baseline): 105 - 2020 (baseline): 107 - 2021 (baseline): 109 - 2019 (adverse): 82 - 2020 (adverse): 66 - 2021 (adverse): 62 - Bank of Canada policy rate - 2017: 0.7 - 2018: 1.4 - 2019 (baseline): 2.1 - 2020 (baseline): 2.5 - 2021 (baseline): 2.6 - 2019 (adverse): 3.3 - 2020 (adverse): 2.9 - 2021 (adverse): 1.2 - 10-year government bond yield - 2017: 1.9 - 2018: 2.4 - 2019 (baseline): 3.1 - 2020 (baseline): 3.4 - 2021 (baseline): 3.5 - 2019 (adverse): 3.7 - 2020 (adverse): 3.2 - 2021 (adverse): 2.6 - Interbank spread (in percentage points) - 2017: 0.5 - 2018: 0.5 - 2019 (baseline): 0.5 - 2020 (baseline): 0.5 - 2021 (baseline): 0.5 - 2019 (adverse): 1.8 - 2020 (adverse): 1.7 - 2021 (adverse): 1.2 - Business credit growth - 2017: 8.8 - 2018: 7.8 - 2019 (baseline): 6.3 - 2020 (baseline): 4.6 - 2021 (baseline): 2.6 - 2019 (adverse): 0.8 - 2020 (adverse): -1.4 - 2021 (adverse): 0.4 - Household credit growth - 2017: 5.3 - 2018: 3.4 - 2019 (baseline): 4.1 - 2020 (baseline): 4.1 - 2021 (baseline): 3.0 - 2019 (adverse): -4.1 - 2020 (adverse): -4.2 - 2021 (adverse): -2.5 *Italicized source: IMF staff estimates and analysis from the Canada FSAP chapter as provided in the source content.* ### 30. House price-to-income and house price-to-rent ratios in major cities (e.g., Toronto and ### 1canea2020001 - 30. House price-to-income and house price-to-rent ratios in major cities (e.g., Toronto and Vancouver) ### House price valuations and aggregate risk estimates - House price-to-income ratio for the Canada-wide housing market is about 1.3 times the long-term average. - House prices have largely stabilized over the past year, but imbalances persist in major metropolitan areas such as the Greater Toronto Area (GTA) and Greater Vancouver Area (GVA); GTA is the most expensive. - House price-at-risk (HaR) analysis: - Based on individual cities using quarterly data from 1980 to 2018. - Downside risks to house prices in Canada have substantially increased over the past decade, partially reflecting higher house price overvaluations. - Downside risks are defined as the 5 percent value-at-risk for the one-year-ahead horizon (and similar analysis conducted for the three-years ahead horizon). - At longer horizons (three years ahead), Canada-wide downside risks have substantially increased in the past two decades, surpassing levels seen around the GFC. - The average three years-ahead house price-at-risk across Canada stood at about -12 percent (annualized) at end-2018 Q3 — i.e., there is a 5 percent likelihood that Canada-wide real house prices will decline by 12 percentage points or more three years from that date. - Tail risks vary substantially across cities: Toronto and Vancouver tail risks were close to -24 percent and -25 percent, respectively, over the same period. ### Drivers and mitigating factors - Supply-side: - Supply-side drivers such as residential investment play an important mitigating role to forward-looking house price risks. - Valuation net of supply-side factors (e.g., residential investment) is a main driver that amplifies downside risks to Canada-wide house prices. - Capital flows: - Capital flows are associated with downside risks to key residential housing markets, but the net effect depends on the type of flow and varies across cities. - Foreign direct investment (FDI): - Typically long-term investment and generally associated with lower future downside risks to several Canadian housing markets. - A one percentage point increase in Canada’s FDI inflows is significantly associated with a reduction of about 0.5–1 percentage point in house price-at-risk of cities such as Calgary, Quebec City, Ottawa, and Toronto. - Other capital inflows (not FDI or portfolio flows), generally attributed to foreign bank transactions, are found to amplify downside risks to house prices in cities such as Toronto, Vancouver, and Calgary. - The largest effects are seen in Vancouver, where speculative capital inflows might have partially contributed to the recent housing boom. - Financial conditions: - Tighter financial conditions (encompassing monetary policy, macroprudential measures, and other factors) are negatively associated with downside risks to house prices — i.e., tighter financial conditions reduce downside risks, typically via lending standards and mortgage costs. - Although sensitivities to financial conditions are statistically significant for most cities, their economic magnitude is marginal. - The most sensitive cities to financial conditions are Toronto, Ottawa, and Calgary. ### City-level developments and heterogeneity - Temporal patterns: - Downside risks across main metropolitan areas in Canada show several cycles over the past four decades, suggesting susceptibility to booms and busts. - There is considerable dispersion in downside risks across cities, reflected by volatility of the lower bound (tenth percentile) of the cross-sectional distribution. - City-specific notes: - GTA: - House price-at-risk has substantially deteriorated in GTA over the past year, largely due to city-specific factors such as higher valuations and constrained supply. - GVA (Greater Vancouver Area): - Downside risks have steeply risen over the past five years, partially reflecting higher price valuations and the negative role of other capital inflows. - The negative contribution of other capital inflows to GVA downside risks seems to be ameliorating over the past two years, likely reflecting housing-related measures by local government. - Ottawa: - House price-at-risk appears to have improved in the past five years, with loose financial conditions and balanced valuations playing important roles. - Montreal: - FDI inflows and financial conditions seem to be important mitigating factors to downside risks. - Calgary: - Housing market risks are driven by excess supply and overvaluation, with downside risks recently improving. - Valuations: - Compared to long-term averages, housing market valuations in Canada have become more stretched over the past five years, especially in GVA and GTA. - Toronto is described as about one and a half times overvalued compared to its long-term average. ### Methodology and data notes - Data and sample: - City-level analysis uses quarterly data from 1980 to 2018. - Risk definition and sensitivity: - Downside risks are measured as the 5 percent value-at-risk for one-year ahead; similar three-year ahead HaR analysis was conducted. - As part of sensitivity analysis, the 10 percent value-at-risk is also depicted and tracks the 5 percent house price-at-risk. - Contribution decomposition: - Contributions to HaR include valuation (net of residential investment), capital flows (FDI and other capital flows), financial conditions index (FCI), and other factors (including oil prices and household debt). - City-level contributions use house price-to-income ratio and residential investment in each city/province. *Source: INTERNATIONAL MONETARY FUND (Canada chapter).* ### 1.5 percentage point (a proxy for potential funding strains). The earnings shock on its own would ### 1canea2020001 - 1.5 percentage point (a proxy for potential funding strains). The earnings shock on its own would ### Corporate debt-at-risk and shocks - The earnings shock on its own would generate an increase in debt-at-risk of about 3.6 percentage points. - The interest rate shock generates an increase of 2.6 percentage points. - Together, they generate an increase of 3.6 percentage points in debt-at-risk, the same size as from the earnings shock alone. - A standalone oil price shock would have a limited effect on energy sector debt-servicing capacity in the baseline analysis, but the analysis may underestimate potential vulnerabilities. - During the oil price shock of 2014–15, when oil prices fell 75 percent, aggregate trailing 4-quarters EBITDA of Canadian energy-sector firms in the Capital IQ sample reported a decline of about 45 percent peak-to-trough (over six quarters). - If energy firms faced a similar 45 percent decline in EBITDA, the share of sector debt-at-risk would rise from about 0.6 percent to 2.0 percent, reflecting that 89 percent of sector debt is owed by firms with interest coverage ratios above 3. - Historical observation: in 2014, 93 percent of energy sector debt was owed by firms with interest coverage ratios above 3, yet sector debt-at-risk rose from 1.6 percent in 2014 to 7.3 percent in 2016. ### Corporate sector structure and vulnerabilities - Two-fifths of corporate borrowing is in the form of liabilities to affiliates. - Construction and real estate firms are relatively leveraged. - Oil and gas, and mining firms have weak earnings, aggravating their debt burden. - Real estate-related firms account for the largest share of growth in firm debt. - Corporate debt-at-risk, though rising, remains small overall, with firms in the utilities and materials sectors among the most vulnerable. - A severe shock consistent with the FSAP adverse scenario would likely double the share of corporate debt-at-risk. ### Data limitations and measurement risks - Statistics Canada data shows private nonfinancial firms have inter-affiliate debt liabilities equivalent to about 50 percent of all other borrowings. - Potential measurement issues: - Inter-affiliate classification may capture arms-length financing (e.g., private equity and debt funds) or obligations issued outside the statistical perimeter, implying arms-length debt in national accounts may be understated. - Challenges exist in understanding the amount of foreign currency borrowing and the degree to which exposures are hedged via revenue streams or derivatives. - The firm-level dataset is based on globally consolidated operations of firms reporting to securities regulators, whereas Statistics Canada data may not capture foreign activities and debt of certain non-resident Canada-incorporated firms. - The degree of overlap between the two data sources is unknown, potentially understating corporate debt vulnerabilities. ### Banking sector: recent performance and resilience - The FSAP assessed banking sector resilience, including solvency stress testing for six domestic systemically important banks (D-SIBs) and Quebec’s D-SIFI, which account for more than 90 percent of deposit-taking institutions’ total assets. - Since end-2013, the combined size of Canadian banks’ balance sheets increased by 52 percent. - Total claims on nonresidents increased to 39 percent of banking sector assets (from 31 percent). - Internal income generation remains robust, aided by exceptional non-interest revenues and contained operational expenses. - Retained earnings could fully finance bank expansion plans while supporting significant capital buffers above minimum ratios. - Risks to future profitability: - Maintaining net interest margin could become more challenging. - Growth of non-domestic asset classes may be more expensive in funding and capital charges. - Domestic rising competition and market maturation effects, especially in the mortgage market, could create challenges. - Potential downside credit risks (credit losses or higher risk weights) could materially impact sector profitability. - Solid current capital ratios could mitigate adverse developments in an economic downturn. ### IMF top-down solvency stress tests: methodology and assumptions - Exercise scope and data: - Conducted using two macrofinancial scenarios (baseline and adverse). - Based on calendar 2018Q3 regulatory reports. - Three-year stress testing horizon. - Risks captured: credit risk (credit impairments, impact on RWAs, funding, interest rate risk), market risk (repricing and spread risks, equity and commodity risks), and non-interest income risks. - Balance sheet and growth assumptions: - Balance sheet allocation and funding composition remain constant as of the cut-off date. - Credit exposures projected to grow in line with scenario projections for retail and business credit growth. - Interest-sensitive liability segments proportionally adjusted to reflect asset changes. - Exchange rate adjustments incorporated. - Maturing capital instruments assumed not rolled over under the adverse scenario. - IFRS 9 treatment: - Accounting provisions projected using stressed stage transition probability matrices by asset segment. - Lifetime Expected Credit Losses projected for Stage 2 assets. - Excess credit accounting provisions over regulatory expected losses were added back to tier-2 capital up to the cap of 0.6 percent of credit RWAs. - Regulatory expected losses in excess of accounting provisions were deducted from common equity tier-1 (CET1) capital to reflect IRB shortfall or excess. - RWAs projection: - Accounted for balance sheet growth, defaulted credit exposures formation, and exchange rate changes. - Additional impairments considered under STA and AIRB. - For market risk, VaR RWA replaced with stressed VaR RWA charge in the adverse scenario. - Operational risk and residual RWA charges kept constant at cut-off date levels. - Income projection: - Total net income reflected projections for net interest income, non-interest income and expenses, trading income, revaluation effects (FVOCI and FVTPL), credit loss provisions, tax charges, and dividend distributions. - Capital requirements and buffers: - Minimum capital requirements based on actual supervisory requirements during 2019–2021. - Capital conservation buffer (CCB) of 2.5 percent (fully phased in 2019). - D-SIB surcharge of 1 percent for all 7 banks in scope. - An additional domestic stability buffer (DSB) of 1.75 percent is required for the six federally regulated D-SIBs. - Under the baseline, banks expected to maintain capital ratios above total target requirements; in the adverse scenario, preservation of the CCB and DSB are relaxed. - Dividend policy: - Dividends assumed constant, anchored to the last fiscal year; distributions restricted only if ending capital ratios fall below regulatory thresholds. - Dividend distribution restrictions by CET1 capital ratio: - 5.5–6.125 percent: Maximum of profits that cannot be distributed = 100 percent - >6.125–6.75 percent: 80 percent - >6.75–7.375 percent: 60 percent - >7.375–8.0 percent: 40 percent - >8.0 percent: 0 percent - Note: The 1 percent D-SIB surcharge extends the capital conservation buffer for capital conservation purposes. ### Credit risk profile and concentration - Credit exposures represent almost 51.6 percent of total assets. - By type of exposure: - Retail real estate-linked products account for about 40 percent of total exposures at default (EAD) and almost 50 percent of total drawn lines. - Corporate loans correspond to 32 percent of credit exposures. - Consumer credit (including credit cards) corresponds to 15.8 percent. - By geography: - 68.3 percent of total exposures are domestic. - 21 percent of total exposures are in the United States. - Regulatory approach: - Approximately 88 percent of exposures are reported under AIRB and 12 percent under STA. - RWAs composition: - Credit-related RWAs account for almost 83 percent of total RWAs (including securitization and counterparty credit risk charges). - Aggregate asset side composition highlights (market aggregates): - Cash and cash equivalents; Interbank deposits; Securities; Mortgage loans; Business and Government loans; Consumer loans; Reverse repos; Derivatives assets; Other non-interest bearing items. - Exposure at Default (EAD) breakdown by type (market aggregates, in percentage points): - Insured mortgages, Uninsured mortgages, HELOC, Consumer credit, Credit cards, Financial institutions, Corporates, Sovereign (detailed percentage-point breakdowns shown in source figures). - Geographical concentration (market aggregates, in percentage points): - Canada 68.3, United States 21.0, Rest of the World 10.6. _International Monetary Fund — Canada FSAP (excerpt)_. ### 57. A variety of modelling approaches were used to model credit parameters under the ### 1canea2020001 - 57. A variety of modelling approaches were used to model credit parameters under the ### Credit risk segmentation and modelling approach - Exposure segmentation scheme: - 16 asset exposure classes across 3 geographies (Canada, United States, Rest of the World) and 2 regulatory approaches (AIRB and STA), resulting in a total of 96 asset-class segments. - Retail real-estate related classes: Residential mortgages (insured), Residential mortgages (uninsured), HELOCs. - Consumer retail classes: Consumer credit, Credit cards. - Corporate exposure classes: Corporates broken down into 11 industrial asset classes (Financial institutions; Agriculture, logging, forestry, fishing, and trapping; Mining, quarrying, and oil wells; Manufacturing; Construction and real estate; Transportations, communications, and other utilities; Wholesale trade; Retail trade; Services; Non-residential mortgages; Other business loans). - Sovereign. - Credit parameter modelling: - Credit parameters collected for all asset types and segments; models estimated for scenario-dependent projection-of-credit parameters: probability of default (PD) point-in-time (PiT), and loss given default (LGD) PiT. - Given lack of historical LGD data from regulatory returns, simplified approach used for all segments except retail real-estate related segments: where no explicit PiT LGD model exists, regulatory downturn LGDs used as basis to provide adjusted estimates for both adverse and baseline scenarios. - PD projection methodology: - Satellite models used to project scenario-dependent forward PD paths. - Bayesian Model Averaging (BMA) was used as the core econometric approach for PD projection under both scenarios. - BMA framework applied to all 11 corporate segments and the consumer and credit card segments for Canada and the United States. - Corporate PD PiT satellite models used historical default rates by industry obtained from third party data providers (Moody’s CreditEdge used for Canadian and U.S. industrial segments; U.S. models used as proxy for ROW). - Credit card historical delinquency and loss statistics provided by the Canadian Bankers Association statistical data warehouse were used to calibrate regression models. - PD path for each scenario and asset segment/class generated from estimated econometric models. - Details of BMA PD PiT satellite models and estimation are in Appendix III. ### Accounting layer and IFRS 9 integration - PiT stage transition rates: - Collected from regulatory returns for all banks and all segments for the first three quarters of 2018. - Based on PiT PD paths, stressed transition rate matrices were produced for each year of the stress testing horizon to project exposure transitions across stages. - Additional calculation layer estimated lifetime losses for Stage II assets. - Accounting layer enables full projection of accounting impairments for each asset type/segment. - Scenario assumptions for provisioning projections: - Under the baseline scenario, macroeconomic variables assumed to remain constant after year 3 of the scenario. - Under the adverse scenario, macroeconomic variables assumed to gradually converge back to the baseline path within a period of six years. - Perfect scenario foresight was assumed to simplify provisioning projections. ### Structural model for mortgage and real-estate retail portfolios - Rationale: - Major loss events in residential housing loan portfolios are practically non-observable in Canadian historical data; currently observable default rates are particularly small, making conventional econometric satellite models inadequate to capture severe downturn dynamics. - Structural model can capture bimodal dynamics: small loss rates during rising housing markets and large NPL formations under severely adverse conditions. - Model features: - Uses