## Preface

## Source details

**Canonical URL:** [Preface](https://www.imf.org/-/media/files/publications/tar/2024/english/tarea2024075-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/tar/2024/english/tarea2024075-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/tar/2024/english/tarea2024075-print-pdf.pdf.json)

---

### Mission scope and participants
- Two Monetary and Capital Markets Department (MCM) missions held in Panama City: January 9 to January 13, 2023, and April 3 to April 8, 2024.
- Technical assistance (TA) provided by Torsten Wezel (MCM, activity lead), Manuel Luy (external expert, 2023 mission), and Eduardo Bastante (external expert, 2024 mission).
- Principal counterpart: Superintendency of Banks of Panama (SBP), including staff of the Financial Studies Department (Mr. Javier Motta, Ms. Ingrid Arboleda, Ms. Libny Gonzales), the Banking Supervision Department, Superintendent of Banks (Mr. Amauri Castillo), and Director of Risk Management (Ms. Nahila Melgar).
- Focus areas: credit risk, liquidity risk, market risk, corporate risk, and implementation/training on stress-testing methodologies.

### Executive summary — core deliverables
- Implemented tools/methodologies in credit risk, liquidity risk, market risk, and corporate health monitoring across 2023–24 missions.
- Follow-up on May 2022 recommendations on PD estimation models:
  - Updated and improved heat map (PD estimation tool at the borrower level) for large borrowers.
  - For retail portfolios: re-estimation of econometric PD models, segmentation for certain portfolios (e.g., electricity and credit card), and lifetime PD calculations for horizons greater than four years.
- Liquidity stress training:
  - Used IMF cash flow stress template (“Treatment of Liquidity Risk in Stress Tests” by M. Catalán, 2015).
  - Introduced MCM Systemwide Liquidity Stress Testing Tool (“Systemwide Liquidity Stress Testing Tool” by H. Oura, 2022) for contagion effects across banks, non-bank financial institutions, and non-financial sectors.
- Market risk and corporate risk preparatory and implementation work:
  - April 3–8, 2024 mission implemented a market risk model (modified duration + differentiated interest rate shocks), corporate risk indicators, and explained the 2023 FSAP liquidity stress methodology.
  - Corporate risk indicators calculated: Debt-to-EBITDA, Interest Coverage Ratio (ICR), and change in cash (cash-flow impacts).
- Implementation status:
  - Methodologies largely implemented; some workstreams pending data collection, model updates, or regulatory action.
  - Noted potential delays due to absorption of resources during the 2023 FSAP and other challenges (e.g., staffing levels).

### Key recommendations (summary)
- Market Risk Stress Testing
  - Implement the market risk stress tool in the analysis of the solvency of financial institutions according to the proposed methodology (paragraph 29).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
  - Evaluate the incorporation of convexity in the reporting of information to the SBP and ensure the quality of investment information submitted through Reporting Form 15 (paragraph 30).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
- Corporate Stress Testing
  - Collect key financial statement data of the largest firms for the calculation of representative corporate risk indicators (paragraph 36).
    - Authority in charge: SBP
    - Priority: Medium
    - Time: MT
- Liquidity Stress Testing
  - Continue to compile all relevant detailed data to allow for a comprehensive liquidity risk analysis of Panamanian banks (paragraph 42).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
- Timeframe definitions:
  - ST: less than 12 months
  - MT: between 12 and 24 months

### Introduction — background and work summary
- Report covers two TA stress testing missions (January 2023 and April 2024); 2023 mission followed up on May 2022 TA that implemented the solvency stress test methodology and provided liquidity/corporate risk training.
- Prior SBP progress:
  - Solvency risk tool developed with IMF between 2011 and 2014 and modified subsequently; further TA in April–May 2021 and May 2022 to improve PD projection models.
  - Market risk quantitative tool originally developed in 2017.
- 2024 mission deliverables:
  - Market risk model using modified duration and differentiated interest rate shocks based on investment risk profiles reported monthly to SBP.
  - Corporate risk indicators: Debt-to-EBITDA; Interest Coverage Ratio (earnings before tax and interest to interest expense); change in cash (company cash-flow impacts).
  - Explanation and enablement of the liquidity stress test methodology used in the 2023 FSAP.

### Strengthening and updating of stress tests (2023 mission)
- Credit Risk (large borrowers)
  - Heat map PD calibration methodology steps:
    - Select macroeconomic, sectoral and idiosyncratic variables reflecting debtor creditworthiness.
    - Calibrate thresholds to determine risk buckets; align number of levels with information dispersion.
    - Assign to each bucket a PD based on historical transition matrices or external ratings; align with Panamanian economy characteristics and portfolio default rate.
    - Calculate debtor intrinsic PD as weighted average of PDs of each variable.
  - Progress achieved in 2023:
    - Increased number of borrowers modeled from 6 to 23.
    - Updated model with information through 2022.
    - Added borrower ability-to-pay variables (profitability: ratio of net income to sales income; liquidity: ratio of cash to accounts payables).
    - Recalibrated heat map weights after adding new variables.
    - SBP supervisory unit project to further increase coverage of financial statements for large business borrowers is ongoing.
- Credit Risk (retail portfolios)
  - Econometric panel-data PD models used across nine credit-type models and by financial institution to capture heterogeneity.
  - Recommendations:
    - Re-estimate default models after normalization of default series post-pandemic (migration of “modified portfolio” to fully performing completed December 2022; new equilibrium estimated to be reached in 2023).
    - Improve segmentation for credit card and industry sector portfolios.
    - Estimate lifetime PD series for horizons greater than four years to analyze convergence to long-run stable level.

