## _100716

## Source details

**Canonical URL:** [_100716](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2016/_100716.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2016/_100716.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2016/_100716.pdf.json)

---

### Executive summary — key findings on tax-induced leverage risks
- Taxes that favor debt finance (“debt bias”) significantly amplify macroeconomic stability risks posed by excessive private leverage.
- Household debt bias from mortgage interest deductions is a long-standing concern; other housing-related taxes have been increasingly used with purposive stability objectives.
- Corporate debt bias induced by deductibility of interest but not equity costs for the corporate income tax (CIT) remains a key stability concern.
- Corporate debt bias increases debt ratios by on average 7 percent of total assets, including for financial institutions.
- An Allowance for Corporate Equity (ACE) has proved effective in mitigating debt bias with large robust effects on corporate leverage, including for banks.
- The base-narrowing features of ACE could reduce CIT revenue by up to 12 percent; the loss is much smaller if ACE is granted only to new equity.
- No major implementation problems have arisen with real-world ACE systems.
- Stability risks from corporate debt bias are a particular concern in the financial sector and may call for sector-specific tax measures (e.g., sector-specific ACE, bank levies).

### Trends and context
- Gross corporate debt of non-financial corporations has reached 90 percent of GDP on average in advanced economies (AEs).
- China comprised about 38 percent of this ratio for EMEs in 2008–12 and 60 percent in 2013–15, on average.
- Aggregate household gross debt as a share of GDP has declined somewhat since the GFC in AEs but has rapidly expanded in EMEs.
- Private debt accumulation can hold back economic recovery in AEs and poses increasing threat to macroeconomic and financial stability in EMEs.

### Household debt bias — findings and evidence
- Mortgage interest deductibility generally provides incentives for household leverage and can amplify financial stability risks.
- Approximately two-thirds of AEs and a little over half of EMEs surveyed in IMF (2011b) allow a deduction or credit for mortgage interest against the personal income tax (PIT).
- Because imputed rents and capital gains on primary residences are rarely taxed, mortgage interest deductibility functions as a net subsidy to housing and can encourage higher mortgage debt.
- Several countries have reduced mortgage interest deductibility:
  - Ireland and Spain have eliminated mortgage interest deductibility on new loans.
  - Denmark and the Netherlands are gradually reducing them.
- Reforming mortgage interest deductions requires a cautious and gradual approach because house prices can respond rapidly with consequent risks to macroeconomic stability.

### Housing taxes and house-price volatility — empirical result
- Recurrent property taxes curb house-price volatility.
- New research (Annex 1): a 1 standard deviation increase in property taxes (equivalent to 0.48 percentage points) would reduce the standard deviation of house price growth from 7.4 to 4.4 percent.
- Annex I (United States, 2005–14) empirical findings:
  - Average effective property tax rate in the sample is 1 percent, with standard deviation of 0.5 percent.
  - Average volatility (standard deviation) of house prices is 5–6 percent.
  - A 1 standard deviation increase in property tax rates (0.48 percentage point) leads to:
    - 1.3–1.7 percent reduction in volatility (growth-rate measure).
    - 0.5–0.6 percent reduction in volatility (de-trended measure).
  - Average effect across regressions: increase in property tax rates by 0.48 percentage points → 3 percent reduction in house price volatility.
  - Selected property tax coefficients (columns I–VIII): -3.542***, -2.784***, -3.210***, -2.802***, -1.283**, -1.111, -1.347**, -1.177.
  - Observations: 459; # states: 51.

### Corporate debt bias — nature, measurement, and trends
- Two distinct distortions in typical CIT systems:
  - Debt bias: preference for debt financing over equity beyond the level otherwise chosen.
  - Debt shifting: tax planning across countries by multinationals.
- Personal income taxes (PIT) can mitigate debt bias but often do not fully offset the CIT-induced bias.
- Empirical findings on effective marginal tax rates (EMTRs):
  - For PIT-taxed investors at the top marginal tax rate, EMTRs can be about equal for debt and equity in some countries (Canada, Malta, Norway, Slovakia, Switzerland).
  - For most countries, EMTR is significantly lower for debt than for equity.
  - For PIT-exempt investors, debt bias is determined by the CIT alone and is usually larger; in several countries EMTR on debt for PIT-exempt investors is "even negative."
- Trends 2005 vs 2012:
  - For PIT-exempt investors, debt bias indicator fell in 23 out of 35 countries and increased in 4 countries.
  - For PIT-taxed investors, debt bias decreased in 19 countries and increased in 14.
  - Overall change small; for PIT-exempt investors METR differences still exceeded 30 points in 2012.
- OECD 2012–2016 context:
  - Mean CIT rate in OECD declined from 25.3 to 24.7 percent between 2012 and 2016.
  - Top statutory PIT rates on interest income rose on average by 1.7 percentage point for bank interest and 1.4 percentage point for bond interest; dividends top rate rose by 1.1 percentage point.

### Corporate debt bias — empirical magnitude and sectoral heterogeneity
- Meta-studies consensus: impact of CIT rate on external debt-asset ratio = 0.28.
- Example implication: CIT rate of 25 percent might be responsible for leverage ratios around 7 percentage-points higher compared to a system neutral between debt and equity.
- Tax systems increase corporate debt ratios by between 4½ and 9 percent of total assets, including in financial institutions.
- Empirical bank findings (Annex II):
  - In the long-run, a 10 percentage point higher CIT rate increases the leverage ratio (debt over assets) of an average bank by between 1.9 and 3.6 percent.
  - Eliminating the debt bias from a 25 percent CIT might reduce an average bank’s leverage ratio by 4½ to 9 percent of its assets.
- Heterogeneity:
  - Large banks less sensitive to tax but more systemically important.
  - Non-bank finance companies show tax responses similar to banks pre-crisis.
  - Investment banks: significant tax effects only before the GFC.
  - Insurance companies: life insurers each 10 percentage points of CIT increases leverage by 2.8–4.8 percentage points in the long run; non-life insurers 1.4–2.4 percentage points.

### Thin Capitalization Rules (TCRs) and earnings-stripping rules — findings
- TCR designs vary: fixed debt-to-equity caps or earnings-stripping (interest-to-EBITDA) rules; distinction whether rules apply to all debt or only related-party debt.
- BEPS Action 4 guidance: earnings-stripping rule denying interest deductibility if net total interest expense exceeds chosen percentage of EBITDA in range between 10 and 30 percent; option for group ratio rule.
- Empirical findings (Annex III / Annex IV):
  - TCRs that apply only to related-party debt have no discernable impact on consolidated (external) debt ratios.
  - Comprehensive TCRs that apply to all debt are associated with reduction in consolidated debt ratio of around 5 percentage points (statistically significant and robust).
  - thinCap–all = -1.333***; thinCap–total = -5.129***; thinCap–related-party = -0.590*** (from Annex table reporting).
  - TCR that applies to all debt reduces probability of high bankruptcy risk by 5 percentage points; related-party-only TCR has no significant effect.
- Drawbacks:
  - Earnings-stripping rules volatile as interest-to-earnings ratios vary with earnings; a 30 percent EBITDA rule would bind for 19 percent of firms on average, with less than one-in-five of those actually thinly capitalized (debt-equity ratio > 2). 9 percent of non-restricted firms have debt-equity ratio above 2:1.

### Comprehensive Business Income Tax (CBIT) — features and challenges
- CBIT denies all net interest deductions for all corporations and exempts interest received by corporations.
- CBIT would broaden CIT base and could allow statutory CIT rate reductions as part of revenue-neutral reform.
- Drawbacks:
  - Raises cost of capital and could hurt investment unless accompanied by measures reducing cost of capital (e.g., immediate deduction for all investment).
  - Unilateral CBIT could induce international distortions; new international agreements required.
  - No country has adopted full CBIT, R-base, or R+F-base cash-flow tax to date.

### Allowance for Corporate Equity (ACE) — design, evidence, and revenue effects
- ACE supplements interest deductibility with a notional deduction for normal return on equity; neutralizes debt-bias and renders CIT neutral for projects returning cost of capital.
- Empirical evidence and experience:
  - Belgium ACE (2006) reduced average consolidated debt ratio in Belgium by almost 20 percentage points for non-financial firms (relative to counterfactual) and by more than 10 percentage points for banks.
  - Belgium, Cyprus, Italy, and Turkey apply ACE variants without major implementation difficulties; Denmark and Switzerland plan ACE.
- Revenue simulations (Annex VII basis):
  - ACE granted to all equity at ongoing 10-year government bond rate reduces sectoral CIT base in 2012 and 2013 by between 5 and 12 percent, depending on sector (reduction least in extractive industries).
  - Incremental ACE (allowance only for equity increases relative to base year): first-year impact between 0.9 and 1.3 percent of pre-tax profits.
  - Financial-sector-targeted ACE:
    - CIT base narrows by between 6 and 12 percent on average across countries when ACE applies to equity in excess of minimum regulatory capital requirement.
    - If financial sector accounts for 17.5 percent of CIT receipts, total CIT revenue would fall by between 1 and 2 percent.
- Design mitigations:
  - Limit ACE to new equity; combine ACE with limitations to interest deductibility; apply ACE selectively to sectors.

### ACE tax planning risks and countermeasures
- ACE can create international tax planning opportunities (cascading allowances, asymmetric adoption across countries).
- Anti-cascading rule: subtract equity represented by participations from ACE base.
- Countermeasures: anti-avoidance rules, thin capitalization rules, denial of interest deductions for debt used to finance participations in foreign companies.

### Financial sector considerations, regulation, and bank levies
- Debt bias significant in regular and shadow banks; non-bank financial institutions can be large (in the United States non-bank sector larger than banking sector).
- Regulatory tension: Basel III increases equity buffers; tax incentives encourage debt—coordination needed.
- Basel III key ratios (in percent of RWA): Common Equity Ratio 4.5; Capital Conservation Buffer 2.5; Tier 1 Capital 6; Total Capital 8; Total Capital + Conservation Buffer 10.5; systemic risk buffer country-dependent.
- Reported shortfall relative to crisis-era loss absorption: total loss absorption capacity should have been between 15–23 percent of RWA to fully absorb past banking crisis losses in AEs.
- Bank levies (selected European examples and rates):
  - Belgium 2012: base total liabilities net of equity and insured deposits; rate 0.035 percent; transformed into New Bank Tax with 2016 rate 0.13231 percent.
  - Cyprus 2011: base total bank deposits; initial rate 0.095 percent; increased to 0.15 percent in 2015; reduced to 0.0375 percent in 2016 as phase-out began.
  - United Kingdom 2011: base total liabilities net of equity and insured deposits with exemption up to £20 billion; initial rates 0.088 percent (short-term) and 0.044 percent (long-term); increased to 0.21 and 0.105 percent in 2015; reductions announced to 0.10 percent and 0.05 percent during 2015–21; new tax on banking sector profit introduced in 2016 at 8 percent.
- Evidence on bank levies:
  - Average bank levy induced increase in regulatory capital ratio by around 0.9 percentage points (Devereux, Johannesen, and Vella 2013).
  - Some reduction in funding risk offset by increases in portfolio risk.

### Empirical methods, samples, and firm-level findings (Annex II and Annex Tables)
- Banking sample (Bankscope): about 151,463 observations; mean leverage 86.5 (pct total assets); CIT rate mean 31.2 (pct); sample covers 131 countries.
- Insurance sample (Orbis Insurance Focus): 19,469 observations; mean insurance leverage 11.5 (pct total assets); CIT rate mean 28.3 (pct).
- Firm-level dynamic panel estimation: Blundell and Bond (1998) system GMM used; long-term effect computed as β2 /(1−β1).
- Regular banks regression highlights:
  - Lagged leverage coefficients (system GMM): 0.34*** (full period), 0.36*** (before 2008), 0.69*** (2008 onwards).
  - CIT rate (system GMM): 0.11** (full); 0.24*** (before 2008); 0.05 (2008 onwards).
  - Long-term effects reported: overall 0.23***; full period GMM 0.17**; before 2008 0.38***; 2008 onwards 0.17.
- Extended sample (regular + investment + finance companies):
  - CIT rate (system GMM basic): 0.12** (full); 0.23*** (before 2008); 0.07 (2008 onwards).
  - Outlier treatment (winsorized 10 percent) yields CIT coefficients significant and positive post-crisis as well.
- Insurance subsamples:
  - Life insurers long-run coefficient ~0.28 (full period); short-term effect 0.08 (full), 0.26 (pre-crisis), 0.09 (post-crisis).
  - Non-life insurers long-run effects 0.14–0.24; short-term ~0.05 (full), 0.16 (pre-crisis), 0.06 (post-crisis).

### Non-financial sector externalities and macro stability risks
- Debt bias causes welfare losses from distorted investment and risk-taking; pre-GFC estimated deadweight loss small relative to tax revenue (examples: 2–3 percent of corporate tax revenue in Norway).
- Since GFC focus shifted to externalities: excess private sector debt is a systemic credit externality.
- Empirical findings:
  - Giroud and Mueller (forthcoming): employment declines during GFC more pronounced in highly-leveraged firms.
  - Sutherland and Hoeller (2012): higher corporate sector leverage associated with higher probability of recession.
  - Jordà and others (2013): credit buildup during expansions tends to make subsequent recessions more likely, deeper and longer lasting.

### Policy implications and recommendations
- Tax-induced debt bias should feature prominently in national tax reform agendas.
- Effective approaches include:
  - Broad restrictions on interest deductibility (targeting all debt) when feasible.
  - Introduction of an ACE, with consideration of limiting ACE to new equity to reduce immediate revenue loss.
  - Sector-specific tax instruments (ACE variants, bank levies) to address amplified financial-sector stability risks.
- Reforms to mortgage interest deductibility and housing taxation should be implemented cautiously and gradually to avoid rapid adverse house-price responses.
- Taxes could complement regulatory efforts to induce higher equity in banks, especially where regulators have limited information or tighter regulation drives intermediation toward shadow banks.
- Consideration of leveling the playing field with shadow banks via taxes or extending bank levies with rate differentiation.

