## _wp15285

## Source details

**Canonical URL:** [_wp15285](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15285.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15285.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15285.pdf.json)

---

### I. Research question, scope, and data
- Research focus:
  - Differential responses of corporate investment to changes in global interest rates and financial volatility depending on firms’ financial strength.
  - Identification of two channels: a financing channel (cost of external finance) and a wait-and-see / real-options channel (non-convex adjustment costs).
  - Exploration of country heterogeneity to gauge whether macroeconomic fundamentals influence these channels.
- Data and measures:
  - Sample: 17,000 nonfinancial listed firms in 38 emerging markets over 1990–2013.
  - External financial-condition measures: changes in (10–year) U.S. government bond yields and uncertainty/volatility measured by the VIX.
  - Strategy: standard Q model of investment augmented with financial-strength variables and interaction terms with changes in interest rates and volatility.
- Key firm-level measures (definitions preserved):
  - Net leverage (xNLev): total debt net of cash stocks, divided by total common equity; expected sign in investment regressions: negative.
  - Cash flow ratio (xCF): net income plus depreciation and depletion within the year, normalized by total (net) property, plant, and equipment at the beginning of the year; treated as contemporaneous.
  - Investment ratio (ι): capital expenditures within the year normalized by capital stock (total net property, plant, and equipment) at the beginning of the year.
  - Tobin’s Q: (market capitalization minus total common equity plus total assets) divided by total assets.
- Sample processing:
  - More than 17,000 listed firms (in 14 sectors) and around 130,000 firm-(country)-year observations after exclusions.
  - Accounting variables winsorized at the 5th and 95th percentiles; macro variables trimmed at the 1st and 99th percentiles for each emerging country.

### II. Key empirical findings (firm-level)
- Coexistence and heterogeneity of channels:
  - Evidence for coexistence of both financing and wait-and-see channels; intensity depends on shock type and firms’ balance sheets.
- Interest-rate shocks:
  - Results consistent with the financing channel.
  - Firms with higher leverage reduce investment disproportionately more in response to an increase in interest rates.
  - The marginal propensity to invest (MPI) out of cash flows increases in response to higher interest rates.
- Volatility/uncertainty shocks (VIX and alternatives):
  - Mixed evidence—both channels operate:
    - Financing channel: more levered firms reduce investment disproportionately more, especially when closer to default boundaries (higher net leverage).
    - Wait-and-see channel: MPI out of cash flows decreases, consistent with increased option value of waiting.
  - Dominance of channel depends on firm balance sheets:
    - Financing channel more significant and wait-and-see less significant the more leveraged firms are.
    - Low-leverage firms cut investment voluntarily to wait and see; MPI decreases when volatility spikes.
- Heterogeneity magnitude:
  - A firm at the top 5th percentile of net leverage exhibits an investment response 7 times larger than a firm at the bottom 5th percentile (illustrated by p95/p5 for VIX = 6.67).
  - A one standard-deviation increase in interest rates or volatility leads to a median reduction in investment equivalent to 10–15 percent of the observed average reduction in investment during the Global Financial Crisis (GFC).
  - A one-standard deviation shock could reduce investment-to-capital ratio for median firm by around 0.5 or 0.7 percentage points for interest rate or uncertainty, respectively.

### III. Numerical summary statistics (selected, preserved exactly as in source Table 1)
- Panel-A: Corporate Variables (X)
  - Investment Ratio: Obs. 92,901; Mean 0.206; SD 0.179; Min 0.018; Max 0.66.
  - Tobin’s Q: Obs. 109,085; Mean 1.293; SD 0.61; Min 0.606; Max 2.888.
  - Net Leverage: Obs. 92,119; Mean 0.336; SD 0.583; Min -0.526; Max 1.657.
  - Cash Flow Ratio: Obs. 87,790; Mean 0.313; SD 0.333; Min -0.163; Max 1.228.
- Panel-B: Shock Variables (S)
  - Yield Shocks: Obs. 95,277; Mean -0.002; SD 0.005; Min -0.011; Max 0.012.
  - Yield Shocks (GSW): Obs. 95,277; Mean -0.003; SD 0.005; Min -0.012; Max 0.01.
  - VIX: Obs. 127,558; Mean 0.211; SD 0.065; Min 0.124; Max 0.325.
  - U.S. Realized Volatility (RVOL): Obs. 127,558; Mean 0.186; SD 0.086; Min 0.074; Max 0.4.
  - JLN Measure (Macro): Obs. 127,558; Mean 0.978; SD 0.06; Min 0.899; Max 1.115.
  - JLN Measure (Firm): Obs. 108,330; Mean 0.921; SD 0.084; Min 0.768; Max 1.086.
- Panel-C: Macro Variables (F)
  - Foreign Reserves/ST External Debt: Obs. 104,890; Mean 4.52; SD 4.333; Min 0.464; Max 74.92.
  - Public Debt/GDP: Obs. 120,827; Mean 0.452; SD 0.206; Min 0.082; Max 1.032.
  - Private Bank Credit/GDP: Obs. 89,098; Mean 0.739; SD 0.341; Min 0.135; Max 1.394.
  - Chinn-Ito Index of Openness: Obs. 110,356; Mean 0.413; SD 0.288; Min 0; Max 1.