### Training in liquidity stress tools
- Training objectives:
  - Assess ability of individual banks and the system to withstand extreme but plausible funding shocks.
  - Identify priority policy actions: reduce exposures or build liquidity buffers.
- Tools and differences:
  - Main tool: IMF cash flow stress template (Catalán, 2015) — Excel-based, simulates liquidity inflows/outflows across different time horizons (vs. Basel III LCR’s fixed 30-day outflow simulation).
  - Introduced MCM Systemwide Liquidity Stress Testing Tool (Oura, 2022) for contagion effects; provided recommendations to enable future implementation.
- Key assumptions in cash flow stress template:
  - Run-off rates: fraction of liabilities withdrawn (or not rolled over) per time horizon.
  - Roll-off rates: fraction of assets convertible into liquidity (and not rolled over) per time horizon.
  - Haircut rates: discounted sales price of securities sold to cover negative liquidity gaps (counterbalancing capacity).

---

### Market Risk Stress Testing and Corporate Risk Indicators (2024 Mission)

### Systemic liquidity stress tool and preparatory steps
- Mission presented “Systemwide Liquidity Stress Testing Tool” by H. Oura, 2022, which assesses impact of aggregate liquidity stress across the economy and evaluates contagion among banks, non-bank financial institutions, and non-financial corporates; methodology could be implemented at a future stage.
- 2023 mission recommended preparatory steps for implementation at individual-bank level: adjust model parameters (run-off rates, roll-off rates, haircuts) based on historical liquidity crisis information and Panama’s financial system characteristics.

### Updating of market risk stress tools — background
- 2017 TA designed a quantitative market risk monitoring model calibrating risk indicators: Value-at-Risk (VaR), Stressed Value-at-Risk (SVaR), Conditional Value-at-Risk (CVaR, or expected shortfall, ES), and Stressed Conditional Value-at-Risk (SCVaR) at institution and system level.
- 2023 mission identified institutions most exposed to market risk to design a regulatory capital framework.

### Market Risk Stress Testing (2024 mission) — data and methodology
- Investment portfolios: consist mainly of fixed-income instruments measured at market value and amortized cost. Permanent investment positions were excluded. Total investment portfolio = sum of exposures measured at market value and held-to-maturity investments. Equity instruments account for 2 percent of the investment portfolio; methodology focuses on fixed-income instruments.
- Portfolio grouping and haircuts (paragraph 19):
  - 1st group: Equity instruments and those that do not have a specific classification — haircut factor of 30 percent to reported book value.
  - 2nd group: Fixed-income instruments that do not have a market price or subordinated instruments — haircut factor of 20 percent to book value.
  - 3rd group: Fixed-income investment portfolio measured at market value — modified duration-based methodology applied.
- Sensitivity analysis segmentation (paragraph 20): four groups for interest rate risk:
  - (i) Panama’s sovereign bonds;
  - (ii) sovereign bonds of other countries;
  - (iii) locally rated corporate instruments;
  - (iv) internationally rated corporate instruments.
- Shock determination (paragraphs 21–22):
  - Adverse shocks derived from interest rate behavior between December 2021 and December 2022, differentiating for risk profile based on modified duration and initial yields.
  - Process based on historical relationship between variation in the Federal Reserve’s (FED) effective interest rate and variation in bond yield during 2022, which was 75 percent.
  - Methodology proposes use of historical relationship between annual variation of FED's effective rate and variation of average bond yields, then apply sensitivity percentages by group, tenor and risk profile; may be improved by incorporating a rate term structure analysis rather than just FED’s effective rate.
- Impact calculation (paragraphs 23–25):
  - Stress impact of each bond position = interest rate shocks (by group and risk profile) × modified duration.
  - Modified duration, initial interest rates, and market value of bonds come from financial institutions’ reports; data quality is key.
  - Convexity would overcome some limitations of modified duration, but convexity data not available; use modified duration is recommended for the time being.
  - Instruments measured at amortized cost must be adjusted to market value before applying price change from interest rates; under stress held-to-maturity may need liquidation.
  - Recommendation: losses from impairment of amortized cost instruments be incorporated into liquidity stress test; losses on instruments at fair value be incorporated into market risk stress test.