borrower affordability metrics (i.e., DSR) and their scenario-driven evolution to infer default probabilities. - House price shocks used to estimate losses given default. - Scenario shocks linked to household microdata sources to simulate borrower behavioral elements under distress. - Joint distribution simulation of DSR, loan-to-value (LTV) ratio, and borrower liquid wealth to represent portfolio credit quality. - Scenario shocks on interest rates, unemployment, and house prices determine borrower impacts, default triggers, and loss linkage. - Detailed implementation described in Appendix IV. - Uses of the structural model in the FSAP: - Primary model for projecting PD PiTs and LGD PiTs for mortgage portfolios. - Disentangling two-layer architecture of banks and mortgage insurers; provided estimates for insured and uninsured mortgage portfolios. - Provided bank-specific starting points and projections after adjusting for bank-specific portfolio quality. - Guided assessment of adequacy of capital requirements for both market layers under current and post-stress conditions. - Enabled sensitivity / impact attribution analysis under different assumptions for interest rate, unemployment, and house price shocks. ### Market risk analysis - Covered valuation risks from changes in risk-free interest rates, credit spreads for interest-sensitive instruments, and equity and commodity risks. - Data collection: - Banks reported debt, equity, commodity and fund portfolios by accounting category and maturity bucket (where applicable). - Macroeconomic scenarios extrapolated to produce financial variable shocks, mainly interest rate spread shocks by security type. - Methodology for debt portfolios: - Interest rate and spread shocks applied using modified duration approach by debt portfolio type (CAN sovereign, other sovereign issuers, provincial and municipal governments, MBSs, corporate bonds and paper, ABSs and ABCPs). - Analysis captured re-pricing losses in FVOCI and FVTPL books due to shocks to sovereign yield curves and valuation impact from spread shocks across accounting books. - Spread projections on provincial government debt, corporate bonds, and ABS/MBS computed based on average yield per maturity tenor from Bloomberg and anchored to scenario macrofinancial conditions. - Accounting portfolio impacts: - Analysis covered amortized cost (AC), FVOCI, and FVPL portfolios. - Losses from AC and FVPL assumed to impact regulatory capital through net profits; FVOCI losses affected capital through other comprehensive income. - For interest-rate sensitive positions in the AC portfolio, only impact of credit spread changes was accounted for. - For conservatism, existing hedges assumed ineffective during the scenario horizon (with specified notable exceptions discussed in the source). - Securities portfolio composition and allocations: - Debt securities primarily held in FVOCI and dominated by sovereign debt. - Sovereign debt accounted for 57 percent of the FVOCI portfolio and 40 percent of the total securities portfolio. - Provincial government debt added an additional 12 percent in the total general government securities share. - Corporate debt constituted 14 percent of the total securities portfolio. - Asset-backed securities (ABS) and mortgage-backed securities (MBS) accounted for 12 percent. - Portfolio allocation by accounting book: 12.4 percent of securities in the AC portfolio, 33.6 percent in the FVOCI portfolio, and the residual in the FVPL book (chart showed 12.4, 33.6, 53.9). - Commodity and mutual fund risks: - Accounted for using the scenario shocks. - Banking sector exposure to commodity and mutual fund risks relatively limited compared to interest rate and spread risks. ### Net interest income (NII) projections - Effective interest rate projection methodology: - Econometric models estimated for six asset and six liability interest-bearing segments (twelve segments in total) using quarterly historical data on effective balances and interest rates. - BMA techniques used to calibrate econometric models for each segment; BMA pools hundreds or thousands of equations and produces a posterior model equation with weighted predictive performance. - Predictor pool included variables such as real GDP, investment, consumption, price inflation, short- and long-term interest rates, and credit spreads. - Detailed satellite estimations in Appendix V. - Segments used (Table 9): - Assets: A1 Interbank deposits; A2 Securities; A3 Mortgage loans; A4 Business and government loans; A5 Consumer loans; A6 Reverse repos. - Liabilities: L1 Personal demand deposits; L2 Personal term deposits; L3 Other demand deposits; L4 Other term deposits (wholesale); L5 Subordinated debt; L6 Other interest-bearing liabilities. - Bank-specific adjustments: - Bank-specific starting point distance by segment informed bank-specific interest rate forward paths. - Bank-specific interest rate percentage deltas from system-wide composite interest rate at the starting point used to adjust composite projections and produce bank-specific interest rate projection paths for both scenarios. - Pass-through and repricing constraints: - Maturity gap analysis used to reflect pass-through constraints for mortgages and consumer credit. - Most liability segments have short-term maturity profiles implying fast repricing; for these, direct NII/expense projection used the satellite-model effective interest rate. - For mortgages and consumer credit, additional constraints employed due to longer-term repricing gaps: - New mortgage origination rates from scenarios used with repricing gap to derive bank-specific effective mortgage interest rate paths. - An additional cap of 330 basis points for the spread of the mortgage rate over the personal demand deposit rate introduced to ensure conservatism and reflect structural competition factors. - For consumer credit, the effective interest rate projected by the econometric model used as the proxy for the portion of the portfolio repriced during the period (no additional constraints). - After these pass-through constraints, tighter interest margins observed in the adverse scenario during the first two years of the horizon. ### Non-interest income projections - Importance: - Non-interest income is a significant component of Canadian banks’ profitability and internal capital generation. - Non-interest total income reached a level of almost 75 percent compared to the sector’s net interest income in 2018 and has been a particularly stable source of profitability over the past seven years. - Stressing approach in the adverse scenario: - Historical data on total non-interest income were collected and broken down by specific activity for the last five years. - Conservative bank-specific projected income produced by adjusting yearly profit by activity: under the adverse scenario, profits from each business activity projected equal to the latest income minus one standard deviation of historical variability of the income. - Minimum drop in yearly income introduced for each activity to account for sensitivity to adverse macrofinancial conditions. - Bank-specific standard deviations calculated after dividing historical profits by bank total assets each year to account for non-domestic expansion. - Foreign-currency composition of each revenue stream accounted for by applying scenario exchange rate movements to currency breakdown at cut-off date. - Operating expenses assumed remain constant through the scenario, allowing only exchange rate effects. - Different sub-components stressed according to degree of market sensitivity or historical variability; income breakdown across activities generates different projections per bank. - Additional details on business activity lines and minimum scaling factors in Appendix VI. ### Key methodological timeframes and assumptions - PiT stage transition rates data period: first three quarters of 2018. - Baseline scenario macroeconomic variables: assumed constant after year 3. - Adverse scenario macroeconomic variables: converge back to baseline within six years. - Perfect scenario foresight assumed for provisioning projections. - Pass-through constraints effect: tighter interest margins in the adverse scenario during the first two years of the horizon. - Mortgage rate spread cap: 330 basis points. *Source: IMF staff compilation from the presented FSAP chapter content.* ### 74. Banks appear resilient to severe macrofinancial shocks, with all banks meeting the ### Banks appear resilient to severe macrofinancial shocks ### Stress test aggregate solvency outcomes - In the adverse scenario, the aggregate CET1 capital ratio would decline by 4.8 percentage points to 7.4 percent in 2020 before recovering to 9.6 percent in 2021. - The difference between the baseline and adverse aggregate CET1 capital ratio at the end of the stress testing horizon is 5.6 percentage points. - In the baseline, the system-wide aggregate CET1 capital ratio would be on an upward trajectory due to banks’ solid revenue-generating capacity. - During the stress testing horizon, most banks would run down conservation capital buffers, subjecting them to dividend restrictions (partially or fully). - The worst outcome in 2020 largely reflects the adverse scenario dynamics: a strong recovery and a more favorable interest rate environment in 2021. - Cumulative pre-loss income during 2019–20 in the adverse scenario would amount to 2.9 percent of RWAs, compared with losses due to credit and market risks of 3.9 and 1.8 percent of RWAs, respectively. - Deteriorating credit quality, as reflected in higher RWAs, would also have an impact of 1.2 percentage points in the final CET1 ratio. ### Accounting considerations and IFRS 9 - Stress test results highlight the importance of properly capturing accounting considerations, particularly related to IFRS 9. - Under the adverse scenario, aggregate Stage 2 assets and Stage 3 assets would account for 13.6 and 8.3 percent of total exposures, respectively. - Accounting impairment charges in the adverse scenario would exceed regulatory provisions for most banks. - Excess provisions are added back to tier-2 capital, up to the cap of 0.6 percent of credit RWAs, to derive a more accurate projection of realistic CET1 projections. ### Credit losses and impairment distribution - Credit losses are the main driver of the solvency result; impairments in the adverse scenario are significant across all asset classes. - Cumulative default rates at the end of the stress testing horizon (adverse vs baseline): - Uninsured mortgages and HELOCs: 6.4 percent (adverse) versus 0.7 percent (baseline). - Consumer retail exposures: 13.3 percent (adverse) versus 8.3 percent (baseline). - Corporate exposures: 8 percent (adverse) versus 3.8 percent (baseline). - Additional cumulative impairment flow under the adverse scenario reaches approximately Can$90 billion. - Of the Can$90 billion, 37 percent correspond to corporates, 34 percent to consumer credit asset classes, and 29 percent to uninsured mortgages and HELOCs. ### Net interest income and non-interest income impacts - Net interest income (NII) experiences a significant drop in the adverse scenario due to imperfect pass-through: - Average yearly NII is approximately 24 percent lower when compared to the year preceding the cut-off date (2018) and the average NII in the baseline. - Aggregate net interest income reaches the lowest value during Year 2 when net interest expense is at its maximum; at this point NII is 34 percent below the cut-off year level, and broadly recovers thereafter. - Non-interest income projections are approximately 20 percent lower for the system aggregate in the adverse scenario relative to the baseline. - Banks still manage to generate Can$168 billion of non-interest profits during the three years, or an average of Can$56 billion per year, versus a yearly starting point of Can$70 billion. ### Market risk, revaluation losses, and RWAs - The total market risk impact largely depends on the scenario path, specifically the interest rate path. - In the adverse scenario: - Revaluation market risk losses in Year 1 reach Can$43 billion and are mostly partly recovered during Years 2 and 3. - Profit-and-loss and other comprehensive income losses are not substantial (less than Can$10 billion, or approximately 0.5 percent of RWAs), but timing matters. - In the baseline, interest rate losses are more protracted: at the end of the stress testing horizon market risk losses account for Can$32.4 billion, or 1.6 percent of RWAs. - Unfavorable macrofinancial conditions materially increase credit risk RWAs in the adverse scenario: - Credit risk RWAs increase to Can$1.66 trillion in Year 2 versus a starting level of Can$1.29 trillion. - Compared with the starting point, credit RWAs increase by 29 percent by Year 2 and 17 percent by Year 3. - The increase in credit RWAs has a negative contribution to CET1 depletion of 1.2 percent and 0.8 percent of total RWAs in Year 2 and Year 3, respectively. - The impact of market risk RWA changes in the adverse scenario is immaterial given the small relative size of market risk RWAs. - Derivatives exposures were not explicitly stressed due to unavailability of sensitivity data; counterparty credit risk and CVA risk were excluded due to data limitations. ### Offsetting factors: taxes and dividend restrictions - Incurred losses under the adverse scenario are partially offset by lower provisions for tax and assumed restrictions on dividend distribution relative to the baseline. - Tax-linked outflows are lower by 0.9 percent of starting RWAs. - Dividend distributions are lower by 0.7 percent of RWAs. - These offsets partially compensate for the more severe CET1 depletion in the adverse scenario. - Note: The FSAP annual proxy for dividend restriction may be slower to reflect the actual quarterly regulatory restriction timing. ### Sensitivity analysis — undrawn exposures, IFRS 9 tenor, and risk-based pricing - Stressed undrawn exposures (quasi-static balance sheet limitation acknowledged): - Assumed annual drawdown rates on top of the drawn exposure path: uninsured mortgages 19 percent, HELOCs 13 percent, consumer credit 3 percent, credit cards 6 percent, corporate 9 percent. - Under these assumptions, total impairments in the adverse scenario are Can$18.5 billion higher than the main constrained-balance-sheet results. - This corresponds to an additional CET1 impact of 91 basis points in the adverse scenario by the end of Year 3. - Large undrawn credit lines, particularly HELOCs, may mask borrower distress and amplify household vulnerabilities; close monitoring or preventive action may be required. - Stressed residual amortization tenor for IFRS 9 lifetime expected credit loss: - Alternative assumption: average lifetime for mortgages set at 15 years versus 2.5 years used in the primary approach (yearly portfolio maturation rate assumed 4 percent, corresponding to a 25-year amortization period). - Under this assumption, Stage II mortgages in the adverse scenario increase to 16.4 percent from 13.6 percent under the primary approach. - The average coverage ratio across all asset segments almost doubles to 3.4 percent. - Total credit impairment flows increase by Can$5.2 billion to Can$94.8 billion for Year 3 of the adverse scenario. - This increase corresponds to an additional decline of the CET1 capital ratio by 26 basis points at the end of the stress testing horizon. - Results should be interpreted with caution because of limited data to differentiate the quality of exposures flowing into Stage II; lifetime expected credit losses might be underestimated versus bottom-up projections. - Risk-based pricing for mortgages may have a more substantial effect on mortgage default rates under the adverse scenario (not quantified here). *Source: IMF staff estimates contained in the provided FSAP stress testing text.* ### 90. Risk-based pricing analysis examines the impact that additional capital charge costs ### 1canea2020001 - 90. Risk-based pricing analysis examines the impact that additional capital charge costs ### Risk-based pricing methodology - Simulation modified the satellite structural model for mortgage defaults to include an additional variable reflecting assumed risk-based sensitive spread over the lending rate. - Spread adjustments by DSR band applied to all borrowers irrespective of LTV: - DSR ratios higher than 0.4: additional spread of 100 basis points. - DSR>0.3 (next band): adjusted higher spread of 50 basis points. - Prime borrowers (DSR<0.15): spread tighter by 50 basis points to the applied benchmark interest rate change. - Effective interest rates for the other DSR bands were not changed. ### Impact on projected defaults and losses - PD PiTs for both insured and uninsured segments increase by 40 to 50 percent under the risk-based pricing assumption. - Loss rates increase by a similar magnitude as PD PiTs. - Several model simulations with different risk-based pricing spreads confirm an increase in loss rates by some magnitude. ### Systemic implications and mitigation - Importance of mechanisms to prevent “wrong-way” risks where elevated RWAs and additional capital charge costs amplify borrower affordability deterioration during downturns. - Recommendation: build capital buffers during good times to mitigate amplifying correlation risks when house price developments and affordability concerns contribute to credit quality deterioration. - Shaping the institutional landscape toward countercyclical risk-based pricing for all mortgages is suggested as a first step. ### Mortgage portfolio amortization vulnerability - Significant concentration of exposures with average remaining amortization tenors between 20 and 30 years (Figure 17). - Concerns: - Skewed remaining amortization profile may reflect extension of amortization periods during contractual renewals. - Remaining amortization tenor profile may mask increased household balance sheet vulnerabilities. - Limits availability of the extension of amortization period as temporary forbearance in a severe housing shock. - Contributes to amplification of systemic risks associated with renewal risk, particularly if banks must deleverage. ### Concentration risk analysis - Large credit exposures to non-financial corporations assessed using simultaneous default scenarios with aggregate LGD of 50 percent: - Default of the single largest exposure: impact of 38 basis points in the average CET1 capital ratio. - Concurrent default of the five largest exposures: reduction in CET1 capital ratio by 1.2 percentage points. - Simultaneous default of the ten largest exposures: decline in CET1 capital ratio by 1.8 percentage points. - In each theoretical case, no individual bank would fall below any regulatory CET1 capital minimum. - Name-specific review found most significant exposures toward well-established domestic and global corporate entities. - Interbank unsecured counterparty exposures show scattered concentrations at the operational level (custodian banks, clearing houses, central counterparties); review did not find elevated or unexpected concentration risks for the Canadian banking sector. ### Top-down stress tests and Bank of Canada (BOC) framework - BOC stress test results largely aligned with FSAP stress test results on overall capital impact and relative bank performance. - BOC framework evolution: - Originates in robust analysis of second-round contagion and spillover effects (MacroFinancial Risk Assessment Framework module). - Development of a unified top-down macro stress testing framework (Macro Stress Testing module) that can combine with second-round analysis. - Further enhancements required: - Satellite models should accommodate bank-specific