## Chapter 3 — Online Annexes: COVID-19 and the Corporate-Sector Outlook

## Source details

**Canonical URL:** [Chapter 3 — Online Annexes: COVID-19 and the Corporate-Sector Outlook](https://www.imf.org/-/media/files/publications/reo/mcd-cca/2021/october/reo-october-2021-corporate-chapter-online-annexes.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/reo/mcd-cca/2021/october/reo-october-2021-corporate-chapter-online-annexes.pdf.md)
- [Structured JSON version](/-/media/files/publications/reo/mcd-cca/2021/october/reo-october-2021-corporate-chapter-online-annexes.pdf.json)

---

### Online Annex 3.1 — Digitalization and Resilience during COVID-19: Firm-Level Evidence

- Data sources
  - Pre-pandemic firm-level data: EBRD-EIB-World Bank Business Environment and Enterprise Performance Survey (BEEPS), last survey (2018-20).
  - Pandemic-period firm-level data: World Bank COVID-19 Follow-up Enterprise Survey (ES COVID Survey) and Business Pulse Survey (BPS).
  - Coverage: 13 countries and territories in ME&CA; sample of over 18,000 firms.
  - Digital proxy: digital dummy constructed from a principal component analysis on two pre-pandemic proxies: internet presence (use of a website) and adoption of foreign technologies.

- Identification strategy
  - Difference-in-differences framework using COVID-19 as exogenous supply shock.
  - Estimated equation (variables described in text): dependent variable Y_i,t = log(total sales); Digital_i = dummy for digitally enabled firms pre-pandemic; COVID_t = 1 in 2020, 0 otherwise; X_i,t = firm controls (including firm size and age); regressions include country- and industry-fixed effects.

- Key empirical finding on role of digital connectivity
  - COVID-19 crisis negatively affected sales across firms.
  - Digitally enabled firms have, on average, higher sales than digitally constrained peers.
  - Digital connectivity mitigated the pandemic’s impact: digitally enabled firms experienced a decline in their sales that is about 4 percentage points lower than that of digitally constrained firms.
  - Results robust to controls for access to credit, firm ownership, and innovation.

- Baseline regression highlights (Annex Table 3.1.1)
  - COVID-19 coefficient examples: -0.167***, -0.156***, -0.159***, -0.168***, -0.170***, -0.167***, -0.168***, -0.168*** (standard errors reported in parentheses).
  - Digital coefficient examples: 2.266***, 2.242***, 1.819***, 2.207***, 1.793***, 1.757***, 1.786***, 1.748***.
  - COVID-19 * Digital interaction: 0.044***, 0.038***, 0.041***, 0.035***, 0.038***, 0.033***, 0.038*** (standard errors in parentheses).
  - Controls: small firm and young firm dummies reported (examples: Small -1.346***, Young -0.432***).
  - Observations reported across specifications: 18074, 18074, 18074, 18074, 18074, 18070, 18070.
  - R-squared examples: 0.002, 0.300, 0.301, 0.395, 0.312, 0.404, 0.423, 0.410, 0.428.
  - Notes: Robust standard errors in parentheses. *, **, *** indicate statistical significance at the 10%, 5% and 1% levels.

### Online Annex 3.2 — Corporate-Sector Stress Testing

- Purpose and scope
  - Framework for assessing corporate performance under stress-test scenarios using publicly listed firms (data availability constraint).
  - Sample: firm-level data from S&P Capital IQ (Compustat), historical financial statements 2002-20.
  - Sample size: more than 660 publicly listed firms from 11 countries.
  - Aggregation dimensions: country groups (oil-importing and oil-exporting), sectors (high- and low-contact-intensive), firm size (small and large).

- Country and industry coverage (Annex Table 3.2.1)
  - Country coverage split: Oil Importing — Egypt, Jordan, Morocco, Pakistan, Tunisia. Oil Exporting — Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, United Arab Emirates.
  - Industry classification (Fama-French) mapped to low-contact-intensity and high-contact-intensity groups (examples: Consumer Non-Durables in low-contact-intensity; Shops, Wholesale, Retail in high-contact-intensity).

- Empirical strategy and projection model
  - Adaptation of Tressel and Ding’s (2021) scenario-based stress-testing tool.
  - Regression sample split: oil-exporting and oil-importing countries; estimation sample ends in 2019.
  - Dynamic OLS panel-data regression specification:
    - Dependent variable Y_i,j,c,t = ROA or annual sales growth.
    - X_i,j,c,t-1 includes lagged firm-level controls: firm size, asset tangibility, sales turnover, leverage, sales growth, ROA.
    - G_c,t = annual real GDP growth (country-year).
    - α_c and α_j = country- and industry-fixed effects.

- Variable definitions (Annex Table 3.2.2)
  - Real GDP Growth: Country-level annual growth of real GDP (IMF World Economic Outlook).
  - ROA: Pre-tax net income to total assets (Capital IQ and IMF Staff calculations).
  - Leverage: Total debt to total assets.
  - Sales Growth: Annual growth rate of sales revenues.
  - Firm Size: Natural logarithm of total assets.
  - Asset Tangibility: Fixed assets to total assets.
  - Asset Turnover: Annual sales revenues to total assets.