### Market Risk Stress Testing — results and recommendations
- Policy rate shock scenario:
  - Applied shock: 300 basis points to the Federal Reserve’s effective rate on the investment portfolio as of December 2023.
  - Estimated impact of market risk stress: $2,553.5 million.
  - Investment portfolio measured at market value: $1,107.8 million.
  - Investment portfolio measured at amortized cost: $1,445.7 million.
- Heterogeneity of impact:
  - Impact on financial institutions ranged from 0.3 percent to 30 percent of the book value of the investment portfolio.
  - Of 38 financial institutions analyzed, average impact = 8.9 percent of the portfolio’s book value.
  - Eight institutions exceeded 15 percent impact on market value.
- Recommendations:
  - Implement the market risk stress tool developed by the mission and incorporate into solvency analysis and liquidity assumptions to obtain integrated stress test results.
  - Ensure and improve quality of information on investments received by SBP; enhance data quality controls, validate information among institutions, and perform on-site verifications.
  - Evaluate inclusion of convexity as an additional indicator for greater precision.
- Table 2 — Main Results of Market Risk Stress Test (as of December 2023) — key figures:
  - Securities at Fair Value:
    - 1st group: Book Value 1,111.3 (US$ million); Impact 333.4 (US$ million); Impact (perc. RWA) 0.3%
    - 2nd group: Book Value 1,043.6; Impact 208.7; Impact (perc. RWA) 0.2%
    - 3rd group: Book Value 10,338.0; Impact 565.7; Impact (perc. RWA) 0.6%
    - Subtotal: Book Value 12,492.9; Impact 1,107.8; Impact (perc. RWA) 1.1%
  - Securities at Amortized Cost:
    - 1st group: Book Value 1,159.7; Impact 347.5; Impact (perc. RWA) 0.3%
    - 2nd group: Book Value 570.7; Impact 127.7; Impact (perc. RWA) 0.1%
    - 3rd group: Book Value 8,036.2; Impact 970.5; Impact (perc. RWA) 1.0%
    - Subtotal: Book Value 9,766.6; Impact 1,445.7; Impact (perc. RWA) 1.5%
  - Total: Book Value 22,259.5; Impact 2,553.5; Impact (perc. RWA) 2.6%
  - Note: US$448.5 million (0.6% of RWA) are due to the gap between book and market value.
- Operational recommendations:
  - Perform detailed analysis and on-site verification for institutions where impact is material relative to total assets.
  - Given staff and information constraints, rely on modified duration calculated by institutions but strengthen validation; consider convexity when data become available.

### Corporate Risk Indicators (2024 mission) — data and methodology
- Implementation status:
  - 2024 mission implemented corporate risk indicators in a limited fashion due to data restrictions.
  - Sample: small sample of only 22 firms collected by supervision during on-site inspections; biased and not representative (firms vary widely in size, not the largest; half of firms are in special mention or lower; data completeness and reliability issues).
- Three risk indicators calculated and stressed:
  - Net Debt-to-EBITDA ratio:
    - Net debt = gross debt − cash and cash-like instruments.
    - Relevant debt includes short- and long-term obligations (typically not transitory positions like accounts payable).
    - Debt stock considered sustainable if ratio ≤ 5 (or ≤ 7 for capital-intensive firms). Ratios beyond these thresholds or negative ratios due to negative EBITDA imply non-sustainable debt.
  - Interest Coverage Ratio (ICR):
    - ICR = EBIT / interest payments.
    - Absolute minimum for debt service sustainability is ratio ≥ 1; some studies use threshold of 1.5.
    - ICR can be stressed by lowering EBIT and/or increasing interest payments.
  - Sustainable cash balance:
    - Starting balance uses cash only (sample lacks short-term investments).
    - Intra-year variation determined by:
      (i) income from operations (EBIT);
      (ii) net cost of financing (interest payments, disaggregated by short- and long-term);
      (iii) net inflow of financing (rollover of maturing funding);
      (iv) depreciation and amortization;
      (v) capital expenditure (CAPEX).
    - Depreciation and amortization and cost of long-term debt assumed pre-determined; depreciation & amortization not stressed.
    - Capital expenditure is capped by depreciation and amortization (lower of the two used).

