## 1. Summary Statistics of the Regression Sample

## Source details

**Canonical URL:** [1. Summary Statistics of the Regression Sample](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16248.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16248.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16248.pdf.json)

---

### I. Introduction — scope and research question
- Objective: Investigate whether corporate indebtedness and overall balance sheet soundness contributed to the investment cycle in Cyprus.
- Data: Firm-level panel of Cypriot non-financial firms over the 2004–14 period covering leverage, cash, earnings, and debt maturity.
- Identification/estimation: System general method of moments (GMM) model for panel data.

- Key mechanisms investigated:
  - High leverage: tight financial constraints reducing ability to invest.
  - High cash holdings: may offset leverage (precautionary) or reflect agency problems/poor investment opportunities.
  - Low earnings relative to debt: debt overhang where marginal benefits of investment accrue to debt holders, deterring equity-financed investment.
  - Debt maturity: affects tradeoff between investing and reducing indebtedness.

- Principal empirical finding (firm level):
  - Overall corporate indebtedness (total debt to assets, and net debt [total debt minus cash] to assets) is negatively associated with investment over the 2004–14 boom-bust cycle.
  - The negative effect of indebtedness on investment is weaker since the Cypriot banking crisis than before the crisis (interpretation: post-crisis excess capacity may allow utilization without additional credit).
  - Corporate cash holdings are negatively associated with investment, consistent with agency-theory interpretation of retained cash reflecting poor investment opportunities.

- Magnitude of estimated effect (firm-level → macro extrapolation):
  - All else equal, a 10 percentage point decrease (increase) in total debt to assets ratio is associated with a 3 to 6 percentage point increase (decrease) in investment rate.
  - In the sample:
    - Mean investment rate decreased from a peak of 4 percent in 2008 to -10 percent in 2014.
    - Mean total debt to assets ratio increased from 61 percent to 68 percent over the same period.
  - Extrapolation implies the increase in corporate leverage (total debt to assets) can explain 1/6 to 1/3 of the decline in mean corporate investment rate in the sample.

### II. Stylized facts about investment and corporate balance sheet in Cyprus

- A. Fixed investment — boom and bust
  - Cyprus economy:
    - Expanded by 24 percent over the period between EU accession in 2004 and its peak in 2008.
    - Brief contraction after 2008 GFC and collapsed during the Cypriot banking crisis over 2012–14 with output contracting by more than 10 percent over this three-year period.
    - GDP in 2014 was 10 percent below its 2008 peak.
    - Economy recovered with 1.7 percent of growth in 2015; robust tourism and professional services were major contributors.
  - Fixed capital investment:
    - Share in GDP increased from 21 percent in 2004 to 27 percent in 2008.
    - Since the GFC, the share dropped to 13 percent in 2015, with the level of fixed investment at half its 2008 peak.
    - The contraction of fixed investment contributed more than 15 percentage points of GDP to the contraction.
    - During the boom years over 2004–08 with growth averaging 4¼ percent per year, fixed investment contributed 2¼ percentage points per year despite being about one quarter of GDP on average.
  - Sectoral composition:
    - Housing construction accounted for more than three-quarters of the decline in fixed investment since 2008.
    - Investment in metal product and machinery equipment was much less affected.
  - Housing market specifics:
    - Housing prices jumped by 135 percent over 2003–08.
    - Number of newly completed dwellings rose from 8,700 in 2003 to 18,200 in 2008.
    - Housing prices plunged by 30 percent from their peak.
    - Number of new dwellings shrank to one-quarter of its peak while stock of housing continued to rise.
    - The collapse of housing construction explains a large share of the investment decline.

