## wpiea2020223-print-pdf

## Source details

**Canonical URL:** [wpiea2020223-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020223-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020223-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020223-print-pdf.pdf.json)

---

### I. Introduction — scope and motivation
- Non-financial corporate sector in advanced economies shifted from net borrower to net lender in the 2000s; rising profits did not translate into higher investments, resulting in growing cash holdings of firms. Similar patterns observed in ASEAN5.
- ASEAN5: Indonesia, Malaysia, Philippines, Singapore, and Thailand.
- In ASEAN5, large profits have tended to increase net saving rather than raise dividends or investment.
- Potential drivers and distinguishing features for ASEAN5 corporate saving behavior:
  - Less developed financial markets in Indonesia, Malaysia, Philippines, and Thailand; use of measures such as FXI and capital flow measures to tackle external shocks.
  - Binding borrowing constraints induce precautionary saving to (i) reduce future transaction costs when external funding costs are high (Almeida et al., 2004) or (ii) manage potential liquidity shortfalls when external financing is not available (Han and Qiu, 2007).
  - Export-oriented firms’ foreign earnings provide a hedge, but capital flow measures (e.g., surrender and repatriation requirements) can induce excess precautionary saving.
- Central firm-level characteristic: external financing dependence (EFD), roughly equal to total capital expenditure less cash flows from operations (rigorous definition in Section III).

### Main findings (paper summary)
- Industries with high external financing dependence grew relatively faster in countries with more developed financial systems in ASEAN5 over the past two decades.
- Within a given ASEAN5 country, industries more dependent on external financing tend to save more, ceteris paribus, suggesting binding financial constraints.
- Capital account (KA) openness and exchange rate (ER) flexibility have heterogeneous effects by EFD and export orientation:
  - Greater KA openness or ER depreciation reduces the average saving rate of industries with low dependence on external funds but increases the saving rate of industries with high dependence.
  - KA opening-up or ER depreciation disproportionately reduces the saving of export-oriented industries, except for exporters that are highly dependent on external funds.
- Improvements in banking sector competitiveness or lending efficiency lower the average corporate saving rate.
- Greater political stability is associated with lower corporate saving of all firms.

### II. Data and stylized facts — aggregate and firm-level evidence
- Data and coverage:
  - Firm-level balance sheet and income statement data for a total of 3503 publicly listed firms over 2000–2017 from the Thompson WorldScope Database.
  - 2530 firms listed as of end-2017; total assets account for 75 percent of ASEAN5 GDP.
  - Variables winsorized at 1 percent; non-financial corporations only.
- Key variable definitions:
  - 푃푟표푓푖푡 = 퐺푟표푠푠 푂푝푒푟푎푡푖푛푔 푆푢푟푝푙푢푠 − 푇푎푥푒푠 표푛 푃푟표푓푖푡푠 − 퐼푛푡푒푟푒푠푡
  - 퐺푟표푠푠 푆푎푣푖푛푔 = 푃푟표푓푖푡 − 푁푒푡 퐷푖푣푖푑푒푛푑푠
  - 푁푒푡 푆푎푣푖푛푔 = 퐺푟표푠푠 푆푎푣푖푛푔 – 퐶푎푝퐸푥
- Aggregate stylized facts for ASEAN5:
  - Manufacturing industries are heavily export-oriented; TiVA shows more than half of manufacturing industries have export-orientation rates over 50 percent (Electrical and Electronic Products, Basic Metals, Machinery and Equipment, Chemical and Pharmaceutical Products highest).
  - Malaysia and Singapore exhibit the highest manufacturing export-orientation rates, followed by Thailand.
  - Export orientation rates decreased slightly over TiVA sample period (2005-2015).
  - Indonesia, Malaysia, Thailand, and Philippines have capital flow measures; overall KA openness of ASEAN4 significantly lower than advanced economies, emerging Europe and LACs throughout the sample, with a temporary reversal after the global financial crisis.
  - Financial market metrics: ASEAN4 financial markets less developed than advanced economies but more developed on average than rest of emerging Asia, emerging Europe, and LACs; heterogeneity:
    - Total credit to private sector exceeds 110 percent of GDP in Thailand or Malaysia.
    - Credit-to-GDP ratios below 50 percent in Indonesia or the Philippines.
- Firm-level stylized facts:
  - Gross and net saving rates exhibit cyclical fluctuations during 2000-2017.
  - Upward trend in corporate saving rates ceased in 2008 as average firm profit was halved during the global financial crisis; dip in gross and net saving rates; recovery in 2009–2010; both rates declined since 2011 for 5 years and picked up since 2015.
  - Gross and net saving rates cumulatively increased by 1.8 and 1.9 percentage points, respectively, since 2000.
  - Within-between decomposition (Chen et al., 2017) shows cumulative change in aggregate gross saving rate is entirely driven by the within-age group component; most changes accounted for by within-size group component.
  - Cross-sectional evidence:
    - Large profits lead firms to increase net saving rather than pay higher dividends or increase investment.
    - Strong cross-sectional relationship between trends in gross saving rate and trends in profit relative to total assets; trends in investment uncorrelated with trends in profit, producing positive relationship between profit and net saving rate.