### Corporate Risk Indicators — results and recommendations
- Illustrative results (sample of 22 firms; not representative):
  - Debt-to-EBITDA (stress factor applied: 20% -30%):
    - Many firms are overindebted even before stress: almost all firms have ratio > 5 (or 7) or negative earnings under baseline.
    - Example rows:
      - Firm 1: Net debt 126,800,466; EBITDA 40,627,990; Debt-to-EBITDA Ratio (baseline) 3.12 → Net debt stressed 152,160,559; EBITDA stressed 28,439,593; Debt-to-EBITDA Ratio (stressed) 5.35
      - Firm 3: Net debt 453,682,000; EBITDA 77,255,000; Debt-to-EBITDA Ratio (baseline) 5.87 → Net debt stressed 544,418,400; EBITDA stressed 54,078,500; Debt-to-EBITDA Ratio (stressed) 10.07
      - Firm 12: Net debt 87,529,378; EBITDA 658,297; Debt-to-EBITDA Ratio (baseline) 132.96 → Net debt stressed 105,035,254; EBITDA stressed 460,808; Debt-to-EBITDA Ratio (stressed) 227.94
  - Interest Coverage Ratio (stress factor applied: -40% (EBIT) and 5% (interest payments)):
    - Under baseline and stressed scenarios, about half the sample firms are above threshold of 1.5 under baseline conditions (two firms do not report interest payments).
    - Example rows:
      - Firm 1: EBIT 27,285,011; Interest payments 3,558,962; Interest Coverage Ratio (baseline) 7.67 → EBIT stressed 16,371,007; Interest payments stressed 3,736,910; Interest Coverage Ratio (stressed) 4.38
      - Firm 3: EBIT 25,144,000; Interest payments 11,273,000; Interest Coverage Ratio (baseline) 2.23 → EBIT stressed 15,086,400; Interest payments stressed 11,836,650; Interest Coverage Ratio (stressed) 1.27
      - Firm 22: EBIT 198,855,621; Interest payments 28,056,312; Interest Coverage Ratio (baseline) 7.09 → EBIT stressed 119,313,373; Interest payments stressed 29,459,128; Interest Coverage Ratio (stressed) 4.05
  - Cash balance stress:
    - Assuming a 40 percent drop in EBIT, a five percent higher cost of short-term financing, and a 95 percent rollover of short-term debt, about one-third of sample firms exhaust their cash under stress.
- Recommendations to improve corporate stress analysis:
  - Revitalize effort to collect financial statement data of the largest Panamanian firms to obtain a more representative sample.
  - SBP should:
    - Collect necessary data for the 100 largest firms as measured by the amount of bank credit.
    - Obtain more granular data on short-term investments and short-term obligations; include economic sector and conglomerate membership.
    - Identify and correct data issues: missing observations, implausible zeros, and observations with unexpected (negative) sign.
  - Once data collection is finished, expand analysis:
    - With 100 observations obtain meaningful distributions of indicators (as in referenced corporate risk analyses).
    - Segment sample by firms mainly bank-financed vs. more diversified funding, or by main economic sector.
    - Use firm-level corporate stress test results as early warning indicators in SBP supervisory process.

---

### Enhanced liquidity stress testing tool (content unit: 38)

### Description of the tool
- Tool expands cash flow based liquidity stress test models by:
  - (i) properly computing forward flows of maturing and reinvested/rolled over assets and liabilities along the liquidity simulation horizon;
  - (ii) operating with only one counterbalancing capacity and its underlying detailed asset components to serve as buffers and as cash inflow-generating at the same time;
  - (iii) incorporating a calculation for losses that may arise when bonds in the held-to-maturity investment category would have to be sold, and therefore result in adverse feedback to banks’ solvency ratios.
- Tool was used in the 2023 Panama FSAP and for other recent FSAPs.
- One session on the liquidity stress test model topic was delivered by Mr. Marco Gross (Monetary and Capital Markets Department of the IMF).

### Versions and data requirements
- Three versions of the tool suite provided, differing by required input data:
  - Simple version: Requires no maturity profile data as inputs for assets or liabilities.
  - Intermediate version: Requires maturity profile data for liabilities.
  - Advanced version: Requires maturity ladder data for both assets and liabilities.
- All three versions require selected market risk parameters as inputs, including:
  - current market values,
  - book values,
  - current yields,
  - durations.
- Advanced model version was used for the 2023 FSAP and is most likely the one the SBP will operate going forward.
- A guidance note detailing the methodology will be made available to the SBP in due course.

### Data granularity and special considerations
- Tool requires detailed information on:
  - various types of funding (deposits, available credit lines and other funding), disaggregated by several maturity buckets up to 180 days;
  - cash and cash equivalents;
  - bonds (separated by maturity and by fair value and amortized cost);
  - other securities, including stocks.
- Special attention item for the SBP:
  - liquidity generating credit lines that Panamanian banks maintain with their foreign correspondent banks (given that Panama has no central bank).
- Bespoke counterfactuals can be and were being conducted with the model during the FSAP in this context.

### Illustrative results and outputs
- Assuming a certain pecking order of asset use to meet outflows, the tool calculated:
  - the stage at which the sample bank ran out of liquid assets in a certain maturity bucket;
  - the liquidity gap;
  - losses caused by fire sales of HTM securities that should feed into the solvency stress test.

### Recommendations for SBP surveillance and next steps
- Continue compiling all relevant, detailed data to allow for a comprehensive liquidity risk analysis of Panamanian banks, supported by the liquidity stress test models provided as part of the TA.
- The Financial Research Directorate will familiarize itself with the tool and then use it to conduct liquidity stress tests. (See Appendix I: Proposed Methodology for Liquidity Stress — Completed.)

*Source: Preface and selected sections of the IMF Technical Assistance Report (tarea2024075-print-pdf).*

### Preface _________________________________________________________________________ 5

### Preface — IMF Technical Assistance Report (tarea2024075-print-pdf) — Preface

### Mission Scope and Participants
- Two Monetary and Capital Markets Department (MCM) missions held in Panama City: January 9 to January 13, 2023, and April 3 to April 8, 2024.
- Technical assistance (TA) provided by Torsten Wezel (MCM, activity lead), Manuel Luy (external expert, 2023 mission), and Eduardo Bastante (external expert, 2024 mission).
- Principal counterpart: Superintendency of Banks of Panama (SBP), including staff of the Financial Studies Department (Mr. Javier Motta, Ms. Ingrid Arboleda, Ms. Libny Gonzales), the Banking Supervision Department, Superintendent of Banks (Mr. Amauri Castillo), and Director of Risk Management (Ms. Nahila Melgar).
- Focus areas: credit risk, liquidity risk, market risk, corporate risk, and implementation/training on stress-testing methodologies.