characteristics and starting points; collect wider bank-specific historical data; extend models/estimation techniques to capture idiosyncratic attributes. - Increase granularity to capture important domestic data attributes (e.g., geographies of exposures, significant currencies); avoid a single “blended” mix across geographies or currencies. - Adapt satellite models to produce bank-specific projections anchored to bank-specific parameters; support by regular bank reporting of required data. - Wider coverage of risks (market risk, operational risk), scenario-consistent balance sheet evolution (quasi-static or dynamic), more granular RWA tracking for performing and nonperforming portfolios, full coverage of nonperforming exposures, and development of an accounting layer to distinguish economic losses from accounting reality (fully fledged IFRS 9 accounting layer). ### Policy recommendations (summarized) - Additional required capital for mortgage exposures is desirable: - Banks’ overall capital buffers are adequate, but lenders’ risk weights for mortgage exposures should be higher. - Authorities should develop and communicate a minimum set of attributes sufficient to characterize a historical period as a full economic cycle; enforce use of downturn proxies if historical downturn episodes are lacking. - Avoid extreme procyclicality by incorporating buffers in the system (including for banks and mortgage insurers) during good times. - Enforce higher levels of RWA densities for mortgage portfolios to mitigate risk amplifications during downturns. - Use of “challenger” models: - Conceptually orthogonal models to classical econometric techniques may provide end-state visibility after simulated shocks and artificially enlarge the available data space using simulation. - Banks may be guided to use similar approaches in internal models to ensure stressed history is included in calibrations. - Stress testing infrastructure and data reporting: - Develop stress testing infrastructure and redesign relevant data needs. - Reporting should better capture country-specific attributes and foreign-currency linked dependencies as geographies outside Canada become more important. - System-wide risks should be accounted for; additional bank-specific reporting that captures vulnerabilities (post aggregation) at the system level may be needed. - Monitoring and reporting priorities given system size and complexity: - Focus on large derivatives exposures (foreign-currency exposure mismatches, interest rate exposures), significant counterparty risks, impact and inefficiencies of existing hedging strategies, and funding or currency mix concentration. - Recommended: BOC, OSFI, and AMF agree on a common monitoring and reporting framework to provide a “bird’s eye view” of the system and selectively include these themes in top-down or bottom-up stress tests. ### Bank liquidity stress tests — Overview and LCR-based results - Two liquidity risk approaches used: LCR-based test and cash-flow analysis. Scenarios derived from historical liquidity crisis episodes targeting domestic market vulnerabilities. - Current liquidity conditions: - Canadian banks well diversified in funding mix: retail and wholesale deposits, repo, and derivative-linked segments. - Significant derivatives-related liabilities raise concerns on foreign-currency funding mix; asset side shows no major mismatches when considered. - Notional size of FX and cross-currency swaps warrants thorough monitoring of cross-currency funding. - LCR-based tests: - LCR regulatory standard: 100 percent over a 30-day stress scenario. - Three scenarios calibrated: - Scenario 1 (Retail): stressed run-offs on retail deposits. - Scenario 2 (Wholesale): increased run-offs on wholesale sources, especially wholesale/corporate demand deposits. - Scenario 3 (Combined): stressed run-offs applied in parallel (considered less plausible). - Tests performed per significant currency and aggregated LCR. - Aggregate LCR outcomes: - Starting point: 133.3 percent. - Scenario 1 (Retail): aggregate average LCR of 127.6 percent. - Scenario 2 (Wholesale): aggregate average LCR of 104.4 percent. - Scenario 3 (Combined): aggregate LCR falls to 90.3 percent (not regarded as plausibly binding from a prudential angle). - Significant currency LCR movements: - U.S. dollar LCR: from 121.5 percent to 118.7 percent (retail scenario) and 98.3 percent (wholesale scenario). - Canadian dollar LCR: from 116 percent to 109.8 percent (retail scenario) and 90.7 percent (wholesale scenario). - Conclusion: ample space for banks to accommodate outflows given solid buffers in security holdings and significant short-term inflows from the loan book. *Source: IMF staff estimates based on data provided by OSFI, AMF, BOC.* ### 110. The cash-flow analysis is based on the assessment of banks’ ability to withstand ### 1canea2020001 - 110. The cash-flow analysis is based on the assessment of banks’ ability to withstand ### Cash-flow analysis: methodology - Liquidity outflows over a forward-looking three-month period are assessed using the contractual maturity ladder for assets and liabilities; a liquidity stress duration of three months is assumed. - A granular set of severity scenarios is defined; each severity level is associated with scenario-specific run-off rates on liabilities and fair value and discount haircuts for liquid assets. - As scenario severity increases, higher run-offs are applied and an overall funding shortfall is estimated; utilization of liquidity and credit off-balance sheet commitments is accounted for. - Counterbalancing capacity is measured by estimating additional liquidity generation across all liquid asset classes after applying scenario-specific fair value and collateral haircuts; counterbalancing capacity is compared to funding needs to establish a liquidity surplus or shortfall. - For each scenario a full collateral revaluation is performed, including collateral used in repo and reverse repo transactions. - Assumptions and conservatism: - Net Cumulative Cash Flow (NCCF) regulatory template was used at a bank consolidated level. - NCCF returns for U.S. subsidiaries were not incorporated into the calculations. - Roll-over rates of maturing assets are assumed to be 100 percent (banks are not allowed to counterbalance outflows by deleveraging or by not extending new credit facilities). - Collateral haircuts for eligible collateral are anchored to standing liquidity facilities haircuts of the BOC and assumed to be constant during a liquidity stress period. - Second-round effects capturing additional collateral needs for existing collateralized funding positions are estimated. - The analysis ignores any positive inflows from the loan portfolio under the constraint that banks are not allowed to reduce credit intermediation during the liquidity episode. - The process estimates additional collateral required to compensate for falling collateral valuations but does not perform a full repricing of derivatives positions. ### Cash-flow analysis: results and key statistics - The cash-flow analysis identifies small liquidity shortfalls for several banks under some severity levels close to the severely adverse case (SL 21). - Aggregate shortfalls: - Combined liquidity shortfall (if only banks with a shortfall are included) in the severely adverse scenario reaches Can$91 billion. - The highest shortfall for an individual bank stands at Can$26 billion. - Interpretation: - Results suggest a large outflow shock would be needed to generate liquidity shortfalls for some Canadian banks. - Even in extreme cases, the shortfalls are considered of a manageable magnitude given the ample space of the BOC to provide liquidity through standard facilities and extraordinary measures. - Drivers of resilience and risks: - Banks’ ability to withstand liquidity shocks is attributed to elevated securities holdings (a significant part of which is currently or could become eligible for BOC liquidity facilities) and absence of significant mismatches in repo books. - The balanced picture assumes repo and reverse-repo run-offs are matched. - On the downside, significant concentrations in wholesale or corporate deposits should be closely monitored due to their unstable nature. - Data gaps remain in identifying eligibility of encumbered and unencumbered collateral. - Liquidity impact of high derivatives exposures (e.g., currency swap liquidity risks and total return swaps) could not be assessed due to data limitations. ### Assessment of OSFI’s liquidity monitoring - Current liquidity templates used by competent authorities capture rich and granular data; however, additional and more sophisticated pre-validation is needed because several submitted liquidity templates failed basic validation checks. - Recommendations to improve monitoring data utility: - Cross-check all past submissions before acceptance to better utilize the information repository. - Complement contractual cash flow reporting with a behavioral flow add-on. - Enhance NCCF reporting with better information on collateral eligible for BOC liquidity facilities to make utilization measures per bank and system-wide readily available. - Integrate behavioral flows on utilization of off-balance sheet credit and liquidity commitments into NCCF calculation; although reported, commitments are not associated with specific outflows. - Ensure classification and reporting of outstanding balances by liability segment follow strict guidelines to minimize subjective interpretation. ### Policy recommendations (liquidity and market functioning) - Closely monitor substantial notional amounts of both liability and asset-related derivatives: - Current liquidity monitoring does not provide insight into nature of notional exposures, relevant sensitivities, and how dysfunctional markets might impact banks’ liquidity profiles. - Additional reporting and monitoring should focus on first- and second-order sensitivities of these positions to different risk factor shocks to identify potential risks on both sides of the balance sheet and to establish a system-wide view on use of such products. - Implement regular benchmarking exercises to identify outliers and enhance risk-based supervision. - Monitor risks arising from repo market functioning: - Canadian banks are very active in repo markets and use a large proportion of securities holdings as collateral in repo/reverse-repo transactions. - Constant monitoring of repo market functioning and the quality of unencumbered collateral is recommended to identify early warning signals of liquidity or financial distress that could impair market confidence. ### Life insurance solvency stress tests — overview and scope - FSAP, in collaboration with OSFI and AMF, conducted life insurance solvency stress tests covering the five largest life insurers (three federally regulated globally active entities and two provincially regulated entities in Québec). - Coverage: - These five insurers correspond to 90 percent of life insurers’ total assets and approximately 80 percent of life insurers’ total net premiums. - Exercise: - Using regulatory and supervisory data, the exercise assessed instantaneous impact of macrofinancial shocks (consistent with the adverse scenario) on life insurers’ solvency position. - Sensitivity of the Life Insurance Capital Adequacy Test (LICAT) ratio was assessed for macrofinancial conditions in 2019Q3 (most severe financial market stress) and 2021Q4 (lowest interest rates) in the adverse scenario. ### Life insurance sector: recent performance and portfolio composition - Profitability and solvency: - Life insurers have maintained high profitability in a low interest rate environment; return on equity has remained stable. - Life insurers maintain strong solvency positions, holding capital buffers in excess of supervisory targets. - Geographic and asset composition: - While Canada accounts for over 40 percent of net premiums, life insurers have significantly expanded overseas operations, particularly in Asia and Europe. - Investment assets predominantly consist of bonds; equity, mortgages, and properties account for less than a quarter of the investment portfolios of the five largest life insurers. - Over 60 percent of assets are invested abroad, with 26 percent of assets invested in the United States. - Canadian bonds account for approximately 41 percent of the aggregate bond portfolio and 55 percent of AAA bonds (including bonds eligible for 0 percent LICAT factor). - Credit quality: - Two-thirds of bonds are rated A or higher. - 16 percent of bonds are either unrated or have a rating below B. - Nearly half of bonds with rating below B or unrated are held in Canada and about 30 percent in the United States. - Investment portfolio headline figures (based on five largest life insurers): - Bonds: C$483bn, (77) - Equity: C$36bn, (6) - Mortgages: C$70bn, (11) - Property: C$36bn, (6) - Assets by region: Canada C$278bn, (39); USA C$182bn, (26); UK C$51bn, (7); Europe (Excl. UK) C$93bn, 13; Japan C$34bn, (5); Other C$73bn, (10) ### Life insurance stress test: methodology - Channels affecting capital ratios in the adverse scenario: - Shock to risk-free rates affects present value of liability cash-flows through the discount factor. - Asset price shocks affect value of the investment portfolio and market-risk capital requirements. - Assets, liabilities and capital requirements are calculated for each of six LICAT regions: Canada, the United States, the United Kingdom, Europe (excluding the United Kingdom), Japan and “other regions.” Aggregation is in Canadian dollars. - Supervisory hurdle rates: - Supervisory minimum requirements used as hurdle rates are 90 percent for the total capital ratio and 50 percent for the core capital ratio. - Total capital ratio = (available capital (tier-1 + tier-2) + surplus allowance + eligible deposits) / base solvency buffer. - Core capital ratio = (tier-1 capital + surplus allowance (70 percent) + eligible deposits (70 percent)) / base solvency buffer. - Base solvency buffer = sum of aggregate capital requirement net of credits for each of six regions multiplied by scalar 1.05. - Aggregate capital requirement within a geography includes credit risk, market risk, insurance risk, segregated funds guarantee risk and operational risk. - Surplus allowance is based on provisions for adverse deviations (PfADs), assumed to be a fixed proportion of liabilities. - Discount rate assumptions: - Liability discount factor consists of a risk-free and a spread component. - Shock to the risk-free component of the discount rate is derived from deviation in government bond yields in the adverse scenario compared to the baseline. - Spread component of the liability discount rate is assumed constant (exercise assumes changes in bond yield spreads are entirely due to changes in level and pricing of credit risk). - This differs from CALM, under which appointed actuaries assess extent to which bond yield spreads are due to liquidity risk versus credit risk. - Asset valuation: - Fixed income assets are affected by shocks to risk-free rates and to credit spreads; shocks vary by years to maturity, credit rating and geography. - Alternative assumption: apply same discount rate for assets and liabilities (more similar to CALM), which assumes no changes in bond yield spreads. - Exchange rate effects: bilateral spot rates are used to calculate exchange rate effects on assets and liabilities; percentage change in the U.S. dollar/Canadian dollar spot rate used as a proxy for “other regions.” - Mark-to-market approach for equity and real estate: equity values use national equity price indices at 2019Q3 and 2021Q4; real estate follows national house price indices. - Capital requirements adjustments: - Changes in capital requirements only reflect changes in market risk due to changes in exposures. - For market risk, capital requirements for interest rate risk, equity risk and real estate are assumed to remain a constant proportion of respective asset values. - Capital requirement for credit risk is unaffected by local currency market value of fixed income assets since bonds are assumed to keep same credit rating. *Source: IMF staff estimates.* ### 131. Large life insurers are somewhat vulnerable to macrofinancial shocks (Figure 23). In the ### Large life insurers are somewhat vulnerable to macrofinancial shocks (Figure 23) ### Life insurers: Top-down stress test results (adverse scenario) - Aggregate core capital ratio of the five largest life insurers would decline by 34 percentage points to 61 percent by 2019Q3, largely driven by the impact of widening credit spreads and falling equity prices. - Some entities would have capital falling below the 50 percent regulatory minimum. - Aggregate total capital ratio would fall by 32 percentage points to 104 percent by 2019Q3; total capital ratios of some life insurers would fall below the 90 percent hurdle rate. - By 2021Q4: - Aggregate core capital ratio would reach 90 percent. - Aggregate total capital ratio would reach 132 percent. - All life insurers would maintain their capital ratios above regulatory minimums. ### Alternative discount-rate assumption (liability discount rate applied to assets) - Applying the same discount rate to assets and liabilities (using the liability discount rate similar to CALM): - Aggregate core capital ratio would increase to 110 percent in 2019Q3. - Aggregate total capital ratio would increase to 155 percent in 2019Q3. - This alternative does not consider widening bond credit spreads. - By 2021Q4, results are the same for both methodologies because shocks to credit spreads are assumed to have fully dissipated. ### Asset and liability drivers of stress-test outcomes - Asset value changes in the adverse scenario (compared to 2018Q4): - Asset values fall by Can $75 billion largely due to a decrease in the bond portfolio value in 2019Q3. - Aggregate bond portfolio falls by Can$50 billion in 2019Q3, due to the increase in the risk-free rate and widening credit spreads. - Aggregate equity holdings value would fall by over Can $20 billion (includes equity in investment portfolio and “other loans and invested assets”). - Mortgages value falls in domestic currency; Canadian dollar depreciation vs. U.S. dollar implies a moderate increase in the Canadian dollar value of U.S. mortgages. - By 2021Q4: - Asset values increase by Can $10 billion compared to 2018Q4, largely due to the bond portfolio. - Bond portfolio value increases by Can $14 billion as credit spreads abate and risk-free rates decline. - Equity prices almost recover to 2018Q4 levels in local currency; exchange rate effect implies increase in Canadian dollar value of equity portfolio. ### Sensitivity to risk-free rates and discount-rate assumptions - Life insurers are sensitive to falling risk-free interest rates. - In the adverse scenario, increases in risk-free rates by 2019Q3 cause the value of liabilities and PfADs in the surplus allowance to fall, which more than offsets the negative effect of rising risk-free rates on the fixed income portfolio—producing a positive total effect on the numerator of the capital ratio in 2019Q3. - The exchange rate has a positive effect on the numerator; credit spread and equity price shocks have negative effects. - A lower risk-free yield curve in 2021Q4 affects the capital ratio negatively. - Assuming the same discount rate for assets and liabilities, a parallel downward shift in the discount yield curve by one percentage point would reduce the core capital ratio by 40 percentage points. ### Policy recommendations for life insurers - Insurance supervisors should continue to monitor life insurers’ portfolios of low-rated and unrated bonds. Search for yields