### III. External financing dependence (EFD)
- Definition and measurement:
  - Firm-level EFD: sum over a 10-year period of (CapEx − Opt. CashFlow) divided by sum of CapEx over same period, i.e., (CapEx − Opt. CashFlow)/CapEx.
  - Industry-level EFD: industry median of firm-level EFDs.
  - Footnotes: 10-year cumulative cash flows approximate cumulative cash stock unless large initial cash stock; large and small firms treated equally.
- Rationale for benchmarking to U.S. data:
  - Production functions similar across countries; U.S. capital markets relatively frictionless so U.S. EFD reflects technological demand for financing.
- Evolution and calibration:
  - U.S. observations: young firms depend more on external funds (median EFD declines with age); distribution of EFD shifting downward over time; NASDAQ firms have larger EFD than NYSE.
  - Benchmark: desired EFD for ASEAN5 industry-by-industry proxied by actual EFD of U.S. firms listed on NASDAQ in 1990s.
- Cross-country comparison:
  - Statistically positive relationship between actual EFD in ASEAN5 and desired EFD proxied by U.S. counterparts.
  - More industries in ASEAN5 rely more on internal funds than U.S. counterparts (below 45-degree line in Figure A10), implying either financial constraints or lower technological demand for external financing.
  - Industry EFD ranking matters most for empirical analysis; magnitude second order.

### IV. Empirical analysis
- Models and identification:
  - Framework based on Rajan and Zingales (1998): test whether industries more dependent on external financing grow faster in countries with more developed financial systems.
  - Dynamic panel for firm growth:
    - g_ijkt = α + α_t + α_i + ρ1 g_ijkt−1 + ρ2 g_ijkt−2 + X_it′β + Y_jt′γ + (F_j · m_kt′) δ1 + ε_ijkt
    - g_ijkt: percent change in firm’s total asset per annum.
    - Estimation: Arellano and Bond (1991) GMM.
  - Dynamic panel for net saving:
    - s_ijkt = α + ρ1 s_ijkt−1 + ρ2 s_ijkt−2 + X_it′β + Y_jt′γ + P_kt′δ1 + (F_j · P_kt′) δ2 + ε_ijkt
    - s_ijkt: net saving rate.
    - P_kt includes KA openness (Chinn-Ito Index updated to 2016), ER flexibility (previous-period depreciation in percent), banking sector indicators (Global Financial Development Database), political stability (ICRG).
    - X_it includes ln(total assets), quartile of firm age since incorporation, Tobin’s Q = (market capitalization + total debt)/total assets.
  - Note: F_j omitted separately due to multicollinearity with firm fixed effects; WorldScope uses consolidated accounts; intra-group financing not included.
- Key empirical results:
  - Financial constraints evidence (Table A2):
    - Interaction coefficients between industry EFD and financial development proxies are uniformly significant at the 5-percent level except for total capitalization.
    - Example magnitudes:
      - Column (1): a 1 percentage point increase in the credit-to-GDP ratio associated with Oil and Gas Extraction growing 0.3 ppts per year faster than Rubber and Plastic Products (25th vs 75th percentile example).
      - Column (3): a 1 percentage point reduction in interest rate spread associated with Oil and Gas Extraction growing 1.3 percentage points per year faster than Rubber and Plastic Products.
    - Conclusion: corporates in ASEAN5 may face financial constraints in 2000-2017.
  - Drivers of net corporate saving (Tables A3–A8):
    - EFD × Tobin’s Q: positive relationship — in constrained environments, firms with higher EFD save more, especially when Tobin’s Q is higher.
    - Tobin’s Q (direct): negative relationship with net saving; interaction with EFD positive.
    - Firm size: positive relationship with net saving (after lag controls).
    - Firm age: young firms more dependent on external funds; coefficient on age quartile negative after size controls.
    - Industry average profit: positive coefficient on profit; negative coefficient on square of profit (more profitable industries invest larger share and save less).
    - KA openness × EFD:
      - Greater KA openness reduces net saving for low-demand industries; industries with high EFD save more as KA openness increases (precautionary saving).
    - KA openness × export-orientation:
      - KA openness disproportionately reduces precautionary saving of export-oriented industries; tightening of capital controls would prompt exporters to save more relative to domestic-oriented industries.
    - KA openness × EFD × export orientation (quantile regressions):
      - For export orientation below 60 percent, greater KA openness does not negatively impact saving.
      - For export orientation over 60 percent, firms more dependent on external funds tend to save more as KA becomes more open.
    - ER flexibility × EFD:
      - ER depreciation reduces net saving for industries with low EFD.
      - Above an EFD threshold, depreciation increases net saving — balance-sheet channel dominates competitiveness channel as EFD increases.
      - Export-oriented firms reduce saving more than non-exporters in response to depreciation due to natural hedging.
    - Banking sector efficiency/competitiveness:
      - Increased banking competition or improved lending efficiency lowers corporate saving rate on average.
      - Competitiveness measures: spread between lending and deposit rates, H-statistics, Lerner index, Boone indicator, assets of top three banks as share of total banking assets.
    - Political stability (policy uncertainty proxy):
      - Improvement in political stability reduces corporate saving across all firms; effect does not discriminate by EFD.

