## 2.1–2.3 Empirical Strategy and Stylized Facts

## Source details

**Canonical URL:** [2.1–2.3 Empirical Strategy and Stylized Facts](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020025-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020025-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020025-print-pdf.pdf.json)

---

### Empirical framework
- Baseline approach: differences-in-differences comparing changes in intangible investment between firms with different levels of pre-crisis financial vulnerabilities, before (2002-2007) and after (2008-2013) the Lehman-induced credit tightening.
- Primary regression (first stage):
  - ∆Intangible Investment_isc = β1 Financial Vulnerability_isc + δ_sc + γX_isc + ε_isc
  - ∆Intangible Investment_isc: difference in average investment in intangible assets (scaled by total assets) of firm i, in industries s and country c between pre-crisis (2002-2007) and post-crisis (2008-2013).
  - Financial Vulnerability_isc: pre-crisis balance sheet vulnerability measure.
  - δ_sc: country-sector fixed effects (4-digit NACE level).
  - X_isc: firm-level controls including age, total assets and cash flows (ratio of cash flows to assets) in the pre-crisis period.
- Second-stage (macro policy interaction):
  - ∆Intangible Investment_isc = β1 Financial Vulnerability_isc + β2 Financial Vulnerability_isc × Expansionary Conditions_c + δ_sc + γX_isc + ε_isc
  - Expansionary Conditions_c: measure of post-GFC easing in monetary conditions (baseline: average forecast error of long-term (10 year) government bond yields).
  - Interpretation: β1 < 0 and β2 > 0 imply expansionary monetary conditions mitigate the adverse effect of financial vulnerability on intangible investment.
- Third-stage (interaction with product market competition):
  - Triple interaction included: Financial Vulnerability_isc × Expansionary Conditions_c × Weak Competition_sc.
  - Weak Competition_sc: proxy for degree of product market competition (baseline: median Lerner index for each country-sector).
  - Testable implications from Aghion, Farhi and Kharroubi (2019):
    - β3 > 0: adverse effect of balance sheet weakness on intangible investment is greater for firms facing higher competition.
    - β4 < 0: beneficial impact of expansionary monetary conditions is greater for firms facing higher competition.
- Standard errors clustered at the country-sector level.

### Data scope and measurement
- Dataset: ORBIS (Bureau van Dijk); data cleaning follows Diez et al. (2019), Kalemli-Özcan et al. (2015), and Gopinath et al. (2017).
- Countries included (17 OECD): Austria, Belgium, Czech Republic, Germany, Finland, France, Greece, Hungary, Ireland, Italy, Korea, Norway, Poland, Portugal, Slovakia, Spain, UK.
- Sectoral coverage: non-farm, non-financial business sector corresponding to two-digit industry codes 5-82 in NACE Revision 2.
- Dependent variable: change in intangible investment = difference in average investment in intangibles as a share of total assets between pre- and post-crisis periods; measured on a net basis (net of depreciation/amortization) as change in real intangible assets.
- Measurement concerns:
  - Intangible assets reporting likely imperfect and under-reported (acquired assets like patents/software vs. internally-generated items at research stage such as R&D).
  - Likely classical measurement error in the dependent variable leading to attenuation bias (conservative results).
  - Robustness checks use alternative dependent variable measures less prone to measurement error.

### Firm-level vulnerability and policy measures
- Baseline firm vulnerability: pre-crisis average leverage ratio = (current liabilities + long-term debt) / total assets.
- Robustness vulnerability: interest coverage ratio = total interest paid / EBITDA.
- Baseline monetary conditions measure: average forecast error of long-term (10 year) government bond yields in the post-crisis period relative to OECD forecasts.
- Alternative monetary policy measures:
  - Deviation of the policy rate from its Taylor-rule-implied value.
  - Forecast error of the short-term policy rate orthogonal to unexpected changes in output growth and inflation (Consensus Economics forecasts).
- Fiscal policy shock (extension): forecast error of the ratio of government consumption to GDP (OECD Economic Outlook).
- Product market competition measures:
  - Baseline: median Lerner index in the pre-crisis period, where Lerner index = (EBITDA - Depreciation and Amortization) / Operating Revenue.
  - Alternative: median firm markups (De Loecker and Warzynski (2012)).
  - Economy-wide alternatives: OECD PMR indicators and sub-indicators (barriers to entry; regulatory protection of incumbents; administrative burdens).

### Identification and parallel trends
- Two identification conditions:
  - The 2008 GFC shock is treated as exogenous.
  - Parallel trends: high- and low-leverage firms had similar intangible investment growth prior to the crisis; divergence occurred after the crisis.
- Parallel trends test: regressions of firm-level intangible investment growth on year dummies and country-sector fixed effects run separately for high- and low-leverage firms (split at median leverage within country-sector).
  - Result: intangible investment by high- and low-leverage firms grew at a similar pace until the GFC; after the GFC high-leverage firms experienced a greater decline.
  - Comparable test for tangible investment shows a less pronounced post-crisis divergence.

### Intangible investment and productivity
- TFP estimated using a production function with labor and tangible capital via the GMM procedure proposed by Wooldridge (2009).
- Firms ranked by change in intangible-to-total-assets ratio (percent) between pre- and post-crisis periods; binned into 100 quantiles.
- Strong positive correlation between changes in intangible investment and TFP growth between pre- and post-crisis periods.
- Conclusion: balance-sheet-reported intangible assets, while incomplete, contain relevant information for productivity growth.

### Baseline empirical findings (key statistics and coefficients)
- Table 1 (summary statistics, 664,086 observations):
  - ∆Intangible Investment: Mean -0.0012; Median 0.0005; 25th percentile 0.0023; 75th percentile 0.0231.
  - ∆Tangible Investment: Mean -0.01400; Median -0.0315; 25th percentile 0.0266; 75th percentile 0.2080.
  - Intangible / Total assets ratio: Mean 0.16; Median 0.02; 25th percentile 0.21; 75th percentile 0.25.
  - Leverage ratio: Mean 0.67; Median 0.70; 25th percentile 0.53; 75th percentile 0.84; Standard deviation 0.21.
  - Lerner index: Mean 0.06; Median 0.05; 25th percentile 0.03; 75th percentile 0.08; Standard deviation 0.06.

- Table 2 (baseline vulnerability effects; Observations: 664,086):
  - Vulnerability_isc: -0.005*** (0.000) in column (1).
  - Vulnerability_isc: -0.014*** (0.005) in column (2).
  - Vulnerability_isc × Pre−crisis physical assets ratio_isc: 0.010* (0.006) in column (2).
  - R-squared: 0.061 (col 1) and 0.065 (col 2).