### Annex simulation highlights — ACE base-narrowing (Annex VII)
- Non-financial firms (2012–13 ORBIS sample):
  - ACE granted on book value of equity reduces CIT base by between 5 and 12 percent across sectors (extractive, manufacturing, services).
  - Examples (2012):
    - Manufacturing: ACE used (pct pre-tax profit) = 12.0.
    - Services: ACE used (pct pre-tax profit) = 12.0.
- Financial firms (2012 Bankscope and Orbis):
  - Regular banks (2012): ACE used (pct pre-tax profit) = 14.3; ACE2 (allowance only for equity in excess of 3 percent assets) used (pct pre-tax profit) = 9.6.
  - Insurance companies (2012): ACE used (pct pre-tax profit) = 14.0; ACE2 used (pct pre-tax profit) = 10.8.
- Caveats: calculations are static, cap ACE by pre-tax profit (no refunds), do not model carryforwards, and represent an upper bound due to accounting vs tax data differences.

*Source: EXECUTIVE SUMMARY, TAX POLICY, LEVERAGE AND MACROECONOMIC STABILITY (October 7, 2016) and chapter and annex material from the provided IMF content unit.*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Key findings on tax-induced leverage risks
- Taxes that favor debt finance (“debt bias”) significantly amplify macroeconomic stability risks posed by excessive private leverage.
- Household debt bias from mortgage interest deductions is a long-standing concern; other housing-related taxes have been increasingly used with purposive stability objectives.
- Corporate debt bias induced by deductibility of interest but not equity costs for the corporate income tax (CIT) remains a key stability concern.
- Corporate debt bias increases debt ratios by on average 7 percent of total assets, including for financial institutions.
- An Allowance for Corporate Equity (ACE) has proved effective in mitigating debt bias with large robust effects on corporate leverage, including for banks.
- The base-narrowing features of ACE could reduce CIT revenue by up to 12 percent; the loss is much smaller if ACE is granted only to new equity.
- No major implementation problems have arisen with real-world ACE systems.
- Stability risks from corporate debt bias are a particular concern in the financial sector and may call for sector-specific tax measures (e.g., sector-specific ACE, bank levies).

### Trends and context
- Gross corporate debt of non-financial corporations has reached 90 percent of GDP on average in advanced economies (AEs).
- China comprised about 38 percent of this ratio for EMEs in 2008–12 and 60 percent in 2013–15, on average.
- Aggregate household gross debt as a share of GDP has declined somewhat since the GFC in AEs but has rapidly expanded in EMEs.
- Private debt accumulation can hold back economic recovery in AEs and poses increasing threat to macroeconomic and financial stability in EMEs.

### Household debt bias — findings and evidence
- Mortgage interest deductibility generally provides incentives for household leverage and can amplify financial stability risks.
- Approximately two-thirds of AEs and a little over half of EMEs surveyed in IMF (2011b) allow a deduction or credit for mortgage interest against the personal income tax (PIT).
- Because imputed rents and capital gains on primary residences are rarely taxed, mortgage interest deductibility functions as a net subsidy to housing and can encourage higher mortgage debt.
- Several countries have reduced mortgage interest deductibility: Ireland and Spain have eliminated mortgage interest deductibility on new loans, while Denmark and the Netherlands are gradually reducing them.
- Reforming mortgage interest deductions requires a cautious and gradual approach because house prices can respond rapidly with consequent risks to macroeconomic stability.

### Housing taxes and house-price volatility — new empirical result
- Recurrent property taxes curb house-price volatility, reinforcing their attractiveness as a relatively fair and efficient revenue source.
- New research (Annex 1) reports a significant negative impact: a 1 standard deviation increase in property taxes (equivalent to 0.48 percentage points) would reduce the standard deviation of house price growth from 7.4 to 4.4 percent.

### Corporate debt bias — assessment of policy responses
- Rules restricting interest deductibility for the CIT have only partially addressed debt bias.
  - Where rules target all debt, they have proven effective.
  - Rules that target only borrowing between related parties (to address debt shifting within multinationals) have no discernable impact on external borrowing of corporate groups—the debt most relevant for stability.
- An Allowance for Corporate Equity (ACE) is effective in reducing corporate leverage; empirical evidence finds large robust effects including for banks.
- The estimated reduction in the CIT base from introducing an ACE could be as large as 12 percent, though smaller if ACE applies only to new equity.
- Practical experience with ACE systems has not revealed major implementation problems.

### Financial sector considerations
- Debt bias is significant for both regular and shadow banks; stability risks extend beyond traditional banks to non-bank financial institutions.
- Sector-specific remedies have been used:
  - Sector-specific ACE designs for financial institutions.
  - Bank levies have been introduced and have reduced bank leverage in Europe.
- Addressing debt bias in the financial sector may require tax measures tailored to sector-specific stability externalities.

### Policy implications and recommendations
- Tax-induced debt bias should feature prominently in countries’ tax reform plans.
- Several approaches have proven capable of reducing debt bias without causing significant implementation difficulty:
  - Broad restrictions on interest deductibility (targeting all debt) when feasible.
  - Introduction of an ACE, with consideration of limiting ACE to new equity to reduce immediate revenue loss.
  - Sector-specific tax instruments (ACE variants, bank levies) to address amplified financial-sector stability risks.
- Reforms to mortgage interest deductibility and housing taxation should be implemented cautiously and gradually to avoid rapid adverse house-price responses.

*Source: EXECUTIVE SUMMARY, TAX POLICY, LEVERAGE AND MACROECONOMIC STABILITY (October 7, 2016).*

### 9.      While some tax policies can be used to steer house price developments,

### 9.      While some tax policies can be used to steer house price developments,

### Tax measures, timing, and neutrality
- Tax measures can be powerful instruments of fiscal stimulus because "tax changes are likely to be capitalized as soon as they are announced."
- Implementation lags and gaps between announcement and implementation create risks:
  - Delays can make measures credible only after the immediate need has passed.
  - Announcement–implementation gaps can induce a "flurry of last minute transactions" or delayed transactions in anticipation of reduced taxes.
- Given difficulty distinguishing bubbles from fundamentals-driven asset price movements, tax policy should focus on neutrality in the treatment of differing assets and forms of income.
- Empirical evidence on property taxes shows mixed effects:
  - Successful cyclical interventions have been achieved using property taxes (recurrent or on transactions).
  - Mixed effects are sometimes reported on house price volatility (ECB 2003) and house price growth (Kuttner and Shim 2013).

### Box 1 — Property Taxes and House Prices: country cases and observed effects
- General caveat: "The impact of tax policy measures on house price fluctuations should be interpreted with caution, given the absence of a counterfactual and limited ability to control for other factors driving house price fluctuations."

- Hong Kong
  - House prices more than doubled between 2009 and 2013.
  - November 2010: seller stamp duty (SSD) of 15 percent for properties resold within two years.
  - October 2012: SSD raised to 20 percent and extended to properties resold within three years.
  - Introduced a 15 percent buyer’s stamp duty (BSD) on residential properties acquired by companies and non-locals (which accounted for about 20 percent of transactions).
  - February 2013: doubled rates of existing BSD for transactions of all types of properties, except for local individuals with only one residential property.
  - Empirical evidence suggests these policies "have been effective in constraining housing demand and restraining house price growth (He 2014)."

- Singapore
  - Between 2003 and 2012, house prices almost doubled; annual percent changes peaked in mid-2010.
  - Mortgages peaked at 18 percent of the banking system’s credit to the non-banking sector.
  - From February 2010: series of tax increases, including an SSD on all residential properties sold within one year of purchase at progressive rates between 1 and 3 percent (later extended to sales within four years and rates raised to 16 percent for sales within a year).
  - December 2011: additional BSD imposed at rates of 10 percent on nonresidents and corporate entities and of 3 percent on second subsequent residential property purchase by permanent residents and third or subsequent residential property purchase by Singapore citizens.
  - Subsequent BSD rates increased to:
    - 15 percent for nonresidents and corporate entities;
    - 5 percent (10 percent) for first (second or subsequent) residential property purchase by permanent residents;
    - 7 percent (10 percent) for second (third and subsequent) residential property purchase by Singapore citizens.
  - Macroprudential measures limiting total debt service ratios introduced in mid-2013.
  - As of the second quarter of 2016:
    - Residential property prices in private and public (resale) housing markets were lower by 9½–10 percent from their peaks in 2013.
    - These peaks followed a rise of about 50–60 percent since the troughs in 2008/09 in the public (resale) and private housing market.
  - Transaction volumes:
    - Private property transactions declined by one-third in 2013 and by 50 percent in 2014.
    - Transaction volumes rose in 2015, but housing loan growth continued to moderate to "low single digit."

- China
  - Fiscal stimulus in 2008 fueled a domestic credit boom starting in 2009 and 2010, contributing to real estate overheating and house price growth of around 40 percent.
  - 2010 measures included:
    - Taxes increased on resale of properties within five years of purchase.
    - Exemptions of home purchases from stamp duties and of home sales from income taxes were abolished, except for cases involving a family’s only home.
  - After measures, home sales rose by only 6 percent in early 2011 compared to 30 percent in the same period of 2010.
  - Sales declined sharply in major cities that had seen the largest price increases.

- Sweden
  - Late 1980s boom reflected credit deregulation, strong population growth, and generous property tax allowances/subsidies.
  - 1991 tax reform: portion of mortgage interest payments deductible reduced to 30 percent.
  - Other tax reforms raised holding cost for housing by imposing full VAT on related activities.
  - House prices plummeted by 17 percent in real terms to mid-1980 levels, also affected by the early 1990s banking crisis.
  - During 1995–2015 house prices and household indebtedness rose again; modest decline during the global financial crisis.
  - 2008: virtual abolition of the property tax in favor of a municipal tax set at the lower of SEK 6,825 (around €969) or 0.75 percent of the property's assessed value—may have contributed to continued price rises.
  - Policy responses since 2010 mainly macroprudential.

- Ireland
  - Late 1990s housing boom prompted stamp duty adjustments favoring first-time buyers and owner-occupiers.
  - Transaction taxes were hiked (progressive stamp duty on residential property) and an anti-speculative property tax on non-principal private residences was introduced for investors.
  - These measures led to a slowdown in house prices as investors exited.
  - 2002: many measures reversed in response to rental accommodation shortages; investors returned and house prices rose until the GFC, which led to a drop in house prices (more than 25 percent in real terms).
  - Authorities eliminated mortgage interest deductibility on new loans taken out after December 31, 2012.

### Corporate debt bias: nature, measurement, and trends
- Two distinct distortions from typical CIT systems that allow interest but not equity returns to be deducted:
  - Debt bias: preference for debt financing over equity beyond the level otherwise chosen.
  - Debt shifting: tax planning across countries by multinationals (lending from low-tax to high-tax affiliates or locating external borrowing in high-tax countries).
- Main macroeconomic stability concern is debt bias (external borrowing associated with debt bias can be a source of macroeconomic instability).
- Personal income taxes (PIT) can mitigate debt bias but often do not fully offset the CIT-induced bias:
  - Returns on debt and equity may be taxed through PIT and/or withholding taxes, affecting overall tax advantage.
  - Integration of PIT and CIT (imputation systems, e.g., Australia, Canada, Chile) can reduce double taxation of dividends and disadvantage of equity, but often insufficient.
- Empirical findings on effective marginal tax rates (EMTRs):
  - For selected AEs and European EMEs, for PIT-taxed investors at the top marginal tax rate, EMTRs can be about equal for debt and equity in some countries (Canada, Malta, Norway, Slovakia, Switzerland).
  - For most countries, EMTR is significantly lower for debt than for equity (i.e., debt favored).
  - For PIT-exempt investors (e.g., institutional investors), debt bias is determined by the CIT alone and is usually larger.
  - In several countries, EMTR on debt for PIT-exempt investors is "even negative," implying marginal debt-financed investment is effectively subsidized.
- Empirical evidence on PIT effects on corporate leverage is mixed; some studies find higher PIT rates on interest relative to dividends and capital gains reduce leverage ratios of domestic firms (Overesch and Voeller 2010; Lin and Flannery 2013).
- Trends in debt bias (2005 vs 2012):
  - For PIT-exempt investors, debt bias indicator fell in 23 out of 35 countries and increased in 4 countries.
  - For PIT-taxed investors, debt bias decreased in 19 countries and increased in 14.
  - Overall change in debt bias has been small; for PIT-exempt investors the difference in METRs still exceeded 30 points in 2012.
- Further context (2012–2016):
  - Mean CIT rate in OECD countries declined from 25.3 to 24.7 percent between 2012 and 2016—tending to reduce debt bias.
  - During the same period, top statutory PIT rates on interest income rose on average by 1.7 percentage point for bank interest and 1.4 percentage point for bond interest, compared to an increase of 1.1 percentage point for dividends—slightly reducing debt bias for investors subject to top marginal tax rates since 2012.
- No compelling economic reason to favor debt over equity for tax purposes; both are returns to capital. Legal and administrative difficulties also arise from distinguishing debt and equity (e.g., hybrid instruments).

### Debt bias in the financial sector and policy coordination
- Excessive leverage of financial institutions (including shadow banks) is a major macroeconomic stability concern due to higher failure probability and contagion/externalities.
- Non-bank financial institutions ("shadow banks") can perform bank-like roles (maturity transformation, risk sharing, liquidity provision). In some countries the non-bank financial sector (measured by total assets) is large; in the United States it is even larger than the banking sector.
- Fundamental tension: regulatory efforts require financial institutions to hold more capital, while tax incentives (debt bias) encourage less capital (more debt).
  - Financial firms typically hold equity buffers beyond regulatory minima, so taxes can affect leverage.
  - Tax and regulatory policies are often handled by different ministries/agencies, implying a need for policy coordination.