- B. Corporate balance sheet — aggregate dynamics and distress
  - Credit and debt:
    - Bank credit to non-financial corporations (NFCs) doubled over 2006–08.
    - Corporate financial debt peaked at 275 percent of GDP in 2012.
    - Corporate debt represented 57 percent of total liabilities; remainder largely unlisted equity (97 percent of total equity).
    - Corporate leverage (debt-to-equity) ratio was 135 percent at end-2012.
  - Financial assets and net position:
    - Corporate financial assets stood at 300 percent of GDP at end-2012.
    - Given their low weight (18 percent of total assets), corporate net financial assets were -184 percent of GDP at end-2012.
    - From December 2012 to December 2015, financial assets fell by 9 percent (mainly due to decline in bank deposits), and corporate net financial asset position widened to 226 percent of GDP as of end-December 2015.
  - Non-performing loans (NPLs) and loan quality:
    - System-wide NPLs to total loans ratio: jumped from 5 percent in early 2010 to 16 percent in the first quarter of 2013; from 30 percent in the second quarter of 2013 to 45 percent at end-2015 (partly due to definitional/aggregation changes).
    - Corporate NPLs accounted for half of system-wide NPLs; corporate NPL ratio reached 55 percent at end-2015.
    - Loans to the two housing-related sectors—construction and real estate services—account for 43 percent of total domestic bank credits at end-2015.
    - Three-quarters of bank credit to the construction sector and 55 percent of credit to the real estate services sector became impaired.
    - Construction and real estate services together were responsible for 52 percent of total corporate NPLs at end-2015.

### III. Firm-level analysis: role of balance sheet strength on investment boom and bust

- Data advantages and caveats
  - Advantages:
    - Large number of observations permits firm fixed effects and control of confounders.
    - Coverage includes small and medium-size firms as well as large firms.
    - Allows examination of multiple balance sheet indicators: leverage, cash, earnings, debt maturity.
  - Caveats:
    - Many observations are consolidated accounts (domestic and foreign activities); consolidation may be appropriate if firms make consolidated decisions.
    - Some firms dropped from regressions due to missing data and lagged estimation structure.
    - Important indicators such as interest coverage ratio are not included due to limited coverage.
    - New firms (less than 3 years old at end of sample) are excluded by data and estimation methodology.

- Empirical specification and interpretation
  - Estimation approach: system GMM panel for 2004–14.
  - Main interpreted channels:
    - Leverage: negative association with investment consistent with financial constraints and debt overhang.
    - Cash holdings: negative association with investment consistent with agency-theory interpretation (retention of cash when investment opportunities are poor).
  - Temporal heterogeneity:
    - Effect of indebtedness on investment is stronger in the pre-crisis period than during/after the Cypriot banking crisis.
    - Possible interpretation: post-crisis excess capacity reduces need for credit to increase utilization.

- Limitations on causal interpretation and sectoral heterogeneity
  - Alternative explanations remain, notably for housing construction:
    - Collapse of housing market may produce simultaneous falling investment and rising debt as developers draw on previously approved credit lines and cannot service loans.
    - Limited firm-level data in housing construction sector prevents formal testing of this channel.

### A. Data and Measurement
- Data source: Orbis database by Bureau van Dijk (commercial dataset, national business registers; includes financial and ownership information on publicly listed and private companies including Cyprus).
- Sample period: 2005 to 2014.
  - Phases covered: post-EU membership expansion period (pre-2008), the GFC and great recession period (2008–11), and the Cypriot banking crisis and recovery period (2012–14).
  - Year 2012 defined as start of Cypriot banking crisis (methodology of Laeven and Valencia (2013)). Years prior to 2012 = pre-crisis period; 2012 onward = post-crisis period.
- Sector classification: NACE 4-digit classification grouped into 19 industries.
- Sample coverage: Orbis sample ~2,000 firms with various period coverage; concentrated in services—close to half in wholesale and retail trade; 15 percent in finance and insurance sector.
- Investment measure:
  - Net investment rate = annual change in fixed tangible assets divided by fixed tangible assets.
  - Net investment preferred over gross investment for relation to capital capacity and future productivity; maximizes sample coverage (includes firms missing investment expenditure or depreciation).
  - Note: gross investment also calculated as net investment plus depreciation; using gross investment does not change main results.
- Balance sheet explanatory variables:
  - Leverage: ratio of debt to assets. Two debt measures:
    - Total debt = sum of long-term debt, loans, credit, and other current liabilities.
    - Net debt = total debt minus cash and cash equivalent.
  - Cash = ratio of cash and cash equivalent to assets.
  - Earnings = EBITDA to debt (captures debt overhang; earnings are more transitory than stock of debt or assets).
  - Debt maturity = share of long-term debt in total debt (or ratio of long-term debt to net debt).
- Controls and transformations:
  - Firm-level controls X include sales growth, total assets (size), and Tobin’s q.
  - Nominal variables converted into Euros using year-end exchange rate and then into real using Cyprus GDP deflator (2010 base year).
- Sample restrictions for regressions:
  - Exclude financial and insurance firms, public sector firms, and firms with missing data.
  - Include only firms with three or more years of observations due to lag structure of GMM estimator.