### V. Conclusion and policy implications
- Summary:
  - Evidence that non-financial corporations in ASEAN5 may be subject to binding financial constraints: high-EFD industries grow faster in countries with more developed financial systems; high-EFD industries save more within countries.
  - KA openness and ER flexibility have heterogeneous effects by EFD and export orientation.
  - Banking sector competition and lending efficiency reduce corporate saving on average.
  - Greater political stability associated with lower net saving rates across firms.
- Open questions and suggested way forward:
  - Whether actual corporate saving exceeds desired saving and how that gap affects external imbalance remains unanswered.
  - Conceptual and empirical challenge: define and estimate desired corporate saving.
  - Suggested considerations for desired saving estimation:
    - Desired saving should be a function of EFD determined by industry production technology.
    - Firm maturity matters: young firms are more dependent on external funds.
    - Firm-level characteristics (export-orientation, ownership) may be relevant.
    - Need to define desirable macro/financial policy levels (financial development, KA openness, ER flexibility).
  - Further research recommended to deepen understanding of external imbalances and self-financing behavior of credit-constrained firms in fast-growing emerging markets.

### Key statistics and coverage (Table A1 highlights)
- Sample totals and shares (in percent of GDP)
  - IDN: No. of Firms = 451; Total Asset = 28.5%; Gross Saving = 1.8%; Investment = 1.4%; Net Saving = 0.3%
  - MYS: No. of Firms = 820; Total Asset = 170.2%; Gross Saving = 9.9%; Investment = 8.4%; Net Saving = 1.5%
  - PHL: No. of Firms = 163; Total Asset = 79.2%; Gross Saving = 4.1%; Investment = 3.2%; Net Saving = 0.9%
  - SGP: No. of Firms = 516; Total Asset = 110.9%; Gross Saving = 3.5%; Investment = 4.2%; Net Saving = -0.7%
  - THA: No. of Firms = 580; Total Asset = 83.3%; Gross Saving = 5.9%; Investment = 4.3%; Net Saving = 1.6%
  - Avg.: No. of Firms = 506; Total Asset = 75.0%; Gross Saving = 4.1%; Investment = 3.5%; Net Saving = 0.7%

*Source — wpiea2020223-print-pdf - References*

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

### wpiea2020223-print-pdf - References

### I. Introduction — scope and motivation
- The non-financial corporate sector in advanced economies shifted from a net borrower to a net lender in the 2000s; rising profits did not translate into higher investments, resulting in growing cash holdings of firms. Similar patterns are observed in ASEAN5 countries.
- ASEAN5 countries include Indonesia, Malaysia, Philippines, Singapore, and Thailand.
- In ASEAN5, large profits have tended to increase net saving, rather than raise dividends or investment.
- Potential drivers and distinguishing features for ASEAN5 corporate saving behavior:
  - Less developed financial markets in Indonesia, Malaysia, Philippines, and Thailand, and use of measures such as FXI and capital flow measures to tackle external shocks.
  - Firms facing binding borrowing constraints may save more for precautionary motives to (i) reduce future transaction costs when external funding costs are high (Almeida et al., 2004) or (ii) manage potential liquidity shortfalls when external financing is not available (Han and Qiu, 2007).
  - Export-oriented firms’ foreign earnings provide a hedge against financial shocks, but capital flow measures (e.g., surrender and repatriation requirements) can induce excess precautionary saving.
- Industry- and firm-level characteristics interact with country-level features to determine saving behavior; a central firm-level characteristic is external financing dependence (EFD), following Rajan and Zingales (1998).
  - EFD roughly refers to a firm's use of external finance, including borrowings and equity issues; it equals total capital expenditure less cash flows from operations (a rigorous definition is provided in Section III of the source).
  - Industries with high EFD (e.g., mining, refinery, construction, machinery) are expected to be more financially constrained and tend to save more in less developed financial systems.

### Main findings (paper summary)
- In ASEAN5, industries with high external financing dependence grew relatively faster in countries with more developed financial systems over the past two decades.
- Within a given ASEAN5 country, industries more dependent on external financing tend to save more, ceteris paribus, suggesting binding financial constraints.
- The impacts of capital account (KA) openness and exchange rate (ER) flexibility vary across industries by EFD and export orientation:
  - Greater KA openness or ER depreciation reduces the average saving rate of industries with low dependence on external funds but increases the saving rate of industries with high dependence.
  - KA opening-up or ER depreciation disproportionately reduces the saving of export-oriented industries, except for exporters that are highly dependent on external funds.
- Improvements in the banking sector’s competitiveness or lending efficiency lower the average corporate saving rate.
- Greater political stability is associated with lower corporate saving of all firms, presumably due to greater policy clarity.