- Table 3 (leverage, monetary conditions, competition; Observations: 664,086):
  - Vulnerability_isc: -0.005*** (0.000) ; -0.007*** (0.001) ; -0.009*** (0.001) across columns (1)–(3).
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.005*** (0.001) in col (1); 0.009*** (0.001) in col (2).
  - Vulnerability_isc × Weak competition_sc: 0.047*** (0.011) in col (2); 0.061*** (0.013) in col (3).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.074*** (0.019) in col (3).
  - R-squared: 0.062, 0.061, 0.062 across columns.

- Magnitudes and interpretations:
  - A 10 percentage point higher leverage ratio is associated with a reduction in intangible investment rate of around 0.05 percentage point (column 1, baseline).
  - The decline between the 75th percentile and 25th percentile firms in leverage amounts to 0.15 percentage point.
  - Interaction with tangible-asset share significant at the 10% level: firms with larger shares of physical assets were less affected by tighter credit conditions.
  - A more-than-expected reduction in long-term external financing cost of 100 basis points fully negates the adverse impact on intangible investment stemming from debt overhang risk (Table 3 interpretation).
  - Scenario: expansionary/contractionary monetary conditions by 50 bps used for illustrative comparisons (Figure 4); adverse effects of financial frictions on intangible investment can be reduced by one-third under cases considered.

- Back-of-the-envelope output implications:
  - Using elasticity of output w.r.t. intangible capital in the range 20 to 40 percent.
  - A 100 basis points negative surprise on monetary conditions could prevent a cumulative output loss of 1 to 2 percent for high-leverage firms (75th percentile) vis-`a-vis low-leverage counterparts over the five years after the crisis.

### Competition, complementarities, and heterogeneity
- Competition amplifies financial-friction effects:
  - Regression evidence: financial constraints have a larger adverse effect on intangible investment when competition is stronger (lower Lerner index).
  - Table 3 and Figure 5 illustrate complementarity:
    - Triple interaction coefficient negative and statistically significant: expansionary monetary conditions alleviate adverse impacts especially when firms face stronger competition.
    - Magnitudes illustrated:
      - At the 25th percentile of pre-crisis Lerner indices (stronger competition): estimated cut in intangible investment for more-levered vs less-levered firms reduced from 0.25 to 0.08 percentage point when monetary conditions are more (rather than less) expansionary.
      - At the 75th percentile of Lerner indices (weaker competition): corresponding numbers are 0.16 and 0.05.
      - Differences: 0.25-0.08=0.17 versus 0.16-0.05=0.11 imply an overall 50% stronger effect of counter-cyclical macroeconomic policies in more competitive environments (comparison of 0.17 vs 0.11).

### Asset-type specificity and robustness
- Asset specificity (Table 4):
  - Column (1) (Intangible to total assets ratio; Observations: 664,084):
    - Vulnerability_isc: -0.022*** (0.004).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.027*** (0.006).
    - Vulnerability_isc × Weak competition_sc: -0.141*** (0.053).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.233*** (0.079).
  - Column (2) (Investment in tangible assets; Observations: 674,266):
    - Vulnerability_isc: -0.017*** (0.003).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.002 (0.005).
    - Vulnerability_isc × Weak competition_sc: -0.001 (0.048).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: 0.007 (0.075).
  - Interpretation: channel through which counter-cyclical macroeconomic policies and product market competition shape responses to credit conditions is more pronounced for intangible assets than for tangible ones.

- Alternative macro shock measures (Table 5):
  - Columns use: Deviation from Taylor rule; Forecast errors in short-term policy rate; Forecast errors in gov’t consumption.
  - Key coefficients (examples):
    - Vulnerability_isc: -0.005*** (0.001) in col (1); -0.011*** (0.001) in col (2); -0.015*** (0.001) in col (3).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.004*** (0.000) in col (1); 0.010*** (0.003) in col (2); 0.014*** (0.001) in col (3).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.016** (0.006); -0.072* (0.038); -0.121*** (0.017) across cols (1)–(3).
  - Results qualitatively identical to baseline; triple interaction borderline significant (10% level) when using forecast error of the short-term rate.

- Alternative competition measures (Table 6):
  - Columns (1)–(4) use: median markup; Product market regulation (PMR); Regulatory protection of incumbents (RPI); Administrative burden for startups (ABS).
  - Key patterns (all columns, Observations: 664,086):
    - Vulnerability_isc coefficients negative and significant across columns.
    - Vulnerability_isc × Expansionary monetary conditions_c positive and significant across columns.
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc negative and significant across columns.
  - Qualitatively identical results; relevant effects significant at the 1% confidence level.

- Additional robustness checks (Table 7):
  - Column (1) Shorter window (2005-2010; Observations: 615,143):
    - Vulnerability_isc: -0.013*** (0.001).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.011*** (0.002).
    - Vulnerability_isc × Weak competition_sc: 0.093*** (0.015).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.100*** (0.024).
    - R-squared: 0.074.
  - Column (2) Alternative vulnerability (interest coverage; Observations: 664,453):
    - Vulnerability_isc: -0.002*** (0.000).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.004*** (0.001).
    - Vulnerability_isc × Weak competition_sc: 0.012* (0.007).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.028*** (0.009).
    - R-squared: 0.060.
  - Column (3) Linear probability model (probability; Observations: 664,086):
    - Vulnerability_isc: -0.175*** (0.013).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.078*** (0.023).
    - Vulnerability_isc × Weak competition_sc: 0.879*** (0.209).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.623* (0.322).
    - R-squared: 0.087.
  - Column (4) Excluding zeros (no intangible assets; Observations: 518,048):
    - Vulnerability_isc: -0.010*** (0.001).
    - Vulnerability_isc × Expansionary monetary conditions_c: 0.011*** (0.002).
    - Vulnerability_isc × Weak competition_sc: 0.069*** (0.016).
    - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.096*** (0.023).
    - R-squared: 0.072.
  - Across specifications, higher pre-crisis firm vulnerability (higher debt-to-assets or lower interest coverage) is associated with a negative main effect and with interactive patterns: positive two-way interactions with Expansionary monetary conditions and Weak competition and negative three-way interaction.

### Policy implications and conclusions
- Core interpretation:
  - Intangible investment is increasingly important for growth but is vulnerable to financial frictions because it is non-pledgeable and hard to liquidate.
  - Counter-cyclical macroeconomic policy can shelter intangible investment from adverse shocks, strengthening longer-term growth.
  - Pro-competitive product market deregulation can interact with macro policy: when competition is strong, counter-cyclical macro policy is more effective at preventing slowdowns in intangible investment by constrained firms.
- Policy recommendations implied by results:
  - Strengthen counter-cyclical macro policies (monetary and fiscal) alongside product market deregulation to foster intangible investment and growth.
  - Recognize that product market competition may exacerbate the negative effects of credit constraints in downturns unless accompanied by stronger counter-cyclical macro policies.
- Broader implication:
  - Counter-cyclical macroeconomic policy could persistently or even permanently affect growth through its protection of intangible investment.