### B. Data Summary
- Investment and balance sheet trends (aggregate firm-level):
  - Average net investment rate: 4 percent in 2008 → -2 percent in 2009 → -10 percent in 2014.
    - Interpretation: since 2009, investment has not covered depreciation of capital.
  - Net investment rates negative across key sectors over 2012–14; strongest drop in construction sector.
  - Debt-to-assets ratio: average 68 percent in 2014, up from 58 percent in 2010 (10 percentage-point increase).
    - Construction and real estate sector had largest increase.
  - Cash-to-assets ratio: average below 10 percent.
  - Earnings-to-debt ratio: improved and turned positive in 2014 from -5 percent in 2013 (mainly due to wholesale and retail sector).
  - Corporate sales growth: declined from 14 percent in 2008 to -4 percent in 2013.
  - Debt maturity: long-term debt share rose to 42 percent of total debt in 2014, up from 24 percent in 2005.
  - Debt concentration: in 2014, about one-third of corporate debt was held by illiquid or insolvent firms, up from less than 8 percent in 2010.
- Regression sample:
  - Remaining sample: 80 firms and about 300 observations after exclusions.
  - 85 percent of observations are publicly listed companies.
  - Average firm size: 350 million USD; median 180 million USD (sample skewed to small-sized observations).

- Selected exact Table 1 summary statistics:
  - Net investment rate: Mean 0.012; St. Dev. 0.312; Min -2.319; Median -0.018; Max 0.969; N 307.
  - Total debt / Assets: Mean 0.495; St. Dev. 0.252; Min 0.058; Median 0.492; Max 1.875; N 307.
  - Net debt / Assets: Mean 0.427; St. Dev. 0.281; Min -0.686; Median 0.444; Max 1.855; N 304.
  - Cash / Assets: Mean 0.067; St. Dev. 0.085; Min 0.000; Median 0.037; Max 0.521; N 302.
  - Earnings / Debt: Mean 0.228; St. Dev. 0.451; Min -3.363; Median 0.121; Max 2.395; N 307.
  - Long-term debt / Debt: Mean 0.400; St. Dev. 0.290; Min 0.000; Median 0.397; Max 0.918; N 307.
  - Sales growth: Mean 0.028; St. Dev. 0.347; Min -1.021; Median 0.000; Max 2.342; N 307.
  - ln(Assets): Mean 11.840; St. Dev. 1.492; Min 7.746; Median 11.978; Max 14.696; N 307.
  - Tobin’s q: Mean 2.212; St. Dev. 3.603; Min -1.788; Median 1.027; Max 24.079; N 250.

### C. Empirical Methodology
- Baseline investment regression:
  - I_it = α + β Leverage + γ Cash + δ Earnings + η Maturity + × X_it + firm and year effects + ε_it
  - I_it is net investment rate; Leverage, Cash, Earnings, Maturity defined as in Section III A; X includes sales growth, total assets, Tobin’s q.
- Estimation challenges and solution:
  - Endogeneity of explanatory variables; firm-specific effects and idiosyncratic shocks present.
  - Two-step system GMM estimator (Blundell and Bond, 2000) used with lagged endogenous variables as instruments.
- Diagnostics reported:
  - Arellano and Bond AR(1) statistic (first-order serial correlation).
  - Arellano and Bond AR(2) statistic (absence of second-order serial correlation).
  - Hansen statistic (overidentification/joint validity of instruments).
- Implementation details:
  - Endogenous variables: contemporaneous values of leverage, cash, earnings, maturity, sales growth, assets size, Tobin’s q.
  - Year and industry dummies included as exogenous variables.
  - Instruments: lagged levels for first-difference equations; lagged first differences for level equations (one lag used).
  - Firms with less than three years of observations dropped.