*Source: _100716 - 9.      While some tax policies can be used to steer house price developments,*

### 18.      The impact of debt bias on financial stability depends on the initial capitalization of

### 18.      The impact of debt bias on financial stability depends on the initial capitalization of

### Impact of debt bias on financial stability
- A marginal increase in a bank’s leverage ratio induced by debt bias has a larger effect on its default risk, and thus the likelihood of crisis, at high levels of debt (i.e., low capitalization).
- De Mooij, Keen, and Orihara (2014) find that eliminating debt bias could have reduced the probability of a financial crisis in some countries by up to 40 percent (left hand panel of Figure 6).
- The associated expected output gain from eliminating debt bias ranges between 0.6 and 13.3 percent of GDP over a four-year period.
- Langedijk and others (2015) find that elimination of debt bias could have reduced expected direct bailout costs by between 17 and 77 percent for European economies.
- At lower bank leverage ratios prevailing in 2016, the marginal impact of debt bias on the probability of a financial crisis is significantly reduced, but remains non-negligible (Figure 6, right hand panel).

### Taxation and debt spillovers to the financial sector (Box 2)
- Risks from excessive debt in non-financial corporates may spill over to the financial sector via higher probabilities of corporate distress or failure increasing non-performing loans (NPLs) on financial institutions’ balance sheets.
- The importance of this spillover depends on:
  - The tax treatment of NPLs:
    - Provisions against, and write-offs of, impaired debts are generally income tax deductible, with claw back if a loan is subsequently repaid.
    - Governments often limit provisions and write-offs that qualify for tax deductibility to prevent tax avoidance.
    - If conditions for deductibility are too strict, banks may retain large portfolios of NPLs that inhibit new lending and amplify financial stability risks.
    - If conditions are too loose, provisioning for NPLs may open loopholes in tax collection.
    - IFRS 9 aims to align criteria for provisions and write-offs internationally, based on an expected credit loss model; tax rules generally follow accounting rules.
  - The treatment of deferred tax assets (DTAs):
    - Large provisions or write-offs create DTAs appearing on banks’ balance sheets as reductions in future tax liabilities; DTAs have no loss-absorption capacity when an institution makes losses.
    - Since the GFC, regulators generally no longer qualify DTAs as bank capital.
    - Some governments have transformed DTAs into “deferred tax credits” (DTCs), claims on the government independent of future profitability; DTCs qualify as regulatory capital but create contingent fiscal liabilities and reinforce bank-sovereign linkages.

### Regulatory context and capitalization (Box 3 and related text)
- New Basel III standards increase the level and quality of required equity buffers; provisions to be fully phased in by 2019.
- Main Basel III instruments:
  - Minimum capital adequacy ratio (CAR) relative to risk-weighted assets (RWA).
  - Minimum leverage ratio specifying capital to total assets of at least 3 percent.
  - Liquidity requirements: liquidity coverage ratio (LCR) and net stable funding ratio (NSFR).
  - Country authorities can require additional buffers (e.g., systemic risk buffer of 1–3.5 percent of RWA; countercyclical buffer of 0–2.5 percent).
- Reported shortfall relative to crisis-era loss absorption:
  - Dagher and others (2016) report that total loss absorption capacity should have been between 15–23 percent of RWA to fully absorb past banking crisis losses in AEs, which is well above the new minimum standards.
- Shadow banks:
  - Generally less tightly regulated, may operate with higher leverage, implying higher marginal impact of debt bias on default probability.
  - Distributional evidence (Figure 7): while most finance companies are less leveraged than banks, a relatively large fraction have very high leverage ratios between 95 and 100 percent.
  - Increased interconnectedness of banks and shadow banks implies debt bias in non-bank financial institutions can contribute significantly to systemic risk.

- Basel III capital ratios (in percent of risk-weighted assets) as presented:
  - Common Equity Ratio 4.5
  - Capital Conservation Buffer 2.5
  - Tier 1 Capital 6
  - Total Capital 8
  - Total Capital + Conservation Buffer 10.5
  - Systemic Risk Buffer (for SIFIs) country-dependent
- Note: total capital + conservation buffer needs to be at least 10.5 percent of RWA; capital needs to be above 8 percent of RWA, of which at least 6 points is Tier 1 capital, of which at least 4.5 points is common equity.

### Empirical evidence on tax effects on financial institutions’ leverage (Annex 2 summary and related findings)
- Corporate taxes raise bank leverage ratios significantly:
  - In the long-run, a 10 percentage point higher CIT rate increases the leverage ratio (debt over assets) of an average bank by between 1.9 and 3.6 percent.
  - Eliminating the debt bias from a 25 percent CIT rate might thus reduce an average bank’s leverage ratio by 4½ to 9 percent of its assets.
- Time-varying responses:
  - The response of financial institutions’ leverage to tax was strong before the GFC—and remains positive after the GFC.
  - Responses after the crisis are somewhat smaller than before it, and only significant in a sample excluding outliers.
    - If the bottom and top 10 percent of observations in the leverage distribution are removed, each 10 percentage point of CIT raises bank leverage by 1.5 percentage points in the long run (post-crisis).
    - If the top and bottom 1 percent are removed, the coefficient for the tax rate is positive but no longer statistically significant in the post-GFC period.
  - The smaller post-crisis effects may be transitory due to rebuilding buffers and implementation of Basel III requirements that forced many banks to increase capital irrespective of tax regimes.
  - Evidence suggests the responsiveness of bank leverage to tax increases with the size of the buffer vis-à-vis the minimum capital requirement (De Mooij and Keen 2016, Bond and others 2016).
- Heterogeneity across institution types:
  - Large banks are noticeably less sensitive to tax, but still pose the greatest concern for financial stability because even small changes in leverage of very large banks can materially affect distress or failure probabilities and systemic risk.
  - Debt bias is evident in non-bank finance companies: in the pre-crisis period, tax responses of non-bank finance companies (consumer finance, factoring and leasing, trade finance and credit card companies) are not significantly different from those of banks.
  - For investment banks, significant tax effects are found only before the GFC; post-crisis the impact of taxes on their debt is no longer significant.
  - Insurance companies:
    - Leverage largely reflects business mix; technical reserves comprise some 65 percent of total liabilities on average, with equity comprising another 23 percent on average for life insurers.
    - For the remaining liabilities, debt bias is prevalent and more pronounced than in banks: in life insurance companies, each 10 percentage points of CIT increases the leverage ratio by between 2.8 and 4.8 percentage points in the long run.
    - For non-life insurers the effect is 1.4–2.4 percentage points.
    - Because insurers are generally less systemically important than banks, the financial stability impact of debt bias in insurance companies is more limited.

### Non-financial sector: externalities and macro stability risks
- Traditional welfare concern from debt bias in non-financial corporations centers on distortions in risk behavior and investment decisions, causing welfare losses (e.g., due to excessive agency costs).
  - Sørensen (2014) estimates deadweight loss from debt bias at 2-3 percent of corporate tax revenue in Norway.
  - Similar exercises for Germany and the United States suggest even smaller welfare costs (Weichenrieder and Klautke 2008, Gordon 2010).
- Since the GFC, focus has shifted to externalities associated with excess leverage and amplified macro stability risks:
  - Excess private sector debt can be seen as a systemic credit externality (Bianchi 2011).
  - Micro-level effects: high debt ratios increase probability of firm bankruptcy in adverse shocks and amplify liquidity constraints post-shock via rollover risks and debt overhang.
  - Empirical findings:
    - Giroud and Mueller (forthcoming) report that employment declines during the GFC were significantly more pronounced in highly-leveraged than in lightly-leveraged firms.
    - Sutherland and Hoeller (2012) find higher leverage ratios in the non-financial corporate sector are associated with a significantly higher probability of recession.
    - Jordà and others (2013) find that credit buildup during expansions tends to make subsequent recessions more likely, deeper and longer lasting.
- Additional note: Taxes might also affect debt maturity structures (e.g., leading to shorter debt maturity and amplifying rollover risks), but empirically there is no strong evidence for maturity effects (Graham 2008).

*Source: IMF (chapter text provided).*

### 23.      Debt bias in non-financial corporations is significant, although its extent varies

### 23.      Debt bias in non-financial corporations is significant, although its extent varies

### Extent and heterogeneity of debt bias
- Meta-studies derive a consensus value for the impact of the CIT rate on the external debt-asset ratio of 0.28.
- Example implication: a CIT rate of 25 percent might be responsible for leverage ratios that are around 7 percentage-points higher compared to a system neutral between debt and equity.
- The debt bias is larger in countries with higher CIT rates.
- Debt bias is higher in sectors characterized by high tangible assets typically put up as collateral.
- Larger firms appear to be more responsive to tax-induced debt bias, which matters for macroeconomic stability because large firms can induce bigger spillovers.

### Empirical evidence on debt bias and policy tools
- Most empirical studies use unconsolidated accounting data; new analysis exploits consolidated accounts of non-financial multinational companies.
- The empirical exercise is based on a large panel of firms in 42 countries between 2005 and 2014.
- About 60 countries now impose some form of thin capitalization rules (TCRs).

### Neutralizing corporate debt bias — overview of options
- Broad policy options:
  - Limit interest deductibility to treat debt more like equity.
  - Introduce a deduction for corporate equity returns to treat equity more like debt.
  - Combine elements of both.
- For the financial sector, there may be case for tax disincentives to debt finance because of larger negative externalities.

### A. Limiting interest deductibility — Thin Capitalization Rules (TCRs) and earnings-stripping rules
Findings
- TCRs cap interest deductibility beyond a fixed ratio of debt relative to equity or by using earnings-stripping (interest-to-earnings) rules.
- Key design distinctions:
  - Whether fixed ratio refers to debt-to-equity or to interest-to-earnings (an ‘earnings-stripping’ rule).
  - Whether the rule applies to all debt or only to related-party debt.
- New G20/OECD guidelines (BEPS Action 4) propose an earnings-stripping rule denying interest deductibility if net total interest expense exceeds a chosen percentage of EBITDA in the range between 10 and 30 percent.
- The guidelines include an option for a group ratio rule tying affiliate interest deductibility to the worldwide group net interest/EBITDA ratio.
- TCRs that apply only to related-party debt have no discernable impact on consolidated (external) debt ratios.
- Comprehensive TCRs that apply to all debt are associated with a reduction in the consolidated debt ratio of around 5 percentage points (statistically significant and robust).
- A comprehensive TCR reduces the probability of a firm facing high bankruptcy risk by 5 percentage points; a TCR that targets related-party debt has no significant effect on bankruptcy risk.

Drawbacks and trade-offs
- Earnings-stripping rules are more volatile because interest-to-earnings ratios vary with earnings; they can bind for firms with temporarily low earnings and thus amplify financial constraints during downturns.
- Illustration statistics for a 30-percents interest-EBITDA rule:
  - On average, 19 percent of firms would be restricted by a 30 percent EBITDA rule.
  - Within the restricted firms, less than one-in-five are actually thinly capitalized (debt-equity ratio greater than 2).
  - 9 percent of the non-restricted firms have a debt-equity ratio above 2:1.
- Advantage of a fixed debt-equity ratio cap: less sensitive to cyclical fluctuations than an interest-to-earnings rule and might avoid creating a new source of financial distress.
- Earnings-stripping rules are effective for mitigating debt shifting (including intracompany interest manipulation); IMF technical assistance has generally encouraged adoption of earnings-stripping rules to counter debt shifting.

### Full denial of interest deductibility — Comprehensive Business Income Tax (CBIT)
Findings and features
- CBIT denies all net interest deductions for all corporations and exempts interest received by corporations to avoid double taxation.
- CBIT is consistent with a broad source-based tax on all capital income (including interest), withheld at the level of the firm.
- CBIT would significantly broaden the CIT base and could allow statutory CIT rate reductions as part of revenue-neutral reform.

Drawbacks and implementation challenges
- CBIT will raise the cost of capital and thus hurt investment unless accompanied by measures reducing the cost of capital (for example, immediate deduction for all investment, transforming CBIT into a real-base (R-base) cash-flow tax).
- A unilateral CBIT might induce significant international distortions because it relies on source taxation of interest while most countries tax interest on a residence basis; new international agreements would be required to avoid double taxation or non-taxation of cross-border interest income.
- Under CBIT, banks’ margins would be effectively untaxed at the bank level; one response is an R+F-base cash flow tax that includes financial intermediation spreads in the tax base.
- CBIT involves long and complex transition issues to cope with pre-existing debt and, if combined with full expensing, the treatment of prior investments already benefiting from depreciation allowances.
- No country has adopted a full CBIT, R-base, or R+F-base cash-flow tax to date; cash-flow taxes have appeared in limited contexts (Mexico as a minimum tax, Estonia’s S-base variant, special regimes in extractive industries and for SMEs).

Selective applications and alternatives
- Applying CBIT logic selectively to intracompany debt has rationale: for a parent fully owning a subsidiary, debt versus equity distinctions are moot and intracompany interest could be re-characterized as equity or taxed at source.
- Risks of unilateral intracompany CBIT include international double taxation or double non-taxation.
- Alternative: allow interest on intracompany debt to be deducted only at the tax rate at which the interest income is taxable in the lender’s country.
- Another option to neutralize debt shifting is formula apportionment of the group’s worldwide interest expense across affiliates.

### B. Allowing a Deduction for Corporate Equity — Allowance for Corporate Equity (ACE)
Findings and experience
- ACE supplements current interest deductibility with a notional deduction for the normal return on equity.
- ACE preserves interest deductibility and introduces a similar deduction for equity returns.
- ACE has been proposed and advocated in multiple contexts and has been adopted in variants by several countries.
- Calculation of taxable profit under ACE can use existing corporate financial accounts (dividends, new share issues, acquisitions of shares) to determine the equity base.
- ACE rate might be based on an average corporate or government bond rate.
- Belgium, Cyprus, Italy, and Turkey apply variants of ACE without major implementation difficulties; Denmark and Switzerland plan to introduce ACE.
- Most countries grant ACE not only to large corporations but also to smaller firms, including partnerships.