### IV. Empirical strategy and main estimated coefficients (preserved examples)
- Estimation approach:
  - Panel regressions (firm-year) with firm fixed-effects; standard errors clustered at country level; robustness with two-way fixed-effects and two-way clustering.
- Representative coefficients (Table 2, column (1)):
  - (Lag) Tobin’s Q 0.059*** (0.008).
  - (Lag) Net Leverage -0.064*** (0.006).
  - Cash Flow Ratio 0.133*** (0.010).
  - ∆ U.S. Interest rates -0.737*** (0.099).
  - Observations 61,320; R-squared 0.121; Number of Firms 11,282.
- Interaction findings (Equation (2), selected reported coefficients):
  - Cash Flow * interest-rate shock example: 0.698** (0.321) in baseline (positive, financing-channel-consistent).
  - Cash Flow * uncertainty shock example: -0.136** (0.055) in baseline (negative, wait-and-see-consistent).
  - Subsample (net leverage above 75th percentile): (Lag) Net Leverage * interest-rate shock -0.907*** (0.321).
  - Joint model (Table 6, column (2) with lagged VIX): S1 (interest rate shock) -0.944*** (0.119); S2 (uncertainty shock) -0.069*** (0.024); (Lag) Net Leverage * S2 -0.104*** (0.029); Cash Flow * S2 -0.129** (0.053).

### V. Economic-significance table excerpts (Table 7, preserved exactly)
- Percentage point reduction in investment rates in response to one standard deviation increase in U.S. interest rates or the VIX:
  - Net Leverage percentiles and impacts:
    - p5: Interest Rates -0.41; VIX -0.22.
    - p25: Interest Rates -0.45; VIX -0.48.
    - p50 (median): Interest Rates -0.47; VIX -0.67.
    - p75: Interest Rates -0.50; VIX -0.92.
    - p95: Interest Rates -0.58; VIX -1.46.
  - Ratios:
    - p75 / p25: Interest Rates 1.13; VIX 1.93.
    - p95 / p5: Interest Rates 1.40; VIX 6.67.

### VI. Country-level heterogeneity: role of macro fundamentals (estimation and interpretation)
- Macro fundamentals examined: Foreign Reserves/ST External Debt (F1), Public Debt/GDP (F2), Private Bank Credit/GDP (F3), and financial account openness (Chinn-Ito).
- Key regression outputs (Table 8, selected coefficients preserved):
  - (Lag) Q: 0.057*** (0.009) through 0.056*** (0.011) in interest-rate columns; 0.055*** (0.007) through 0.051*** (0.009) in uncertainty columns.
  - Shocks (examples): S: -1.485*** (0.430) and -2.292** (1.017) in interest-rate columns; S: -0.070 (0.049) in an uncertainty column.
  - F2: Public Debt/GDP: -0.172*** (0.044) and -0.181*** (0.053) in relevant columns.
  - (Lag) Net Leverage: -0.068*** (0.008) and -0.051** (0.023) in interest-rate columns.
  - Cash Flow Ratio: 0.132*** (0.011) and 0.179*** (0.034) in interest-rate columns.
  - Observations and fit examples: Observations: 52,315; 58,959; 39,129 (columns vary). R-squared: 0.119, 0.123, 0.118.
- Illustration of quantitative heterogeneity:
  - Following an increase in U.S. long-term interest rates, the contraction in investment by a firm with median net leverage is three times larger if the firm is in a country with public debt of 75 percent of GDP than if located in a country with public debt of around 32 percent of GDP.
- Formal slope-heterogeneity tests (Table 9, selected coefficients preserved):
  - Interest-rate shocks with all macro variables and openness included:
    - Foreign Reserves: 0.261*** (0.084).
    - Public Debt: -7.346*** (1.814).
    - Credit Deepening: 1.677** (0.757).
  - Uncertainty shocks: no statistically significant heterogeneity across country fundamentals (example coefficients: Foreign Reserves: 0.002 (0.006); Public Debt: 0.123 (0.076); Credit Deepening: -0.014 (0.068)).
- Interpretation:
  - Low public debt, high international reserves, and deeper financial markets mitigate firms’ investment responses to increases in U.S. interest rates.
  - No statistically significant role for macroeconomic fundamentals in shaping the response of investment to higher uncertainty.