### Executive Summary — Core Deliverables
- Implemented tools/methodologies in credit risk, liquidity risk, market risk, and corporate health monitoring across 2023–24 missions.
- Follow-up on May 2022 recommendations on PD estimation models:
  - Updated and improved heat map (PD estimation tool at the borrower level) for large borrowers.
  - For retail portfolios: re-estimation of econometric PD models, segmentation for certain portfolios (e.g., electricity and credit card), and lifetime PD calculations for horizons greater than four years.
- Liquidity stress training:
  - Used IMF cash flow stress template (“Treatment of Liquidity Risk in Stress Tests” by M. Catalán, 2015).
  - Introduced MCM Systemwide Liquidity Stress Testing Tool (“Systemwide Liquidity Stress Testing Tool” by H. Oura, 2022) for contagion effects across banks, non-bank financial institutions, and non-financial sectors.
- Market risk and corporate risk preparatory and implementation work:
  - April 3–8, 2024 mission implemented a market risk model (modified duration + differentiated interest rate shocks), corporate risk indicators, and explained the 2023 FSAP liquidity stress methodology.
  - Corporate risk indicators calculated: Debt-to-EBITDA, Interest Coverage Ratio (ICR), and change in cash (cash-flow impacts).
- Implementation status:
  - Methodologies largely implemented; some workstreams pending data collection, model updates, or regulatory action.
  - Noted potential delays due to absorption of resources during the 2023 FSAP and other challenges (e.g., staffing levels).

### Key Recommendations (summary of Table 1)
- Market Risk Stress Testing
  - Implement the market risk stress tool in the analysis of the solvency of financial institutions according to the proposed methodology (paragraph 29).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
  - Evaluate the incorporation of convexity in the reporting of information to the SBP and ensure the quality of investment information submitted through Reporting Form 15 (paragraph 30).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
- Corporate Stress Testing
  - Collect key financial statement data of the largest firms for the calculation of representative corporate risk indicators (paragraph 36).
    - Authority in charge: SBP
    - Priority: Medium
    - Time: MT
- Liquidity Stress Testing
  - Continue to compile all relevant detailed data to allow for a comprehensive liquidity risk analysis of Panamanian banks (paragraph 42).
    - Authority in charge: SBP
    - Priority: High
    - Time: ST
- Timeframe definitions from Table 1:
  - ST: less than 12 months
  - MT: between 12 and 24 months

### Introduction — Background and Work Summary
- Report covers two TA stress testing missions (January 2023 and April 2024); 2023 mission followed up on May 2022 TA that implemented the solvency stress test methodology and provided liquidity/corporate risk training.
- Prior SBP progress:
  - Solvency risk tool developed with IMF between 2011 and 2014 and modified subsequently; further TA in April–May 2021 and May 2022 to improve PD projection models.
  - Market risk quantitative tool originally developed in 2017.
- 2024 mission deliverables:
  - Market risk model using modified duration and differentiated interest rate shocks based on investment risk profiles reported monthly to SBP.
  - Corporate risk indicators: Debt-to-EBITDA; Interest Coverage Ratio (earnings before tax and interest to interest expense); change in cash (company cash-flow impacts).
  - Explanation and enablement of the liquidity stress test methodology used in the 2023 FSAP.

### Strengthening and Updating of Stress Tests (2023 Mission)
- Credit Risk (large borrowers)
  - Heat map PD calibration methodology steps:
    - Select macroeconomic, sectoral and idiosyncratic variables reflecting debtor creditworthiness.
    - Calibrate thresholds to determine risk buckets; align number of levels with information dispersion.
    - Assign to each bucket a PD based on historical transition matrices or external ratings; align with Panamanian economy characteristics and portfolio default rate.
    - Calculate debtor intrinsic PD as weighted average of PDs of each variable.
  - Progress achieved in 2023:
    - Increased number of borrowers modeled from 6 to 23.
    - Updated model with information through 2022.
    - Added borrower ability-to-pay variables (profitability: ratio of net income to sales income; liquidity: ratio of cash to accounts payables).
    - Recalibrated heat map weights after adding new variables.
    - SBP supervisory unit project to further increase coverage of financial statements for large business borrowers is ongoing.
- Credit Risk (retail portfolios)
  - Econometric panel-data PD models used across nine credit-type models and by financial institution to capture heterogeneity.
  - Recommendations:
    - Re-estimate default models after normalization of default series post-pandemic (migration of “modified portfolio” to fully performing completed December 2022; new equilibrium estimated to be reached in 2023).
    - Improve segmentation for credit card and industry sector portfolios.
    - Estimate lifetime PD series for horizons greater than four years to analyze convergence to long-run stable level.