- 2019 summary statistics (Annex Table 3.2.3)
  - Oil Importers Subsample (variables: 1st Quartile / Median / 3rd Quartile / Mean / Standard Deviation)
    - Real GDP Growth (%) : 1.91 / 1.91 / 5.55 / 2.89 / 1.65
    - ROA (%) : 0.25 / 5.31 / 10.33 / 5.5 / 9.66
    - Leverage (%) : 7.48 / 25.35 / 40.82 / 26.34 / 20.59
    - Sales Growth (%) : -5.41 / 6.89 / 20.11 / 7.88 / 26.73
    - Firm Size : 6.54 / 8.14 / 9.42 / 7.88 / 2.11
    - Asset Tangibility (%) : 23.63 / 38.87 / 55.35 / 40.62 / 22.26
    - Asset Turnover (%) : 0.48 / 0.8 / 1.11 / 0.85 / 0.54
  - Oil Exporters Subsample (variables: 1st Quartile / Median / 3rd Quartile / Mean / Standard Deviation)
    - Real GDP Growth (%) : 0.33 / 0.33 / 0.43 / 0.52 / 0.64
    - ROA (%) : -0.39 / 3.13 / 7.57 / 3.22 / 8.08
    - Leverage (%) : 7.45 / 21.57 / 36.22 / 24.22 / 18.6
    - Sales Growth (%) : -10.47 / 1.12 / 12.43 / 2.68 / 24.56
    - Firm Size : 5.54 / 7.18 / 8.27 / 6.95 / 2.09
    - Asset Tangibility (%) : 27.12 / 46.44 / 60.57 / 44.0 / 22.35
    - Asset Turnover (%) : 0.26 / 0.43 / 0.75 / 0.53 / 0.39

- Projection mechanics
  - Project ROA and sales growth over 2021-23 using estimated regression coefficients, precrisis (2019) firm structural characteristics, 2020 actual values of ROA and sales growth, IMF WEO GDP growth forecasts for 2021 onward, and estimated country- and industry-fixed effects.
  - Iterative projection: 2021 projection uses 2020 actuals; 2022 projection uses projected 2021 values; 2023 projection uses projected 2022 values.
  - Accounting identities to derive cash balance, debt evolution, interest coverage ratio (ICR), and equity:
    - Cash Balance_t+1 = EBIT_t+1 – Interest Expenses_t+1 – Taxes_t+1 + Cash Balance_t (negative value indicates liquidity shortage).
    - Debt_t+1 = Debt_t - Cash Balance_t+1 if Cash Balance_t+1 < 0; equal to Debt_t otherwise.
    - ICR_t = EBIT_t / Interest Expenses_t.
    - Equity_t+1 = Equity_t + Retained Earnings_t+1 (retained earnings = net income under assumption of no new equity issuance and no dividends).

- Baseline scenario assumptions
  - Projected pre-tax net income obtained by multiplying projected ROA by 2019 total assets.
  - Taxes projected based on 2020 effective income tax rate.
  - Retained earnings calculated as net income after interest expenses and taxes (using projected ROA times 2019 total assets minus projected taxes).

- Adverse scenario assumptions
  - Subdued recovery and symmetric withdrawal of policy support starting in 2022.
  - Real GDP growth assumed at one standard deviation below the IMF’s World Economic Outlook forecast.
  - Policy withdrawal modeled as:
    - Return to precrisis effective income tax rates.
    - Rise in the effective interest rate by 200 basis points relative to 2020 rates.

- Vulnerability indicators and definitions
  - Liquidity stress: firms with negative cash balances (would need to borrow or adjust to avoid default).
  - Insolvency risk: firms with negative equity positions (liabilities larger than assets).
  - “Firms at risk”: share of firms with cash balances or equity below zero.
  - “Debt at risk”: debt of firms at risk as a fraction of total non-financial corporate debt.
  - Calculation formulas provided for firm-at-risk and debt-at-risk metrics using indicator 1(.) for negative cash balance and summations over group firms (N_g).

- Regression results (Annex Table 3.2.4) — main coefficients and interpretation
  - Real GDP Growth coefficients (association with ROA and Sales Growth):
    - ROA: 0.139*** (oil exporters), 0.246*** (oil importers).
    - Sales Growth: 1.55*** (oil exporters), 1.074*** (oil importers).
    - Interpretation example in text: A one percentage point increase in real GDP growth is associated with a rise in firm profitability by 0.14 and 0.25 percentage point in oil exporters and oil importers, respectively; and with additional sales growth of 1.6 and 1.1 percentage points in oil exporters and oil importers, respectively.
  - Persistence of profitability shocks:
    - Lagged ROA coefficients: 0.723*** (oil exporters), 0.663*** (oil importers) — indicating large persistence of shocks to ROA.
  - Other reported coefficients (examples):
    - Lagged Sales Growth: 0.011** (oil exporters), -0.003 (oil importers).
    - Lagged Leverage: -0.022** (oil exporters), -0.037*** (oil importers).
    - Lagged Firm Size: 0.133 (oil exporters), 0.410*** (oil importers) for ROA; -1.390*** (oil exporters), -0.933*** (oil importers) for Sales Growth.
    - Lagged Asset Turnover: 1.588*** (oil exporters), 0.855*** (oil importers) for ROA; -3.47*** (oil exporters), -3.417*** (oil importers) for Sales Growth.
    - Lagged Asset Tangibility: 0.018*** (oil exporters), -0.0009 (oil importers) for ROA; 0.02 (oil exporters), 0.018 (oil importers) for Sales Growth.
  - Regression diagnostics:
    - Observations: ROA regressions — 2,593 (oil exporters), 5,378 (oil importers); Sales Growth regressions — 2,817 (oil exporters), 5,991 (oil importers).
    - R-squared examples: ROA — 0.637 (oil exporters), 0.568 (oil importers); Sales Growth — 0.077 (oil exporters), 0.02 (oil importers).
    - Note: t-statistics reported in parentheses in table; ***, **, * indicate statistical significance at the 1, 5 and 10 percent levels.

*Source: Online annexes to Chapter 3, Regional Economic Outlook: Middle East and Central Asia, October 2021 (IMF).*

---


_Source: https://www.imf.org/-/media/files/publications/reo/mcd-cca/2021/october/reo-october-2021-corporate-chapter-online-annexes.pdf_