### D. Results
- Main results (full sample):
  - Coefficients on leverage (both total debt and net debt) are negative and significant at the 1 percent level in all specifications.
    - Estimated magnitude: a 10 percentage point decrease (increase) in total debt to assets ratio is associated with a 3 to 6 percentage point increase (decrease) in investment rate.
  - Cash holdings coefficients are negative and significant at the 10 percent level in some specifications, implying firms with more cash do not invest more (consistent with agency theory).
  - Demand (sales growth) is an important positive driver of investment (sales growth coefficient significant and positive).
  - After controlling for leverage, debt maturity and the ability to pay debt by cash or earnings play a much smaller role.
- Diagnostics:
  - First order serial correlation (AR(1)) is negative as expected; p-values significant at 5 or 10 percent in all results.
  - No evidence of second order serial correlation (AR(2) p-values not significant).
  - Hansen overidentification tests cannot reject instruments, though p-values suggest instruments may be weak.

- Solvent firms (restricted sample):
  - Sample restricted to firms with total debt < assets (about 5 percent of firms were insolvent and excluded).
  - Results similar to baseline when Tobin’s q is not included.
  - When Tobin’s q is included, leverage coefficients remain negative but are not statistically significant (likely due to negative correlation between leverage and Tobin’s q).
  - Cash coefficients remain negative and significant in all specifications.

- Subsample and temporal results:
  - Pre-crisis (2004–11):
    - Coefficients on leverage remain negative and significant when Tobin’s q not included.
    - Estimated magnitude: before the crisis, a 10 percentage point decrease (increase) in leverage associated with a 6 to 10 percentage point increase (decrease) in investment rate.
  - Post-GFC, Pre-Crisis (2008–11): Very similar results to pre-crisis subsample.
  - Post-crisis effects (interaction with crisis dummy):
    - Linear leverage term remains negative and significant in all specifications.
    - Interaction term total debt × Crisis is positive and significant at the 10 percent level in one specification, while overall effect of leverage remains negative post-crisis.
    - None of the interaction terms with the crisis dummy are significant except earnings in one specification.
    - Conclusion: main balance sheet driver of investment (leverage) remains unchanged after the crisis.
  - Interpretation: effect of indebtedness on investment appears smaller after the crisis than before; possible interpretation is that post-crisis spare capacity can be utilized without much additional credit, implying a creditless recovery is possible.

### IV. Conclusion
- Main empirical finding:
  - Strong and negative effect of corporate indebtedness on investment over the boom-bust cycle in Cyprus.
  - Estimated magnitude: a 10 percentage point decrease (increase) in leverage (total debt to assets) is associated with a 3 to 6 percentage point increase (decrease) in investment rate over the last decade.
  - Extrapolation: the increase in corporate leverage may account for 1/6 to 1/3 of the decline in corporate investment from its 2008 peak.

- Policy implications and recommendations:
  - Need to repair corporate balance sheets to support faster recovery, sustainable rise in investment, and macrofinancial stability.
  - Cyprus progress:
    - New insolvency framework allows over-indebted borrowers to restructure debt, providing viable companies opportunity to repair balance sheets.
    - Banks have internal workout policies to facilitate debt restructuring.
  - Overall recommendation: a comprehensive policy effort to reduce corporate debt and improve balance sheet strength would contribute to faster recovery, higher investment, and macrofinancial stability.

* _wp16248 - 1. Summary Statistics of the Regression Sample (PDF chapter), canonical URL provided in the content unit._

### 1. Summary Statistics of the Regression Sample.....................................................................17

### 1. Summary Statistics of the Regression Sample

### I. Introduction — scope and research question
- Objective: Investigate whether corporate indebtedness and overall balance sheet soundness contributed to the investment cycle in Cyprus.
- Data: Firm-level panel of Cypriot non-financial firms over the 2004–14 period covering leverage, cash, earnings, and debt maturity.
- Identification/estimation: System general method of moments (GMM) model for panel data.