### Training in Liquidity Stress Tools
- Training objectives:
  - Assess ability of individual banks and the system to withstand extreme but plausible funding shocks.
  - Identify priority policy actions: reduce exposures or build liquidity buffers.
- Tools and differences:
  - Main tool: IMF cash flow stress template (Catalán, 2015) — Excel-based, simulates liquidity inflows/outflows across different time horizons (vs. Basel III LCR’s fixed 30-day outflow simulation).
  - Introduced MCM Systemwide Liquidity Stress Testing Tool (Oura, 2022) for contagion effects; provided recommendations to enable future implementation.
- Key assumptions in cash flow stress template:
  - Run-off rates: fraction of liabilities withdrawn (or not rolled over) per time horizon.
  - Roll-off rates: fraction of assets convertible into liquidity (and not rolled over) per time horizon.
  - Haircut rates: discounted sales price of securities sold to cover negative liquidity gaps (counterbalancing capacity).

*Italic — Source: Preface and Prefatory sections of the IMF Technical Assistance Report (tarea2024075-print-pdf)*

### 14. T he mission also informed about a new systemic liquidity stress tool. This tool (“Systemwide

### Market Risk Stress Testing and Corporate Risk Indicators (2024 Mission)

### Systemic liquidity stress tool and preparatory steps
- The mission informed about a new systemic liquidity stress tool: “Systemwide Liquidity Stress Testing Tool” by H. Oura, 2022, which assesses the impact of a high degree of aggregate liquidity stress across the economy and evaluates contagion among banks, non-bank financial institutions, and the non-financial corporate sector. The methodology could be implemented at a future stage.  
- The 2023 mission recommended preparatory steps for implementation of the liquidity testing tool at the level of individual banks: adjust model parameters (run-off rates, roll-off rates, and haircuts) based on historical liquidity crisis information and Panama’s financial system characteristics.

### Updating of market risk stress tools — background
- The 2017 TA mission designed a quantitative market risk monitoring model (investment portfolio) calibrating risk indicators: Value-at-Risk (VaR), Stressed Value-at-Risk (SVaR), Conditional Value-at-Risk (CVaR, or expected shortfall, ES), and Stressed Conditional Value-at-Risk (SCVaR) at institution and system level.  
- The 2023 mission identified institutions most exposed to market risk to design a regulatory capital framework to mitigate these risks.

### Market Risk Stress Testing (2024 Mission) — Data and methodology
- Investment portfolios: Panamanian banks’ investment portfolios consist mainly of fixed-income instruments measured at market value and amortized cost. Permanent investment positions were excluded. Total investment portfolio = sum of exposures measured at market value and held-to-maturity investments. Equity instruments account for 2 percent of the investment portfolio; methodology focuses on fixed-income instruments.
- Portfolio grouping and haircuts (paragraph 19):
  - 1st group: Equity instruments and those that do not have a specific classification — haircut factor of 30 percent to reported book value.
  - 2nd group: Fixed-income instruments that do not have a market price or subordinated instruments — haircut factor of 20 percent to book value.
  - 3rd group: Fixed-income investment portfolio measured at market value — modified duration-based methodology applied.
- Sensitivity analysis segmentation (paragraph 20): four groups for interest rate risk:
  - (i) Panama’s sovereign bonds;
  - (ii) sovereign bonds of other countries;
  - (iii) locally rated corporate instruments;
  - (iv) internationally rated corporate instruments.
- Shock determination (paragraphs 21–22):
  - Adverse shocks derived from interest rate behavior between December 2021 and December 2022, differentiating for risk profile based on modified duration and initial yields.
  - Process based on historical relationship between variation in the Federal Reserve’s (FED) effective interest rate and variation in bond yield during 2022, which was 75 percent.
  - Methodology proposes use of historical relationship between annual variation of FED's effective rate and variation of average bond yields, then apply sensitivity percentages by group, tenor and risk profile; may be improved by incorporating a rate term structure analysis rather than just FED’s effective rate.
- Impact calculation (paragraphs 23–25):
  - Stress impact of each bond position = interest rate shocks (by group and risk profile) × modified duration.
  - Modified duration, initial interest rates, and market value of bonds come from financial institutions’ reports; data quality is key.
  - Convexity would overcome some limitations of modified duration, but convexity data not available; use modified duration is recommended for the time being.
  - Instruments measured at amortized cost must be adjusted to market value before applying price change from interest rates; under stress held-to-maturity may need liquidation.
  - Recommendation: losses from impairment of amortized cost instruments be incorporated into liquidity stress test; losses on instruments at fair value be incorporated into market risk stress test.