### VII. Robustness checks
- Alternative shock measures: GSW yields and realized volatility (RVOL) produce nearly identical results (Table 4, columns (2) and (8)).
- Fixed-effects: results robust to year fixed effects and country-year fixed effects (Table 4, columns (3), (4), (9), (10)).
- Clustering: two-way clustering (country and year) does not change statistical significance (Table 4, columns (5) and (11)).
- Sample composition: excluding all Chinese and Indian firms yields similar results (Table 4, columns (6) and (12)).
- Alternative uncertainty measures: JLN (macro) and JLN (firm) produce qualitatively similar results (Table 5); example: (Lag) Net Leverage * Uncertainty -0.065* (0.035) for JLN (macro) and -0.106*** (0.028) for JLN (firm).
- Endogeneity: model passes Sargan-Hansen endogeneity test for right-hand-side firm-level variables when firm fixed effects are included (results available from authors upon request).

### VIII. Theoretical framing and intuition
- Investment vs net leverage:
  - Downward-sloping relationship between investment and net leverage consistent with financial market imperfections.
  - External shocks cause non-parallel shifts: more-levered firms face larger increases in external finance costs and thus larger cuts in investment.
- Investment vs cash flows (MPI):
  - Interest-rate shock: MPI increases (financing channel).
  - Uncertainty shock: financing channel raises MPI while wait-and-see channel lowers MPI; net effect depends on firm leverage.
  - Empirical illustrations: OLS and Nadaraya-Watson fits of investment on lagged net leverage during high vs. low uncertainty show higher net leverage firms reduce investment more when uncertainty increases.

### IX. Policy recommendations (preserved)
- Maintain strong macroeconomic fundamentals—specifically:
  - Keep public debt low.
  - Preserve high international reserves relative to short-term external obligations.
  - Promote deeper financial markets (higher private credit by deposit banks to GDP).
- Rationale: these policies help mitigate the effect of higher global interest rates on corporate investment through the financing channel by reducing aggregate risk, lessening financial frictions, and improving ability to substitute domestic financing or conduct countercyclical policies.

*Source: _wp15285 - References and Section IV excerpt (PDF chapter/section).*

### References .............................................................................................................

### _wp15285 - References

### I. Introduction — research question and scope
- Periods of tighter global financial conditions are often followed by lower corporate investment in emerging markets.
- Research focus:
  - Differential responses of corporate investment to changes in global interest rates and financial volatility depending on firms’ financial strength.
  - Identification of two channels: a financing channel (cost of external finance) and a wait-and-see / real-options channel (non-convex adjustment costs).
  - Exploration of country heterogeneity to gauge whether macroeconomic fundamentals influence these channels.
- Data and measures:
  - Sample: 17,000 nonfinancial listed firms in 38 emerging markets over 1990–2013.
  - External financial-condition measures: changes in (10–year) U.S. government bond yields and uncertainty/volatility measured by the VIX.
  - Strategy: standard Q model of investment augmented with financial-strength variables and interaction terms with changes in interest rates and volatility.