*Source: wpiea2020025-print-pdf (IMF Working Paper).*

### 2.1    Empirical Strategy

### wpiea2020025-print-pdf - 2.1    Empirical Strategy

### Empirical framework
- Baseline approach: differences-in-differences comparing changes in intangible investment between firms with different levels of pre-crisis financial vulnerabilities, before (2002-2007) and after (2008-2013) the Lehman-induced credit tightening.
- Relation to prior literature: methodology similar to Giroud and Mueller (2017), Kalemli-Özcan et al. (2018), and Duval, Hong and Timmer (2020).
- Primary regression (first stage):
  - ∆Intangible Investment_isc = β1 Financial Vulnerability_isc + δ_sc + γX_isc + ε_isc
  - ∆Intangible Investment_isc: difference in average investment in intangible assets (scaled by total assets) of firm i, in industries s and country c between pre-crisis (2002-2007) and post-crisis (2008-2013).
  - Financial Vulnerability_isc: pre-crisis balance sheet vulnerability measure (detailed below).
  - δ_sc: country-sector fixed effects (4-digit NACE level).
  - X_isc: firm-level controls including age, total assets and cash flows (ratio of cash flows to assets) in the pre-crisis period.
  - Identification: country-sector fixed effects and firm controls absorb country-industry common shocks; reverse causality not a concern because the Lehman shock was unforeseen.
- Second-stage (macroeconomic policy interaction):
  - ∆Intangible Investment_isc = β1 Financial Vulnerability_isc + β2 Financial Vulnerability_isc × Expansionary Conditions_c + δ_sc + γX_isc + ε_isc
  - Expansionary Conditions_c: measure of post-GFC easing in monetary conditions (detailed below).
  - Interpretation: β1 < 0 and β2 > 0 imply expansionary monetary conditions mitigate the adverse effect of financial vulnerability on intangible investment.
- Third-stage (interaction with product market competition):
  - ∆Intangible Investment_isc = β1 Financial Vulnerability_isc + β2 Financial Vulnerability_isc × Expansionary Conditions_c + β3 Financial Vulnerability_isc × Weak Competition_sc + β4 Financial Vulnerability_isc × Expansionary Conditions_c × Weak Competition_sc + δ_sc + γX_isc + ε_isc
  - Weak Competition_sc: proxy for degree of product market competition (detailed below).
  - Testable implications from Aghion, Farhi and Kharroubi (2019):
    - β3 > 0: adverse effect of balance sheet weakness on intangible investment is greater for firms facing higher competition.
    - β4 < 0: beneficial impact of expansionary monetary conditions is greater for firms facing higher competition.
  - Standard errors clustered at the country-sector level.

### Data scope and sample
- Dataset: ORBIS (Bureau van Dijk), cross-country longitudinal dataset of listed and unlisted firms; data cleaning follows Diez et al. (2019), Kalemli-Özcan et al. (2015), and Gopinath et al. (2017).
- Countries included (17 OECD): Austria, Belgium, Czech Republic, Germany, Finland, France, Greece, Hungary, Ireland, Italy, Korea, Norway, Poland, Portugal, Slovakia, Spain, UK.
- Sectoral coverage: non-farm, non-financial business sector corresponding to two-digit industry codes 5-82 in NACE Revision 2 (manufacturing and selected service sectors).
- Dependent variable: change in intangible investment = difference in average investment in intangibles as a share of total assets between pre- and post-crisis periods; measured on a net basis (net of depreciation/amortization) as change in real intangible assets.
- Measurement concerns:
  - Intangible assets reporting likely imperfect and under-reported (acquired assets like patents/software vs. internally-generated items at research stage such as R&D).
  - Likely classical measurement error in the dependent variable leading to attenuation bias (conservative results).
  - Robustness checks: alternative dependent variable measures less prone to measurement error.

### Firm-level financial vulnerability measures
- Baseline measure: pre-crisis average leverage ratio capturing degree of debt overhang risk.
  - Calculated as (current liabilities + long-term debt) / total assets over corresponding periods.
- Robustness measure: interest coverage ratio prior to the crisis.
  - Defined as total interest paid by the firm over EBITDA (earnings before interest, taxes, depreciation and amortization).
  - Intended to capture drags on financing from debt payments.

### Macroeconomic policy condition measures
- Baseline measure of post-crisis monetary conditions:
  - Average forecast error of long-term (10 year) government bond yields in the post-crisis period, relative to OECD forecasts for the year (Fall issue of the OECD Economic Outlook in the previous year).
  - Captures surprise component of monetary conditions and incorporates conventional and unconventional monetary policy effects.
- Alternative monetary policy shock measures:
  - Deviation of the policy rate from its simple Taylor-rule-implied value in the post-crisis period (using standard Taylor rule parameters).
  - Forecast error of the short-term policy rate orthogonal to unexpected changes in output growth and inflation:
    - Step 1: compute differences between actual and forecast short-term rate, inflation, and GDP growth using forecasts from Consensus Economics in October of the same year.
    - Step 2: regress forecast errors for the short-term rate on forecast errors for inflation and GDP growth; residuals are the exogenous monetary policy shock.
  - These alternatives address concerns that baseline measure may reflect correlated aggregate GDP growth conditions.
- Fiscal policy shock (extension):
  - Forecast error of the ratio of government consumption to GDP (following Auerbach and Gorodnichenko (2012)), forecasts taken from the Fall issue of the OECD Economic Outlook in the same year.

### Product market competition measures
- Three approaches to measure degree of competition; two derived from firm-level market power:
  - Baseline: median Lerner index for each country-sector in the pre-crisis period.
    - Lerner index = (EBITDA - Depreciation and Amortization) / Operating Revenue.
    - Represents an inverse measure of country-sector-level competition.
  - Alternative: firm-level markups computed following De Loecker and Warzynski (2012); take median firm markups in each country-sector as a measure of (weak) competition.
- Economy-wide alternative:
  - OECD product market regulation (PMR) indicators (overall PMR indicator and sub-indicators on barriers to entry and regulatory protection of incumbents) used for robustness.
- Rationale: firm-level measures preferred because they directly measure market competition at the country-industry level.

*Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020025-print-pdf.pdf*

### 2.3    Stylized Facts

### 2.3    Stylized Facts

### Identification and key assumptions
- Two conditions for the differences-in-differences strategy to be valid:
  - The 2008 GFC shock should be exogenous.
  - Firms with different levels of financial vulnerabilities should have had a parallel trend in intangible investment growth prior to the crisis, with any divergence materializing only after the crisis.
- Evidence on exogeneity: the GFC was unforeseen by individual firms (example: managers in the securitized finance industry failed to identify the housing bubble).
- Parallel trends test:
  - Regressions of firm-level intangible investment growth on year dummies and four-digit country-sector fixed effects are run separately for high-leverage and low-leverage firms.
  - Leverage split: in each country-sector, firms are split at the median of the average leverage ratio over the pre-crisis period.
  - Result: intangible investment by high- and low-leverage firms grew at a similar pace until the GFC; after the GFC high-leverage firms experienced a greater decline.
  - Comparable test for tangible investment shows a less pronounced post-crisis divergence, supporting the premise that intangible investment is more dependent on continuous finance due to limited pledgeability.