### Market Risk Stress Testing — Results and recommendations
- Policy rate shock scenario:
  - Applied shock: 300 basis points to the Federal Reserve’s effective rate on the investment portfolio as of December 2023.
  - Estimated impact of market risk stress: $2,553.5 million.
  - Investment portfolio measured at market value: $1,107.8 million.
  - Investment portfolio measured at amortized cost: $1,445.7 million.
- Heterogeneity of impact:
  - Impact on financial institutions ranged from 0.3 percent to 30 percent of the book value of the investment portfolio.
  - Of 38 financial institutions analyzed, average impact = 8.9 percent of the portfolio’s book value.
  - Eight institutions exceeded 15 percent impact on market value.
- Recommendations:
  - Implement the market risk stress tool developed by the mission and incorporate into solvency analysis and liquidity assumptions to obtain integrated stress test results.
  - Ensure and improve quality of information on investments received by SBP; enhance data quality controls, validate information among institutions, and perform on-site verifications.
  - Evaluate inclusion of convexity as an additional indicator for greater precision.
- Table 2 — Main Results of Market Risk Stress Test (as of December 2023) — key figures preserved:
  - Securities at Fair Value:
    - 1st group: Book Value 1,111.3 (US$ million); Impact 333.4 (US$ million); Impact (perc. RWA) 0.3%
    - 2nd group: Book Value 1,043.6; Impact 208.7; Impact (perc. RWA) 0.2%
    - 3rd group: Book Value 10,338.0; Impact 565.7; Impact (perc. RWA) 0.6%
    - Subtotal: Book Value 12,492.9; Impact 1,107.8; Impact (perc. RWA) 1.1%
  - Securities at Amortized Cost:
    - 1st group: Book Value 1,159.7; Impact 347.5; Impact (perc. RWA) 0.3%
    - 2nd group: Book Value 570.7; Impact 127.7; Impact (perc. RWA) 0.1%
    - 3rd group: Book Value 8,036.2; Impact 970.5; Impact (perc. RWA) 1.0%
    - Subtotal: Book Value 9,766.6; Impact 1,445.7; Impact (perc. RWA) 1.5%
  - Total: Book Value 22,259.5; Impact 2,553.5; Impact (perc. RWA) 2.6%
  - Note: US$448.5 million (0.6% of RWA) are due to the gap between book and market value.
- Additional operational recommendations:
  - Perform detailed analysis and on-site verification for institutions where impact is material relative to total assets.
  - Given staff and information constraints, rely on modified duration calculated by institutions but strengthen validation; consider convexity when data become available.

### Corporate Risk Indicators (2024 Mission) — Data and methodology
- Implementation status:
  - After introducing the methodology during the 2023 visit, the 2024 mission implemented corporate risk indicators in a limited fashion due to data restrictions.
  - Sample: small sample of only 22 firms collected by supervision during on-site inspections; biased and not representative (firms vary widely in size, not the largest; half of firms are in special mention or lower; data completeness and reliability issues).
- Three risk indicators calculated and stressed:
  - Net Debt-to-EBITDA ratio:
    - Net debt = gross debt − cash and cash-like instruments.
    - Relevant debt includes short- and long-term obligations (typically not transitory positions like accounts payable).
    - Debt stock considered sustainable if ratio ≤ 5 (or ≤ 7 for capital-intensive firms). Ratios beyond these thresholds or negative ratios due to negative EBITDA imply non-sustainable debt.
  - Interest Coverage Ratio (ICR):
    - ICR = EBIT / interest payments.
    - Absolute minimum for debt service sustainability is ratio ≥ 1; some studies use threshold of 1.5.
    - ICR can be stressed by lowering EBIT and/or increasing interest payments.
  - Sustainable cash balance:
    - Starting balance uses cash only (sample lacks short-term investments).
    - Intra-year variation determined by:
      (i) income from operations (EBIT);
      (ii) net cost of financing (interest payments, disaggregated by short- and long-term);
      (iii) net inflow of financing (rollover of maturing funding);
      (iv) depreciation and amortization;
      (v) capital expenditure (CAPEX).
    - Depreciation and amortization and cost of long-term debt assumed pre-determined; depreciation & amortization not stressed.
    - Capital expenditure is capped by depreciation and amortization (lower of the two used).