during a prolonged period of low interest rates has resulted in large exposures to lower quality and unrated fixed income assets, making life insurers particularly vulnerable to credit spread shocks. - In preparation for IFRS 17, OSFI and AMF should carefully consider how risk margins interact with the regulatory solvency framework for life insurers: - Under the current framework (CALM), liability cash-flows are discounted using asset discount rates minus allowances for expected defaults, a market shock and a margin for adverse deviation in the expected default assumption. - Fixed income portfolios of life insurers imply credit risk and sensitivity to macrofinancial conditions, while liability cash-flows are largely independent of such factors. - A framework that takes this into consideration would allow for more accurate estimates of the present value of cash-flows during periods of elevated credit risk. ### Mortgage insurance stress tests: overview and methodology - Exercise covered all three mortgage insurers and used a balance-sheet approach with regulatory data, consistent with bank solvency stress test methodology. - Two scenarios over the 3-year horizon 2019Q1-2021Q4: baseline (October 2018 WEO) and adverse scenario. - Supervisory target capital ratio: Mortgage Insurance Capital Adequacy Test (MICAT) of 150 percent (hurdle rate). - Key modelling assumptions and approaches: - Adverse scenario affects mortgage insurers via increased default rates, higher loss given default from house price declines (increasing total claims), lower credit growth (reducing net premiums written), and asset price shocks affecting investment portfolios and income. - Total claims projected to mirror banks’ potential losses on mortgage exposures; PD and LGD parameters assumed to coincide with bank stress test. - Unearned premiums assumed proportionate to mortgage insurance-in-force (no provisions held for mortgages already defaulted). - Mark-to-market approach for asset price shocks; investment portfolios are above 90 percent domestic interest-sensitive assets. - Investment income projected via simple regressions on the size of the investment portfolio. - Taxes: mortgage insurers assumed to obtain tax refunds from reported operating losses (taxes paid within previous three years eligible for refund). - Capital requirements: - For credit and market risks: proportionate to asset values. - Operational risk: 20 percent of the sum of capital requirements for insurance, market and credit risk. - Sensitivity analysis on MICAT insurance risk requirement: two methodologies - Primary: capital requirement proportionate to mortgage insurance in-force outstanding (consistent with MICAT). - Alternative: constant capital requirement ratios within 10 LTV buckets with LTVs updated quarterly with the national house price index. ### Mortgage insurance stress-test results (adverse scenario) - Mortgage insurers are vulnerable to severe macroeconomic downturns with significant house price declines; aggregate capital would turn negative. - Mortgage insurers would need additional capital of Can$15 billion to meet the supervisory solvency target (150 percent MICAT). - Cumulative insurance claims projected at Can$25 billion (consistent with banks’ insured mortgage portfolio credit losses); compared to Can$1 billion during 2015Q4-2018Q3. - PD assumed 2.8 percent per year on average 2019–2021; LGD assumed 43 percent over the same period. - Investment portfolios’ market value remains largely unchanged (yield curve shifts out in 2019–20 and back in 2021). - Revenue and income impacts: - Total revenues fall from Can$3.2 billion per year in 2016–18 to Can$1.6 billion a year in 2019–21. - Tax refunds of Can$1.5 billion over the 3-year period. - Total underwriting revenue falls from Can$2.4 billion a year to Can$0.5 billion amid negative credit growth and a 5 percent decrease in insured mortgage portfolio. - Investment income falls from Can$2.5 billion for 2016–18 to Can$1.5 billion for 2019–21. - Aggregate losses and capital trajectory: - Total costs amount to Can$24.3 billion and total revenue to Can$4.8 billion over the stress horizon → total losses of Can$19.5 billion. - Capital available falls from Can$18 billion in 2018Q3 to negative Can$1 billion by 2021Q4 (driven by Can$19.5 billion in losses, roughly unchanged investment portfolio value, and Can$0.8 billion decrease in liabilities due to smaller outstanding insured mortgage portfolio). - Government guarantee would mitigate losses sustained by banks despite mortgage insurers’ negative capital in the test. - Alternative assumption (updated LTVs): - Required capital would be Can$8.5 billion higher. - MICAT capital requirement stable, reaching Can$13.9 billion by end-2021 (insurance risk ~80 percent). - Updating insurance capital requirement with current house prices would increase requirement to Can$22.3 billion by end-2021 → implied capital shortfall of approximately Can$23 billion. - System-wide capital buffer assessment: - Seven D-SIFIs have capital buffers for insured mortgage exposures equivalent to 0.17 percent of outstanding insured mortgages (as of October 2018, these seven entities had insured mortgage exposures of Can$453 billion). - Accounting for mortgage insurers’ required capital for insurance risk, system-wide capital buffers would amount to 1.96 percent. - In an adverse scenario, these buffers should rise to 4.32 percent, implying additional capital need of Can$28 billion to cover expected and unexpected losses for insured mortgage exposures. *Sources: AMF; OSFI; IMF staff calculations.* ### 156. Mortgage insurers should hold more capital so that they are more resilient to severe ### 1canea2020001 - 156. Mortgage insurers should hold more capital so that they are more resilient to severe ### Mortgage insurers: capital adequacy and policy recommendation - Mortgage insurers may face a capital shortfall in a severely adverse scenario as illustrated by stress tests. - Recommendation: Mortgage insurers should hold more capital so that they are more resilient to severe macrofinancial shocks. - Recommendation: Steps should be taken to ensure that required capital of mortgage insurers is enough to absorb tail-risk shocks. - Recommendation: Capital requirements for mortgage exposures at lenders and mortgage insurers should be tightened to properly account for through-the-cycle credit risk that may exceed Canadian historical experience. ### Overview of risk-taking in nonbanks and financial markets - The FSAP assessed risks in nonbank financial institutions and financial market activities. - The nonbank sector has grown rapidly since the 2014 FSAP, driven largely by investment funds. - Some financially risky private market activities may not be captured in official sector monitoring, increasing the importance of improving data coverage. ### Landscape of nonbank and financial market participants — key size and growth figures - Assets of nonbank financial institutions (NBFIs) excluding insurers accounted for almost two-thirds of the Can$13.9 trillion in financial system assets (635 percent of GDP). - NBFIs excluding insurers accounted for about four-fifths of the 71 percentage point increase in financial system assets relative to GDP between 2013 and 2018Q3. - Mutual fund sector: Can$2.3 trillion (now 105 percent of GDP), includes exchange-traded funds (ETFs). - Pension fund assets: 113 percent of GDP (noted as 114 percent of GDP elsewhere in the chapter). - Other investment funds: Can$570 billion in gross assets; hedge funds: Can$110 billion in gross assets. - According to the FSB definition, nonbank financial intermediation in Canada was US$1.57 trillion at end-2017. - Growth rates: roughly 11 percent a year during 2011–16, slowing to 3.5 percent in 2017. - FSB-defined “nonbank financial intermediation” in Canada reached Can$2.1 trillion (investment funds with fixed income holdings represent about three quarters of this). - ETFs: about Can$160 billion in assets (10 percent of publicly offered fund assets) as of mid-2018. - Foreign holdings contributed: foreign equities 50 percent and foreign bonds 17 percent of total sector asset growth excluding mutual fund shares. - Bond funds and bond fund of funds accounted for slightly less than one-third of publicly offered fund assets. - Pension funds hold over 30 percent of outstanding provincial government bonds and foreign assets equivalent to about a third of total assets under management. - Six large public pension funds account for roughly two-fifths of all pension fund assets. - Privately offered investment funds (OSC survey): Can$570 billion in gross assets as of end-2017; fund net assets grew 14 percent in 2017. - Real estate investment trusts (REITs): Can$145 billion (as of end-2017). - NBFIs engaged in credit intermediation beyond asset management estimated at 12 percent of GDP. - Mortgage finance companies (MFCs): about Can$60 billion in assets. - Mortgage investment companies (MICs): about Can$15 billion in assets as of end-2017. - Consumer and business leasing sector: about Can$185 billion. - Captive financial institutions and money lenders (CFIML) account for almost one quarter of financial system assets (156 percent of GDP). - Other private financial institutions total: 156 percent of GDP (category includes low-risk entities and potentially significant financial risk-taking entities). ### Risk-taking among institutional investors — exposures, leverage, and liquidity concerns - Institutional investors have increased risk-taking via private market strategies and derivatives, adding leverage and liquidity risks. - Alternative and private market investments contributed 40 percent of the increase in pension fund assets since 2013; foreign equities contributed 39 percent. - Share of Level 3 (illiquid, difficult-to-value) assets in six largest public pension funds rose from 31 to 36 percent over the period examined. - Reported gross leverage for six large public pensions: about 130 percent post-crisis (declining somewhat in recent years). - Adjusted for alternative and private market assets, estimated effective leverage in publicly traded portfolios of the six large pension funds is closer to 150 percent. - Six large public pension funds report gross notional derivatives positions outstanding reaching nearly 200 percent as of 2017. - About three-fourths of pension fund derivative exposures captured by OSC data have foreign counterparties. - Institutional investors’ use of non-plain vanilla OTC derivatives (e.g., bond forwards, equity swaps) has grown, adding leverage and liquidity risks. - Broker-dealer lending and short-term wholesale funding have grown significantly, supporting increased market risk-taking. ### Asset price valuations and market risks - Fixed income asset valuations are stretched by historical norms; term premiums have reached multi-decade lows and, by some models, turned negative. - Canada stands out for having among the flattest yield curves, partly reflecting large pension and insurer demand for longer-dated debt. - The 30-year/2-year yield differential for government bonds has been lower than major bond markets for much of the post-GFC period. - Corporate bond duration-adjusted credit spreads are near historically low levels, close to the low levels seen in the United States. - The implied price of volatility in government bonds has fallen to multi-year lows. *Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1canea2020001.pdf* ### 173. Commercial real estate prices are stretched. Capitalization rates—a measure of estimated ### Commercial real estate prices are stretched. ### Asset valuations and market risks - Capitalization rates (annual net income as a percentage of purchase price) have been steadily trending downwards since the end of the GFC, according to CBRE Group. - Certain CRE property types and major metropolitan areas show particularly stretched valuations, e.g., multifamily residential (capitalization rates at or below 4 percent) and office real estate. - Domestic equity market valuations are close to historical averages; ratio of stock prices to analyst-estimated forward earnings are well within historical norms. - Government bond yield curves are consistently flatter than major market peers, reflecting pension demand for long-term bonds. - Corporate bond spreads have trended toward historic lows in recent years. ### Foreign ownership and corporate bond issuance - Non-resident investors absorbed 85 percent of the net increase in Canadian corporate bond issuance since end-2012. - Nonresidents’ ownership share of Canadian corporate issuance increased from 32 to 48 percent as of 2018Q4. - Much of the issuance is likely denominated in foreign currency: such debt accounted for 91 percent of the net issuance between 2014 and end-2018. - Financial firms, largely the D-SIBs, account for nearly 60 percent of this issuance. - A large portion of financial sector issuance supports foreign-currency assets; data on currency mismatch is lacking. - Even assuming currency hedging, sustained depreciation of the exchange rate could decrease the attractiveness of foreign funding and lead to deleveraging over time. ### Investment fund sector growth and vulnerabilities - Investment funds (primarily open-ended mutual funds and exchange traded funds) have been the fastest growing segment of the Canadian financial system since 2013. - Fixed income assets accounted for roughly 28 percent of these funds’ underlying assets (including funds classified as mixed or balanced funds). - In the funds examined in the investment fund stress test: - Holdings of debt rated BBB and below rose from 11 to 23 percent of assets since 2008. - Cash fell from 7 to 3 percent of assets. - Portfolio duration has increased. - In a period of stress with large redemptions, forced liquidations of less liquid debt could be significant, increasing risk premia and potentially triggering further selling. ### Investment fund liquidity stress tests — overview and sample - The FSAP, in collaboration with the BoC, conducted stress tests of open-ended mutual funds with large corporate bond allocations using a single sample of over Can$323 billion in mutual fund assets, which held about 25 percent of the Canadian dollar-denominated corporate bond market as of 2018. - The stress test fund sample includes 238 Canadian domestic-currency corporate bond-focused funds. - Sample characteristics: - Market value of assets: Can$323 billion. - Sample funds represent about one-fifth of assets of the total investment fund sector (excluding fund of funds). - Sample funds account for about three-quarters of the sector’s domestic-currency corporate bond holdings, and about one quarter of all Canadian dollar-denominated corporate bond issuance outstanding. - Corporate bonds were 46 percent of AUM for the sample and 28 percent of AUM for mixed funds. - Government and other fixed income holdings accounted for 41 and 14 percent of AUM at bond and mixed funds, respectively. - 10 percent of fund sample holdings were bonds with a rating of BB or lower (high yield) and 13 percent were rated BBB. - Historical fund flow dataset: - Includes 54 funds which wound down before end-2018Q3. - 29,321 monthly observations in total; average of 97 observations per fund (about 8 years). - Excludes fund flow observations where implied fund return exceeds 20 percent or is below -20 percent. ### Stress test methodology - Three calibrations for redemption shocks: 1. First-percentile historical fund flows of the entire fund sample by mandate (bond or mixed), producing a single redemption shock per fund type. 2. First-percentile historical fund flow for each individual fund, producing a unique redemption shock per fund. 3. BOC calibration: assumes a 100 basis points parallel increase to the yield curve over one quarter, translated into an initial negative shock to bond values using fund-level duration as of 2018Q3; then uses empirically estimated performance-flow sensitivities to gauge redemptions. - The BOC approach is intended to test the impact of a large but plausible shock that exceeds anything in recent Canadian history. - The analysis incorporates bond credit rating information of bonds held by sample funds to identify which segments would be most affected by outflows. ### Stress test results - Aggregate outflows: - First calibration generated aggregate outflows of 7.2 percent (sample-wide rate of -7.1 percent using first calibration; -8.6 percent for bond funds; -5.6 percent for mixed funds). - Second calibration generated aggregate outflows of 4.3 percent (weighted average of the first percentile fund flow of all funds). - Forced selling under stressed conditions could reach the equivalent of 1 to 5 percent of the outstanding stock of corporate bonds, depending on calibration and liquidation preferences. - On a credit-rating basis: - Sample funds held 44 percent of the BBB and below corporate bond segment; forced selling would disproportionately affect BBB and below. - Sample funds hold 17 percent share of AA-rated bonds (largely provincial government bonds); forced sales could be equivalent to 3 percent of that market. - BOC estimate: forced corporate bond sales equivalent to 5 percent of the outstanding stock would generate 93 basis points in increased excess bond premium (measure of yield required in excess of CDS-implied credit spread). - These results could vary by 30-40 basis points depending on liquidation preference assumptions and opportunistic buying by other long-term investors. - Historical outflows context: - Largest monthly historical outflow within the sample was -1.6 percent in December 2008. - Largest outflow experience of U.S. bond funds dating back to 2000 was -3.2 percent. ### Policy recommendations - Strengthen capacity to monitor risks in the complex and fast-evolving nonbank financial sector: - Address data limitations in monitoring growth and risk profile of private consumer and real estate lenders. - Improve tracking of risk-taking and leverage in private NBFIs and funds they set up. - Strengthen data collection and increase surveillance of the nonbank sector, given growth in nonfinancial real estate firms’ debt and difficulty tracking private fund structures. - Enhance oversight and transparency of large pension funds given systemic relevance: - Increase the detail, standardization, and reporting frequency of pension fund financial disclosures to better monitor and assess risk-taking. - Implement standardized liquidity stress-testing techniques and assumptions to be used alongside in-house techniques to increase transparency and comparability around liquidity risk management. *Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1canea2020001.pdf* ### 190. The BOC framework generated a larger redemption shock at 18.0 percent. The median ### 1canea2020001 - 190. ### Redemption Shock Scenarios - The BOC framework generated a larger redemption shock at 18.0 percent. - The median fund-level redemption was 17.5 percent. - At the fund level, the shock (18.0 percent) is larger than the historical first-percentile largest monthly historical redemption for 97 percent of funds in the sample. - Comparison across calibrations: - The first calibration is larger than 69 percent of funds’ first-percentile redemptions. - The second calibration is larger than just half of first-percentile redemptions. - The third approach’s magnitude partly reflects an assumed 100 basis points interest rate shift. - This 100 basis points shift has occurred only once for 5-year Canadian government bonds since the sample began (2002Q2). - This shock is triple the standard deviation for 3-month changes to the 5-year yield over the last decade (33 basis