### Intangible investment and productivity
- TFP estimation:
  - TFP series derived from a production function with labor and tangible capital, using the GMM procedure proposed by Wooldridge (2009).
- Empirical link:
  - Firms are ranked by the change in intangible-to-total-assets ratio (percent) between pre- and post-crisis periods; data broken into 100 quantiles.
  - Strong positive correlation between changes in intangible investment and TFP growth between pre- and post-crisis periods.
  - The intangible assets reported on balance sheets, while not fully comprehensive, contain relevant information for productivity growth.

### Financial frictions and intangible investment (baseline results)
- Main finding (Table 2):
  - A 10 percentage point higher leverage ratio is associated with a reduction in intangible investment rate of around 0.05 percentage point (column 1).
  - Difference illustrated: the decline in intangible investment rate between the 75th percentile and 25th percentile firms (in leverage) amounts to 0.15 percentage point.
- Role of asset pledgeability:
  - Interaction of leverage with share of tangible assets (pre-crisis) is statistically significant at the 10% confidence level (column 2).
  - Interpretation: firms with a larger share of physical assets were less affected by tighter credit conditions; those with smaller shares were more affected.

### Role of expansionary monetary conditions
- Table 3 (specification (2), column 1):
  - The interaction term indicates that more expansionary monetary conditions can dampen or prevent cuts in intangible investment by financially-constrained firms.
  - Magnitude: a more-than-expected reduction in the long-term external financing cost of some 100 basis points fully negates the adverse impact on intangible investment stemming from debt overhang risk.
- Scenario comparison (Figure 4):
  - Two scenarios: more-than-expected expansionary and contractionary monetary conditions by 50bps.
  - Adverse effects of financial frictions on intangible investment can be reduced by one-third under the cases considered (red bars in the middle).
- Output implications (back-of-the-envelope):
  - Use of an elasticity of output with respect to intangible capital in the range 20 to 40 percent.
  - A 100 basis points negative surprise on monetary conditions could prevent a cumulative output loss of 1 to 2 percent for high-leverage firms (75th percentile) vis-`a-vis low-leverage counterparts over the five years after the crisis.

### Competition, financial frictions, and complementarity with monetary policy
- Competition amplifies financial-friction effects:
  - Premise: adverse effects of financial frictions on intangible investment are larger for firms facing high competition because they cannot use monopoly rents to self-finance.
  - Regression results (Table 3, column 2): financial constraints have a larger adverse effect on intangible investment when competition is stronger (country-sector median Lerner index is lower).
- Complementarity test (specification (3), Table 3 column 3 and Figure 5):
  - Triple interaction term coefficient is negative and statistically significant, indicating expansionary monetary conditions alleviate adverse impacts especially when firms face stronger competition.
  - Magnitudes illustrated:
    - At the 25th percentile of pre-crisis Lerner indices (stronger competition), the estimated cut in intangible investment by more-leveraged firms relative to less-leveraged counterparts can be reduced from 0.25 to 0.08 percentage point when monetary conditions are more (rather than less) expansionary than expected (green shaded bars on the left).
    - At the 75th percentile of Lerner indices (weaker competition), corresponding numbers are 0.16 and 0.05 (green shaded bars on the right).
    - Differences: 0.25-0.08=0.17 versus 0.16-0.05=0.11 (red solid and shaded bars in the middle), implying an overall 50% stronger effect of counter-cyclical macroeconomic policies in more competitive environments.

### Asset-type specificity and additional checks
- Asset specificity (Table 4):
  - Replacing the dependent variable with:
    - Difference in average ratio of intangible to tangible assets (column 1).
    - Difference in average investment in tangible assets (scaled by total assets; column 2).
  - Result: the channel through which counter-cyclical macroeconomic policies and product market competition shape responses to credit conditions is more pronounced for intangible assets than for tangible ones.
- Robustness (Section 4):
  - Alternative measures of expansionary monetary conditions (Table 5):
    - Deviation of actual policy rate from Taylor-rule implied rate (column 1).
    - Forecast error of actual short-term rate orthogonal to changes in inflation and output growth (column 2).
    - Forecast error of government consumption expenditure to GDP (column 3).
    - Qualitatively identical results to baseline; double and triple interaction terms remain statistically significant. The triple interaction is borderline significant (10% level) when using the forecast error of the short-term rate.
  - Alternative measures of product market competition (Table 6):
    - Country-sector median of estimated markups (column 1).
    - OECD overall PMR indicator (column 2).
    - OECD sub-indicator on regulatory protection of incumbents (column 3).
    - OECD sub-indicator on administrative burdens on start-ups.
    - Qualitatively identical results; relevant effects significant at the 1% confidence level.
  - Additional sample/specification checks (Table 7):
    - Shorten time window to 2005-2010 (two years before versus two years after the GFC), removing 2010-2012 euro area crisis: slightly larger point estimates in absolute terms (column 1).
    - Alternative firm-level vulnerability: interest coverage ratio yields qualitatively identical results (column 2).
    - Dependent variable as dummy for non-negative change in net intangible investment between pre- and post-crisis periods (linear probability model): qualitatively identical results (column 3).
    - Excluding firms that never had intangible assets (~10 percent of full sample): estimation results almost identical with slightly bigger point estimates (column 4).

### Policy implications and conclusion
- Core interpretation:
  - Intangible investment is increasingly important for growth but is vulnerable to financial frictions because it is non-pledgeable and hard to liquidate.
  - Counter-cyclical macroeconomic policy can shelter intangible investment from adverse shocks, strengthening longer-term growth.
  - Pro-competitive product market deregulation can interact with macro policy: when competition is strong, counter-cyclical macro policy is more effective at preventing slowdowns in intangible investment by constrained firms.
- Policy recommendations implied by results:
  - Strengthen counter-cyclical macro policies (monetary and fiscal) alongside product market deregulation to foster intangible investment and growth.
  - Be cautious that product market competition may backfire when credit constraints bite in downturns unless accompanied by stronger counter-cyclical macro policies.
- Broader implication:
  - Counter-cyclical macroeconomic policy could persistently or even permanently affect growth through its protection of intangible investment.