### Key empirical findings
- Evidence for coexistence of both financing and wait-and-see channels; intensity depends on shock type and firms’ balance sheets.
- Interest-rate shocks:
  - Results consistent with the financing channel.
  - Firms with higher leverage reduce investment disproportionately more in response to an increase in interest rates.
  - The marginal propensity to invest (MPI) out of cash flows increases in response to higher interest rates.
- Volatility/uncertainty shocks (VIX):
  - Mixed evidence—both channels operate:
    - Financing channel: more levered firms reduce investment disproportionately more, especially when closer to default boundaries (higher net leverage).
    - Wait-and-see channel: MPI out of cash flows decreases, consistent with increased option value of waiting.
  - Dominance of channel depends on firm balance sheets:
    - Financing channel more significant and wait-and-see less significant the more leveraged firms are.
    - Low-leverage firms cut investment voluntarily to wait and see; MPI decreases when volatility spikes.
- Magnitude and heterogeneity:
  - A one standard-deviation increase in interest rates or volatility leads to a median reduction in investment equivalent to 10–15 percent of the observed average reduction in investment during the Global Financial Crisis (GFC).
  - Differential response: a firm at the top 5th percentile of net leverage exhibits an investment response 7 times larger than a firm at the bottom 5th percentile.

### Role of country macroeconomic fundamentals
- Macroeconomic fundamentals that weaken the contraction of investment in response to higher interest rates:
  - Lower public debt.
  - Higher foreign reserves.
  - Deeper financial markets.
- Quantitative example:
  - Following an increase in U.S. long-term interest rates, the contraction in investment by a firm with median net leverage is three times larger if the firm is located in a country with public debt of 75 percent of GDP than if located in a country with public debt of around 32 percent of GDP.
- Possible mechanisms:
  - Firms in countries with weaker fundamentals face relatively higher downside risks, exacerbating financial frictions and sensitivity to higher interest rates via the financing channel.
  - In countries with stronger fundamentals:
    - Deeper financial markets may offer better opportunities to substitute domestic financing for external financing when global credit conditions tighten.
    - Global investors may rebalance portfolios toward firms in countries with stronger fundamentals.
    - Governments may have higher ability to conduct countercyclical macroeconomic policies.
- For uncertainty shocks:
  - No statistically significant role for macroeconomic fundamentals in shaping the response of investment to higher uncertainty.
  - Suggested intuition: as fundamentals strengthen, the financing channel weakens while the real-option (wait-and-see) channel strengthens, leaving aggregate effects of uncertainty on investment similar across countries.

### Theoretical framing and graphical intuition
- Investment vs net leverage:
  - Downward-sloping relationship between investment and net leverage consistent with financial market imperfections.
  - External shocks can cause non-parallel shifts: more-levered firms face larger increases in external finance costs and thus larger cuts in investment.
- Investment vs cash flows (MPI):
  - Slope interpreted as marginal propensity to invest (MPI) out of cash flows.
  - Interest-rate shock: MPI increases (financing channel).
  - Uncertainty shock: two opposing effects
    - Financing channel raises MPI (as firms with more cash cut less).
    - Wait-and-see (real-options) channel lowers MPI because firms with stronger cash flows can afford to wait, reducing responsiveness to extra cash.
  - Net effect on MPI depends on firm leverage; more-levered firms exhibit financing-channel dominance, less-levered firms exhibit wait-and-see dominance.
- Empirical illustrations:
  - OLS and Nadaraya-Watson non-parametric fits of investment on lagged net leverage during high vs. low uncertainty periods show that firms with higher net leverage reduce investment more than firms with lower net leverage when uncertainty increases.
  - High vs low uncertainty defined by whether the VIX is above or below the median.

### Empirical strategy (model and variables)
- Starting model: neoclassical Q-model of investment augmented by firm balance-sheet variables.
- Key regression variables:
  - Investment-to-capital-stock rate of firm i in country j during year t.
  - Firm average Tobin’s Q (proxied by market-to-book), lagged (t-1).
  - Net leverage: total debt net of cash stock, as a share of total common equity.
  - Cash flows normalized by capital stock.
  - S: volatility (VIX) or changes in U.S. benchmark (10–year) interest rates.
- Identification strengths:
  - Use of common and exogenous global measures (U.S. yields and VIX) to mitigate reverse causality concerns.
  - Focus on firm-differential behavior along financial-strength dimension reduces concern that results are driven solely by a standard neoclassical cost-of-capital channel where firm financial structure plays no role.