Final empirical and policy takeaways
- Consensus estimate: impact of CIT rate on external debt-asset ratio is 0.28.
- A CIT rate of 25 percent might be responsible for about 7 percentage-points higher leverage relative to neutrality.
- TCRs that apply to all debt reduce consolidated debt ratios by around 5 percentage points and reduce likelihood of high bankruptcy risk by 5 percentage points; TCRs limited to related-party debt do not significantly affect consolidated debt or bankruptcy risk.
- Earnings-stripping rules in the BEPS guidance allow countries to choose an EBITDA threshold in the range between 10 and 30 percent; a 30 percent EBITDA rule would bind for 19 percent of firms on average, with less than one-in-five of those constrained firms actually thinly capitalized (debt-equity ratio greater than 2), while 9 percent of non-restricted firms have a debt-equity ratio above 2:1.
- Net overall revenue effect of interest deductibility estimated in the range of 0.5 to 0.8 percent of GDP in major European countries in recent years.

*Source: IMF staff analysis in “Debt bias in non-financial corporations is significant, although its extent varies,” chapter 23 of the provided IMF PDF content.*

### Annex 5

### Annex 5

### Main features and neutrality properties of the Allowance for Corporate Equity (ACE)
- ACE allows a deduction for both interest and the normal rate of return on equity, charging no tax on projects with a return that matches the cost of capital; it is therefore a tax on economic rents.
- Neutrality properties emphasized:
  - Neutralizes debt-bias.
  - Renders the CIT neutral with respect to marginal investment decisions (effective marginal tax rate is zero for projects returning the cost of capital).
  - Removes investment distortions induced by differences between economic depreciation and depreciation for tax purposes (present value of depreciation allowance plus ACE allowance is independent of tax-account depreciation rates).
- Caveat: Evidence on investment effects is scarce and empirical outcomes are mixed.

### Empirical evidence on leverage (Belgian case and broader findings)
- Belgian ACE implementation in 2006:
  - Reduced the average consolidated debt ratio in Belgium by almost 20 percentage points for non-financial firms (relative to the counterfactual).
  - Reduced the average consolidated debt ratio by more than 10 percentage points for banks (relative to the counterfactual).
- Empirical studies cited: Princen 2012; Hebous and Ruf 2015; Schepens 2016.
- Banks sample for analysis: 18,551 observations for 49 countries between 2005–13 (includes commercial banks, cooperative banks, and investment banks).

### Tax planning, international spillovers, and countermeasures
- ACE can create international tax planning opportunities similar to existing debt-shifting schemes; these can be mitigated by countermeasures (see Annex 6).
- ACE can reduce some avoidance driven by differential tax treatment of debt and equity (e.g., excessive lending to subsidiaries in ACE countries becomes less advantageous).
- Asymmetric adoption of ACE (only some countries) can induce new forms of tax planning, such as intracompany loans from ACE countries financed by equity to take interest deductions in non-ACE countries.
- Countries without ACE and with relatively high CIT rates will find Thin Capitalization Rules (TCRs) more important if some countries adopt ACE or very low CIT rates.

### Revenue effects and mitigation options
- Simulations for a group of OECD countries (Annex 7 basis):
  - ACE granted to all equity at the ongoing 10-year government bond rate reduces the sectoral CIT base in 2012 and 2013 by between 5 and 12 percent, depending on the sector (reduction least in extractive industries).
- Mitigation options:
  - Incremental ACE (apply to equity increases relative to a base year):
    - First-year impact based on simulations lies between 0.9 and 1.3 percent of pre-tax profits.
    - Revenue effect will gradually increase over time as new equity stock expands.
  - Combine ACE with a limitation to interest deductibility (calibrate ACE rate and debt constraint to ensure revenue neutrality).
  - Implement ACE in specific sectors where economic rents are important, perhaps in addition to the ordinary CIT (examples: Norway petroleum tax scheme with allowance of 7.5 percent of investment cost for first four years; Israel mining regime with a “yield on the amortized cost of capital”).

### Revenue assessment for financial-sector–targeted ACE
- Imposing ACE only on financial institutions can neutralize debt bias where macro stability risks are largest while limiting revenue loss.
- Revenue simulation findings (Annex 7 basis):
  - CIT base narrow by between 6 and 12 percent on average across countries when ACE applies to equity in excess of the minimum regulatory capital requirement.
  - If the financial sector accounts for 17.5 percent of CIT receipts, total CIT revenue would fall by between 1 and 2 percent.
- Design note: ACE in the financial sector can be limited to equity in excess of the minimum regulatory capital requirement (minimum capital requirement example used in simulations: 3 percent of total assets).

### Taxes, regulation, and use of tax instruments to curb leverage
- Taxes could complement regulatory efforts to induce higher equity in banks, especially where regulators have limited information or tighter regulation drives intermediation toward shadow banks.
- Debate remains on the relative merits of taxation versus regulation; tax incentives can be useful to supplement regulation where regulation alone may be insufficient.

### Recent developments in bank levies (selection of European reforms)
- Bank levies introduced mainly in Europe between 2010 and 2012; several countries reformed these taxes since 2015.
- Country examples and exact figures cited:
  - Belgium:
    - 2012 Financial Stability Contribution base: total liabilities net of equity and insured deposits; rate 0.035 per cent.
    - Transformed into “New Bank Tax” with predetermined revenue target of €805m; €500m reserved for SRF. Tax base: average of reported Belgian “debt towards clients” of preceding year. For 2016 the tax rate is 0.13231 percent.
  - Cyprus:
    - 2011 Special Tax on Credit Institutions base: total bank deposits in previous year (restricted to max 20 percent of total taxable profits; excludes deposits from foreign financial institutions and interbank deposits).
    - Rate initially 0.095 percent; increased to 0.15 percent in 2015; rate reduced to 0.0375 percent in 2016 as phase-out began.
  - United Kingdom:
    - 2011 Bank Levy on total liabilities net of equity and insured deposits with exemption up to £20 billion. Initial rates: 0.088 percent for short-term unsecured liabilities and 0.044 per cent for long-term unsecured liabilities.
    - Rates increased to 0.21 and 0.105 percent respectively in 2015.
    - Announced reductions to 0.10 percent (short-term) and 0.05 percent (long-term) during 2015–21; new tax on banking sector profit introduced in 2016 at a rate of 8 percent.
- Evidence on bank levies:
  - Average bank levy induced an increase in regulatory capital ratio by around 0.9 percentage points (Devereux, Johannesen, and Vella 2013).
  - Some reduction in funding risk was offset by increases in portfolio risk; funding risk reductions occurred mainly for banks with relatively high capital ratios.
  - Schepens (2016) finds the Belgian ACE reduced risk taking mostly by the least capitalized Belgian banks.

### Shadow banking and leveling the playing field
- Banks may have lower funding costs due to implicit/explicit government insurance; regulatory constraints may raise bank funding costs relative to shadow banks.
- To prevent regulatory arbitrage, options include:
  - Introduce taxes specifically on shadow banks to level the playing field.
  - Extend bank levies to the shadow banking sector with rate differentiation reflecting funding cost differences induced by regulatory regimes.

### Conclusions (policy implications)
- Systematic tax distortions favoring debt over equity remain a significant concern for macroeconomic stability; debt bias affects both households and firms.
- Household debt bias:
  - Primary source can be mitigated by gradual reform of generous tax deductions for mortgage interest.
  - Reforms should be gradual to avoid house-price disruptions in the short run.
  - Housing-related taxes (such as transaction taxes) have been used to stabilize house prices with mixed success; timing and calibration are hard.
  - Prefer designing stable and neutral tax systems with respect to housing; recurrent property taxes can help mitigate house-price volatility risks and have additional efficiency and fairness merits.

*Source: Annex 5 (excerpt).*

### 47.      Corporate debt bias remains a stability concern, including in the financial sector—

### 47.      Corporate debt bias remains a stability concern, including in the financial sector—

### Corporate debt bias: summary finding
- Tax systems increase corporate debt ratios by between 4½ and 9 percent of total assets, including in financial institutions.
- New evidence shows that this is also true for non-bank financial institutions, including finance companies and insurance companies.
- Corporate debt bias remains a macroeconomic stability concern despite strengthened regulation of financial institutions.

### Policy options to address corporate debt bias
- Limit interest deductibility:
  - Thin capitalization rules are one approach, but are effective in addressing debt bias only if applied to all debt (not only intragroup debt of multinationals) and should be based on a fixed ratio of debt-to-equity.
- Cash-flow taxation:
  - Some form of cash-flow taxation is a possible alternative, although experience with these is limited.
- Allowance for Corporate Equity (ACE):
  - An ACE system can neutralize corporate debt bias and has been effectively applied in several countries without creating major implementation problems.
  - The commonly cited drawback—revenue loss due to narrowing of the corporate tax base—can be mitigated by design choices (for example, granting the allowance only for incremental equity, or combining ACE with limitations on interest deductibility).
  - New opportunities for tax avoidance under an ACE can be mitigated by countermeasures.
- Sector-specific measures for financial institutions:
  - Sector-specific tax measures may be warranted for financial institutions because of particularly large externalities from over-leverage.
  - Bank levies on non-core funding have been effectively used in Europe to reduce bank leverage.
  - Consideration could be given to a sector-specific allowance for corporate equity (above the regulatory minimum capital) to encourage capitalization and resolve the tension between regulatory policies that encourage equity finance and tax policies that encourage debt finance.

### Policy guidance and timing
- Addressing debt bias should feature prominently in national tax reform agendas to mitigate macroeconomic stability risks.
- Several approaches have proved effective; their design can be tailored to countries’ needs and circumstances.
- The current low interest rate environment facilitates the feasibility of these reforms by mitigating implications for both public and corporate finances.

### Annex I — Property taxes and house price volatility (United States, 2005–14)
- Data and measures:
  - Sample: 51 U.S. states and several MSAs located in those states, period 2005–14.
  - Effective property tax rate: ratio of the median annual property tax payment to the median property value (American Community Survey).
  - House price data: Federal Housing Finance Agency; annual house prices estimated as averages of four quarters.
  - House price volatility measured as the standard deviation using a five-year backward moving window, based on (i) annual growth rates and (ii) percentage deviations from the HP-filtered value (HP smoothing parameter set to 100).
  - Panel is balanced with 510 state-year observations.
  - Average effective property tax rate in the sample is 1 percent, with standard deviation of 0.5 percent.
  - Average volatility (standard deviation) of house prices is 5–6 percent, depending on measure.
- Empirical prediction:
  - Theory predicts a negative slope coefficient on property tax rate for price volatility (β < 0): higher property taxes flatten housing demand and reduce price responsiveness to demand shocks.
- Main empirical results (Annex Table 1.2):
  - In all specifications the slope coefficient of the property tax variable is negative (and in most cases significantly so).
  - Economic significance: a 1 standard deviation increase in property tax rates (0.48 percentage point) leads to:
    - 1.3–1.7 percent reduction in volatility based on the standard deviation of the growth rate measure.
    - 0.5–0.6 percent reduction in volatility based on the de-trended measure.
  - Poghosyan (2016) robustness: results are robust to endogeneity (using average property tax rates in neighboring states as an instrument or using dynamic GMM).
  - A difference-in-difference approach shows higher state property taxes reduce house price volatility more in MSAs with more rigid housing supply.
  - The average economic effect across all regressions suggests an increase in property tax rates by 0.48 percentage points leads to a 3 percent reduction in house price volatility.
- Selected coefficients (from Annex Table 1.2):
  - Property tax rate coefficients (columns I–VIII): -3.542***, -2.784***, -3.210***, -2.802***, -1.283**, -1.111, -1.347**, -1.177.
  - Lagged dependent variable coefficients: 0.579***, 0.551***, 0.497***, 0.526***, 0.466***, 0.462***, 0.465***, 0.474***.
  - Economic significance entries (columns I–VIII): -1.7, -1.3, -1.5, -1.3, -0.6, -0.5, -0.6, -0.6.
  - Observations: 459 (for each column); # states: 51. R2 ranges reported (e.g., 0.365, 0.588, 0.434, 0.598, 0.215, 0.460, 0.219, 0.465).