*Source: wpiea2020025-print-pdf - 2.3    Stylized Facts*

### References

### References

### Scholarly and working paper citations
- Ackerberg, D., K. Caves, and G. Frazer. 2015. “Identification Properties of Recent Production Function Estimators.” Econometrica 83(6): 2411–2451.
- Aghion, P. G. Angeletos, A. Banerjee and K. Manova. 2010. “Volatility and Growth: Credit Constraints and the Composition of Investment.” Journal of Monetary Economics, 57(3): 246-65.
- Aghion, P., P. Askenazy, R. Bourl`es, G. Cette, and N. Dromel. 2009a. Education, Market Rigidities and Growth. Economics Letters, 102 (1): 62-65.
- Aghion, P., N. Bloom, R. Blundell, R. Griffith, and P. Howitt. 2005. “Competition and Innovation: an Inverted-U Relationship.” Quarterly Journal of Economics, 120(2): 701-28.
- Aghion, P., R. Blundell, R. Griffith, P. Howitt, and S. Prantl. 2009b. The Effects of Entry on Incumbent Innovation and Productivity. Review of Economics and Statistics, 91(1): 20-32.
- Aghion, P., R. Blundell, R. Griffith, P. Howitt, and S. Prantl 2004. Entry and Productivity Growth: Evidence from Microlevel Panel Data. Journal of the European Economic Association, 2(2-3): 265-76.
- Aghion, P., R. Burgess, S. Redding and F. Zilibotti. 2008. “The Unequal Effects of Liberalization: Evidence from Dismantling the License Raj in India.” American Economic Review, 98(4): 1397-1412.
- Aghion, P., E. Farhi and E. Kharroubi. 2019. “Monetary Policy, Product Market Competition and Growth.” Economica, 86(343): 431-470.
- Aghion, P., D. Hemous and E. Kharroubi. 2014. “Cyclical Fiscal Policy, Credit Constraints, and Industry Growth.” Journal of Monetary Economics, 62(C): 41-58.
- Alesina, A., S. Ardagna, G. Nicoletti, and F. Schiantarelli. 2005. Regulation and Investment. Journal of the European Economic Association, 3(4):791-825.
- Almeida, H., M. Campello, B. Laranjeira and S. Weisbenner. 2012. “Corporate Debt Maturity and the Real Effects of the 2007 Credit Crisis.” Critical Finance Review, 1(1): 3-58.
- Arnold, J., B. Javorcik, M. Lipscomb and A. Mattoo. 2016. Services Reform and Manufacturing Performance: Evidence from India. Economic Journal, 126(590): 1-39.
- Aw, B., M. Roberts, D. Xu. 2011. “R&D Investment, Exporting, and Productivity Dynamics.” American Economic Review, 101(4): 1312-44.
- Barone, G. and F. Cingano (2011). Service Regulation and Growth: Evidence from OECD Countries. Economic Journal, 121(555): 931-957.
- Bassanini, A. and R. Duval. 2009. Unemployment, Institutions, and Reform Complementarities: Reassessing the Aggregate Evidence for OECD Countries. Oxford Review of Economic Policy, 25(1): 40-59.
- Benmelech, E., N. Bergman and A. Seru. 2011. “Financing Labor,” NBER Working Paper No. 17144.
- Blanchard, O., E. Cerutti and L. Summers. 2015. Inflation and Activity: Two Explorations and their Monetary Policy Implications. IMF Working Paper No. 15/230.
- Blanchard, O. and F. Giavazzi. 2003. Macroeconomic Effects of Regulation and Deregulation in Goods and Labor Markets. Quarterly Journal of Economics, 118(3): 879-907.
- Bordo, M., and J. Landon-Lane. 2013. “Does Expansionary Monetary Policy Cause Asset Price Booms: Some Historical and Empirical Evidence”, NBER Working Paper No. 19585, National Bureau of Economic Research.
- Bouis, R., R. Duval and J. Eugster. 2019. “How Fast Does Product Market Deregulation Pay Off? New Evidence from Non-Manufacturing Industry Deregulation in Advanced Economies.” Journal of Comparative Economics.
- Bourl`es, R., G. Cette, J. Lopez, J. Mairesse, and G. Nicoletti. 2013. Do Product Market Regulations in Upstream Sectors Curb Productivity Growth? Panel Data Evidence for OECD Countries. Review of Economics and Statistics, 95(5): 1750-1768.
- Bruno, V., and H. Shin. 2015. “Capital Flows and the Risk-Taking Channel of Monetary Policy.” Journal of Monetary Economics, 71(C): 119-132.
- Cerra, V. and S. Saxena. 2008. “Growth Dynamics: The Myth of Economic Recovery,” American Economic Review, 98(1): 439-457.
- Cheng, I.-H., Raina, S., Xiong, W. (2014). Wall street and the housing bubble. American Economic Review, 104(9): 2797–2829.
- Chodorow-Reich, G. 2014. “The Employment Effects of Credit Market Disruptions: Firm-Level Evidence from the 2008-9 Financial Crisis,” Quarterly Journal of Economics, 129(1): 1-59.
- Coibion, O., Y. Gorodnichenko, L. Kueng and J. Silvia. 2017. “Innocent Bystanders? Monetary Policy and Inequality.” Journal of Monetary Economics, 88(C): 70-89.
- Conway, P., D. de Rosa, F. Steiner and G. Nicoletti. 2006. Product Market Regulation and Productivity Convergence. OECD Economic Studies, 43(2), pp. 39-76.
- Corrado, C., J. Haskel, C. Jona-Lasinio and M. Iommi. 2016. “Intangible Investment in the EU and US Before and Since the Great Recession and its Contribution to Productivity Growth.” EIB Working Paper No. 2016/08.v
- Corrado, C., C. Hulten and D. Sichel, 2009. “Intangible Capital and U.S. Economic Wealth.” Review of Income and Wealth, 55(3): 661-685.
- Corrado, C., C. Hulten and D. Sichel, 2005. Measuring Capital and Technology: An Expanded Framework. In J. Haltiwanger, C. Corrado and D. Sichel (eds), Measuring Capital in New Economy, Chicago, IL: University of Chicago Press: 114-146.
- Correa-Caro. C, L. Medina, M. Poplawski-Ribeiro, B. Sutton. 2018. “Fiscal Stimulus Impact on Firms’ Profitability During the Global Financial Crisis.” IMF Working Paper No.18/251.
- Dell’Ariccia, G., L. Laeven, and G. Suare. 2017. “Bank Leverage and Monetary Policy’s Risk-Taking Channel: Evidence from the United States.” Journal of Finance, 72(2): 613-654.
- De Loecker, J., J. Eeeckhout. And G. Unger. Forthcoming. “The Rise of Market Power.” Quarterly journal of Economics.
- De Loecker, J., and F. Warzynski. 2012. “Markups and Firm-Level Export Status.” American Economic Review, 102(6): 2437-2471.
- de Ridder, Martin. 2016. “Investment in Productivity and the Long-Run Effect of Financial Crises on Output.” Cambridge-INET Working Paper Series 2016/18; Cambridge Working Paper Economics 1659, University of Cambridge, UK.
- Diez. F., J. Fan and C. Villegas-Sanchez. 2019. “Global Declining Competition.” IMF Working Paper.
- Dinlersoz, E., S. Kalemli-Özcan, H. Hyatt, and V. Penciakova. 2018. “Leverage over the Life Cycle and Implications for Firm Growth and Shock Responsiveness.” NBER Working Paper No. 25226.
- Doraszelski, U., and J. Jaumandreu. 2013. “R&D and Productivity: Estimating Endogenous Productivity.” Review of Economic Studies, 80(4): 1338-1383.
- Draghi, M. 2015. Monetary Policy and Structural Reforms in the Euro Area. Speech, Prometeia 40, Bologno, December 14th, available at https://www.ecb.europa.eu/press/key/date/2015/html/sp151214.en.html
- Duval, R., G. Hong and Y. Timmer. 2020. “Financial Frictions and the Great Productivity Slowdown.” Review of Financial Studies, 33(2): 475-503.
- Duval, R., D. Furceri. 2018. “The Effects of Labor and Product Market Reforms: The Role of Macroeconomic Conditions and Policies.” IMF Economic Review, 66(1): 31-69.
- Ebell, M., and C. Haefke. 2009. Product Market Deregulation and the U.S. Employment Miracle. Review of Economic Dynamics, 12(3): 479-504.
- Fang, L., and R. Rogerson. 2011. Product Market Regulation and Market Work: A Benchmark Analysis. American Economic Journal: Macroeconomics, 3(2): 163-88.
- Felbermayr, G. and J. Prat. 2011. Product Market Regulation, Firm Selection, and Unemployment. Journal of the European Economic Association, 9(2): 278-317.
- Fiori, G., G. Nicoletti, S. Scarpetta, and F. Schiantarelli. 2012. Employment Outcomes and the Interaction between Product and Labor Market Deregulation: Are They Substitutes or Complements? Economic Journal, 122 (558): 79-104.
- Gal, P. 2013. “Measuring Total Factor Productivity at the Firm Level using Orbis,” OECD Economics Department Working Paper No. 1049.
- Gal, P. and A. Hijzen. 2016. “The Short-Term Impact of Product Market Reforms: A Cross-Country Firm-Level Analysis.” IMF Working Paper Series No. 16/116.
- Giroud, X., and H. Mueller. 2017. “Firm Leverage, Consumer Demand, and Employment Losses during the Great Recession.” Quarterly Journal of Economics, 132(1): 271-316.
- Holmstrom, B. and J. Tirole. 1997. “Financial Intermediation, Loanable Funds, and the Real Sector.” Quarterly Journal of Economics, 112(3): 663-691.
- Hall, B. and J. Lerner. 2010. “The Financing of R&D and Innovation.” In Handbook of the Economics of Innovation. Vol. 1, edited by Bronwyn H Hall and Nathan Rosenberg. Elsevier, 2010.
- Hall, B., J. Mairesse, and P. Mohnen. 2010. “Measuring the Returns to R&D.” In Handbook of the Economics of Innovation. Vol. 2, edited by Bronwyn H Hall and Nathan Rosenberg. Elsevier, 2010.
- Huber, K. 2018. “Disentangling the Effects of a Banking Crisis: Evidence from German Firms and Counties,” American Economic Review, 108(3): 868-898.
- Inklaar, R. and M. Timmer, 2014. “The Relative Price of Services” Review of Income and Wealth, 60(4): 727-746.
- Inklaar, R., M. Timmer, and B. Van Ark. 2008. “Market Services Productivity Growth across Europe and the US,” Economic Policy, 23(53): 139-94.
- International Monetary Fund. 2016. Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies. World Economic Outlook, Chapter 3, April, Washington.
- International Monetary Fund. 2015. Where Are We Headed? Perspectives on Potential Output. World Economic Outlook, Chapter 3, April, Washington.
- Jord`a, `O., M. Schularick and A. Taylor. 2013. “When Credit Bites Back,” Journal of Money, Credit and Banking, 45(s2): 3-28.
- Kalemli-Özcan, S., L. Laeven and D. Moreno. 2018. “Debt Overhang, Rollover Risk, and Corporate Investment: Evidence from the European Crisis,” NBER Working Paper Series 24555, National Bureau of Economic Research.
- Kalemli-Özcan, S., B. Sorensen, C. Villegas-Sanchez, V. Volosovych and S. Yesiltas, 2015. “How to Construct Nationally Representative Firm Level Data from the Orbis Global Database,” NBER Working Paper Series 21558, National Bureau of Economic Research.
- Koske, I., I. Wanner, R. Bitetti and O. Barbiero. 2015. “The 2013 update of the OECD product market regulation indicators: policy insights for OECD and non-OECD countries”, OECD Economics Department Working Papers, No. 1200.
- Nicoletti, G. and S. Scarpetta. 2003. Regulation, Productivity and Growth: OECD Evidence. Economic Policy, 18(36): 9-72.
- Organisation for Economic Cooperation and Development, OECD. 2016. Economic Policy Reforms: Going for Growth, Paris.
- Ramey, G. and V. Ramey. 1995. “Cross-Country Evidence on the Link between Volatility and Growth.” American Economic Review, 85 (5): 1138-51.
- Reinhart, C. and K. Rogoff. 2009. “The Aftermath of Financial Crises,” American Economic Review, 99 (2): 466-72.
- Romer, C. and D. Romer. 2017. “New Evidence on the Aftermath of Financial Crises in Advanced Countries,” American Economic Review, 107(10): 3072-3118.
- Siemer, M. 2016. “Firm Entry and Employment Dynamics in the Great Recession,” Finance and Economics Discussion Series 2014-56, Board of Governors of the Federal Reserve System.
- Wooldridge, J. M. 2009. “On Estimating Firm-level Production Functions Using Proxy Variables to Control for Unobservables.” Economics Letters, 104 (3): 112–114.