### Key mechanisms investigated
- High leverage: tight financial constraints reducing ability to invest.
- High cash holdings: may offset leverage (precautionary) or reflect agency problems/poor investment opportunities.
- Low earnings relative to debt: debt overhang where marginal benefits of investment accrue to debt holders, deterring equity-financed investment.
- Debt maturity: affects tradeoff between investing and reducing indebtedness.

### Principal empirical finding (firm level)
- Overall corporate indebtedness (total debt to assets, and net debt [total debt minus cash] to assets) is negatively associated with investment over the 2004–14 boom-bust cycle.
- The negative effect of indebtedness on investment is weaker since the Cypriot banking crisis than before the crisis (interpretation: post-crisis excess capacity may allow utilization without additional credit).
- Corporate cash holdings are negatively associated with investment, consistent with agency-theory interpretation of retained cash reflecting poor investment opportunities.

### Magnitude of estimated effect (firm-level → macro extrapolation)
- All else equal, a 10 percentage point decrease (increase) in total debt to assets ratio is associated with a 3 to 6 percentage point increase (decrease) in investment rate.
- In the sample:
  - Mean investment rate decreased from a peak of 4 percent in 2008 to -10 percent in 2014.
  - Mean total debt to assets ratio increased from 61 percent to 68 percent over the same period.
- Extrapolation implies the increase in corporate leverage (total debt to assets) can explain 1/6 to 1/3 of the decline in mean corporate investment rate in the sample.

---

### II. Stylized facts about investment and corporate balance sheet in Cyprus

H3: A. Fixed investment — boom and bust
- Cyprus economy:
  - Expanded by 24 percent over the period between EU accession in 2004 and its peak in 2008.
  - Brief contraction after 2008 GFC and collapsed during the Cypriot banking crisis over 2012–14 with output contracting by more than 10 percent over this three-year period.
  - GDP in 2014 was 10 percent below its 2008 peak.
  - Economy recovered with 1.7 percent of growth in 2015; robust tourism and professional services were major contributors.
- Fixed capital investment:
  - Share in GDP increased from 21 percent in 2004 to 27 percent in 2008.
  - Since the GFC, the share dropped to 13 percent in 2015, with the level of fixed investment at half its 2008 peak.
  - The contraction of fixed investment contributed more than 15 percentage points of GDP to the contraction.
  - During the boom years over 2004–08 with growth averaging 4¼ percent per year, fixed investment contributed 2¼ percentage points per year despite being about one quarter of GDP on average.
- Sectoral composition:
  - Housing construction accounted for more than three-quarters of the decline in fixed investment since 2008.
  - Investment in metal product and machinery equipment was much less affected.
- Housing market specifics:
  - Housing prices jumped by 135 percent over 2003–08.
  - Number of newly completed dwellings rose from 8,700 in 2003 to 18,200 in 2008.
  - Housing prices plunged by 30 percent from their peak.
  - Number of new dwellings shrank to one-quarter of its peak while stock of housing continued to rise.
  - The collapse of housing construction explains a large share of the investment decline.

H3: B. Corporate balance sheet — aggregate dynamics and distress
- Credit and debt:
  - Bank credit to non-financial corporations (NFCs) doubled over 2006–08.
  - Corporate financial debt peaked at 275 percent of GDP in 2012.
  - Corporate debt represented 57 percent of total liabilities; remainder largely unlisted equity (97 percent of total equity).
  - Corporate leverage (debt-to-equity) ratio was 135 percent at end-2012.
- Financial assets and net position:
  - Corporate financial assets stood at 300 percent of GDP at end-2012.
  - Given their low weight (18 percent of total assets), corporate net financial assets were -184 percent of GDP at end-2012.
  - From December 2012 to December 2015, financial assets fell by 9 percent (mainly due to decline in bank deposits), and corporate net financial asset position widened to 226 percent of GDP as of end-December 2015.
- Non-performing loans (NPLs) and loan quality:
  - System-wide NPLs to total loans ratio: jumped from 5 percent in early 2010 to 16 percent in the first quarter of 2013; from 30 percent in the second quarter of 2013 to 45 percent at end-2015 (partly due to definitional/aggregation changes).
  - Corporate NPLs accounted for half of system-wide NPLs; corporate NPL ratio reached 55 percent at end-2015.
  - Loans to the two housing-related sectors—construction and real estate services—account for 43 percent of total domestic bank credits at end-2015.
  - Three-quarters of bank credit to the construction sector and 55 percent of credit to the real estate services sector became impaired.
  - Construction and real estate services together were responsible for 52 percent of total corporate NPLs at end-2015.