### A. Literature review — context and mechanisms
- Global structural changes in the non-financial corporate sector: decline in labor income shares, higher market concentration, declining investment and productivity growth (IMF, 2019a); profits rose but investment did not, shifting firms toward net lending (Chen et al., 2017).
- In advanced economies, the saving shift has been driven largely by high-tech companies (Bates et al., 2009; Booth and Zhou, 2013; Begenau and Palazzo, 2017).
- In emerging markets, literature emphasizes credit constraints; examples:
  - Financial reforms reduced saving of previously credit constrained firms relative to unconstrained ones, but corporate sector saving increased post-reform, especially in industries more dependent on external finance (Fan and Kalemli-Özcan, 2016).
  - Bacchetta and Benhima (2015) model complementarity between demand for liquid foreign bonds and domestic investment in fast-growing emerging markets facing liquidity constraints.
- Exchange rate channels on corporate saving:
  - Competitiveness channel: ER depreciation reduces exporters’ saving by improving competitiveness and boosting investment.
  - Balance-sheet channel: Depreciation weakens firms with foreign currency liabilities, reducing access to external finance and raising precautionary internal saving.
  - Competitiveness channel is less effective where invoicing is in a dominant currency or economies are heavily integrated into global supply chains — both features applicable to ASEAN5 (Casas et al., 2016; IMF, 2019b; Amiti et al., 2014).

### II. Data and stylized facts — aggregate and firm-level evidence
- Data and coverage:
  - Firm-level balance sheet and income statement data for a total of 3503 publicly listed firms over 2000–2017, drawn from the Thompson WorldScope Database.
  - 2530 firms are listed as of end-2017, with total assets accounting for 75 percent of ASEAN5 GDP (see Table A1).
  - All variables are winsorized at 1 percent to remove possible effect of outliers.
  - Analysis focuses on non-financial corporations only.
- Key variable definitions (as provided):
  - 푃푟표푓푖푡 = 퐺푟표푠푠 푂푝푒푟푎푡푖푛푔 푆푢푟푝푙푢푠 − 푇푎푥푒푠 표푛 푃푟표푓푖푡푠 − 퐼푛푡푒푟푒푠푡;
  - 퐺푟표푠푠 푆푎푣푖푛푔 = 푃푟표푓푖푡 − 푁푒푡 퐷푖푣푖푑푒푛푑푠;
  - 푁푒푡 푆푎푣푖푛푔 = 퐺푟표푠푠 푆푎푣푖푛푔 – 퐶푎푝퐸푥.
- Aggregate stylized facts for ASEAN5:
  - Manufacturing industries are heavily export-oriented; according to TiVA, more than half of manufacturing industries in ASEAN5 have export-orientation rates over 50 percent, with Electrical and Electronic Products, Basic Metals, Machinery and Equipment, and Chemical and Pharmaceutical Products topping the list.
  - Among ASEAN5, Malaysia and Singapore exhibit the highest manufacturing export-orientation rates, followed by Thailand.
  - Export orientation rates decreased slightly over the TiVA sample period (2005-2015) across almost all countries and industries.
  - Indonesia, Malaysia, Thailand, and Philippines (ASEAN4) have capital flow measures in place; overall KA openness of ASEAN4 has been significantly lower than advanced economies, emerging Europe and LACs throughout the sample, with a temporary reversal following the global financial crisis.
  - Financial market metrics (Global Financial Development Database) show ASEAN4 financial markets are less developed than advanced economies but more developed on average than rest of emerging Asia, emerging Europe, and LACs; heterogeneity exists within ASEAN5:
    - Total credit to private sector exceeds 110 percent of GDP in Thailand or Malaysia (above the average level in advanced economies).
    - Credit-to-GDP ratios are below 50 percent in Indonesia or the Philippines.
- Firm-level stylized facts:
  - Both gross and net saving rates of non-financial firms in ASEAN5 exhibit cyclical fluctuations during 2000-2017 (Figure A1).
  - The upward trend in corporate saving rates ceased in 2008 as average firm profit was halved during the global financial crisis, causing a sudden dip in gross and net saving rates; after recovery in 2009–2010, both rates declined since 2011 for 5 years and picked up since 2015.
  - Gross and net saving rates have cumulatively increased by 1.8 and 1.9 percentage points, respectively, since 2000.
  - Within-between decomposition (following Chen et al., 2017) of changes in aggregate saving rates:
    - Firms divided into groups i=1,2,...,I by quartiles of size or age.
    - Decomposition formula:
      ∆s_t = 1/2 ∑(ω_i,t + ω_i,t−1) ∆s_i,t + 1/2 ∑(s_i,t + s_i,t−1) ∆ω_i,t,
      where ∆s_i,t = s_i,t − s_i,t−1 and ω_i,t is the share of group i in total asset in period t.
    - The cumulative change in aggregate gross saving rate is entirely driven by the within-age group component; most changes are accounted for by changes in the within-size group component rather than changes in the share of old or large firms.
  - Cross-sectional evidence:
    - In cases of large profits, non-financial firms in ASEAN5 tend to increase net saving rather than pay higher dividends or increase investment.
    - Top panels of Figure A7 show a strong cross-sectional relationship between trends in gross saving rate and trends in profit relative to total assets, partly due to weak correlation between profit and dividends.
    - Bottom panels of Figure A7 show that trends in investment are uncorrelated with trends in profit, producing a positive relationship between profit and net saving rate.