### Additional working papers and notes cited
- Benmelech, E., N. Bergman and A. Seru. 2011. “Financing Labor,” NBER Working Paper No. 17144.
- Bordo, M., and J. Landon-Lane. 2013. NBER Working Paper No. 19585.
- Corrado, C., J. Haskel, C. Jona-Lasinio and M. Iommi. 2016. EIB Working Paper No. 2016/08.v
- Correa-Caro. C, L. Medina, M. Poplawski-Ribeiro, B. Sutton. 2018. IMF Working Paper No.18/251.
- De Loecker, J., J. Eeeckhout. And G. Unger. Forthcoming. Quarterly journal of Economics.
- Diez. F., J. Fan and C. Villegas-Sanchez. 2019. IMF Working Paper.
- Dinlersoz, E., S. Kalemli-Özcan, H. Hyatt, and V. Penciakova. 2018. NBER Working Paper No. 25226.
- Duval, R., G. Hong and Y. Timmer. 2020. Review of Financial Studies, 33(2): 475-503.
- Kalemli-Özcan, S., L. Laeven and D. Moreno. 2018. NBER Working Paper Series 24555.
- Kalemli-Özcan, S., B. Sorensen, C. Villegas-Sanchez, V. Volosovych and S. Yesiltas, 2015. NBER Working Paper Series 21558.
- Siemer, M. 2016. Finance and Economics Discussion Series 2014-56.