### Corporate Risk Indicators — Results and recommendations
- Illustrative results (sample of 22 firms; not representative):
  - Debt-to-EBITDA (Table 3):
    - Many firms are overindebted even before stress: almost all firms have ratio > 5 (or 7) or negative earnings under baseline.
    - Example rows preserved exactly (selected):
      - Firm 1: Net debt 126,800,466; EBITDA 40,627,990; Debt-to-EBITDA Ratio (baseline) 3.12 → Net debt stressed 152,160,559; EBITDA stressed 28,439,593; Debt-to-EBITDA Ratio (stressed) 5.35
      - Firm 3: Net debt 453,682,000; EBITDA 77,255,000; Debt-to-EBITDA Ratio (baseline) 5.87 → Net debt stressed 544,418,400; EBITDA stressed 54,078,500; Debt-to-EBITDA Ratio (stressed) 10.07
      - Firm 12: Net debt 87,529,378; EBITDA 658,297; Debt-to-EBITDA Ratio (baseline) 132.96 → Net debt stressed 105,035,254; EBITDA stressed 460,808; Debt-to-EBITDA Ratio (stressed) 227.94
    - Stress factor applied: 20% -30%.
  - Interest Coverage Ratio (Table 4):
    - Under baseline and stressed scenarios (stress factor: -40% EBIT, 5% interest cost increase), about half the sample firms are above threshold of 1.5 under baseline conditions (two firms do not report interest payments).
    - Example rows preserved exactly (selected):
      - Firm 1: EBIT 27,285,011; Interest payments 3,558,962; Interest Coverage Ratio (baseline) 7.67 → EBIT stressed 16,371,007; Interest payments stressed 3,736,910; Interest Coverage Ratio (stressed) 4.38
      - Firm 3: EBIT 25,144,000; Interest payments 11,273,000; Interest Coverage Ratio (baseline) 2.23 → EBIT stressed 15,086,400; Interest payments stressed 11,836,650; Interest Coverage Ratio (stressed) 1.27
      - Firm 22: EBIT 198,855,621; Interest payments 28,056,312; Interest Coverage Ratio (baseline) 7.09 → EBIT stressed 119,313,373; Interest payments stressed 29,459,128; Interest Coverage Ratio (stressed) 4.05
    - Stress factor applied: -40% (EBIT) and 5% (interest payments).
  - Cash balance stress:
    - Assuming a 40 percent drop in EBIT, a five percent higher cost of short-term financing, and a 95 percent rollover of short-term debt, about one-third of sample firms exhaust their cash under stress.
- Recommendations to improve corporate stress analysis:
  - Revitalize effort to collect financial statement data of the largest Panamanian firms to obtain a more representative sample.
  - SBP should:
    - Collect necessary data for the 100 largest firms as measured by the amount of bank credit.
    - Obtain more granular data on short-term investments and short-term obligations; include economic sector and conglomerate membership.
    - Identify and correct data issues: missing observations, implausible zeros, and observations with unexpected (negative) sign.
  - Once data collection is finished, expand analysis:
    - With 100 observations obtain meaningful distributions of indicators (as in referenced corporate risk analyses).
    - Segment sample by firms mainly bank-financed vs. more diversified funding, or by main economic sector.
    - Use firm-level corporate stress test results as early warning indicators in SBP supervisory process.

*Source: SBP data and mission calculations; IMF Technical Assistance Report (excerpts provided).*

### 38. An enhanced liquidity stress testing tool was provided to the SBP.

### 38. An enhanced liquidity stress testing tool was provided to the SBP.

### Description of the tool
- The tool expands on previously existing cash flow based liquidity stress test models in three dimensions:
  - (i) by properly computing the forward flows of maturing and reinvested/rolled over assets and liabilities along the liquidity simulation horizon;
  - (ii) by operating with only one counterbalancing capacity, and its underlying detailed asset components, to serve as buffers and as cash inflow-generating at the same time; and
  - (iii) by incorporating a calculation for losses that may arise when bonds in the held-to-maturity investment category would have to be sold, and therefore result in adverse feedback to banks’ solvency ratios.
- The tool was used in the 2023 Panama FSAP and for other recent FSAPs.
- One session on the liquidity stress test model topic was delivered by Mr. Marco Gross (Monetary and Capital Markets Department of the IMF).7

### Versions and data requirements
- Three versions of the tool suite were provided, differing by required input data:
  - Simple version:
    - Requires no maturity profile data as inputs for assets or liabilities.
  - Intermediate version:
    - Requires maturity profile data for liabilities.
  - Advanced version:
    - Requires maturity ladder data for both assets and liabilities.
- All three versions require selected market risk parameters as inputs, including:
  - current market values,
  - book values,
  - current yields,
  - durations.
- The advanced model version was used for the 2023 FSAP and is most likely the one the SBP will operate going forward.
- A guidance note detailing the methodology will be made available to the SBP in due course.

### Data granularity and special considerations
- The liquidity tool uses a granular set of data and requires detailed information on:
  - various types of funding (deposits, available credit lines and other funding), disaggregated by several maturity buckets up to 180 days;
  - cash and cash equivalents;
  - bonds (separated by maturity and by fair value and amortized cost);
  - other securities, including stocks.
- Special attention item for the SBP:
  - liquidity generating credit lines that Panamanian banks maintain with their foreign correspondent banks (given that Panama has no central bank).
- Bespoke counterfactuals can be and were being conducted with the model during the FSAP in this context.

### Illustrative results (mock sample) and model outputs
- Assuming a certain pecking order of asset use to meet outflows, the tool calculated:
  - the stage at which the sample bank ran out of liquid assets in a certain maturity bucket;
  - the liquidity gap;
  - losses caused by fire sales of HTM securities that should feed into the solvency stress test.

### Recommendations for SBP surveillance and next steps
- Continue compiling all relevant, detailed data to allow for a comprehensive liquidity risk analysis of Panamanian banks, supported by the liquidity stress test models provided as part of the TA.
- The Financial Research Directorate will familiarize itself with the tool and then use it to conduct liquidity stress tests. (See Appendix I: Proposed Methodology for Liquidity Stress — Completed.)

*IMF Technical Assistance Report — content unit: 38. An enhanced liquidity stress testing tool was provided to the SBP.*

---


_Source: https://www.imf.org/-/media/files/publications/tar/2024/english/tarea2024075-print-pdf.pdf_