*Italic: Source — content unit: wpiea2020223-print-pdf - References*

### conclusions are robust if the 10-year trend is replaced with a 5-year trend.

### wpiea2020223-print-pdf - conclusions are robust if the 10-year trend is replaced with a 5-year trend.

### III. EXTERNAL FINANCING DEPENDENCE
- Definition and measurement
  - External financing dependence (EFD) at the firm level: sum over a 10-year period of a firm’s use of external finance (borrowings and equity issues, which equals total capital expenditure less cash flows from operations) divided by the sum of capital expenditure over the same period, i.e., (CapEx − Opt. CashFlow)/CapEx.
  - Industry-level EFD: industry median of firm-level EFDs.
  - Footnotes:
    - The sum of cash flows over a period of 10 years could be a good approximation of cumulative cash stock, unless the firm had a large initial cash stock in the beginning of the period (e.g., mature firms).
    - Large and small firms are treated equally, preventing large mature firms from swamping information from small firms.
- Rationale for benchmarking to U.S. data
  - Desired dependence on external funds is identified using data on U.S. firms and applied to other countries because:
    - Production functions of the same industry are similar across countries, especially in manufacturing.
    - U.S. capital markets are relatively frictionless, so actual external funds raised reflect technological demand for financing rather than supply constraints.
- Evolution and calibration of EFD
  - Observations from U.S. data:
    - Young firms tend to depend more on external funds (declining median EFD for U.S. firms over time).
    - Distribution of EFD is generally shifting downward over time for firms listed on NASDAQ and NYSE.
    - NASDAQ-listed firms have larger EFD than NYSE-listed firms due to higher shares of young and small firms on NASDAQ.
    - Industry rankings of EFD are relatively stable across decades, with some industries showing larger EFD in the 2010s relative to the 2000s.
  - Benchmark choice: desired EFD for ASEAN5 benchmarked industry-by-industry to the actual EFD of U.S. firms listed on NASDAQ in 1990s.
- Cross-country comparison
  - At industry level, a statistically positive relationship exists between actual EFD in ASEAN5 countries and desired EFD proxied by U.S. counterparts.
  - More industries in ASEAN5 rely more on internal funds (savings) than their U.S. counterparts did (i.e., below the 45-degree line in Figure A10), implying either financial constraints or lower technological demand for external financing relative to the U.S. two decades ago.
  - The EFD ranking across industries matters most for empirical analysis; magnitude is second order.

### IV. EMPIRICAL ANALYSIS
#### A. Models
- Objective: study corporate saving behavior in ASEAN5 through EFD; test whether ASEAN5 firms are financially constrained.
- Identification strategy (Rajan and Zingales (1998) framework):
  - Test hypothesis: industries more dependent on external financing grow faster in countries with more developed financial systems.
  - Financial development indicators, m_k, drawn from the Global Financial Development Database (measure depth, efficiency, stability, competitiveness).
  - Financially constrained defined as inability to borrow adequate resources or borrowing at higher cost than risk-adjusted frictionless rate.
- Dynamic panel regression for firm growth (equation (1)):
  - g_ijkt = α + α_t + α_i + ρ1 g_ijkt−1 + ρ2 g_ijkt−2 + X_it′β + Y_jt′γ + (F_j · m_kt′) δ1 + ε_ijkt
  - g_ijkt: growth rate of firm i in industry j in country k at time t (firm growth rate defined as percent change in firm’s total asset per annum).
  - X_it and Y_jt: firm-level and industry-level characteristics.
  - F_j · m_kt: interaction between industry j’s desired EFD and country financial development.
  - Estimation method: Arellano and Bond (1991) GMM for dynamic panels with lagged dependent variables.
- Dynamic panel regression for net saving (equation (2)):
  - s_ijkt = α + ρ1 s_ijkt−1 + ρ2 s_ijkt−2 + X_it′β + Y_jt′γ + P_kt′δ1 + (F_j · P_kt′) δ2 + ε_ijkt
  - s_ijkt: net saving rate of company i in industry j in country k at time t.
  - P_kt: macro-financial and structural factors in country k (includes KA openness, ER flexibility, banking sector efficiency and competitiveness, political stability).
  - F_j · P_kt′: interaction of industry j’s desired EFD with macro-financial and structural factors.
  - Industry vector Y_jt includes industry average profit as share of total asset and its square.
  - Firm vector X_it includes ln(total assets), quartile of firm age since incorporation, and Tobin’s Q (market capitalization + total debt)/total assets.
  - Policy variables: KA openness (Chinn-Ito Index updated to 2016), ER flexibility measured by depreciation of local currency in percent in the previous period, banking sector indicators from Global Financial Development Database, political stability from ICRG.
- Notes:
  - F_j cannot be added directly due to multicollinearity with firm fixed effects.
  - WorldScope uses consolidated account data; firm-level intra-group financing variable not included.