---

### III. Firm-level analysis: role of balance sheet strength on investment boom and bust

H3: Data advantages and caveats
- Advantages:
  - Large number of observations permits firm fixed effects and control of confounders.
  - Coverage includes small and medium-size firms as well as large firms.
  - Allows examination of multiple balance sheet indicators: leverage, cash, earnings, debt maturity.
- Caveats:
  - Many observations are consolidated accounts (domestic and foreign activities); consolidation may be appropriate if firms make consolidated decisions.
  - Some firms dropped from regressions due to missing data and lagged estimation structure.
  - Important indicators such as interest coverage ratio are not included due to limited coverage.
  - New firms (less than 3 years old at end of sample) are excluded by data and estimation methodology.

H3: Empirical specification and interpretation
- Estimation approach: system GMM panel for 2004–14.
- Main interpreted channels:
  - Leverage: negative association with investment consistent with financial constraints and debt overhang.
  - Cash holdings: negative association with investment consistent with agency-theory interpretation (retention of cash when investment opportunities are poor).
- Temporal heterogeneity:
  - Effect of indebtedness on investment is stronger in the pre-crisis period than during/after the Cypriot banking crisis.
  - Possible interpretation: post-crisis excess capacity reduces need for credit to increase utilization.

H3: Limitations on causal interpretation and sectoral heterogeneity
- Alternative explanations remain, notably for housing construction:
  - Collapse of housing market may produce simultaneous falling investment and rising debt as developers draw on previously approved credit lines and cannot service loans.
  - Limited firm-level data in housing construction sector prevents formal testing of this channel.

---

*Italic source attribution: _wp16248 - 1. Summary Statistics of the Regression Sample (PDF chapter), canonical URL provided in the content unit.*

### Section III C for details). Thus our results are silent on the behavior of new firms.

### _wp16248 - Section III C for details). Thus our results are silent on the behavior of new firms.

### A. Data and Measurement
- Data source: Orbis database by Bureau van Dijk (commercial dataset, national business registers; includes financial and ownership information on publicly listed and private companies including Cyprus).
- Sample period: 2005 to 2014.
  - Phases covered: post-EU membership expansion period (pre-2008), the GFC and great recession period (2008–11), and the Cypriot banking crisis and recovery period (2012–14).
  - Year 2012 defined as start of Cypriot banking crisis (methodology of Laeven and Valencia (2013)). Years prior to 2012 = pre-crisis period; 2012 onward = post-crisis period.
- Sector classification: NACE 4-digit classification grouped into 19 industries.
- Sample coverage: Orbis sample ~2,000 firms with various period coverage; concentrated in services—close to half in wholesale and retail trade; 15 percent in finance and insurance sector.
- Investment measure:
  - Net investment rate = annual change in fixed tangible assets divided by fixed tangible assets.
  - Net investment preferred over gross investment for relation to capital capacity and future productivity; maximizes sample coverage (includes firms missing investment expenditure or depreciation).
  - Note: gross investment also calculated as net investment plus depreciation; using gross investment does not change main results.
- Balance sheet explanatory variables:
  - Leverage: ratio of debt to assets. Two debt measures:
    - Total debt = sum of long-term debt, loans, credit, and other current liabilities.
    - Net debt = total debt minus cash and cash equivalent.
  - Cash = ratio of cash and cash equivalent to assets.
    - Theoretical ambiguity: precautionary motive (Almeida and others, 2004) vs. agency theory (Jensen, 1986). Empirical evidence mixed (Miccelson and Partch (2003); Dittmar and others (2003); Opler and others (1999)).
  - Earnings = EBITDA to debt (captures debt overhang; earnings are more transitory than stock of debt or assets).
  - Debt maturity = share of long-term debt in total debt (or ratio of long-term debt to net debt).
- Controls and transformations:
  - Firm-level controls X include sales growth, total assets (size), and Tobin’s q.
  - Nominal variables converted into Euros using year-end exchange rate and then into real using Cyprus GDP deflator (2010 base year).
- Sample restrictions for regressions:
  - Exclude financial and insurance firms, public sector firms, and firms with missing data.
  - Include only firms with three or more years of observations due to lag structure of GMM estimator.