### Contribution to literature
- Differentiates from prior work that focused on average or aggregate responses by studying firm-level heterogeneity tied to balance-sheet strength.
- Reconciles two strands of the uncertainty-transmission debate:
  - Canonical real-options view (Bloom and others): higher volatility raises option value of waiting and reduces voluntary investment.
  - Financial-frictions view (Arellano, Bai, and Kehoe; Christiano, Motto, and Rostagno; others): higher volatility increases costs of funds via market imperfections and depresses investment.
  - This paper shows both channels coexist and dominance depends on firms’ balance sheets.
- Extends the literature on investment-to-cash sensitivity by showing sensitivity is state-dependent and influenced by changes in global interest rates and uncertainty.
- Novel quantification of the role of macroeconomic fundamentals in shaping corporate investment responses to higher global interest rates or volatility.

*Source: _wp15285 - References (excerpt of PDF content).*

### Section IV below. As suggested by the existing empirical literature, we expect   to be

### _wp15285 - Section IV below. As suggested by the existing empirical literature, we expect   to be

### Firm financial measures and hypotheses
- Net leverage (xNLev) defined as total debt net of cash stocks, divided by total common equity; expected sign in investment regressions: negative.
- Cash flow ratio (xCF) defined as net income plus depreciation and depletion within the year, normalized by total (net) property, plant, and equipment at the beginning of the year; treated as contemporaneous.
- Leverage is lagged for two stated reasons: (i) to avoid mechanical feedback between investment and net leverage when borrowing to invest, and (ii) to reflect timing of firm decisions where investment in year t reflects plans initiated in previous period.
- Hypothesis 1: Firms with relatively higher net leverage reduce investment more aggressively in response to higher interest rates or volatility (interaction term between shock S and lagged net leverage expected negative).
- Hypothesis 2: Higher interest rates increase the marginal propensity to invest (MPI) out of cash flows (interaction between shock S and cash flow, ς2, expected positive).
- Hypothesis 3: Higher uncertainty could increase or decrease MPI out of cash flows: under financing channel ς2 positive; under wait-and-see/real-options channel ς2 negative; wait-and-see expected stronger for firms with stronger cash flows.

### Data and sample
- Data source: World Scope; sample covers 38 emerging markets from 1990 to 2013 (annual frequency), compiled in Magud and Sosa (2015).
- Exclusions: firms with negative core accounting variables and firms in financial, utility, and public sectors.
- Sample size after exclusions: more than 17,000 listed firms (in 14 sectors) and around 130,000 firm-(country)-year observations.
- Winsorization and trimming: accounting variables winsorized at the 5th and 95th percentiles; macroeconomic variables trimmed at the 1st and 99th percentiles for each emerging country.
- Core accounting variables (used to screen): total asset, total (net) property, plant, and equipment, total common equity, cash stock, and capital expenditures.

### Variable definitions and shock measures
- Investment ratio (ι): capital expenditures within the year normalized by capital stock (total net property, plant, and equipment) at the beginning of the year.
- Tobin’s Q: (market capitalization minus total common equity plus total assets) divided by total assets.
- Interest-rate shocks: change in the annual average of the daily yield of the 10-year, constant-maturity U.S. Treasury bond from FRED; robustness uses change in zero-coupon, continuously compounding yields (Gurkaynak, Sack, and Wright (2006), denoted GSW).
- Uncertainty shocks: VIX (annual average of daily values, divided by 100 for scale comparability) or realized volatility (annual standard deviation of daily returns of CRSP total equity market portfolio); robustness uses JLN measures (Jurado and others (2015)): JLN (macro) and JLN (firm) (both from U.S. data).
- Macro controls: U.S. real GDP growth; country-level data from IMF World Economic Outlook and International Financial Statistics; bank credit to GDP from World Bank Financial Development database; Chinn-Ito index normalized between 0 and 1 for financial account openness.