point). ### Asset Sales: Approaches and Assumptions - Two approaches to calculate portfolio sales composition: - Pro-rata sales (vertical slicing) - Assets sold in proportion to the fund’s current asset allocation to maintain allocation structure. - Assumed default for most funds in the stress test because: - Mixed funds and indexed funds prefer to maintain asset allocation given mandates. - In market stress, fund managers may preserve liquid assets anticipating future redemptions. - Considered most appropriate for this analysis given severe market stress across scenarios. - Waterfall sales (horizontal slicing) - Fund managers use most liquid assets first, in order of liquidity characteristics. - Liquid asset hierarchy follows a specification loosely based on the Basel III LCR haircut schedule. - Assumed sale order: (i) cash; (ii) AAA and AA bonds; (iii) equities; (iv) other bonds in order A, BBB, BB, B, and so on; (v) other assets. - Provided for comparison purposes. ### Estimated Asset Sales Under Stress (Pro Rata and Waterfall) - Under pro rata sales and the three calibrations: - Forced selling of corporate bonds would reach 5 percent of the total stock outstanding (Figure 31 statement). - Sample funds would sell: - Can $11.9 billion in corporate bonds under the first calibration. - Can $6.5 billion in corporate bonds under the second calibration. - Can $29.7 billion in corporate bonds under the third calibration. - These sales are equivalent to roughly 2, 1, and 5 percent of domestic-currency corporate bonds for calibrations 1, 2, and 3, respectively. - Government bond sales under the third calibration are about 1.5 percent of the outstanding market. - Pro rata sales by credit rating implications: - Sample funds’ bond holdings are skewed towards lower-yielding bonds compared to the overall Canadian dollar-denominated bond market. - Relatively small high-yield market would see the largest impact because sample-reported speculative-grade holdings exceed the Bank of America Merrill Lynch Canadian dollar-denominated high yield corporate bond index. - Investment fund holdings by segment: - Funds hold roughly 17 percent of the AA-rated bond segment (about three quarters provincial government bonds). - Funds hold 23 percent of the BBB market (about three quarters nonfinancial corporate papers). - Under the third calibration, forced selling of AA- and BBB-rated bonds would be equivalent to three and five percent of those market segments, respectively. - Waterfall sales implications: - Would limit overall impact on corporate bond and lower-rated markets but concentrate selling in government-issued and higher-rated bonds. - Government bond sales under waterfall would more than double relative to pro rata assumptions. - Corporate bond sales would decline by a larger factor relative to pro rata assumptions. - Under the third calibration, AAA bond sales would rise from 1.5 percent of AAA bonds outstanding to 4 percent. ### Price and Liquidity Impact Estimates - The BOC bond market trading liquidity supply model estimates: - Investment fund corporate bond sales of Can $31.7 billion would increase corporate bond liquidity risk premiums by 93 basis points. - If long-term institutional investors did not absorb increased supply, estimated liquidity risk premiums would rise to around 140 basis points. - This is close to the 165 basis point premium observed in Canadian markets during the GFC. - Sensitivities that could amplify price impact: - Pace of redemptions (weekly concentration vs. monthly data). - Breadth and specific bonds sold (smaller and less frequently traded bonds require deeper discounts). - If fund flows occur within a week, redemption-triggered bond sales would be quite large relative to average volumes, magnifying downward price pressures. - Estimated implied corporate bond sales (pro rata) equal roughly one to three times the average weekly corporate bond trading volumes in 2018 (institutional, retail, and non-resident investors). ### Key Quantitative Results (from Table 11 and narrative) - Redemption shocks and sample statistics: - BOC framework shock: 18.0 percent. - Median fund-level redemption: 17.5 percent. - Corporate bond sales by calibration (pro rata): - Calibration 1: Can $11.9 billion. - Calibration 2: Can $6.5 billion. - Calibration 3: Can $29.7 billion. - Price impact model estimates: - Can $31.7 billion corporate bond sales → 93 basis points liquidity premium increase. - No long-term investor absorption → ~140 basis points liquidity premium increase. - GFC observed premium: 165 basis points. - Market-share and holdings highlights (from narrative): - Funds hold roughly 17 percent of AA-rated segment. - Funds hold 23 percent of BBB market. - High-yield holdings in sample exceed Bank of America Merrill Lynch Canadian dollar-denominated high yield corporate bond index. ### Policy Recommendations - Continue to strengthen risk surveillance for investment funds: - Regularly monitor funds’ bond portfolios on a fund-level or bond-level basis for systemic and market risk monitoring. - Provide a more granular and higher-frequency understanding of the distribution of credit and liquidity risk exposures within the investment fund sector. - Assess and enhance liquidity risk management: - More actively assess funds’ liquidity risk management capabilities, including for investment grade assets. - Consider requiring regular disclosures to supervisors or investors on relevant policies and assessments. - Encourage adoption of liquidity management tools: - Consider encouraging greater adoption of redemption gates, fees, and swing pricing to limit incentives for investors to withdraw money when they see other investors exiting. *Source: 1canea2020001 - 190. (extracted content)* ### 2. Channels of risk ### 2. Channels of risk ### Satellite models for macrofinancial linkages - Several empirical models are used, together with the Bayesian Model Average technique, to estimate scenario dependent point-in-time (PiT) PDs for other credit exposures. - Model outputs depend on bank-specific starting points. - For the cash-flow analysis, asset haircuts reflect two components: (i) shocks to interest rates and asset prices as captured in the macrofinancial scenarios; and (ii) additional haircuts required by counterparties to accept specific assets as collateral for secured funding transactions. - N/A for life insurance satellite models. ### Stress test horizon and scenario analysis - Stress test horizon: Three years (2018Q4–2021Q4) for credit/solvency and life insurance exercises. - LCR test horizon: 30 days. - Cash-flow analysis horizon: Normally 3 months. - Scenarios: - Baseline scenario: Based on October 2018 World Economic Outlook (WEO) projections. - Adverse scenario: Simulated using the Global Macrofinancial Model (GFM). Features: - Severe recession concurrent with significant financial market stress and a sharp housing market correction. - Main triggers: external developments, including global trade disruptions, tightening global financial conditions, and weaker-than-expected global economic activity, as described in the Risk Assessment Matrix (RAM). - Central feature: tighter-than-expected monetary policy by some major central banks in response to potential de-anchoring of inflation expectations induced by disruption in international trade and global production chains. - Disorderly monetary tightening leads to significant global financial market stress, global housing market and credit cycle downturns, amplified by existing macrofinancial vulnerabilities (housing market imbalances and high household debt), resulting in a sharp housing market correction and deterioration in bank asset quality. Significant financial stress would also materialize in Canada. - Under the adverse scenario for Canada: - The Canadian economy would encounter two years of output contraction (2018 and 2019), with cumulative real GDP growth of -2 percent during 2019–21, equivalent to 3 standard deviation. - Based on growth-at-risk analysis, its likelihood is 3.8 percent. - LCR test scenarios (liquidity): - (i) a run on retail deposits, with higher run-off rates for retail deposits; - (ii) a run on wholesale funding, with higher run-off rates for corporate deposits and other wholesale funding; - (iii) a combination of runs on retail deposits and wholesale funding. - Cash-flow analysis scenarios: a series ranging from mild to severely adverse liquidity conditions, considering both funding and market liquidity risks. - Life insurance adverse scenario: exercise considers two periods of the adverse scenario: 2019Q3 (most severe financial market stress) and 2021Q4 (lowest interest rates). ### Tail shocks and sensitivity analysis - Sensitivity exercises explored include: - Shocks to household affordability that critically affect PDs of mortgage exposures under an alternative assumption where higher-risk borrowers are asked for additional interest rate spreads to compensate banks for larger capital charges (amplifying stress on households’ affordability). - Assessment of the impact on accounting lifetime expected credit loss for Stage II mortgage exposures because of a shorter contractual lifetime vs a longer average amortization lifetime. - Assessment of the relative impact on losses due to the utilization of committed credit lines in an adverse scenario with no balance sheet restrictions (impact of large undrawn EADs). - Credit exposure concentration risks are assessed, taking into account market structure specificities. - For life insurers: alternative assumptions regarding changes in risk-free interest rates that would generate a more material impact are examined. - Sensitivity analysis: N/A for some liquidity LCR testing. ### Risks, buffers, and behavioral adjustments (credit, market, interest rate, liquidity) - Credit risk: - Captures all drawn and undrawn balances associated with on-balance sheet loan portfolios and exposures at amortization cost. - On-balance sheet credit EADs evolve broadly based on credit growth assumption in scenarios. - Off-balance sheet credit EADs associated with undrawn balances are assumed to grow in proportion with drawn balances such that utilization remains fixed; all facilities assumed contractually irrevocable to extend funds in the future. - For the FSAP team top-down: Off-balance sheet credit EADs are assumed to evolve in alignment with credit growth assumption in scenarios. - Market risk: - Reflected in valuation effects of FVTPL and FVOCI positions, as well as net open financial positions (i.e., currencies, equities, and commodities). - For life insurers: market risk and credit risk affect the valuation of investment portfolios. - Net interest income: - Affected by the margin rate implied by asset-side and liability-side interest rates (BOC top-down); FSAP team top-down specifies change in the reference rate and margin rate implied by asset- and liability-side interest rates. - Liquidity risk: - Funding liquidity reflected in funding run-off rates and asset roll-over rates. - Market liquidity reflected in asset haircuts, which could be influenced by market movements, fire sales, and collateral supply constraints. - Liquidity from the Bank of Canada’s Emergency Lending Assistance is not considered. - Behavioral adjustments and contagion: - Framework features a contagion module based on the MacroFinancial Risk Assessment Framework (MFRAF) to analyze contagion effects, including interaction between solvency and liquidity conditions. - Dividends: - Equal to the greater of (i) the most recent historical dividend (2018Q4) or (ii) the dividend implied by a fixed ratio of dividends to post-tax net income. - Dividends are not reduced unless there is a breach of the capital conservation buffer. Should this occur, the maximum dividend payout is subject to regulatory restrictions. - Dividends anchored to the previous fiscal year’s level (before the cut-off date) are assumed to be paid by banks. When the capital conservation buffer is breached, restrictions on dividend distributions are aligned with the regulatory framework. - EADs linked to securities holdings remain constant in solvency modules. - If relevant, maturing assets are replaced by exposures of the same type of risk. - Maturing capital instruments are generally not allowed to be renewed in the adverse scenario; some exceptions may be granted for certain capital instruments issued in foreign currency by overseas subsidiaries. - If a bank’s capital falls below regulatory requirements, no prompt corrective action is assumed in the BOC top-down setup. ### Calibration of risk parameters and accounting/regulatory standards - Scenario dependent forward paths for PiT PDs and LGDs are estimated for each asset class and geography based on: - Historical non-performing loan ratios (for the corporate sector). - A structural model based on microsimulations using the HRAM (for the household sector). - Given limited availability of PiT LGD data, some proxies are used (described as "very basic proxies" in places). - For internal ratings-based (IRB) exposures: - Risk-weighted assets are projected on the basis of updated regulatory through-the-cycle (TTC) PDs and downturn LGDs, using appropriate scaling multipliers from the PiT parameters. - For standardized approach (STA) exposures: - Risk-weighted assets are projected based on constant risk weight densities. - IFRS 9: - Expected credit losses for mortgage exposures are projected by estimating stage transition matrices based on projected PiT PDs and stage transition rates. - Estimations of lifetime expected losses assume a transition to baseline parameters after 5 years. - Life insurance calibration: - Risk-free interest rates and asset prices (credit spreads, equity prices, and house prices) calibrated in line with the macrofinancial scenario. - Yield curves interpolated based on relevant short-term and long-term government bond yields. - Macrofinancial shocks mainly affect available capital and base solvency buffer. - Regulatory/accounting and market-based standards: - Basel III LCR monitoring tool used; LCR hurdle set at 100 percent. - Net cash balance hurdle for cash-flow analysis: to pass, a non-negative net cash balance is required. - In baseline capital hurdles include regulatory minimum, capital conservation buffer, the D-SIFI surcharge, and the applicable countercyclical capital buffer and/or domestic stability buffer. - In adverse scenario, hurdles include the regulatory minimum and the D-SIFI surcharge. - Hurdle rates based on common equity tier-1, tier-1, and total capital ratios. - Life insurance regulatory framework: Canada’s Life Insurance Capital Adequacy Test (LICAT). - Hurdle rates based on regulatory minimums for LICAT’s core and total capital ratios at 50 percent and 90 percent, respectively. - Total capital ratio = (available capital including both tier-1 and tier-2 capital, surplus allowance and eligible deposits) / base solvency buffer. - Core capital ratio = (tier-1 capital + surplus allowance [only 70 percent] + eligible deposits [only 70 percent]) / base solvency buffer. ### Reporting format and outputs - System-wide evolution of common equity tier-1, tier-1 and total capital ratios. - Distribution of bank capital positions and life insurers’ capital positions. - Contribution to key drivers to system-wide net income and capital positions, including differences between baseline scenarios and adverse scenarios. - Number of institutions with capital below the hurdles, and the share of their assets. - Amount of capital shortfalls. - For liquidity: changes in the system-wide liquidity position, including important drivers for cash outflows, cash inflows, and counterbalancing capacity. - Distribution of bank liquidity positions. - Number of institutions with LCR below 100 percent and/or negative net cash balance. - Amount of liquidity shortfalls, including by currencies. - For life insurers: contribution to key drivers to system-wide capital position and number/asset share of institutions below LICAT hurdles. *Source: 1canea2020001 - 2. Channels of risk* ### 1. Institutional ### 1. Institutional ### Institutional perimeter — Mortgage insurers - Institutions included: - All three mortgage insurers, all federally regulated - Market share: - 100 percent of mortgage insurers’ total assets - Data and baseline date: - OSFI: regulatory returns - CMHC: data equivalent to OSFI regulatory returns - Data as of September 2018 - Scope of financial consolidation: mortgage insurance business ### Channels of risk propagation — Mortgage insurers - Methodology: - Balance sheet approach - Projections of key balance sheet, income statement and capital account items - Passive balance sheet assumption - Premium earnings are projected in line with overall insured mortgage credit growth (based on the scenario) and adjusted to reflect institutions’ behavior under stress. - Claims are projected to mirror banks’ potential losses on mortgage exposures. A cross-check with estimated expected credit losses incurred by banks would be made. - Other income and expense components are projected based on their sensitivity to macrofinancial conditions. - The mark-to-market approach is used to assess the impact of macrofinancial shocks on investment portfolios, which would in turn affect available capital. - Adjustments to minimum required capital are made to reflect changes in credit risk, insurance risk, market risk and operational risk. - Insurance risk is adjusted based on LTV that reflects updated house prices (as sensitivity analysis). - Satellite models for macrofinancial linkages: - Empirical models are used to project key components of non-premium income and expenses. - Stress test horizon: - Three years (2018Q4–2021Q4) ### Tail shocks / Scenario analysis — Mortgage insurers - Scenario design: - Based on two common macrofinancial scenarios - The scenarios specify key macrofinancial variables (e.g., real GDP growth, inflation rate, unemployment rates, exchange rates, equity prices, house prices, interest rates, and credit growth) for Canada and important geographies/countries, as well as global variables (e.g., commodity prices). - The baseline scenario is based on October 2018 WEO projections. - Adverse scenario features: - Simulated using the GFM. - Severe recession concurrent with significant financial market stress and a sharp housing market correction. - Main triggers: external developments including global trade disruptions, tightening global financial conditions and weaker-than-expected global economic activity, as described in the RAM. - Central feature: tighter-than-expected monetary policy by some major central banks in response to potential de-anchoring of inflation expectations, induced by disruption in international trade and global production chains. - Disorderly monetary tightening leads to significant global financial market stress, setting off global housing market and credit cycle downturns, amplified by housing market imbalances and high household debt. - Significant financial stress also materializes in Canada. - Quantified adverse impact: - Under the adverse scenario, the Canadian economy would encounter two years of output contraction (2018 and 2019), with cumulative real GDP growth of -2 percent during 2019–21, equivalent to 3 standard deviation. - Based on growth-at-risk analysis, its likelihood is 3.8 percent. ### Risks and buffer — Mortgage insurers - Risks/factors assessed: - Credit and market risks affect the valuation of investment portfolios. - Insurance risk is captured in two aspects—claim payouts and minimum required capital. - New mortgage insurance business evolves in line with overall insured mortgage credit growth (based on the scenario) and