### Figures

### Figure 1: Pre- and Post-GFC Trends in Intangible and Tangible Investment Growth in High-Leverage vs. Low-Leverage Firms
- Panel (a): Average intangible investment growth for high- and low-leverage firms.
- Panel (b): Average tangible investment growth for high- and low-leverage firms.
- Note: Series represent coefficient estimates on year dummy variables from regressions of firm-level intangible (panel a) or tangible (panel b) investment growth on year dummies and country-sector fixed effects separately for high-leverage firms (blue solid line) and low-leverage firms (red solid line). Leverage threshold to split groups is, in each country-sector separately, the median across firms of the average leverage ratio over the pre-crisis period.

### Figure 2: Change in Average TFP Growth and Average Intangible Investment Growth between the Pre-GFC and post-GFC periods
- Note: For each firm, the change in the intangible-to-total-assets ratio (in percent) between pre- and post-crisis periods is calculated and ranked, broken into 100 quantiles. Each dot represents the quantile-median deviation from a country-sector fixed effect (x-axis) plotted against the quantile-median deviation of the change in average TFP growth from a country-sector fixed effect (y-axis). Post- and pre-crisis periods include five years after and before the 2008 crisis, respectively.

### Figure 3: Illustration of the Baseline Estimation Results: Financial Frictions and Intangible Investment
- Note: High (low) leverage corresponds to the 75th (25th) percentile of the cross-firm distribution of pre-crisis average leverage ratio. Green shaded bar indicates the difference in estimated effects for high and low leverage firms. Estimated coefficients are from column (1) in the baseline results table.

### Figure 4: Illustration of the Baseline Estimation Results: The Role of Counter-cyclical Policy
- Note: High (low) leverage corresponds to the 75th (25th) percentile of the cross-firm distribution of pre-crisis average leverage ratio. Green shaded bars indicate the difference in estimated effects for high and low leverage firms, separately for contractionary and expansionary monetary conditions. Estimated coefficients are from column (2) in the baseline results table. Expansionary/contractionary monetary conditions are defined as forecast errors in 10-year gov’t bond yields by +/-50 bps.

### Figure 5: Illustration of the Baseline Estimation Results: Complementarity between Product Market Competition and Counter-cyclical Policy
- Note: High (low) leverage corresponds to the 75th (25th) percentile of the cross-firm distribution of pre-crisis average leverage ratios. Green shaded bars indicate the difference in estimated effects for high and low leverage firms in contractionary and expansionary monetary conditions, respectively; red bars measure the difference between them, separately for strong and weak competition environments. Estimated coefficients are from column (4) in the baseline results table. Expansionary/contractionary monetary conditions are defined as forecast errors in 10-year gov’t bond yields by +/-50 bps. Weak (strong) competition corresponds to the 75th (25th) percentile of the country-sector distribution of pre-crisis average Lerner index values.

### Tables

### Table 1: Summary Statistics
- Variables and distributional statistics (based on 664,086 observations of the baseline estimation sample):
  - ∆Intangible Investment: Mean -0.0012; Median 0.0005; 25th percentile 0.0023; 75th percentile 0.0231.
  - ∆Tangible Investment: Mean -0.01400; Median -0.0315; 25th percentile 0.0266; 75th percentile 0.2080.
  - Intangible / Total assets ratio: Mean 0.16; Median 0.02; 25th percentile 0.21; 75th percentile 0.25.
  - Leverage ratio: Mean 0.67; Median 0.70; 25th percentile 0.53; 75th percentile 0.84; Standard deviation 0.21.
  - Lerner index: Mean 0.06; Median 0.05; 25th percentile 0.03; 75th percentile 0.08; Standard deviation 0.06.
- Note: ∆Intangible Investment is the difference in the average net investment in intangible assets (as a ratio of total assets) between post- and pre-crisis periods. ∆Tangible Investment is similarly defined. Intangible / Total assets ratio is the ratio of intangible assets to total (tangible+intangible) assets in the pre-crisis period. Leverage ratio is defined as the average debt-to-assets ratio in the pre-crisis period. Lerner index is computed as the average of (EBITDA-Depreciation and amortization)/Operating revenue in the pre-crisis period.

### Table 2: Baseline Estimation Results: Intangible Investment and Leverage
- Columns (1) and (2):
  - Vulnerability_isc: -0.005*** (0.000) in column (1); -0.014*** (0.005) in column (2).
  - Vulnerability_isc × Pre−crisis physical assets ratio_isc: 0.010* (0.006) in column (2).
- Observations: 664,086; R-squared: 0.061 (col 1) and 0.065 (col 2).
- Country-Sector FE: Yes.
- Note: Dependent variable is the difference in the average net investment ratio in intangible assets (ratio of intangible to total assets) between post- and pre-crisis periods. Firm-level vulnerability is the average debt-to-assets ratio in the pre-crisis period. The post-crisis period starts in 2008. Firm-specific controls included but not reported. Standard errors clustered at the country-sector level. *: significant at 10% level; **: significant at 5% level; ***: significant at 1% level.

### Table 3: Baseline Estimation Results: Intangible Investment, Leverage and Macro Policies
- Columns (1)–(3):
  - Vulnerability_isc: -0.005*** (0.000); -0.007*** (0.001); -0.009*** (0.001).
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.005*** (0.001) in col (1); 0.009*** (0.001) in col (2).
  - Vulnerability_isc × Weak competition_sc: 0.047*** (0.011) in col (2); 0.061*** (0.013) in col (3).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.074*** (0.019) in col (3).
- Observations: 664,086 in all columns. R-squared: 0.062, 0.061, 0.062.
- Country-Sector FE: Yes.
- Note: Expansionary monetary conditions defined as average OECD forecast error for long term (10-year government bond) interest rate in the post-crisis period. Weak competition measured as median pre-crisis Lerner index in each country-sector. Firm-specific controls included but not reported. Standard errors clustered at country-sector level. Significance: * 10%; ** 5%; *** 1%.