### Annex II — Taxation and financial institutions’ leverage (firm-level analysis)
- Research scope and contribution:
  - Extends prior work by including investment banks, shadow banks (non-bank financial institutions), and insurance companies, with recent data covering seven years since the GFC.
  - Uses firm-level unconsolidated balance sheet data to align tax rate effects with operations within national borders; this reduces tax-measurement error but may capture both debt bias and debt shifting.
- Empirical specification:
  - Dependent variable: leverage ratio (total debt liabilities over total assets) for financial institution i in year t.
  - Key regressor: top statutory corporate income tax (CIT) rate in the country.
  - Controls: lagged leverage (autoregressive term), size (log total assets, linear and quadratic), profitability (pre-tax profits to assets), GDP growth, inflation, riskiness (ratio of risk-weighted assets to total assets), firm fixed effects, year fixed effects.
  - Expectation: CIT has a positive effect on leverage (debt bias coefficient expected positive).
  - Long-term effect computed as β2 /(1−β1) given autoregressive coefficient β1 and short-run tax coefficient β2.
  - Distinction made between regular banks, investment banks, and non-bank financial institutions via interaction terms with CIT.
- Estimation:
  - Use Blundell and Bond (1998) system GMM dynamic panel estimator to correct for bias from lagged dependent variable and endogeneity of risk.
  - Panel dimension: T=15, N>14,000, making system GMM appropriate.
- Data sources and sample:
  - Banks and finance companies: Bankscope (Bureau van Dijk). About 151,000 observations, almost 14,000 distinct banks and finance companies in 131 countries over 2001–15.
    - Composition: almost 93 percent “regular” banks; non-bank finance companies 5 percent; investment banks 2 percent.
  - Insurance companies: Orbis Insurance Focus (Bureau van Dijk). Over 68,000 observations, more than 7,500 firms from 85 countries over 2005–2015.
    - Composition: 30 percent life insurers; 70 percent non-life insurers.
  - Data processing:
    - Unconsolidated firm accounts used.
    - Variables winsorized at the 1 percent level except tax, inflation, and GDP growth.
    - Firms lacking at least three consecutive years of financial results dropped; countries with less than 20 observations dropped.
- Summary statistics (Annex Table 2.1 — Banking):
  - Full sample entries (selected):
    - Leverage, pct total assets (winsorized fraction .01): 151,463 observations; mean 86.5; std. dev. 13.2; min 10.9; max 98.5.
    - CIT rate (pct): 151,174 observations; mean 31.2; std. dev. 6.5; min 8.5; max 60.0.
    - Log assets: mean 5.2; std. dev. 1.7; min 1.8; max 9.9.
    - GDP growth (pct): 151,449 observations; mean 1.9; std. dev. 2.6; min -62.1; max 104.5.
    - CPI inflation (pct): 151,048 observations; mean 2.9; std. dev. 3.9; min -37.2; max 359.9.
    - Pre-tax profit, pct total assets: 151,120 observations; mean 1.0; std. dev. 1.9; min -6.4; max 10.6.
    - Risk (winsorized fraction .01): 100,663 observations; mean 0.7; std. dev. 0.2; min 0.2; max 1.0.
    - Total assets, million USD: mean 1,119; std. dev. 3,194; min 620; max 753,000 (note: table formatting preserves given entries).
  - Subsample examples:
    - Regular banks [13,047 bank IDs; 131 countries].
    - Investment banks [308 bank IDs; 67 countries].
    - Finance companies [635 bank IDs; 85 countries].
  - Additional sample aggregates and statistics are reported in Annex Table 2.1 consistent with the banking sample breakdown.

*Source: IMF, "Tax Policy, Leverage and Macroeconomic Stability" (content unit 47 and Annexes I–II).*

### Annex Table 2.2. Summary Statistics—Insurance

### Annex Table 2.2. Summary Statistics—Insurance

### Sample composition and measurement
- Sample dominated by observations in the United States: "over 60 percent of the banking observations and over 40 percent of the insurance observations."
- Insurance leverage computed using liabilities net of technical reserves; analysis focuses on liabilities in excess of technical reserves.
- Macroeconomic and tax variables:
  - Inflation and GDP growth from IMF’s World Economic Outlook database.
  - Statutory CIT rates from IMF’s Fiscal Affair Department database.

### Key cross-sectional statistics (winsorized fractions as noted)
- Insurance-sample mean leverage and related moment statistics (as reported in the table header):
  - Leverage, pct total assets (winsorized fraction .01): Obs 19,469; Mean 11.5; Std. Dev. 16.1; Min 0.07; Max 73.0
  - CIT rate (pct): Obs 19,419; Mean 28.3; Std. Dev. 7.5; Min 8.5; Max 39.0
  - Log assets (winsorized fraction .01): Obs 19,469; Mean 6.1; Std. Dev. 2.7; Min -0.4; Max 11.0
  - Log assets sq (winsorized fraction .01): Obs 19,469; Mean 44.7; Std. Dev. 32.9; Min 0.2; Max 122.0
  - GDP growth (pct): Obs 19,469; Mean 2.1; Std. Dev. 3.0; Min -15.1; Max 15.2
  - CPI inflation (pct): Obs 17,993; Mean 2.7; Std. Dev. 3.0; Min -21.0; Max 48.7
  - Pre-tax profit, pct total assets (winsorized fraction .01): Obs 19,411; Mean 2; Std. Dev. 7; Min -29; Max 38
  - Total assets, million USD (winsorized fraction .01): Obs 19,469; Mean 5,654; Std. Dev. 13,240; Min 162; Max 574
  - Total equity, million USD (winsorized fraction .01): Obs 19,469; Mean 406; Std. Dev. 1,080; Min 0; Max 6,633
  - Pretax profits, million USD (winsorized fraction .01): Obs 19,411; Mean 481; Std. Dev. 148; Min -948; Max 866

- Insurance subsamples (counts and coverage):
  - Life insurance companies: 2,294 firm IDs; 80 countries
  - Non-life insurance companies: 5,348 firm IDs; 85 countries

- Comparative leverage notes:
  - Average leverage ratio across the regular banks sample: 87 percent.
  - Investment banks leverage: 73 percent.
  - Finance companies leverage: 77 percent.
  - For insurance sample (excess liabilities):
    - Average leverage for life insurers: 11.5 percent.
    - Average leverage for non-life insurers: 13.9 percent.
  - Mean statutory CIT rate in the banking sample: 31.2 percent (Min 8.5 percent; Max 60 percent — Libya in the early 2000s).
  - Mean CIT rate in the insurance sample: 29.4 percent.

### Cross-plot observations (Annex Figure 2.1 summary)
- Top panels: strong positive correlation between leverage and the statutory tax rate for banks (including investment banks and shadow banks), life insurers, and non-life insurers.
- Bottom panels: strong positive correlation between leverage and bank size; weak positive or no correlation between insurance firm size and leverage.

### Regression results — Regular banks (Annex Table 2.3)
- Data sample: Regular banks, unconsolidated accounts. 2001–15.
- Estimation methods and diagnostics:
  - Columns (1)-(2): OLS (with and without fixed effects).
  - Columns (3)-(5): two-step system GMM with lagged dependent variable and risk instrumented. Lag limits = (2,1). Robust standard errors clustered at country level.
- Selected coefficient estimates (system GMM and OLS reported as in table):
  - Lagged leverage (pct total assets):
    - Column (1) OLS no FE: 0.81*** [0.01]
    - Column (2) OLS FE: 0.37*** [0.01]
    - Column (3) System GMM full period: 0.34*** [0.02]
    - Column (4) System GMM before 2008: 0.36*** [0.03]
    - Column (5) System GMM 2008 onwards: 0.69*** [0.08]
  - CIT rate (pct):
    - Column (1) 0.04*** [0.01]
    - Column (2) 0.06* [0.03]
    - Column (3) 0.11** [0.05]
    - Column (4) 0.24*** [0.07]
    - Column (5) 0.05 [0.04]
  - Log assets (winsorized .01):
    - Column (1) 2.35*** [0.68]
    - Column (2) 8.83*** [1.69]
    - Column (3) 6.38*** [0.62]
    - Column (4) 6.74*** [0.61]
    - Column (5) 3.13*** [0.85]
  - Log assets sq (winsorized .01):
    - Column (1) -0.17*** [0.05]
    - Column (2) -0.40*** [0.09]
    - Column (3) -0.44*** [0.07]
    - Column (4) -0.51*** [0.05]
    - Column (5) -0.21*** [0.06]
  - Pre-tax profit, pct total assets (winsorized .01):
    - Column (1) -0.78*** [0.06]
    - Column (2) -0.61*** [0.05]
    - Column (3) -1.21*** [0.08]
    - Column (4) -1.55*** [0.03]
    - Column (5) -0.74*** [0.10]
  - GDP growth (pct):
    - Column (1) 0.07** [0.03]
    - Column (2) 0.11** [0.06]
    - Column (3) -0.07 [0.07]
    - Column (4) 0.09 [0.15]
    - Column (5) 0.03 [0.04]
- Observations and panel dimensions:
  - Observations: Column (1)-(2) 131,118; Column (3) 91,950; Column (4) 38,372; Column (5) 47,104
  - Number of id: 13,024 (no FE sample), 9,970 (FE), 7,405 (full period GMM), 8,563 (before 2008), 7,082 (2008 onwards)
- Diagnostics and long-term effects:
  - Long-term effect (reported): overall 0.23*** [0.05]; FE model 0.09* [0.05]; full period GMM 0.17** [0.07]; before 2008 0.38*** [0.11]; 2008 onwards 0.17 [0.12]
  - Hansen p and AR statistics reported in table.

- Interpretation from text:
  - OLS regressions suggest tax rate has a strongly significant impact on leverage over the period.
  - System GMM confirms positive and statistically significant tax effect on leverage over the full period and pre-crisis period.
  - Short-term tax effect pre-crisis: 0.24 (double the effect during full period).
  - Tax effect not statistically significant post-crisis, possibly due to tighter regulation and higher capital buffers.
  - Long-run effect larger due to sluggish response; long-run effect pre-2008 = 0.38; entire period = 0.17; similar to De Mooij and Keen (2016).

### Regression results — Extended sample: regular + investment banks + finance companies (Annex Table 2.4 and 2.5)
- Data sample: regular banks, investment banks, and non-bank finance companies, unconsolidated accounts. 2001–15.
- Main findings (system GMM, with and without interaction terms):
  - Including investment banks and finance companies does not change overall CIT coefficient observed for regular banks.
  - Interaction term (CIT * Investment bank): -0.17** [0.08] (basic specification) and 0.14*** [0.02] (interaction specification) and -0.14* [0.07] (2008 onwards column) — sign and significance vary by specification and period.
    - Implication: for investment banks over the entire sample period, CIT rate is not a significant driver of leverage (net effect not significantly different from zero).
    - Pre-GFC: investment banks exhibited much higher sensitivity to tax rates than regular banks; post-GFC tax considerations no longer significant for investment bank leverage.
  - Interaction term (CIT * Finance companies): -0.06* [0.04] in one specification; other columns show -0.05 [0.05] and -0.02 [0.03] — generally negative but smaller in absolute value than CIT coefficient.
    - Net effect for finance companies not significant on full or post-crisis sample; pre-GFC net effect positive.
  - CIT rate (pct) in extended sample (system GMM basic specification):
    - Column (1) full period 0.12** [0.05]
    - Column (2) before 2008 0.23*** [0.06]
    - Column (3) 2008 onwards 0.07 [0.05]
  - Lagged leverage in extended sample:
    - 0.34*** [0.02] (full period); 0.37*** [0.03] (before 2008); 0.65*** [0.09] (2008 onwards)
  - Long-term effects (basic): 0.19** [0.07] (full); 0.36*** [0.09] (before 2008); 0.21* [0.11] (2008 onwards)
- Outlier treatment (Annex Table 2.5, winsorized 10 percent):
  - CIT coefficients remain significant and positive; now also significantly different from zero for the post-crisis period.
  - System GMM coefficients with winsorized 0.1:
    - CIT rate (pct): 0.03*** [0.00] (full); 0.06*** [0.01] (before 2008); 0.04*** [0.01] (2008 onwards)
    - Lagged leverage: 0.80*** [0.02] (full); 0.70*** [0.02] (before 2008); 0.77*** [0.05] (2008 onwards)
    - Long-term effect: 0.13*** [0.03] (full); 0.19*** [0.05] (before 2008); 0.15*** [0.02] (2008 onwards)

### Regression results — Insurance subsamples (Annex Tables 2.6 and 2.7)
- Life insurance companies (data sample: unconsolidated accounts. 2005–15):
  - Estimation: Column (1) OLS; Columns (2)-(4) two-step system GMM (lag limits (2,1)).
  - Key coefficients (system GMM, summary from table):
    - Lagged leverage: 0.93*** [0.02] (full OLS reported) and system GMM values in table columns.
    - CIT rate (pct): 0.03* [0.02] (full OLS column); system GMM: 0.08*** [0.03] (before 2008); 0.26*** [0.09] (before 2008 column indicated larger pre-crisis effect); 0.09*** [0.03] (2008 onwards)
    - Log assets (winsorized .01): -0.32*** [0.07] (full); -0.80*** [0.24] (before 2008); -0.79** [0.38] (2008 onwards); combined effect on size implies strong inverse relation for largest firms.
    - Log assets sq: 0.03*** [0.01] (full); 0.06*** [0.02] (before 2008); 0.11** [0.05] (before 2008 column)
  - Sample size and diagnostics:
    - Observations 17,015 (full); R-squared 0.88 (OLS).
    - Number of id 2,279; Hansen p reported 0.51 (full).
    - Long-term effect reported: 0.40 [0.25] (OLS long-term effect); system GMM long-term effects: 0.28*** [0.09] (before 2008); 0.43*** [0.14] (2008 onwards); 0.48*** [0.14] (other column)
  - Text interpretation:
    - Short-term effect for life insurers over 2005–2015: 0.08.
    - Splitting sample: pre-crisis coefficient 0.26; post-crisis coefficient 0.09.
    - Long-run coefficient on full period: around 0.28 (larger than 0.17 for regular banks).
    - Asset size effect: combined coefficient on log assets equals -1.5 for largest firm example (calculation shown in text: -0.8 * 11 + 0.06 * 122 = -1.5), indicating larger firms associated with lower leverage.

- Non-life insurance companies (data sample: unconsolidated accounts. 2005–15):
  - Estimation: Column (1) OLS; Columns (2)-(4) two-step system GMM (lag limits (2,1)).
  - Key coefficients (system GMM and OLS summarized):
    - Lagged leverage: 0.86*** [0.01] (OLS); system GMM columns report 0.64*** [0.03] (before 2008); 0.34*** [0.12] (2008 onwards); 0.63*** [0.03] (2008 onwards column)
    - CIT rate (pct): 0.02 [0.01] (full OLS); 0.05** [0.03] (before 2008); 0.16** [0.08] (before 2008); 0.06* [0.03] (2008 onwards)
    - Log assets (winsorized .01): 0.17 [0.15] (full OLS); system GMM shows small/insignificant size coefficients; log assets sq small negative.
    - Pre-tax profit, pct total assets: -0.02** [0.01] (full); -0.03*** [0.01] (before 2008); -0.08*** [0.02] (2008 onwards)
  - Sample size and diagnostics:
    - Observations 42,243 (full); R-squared 0.74 (OLS).
    - Number of id 5,320; Hansen p reported 0.03 (full), with AR diagnostic p-values reported.
    - Long-term effect reported: 0.14 [0.09] (full); 0.14** [0.07] (before 2008); 0.24** [0.11] (2008 onwards); 0.15* [0.08] (other column)
  - Text interpretation:
    - Non-life insurers show similar patterns to life insurers but with slightly smaller coefficient estimates: 0.05 (whole sample), 0.16 (pre-crisis), 0.06 (post-crisis).
    - Long-run coefficients range between 0.14–0.24.
    - GMM diagnostics: problematic for some specifications — tests reject AR(1) errors in Arellano-Bond Model as expected; Hansen test and AR(2) indicate autocorrelation concerns for some specifications.