#### B. Results
- Are ASEAN5 corporates financially constrained?
  - Estimation of equation (1) reported in Table A2 (Appendix II); financial development proxies include domestic credit-to-GDP ratio, total capitalization (percent of GDP), bank interest rate spread, bank NPL rate, bank concentration.
  - Columns (1)-(5) baseline: only one proxy included at a time.
  - Except for total capitalization, coefficients of interaction term are uniformly significant at the 5-percent level and signs align with hypothesis that financial development facilitates growth of EFD industries.
  - Interpretation example:
    - Industries at 25th and 75th percentiles of dependence among manufacturing industries in ASEAN5: Rubber and Plastic Products (25th) and Oil and Gas Extraction (75th).
    - Column (1): a 1 percentage point increase in the credit-to-GDP ratio associated with Oil and Gas Extraction growing 0.3 ppts per year faster than Rubber and Plastic Products.
    - Column (3): a 1 percentage point reduction in interest rate spread associated with Oil and Gas Extraction growing 1.3 percentage points per year faster than Rubber and Plastic Products.
  - Conclusion: corporates in ASEAN5 may be facing financial constraints in sample period 2000-2017.
- What drives net corporate saving rate in ASEAN5? (results from Table A3-A8)
  - External financing dependence
    - F_j interacted with Tobin’s Q: positive relationship between net saving rate and EFD (interacted with Tobin’s Q) — in financially constrained environments, firms with higher EFD save more, especially when Tobin’s Q is higher.
  - Tobin’s Q
    - Direct negative relationship between net saving and Tobin’s Q (without interaction) — implying firms with more future investment opportunities do not save more in the current period.
    - After interacting Tobin’s Q with EFD, the interaction coefficient is positive: promising firms in industries with larger EFD have incentive to save more.
  - Firm size and age
    - Positive relationship between net saving rate and firm size (after controlling for lagged terms) consistent with industrial concentration and rising market power raising corporate profits and saving.
    - Coefficient on firm age quartile is negative after controlling for firm size; young firms are more dependent on external funds than mature firms.
  - Industry average profit
    - Positive coefficient on industry-specific average profit: firms save part of operating cash flow.
    - Negative coefficient on square of industry profit: more profitable industries invest larger share of cash flow and save less.
  - Interaction effects of macro policies and financial development (Tables A4-A6)
    - KA openness × EFD
      - Greater KA openness reduces net saving rate for firms in industries with low demand for external funds.
      - Industries with high EFD tend to save more as KA openness increases (precautionary saving in good times due to vulnerability to capital outflows in bad times).
    - KA openness × export-orientation
      - KA openness disproportionately reduces precautionary saving of export-oriented industries (negative coefficients on KA openness and interaction term).
      - If capital controls tighten, exporters would respond by saving more relative to domestic-oriented industries.
    - KA openness × EFD × export orientation (quantile regressions, Table A5)
      - For industries with export orientation rate below 60 percent, greater KA openness does not negatively impact saving rates.
      - For highly export-oriented industries (export orientation rate over 60 percent), firms more dependent on external funds tend to save more as KA becomes more open.
    - ER flexibility × EFD (Table A6)
      - For industries with lower dependence on external funds, ER depreciation reduces net saving rate.
      - For industries with EFD breaching a threshold, depreciation increases net saving rate — competitiveness channel eventually offset by negative balance sheet effects as EFD increases.
      - Export-oriented firms reduce saving more than non-exporters in response to exchange rate depreciation due to natural hedging.
    - Banking sector efficiency and competitiveness (Table A7)
      - Increased banking sector competition or improved banks’ lending efficiency lowers corporate saving rate on average.
      - Banking sector competitiveness measured by spread between lending and deposit rates, H-statistics, Lerner index, Boone indicator, and assets of top three banks as share of total banking assets.
    - Policy uncertainty (Table A8)
      - Political instability index from ICRG used as proxy for economic policy uncertainty.
      - Improvement in political stability reduces corporate saving, consistent with precautionary motive.
      - Effect of political stability does not discriminate against firms with high EFD.