### Table 4: Baseline Estimation Results: Intangible Investment vs. Tangible Investment
- Column (1) Intangible to total assets ratio:
  - Vulnerability_isc: -0.022*** (0.004).
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.027*** (0.006).
  - Vulnerability_isc × Weak competition_sc: -0.141*** (0.053).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.233*** (0.079).
  - Observations: 664,084; R-squared 0.012; Country-Sector FE Yes.
- Column (2) Investment in intangible assets:
  - Vulnerability_isc: -0.017*** (0.003).
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.002 (0.005).
  - Vulnerability_isc × Weak competition_sc: -0.001 (0.048).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: 0.007 (0.075).
  - Observations: 674,266; R-squared 0.033; Country-Sector FE Yes.
- Note: Dependent variables: column (1) difference in average ratio of intangible assets to total assets between post- and pre-crisis periods; column (2) difference in average net investment in tangible assets (as ratio of total assets) between post- and pre-crisis periods. Expansionary monetary conditions and Weak competition defined as in previous tables. Firm-specific controls included but not reported. Standard errors clustered at country-sector level. Significance: * 10%; ** 5%; *** 1%.

### Table 5: Alternative Measures of Macroeconomic Policy Shocks
- Expansionary policy measures across columns (1)–(3): Deviation from Taylor rule; Forecast errors in short term policy rate; Forecast errors in gov’t consumption.
- Key coefficients:
  - Vulnerability_isc: -0.005*** (0.001) in col (1); -0.011*** (0.001) in col (2); -0.015*** (0.001) in col (3).
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.004*** (0.000) in col (1); 0.010*** (0.003) in col (2); 0.014*** (0.001) in col (3).
  - Vulnerability_isc × Weak competition_sc: 0.019*** (0.010); 0.062*** (0.015); 0.099*** (0.015).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.016** (0.006); -0.072* (0.038); -0.121*** (0.017).
- Observations: 664,086 (col 1); 571,482 (col 2); 647,836 (col 3). R-squared: 0.063, 0.062, 0.063.
- Note: Expansionary monetary conditions defined differently across columns as specified. Weak competition measured as median pre-crisis Lerner index. Firm-specific controls included but not reported. Standard errors clustered at country-sector level. Significance: * 10%; ** 5%; *** 1%.

### Table 6: Alternative Measures of Product Market Competition
- Competition measures across columns (1)–(4): Median markup (country-sector); Product market regulation; Regulatory protection of incumbents; Administrative burden for startups.
- Key coefficients (all columns use 664,086 observations; R-squared 0.062; Country-Sector FE Yes):
  - Vulnerability_isc: -0.011*** (0.001); -0.018*** (0.002); -0.042*** (0.003); -0.030*** (0.003) for columns (1)–(4) respectively.
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.011*** (0.002); 0.041*** (0.003); 0.067*** (0.005); 0.052*** (0.005).
  - Vulnerability_isc × Weak competition_sc: 0.003*** (0.001); 0.008*** (0.001); 0.030*** (0.003); 0.009*** (0.001).
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.004*** (0.001); -0.021*** (0.002); -0.049*** (0.004); -0.017*** (0.002).
- Note: Dependent variable is the difference in the average net investment in intangible assets (as a ratio of total assets) between post- and pre-crisis periods. Expansionary monetary conditions is the average OECD forecast error for long term (10-year government bond) interest rate in the post-crisis period. Weak competition measured as median markup (col 1); OECD PMR indicator in 2008 (col 2); OECD RPI in 2008 (col 3); OECD ABS in 2008 (col 4). Firm-specific controls included but not reported. Standard errors clustered at country-sector level. Significance: * 10%; ** 5%; *** 1%.

*References and statistical tables reproduced from the source content.*

### 4.  The post-crisis period starts in 2008.  Firm-specific controls (included in regressions but not reported ) are firm 

### Table 7: Further Robustness Checks

### Regression specification and dependent variable
- Dependent variable: the difference in the average net investment in intangible assets (as a ratio of total assets) between post- and pre-crisis periods.
- Post-crisis period starts in 2008.
- Firm-level Vulnerability:
  - Measured as the average debt-to-assets ratio in the pre-crisis period in columns (1), (3), and (4).
  - Measured as the interest coverage ratio (the average ratio of interest payments to earnings (EBITDA)) in the pre-crisis period in column (2).
- Expansionary monetary conditions: the average OECD forecast error for long term (10-year government bond) interest rate in the post-crisis period (measure of more-than-expected policy loosening).
- Weak competition: median pre-crisis Lerner index value in each country-sector (reflecting degree of profitability).
- Firm-specific controls (included but not reported): firm age, total assets, and cash-flow/assets ratio as well as their interaction terms with Expansionary monetary conditions and/or Weak competition measures.
- All columns include country-sector fixed effects.
- Standard errors are clustered at the country-sector level.
- Significance markers: * significant at 10% level; ** significant at 5% level; *** significant at 1% level.

### Key coefficient estimates (by column)
- Column (1) Shorter window (2005-2010)
  - Vulnerability_isc: -0.013***
    - (0.001)
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.011***
    - (0.002)
  - Vulnerability_isc × Weak competition_sc: 0.093***
    - (0.015)
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.100***
    - (0.024)
  - Observations: 615,143
  - R-squared: 0.074
  - Country-Sector FE: Yes

- Column (2) Alternative vulnerability (interest coverage) (664,453 observations)
  - Vulnerability_isc: -0.002***
    - (0.000)
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.004***
    - (0.001)
  - Vulnerability_isc × Weak competition_sc: 0.012*
    - (0.007)
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.028***
    - (0.009)
  - Observations: 664,453
  - R-squared: 0.060
  - Country-Sector FE: Yes

- Column (3) Linear probability model (probability) (664,086 observations)
  - Vulnerability_isc: -0.175***
    - (0.013)
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.078***
    - (0.023)
  - Vulnerability_isc × Weak competition_sc: 0.879***
    - (0.209)
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.623*
    - (0.322)
  - Observations: 664,086
  - R-squared: 0.087
  - Country-Sector FE: Yes

- Column (4) Excluding zeros (no intangible assets) (518,048 observations)
  - Vulnerability_isc: -0.010***
    - (0.001)
  - Vulnerability_isc × Expansionary monetary conditions_c: 0.011***
    - (0.002)
  - Vulnerability_isc × Weak competition_sc: 0.069***
    - (0.016)
  - Vulnerability_isc × Expansionary monetary conditions_c × Weak competition_sc: -0.096***
    - (0.023)
  - Observations: 518,048
  - R-squared: 0.072
  - Country-Sector FE: Yes

### Robustness tests and interpretations (as presented)
- Column 1: shorter window between 2005 and 2010.
- Column 2: alternative vulnerability measure using interest coverage (average ratio of interest payments to earnings (EBITDA)) in the pre-crisis period.
- Column 3: linear probability model by replacing the non-negative dependent variable with 1 (and 0 otherwise).
- Column 4: excludes observations without intangible assets in both periods (hence no change in intangible investment during the periods).
- Across specifications, higher pre-crisis firm vulnerability (higher debt-to-assets or lower interest coverage) is associated with a negative coefficient on Vulnerability_isc and positive coefficients on the two-way interactions with Expansionary monetary conditions and Weak competition, with a negative coefficient on the three-way interaction.

*Source: wpiea2020025-print-pdf.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020025-print-pdf.pdf_