### Comparative findings and interpretation
- Tax rates are associated with higher leverage across banking and insurance samples, with stronger effects pre-GFC than post-GFC in many specifications.
- Magnitude comparisons:
  - Regular banks: long-run effect ~0.17 (entire period) and 0.38 (pre-2008).
  - Life insurers: long-run coefficient on full period ~0.28; short-term effect ~0.08 (full), 0.26 pre-crisis, 0.09 post-crisis.
  - Non-life insurers: short-term effects ~0.05 (full), 0.16 (pre-crisis), 0.06 (post-crisis); long-run 0.14–0.24.
- Firm size relation to leverage:
  - Life insurers: larger firms associated with lower leverage (combined coefficient example yields -1.5 for largest firm).
  - Banks and extended sample (regular + investment + finance): size and leverage positive (example calculation yields 21.3 for largest firm).

*Annex Table 2.2. Summary Statistics—Insurance (excerpted content).*

### Annex III. Thin Capitalization Rules

### Annex III. Thin Capitalization Rules

### Overview of TCR varieties and prevalence
- Approximately two-thirds of thin capitalization rules (TCRs) restrict interest deduction only for related-party debt.
- 50 countries employ a fixed debt-equity rule that denies deductibility if the debt-to-equity ratio exceeds a fixed value; common ratios range from 1:1 to 5:1, with 3:1 most common.
- By 2016, eight countries had replaced fixed debt-equity TCRs with fixed interest/income (earnings stripping) rules.
- Earnings-stripping rules commonly cap interest deductions to 30 percent of earnings before interest, taxes, depreciation and amortization (EBITDA).
- Arm’s-length (worldwide gearing) rules compare a subsidiary’s debt-to-equity ratio with that of the global enterprise; examples include:
  - Australia: permits up to 100 percent of the global benchmark.
  - New Zealand: permits up to 110 percent of the global benchmark.
  - Kazakhstan, South Africa, United Kingdom adopt worldwide gearing tests (United Kingdom proposed replacement by an earnings stripping rule).
- Some countries combine approaches (e.g., Denmark, France, United States). Example: United States defines excess interest expense as net interest exceeding 50 percent of taxable income and disqualifies that interest for deduction only if the corporation’s debt to equity ratio exceeds 1.5:1.
- An increasing number of countries allow alternative or combined rules and different computation methods for denied interest (e.g., use of different interest rates to compute denied portions).

### Other design features and exceptions
- Carry forwards:
  - Some rules allow indefinite carry-forward of denied interest.
  - Other rules limit carry-forward to a number of years.
  - Some countries re-characterize denied interest as dividends with no carry-forward.
- Exceptions:
  - In only 15 countries listed in Annex Table 3.1 does the TCR apply to financial institutions (often with a modified ratio); in all other countries the financial sector is exempt.
  - In many countries the TCR applies only to firms that exceed a threshold of net interest payment.
- Table footnotes (selected specific features preserved as stated):
  - Kenya applies interest-stripping rules to foreign-controlled companies.
  - Romania does not distinguish between related and unrelated party debt but allows full deduction of interest on loans from banks and other financial institutions.
  - Net interest expense is fully deductible if it does not exceed €1million even if the interest-EBITDA ratio is above the 30 percent.

### Empirical strategy to assess TCRs and corporate sector stability
- Focus: effect of TCRs on consolidated corporate debt ratios of non-financial corporations and on firms’ financial distress (Altman Z-score).
- Benchmark firm-panel model (equation (1)) estimates consolidated total debt-asset ratio with:
  - thinCap indicator for TCR presence,
  - statutory CIT rate (citRate),
  - vector of controls including macro variables,
  - year fixed effects and sector fixed effects; variant includes industry-year fixed effects.
- Distinction in analysis between:
  - total-debt TCR (applies to total debt), and
  - related-party TCR (applies only to related-party debt).
- Difference-in-difference specification (equation (2)) interacts industry tangibility (Tang_s) with tax variables to capture differential responsiveness by industry tangibility.
- Financial distress measured using Altman Z-score:
  - Z = 3.25 + 6.56*X1 + 3.26*X2 + 6.72*X3 + 1.05*X4
    - X1 = working capital / total assets
    - X2 = retained earnings / total assets
    - X3 = earnings before interest and taxes / total assets
    - X4 = book value of equity / book value of total liabilities
  - Binary distress indicator = 1 if Z-score < 2, 0 otherwise (firms below this cut-off correspond to the lowest 7th percentile, indicating high bankruptcy risk).

### Data used
- Firm-level source: ORBIS database (consolidated accounts).
- Sample: about 263,000 observations corresponding to 59,126 distinct firms in 42 countries, period 2005 to 2014.
- Winsorization: variables winsorized at the 0.5 percent level.
- Key sample statistics:
  - Average consolidated debt-to-assets ratio: 62 percent.
  - Median consolidated debt-to-assets ratio: 58.7 percent.
  - Mean statutory CIT rate: 22.2 percent.
  - Minimum statutory CIT rate: 8.5 percent.
  - Maximum statutory CIT rate: 37.8 percent.
- Macroeconomic controls sourced from IMF World Economic Outlook and World Bank WDI; TCRs and CIT rates from IMF Fiscal Affairs Department database.

### Main empirical findings (Annex Table 4.2 and related results)
- Regression context: dependent variable = consolidated debt ratio; robust standard errors; various fixed-effect specifications.
- Key estimated coefficients (preserved exactly from table and text):
  - thinCap–all = -1.333*** (column (1); standard error (0.178)).
  - thinCap–total = -5.129*** (column (2); standard error (0.171)).
  - thinCap–related-party = -0.590*** (column (3); standard error (0.180)).
  - citRate ≈ 0.301*** to 0.313*** across columns (exact values shown in table):
    - 0.301*** (column (1); (0.0182))
    - 0.308*** (column (2); (0.0181))
    - 0.313*** (column (3); (0.0182))
  - gdpGrowth coefficients (examples):
    - -0.178*** (column (2); (0.0105))
    - -0.203*** (column (4); (0.0438))
    - -0.210*** (column (6); (0.0422))
  - inflationCpi examples:
    - 0.921*** (column (1); (0.0408))
    - 0.661*** (column (3); (0.0431))
  - ebitdaShareOfAssets ≈ -0.728*** across specifications (e.g., (0.00663)).
  - interestRateReal ≈ 1.587*** to 1.681*** across columns.
- Interpretation and robustness:
  - Presence of some TCR (thinCap–all) reduces consolidated debt ratio by about -1.3 percentage points (column (1)).
  - Total-debt TCRs have a much stronger negative effect on corporate debt: coefficient -5.1 (column (2)), robust across specifications (e.g., -5.360*** in column (6)).
  - Related-party TCRs have a much smaller effect: coefficient -0.590*** (column (3)) and become statistically insignificant when industry-year fixed effects are included (columns (4)–(6)).
  - A Wald test strongly rejects the hypothesis of no difference between total-debt and related-party TCRs (p-value 0.00).
  - The coefficient for statutory CIT rate is positive and significant with values around 0.3, consistent with literature on tax-induced debt bias.
- Correlations noted in descriptive plots:
  - Positive correlation between corporate debt ratios and CIT rates.
  - Strong negative correlation between Z-score and debt ratio (higher leverage associated with lower Z-score and higher bankruptcy risk).

### Summary implication
- Total-debt TCRs materially reduce consolidated corporate leverage (large, statistically significant negative coefficients).
- Related-party TCRs have weaker and less robust effects on consolidated leverage.
- Higher statutory CIT rates are associated with higher consolidated debt ratios (debt bias), with estimated coefficients around 0.3.
- High leverage is associated with increased financial distress as captured by Altman Z-score; use of Z-score < 2 identifies firms in the lowest 7th percentile facing high bankruptcy risk.

*Source: Annex III and Annex IV, "Thin Capitalization Rules" and "Thin Capitalization Rules and Corporate Sector Stability" (selected tables and text) from the provided IMF content unit.*

### Annex Table 4.3, reports results from estimating specification (2). In all columns, the estimated effect

### _100716 - Annex Table 4.3, reports results from estimating specification (2). In all columns, the estimated effect

### Estimated effects of TCRs and tangibility (Annex Table 4.3)
- In all specifications, the interaction term between taxes and tangibility on the consolidated debt ratio is positive in the text discussion (debt bias higher in capital-intensive industries), while the table reports interaction coefficients for thin-cap variables with tangibility as follows:
  - thinCapAll × tangibility: −0.0261* (0.0133)
  - thinCapTotal × tangibility: −0.0853*** (0.0226); −0.0888*** (0.0233)
  - thinCapRelatedParty × tangibility: −0.00557 (0.0139)
  - Citrate × tangibility: 0.0064*** (0.00141); 0.0064*** (0.00145); 0.0057*** (0.00133); 0.0057*** (0.00131) across columns
  - ebitdaShareOfAssets: −0.767*** (0.0249); −0.763*** (0.0251); −0.764*** (0.0250); −0.764*** (0.0250)
  - operatingRevenueRatio: 1.978*** (0.158); 1.892*** (0.159); 1.886*** (0.159); 1.886*** (0.159)
  - Constants reported: 21.67*** (2.881); 23.21*** (2.906); 23.59*** (3.011); 23.64*** (2.989)
- Sample and fit:
  - Observations: 369,844; 358,978; 358,978; 358,978 (by column)
  - R2: 0.169 (all columns)
  - Industry fixed effects: Yes; Country-year FE: Yes

### Interpretation and driver of results
- Positive interaction interpretation in text: debt bias is higher in capital-intensive industries because firms relying on tangible assets have easier access to credit and react more sensitively to corporate taxes.
- Column (2) adds interaction between the TCR-variable and tangibility: coefficient statistically significant and negative, implying TCRs reduce debt ratios differentially more in industries with higher tangibility.
- Columns (3) and (4) show the finding is driven solely by TCRs that apply to total debt; related-party TCR coefficient is statistically insignificant.

### Magnitude example (partial effects)
- Estimated β_sh (column 4 example): −0.088.
- Partial effect of thinCapTotal on debt ratio evaluated at:
  - 25th percentile of tangibility: −0.71 percentage points (thinCapTotal decreases the debt ratio by 0.71 percentage points, compared to the base effect and given the 25th percentile level of Tang).
  - 75th percentile of tangibility: −2.2 percentage points (thinCapTotal decreases the debt ratio by 2.2 percentage points, compared to the base effect and given the 75th percentile level of Tang).
- Difference between interaction effects: −1.5 percentage points (−2.2 + 0.71).

### Additional note on multinationals
- For a multinational, results suggest a TCR reduces group-level leverage, but the analysis cannot infer whether debt increases in affiliates located in high-tax countries where no TCR applies.

---

### Thin-capitalization rules and corporate stability (Annex Table 4.4)
- Dependent variable: A binary financial distress indicator based on the Z-score.
- Main coefficients (LPM and Logit columns):
  - thinCap–total:
    - LPM (columns 1 and 2): −0.050*** (0.009); −0.050*** (0.004)
    - Logit (columns 3 and 4): −1.493*** (0.089); −1.495*** (0.090)
  - thinCap–related-party: −0.003 (0.012); −0.003 (0.006); 0.018 (0.049); 0.024 (0.050)
  - citRate: −0.000 (0.001); −0.001 (0.000); −0.004 (0.004); −0.004 (0.004)
  - gdpGrowth: −0.002*** (0.000); −0.002*** (0.000); −0.028*** (0.003); −0.029*** (0.003)
  - interestRateReal: 0.002*** (0.001); 0.002*** (0.001); 0.030*** (0.004); 0.031*** (0.004)
  - Volatility: 0.504*** (0.047); 0.510*** (0.030); 3.787*** (0.269); 3.916*** (0.282)
  - Constants: 0.078*** (0.025); 0.075*** (0.013); −12.82*** (0.574); −12.55*** (0.993)
- Sample and fit:
  - Observations: 44,590; 44,590; 44,589; 44,466 (by column)
  - R2: 0.053 (LPM columns)
  - Fixed effects: Industry and year FE; Industry-year FE in columns 2 and 3
- Interpretation:
  - thinCap–total has a negative sign in all columns: total-debt TCRs reduce financial distress and increase likelihood of being in the safe zone.
  - LPM estimates: a total–debt TCR reduces probability of bankruptcy risk by 5 percentage points.
  - Related-party TCRs show no significant effect; in some columns they appear with the wrong sign.
  - Conclusion: total-debt TCRs are effective in reducing bankruptcy risk associated with debt bias; related-party-only TCRs are not.