### Summary statistics (selected, preserved exactly as in source Table 1)
- Panel-A: Corporate Variables (X)
  - Investment Ratio: Obs. 92,901; Mean 0.206; SD 0.179; Min 0.018; Max 0.66.
  - Tobin’s Q: Obs. 109,085; Mean 1.293; SD 0.61; Min 0.606; Max 2.888.
  - Net Leverage: Obs. 92,119; Mean 0.336; SD 0.583; Min -0.526; Max 1.657.
  - Cash Flow Ratio: Obs. 87,790; Mean 0.313; SD 0.333; Min -0.163; Max 1.228.
- Panel-B: Shock Variables (S)
  - Yield Shocks: Obs. 95,277; Mean -0.002; SD 0.005; Min -0.011; Max 0.012.
  - Yield Shocks (GSW): Obs. 95,277; Mean -0.003; SD 0.005; Min -0.012; Max 0.01.
  - VIX: Obs. 127,558; Mean 0.211; SD 0.065; Min 0.124; Max 0.325.
  - U.S. Realized Volatility (RVOL): Obs. 127,558; Mean 0.186; SD 0.086; Min 0.074; Max 0.4.
  - JLN Measure (Macro): Obs. 127,558; Mean 0.978; SD 0.06; Min 0.899; Max 1.115.
  - JLN Measure (Firm): Obs. 108,330; Mean 0.921; SD 0.084; Min 0.768; Max 1.086.
- Panel-C: Macro Variables (F)
  - Foreign Reserves/ST External Debt: Obs. 104,890; Mean 4.52; SD 4.333; Min 0.464; Max 74.92.
  - Public Debt/GDP: Obs. 120,827; Mean 0.452; SD 0.206; Min 0.082; Max 1.032.
  - Private Bank Credit/GDP: Obs. 89,098; Mean 0.739; SD 0.341; Min 0.135; Max 1.394.
  - Chinn-Ito Index of Openness: Obs. 110,356; Mean 0.413; SD 0.288; Min 0; Max 1.

### Empirical strategy and main estimation results
- Estimation: panel regressions (firm-year) with firm fixed-effects; standard errors clustered at country level; robustness with two-way fixed-effects and two-way clustering presented.
- Equation (1) (average level effects): lagged Tobin’s Q, lagged Net Leverage, and Cash Flow Ratio all statistically significant with expected signs.
  - Example coefficients from Table 2, column (1): (Lag) Tobin’s Q 0.059*** (0.008); (Lag) Net Leverage -0.064*** (0.006); Cash Flow Ratio 0.133*** (0.010); ∆ U.S. Interest rates -0.737*** (0.099).
  - Observations 61,320; R-squared 0.121; Number of Firms 11,282.
- Equation (2) (interactions to capture heterogeneous responses):
  - Interaction (Lag) Net Leverage * S: negative coefficient for uncertainty shocks and statistically significant (supports Hypothesis 1 for uncertainty but not for interest rates in baseline).
  - Interaction Cash Flow * S: statistically significant with opposite signs across shocks:
    - For interest-rate shocks: Cash Flow * S positive (e.g., 0.698** (0.321) in baseline), implying firms with weaker cash flows reduce investment relatively more—consistent with financing channel (Supports Hypothesis 2).
    - For uncertainty shocks: Cash Flow * S negative (e.g., -0.136** (0.055) in baseline), implying higher uncertainty reduces MPI—consistent with wait-and-see channel (Related to Hypothesis 3).
  - Subsample results (Table 3): for firms with net leverage above the 75th percentile, (Lag) Net Leverage * interest-rate shock coefficient is -0.907*** (0.321), indicating heterogeneity by leverage emerges in highly leveraged firms.
  - When both shocks included simultaneously (Table 6, column (2) with lagged VIX): S1 (interest rate shock) -0.944*** (0.119); S2 (uncertainty shock) -0.069*** (0.024); (Lag) Net Leverage * S2 -0.104*** (0.029); Cash Flow * S2 -0.129** (0.053).

### Robustness checks
- Alternative shock measures: GSW yields and realized volatility (RVOL) produce nearly identical results (Table 4, columns (2) and (8)).
- Fixed-effects: results robust to year fixed effects and country-year fixed effects (Table 4, columns (3), (4), (9), (10)).
- Clustering: two-way clustering (country and year) does not change statistical significance (Table 4, columns (5) and (11)).
- Sample composition: excluding all Chinese and Indian firms yields similar results (Table 4, columns (6) and (12)).
- Alternative uncertainty measures: JLN (macro) and JLN (firm) produce qualitatively similar results (Table 5); example: (Lag) Net Leverage * Uncertainty -0.065* (0.035) for JLN (macro) and -0.106*** (0.028) for JLN (firm).
- Model passes Sargan-Hansen endogeneity test for right-hand-side firm-level variables when firm fixed effects are included (results available from authors upon request).