institutions’ behavior under stress (e.g., during the global financial crisis). - Behavioral adjustments: - Net income, through the impact on equity, determines the overall size of investment portfolios. The asset allocation in investment portfolios remains unchanged. - If relevant, maturing assets are replaced by exposures of the same type of risk. - If mortgage insurers’ capital falls below regulatory requirements, no prompt corrective action is assumed. - Dividends are paid if mortgage insurers generate positive post-tax net income and do not need additional capital. ### Regulatory and market-based standards and parameters — Mortgage insurers - Calibration of risk parameters: - Claim payouts are projected based on expected credit losses for mortgage exposures under the bank solvency stress tests. - Changes in insurance risk, which would affect minimum required capital, are estimated to reflect changes in house prices and thus changes in LTV. - The impact of relevant asset prices on investment portfolios in terms of credit and market risks (i.e., credit spreads, interest rates, equity prices) are assessed according to the macrofinancial scenarios. - Regulatory/accounting and market-based standards: - Regulatory capital framework based on Canada’s Mortgage Insurance Capital Adequacy Test (MICAT) - Hurdle rates based on the supervisory target for the MICAT’s capital ratio at 150 percent. - The MICAT’s capital ratio is based on available capital divided by two-thirds of minimum required capital. ### Reporting format for results — Mortgage insurers - Output presentation: - System-wide evolution of MICAT ratio. - Distribution of mortgage insurers’ capital positions. - Contribution to key drivers to system-wide net income and capital position, including differences between the baseline scenario and the adverse scenario. - Number of institutions with capital below the hurdles, and the share of their assets. - Amount of capital shortfalls. --- ### Investment Fund Stress Testing — Institutional perimeter - Institutions included: - Open-ended Canada-domiciled mutual funds with lifetime average allocations to Canadian dollar-denominated corporate bond above 20 percent and lifetime average assets under management of at least Can$50 million - Only funds with fixed income and balanced mandates - Sample funds held Can$323 billion in assets under management as of 2018Q3 - Market share: - 19 percent of total assets of mutual funds - Sample funds held 77 percent of corporate bonds (including both financial and nonfinancial issuers) held by mutual funds, 25 percent of outstanding Canadian corporate bonds, and 8 percent of outstanding Canadian government bonds (including subnational issuers). - Data and baseline date: - Morningstar: monthly fund-level assets (including the allocation) and fund flows (defined as net sales or redemptions of mutual fund shares) - Fund flow data since January 2002 - Asset holding data as of September 2018 ### Channels of risk propagation — Investment funds - Methodology: - The exercise quantifies the amount of asset sales (particularly, corporate bonds) following redemption shocks. - Fund-level redemptions are calibrated by (i) assuming a duration-driven shock to fund performance based on a shift in interest rates; and (ii) estimating investors’ net redemptions in response to that shock, based on an estimated sensitivity of fund flows to fund-level performance. - The exercise also quantifies the price impact on corporate bonds due to the forced sales. - Fund-level redemptions are also calibrated based on historical fund-flows experience. - Satellite models for macrofinancial linkages: - Each fund's sensitivity is estimated based on the observed relationship between fund flows and a CAPM model-derived measure of "alpha," or fund outperformance relative to a benchmark. - The resulting price impact on corporate bonds is estimated as a function of the balance sheet constraints and market conditions faced by broker-dealers and leveraged investors. - Stress test horizon: - N/A ### Tail shocks / Scenario analysis — Investment funds - Scenarios: - A parallel increase in the yield curve by 100 basis points - Two exogenous redemption shocks are calibrated at the fund-level using the first percentile of the historical distribution of monthly fund flows. - The first shock is based on the distribution combining all funds of a given investment mandate (i.e., fixed-income and balanced). - The second shock is based on the distribution of each individual fund. - Sensitivity analysis: - N/A ### Risks and buffers — Investment funds - Risks/factors assessed: - Market risk affects performance of mutual funds, which in turn creates liquidity risk that entails bond sales. Then, performance of mutual funds is affected by resulting bond sales. - Behavioral adjustments: - Sales of corporate bonds and other assets resulting from redemptions are calculated through an assumption that fund managers liquidate fund assets on a pro-rata basis as well as a waterfall basis (i.e., based on the liquidity hierarchy). ### Calibration and reporting — Investment funds - Calibration of risk parameters: - Fund-level redemptions are calibrated based on the fund-level sensitivity of net fund flows to fund performance. - Regulatory/accounting and market-based standards: - N/A - Output presentation: - Amount of forced corporate bond sales, relative to outstanding amounts of bonds and trading volumes by bond types - Resulting price impact on corporate bonds --- ### Appendix III — Credit Risk Satellite Models (overview) - Purpose: - Econometric models estimated to produce scenario dependent forward paths for point-in-time (PiT) probabilities of default (PDs) for different exposure segments. - Data and calibration challenges: - Low default rates in the domestic market make obtaining reliable historical time series challenging due to potential outliers in impairment flows. - Corporate segments modeling: - Historical default rates from Moody’s Analytics Creditedge were used to obtain historical series for all segments considered. - Separate models estimated for Canada and the United States; U.S. models also used as a proxy for the ROW segments. - Mapping using weighted combinations of provider segments was used where necessary. - In total, models for 10 industrial segments were estimated for Canada and 15 industrial segments for the U.S. After applying the necessary segment mapping between basic and composite segments, forward path estimates for the 11 active corporate industrial segments were obtained for all geographies. - Methodology: - A Bayesian Model Averaging (BMA) econometric technique (Bayesian Averaging of Classical Estimates, BACE) was employed to model and project default rates at individual geography and portfolio levels. - BMA operates with a pool of equations (several hundreds or thousands) per dependent variable, assigning weights that reflect relative predictive performance, producing a “posterior model” equation. - Pool of equations considers all possible combinations of predictors from a pool including real GDP, investment, consumption, exports, price inflation, short- and long-term interest rates, among others. - PD transformation and forecasting: - Logit transformation applied to original PD to ensure predictions between 0 and 1: - Y_it = logit(CD_it / (1 − CD_it)) - Logit-transformed PDs modeled as a linear function of exogenous macroeconomic and financial regressors, with autoregressive lags and lags of explanatory variables included. - Conditional PD PiT forecasts for each segment were attached to the starting point of bank PDs in logistic space to generate full path under both the baseline and adverse scenarios. - Explanatory variables and sample: - Variables used include real GDP, investment, inflation, unemployment rate, short-term and long-term interest rate spread, equity prices, private credit by sectors; housing price included for real estate and construction PD models. - Inputs other than interest rates are subject to quarterly growth transformation; interest rates and unemployment rate taken as percentage point changes between periods. - Sample period for calibration ranges between 2000 and 2018, with quarterly frequency. - Model selection criteria: - Five information criteria used: R-square, Durbin Watson statistic, number of significant variables with high posterior inclusion probability, quality of in-sample forecast, and size of impact in forecasting period. - Ideal candidate: relatively high R-square, Durbin Watson between 1.5 to 2.5, small root-mean-square-error, historically consistent size of impact under stress. - Results summary: - Scenario dependent PD projections broadly in line with historical stress episodes. - Forward-looking PD paths: size of impact for Canada in line with stress experienced during the GFC; for the U.S. consistent with both the GFC and the Dot-com bubble. - Salient idiosyncrasies among segments: highest PDs under stress assigned to energy related sectors (mining, etc.) and lowest PDs to communication, transportation, and other utilities. *Source: 1canea2020001 - 1. Institutional* ### 8. Investment, unemployment rate, equity returns, and credit-related measures play ### 8. Investment, unemployment rate, equity returns, and credit-related measures play significant roles in determining underlying credit risks ### Key empirical findings on PD drivers - Investment, unemployment rate, equity returns, and credit-related measures have high posterior inclusion probability and sizable normalized long-run multipliers for corporate segment PDs across Canada and the United States. - Long-term and short-term interest rate spreads weigh more heavily in the determination of U.S.-based corporate PDs. - Volatility in housing prices substantially influences financial resilience in real-estate and construction segments. - For corporate segments, normalized R square values reported include: 0.64, 0.64, 0.56, 0.93, and 0.90 in Table III.1 (Canada). - For U.S. corporate segments, normalized R square values reported include: 0.89, 0.88, 0.63, 0.85, 0.84, 0.66, and 0.60 in Table III.2 (United States). ### Canada: corporate PD PiT model highlights (from Table III.1 and Figure III.1) - Variables showing notable normalized long-run multipliers (stars denote higher posterior inclusion probability): - Real investment growth, yoy: contains a -0.38* entry in one portfolio. - Unemployment rate, percentage point change, qoq: entries include 0.18* and 0.02 in some portfolios. - Equity indices, percentage change, qoq and yoy: multiple negative multipliers (e.g., -0.2*, -0.36*, -0.19*). - Credit to household and corporate, percentage change, yoy and qoq: several negative multipliers (e.g., credit to corporate, yoy includes -0.59*, -0.57*). - Housing price, percentage change, yoy: -0.01 in one portfolio. - Model diagnostics (selected): Durbin Watson reported as 0.82, 0.91, 0.71, 1.82, 1.56 across portfolios; number of lags of independent variables reported as 2 in many equations; number of observations indicated as 6 (repeated formatting in table). ### United States: corporate PD PiT model highlights (from Table III.2 and Figure III.2) - Variables with notable normalized long-run multipliers (stars denote higher posterior inclusion probability): - Real investment growth, yoy: e.g., -0.45* and -0.29* in some portfolios. - Unemployment rate, percent and percentage point change (qoq and yoy): significant positive multipliers in several portfolios (e.g., unemployment rate, percentage point change, yoy includes 0.58* in one portfolio). - Long term sovereign yield spread against U.S. swap, percent: positive multipliers in several portfolios (e.g., 0.24*, 0.55*, 0.66*). - Short term interest rate spread against policy rate, percent: positive multipliers in some portfolios (e.g., 0.13*, 0.02*). - Equity indices, percentage change, qoq and yoy: large negative multipliers in many portfolios (e.g., -0.48, -1.08, -0.67*). - Housing price, percentage change, yoy: -0.3* in one portfolio. - Model diagnostics (selected): Durbin Watson values reported include 0.64, 2.05, 0.55, 1.14, 0.79, 0.30, 0.36 across portfolios; number of lags of independent variables reported as 2; number of observations indicated as 6 (repeated formatting in table). ### Consumer loans: data constraints and modeling approach - Historical regulatory data for consumer loans PD PiTs are very limited; extremely low default/delinquency rates make impairment-flow deduced PDs unreliable and subject to distortion from one-off events. - A consumer model proxy was developed using Canadian Bankers Association public time series on credit card delinquencies and loss rates. - LGD proxy: a LGD of 90 percent was used to back out historical PDs for the credit-card-based proxy. - To preserve degrees of freedom, the consumer proxy model used only minimum and most relevant explanatory variables: principally the unemployment rate and short-term interest rate spreads. - The credit-card model output was checked for consistency with historically observed default rates. - The credit-card model served as a proxy for the broader consumer credit segment by applying a 60 percent scaling factor on the PD PiT forward path to reflect lower sensitivity of the broader consumer segment to economic downturns. - The same PD PiT forward paths were applied across banks for each consumer exposure segment, assuming no major cross-bank differences in credit quality characteristics. ### Consumer model OLS estimates (Table III.3) — Credit cards (dependent variable: year-on-year change in the HH NPL ratio) - Unemployment rate, percentage point change, yoy: coefficient = 0.124*** (t-statistic 3.850) - Short term interest spread, percent: coefficient = 0.092 (t-statistic 0.390) - Constant: -3.218*** (t-statistic -43.80) - R-square = 0.254 - # of observations = 58 - Interpretation: unemployment rate is the most statistically significant explanatory variable for consumer PD movement. ### Scenario comparison for credit-card PDs - Adverse scenario peak forward-looking PD path: 6.3 percent. - Global Financial Crisis (GFC) benchmark forward-looking PD path: 5.3 percent. - Under the adverse scenario, the sharper increase in both unemployment rate and short-term spread generates a more pronounced PD path than during the GFC. ### LGD PiT assumptions and projection methodology - For corporate and consumer segments, PiT LGDs were proxied because historical LGD time series were not available. - Baseline scenario: PiT LGD assumed to be 80 percent of the downturn LGD for the entire scenario horizon. - Adverse scenario: PiT LGD assumed to be 10 percent higher than the downturn LGD, capped at 95 percent. - Mortgage-related segments: a structural model projected a separate LGD PiT path for each scenario. ### Structural model for mortgage defaults (Appendix IV) - Rationale: - Historical mortgage default time series in Canada are heavily skewed toward zero because Canada has not experienced a severe housing market crisis; rising house prices have mitigated observed defaults historically. - Pure econometric models may fail to capture the non-linear nature of mortgage defaults under severe adverse shocks. - Model choice and features: - A structural model (based on the TUI model) was preferred to capture household affordability and behavioral responses to scenario shocks (mortgage interest rates, unemployment rates, house prices, inflation). - The FSAP implementation used household microdata from the National Household Survey (2016) to infer joint distributions of current DSR ratio, current LTV ratio, and current Household Liquid Wealth, and to flag Mortgage Insurance status. - The structural distress probability function (Eq.1) links the probability of household stress to DSR post-stress (DPRt), change in DSR vs cut-off date (ΔDPRt), the unemployment rate (ut), and change in unemployment rate from the cut-off date (Δut), with multiple parameters and exponent terms to capture non-linearity. - The model was enhanced with behavioral reaction elements calibrated from microdata and accounts for Canada-specific two-layer mortgage insurance structure. - Model outputs and interpretation: - The pre-shock distribution is concentrated on safe partitions, but the adverse scenario shock reveals a challenging post-stress distribution with heavily penalized risky partitions and increased projected RWA density. - The DSR distribution has a significant tail above 0.4, associated with lower wealth buffers and higher stress probability for those borrowers. *Source: IMF staff calculations (Canada FSAP chapter).* ### 5. Additional conditions need to be met in order to assume that a mortgage exposure ### 1canea2020001 - 5. Additional conditions need to be met in order to assume that a mortgage exposure reaches default ### Default conditions and probability model - Default is a two-stage outcome: - Stage 1: Household is in distress (post first stage behavioral/stress event). - Stage 2: Default occurs only when: - the household’s Liquid Wealth is not enough to cover servicing needs, and - the value of the loan is higher than the value of the collateral (negative equity condition is a prerequisite for default). - Default occurs only if the post-stress LTV is higher than 1 and if any wealth buffers are not enough to cover servicing needs after a distress event. - Under a positive house price assumption the model captures masking of defaults due to price appreciation; in that positive house price scenario an outright sale would be triggered by distress as opposed to a default event. - Default probability is given by formula (Eq.2), where: - the first bracketed term denotes the probability that the property value after stress minus some liquidation discount is lower than the outstanding loan notional, and - the second term denotes that the stochastic behavioral rule Bt(x)—which accounts for the use of Liquid Wealth OL0 at the cut-off—has failed to save the loan from default. - Behavioral-rule calibration example used in the solvency ST: a linear survival rule where: - the survival probability is linear between a wealth buffer of 8 and 36 months, - buffers below 8 months will not be sufficient to weather a default event, - borrowers with wealth buffers exceeding 36 months would survive the distress event with probability 1. ### Conditional LGD and collateral realization (Eq.3) - The conditional LGD is driven by the discounted sale price of the house. - Sale occurs at time t+s (s = average time to realize the collateral). - Sale proceeds are net of transaction costs (discounted at a rate reflecting the scenario interest rate premium) and assume an additional foreclosure liquidation discount 훿. - LGD formula reference: OLLD_t = 1 − (1 − 훿) P_{t+s} / (L* (1 + i_t + c_s_t)^s) (Eq.3). ### Partitioning, Monte Carlo implementation, and joint distribution - The joint distribution is partitioned along DSR and LTV dimensions. - LTV partitioning values used: 1, 0.95, 0.9, 0.85, 0.8, 0.75, and 0.65. - DSR partitioning values used: 0.15, 0.2, 0.3, and 0.4. - Mapping yields a 5 by 8 partitioned space for the first two dimensions. - For each DSR/LTV partition a Monte Carlo simulation (on house price changes anchored to a central house price shock) produces projected average PDs and LGDs. - Portfolio average PD (or LGD) = weighted average of projected PDs (LGDs) per DSR/LTV density partition. - The fourth joint-distribution dimension (insured flag) is used to produce separate projections for insured and uninsured portfolios. - Household Survey distribution is adjusted to reflect the aggregate current LTV distribution as reported by banks; in practice this shifts parts of the density towards higher LTV