---

### Allowance for Corporate Equity (ACE) — design and mechanics (Annex V)
- Purpose: ACE supplements current deductibility of interest with a deduction for the normal return on equity. Deduction = ACE base × ACE rate.
- ACE base:
  - Base can be entire equity stock or incremental relative to a base year (e.g., last year’s equity).
  - Incremental base roughly equals new equity issues plus retention of after-tax profits relative to last year.
  - Equity participations in other firms should be subtracted from the ACE base to avoid duplication; if equity minus participations is negative, ACE may add to the CIT base rather than deduct.
  - Participations in foreign companies should be excluded from ACE base because foreign equity returns are not subject to domestic tax; relief for double taxation is usually granted.
- ACE rate (notional return):
  - Full tax neutrality requires imputed rate equal to rate at which shareholders discount tax savings from future ACE allowances.
  - If tax law allows full loss offset and CIT rate fixed, discount rate = risk-free rate (e.g., government bond rate).
  - In practice, restrictions on loss offsets introduce risk; a substantial part of the risk tied to bankruptcy probability.
  - Setting imputed rate equal to interest on company’s long-term debt would approximate neutrality, but administratively single notional rate for all firms is common.
  - In countries with developed corporate bond markets, average corporate bond interest rate is a natural benchmark.

### ACE systems in practice (selected entries from Annex Table 5.1)
- Belgium (Since 2006):
  - Base: Book value of equity; not incremental system
  - Rate: Average monthly government bond rate of year preceding fiscal year by 2 years. Rate initially capped at 6.5 percent, lowered to 3 percent in 2013. Special SME rate is 0.5 pp higher.
  - Note: The notional return is deductible. In 2013, carrying forward of unused allowances was eliminated and a tax on distributed dividends of large firms was levied.
- Italy (Since 2012):
  - Base: New equity (increase in equity over a 2010 base); incremental system
  - Rate path: For the first three fiscal years (2011, 2012, 2013): 3 percent; 4 percent for the 2014 fiscal year; 4.5 percent for 2015; 4.75% 2016; subsequent years: rate based on Italian public debt securities' average return and a risk factor, annually set by the Minister of Finance.
  - Note: Italian resident companies and Italian permanent establishments of non-resident companies can deduct the NID (with certain exclusions). New equity does not include profits from that year and may not exceed company equity at end of fiscal year.
- Cyprus (Since January 2015):
  - Base: New equity introduced on or after 1 January 2015 (issued share capital and share premium, fully paid); incremental system
  - Rate: Interest rate of the 10-year government bond yield of country where new equity is invested increased by 3% with a lower limit equal to the 10-year government bond yield of the Republic of Cyprus increased by 3%.
- Turkey (Since July 2015):
  - Base: Cash capital increase of paid-in capital; incremental system
  - Rate: 50 percent of the annual weighted average interest rate applied to Turkish Lira-denominated loans provided by banks.
  - Note: Benefit excluded for companies with high passive income or high financial assets, shares in subsidiaries or participations.

- Additional contextual notes:
  - Portugal: Share contributions during 2010–2013 and SME provisions (notional returns deductible).
  - Footnotes: Swiss parliament adopted incremental ACE in 2016; Denmark proposed incremental ACE in 2017 budget for implementation in 2019. Brazil’s system closer to dividend deduction and thus excluded from table.

---

### Tax planning risks with ACE and countermeasures (Annex VI)
- Main tax planning schemes discussed: (i) cascading of ACE relief; (ii) asymmetries in ACE structures; (iii) asymmetries across countries in ACE application.
- Cascading example and anti-cascading rule:
  - Example: Consolidated account shows assets and equity of 400. With ACE rate 10 percent, allowance = 40.
  - If firm organized as chain of subsidiaries, ACE could cascade: total allowance could expand to 3 x 40 = 120.
  - Anti-cascading rule: subtract equity represented by participations (domestic and foreign) from ACE base. In the example, subtractions reduce ACE base for H and S1 to zero; total ACE base of group reduced to 400 and allowance to 40 (granted to S2). Anti-cascading restores neutrality with respect to organizational structure.
- Other planning concerns (summarized from text):
  - ACE may make existing tax avoidance more attractive (e.g., equity-financed investment in assets yielding private benefits).
  - Incremental ACE can be abused via transformations of old into new equity (e.g., liquidation and asset transfers), though liquidation may trigger capital gains taxation.
  - Anti-avoidance rules (specific or general) can limit abuses; some countries (e.g., Belgium, Italy) implement such safeguards.
  - ACE regimes generally also grant allowance to non-corporate firms; differences across corporate and non-corporate treatment can create planning opportunities.

*Source: IMF — Annex Tables and accompanying annex text as provided in the supplied content unit.*

### Annex Figure 6.1. ACE Duplication and the Anti-Cascading Rule

### Annex Figure 6.1. ACE Duplication and the Anti-Cascading Rule

### Asymmetries in ACE
- Example balance-sheet numbers and ACE outcomes:
  - Consolidated company (Company C): assets of 400, financed by 300 of external debt and 100 of equity.
  - Company C has an ACE base of 100, which gives rise to an allowance of 10.
  - Unconsolidated group (Group U):
    - Subsidiary S: allowance of 40 (10 percent of 400).
    - Holding H: negative ACE base of 300 (corresponding to its debt), which would create an addition of 30 to its tax base if negative ACE base permitted.
  - If negative ACE base is permitted, allowances are equivalent for C and U and ACE is neutral to organizational structure.
  - If ACE base is capped at zero (zero lower bound):
    - ACE base of company H becomes zero instead of – 300, so there is no addition of 30 to H’s tax base.
    - ACE granted to S remains 40.
    - On balance, group U enjoys a larger overall allowance than company C; restructuring becomes attractive and the anti-cascading rule is effectively blunted.

### Anti-avoidance and domestic policy responses
- Policy recommendation to minimize tax planning through ACE:
  - Permit the ACE base to be negative so that ACE can generate tax payment for some companies.
- Political/economic constraint:
  - Countries typically impose a zero lower bound because of fear of exit of companies (such as holding companies) if confronted with an ACE addition.
- Anti-avoidance instruments described:
  - Specific or general anti-avoidance rule to cap total allowance for groups deemed artificially structured to avoid tax.
  - In specific cases where the zero cap is binding, deny interest deductions if the sum of participations exceeds debt (Belgium uses such instruments).
  - Comprehensive thin capitalization rule (earnings stripping rule) to cap interest deductions. Example:
    - In Annex Figure 6.2, company H has a debt-equity ratio of 3:1.
    - If interest deductions were capped at a ratio of 1:1, the interest deduction would be reduced from 30 to 10.

### International tax planning induced by ACE
- Cross-border incentive structure and example numbers:
  - Country A: adopts an ideal ACE with anti-cascading rule that allows ACE base to be negative (allowance can become an addition).
  - Country F: adopts a standard income tax; no ACE applies in country F (neither positive nor negative).
  - Consolidated firm: consolidated assets of 400 fully financed by equity ⇒ consolidated ACE base of 400 and associated allowance of 40 (benchmark).
  - Ownership structure in Annex Figure 6.3:
    - Production subsidiary S holds the assets.
    - Holding H2 owns S and is financed by an intra-company loan of 300 from H1 and an equity injection of 100 from H1.
    - External shareholders hold H1.
  - Tax outcomes when entities located across A and F:
    - In country A, positive ACE allowance is granted to H1 and S of, respectively, 30 and 40, while H2 faces a negative ACE on account of its negative balance.
    - It becomes attractive to locate S and H1 in country A (ACE generates an allowance) and H2 in country F (to avoid an addition to the tax base on account of negative ACE in A).
    - If H2 resides in country F, total allowance becomes 70 instead of 40: 30 to H1 and 40 to S — ACE becomes excessive relative to consolidated treatment.
    - Country A levies tax on intracompany interest receipts of 30 from H1, so country A’s overall tax base is not necessarily reduced; country F suffers base erosion due to an additional interest deduction of 30 that can be offset against other income.
- Policy implications:
  - International ACE-induced planning is analogous to classical debt shifting caused by differences in statutory CIT rates.
  - Countermeasures by non-ACE countries include strengthening thin capitalization rules or denying interest deductions for debt used only to finance participations in foreign companies (example noted: United Kingdom adopts such a rule, with caveats due to ECJ case 168-01).
  - Paradox: adoption of ACE by one country increases the importance for other countries to protect their base through thin capitalization rules; global adoption of ACE would make thin capitalization rules redundant.

### Annex VII — Simulations: Estimated base narrowing from introducing an ACE
- Data and scope:
  - Non-financial firms: firm-level data from ORBIS by Bureau van Dijk, unconsolidated accounts for a sample of over one million firms in OECD countries during 2012–13. Observations grouped into: extractive industries, manufacturing, services.
  - Financial firms: Bankscope data plus insurance company data from ORBIS; unconsolidated accounts for over 10,000 financial institutions in OECD countries during 2012–13. Observations grouped into five subsectors: regular banks; investment banks; bank holdings; finance companies; insurance companies (life and other).
  - Country coverage spans all OECD countries, with Belgium, Italy, and Portugal excluded because they already had some form of ACE in place during the period analyzed.
- Methodology and assumptions:
  - For each firm, ACE allowance = country-specific ACE rate × firm-specific ACE base.
  - ACE rate: set at the long-term government bond yield in each country (mimicking a notional risk-free rate of return to equity).
  - ACE base scenarios:
    - Non-financial firms: (i) book value of total equity (ACE1); (ii) equity increase in 2013 relative to base of 2012 (ACE3).
    - Financial firms: same two scenarios plus (iii) allowance only for equity in excess of minimum regulatory capital of 3 percent of assets (ACE2).
  - For financial firms, calculations correct for participations in other firms.
  - ACE amount per firm is capped at the before-tax profit (no refunds assumed); unused ACE not carried forward (estimations are static).
- Key simulation design notes:
  - ACE base and ACE rates are country- and firm-specific; aggregated ACE allowance is expressed as percent of underlying pre-tax profit to reflect percentage change in CIT base (and, if tax rates unchanged, percentage change in CIT revenues).
  - Calculations represent an upper bound of expected revenue effects because accounting data do not necessarily coincide with tax accounts and, for many firms, there is no information about share participations.
- Selected aggregate statistics from Annex Table 7.1 (2012 sample unless indicated):
  - Extractive industries (2012):
    - Number of firms: 3,533
    - Total assets: 875,250
    - Total equity: 448,101
    - Total Pre-tax Profit: 207,451
    - Total Tax: 103,813
    - Equity-to-Assets ratio (pct): 51.2
    - ACE rate (pct): 2.4
    - CIT rate (pct): 27.1
    - ACE base: 448,101
    - Notional ACE = base*rate: 11,064
    - ACE used: 10,497
    - ACE used (pct pre-tax profit): 5.1
  - Manufacturing (2012):
    - Number of firms: 209,142
    - Total assets: 5,339,895
    - Total equity: 2,572,378
    - Total Pre-tax Profit: 442,000
    - Total Tax: 101,354
    - Equity-to-Assets ratio (pct): 48.2
    - ACE rate (pct): 2.2
    - CIT rate (pct): 25.3
    - ACE base: 2,572,378
    - Notional ACE = base*rate: 57,632
    - ACE used: 52,937
    - ACE used (pct pre-tax profit): 12.0
  - Services (2012):
    - Number of firms: 876,407
    - Total assets: 9,720,080
    - Total equity: 3,892,087
    - Total Pre-tax Profit: 700,579
    - Total Tax: 175,752
    - Equity-to-Assets ratio (pct): 40.0
    - ACE rate (pct): 2.4
    - CIT rate (pct): 26.8
    - ACE base: 3,892,087
    - Notional ACE = base*rate: 98,230
    - ACE used: 84,320
    - ACE used (pct pre-tax profit): 12.0
  - Regular banks (2012):
    - Number of firms: 7,765
    - Total assets: 17,123,715
    - Total equity: 1,414,992
    - Total Pre-tax Profit: 165,400
    - Total Tax: 50,983
    - Equity-to-Assets ratio (pct): 8.3
    - ACE rate (pct): 1.6
    - CIT rate (pct): 29.7
    - ACE base: 1,414,992
    - Notional ACE = base*rate: 23,938
    - ACE used: 23,669
    - ACE used (pct pre-tax profit): 14.3
    - ACE2 base: 909,437
    - ACE2 Notional ACE = base*rate: 15,972
    - ACE2 used: 15,832
    - ACE2 used (pct pre-tax profit): 9.6
  - Insurance companies (2012):
    - Number of firms: 2,406
    - Total assets: 12,592,165
    - Total equity: 1,436,287
    - Total Pre-tax Profit: 199,700
    - Total Tax: 36,438
    - Equity-to-Assets ratio (pct): 11.4
    - ACE rate (pct): 1.9
    - CIT rate (pct): 29.5
    - ACE base: 1,436,287
    - Notional ACE = base*rate: 28,321
    - ACE used: 27,937
    - ACE used (pct pre-tax profit): 14.0
    - ACE2 base: 1,096,089
    - ACE2 Notional ACE = base*rate: 21,798
    - ACE2 used: 21,594
    - ACE2 used (pct pre-tax profit): 10.8
- Selected 2013 comparators (illustrative shifts):
  - Extractive industries (2013): ACE used (pct pre-tax profit) = 7.3 (ACE base 480,439; ACE rate 2.8; Notional ACE 13,385; ACE used 12,642).
  - Manufacturing (2013): ACE used (pct pre-tax profit) = 12.3 (ACE base 2,627,907; ACE rate 2.1; Notional ACE 55,554; ACE used 52,803).
  - Services (2013): ACE used (pct pre-tax profit) = 12.1 (ACE base 4,025,312; ACE rate 2.2; Notional ACE 92,412; ACE used 83,881).
  - Regular banks (2013): ACE used (pct pre-tax profit) = 17.2 (ACE base 1,500,419; ACE rate 1.9; Notional ACE 30,710; ACE used 30,553); ACE2 used (pct pre-tax profit) = 11.8 (ACE2 base 981,099; ACE2 Notional ACE 21,023; ACE2 used 20,936); ACE3 figures include ACE3 base 112,200 with Notional ACE 2,433 and ACE used 2,432 (pct pre-tax profit) = 1.4.
- Interpretation and caveats:
  - Calculations are stylized and represent an upper bound of revenue effects because of data limitations (accounting vs tax accounts; missing participations data for many firms).
  - ACE amounts are capped by pre-tax profit (no refunds assumed); unused ACE carryforward not modeled.
  - Aggregated ACE expressed as percent of pre-tax profit approximates percent change in CIT base and, with unchanged CIT rates, percent change in CIT revenues.