### Economic significance
- Table 7: percentage point reduction in investment rates in response to one standard deviation increase in U.S. interest rates or the VIX (selected entries preserved exactly):
  - Net Leverage percentiles and impacts:
    - p5: Interest Rates -0.41; VIX -0.22.
    - p25: Interest Rates -0.45; VIX -0.48.
    - p50 (median): Interest Rates -0.47; VIX -0.67.
    - p75: Interest Rates -0.50; VIX -0.92.
    - p95: Interest Rates -0.58; VIX -1.46.
  - Ratios:
    - p75 / p25: Interest Rates 1.13; VIX 1.93.
    - p95 / p5: Interest Rates 1.40; VIX 6.67.
- Interpretation: difference in response between bottom and top extremes in net leverage could be as large as about 7 times (illustrated by p95/p5 for VIX = 6.67). A one-standard deviation shock could reduce investment-to-capital ratio for median firm by around 0.5 or 0.7 percentage points for interest rate or uncertainty, respectively.
- Comparison to crisis: median reduction in investment in 2009 was around 4.6 percentage points; a one-standard deviation shock could account for 10–15 percent of the GFC reduction; plugging observed VIX increase during GFC into estimates explains 37 percent of median reduction.

### Key empirical takeaways
- Both financing and wait-and-see channels coexist: higher interest rates tend to raise MPI out of cash flows (financing channel), while higher uncertainty tends to lower MPI (wait-and-see).
- Heterogeneous effects across firms:
  - Leverage dimension: uncertainty shocks generate statistically significant differential responses by leverage in the full sample; interest-rate heterogeneous effects by leverage appear primarily among highly leveraged firms (above 75th percentile).
  - Cash-flow dimension: weaker-cash-flow firms reduce investment more under higher interest rates; stronger-cash-flow firms reduce investment more under higher uncertainty.
- Country-level fundamentals are hypothesized to matter in buffering firms against external shocks; empirical augmentation with country fundamentals (external liquidity, fiscal space, financial development) is introduced for further analysis in subsequent sections.

*Source: IMF working paper content (Section IV and surrounding empirical results) as provided.*

### 1. F1: foreign reserves to short-term external debt ratio

### 1. F1: foreign reserves to short-term external debt ratio

### Model specification and estimation approach
- The paper estimates a specification that includes all possible interactions among shocks, firm-, and country-fundamentals (equation (3)).
- Differential responses of firms in different countries to a common shock are investigated via the differential of equation (3) and a linear combination given in equation (4).
- Financial depth is measured as the ratio of total private credit by deposit banks to GDP. Results are robust to using private credit from all financial institutions.
- Estimation uses firm-level fixed effects with standard errors clustered at the country level (robust to arbitrary heteroskedasticity).

### Key regression outputs (Table 8)
- Columns (1)-(3): interest-rate shocks. Columns (4)-(6): uncertainty shocks.
- Common coefficient highlights:
  - (Lag) Q: 0.057*** (0.009), 0.056*** (0.008), 0.056*** (0.011) for columns (1)-(3); 0.055*** (0.007), 0.055*** (0.007), 0.051*** (0.009) for columns (4)-(6).
  - Shocks: S: -1.485*** (0.430), -0.056 (0.538), -2.292** (1.017) for interest-rate columns; -0.070 (0.049), -0.098 (0.071), -0.012 (0.071) for uncertainty columns.
  - F1: Foreign Reserves/ST External Debt coefficients shown: -0.002 (0.001) and -0.003 (0.004) in relevant columns.
  - F2: Public Debt/GDP: -0.172*** (0.044) and -0.181*** (0.053) in relevant columns.
  - F3: Private Bank Credit/GDP: -0.021 (0.038) and 0.027 (0.061) in relevant columns.
  - (Lag) Net Leverage: -0.068*** (0.008), -0.062*** (0.019), -0.051** (0.023) for interest-rate columns; -0.047*** (0.016), -0.033 (0.037), 0.017 (0.039) for uncertainty columns.
  - Cash Flow Ratio: 0.132*** (0.011), 0.123*** (0.021), 0.179*** (0.034) for interest-rate columns; 0.151*** (0.028), 0.125*** (0.042), 0.196*** (0.063) for uncertainty columns.
  - US Real GDP Growth: 0.280** (0.130), 0.288** (0.139), 0.275* (0.145) in interest-rate columns; 0.051 (0.139), 0.096 (0.153), 0.055 (0.137) in uncertainty columns.
- Model fit and sample:
  - Observations: 52,315; 58,959; 39,129 (columns vary).
  - R-squared: 0.119, 0.123, 0.118 (interest-rate columns) and 0.122, 0.125, 0.123 (uncertainty columns).
  - Number of firms: 10,010; 11,084; 8,499 (columns vary).
- Significance notation: *** p<0.01, ** p<0.05, * p<0.1.