### V. CONCLUSION
- Context and motivation
  - Global trend: non-financial corporate sector shifted from net borrower to net saver over past two decades.
  - Literature on corporate saving in emerging markets is sparse; ASEAN5 is heavily integrated in GVCs and attracts substantial international capital flows.
  - Corporate saving rates in ASEAN5 show cyclical fluctuations, not the upward trend seen in advanced economies — partly due to unique ASEAN5 features.
- Main findings
  - Evidence that non-financial corporations in ASEAN5 may be subject to binding financial constraints:
    - Industries with high EFD grow relatively faster in countries with more developed financial systems (conditional on period).
    - Within a country, industries more dependent on external financing tend to save more, ceteris paribus.
  - KA openness and ER flexibility have heterogeneous effects:
    - Greater KA openness or ER depreciation reduces average saving rate of low-dependence industries but increases saving rate of high-dependence industries.
    - KA opening-up or ER depreciation disproportionately reduces saving rate of export-oriented industries, except for exporters that are highly dependent on external funds.
  - Banking sector competition or improved banks’ lending efficiency reduces corporate saving on average.
  - Greater political stability (lower policy uncertainty) is associated with lower net saving rates for all firms.
- Open question and suggested way forward
  - Whether actual corporate saving rate exceeds desired saving rate and how that gap contributes to overall external imbalance remains unanswered.
  - Conceptual and empirical challenge: define and estimate desired corporate saving.
  - Suggested considerations for desired saving rate estimation:
    - Firm’s desired saving should be a function of its EFD, determined by industry production technology.
    - Firm maturity matters: young firms are more dependent on external funds than mature ones.
    - Firm-level characteristics, such as export-orientation rate and firm ownership, may be relevant.
    - Desirable levels of macro and financial factors/policies (financial development, KA openness, ER flexibility) need definition.
  - Further research needed to deepen understanding around external imbalances and the self-financing behavior of credit-constrained firms in rapidly growing emerging markets.

*Italic: Source — wpiea2020223-print-pdf - conclusions are robust if the 10-year trend is replaced with a 5-year trend.*

### REFERENCES

### REFERENCES

### Bibliographic references
- Almeida, H., Campello, M., & Weisbach, M. S. (2004). The cash flow sensitivity of cash. The Journal of Finance, 59(4), 1777-1804.
- Amiti, M., Itskhoki, O., & Konings, J. (2014). Importers, exporters, and exchange rate disconnect. American Economic Review, 104(7), 1942-78.
- Aoyagi, C., & Ganelli, G. (2014). Unstash the cash!: corporate governance reform in Japan (No. 14-140). International Monetary Fund.
- Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The review of economic studies, 58(2), 277-297.
- Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009). Why do US firms hold so much more cash than they used to?. The journal of finance, 64(5), 1985-2021.
- Bacchetta, P., & Benhima, K. (2015). The demand for liquid assets, corporate saving, and international capital flows. Journal of the European Economic Association, 13(6), 1101-1135.
- Bayoumi, T., Tong, H., & Wei, S. J. (2012). The Chinese corporate savings puzzle: a firm-level cross-country perspective. In Capitalizing China (pp. 283-308). University of Chicago Press.
- Begenau, J., & Palazzo, B. (2017). Firm selection and corporate cash holdings (No. w23249). National Bureau of Economic Research.
- Booth, L. D., & Zhou, J. (2013). Increase in cash holdings: Pervasive or sector-specific?. Frontiers in Finance and Economics, 10(2), 31-62.
- Chen, P., Karabarbounis, L., & Neiman, B. (2017). The global rise of corporate saving. Journal of Monetary Economics, 89, 1-19.
- Dao, M.C., & Maggi, C. (2018). The rise in corporate saving and cash holding in advanced economies: aggregate and firm level trends (No. 18-262). International Monetary Fund.
- DeAngelo, H., DeAngelo, L., & Skinner, D. J. (2004). Are dividends disappearing? Dividend concentration and the consolidation of earnings. Journal of financial economics, 72(3), 425-456.
- Fama, E. F., & French, K. R. (2001). Disappearing dividends: changing firm characteristics or lower propensity to pay?. Journal of Financial economics, 60(1), 3-43.
- Fan, J., & Kalemli-Özcan, Ş. (2016). Emergence of Asia: Reforms, corporate savings, and global imbalances. IMF Economic Review, 64(2), 239-267.
- Casas, C., Díez, F. J., Gopinath, G., & Gourinchas, P. O. (2016). Dominant currency paradigm (No. w22943). National Bureau of Economic Research.
- Han, S., & Qiu, J. (2007). Corporate precautionary cash holdings. Journal of Corporate Finance, 13(1), 43-57.
- Horioka, C. Y., & Terada-Hagiwara, A. (2013). Corporate cash holding in Asia (No. w19688). National Bureau of Economic Research.
- International Monetary Fund. (2019a). World Economic Outlook, April 2019: Chapter 2: The Rise of Corporate Market Power and Its Macroeconomic Effects. International Monetary Fund.
- International Monetary Fund. (2019b). External Sector Report, July 2019: Chapter 2: Exchange Rates and External Adjustment. International Monetary Fund.
- Jiang, B., & Sedik, T. S. (2019). The Turning Tide: How Vulnerable are Asian Corporates?. International Monetary Fund.
- Kuijs, L. (2006). How will China's saving-investment balance evolve?. The World Bank.
- Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of financial economics, 52(1), 3-46.
- Rajan, R. G., & Zingales, L. (1996). Financial dependence and growth (No. w5758). National bureau of economic research.
- Sher, G. (2014). Cashing in for growth: Corporate cash holdings as an opportunity for investment in Japan (No. 14-221). International Monetary Fund.