### B. Data Summary
- Investment and balance sheet trends (aggregate firm-level):
  - Average net investment rate: 4 percent in 2008 → -2 percent in 2009 → -10 percent in 2014.
    - Interpretation: since 2009, investment has not covered depreciation of capital.
  - Net investment rates negative across key sectors over 2012–14; strongest drop in construction sector.
  - Debt-to-assets ratio: average 68 percent in 2014, up from 58 percent in 2010 (10 percentage-point increase).
    - Construction and real estate sector had largest increase.
  - Cash-to-assets ratio: average below 10 percent.
  - Earnings-to-debt ratio: improved and turned positive in 2014 from -5 percent in 2013 (mainly due to wholesale and retail sector).
  - Corporate sales growth: declined from 14 percent in 2008 to -4 percent in 2013.
  - Debt maturity: long-term debt share rose to 42 percent of total debt in 2014, up from 24 percent in 2005.
  - Debt concentration: in 2014, about one-third of corporate debt was held by illiquid or insolvent firms, up from less than 8 percent in 2010.
- Regression sample:
  - Remaining sample: 80 firms and about 300 observations after exclusions.
  - 85 percent of observations are publicly listed companies.
  - Average firm size: 350 million USD; median 180 million USD (sample skewed to small-sized observations).
- Table 1 summary statistics (selected exact values):
  - Net investment rate: Mean 0.012; St. Dev. 0.312; Min -2.319; Median -0.018; Max 0.969; N 307.
  - Total debt / Assets: Mean 0.495; St. Dev. 0.252; Min 0.058; Median 0.492; Max 1.875; N 307.
  - Net debt / Assets: Mean 0.427; St. Dev. 0.281; Min -0.686; Median 0.444; Max 1.855; N 304.
  - Cash / Assets: Mean 0.067; St. Dev. 0.085; Min 0.000; Median 0.037; Max 0.521; N 302.
  - Earnings / Debt: Mean 0.228; St. Dev. 0.451; Min -3.363; Median 0.121; Max 2.395; N 307.
  - Long-term debt / Debt: Mean 0.400; St. Dev. 0.290; Min 0.000; Median 0.397; Max 0.918; N 307.
  - Sales growth: Mean 0.028; St. Dev. 0.347; Min -1.021; Median 0.000; Max 2.342; N 307.
  - ln(Assets): Mean 11.840; St. Dev. 1.492; Min 7.746; Median 11.978; Max 14.696; N 307.
  - Tobin’s q: Mean 2.212; St. Dev. 3.603; Min -1.788; Median 1.027; Max 24.079; N 250.

### C. Empirical Methodology
- Baseline investment regression (equation 1):
  - I_it = α + β Leverage + γ Cash + δ Earnings + η Maturity + × X_it + firm and year effects + ε_it
  - I_it is net investment rate; Leverage, Cash, Earnings, Maturity defined as in Section III A; X includes sales growth, total assets, Tobin’s q.
- Estimation challenges:
  - Endogeneity of explanatory variables; error term contains firm-specific effects and idiosyncratic shocks.
  - Fixed effects or GLS inconsistent if leverage not strictly exogenous.
  - Two-step system GMM estimator (Blundell and Bond, 2000) used with lagged endogenous variables as instruments.
- Diagnostics reported:
  - Arellano and Bond AR(1) statistic: tests first-order serial correlation.
  - Arellano and Bond AR(2) statistic: tests absence of second-order serial correlation in first-differenced errors.
  - Hansen statistic: tests overidentification/joint validity of instruments.
- Implementation details:
  - Endogenous variables: contemporaneous values of leverage, cash, earnings, maturity, sales growth, assets size, Tobin’s q.
  - Year and industry dummies included as exogenous variables.
  - Instruments: lagged levels for first-difference equations; lagged first differences for level equations (one lag used).
  - Firms with less than three years of observations dropped.