### Country heterogeneity in investment responses (Figure 6 and interpretation)
- Figure 6 plots the partial derivative of the investment ratio w.r.t. interest-rate shock from equation (4) across percentiles (1 to 99) of country fundamentals; shows point estimates with 90% confidence intervals.
- Visual inspection suggests slopes are statistically different from zero for interest-rate responses along country fundamentals.
- Illustrative numeric comparison:
  - Following an increase in U.S. long-term interest rates, the contraction in investment by a firm with the median level of net leverage is three times larger if the firm is in a country with public debt of 75 percent of GDP than if it is in a country with public debt of around 32 percent of GDP.
- Qualitative findings:
  - Firms in countries with stronger sovereign balance sheets (lower public debt, higher reserves) or more developed financial markets are more buffered against increases in U.S. interest rates.
  - A firm with net leverage equal to the sample median contracts investment less in response to higher interest rates when located in a country with higher foreign reserves or deeper financial markets.

### Formal tests of slope heterogeneity (Table 9)
- Table 9 reports statistical significance of the slope term in equation (4) for interest-rate and uncertainty shocks.
- Interest-rate shocks (columns represent different specifications):
  - From Table 8: Foreign Reserves: 0.125** (0.052); Public Debt: -2.087*** (0.795); Credit Deepening: 1.836* (0.967).
  - All macro variables included at once: Foreign Reserves: 0.268*** (0.081); Public Debt: -7.307*** (1.771); Credit Deepening: 1.543* (0.800).
  - All + Openness: Foreign Reserves: 0.261*** (0.084); Public Debt: -7.346*** (1.814); Credit Deepening: 1.677** (0.757).
- Uncertainty shocks:
  - From Table 8: Foreign Reserves: 0.002 (0.006); Public Debt: 0.123 (0.076); Credit Deepening: -0.014 (0.068).
- Notes: Column (3) further controls for financial account openness and its interaction with shocks. *** p<0.01, ** p<0.05, * p<0.1.
- Key inference: the effect of public debt on investment triples when all fundamentals and the openness indicator are included (compared to Table 8), reinforcing the role of macro fundamentals for interest-rate responses. No statistically significant heterogeneity is found along country fundamentals for uncertainty shocks.

### Main empirical conclusions
- Significant heterogeneity exists in the response of corporate investment in emerging markets to changes in U.S. interest rates or volatility.
- Balance-sheet dimension:
  - More levered firms reduce capital expenditures proportionally more when facing higher uncertainty or higher interest rates — supporting a financing channel.
  - Firms’ marginal propensity to invest (out of cash flows) increases on average when faced with higher interest rates — consistent with the financing channel.
  - Evidence also supports a wait-and-see channel: marginal propensity to invest decreases when uncertainty increases.
  - The relative strength of financing vs. wait-and-see channels depends on firms’ leverage: for highly leveraged firms the financing channel strengthens while the wait-and-see channel weakens, and vice versa.
- Macro fundamentals:
  - Low public debt, high international reserves, and deeper financial markets mitigate firms’ investment responses to increases in U.S. interest rates.
  - Strong country fundamentals reduce firms’ aggregate risks and the intensity of financial frictions, improving resilience to global interest-rate shocks.
- Heterogeneity along country fundamentals in response to volatility shocks is less significant.

### Policy recommendations
- Maintain strong macroeconomic fundamentals—specifically:
  - Keep public debt low.
  - Preserve high international reserves relative to short-term external obligations.
  - Promote deeper financial markets (higher private credit by deposit banks to GDP).
- These policies help mitigate the effect of higher global interest rates on corporate investment through the financing channel.

*Italic: Source: _wp15285 - 1. F1: foreign reserves to short-term external debt ratio (PDF chapter/section).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15285.pdf_