### Sources cited for figures and tables
- WorldScope and IMF staff calculation.
- IMF, Information Notice System (noted for Exchange Rates figure).

### APPENDIX I. FIGURES

### Figures and themes (listed)
- Figure A1. Corporate Saving in ASEAN5, 2000-2017. Sources: WorldScope and IMF staff calculation.
- Figure A2.1. U.S.: Net Saving by Industry (as share of total asset). Sources: WorldScope and IMF staff calculation.
- Figure A2.2. ASEAN5: Net Saving by Industry (as share of total asset). Sources: WorldScope and IMF staff calculation.
- Figure A3.1. Export-Orientation of Manufacturing Industries in ASEAN5, 2006-2016 (unweighted average). Sources: WorldScope and IMF staff calculation.
- Figure A3.2. Export-Orientation of the Manufacturing Sector by Country, 2006-2016. Sources: WorldScope and IMF staff calculation.
- Figure A4.1. Capital Account Openness, 2000-2017 (normalized Chinn-Ito Index by country group). Sources: WorldScope and IMF staff calculation.
- Figure A4.2. Capital Account Openness in ASEAN5, 2000-2017 (normalized Chinn-Ito Index). Sources: WorldScope and IMF staff calculation.
- Figure A5. Exchange Rate Movements in ASEAN5, 2000-2018.
  - Countries shown: Malaysia, Thailand, Indonesia, Philippines, Singapore.
  - Source: IMF, Information Notice System.
  - Note: An upward movement denotes a domestic currency appreciation.
  - Axis label: Exchange Rates (US Dollar per National Currency, Period Average; Index, 2000=100).
- Figure A6.1. Credit-to-GDP Ratio, 2000-2017 (in percent, unweighted average by country group). Sources: WorldScope and IMF staff calculation.
- Figure A6.2. Credit-to-GDP Ratio in ASEAN5, 2000-2017 (in percent). Sources: WorldScope and IMF staff calculation.
- Figure A6.3. Total Capitalization, 2000-2017 (in percent, unweighted average by country group). Sources: WorldScope and IMF staff calculation.
- Figure A6.4. Spread between Lending and Deposit Rates, 2000-2017 (in percent, unweighted average by country group). Sources: WorldScope and IMF staff calculation.
- Figure A6.5. Total Capitalization, 2000-2017 (in percent, unweighted average by country group). Sources: WorldScope and IMF staff calculation.
- Figure A6.6. Concentration of the Largest Three Banks, 2000-2017 (in percent, unweighted average by country group). Sources: WorldScope and IMF staff calculation.
- Figure A7. Firm’s Profit and Saving Trends. Sources: WorldScope and IMF staff calculation.
- Figure A8. USA: External Financing Dependence by Age (Number of years since IPO).
  - Note: The header H and L refer to industries with high and low external financing dependence, respectively. We define an industry is of high EFD if its desired EFD is above the median EFD across all industries.
- Figure A9. Choice between NYSE and NASDAQ. Sources: WorldScope and IMF staff calculation.
- Figure A10. External Financing Dependence by Industry (Comparison between ASEAN5 and USA). Sources: WorldScope and IMF staff calculation.
- Figure A11. USA: Evolution of External Financing Dependence by Industry. Sources: WorldScope and IMF staff calculation.

### APPENDIX II. TABLES

### Table inventory and highlighted data (Table A1 detailed)
- Table A1: Data Coverage (in percent of GDP)
  - Country | No. of Firms | Total Asset | Gross Saving | Investment | Net Saving
  - IDN | 451 | 28.5% | 1.8% | 1.4% | 0.3%
  - MYS | 820 | 170.2% | 9.9% | 8.4% | 1.5%
  - PHL | 163 | 79.2% | 4.1% | 3.2% | 0.9%
  - SGP | 516 | 110.9% | 3.5% | 4.2% | -0.7%
  - THA | 580 | 83.3% | 5.9% | 4.3% | 1.6%
  - Avg. | 506 | 75.0% | 4.1% | 3.5% | 0.7%

### Other tables listed (titles only)
- Table A2: Firm growth and Financial System Development
- Table A3: Firm/industry Characteristics and Net Saving
- Table A4. Net Saving Rate: KA Openness, External Financing Dependence, and Export Orientation
- Table A5. Interaction between KA Openness and External Financing Dependence by Export Orientation
- Table A6: Net Saving Rate: ER Flexibility, External Financing Dependence, and Export Orientation
- Table A7: Banking Sector Efficiency, Competitiveness, and Net Saving
- Table A8: Political Stability and Net Saving

*Source: wpiea2020223-print-pdf - REFERENCES*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020223-print-pdf.pdf_