buckets. ### Scenario-mode: Expected Loss projections (Table IV.1 results) - Model core horizon: three-year horizon; parameters calibrated to a three-year horizon and produce PDs, loss rates, and LGDs at end of three years. - Year-by-year approximation: start from year 1, apply each end-year macro shock and back out one-year PD and LGD values after accounting for already calculated years. Table IV.1: Canada: Mortgage Default-Scenario Paths for Structural Model (In percent) - Aggregate - Baseline: PD Y1 0.3; PD Y2 0.3; PD Y3 0.2; LGD Y1 24.1; LGD Y2 19.1; LGD Y3 13.9; EL Y1 0.1; EL Y2 0.0; EL Y3 0.0 - Adverse: PD Y1 1.4; PD Y2 2.5; PD Y3 2.8; LGD Y1 27.4; LGD Y2 47.6; LGD Y3 41.1; EL Y1 0.4; EL Y2 1.2; EL Y3 1.1 - Insured only - Baseline: PD Y1 0.6; PD Y2 0.5; PD Y3 0.4; LGD Y1 25.0; LGD Y2 20.9; LGD Y3 14.4; EL Y1 0.1; EL Y2 0.1; EL Y3 0.1 - Adverse: PD Y1 2.1; PD Y2 3.1; PD Y3 3.3; LGD Y1 30.0; LGD Y2 51.5; LGD Y3 45.5; EL Y1 0.6; EL Y2 1.6; EL Y3 1.5 - Uninsured only - Baseline: PD Y1 0.1; PD Y2 0.1; PD Y3 0.1; LGD Y1 19.4; LGD Y2 8.9; LGD Y3 8.2; EL Y1 0.0; EL Y2 0.0; EL Y3 0.0 - Adverse: PD Y1 0.9; PD Y2 2.1; PD Y3 2.4; LGD Y1 22.8; LGD Y2 43.1; LGD Y3 36.5; EL Y1 0.2; EL Y2 0.9; EL Y3 0.9 ### Unexpected loss mode and RWA density estimation - Unexpected losses / RWA densities are estimated by running a full range of macro scenarios; a Basel 99.9 quantile equivalent of the loss distribution corresponds to unexpected loss threshold / capital requirement. - Capital charges can be produced based on the joint distribution of DSR, LTV, Liquid Wealth, and insurance flag. - Model used as challenger for capital charges and RWA densities for banks and mortgage insurers. - Simulation exercise: 10,000 scenarios simulated; the loss value at the 99.9 quantile of the loss distribution defines unexpected losses. - Joint normal distribution assumptions for the three shocks used in the capital-projecting mode: - mean shocks: interest rates mean 1.1 percent; unemployment marginal change mean 0.1 percent; house price change mean 4 percent. - standard deviations: interest rates 1; unemployment 0.5; house prices 15 percent. - correlations: interest rates and house price shocks correlation -0.3; interest rates and unemployment correlation 0.3. Table IV.2: Canada: RWA Densities for DSR and LTV Partitions (In percent) - Columns (LTV buckets): LTV>1, LVT>0.95, LTV>0.9, LVT>0.85, LTV>0.8, LVT>0.75, LVT>0.65, LVT<0.65 - Rows (DSR buckets) with corresponding RWA densities: - DSR>0.4: 310.8, 240.9, 219.0, 201.4, 182.7, 154.9, 126.2, 34.7 - DSR>0.3: 38.9, 30.1, 27.1, 24.9, 22.5, 19.6, 15.3, 4.0 - DSR>0.2: 17.2, 13.1, 12.0, 10.6, 9.5, 8.2, 6.5, 1.6 - DSR>0.15: 8.1, 6.5, 6.0, 5.2, 4.6, 4.0, 3.1, 0.7 - DSR<0.15: 3.2, 2.5, 2.3, 1.9, 1.7, 1.5, 1.1, 0.2 Table IV.3: Canada: Model-projected RWA Densities: Pre- and Post-Stress (In percent) - RWA density pre-stress / RWA density post-stress: - Insured: 24.0 / 53.8 - Non-insured: 14.0 / 35.8 - Aggregate: 18.3 / 43.4 ### Bank-specific and regional extensions - Bank-specific starting points: use bank-specific DSR/LTV distribution to project scenario-specific forward PD and loss-rate paths; bank-specific weights across the 5 by 8 DSR/LTV partitions produce bank-weighted PD and loss-rate projections. - Regional/provincial extension (exploratory): incorporate a fifth variable for regional concentration to generate provincial DSR/LTV distributions; combine with regulatory reporting on regional concentration and region-specific house price shocks (HaR or similar methodologies) to produce regional projections. Note: this exploratory work was not incorporated in the solvency ST exercise. ### Model assessment and key summary metrics - Core horizon: three-year horizon; model targets probabilities of default, loss rates, and LGDs at end of three years. - Under the adverse scenario, cumulative three-year PDs for the entire market reach 6.7 percent. - Cumulative three-year PD decomposition: insured market 8.5 percent; uninsured mortgages 5.3 percent. - Under the baseline scenario, cumulative three-year PD projected at 0.8 percent. - Model strengths: - Captures masking of historical defaults due to rising house prices. - Produces intuitively reliable estimates on loss-rate projections and attributing driving factors. - Versatile: integrates a range of behavioral rules and adjustable granularity. - Multiple use modes: scenario-specific expected-loss mode, unexpected-loss / capital-charge mode, bank-specific starting-point mode. - Computationally efficient for sensitivity analysis and useful where historical distress episodes are rare or to complement classical econometric calibrations. *Source: 1canea2020001 - 5. Additional conditions need to be met in order to assume that a mortgage exposure reaches default* ### Appendix VI. Non-Interest Income Business Activity Segments ### 1canea2020001 - Appendix VI. Non-Interest Income Business Activity Segments ### Non-interest Income Segment Scaling Factors (Adverse Scenario) - Table of scaling factors (In percent) by segment and year - Non-interest Non-trading Income (Market Sensitive) - S1 Investment management and custodial fees: 15 (2019), 10 (2020), 10 (2021) - S2 Mutual fund fees: 25 (2019), 15 (2020), 10 (2021) - S3 Underwriting and advisory fees: 35 (2019), 20 (2020), 10 (2021) - S4 Securities commissions and fees: 20 (2019), 15 (2020), 10 (2021) - Non-interest Non-trading Income (Non-market Sensitive) - S5 Deposit account and other payment service fees: 5 (2019), 5 (2020), 5 (2021) - S6 Debit and credit card fees: 5 (2019), 5 (2020), 5 (2021) - S7 Loan and acceptance fees: 5 (2019), 5 (2020), 5 (2021) - S8 Net insurance revenues: 10 (2019), 10 (2020), 10 (2021) - S9 Other non-interest income: 5 (2019), 5 (2020), 5 (2021) ### Projected Non-Interest Income under the Adverse Scenario — Formula and Definitions - Projected income for segment s and scenario year y_i (notation from source): - I_s^{y_i} = min( I_s^{2018} − σ_s^{2013−2018}, I_s^{2018} * (1 − e_s^{y_i}) * (A_s^{y_i} / A_s^{2018}) * f_l^{y_i} ) - Definitions (as given): - I_s^{2018} denotes the income for segment s for year 2018 - σ_s^{2013−2018} is the standard deviation of income for segment s for the years 2013 to 2018 - e_s^{y_i} is the scaling factor of the table above for segment s for the year C_i (scenario year) - A_s^{y_i} denotes the total assets size of the bank for segment s for the year C_i (scenario year) - f_l^{y_i} denotes the proportion of income—adjusted for currency appreciation—in currency l - Baseline scenario assumptions: - Segment income assumed constant and equal to the segment income during 2018 adjusted only for currency appreciation - Operational expenses assumed to remain constant at the 2018 levels but adjusted to account for the scenario projected currency appreciation *CANADA INTERNATIONAL MONETARY FUND 141* ### Appendix VII. LCR Stress Test Scenarios ### Overview of Scenarios - Scenario ST Scenarios: - 1 - Regulatory - 2 - ST Retail - 3 - ST Wholesale - 4 - ST Combined ### 1. Stock of High-quality Liquid Assets (HQLA) - 1.1. Level 1 Assets — Weight (all scenarios) - Coins and banknotes: 1.00 - Total central bank reserves, of which part drawable in stress: 1.00 - Securities with a 0 percent risk weight (issued/guaranteed by sovereigns, central banks, PSEs, BIS/IMF/ECB/European Community, or MDBs): 1.00 - Sovereign or central bank debt securities issued in domestic currencies by the sovereign or central bank in the country in which the liquidity risk is being taken or in the bank’s home country: 1.00 - Domestic sovereign or central bank debt securities issued in foreign currencies, up to the amount of the bank’s stressed net cash outflows in that specific foreign currency: 1.00 - 1.2. Level 2A Assets — Weight (all scenarios) - Securities with a 20 percent risk weight (issued/guaranteed by sovereigns, central banks, PSEs, MDBs; non-financial corporate bonds rated AA- or better; covered bonds rated AA- or better): 0.85 - Adjusted amount of Level 2A assets: 0.85 - 1.3. Level 2B Assets — Weights (all scenarios) - RMBS rated AA or better: 0.75 - Non-financial corporate bonds rated BBB- to A+: 0.50 - Non-financial common equity shares: 0.50 - Sovereign or central bank debt securities rated BBB- to BBB+: 0.50 - Adjusted amount of Level 2B RMBS assets: 0.75 - Adjusted amount of Level 2B non-RMBS assets: 0.50 ### 2. Net Cash Outflows - 2.1. Cash Outflows - 2.1.1. Retail Deposit Run-off — Weights by scenario (columns correspond to scenarios 1–4) - are in Canada (eligible for a 3 percent run-off rate, transactional accounts): 0.03, 0.05, 0.10, 0.10 - are not in Canada (eligible for a 3 percent run-off rate, transactional accounts): 0.03, 0.05, 0.10, 0.10 - eligible for a 5 percent run-off rate (transactional accounts): 0.05, 0.05, 0.10, 0.10 - in non-transactional accounts with established relationships eligible for a 3 percent run-off rate, are in Canada: 0.03, 0.05, 0.05, 0.05 - in non-transactional accounts with established relationships eligible for a 3 percent run-off rate, are not in Canada: 0.03, 0.05, 0.05, 0.05 - eligible for a 5 percent run-off rate (non-transactional relationship accounts): 0.05, 0.05, 0.05, 0.05 - in non-transactional and non-relationship accounts: 0.10, 0.10, 0.10, 0.10 - Uninsured deposits: 0.10, 0.15, 0.15, 0.15 - Sourced from an unaffiliated third-party: 0.10, 0.15, 0.15, 0.15 - Foreign currency: 0.10, 0.15, 0.15, 0.15 - Term deposits with a remaining maturity of > 30 days: 0.00, 0.00, 0.00, 0.00 - 2.1.2. Unsecured Wholesale Funding Run-off — Selected weights by scenario - eligible transactional accounts (are in Canada): 0.03, 0.08, 0.08, 0.08 - eligible transactional accounts (are not in Canada): 0.03, 0.08, 0.08, 0.08 - eligible for a 5 percent run-off rate: 0.05, 0.08, 0.08, 0.08 - in non-transactional and non-relationship accounts: 0.10, 0.10, 0.10, 0.10 - Uninsured deposits (wholesale): 0.10, 0.10, 0.25, 0.25 - Sourced from an unaffiliated third-party (wholesale): 0.10, 0.10, 0.25, 0.25 - Foreign currency (wholesale): 0.10, 0.10, 0.25, 0.25 - Term deposits with a remaining maturity of > 30 days: 0.00, 0.00, 0.00, 0.00 - Operational deposits provided by non-financial corporates insured with a 3 percent run-off rate: 0.03, 0.03, 0.25, 0.25 - Operational deposits provided by non-financial corporates insured with a 5 percent run-off rate: 0.05, 0.05, 0.25, 0.25 - Operational deposits provided by non-financial corporates Uninsured: 0.25, 0.25, 0.50, 0.50 - Operational deposits provided by sovereigns/central banks/PSEs/MDBs insured with a 3 percent run-off rate: 0.03, 0.03, 0.03, 0.03 - Operational deposits provided by sovereigns/central banks/PSEs/MDBs insured with a 5 percent run-off rate: 0.05, 0.05, 0.05, 0.05 - Operational deposits provided by sovereigns/central banks/PSEs/MDBs Uninsured: 0.25, 0.25, 0.25, 0.25 - Operational deposits provided by banks insured with a 3 percent run-off rate: 0.03, 0.03, 0.25, 0.25 - Operational deposits provided by banks insured with a 5 percent run-off rate: 0.05, 0.05, 0.25, 0.25 - Operational deposits provided by banks Uninsured: 0.25, 0.25, 0.50, 0.50 - Total non-operational deposits provided by other banks: 1.00, 1.00, 1.00, 1.00 - Unsecured debt issuance: 1.00, 1.00, 1.00, 1.00 - Additional balances required to be installed in central bank reserves: 1.00, 1.00, 1.00, 1.00 - Non-operational deposits where entire amount is fully covered by an effective deposit insurance scheme (non-financial corporates): 0.20, 0.20, 0.50, 0.50 - Non-operational deposits where entire amount is not fully covered (non-financial corporates): 0.40, 0.40, 0.50, 0.50 - Non-operational deposits provided by other banks (where entire amount not specified): 1.00, 1.00, 1.00, 1.00 - 2.1.3. Secured Funding Run-off — Weights (all scenarios) - Transactions maturing ≤ 30 days not conducted with domestic central bank and backed by Level 2A assets (involving eligible liquid assets): 0.15 - Transactions maturing ≤ 30 days not conducted with domestic central bank and backed by Level 2B RMBS assets (involving eligible liquid assets): 0.25 - Transactions maturing ≤ 30 days not conducted with domestic central bank and backed by Level 2B non-RMBS assets where counterparties are domestic sovereigns/MDBs/domestic PSEs with a 20 percent risk weight: 0.25 - Same where counterparties are not domestic sovereigns/MDBs/domestic PSEs with a 20 percent risk weight: 0.50 - Transactions maturing ≤ 30 days not conducted with domestic central bank and backed by other assets where counterparties are domestic sovereigns/MDBs/domestic PSEs with a 20 percent risk weight: 0.25 - Same where counterparties are not domestic sovereigns/MDBs/domestic PSEs with a 20 percent risk weight: 1.00 - 2.1.4. Additional Requirements — Weights (all scenarios unless indicated) - Derivatives cash outflow: 1.00 - Increased liquidity needs related to downgrade triggers in derivatives and other financing transactions: 1.00 - Increased liquidity needs related to the potential for valuation changes on posted collateral (for other collateral i.e., all non-Level 1 collateral): 0.20 - Increased liquidity needs related to excess non-segregated collateral that could be called at any time: 1.00 - Increased liquidity needs related to contractually required collateral on transactions not yet demanded by counterparty: 1.00 - Increased liquidity needs related to contracts that allow collateral substitution to non-HQLA assets: 1.00 - Increased liquidity needs related to market valuation changes on derivative or other transactions: 1.00 - Loss of funding on ABS and other structured financing instruments issued by the bank, excluding covered bonds: 1.00 - Loss of funding on covered bonds issued by the bank: 1.00 - Loss of funding on asset-backed commercial paper (ABCP), conduits, SIVs (debt maturing ≤ 30 days): 1.00 - Undrawn committed credit and liquidity facilities to retail and small business customers: 0.05 - Undrawn committed credit facilities to non-financial corporates: 0.10 (scenarios 1–3), 0.25 (scenario 4) - Undrawn committed liquidity facilities to non-financial corporates: 0.30 (scenarios 1–3), 0.50 (scenario 4) - Undrawn committed credit and liquidity facilities provided to banks subject to prudential supervision: 0.40 - Undrawn committed credit facilities provided to other FIs: 0.40 - Undrawn committed liquidity facilities provided to other FIs: 1.00 - Undrawn committed credit and liquidity facilities to other legal entities: 1.00 - Other contractual obligations to extend funds to financial institutions: 1.00 - Retail/small business/non-financial and other clients contractual obligations (retail, small business customers, non-financials and other clients): 1.00 - Non-contractual obligations related to potential liquidity draws from joint ventures or minority investments: 1.00 - Unconditionally revocable "uncommitted" credit and liquidity facilities provided to retail and small business customers: 0.02 (scenarios 1–3), 0.05 (scenario 4) - Unconditionally revocable "uncommitted" credit and liquidity facilities provided to all other customers: 0.05 (scenarios 1–3), 0.10 (scenario 4) - Trade finance-related obligations (including guarantees and letters of credit): 0.03 - Guarantees and letters of credit unrelated to trade finance obligations: 0.05 (scenarios 1–3), 0.10 (scenario 4) - Structured products (non-contractual obligations): 0.05 - Other non-contractual obligations: 0.05 - Non-contractual obligations where customer short positions are covered by other customers’ collateral: 0.50 (scenarios 1–3), 0.75 (scenario 4) - Other contractual cash outflows (including those related to unsecured collateral borrowings and uncovered short positions): 1.00 - 2.2. Cash Inflows - 2.2.1. Secured Lending Including Reverse Repo and Securities Borrowing — Weights - Transactions backed by Level 2A assets (involving eligible liquid assets): 0.15 - Transactions backed by Level 2B RMBS assets (involving eligible liquid assets): 0.25 - Transactions backed by Level 2B non-RMBS assets (involving eligible liquid assets): 0.50 - Margin lending backed by non-Level 1 or non-Level 2 collateral: 0.50 - Transactions backed by other collateral: 1.00 - Re-used (rehypothecated) collateral transactions to cover the reporting institution's outright short positions — Transactions backed by Level 1/2A/2B assets and margin lending: 0.00 - 2.2.2. Other Inflows by Counterparty — Weights - Contractual inflows due in ≤ 30 days from fully performing loans, not reported as secured lending: - Retail customers: 0.50 - Small business customers: 0.50 - Non-financial corporates: 0.50 - Central banks: 1.00 - Financial institutions: operational deposits related to clearing placed by indirect clearers with an OSFI-regulated direct clearer: 0.25 - Financial institutions: other operational deposits: 0.00 - Financial institutions: non-operational demand deposits placed by indirect clearers with an OSFI-regulated direct clearer: 1.00 - Financial institutions: all payments on other loans and deposits due in ≤ 30 days: 1.00 - Other entities: 0.50 - 2.2.3. Other Cash Inflows — Weights - Derivatives cash inflow: 1.00 - Contractual inflows from securities maturing ≤ 30 days, not included elsewhere: 1.00 - Other contractual cash inflows: 1.00 - 2.2.4. Total Cash Inflows - Cap on cash inflows: 0.75 (applied to total cash inflows before cap) *CANADA INTERNATIONAL MONETARY FUND 146–149* ### 3. Collateral Swaps ### 3. Collateral Swaps ### Weight Outflows — Collateral swaps maturing ≤ 30 days - Collateral swaps maturing ≤ 30 days (overall heading; detailed breakdown follows). ### Of which the borrowed assets are not re-used (i.e., are not rehypothecated) - Level 1 assets are lent and Level 1 assets are borrowed, of which: - Involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Not involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Level 1 assets are lent and Level 2A assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 1 assets are lent and Level 2B RMBS assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 1 assets are lent and Level 2B non-RMBS assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 1 assets are lent and other assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2A assets are lent and Level 1 assets are borrowed, of which: - Involving eligible liquid assets: 0.15 0.15 0.15 0.15 - Not involving eligible liquid assets: 0.15 0.15 0.15 0.15 - Level 2A assets are lent and Level 2A assets are borrowed, of which: - Involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Not involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Level 2A assets are lent and Level 2B RMBS assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2A assets are lent and Level 2B non-RMBS assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2A assets are lent and other assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2B RMBS assets are lent and Level 1 assets are borrowed, of which: - Involving eligible liquid assets: 0.25 0.25 0.25 0.25 - Not involving eligible liquid assets: 0.25 0.25 0.25 0.25 - Level 2B RMBS assets are lent and Level 2A assets are borrowed, of which: - Involving eligible liquid assets: 0.10 0.10 0.10 0.10 - Not involving eligible liquid assets: 0.10 0.10 0.10 0.10 - Level 2B RMBS assets are lent and Level 2B RMBS assets are borrowed, of which: - Involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Not involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Level 2B RMBS assets are lent and Level 2B non-RMBS assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2B RMBS assets are lent and other assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Level 2B non-RMBS assets are lent and Level 1 assets are borrowed, of which: - Involving eligible liquid