### D. Results
- Main results (Table 4, full sample):
  - Coefficients on leverage (both total debt and net debt) are negative and significant at the 1 percent level in all specifications.
    - Estimated magnitude: a 10 percentage point decrease (increase) in total debt to assets ratio is associated with a 3 to 6 percentage point increase (decrease) in investment rate.
  - Cash holdings coefficients are negative and significant at the 10 percent level in some specifications, implying firms with more cash do not invest more (consistent with agency theory).
  - Demand (sales growth) is an important positive driver of investment (sales growth coefficient significant and positive).
  - After controlling for leverage, debt maturity and the ability to pay debt by cash or earnings play a much smaller role.
  - Diagnostic tests:
    - First order serial correlation (AR(1)) is negative as expected; p-values significant at 5 or 10 percent in all results.
    - No evidence of second order serial correlation (AR(2) p-values not significant).
    - Hansen overidentification tests cannot reject instruments, though p-values suggest instruments may be weak.
- Solvent firms (Table 5):
  - Sample restricted to firms with total debt < assets (about 5 percent of firms were insolvent and excluded).
  - Results similar to baseline when Tobin’s q is not included.
  - When Tobin’s q is included, leverage coefficients remain negative but are not statistically significant (likely due to negative correlation between leverage and Tobin’s q).
  - Cash coefficients remain negative and significant in all specifications.
- Subsample analysis:
  - Pre-crisis (2004–11, Table 6):
    - Coefficients on leverage remain negative and significant when Tobin’s q not included.
    - Estimated magnitude: before the crisis, a 10 percentage point decrease (increase) in leverage associated with a 6 to 10 percentage point increase (decrease) in investment rate.
  - Post-GFC, Pre-Crisis (2008–11, Table 7):
    - Very similar results to pre-crisis subsample.
  - Post-crisis effects (interaction with crisis dummy, Table 8):
    - Linear leverage term remains negative and significant in all specifications.
    - Interaction term total debt × Crisis is positive and significant at the 10 percent level in one specification, while overall effect of leverage remains negative post-crisis.
    - None of the interaction terms with the crisis dummy are significant except earnings in one specification.
    - Conclusion: main balance sheet driver of investment (leverage) remains unchanged after the crisis.
- Interpretation:
  - Effect of indebtedness on investment appears smaller after the crisis than before.
  - Possible interpretation: post-crisis spare capacity can be utilized without much additional credit → leverage-investment linkage may weaken and a creditless recovery is possible.
  - Cross-country evidence: after a banking crisis and credit boom, recovery would almost certainly be creditless (Abiad and others, 2011); credit-less recoveries on average have output growth a third lower than recoveries with credit (Abiad and others, 2011).

### IV. Conclusion
- Main empirical finding:
  - Strong and negative effect of corporate indebtedness on investment over the boom-bust cycle in Cyprus.
  - Estimated magnitude: a 10 percentage point decrease (increase) in leverage (total debt to assets) is associated with a 3 to 6 percentage point increase (decrease) in investment rate over the last decade.
  - Extrapolation: the increase in corporate leverage may account for 1/6 to 1/3 of the decline in corporate investment from its 2008 peak.
- Policy implications and recommendations:
  - Need to repair corporate balance sheets to support faster recovery, sustainable rise in investment, and macrofinancial stability.
  - Cyprus progress:
    - New insolvency framework allows over-indebted borrowers to restructure debt, providing viable companies opportunity to repair balance sheets.
    - Banks have internal workout policies to facilitate debt restructuring.
  - Overall recommendation: a comprehensive policy effort to reduce corporate debt and improve balance sheet strength would contribute to faster recovery, higher investment, and macrofinancial stability.

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16248.pdf_