assets: 0.50 0.50 0.50 0.50 - Not involving eligible liquid assets: 0.50 0.50 0.50 0.50 - Level 2B non-RMBS assets are lent and Level 2A assets are borrowed, of which: - Involving eligible liquid assets: 0.35 0.35 0.35 0.35 - Not involving eligible liquid assets: 0.35 0.35 0.35 0.35 - Level 2B non-RMBS assets are lent and Level 2B RMBS assets are borrowed, of which: - Involving eligible liquid assets: 0.25 0.25 0.25 0.25 - Not involving eligible liquid assets: 0.25 0.25 0.25 0.25 - Level 2B non-RMBS assets are lent and Level 2B non-RMBS assets are borrowed, of which: - Involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Not involving eligible liquid assets: 0.00 0.00 0.00 0.00 - Level 2B non-RMBS assets are lent and other assets are borrowed, of which: - Involving eligible liquid assets: (no values reported) - Not involving eligible liquid assets: (no values reported) - Other assets are lent and Level 1 assets are borrowed, of which: - Involving eligible liquid assets: 1.00 1.00 1.00 1.00 - Not involving eligible liquid assets: 1.00 1.00 1.00 1.00 - Other assets are lent and Level 2A assets are borrowed, of which: - Involving eligible liquid assets: 0.85 0.85 0.85 0.85 - Not involving eligible liquid assets: 0.85 0.85 0.85 0.85 - Other assets are lent and Level 2B RMBS assets are borrowed, of which: - Involving eligible liquid assets: 0.75 0.75 0.75 0.75 - Not involving eligible liquid assets: 0.75 0.75 0.75 0.75 - Other assets are lent and Level 2B non-RMBS assets are borrowed, of which: - Involving eligible liquid assets: 0.50 0.50 0.50 0.50 - Not involving eligible liquid assets: 0.50 0.50 0.50 0.50 - Other assets are lent and other assets are borrowed: - 0.00 0.00 0.00 0.00 ### Of which the borrowed assets are re-used (i.e., are rehypothecated) in transactions to cover short positions - Level 1 assets are lent and Level 1 assets are borrowed: - 0.00 0.00 0.00 0.00 - Level 1 assets are lent and Level 2A assets are borrowed: - (no values reported) - Level 1 assets are lent and Level 2B RMBS assets are borrowed: - (no values reported) - Level 1 assets are lent and Level 2B non-RMBS assets are borrowed: - (no values reported) - Level 1 assets are lent and other assets are borrowed: - (no values reported) - Level 2A assets are lent and Level 1 assets are borrowed: - 0.15 0.15 0.15 0.15 - Level 2A assets are lent and Level 2A assets are borrowed: - 0.00 0.00 0.00 0.00 - Level 2A assets are lent and Level 2B RMBS assets are borrowed: - (no values reported) - Level 2A assets are lent and Level 2B non-RMBS assets are borrowed: - (no values reported) - Level 2A assets are lent and other assets are borrowed: - (no values reported) - Level 2B RMBS assets are lent and Level 1 assets are borrowed: - 0.25 0.25 0.25 0.25 - Level 2B RMBS assets are lent and Level 2A assets are borrowed: - 0.10 0.10 0.10 0.10 - Level 2B RMBS assets are lent and Level 2B RMBS assets are borrowed: - 0.00 0.00 0.00 0.00 - Level 2B RMBS assets are lent and Level 2B non-RMBS assets are borrowed: - (no values reported) - Level 2B RMBS assets are lent and other assets are borrowed: - (no values reported) - Level 2B non-RMBS assets are lent and Level 1 assets are borrowed: - 0.50 0.50 0.50 0.50 - Level 2B non-RMBS assets are lent and Level 2A assets are borrowed: - 0.35 0.35 0.35 0.35 - Level 2B non-RMBS assets are lent and Level 2B RMBS assets are borrowed: - 0.25 0.25 0.25 0.25 - Level 2B non-RMBS assets are lent and Level 2B non-RMBS assets are borrowed: - 0.00 0.00 0.00 0.00 - Level 2B non-RMBS assets are lent and other assets are borrowed: - (no values reported) - Other assets are lent and Level 1 assets are borrowed: - 1.00 1.00 1.00 1.00 - Other assets are lent and Level 2A assets are borrowed: - 0.85 0.85 0.85 0.85 - Other assets are lent and Level 2B RMBS assets are borrowed: - 0.75 0.75 0.75 0.75 - Other assets are lent and Level 2B non-RMBS assets are borrowed: - 0.50 0.50 0.50 0.50 - Other assets are lent and other assets are borrowed: - 0.00 0.00 0.00 0.00 *Source: 3. Collateral Swaps (Canada), IMF.* ### Appendix VIII. Cash Flow-based Analysis—Stress Test Haircuts ### Appendix VIII. Cash Flow-based Analysis—Stress Test Haircuts and Run-Off Rates ### Liability segments — Run-off rates (In percent) - Retail and Small Business Demand / Notice Deposits: - Type 1 insured, stable demand deposits: Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 7.5 - Type 2 insured, stable demand deposits: Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 10.0 - Insured, less stable demand deposits: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Unaffiliated third-party sourced demand deposits: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Uninsured demand deposits: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Retail and Small Business Cashable Term Deposits: - Type 1 insured, stable: Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 7.5 - Type 1 insured, less stable: Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 10.0 - Type 2 insured, stable: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Type 2 insured, less stable: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Uninsured: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Unaffiliated third-party sourced: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Retail and Small Business Fixed Term Deposits (by original term): - Type 1, insured, stable — (30-day) through (>1 year): Mild Scenario SL1 1.5; Severely Adverse Scenario SL 21 5.0 - Type 1, insured, less stable — (30-day) through (>1 year): Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 7.5 - Type 2, insured, stable — (30-day) through (>1 year): Mild Scenario SL1 1.5; Severely Adverse Scenario SL 21 5.0 - Type 2, insured, less stable — (30-day) through (>1 year): Mild Scenario SL1 2.5; Severely Adverse Scenario SL 21 7.5 - Uninsured — (30-day) through (>1 year): Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Unaffiliated third-party sourced — (30-day) through (>1 year): Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Structured Notes: - Structured Notes - No Cust Call: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Structured Notes - Cust Call: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Commercial, Corporate and Wholesale Demand Deposits: - Operational - insured, within approved jurisdiction: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Operational - insured, outside of approved jurisdiction: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Operational - not Insured: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 25.0 - Non-Operational - insured (FI): Mild Scenario SL1 30.0; Severely Adverse Scenario SL 21 100.0 - Non-Operational - uninsured (FI): Mild Scenario SL1 50.0; Severely Adverse Scenario SL 21 100.0 - Non-Operational - insured (Corp, Sovereigns central banks, PSE, MDB): Mild Scenario SL1 30.0; Severely Adverse Scenario SL 21 50.0 - Non-Operational - uninsured (Corp, Sovereigns, central banks, PSE, MDB): Mild Scenario SL1 50.0; Severely Adverse Scenario SL 21 100.0 - Commercial, Corporate and Wholesale Term Demands: - Operational term, original term <30 days, insured, within approved Jurisdiction: Mild Scenario SL1 5.0; Severely Adverse Scenario SL 21 15.0 - Operational term, original term <30 days, insured, outside of approved jurisdiction: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Operational term, original term <30 days, not Insured: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 30.0 - Operational term, original term >30 days and non-operational term: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Wholesale term: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Other Deposits / Guarantees: - Customers' Bankers acceptances issued — 1 month, 2 month, 3 month: Mild Scenario SL1 75.0; Severely Adverse Scenario SL 21 100.0 - Other deposits - Swapped intrabank deposits: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 20.0 - Other deposits - Deposits from affiliates: Mild Scenario SL1 10.0; Severely Adverse Scenario SL 21 30.0 - Liabilities—Repo and Securities Lent (examples): - Non-FI Issued ABS (High rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - FI Issued ABS (High rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - FI Issued ABS (Medium rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - FI Issued ABS (Low/not rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Bank's own debt not eliminated: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Eligible Non-Financial Common Equity shares: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Sovereign: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Liabilities—Securities Sold Short (examples mirror many repo/ securities-lent entries): - Non-FI Issued ABS (High rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Bank's own debt not eliminated: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Eligible Financial Common Equity: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - State, Provincial (High rated): Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 40.0 - Liabilities—Wholesale Issuance: - Wholesale Secured Debt Issuance — Bank's ABCP: Mild Scenario SL1 15.0; Severely Adverse Scenario SL 21 30.0 - Wholesale Secured Debt Issuance — Bank's ABS and MBS: Mild Scenario SL1 30.0; Severely Adverse Scenario SL 21 50.0 - Wholesale Unsecured Debt Issuance — Bankers' acceptances (BAs): Mild Scenario SL1 30.0; Severely Adverse Scenario SL 21 65.0 - Wholesale Unsecured Debt Issuance — Commercial Paper issued: Mild Scenario SL1 50.0; Severely Adverse Scenario SL 21 65.0 - Wholesale Unsecured Debt Issuance — Senior Unsecured Debt issued: Mild Scenario SL1 50.0; Severely Adverse Scenario SL 21 100.0 - Wholesale Unsecured Debt Issuance — Subordinated Debt issued: Mild Scenario SL1 50.0; Severely Adverse Scenario SL 21 100.0 ### Asset classes — Haircuts and price shocks (In percent) - Market shocks (examples) and haircuts for cash resources: - Coins and Bank Notes — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 0.00; Haircut Sev Adverse SL21 0.00 - Deposits with Central Banks - Mandatory — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 0.00; Haircut Sev Adverse SL21 0.00 - Deposits with FIs - Demand Deposits at FIs — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 0.00; Haircut Sev Adverse SL21 0.00 - Assets — Other assets with extreme haircuts under Severely Adverse: - Cash collateral pledged for exchange-traded derivatives: Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Cash collateral pledged for OTC derivatives: Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Cash collateral pledged for short sales: Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Commodities — Precious Metals and Other Commodities Held: Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Derivative Related Assets (FX and Cross Currency Swap Assets; Other DRA): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Other Assets - Other Assets: Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 0.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Assets — Securities (selected haircuts): - Non-FI Issued ABS (High rated) — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 5.00; Haircut Sev Adverse SL21 10.00 - Non-FI Issued ABCP (High rated) — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 5.00; Haircut Sev Adverse SL21 10.00 - FI Issued ABS (High rated) — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 15.00; Haircut Sev Adverse SL21 30.00 - FI Issued ABCP (High rated) — Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 15.00; Haircut Sev Adverse SL21 30.00 - Non-FI Issued ABS (Medium rated): Market Shock Mild SL1 2.50; Market Shock Sev Adverse SL 21 7.50; Haircut Mild SL1 15.00; Haircut Sev Adverse SL21 30.00 - FI Issued ABS (Medium rated): Market Shock Mild SL1 2.50; Market Shock Sev Adverse SL 21 7.50; Haircut Mild SL1 25.00; Haircut Sev Adverse SL21 50.00 - Non-FI Issued ABS (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 10.00; Haircut Mild SL1 25.00; Haircut Sev Adverse SL21 50.00 - FI Issued ABS (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 10.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 100.00 - Bank's own securities (not eliminated): - Bank's own debt not eliminated: Market Shock Mild SL1 10.00; Market Shock Sev Adverse SL 21 40.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Bank's own equity not eliminated: Market Shock Mild SL1 10.00; Market Shock Sev Adverse SL 21 40.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Corporate bonds and paper (selected entries): - Non-FI issued unsecured bonds and paper (High rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 10.00; Haircut Sev Adverse SL21 20.00 - FI issued unsecured bonds and paper (High rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 10.00; Haircut Mild SL1 10.00; Haircut Sev Adverse SL21 20.00 - FI issued unsecured bonds and paper (Medium rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 7.50; Haircut Mild SL1 25.00; Haircut Sev Adverse SL21 40.00 - FI issued unsecured bonds and paper (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 15.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 100.00 - Equities (haircuts): - Eligible Non-Financial Common Equity shares: Market Shock Mild SL1 10.00; Market Shock Sev Adverse SL 21 40.00; Haircut Mild SL1 30.00; Haircut Sev Adverse SL21 50.00 - Eligible Financial Common Equity: Market Shock Mild SL1 10.00; Market Shock Sev Adverse SL 21 40.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Other Equities: Market Shock Mild SL1 10.00; Market Shock Sev Adverse SL 21 40.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 100.00 - Government securities (selected): - Sovereign (High rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 2.50; Haircut Mild SL1 2.50; Haircut Sev Adverse SL21 5.00 - State, Provincial (High rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 5.00; Haircut Sev Adverse SL21 10.00 - Sovereign (Medium rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 2.50; Haircut Mild SL1 10.00; Haircut Sev Adverse SL21 10.00 - Sovereign (Low/not rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 10.00; Haircut Mild SL1 20.00; Haircut Sev Adverse SL21 30.00 - State, Provincial (Low/not rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 60.00 - State Municipal (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 15.00; Haircut Mild SL1 100.00; Haircut Sev Adverse SL21 100.00 - Supranational and Multilateral Development Bank (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 15.00; Haircut Mild SL1 40.00; Haircut Sev Adverse SL21 60.00 - Mortgage Backed Securities (MBS) (selected): - Agency MBS (High rated): Market Shock Mild SL1 2.50; Market Shock Sev Adverse SL 21 7.50; Haircut Mild SL1 5.00; Haircut Sev Adverse SL21 10.00 - Agency MBS (Medium rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 10.00; Haircut Mild SL1 10.00; Haircut Sev Adverse SL21 20.00 - Agency MBS (Low/not rated): Market Shock Mild SL1 5.00; Market Shock Sev Adverse SL 21 15.00; Haircut Mild SL1 20.00; Haircut Sev Adverse SL21 40.00 - Non-Agency Commercial MBS (Medium rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 30.00; Haircut Sev Adverse SL21 50.00 - Non-Agency Commercial MBS (Low/not rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 100.00 - Non-Agency Residential MBS (Low/not rated): Market Shock Mild SL1 0.00; Market Shock Sev Adverse SL 21 5.00; Haircut Mild SL1 50.00; Haircut Sev Adverse SL21 75.00 *Source: Appendix VIII. Cash Flow-based Analysis—Stress Test Haircuts and Run-Off Rates (Canada).* ### References ### References ### Cited works - Adrian, Tobias, Nina Boyarchenko, and Domenico Giannone. 2019. "Vulnerable Growth." American Economic Review 109(4): 1263-89. - Alter, Adrian, Jane Dokko, and MissDulani Seneviratne. 2018. “House Price Synchronicity, Banking Integration, and Global Financial Conditions.” IMF Working Paper 18/250, International Monetary Fund. - Arora, Rohan, Guillaume Bédard-Pagé, Guillaume Ouellet Leblanc, and Ryan Shotlander. 2019. “Bond Funds and Fixed-Income Market Liquidity: A Stress-Testing Approach.” Bank of Canada (115). - Arora, Rohan and Guillaume Ouellet Leblanc. 2018. “How do Canadian Corporate Bond Mutual Funds Meet Investor Redemptions?” Bank of Canada (2018-14). - Badarinza, Cristian, and Tarun Ramadorai. 2018. "Home away from home? Foreign Demand and London House Prices." Journal of Financial Economics 130(3): 532-555. - Baranova, Yuliya, Liu, Zijun, and Tamarah Shakir. 2017. "Dealer Intermediation, Market Liquidity and the Impact of Regulatory Reform." Bank of England Working Paper (665). - Grieder, Tim, and Claire Schaffter. 2019. “Measuring Non-Financial Corporate Sector Vulnerabilities in Canada.” Bank of Canada (2019-15). - Harrison, I. and Mathew, C. 2008. “Project TUI: A Structural Approach to the Understanding and Measurement of Residential Mortgage Lending Risk”, Reserve Bank of New Zealand. - International Monetary Fund. 2017. “New Zealand: Financial Sector Assessment Program: Technical Note on Stress testing.” - –––––––––––––––––––––. 2017a. “Global Financial Stability Report: Is Growth at Risk?” October. Washington, DC: International Monetary Fund, Washington, DC. - –––––––––––––––––––––. 2018. Brazil: Financial Sector Assessment Program: Technical Note on Stress Testing and Systemic Risk Analysis. IMF Country Report 18/344. - –––––––––––––––––––––. 2019. “Downside Risks to House Prices”, Global Financial Stability Review, Chapter 2, International Monetary Fund, Washington, DC. - –––––––––––––––––––––. 2019b. “Canada: Financial Sector Assessment Program: Technical Note on Housing Finance.”(forthcoming) - Gaa, C., X. Liu, C. MacDonald, X. Shen. 2019. “Assessing the Resilience of the Canadian Banking System.” Staff Analytical Note 2019-16, Bank of Canada. Available athttps://www.bankofcanada.ca/2019/05/staff-analytical-note-2019-16/ - CANADA IN INTERNATIONAL MONETARY FUND 165 - Gross, M., and J. Población. 2017. “Assessing the Efficacy of Borrower-based Macroprudential Policy Using an Integrated Micro-macro Model for European Households.” Economic Modelling 61: 510-528. - Laliotis, D., and J. Población. 2016. “Agent Based Models for the Assessment of the Application LTV/LTI Cap Macroprudential Measures.” Proceedings, 43rd Annual Conference, Eastern Economic Association, New York. - Laliotis, D., Buesa, A., Leber, M., Población J. 2019. “An agent-based model for the assessment of LTV Caps.” ECB Working Paper Series 2293. - Moody’s Investors Service. 2018. “Pension Funds – Canada: Pension Fund Liquidity Improves Despite Greater Illiquid Investment”. Sector In-Depth, 14 November 2018. - Office of the Superintendent of Financial Institutions (OSFI). 2018. "Capital Adequacy Requirements (CAR) 2018.” Chapter 1–Overview. Available at http://www.osfi-bsif.gc.ca/eng/fi-if/rg-ro/gdn-ort/gl-ld/Pages/car18_index.aspx. - Peterson, B. and T. Roberts. 2016. “Household Risk Assessment Model.” Bank of Canada, Technical Report 106. - Sala-I-Martin, X., Doppelhofer, G. and Miller, R.I. 2004., “Determinants of Long-term Growth: A Bayesian Averaging of Classical Estimates (BACE) approach.” American Economic Review, Vol. 94(4): 813-835. - Sá, Filipa. 2016 “The Effect of Foreign Investors on Local Housing Markets: Evidence From The United Kingdom”, CEPR Discussion Paper 11658. - Sá, Filipa. and Tomasz. Wieladek. 2015. “Capital Inflows and the US Housing Boom.” Journal of Money, Credit and Banking, Volume 47(S1): 221-256. - Vitek, F. 2018. “The Global Macrofinancial Model.” IMF Working Paper 18/81, International Monetary Fund, Washington, DC. *Reference list from 1canea2020001 - References* --- _Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1canea2020001.pdf_