*Annex Figure 6.1 and related annex text from the IMF chapter "TAX POLICY, LEVERAGE AND MACROECONOMIC STABILITY."*

### References

### References

### Theoretical foundations, asset bubbles, and macroeconomic dynamics
- Abreu, D., and M. K. Brunnermeier, 2003, “Bubbles and Crashes,” Econometrica, Vol. 71, No. 1, pp. 173–204.
- Acemoglu, D., V. Carvalho, A. Ozdaglar, and A. Tahbaz-Salehi, 2012, “The Network Origins of Aggregate Fluctuations,” Econometrica, Vol. 80, pp. 1977–2016.
- Bianchi, J., 2011, “Overborrowing and Systemic Externalities in the Business Cycle,” American Economic Review, Vol. 101, pp. 3400–26.
- Jordà, Ò, M. Schularick, and A.M. Taylor, 2013, “When Credit Bites Back,” Journal of Money, Credit and Banking, Vol. 45, pp. 3–28.
- Reinhart, C., and K. Rogoff, 2009, “This Time is Different: Eight Centuries of Financial Folly,” (Oxford and Princeton: Princeton University Press).

### Housing, mortgage taxation, and house-price dynamics
- Andrews, D., 2010, “Real House Prices in OECD Countries: The Role of Demand Shocks and Structural and Policy Factors,” OECD Economics Department Working Papers, No. 831 (Paris: OECD).
- Aregger, N., M. Brown, and E. Ross, 2013, “Transaction Taxes, Capital Gains Taxes and House Prices,” Swiss National Bank Working Paper No. 2 (Zurich).
- Best, M.C. and H.J. Kleven, 2016, “Housing Market Responses to Transaction Taxes: Evidence from Notches and Stimulus in the UK”, Stanford University.
- Capozza, D., R. Green, and P. Hendershott, 1996, “Taxes, Mortgage Borrowing, and House Prices,” Wisconsin-Madison CULER Working Paper No. 96-06 (Madison: University of Wisconsin Center for Urban Land Economic Research).
- Chambers, M., C. Garriga and D.E. Schlagenhauf, 2013, “Did Housing Policies Cause the Postwar Boom in Homeownership?” NBER Working Paper 18821.
- Glaeser, E.L., and J.M. Shapiro, “The Benefits of the Home Mortgage Interest Deduction” in J. Poterba (ed.), Tax Policy and the Economy, Vol. 17, pp. 37-82 (NBER).
- Hendershott, P. and G. Pryce., 2006, “The Sensitivity of Homeowner Leverage to the Deductibility of Home Mortgage Interest”, Journal of Urban Economics, Vol. 60, pp. 50-68.
- Jappelli, T. and L. Pistaferri, 2007, “Do People Respond to Tax Incentives? An analysis of the Italian Reform of the Deductibility of Home Mortgage Interest,” European Economic Review, Vol. 51, pp. 247-271.
- Poterba, J. and T. Sinai, 2011, "Revenue Costs and Incentive Effects of the Mortgage Interest Deduction for Owner-Occupied Housing," National Tax Journal, Vol. 64, pp.531-564.
- Raya, J. M., and A. Kucel, 2016, “Did Housing Taxation Contribute to Increase Riskier Borrowing?” Journal of Real Estate Finance and Economics, Vol. 53, pp. 90–113.
- Van Den Noord, P., 2005, “Tax Incentives and House Price Volatility in the Euro Area: Theory and Evidence,” Economie Internationale, Vol. 101, pp. 29–45.
- Wolswijk, G., 2005, “On Some Fiscal Effects on Mortgage Debt Growth in the EU,” ECB Working Paper No. 526.

### Taxation and corporate finance; debt bias, ACE/CBIT, and corporate equity allowances
- Auerbach, A., M.P. Devereux, and H. Simpson, 2010, “Taxing Corporate Income,” in J. Mirrlees, S. Adam, T. Besley, R. Blundell, S. Bond, R. Chote, M. Gammie, P. Johnson, G. Miles and J. Poterba (eds.), Dimensions of Tax Design, Chapter 9, pp. 837–93 (Oxford: Oxford University Press).
- Aus dem Moore, N., 2015, “Corporate Taxation and Investment: Evidence from the Belgian ACE Reform,” Ruhr Economic Papers, Vol. 0531.
- Boadway, R. and N. Bruce, 1984, “A General Proposition on the Design of a Neutral Business Tax”, Journal of Public Economics 24, pp. 231–39.
- Bond, S.R. and M.P. Devereux, 1995, “On the Design of a Neutral Business Tax Under Uncertainty”, Journal of Public Economics, Vol. 58, pp. 57–71.
- De Mooij, R., 2011, “The Tax Elasticity of Corporate Debt: A Synthesis of Size and Variations,” IMF Working Paper 11/95 (Washington: International Monetary Fund).
- De Mooij, R., and M.P. Devereux, 2011, “An Applied Analysis of ACE and CBIT Reforms in the EU”, International Tax and Public Finance, Vol. 18, pp. 93–120.
- De Mooij, R., 2012, “Tax Biases to Debt Finance: Assessing the Problem, Finding Solutions,” Fiscal Studies, Vol. 33, pp. 489–512.
- De Mooij, R., M. Keen, and M. Orihara, 2014, “Taxation, Bank Leverage, and Financial Crises,” in: R. de Mooij and G. Nicodeme (eds.), Taxation and Regulation of the Financial Sector, CESifo Seminar Series, (MIT Press).
- De Mooij, R., and G. Nicodeme (eds.), 2014, “Taxation and Regulation of the Financial Sector,” CESifo Seminar Series, (MIT Press).
- De Mooij, R., and M. Keen, 2016, “Debt, Taxes, and Banks,” Journal of Money, Credit and Banking, Vol. 48, pp. 5–33.
- Desai., M., and D. Dharmapala, 2015, “Interest Deduction in a Multijurisdictional World,” National Tax Journal, Vol. 68, pp. 563–680.
- Keen, M., and J. King, J., 2002, “The Croatian Profit Tax: An ACE in Practice,” Fiscal Studies, Vol. 23, pp. 401–18.
- Klemm, A., 2007, “Allowances for Corporate Equity in Practice,” CESifo Economic Studies, Vol. 53, pp. 229–262.
- Michielse, G. M.M., R. de Mooij, and C. Van Peteghem, 2014, “Allowances for Corporate Equity,” in G.M.M. Michielse, and V. Thuronyi, (eds.) Tax Design Issues Worldwide, Series on International Taxation, Vol. 51, pp. 121–40.
- Princen, S., 2012, “Taxes do Affect Corporate Financing Decisions: The Case of Belgian ACE,” CESifo Working Paper No. 3713 (Munich).
- Zangari, E., 2014, “Addressing the Debt Bias: A Comparison between the Belgian and the Italian ACE Systems,” Taxation Paper No 44 (Brussels: European Commission).

### Taxation, banks, financial institutions, and regulatory interactions
- Blouin, J., H. Huizinga, L. Laeven, and G. Nicodeme, 2014, “Thin Capitalization Rules and Multinational Firm Capital Structure,” IMF Working Paper 14/12, (Washington: International Monetary Fund).
- Buettner, T., M. Overesch, U. Schreiber, and G. Wamser, 2012, “The Impact of Thin-Capitalization Rules on the Capital Structure of Multinational Firms,” Journal of Public Economics, Vol. 96, pp. 930–38.
- Dagher, J., G. Dell’Ariccia, L. Laeven, L. Ratnovski, and H. Tong, 2016, “Benefits and Costs of Bank Capital,” IMF Staff Discussion Note No. 16 (Washington: International Monetary Fund).
- Devereux, M., N. Johannesen, and J. Vella, 2015, “Can Taxes Tame the Banks? Evidence from the European Bank Levies,” Oxford University Center for Business Taxation, Working Paper 1325.
- Groenewegen, J., R. Mosch and P. Wierts, 2016, “Subsidies and Taxes on Bank Debt”, Mimeo, De Nederlandsche Bank, Amsterdam.
- Gu, G.W, R.A. de Mooij and T. Poghosyan, 2015. “Taxation and Leverage in International Banking” International Tax and Public Finance, Vol. 22, pp. 177-200.
- Hemmelgarn, T., and D. Teichmann, 2014, "Tax reforms and the capital structure of banks," International Tax and Public Finance, Vol. 21, pp. 645–93.
- Heckemeyer, J., and R. De Mooij, forthcoming, "Taxation and Corporate Debt: Are Banks Any Different?" National Tax Journal.
- Schepens, G., 2016, “Taxes and Bank Capital Structure”, Journal of Financial Economics, Vol. 120, pp. 585-600
- Thimann, C., 2014, “How Insurers Differ from Banks: A Primer on Systemic Regulation,” SRC Special Paper No. 3, (London: London School of Economics).

### Macroprudential policy, financial stability, and housing-market tools
- Allen, F., A. Babus, and E. Carletti, 2009, “Financial Crises: Theory and Evidence,” Annual Review of Financial Economics, Vol. 1, pp. 97–116.
- Claessens, S., M. Keen, and C. Pazarbasioglu, 2010, “Financial Sector Taxation: The IMF’s Report to the G-20 and Background Material” (Washington: International Monetary Fund).
- Claessens, 2016, “An Overview of Macroprudential Policy Tools,” IMF Working Paper 14/214 (Washington: International Monetary Fund).
- Crowe, C., G. Dell’Ariccia, D. Igan, and P. Rabanal, 2013, “How to Deal with Real Estate Booms: Lessons from Country Experiences,” Journal of Financial Stability, Vol. 9, pp. 300–19.
- Darbar, S., and X. Wu., 2015, “Experiences with Macroprudential Policy-Five Case Studies,” IMF Working Paper 15/123 (Washington: International Monetary Fund).
- He, D., 2014, “The Effects of Macroprudential Policies on Housing Market Risks: Evidence from Hong Kong,” Banque de France, Financial Stability Review No. 18, April (Paris).
- Kuttner, K. and I. Shim, 2013, “Can Non-Interest Rate Policies Stabilize Housing Markets? Evidence from a Panel of 57 Economies.” BIS Working Paper No. 433 (Basel, Switzerland).
- Lim, C., F. Columba, A. Costa, P. Kongsamut, A. Otani, M. Saiyid, T. Wezel, and X. Wu, 2011, “Macroprudential Policy: What Instruments and How to Use Them?” IMF Working Paper 11/238 (Washington: International Monetary Fund).
- Cerutti, E., S. Claessens, and L. Laeven, forthcoming, “The Use and Effectiveness of Macroprudential Policies: New Evidence,” Journal of Financial Stability.
- Cerutti, E., J. Dagher, and G. Dell’Ariccia, 2015, “Housing Finance and Real-Estate Booms: A Cross-Country Perspective”, IMF Staff Discussion Note 15/12 (Washington: International Monetary Fund).

### Empirical methods, data sources, and cross-country studies
- Altman, E.I., 1968, “Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy,” Journal of Finance, Vol. 23, pp. 589–609.
- Blundell, R. and S. Bond, 1998, “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models,” Journal of Econometrics, Vol. 87, pp. 115–43.
- Feld, L. P., J.H. Heckemeyer, and M. Overesch, 2013, “Capital Structure Choice and Company Taxation—A Meta Study,” Journal of Banking and Finance, Vol. 37, pp. 2850–66.
- Laeven, Luc, and Fabian Valencia, 2012, “Systemic Banking Crises Database: An Update,” IMF Working Paper 12/163 (Washington: International Monetary Fund).
- Mian, A. and A. Sufi, 2010, “Household Leverage and the Recession of 2007–09,” IMF Economic Review, Vol. 58, pp. 74-117.
- Sialm, C., 2009, “Tax Changes and Asset Pricing,” American Economic Review, Vol. 99, pp. 1356–83.
- Sutherland, D., and P. Hoeller, 2012, “Debt and Macroeconomic Stability: An Overview of the Literature and Some Empirics,” OECD Economics Department Working Papers 1006 (Paris).

### Institutional reports, databases, and policy reviews
- Bank for International Settlements (BIS), 2016, “86th Annual Report 2015/15” (Basel, Switzerland)
- European Commission (EC), 2012, “Taxation of Housing,” in Quarterly Report on the Euro Area No. 4. (Brussels).
- European Commission (EC), 2015, “Experiences with Cash-flow Taxation and Prospects”, Taxation Papers, No. 55 (Brussels).
- European Commission (EC), 2016a, “Council Directive: laying down rules against tax avoidance practices that directly affect the functioning of the internal market”, COM (2016) final.
- European Commission (EC), 2016b, “Taxes in Europe Database”.
- Financial Stability Board (FSB), IMF, OECD, BIS, IOSCO and WBG, 2015, “Corporate Funding Structures and Incentives”, Final Report for the G20.
- Financial Stability Board (FSB), 2016: “Thematic Review on the Implementation of the FSB Policy Framework for Shadow Banking Entities - Peer Review Report”, Financial Stability Board, Basel.
- International Monetary Fund, multiple reports and working papers including:
  - 2009, “Debt Bias and Other Distortions: Crisis-Related Issues in Tax Policy,” June (Washington).
  - 2010, “A Fair and Substantial Contribution by the Financial Sector” Report for the G20.
  - 2011, “Durable Financial Stability: Getting There from Here”, Global Financial Stability Report, Chapter 3, pp. 111-157 (Washington).
  - 2012, “Fiscal Regimes for Extractive Industries—Design and Implementation”, IMF Policy Paper (Washington).
  - 2013, “Key Aspects of Macroprudential Policy—Background Paper,” (Washington).
  - 2014a, “Spillovers in International Corporation Taxation” IMF Policy Paper, May 2014 (Washington).
  - 2014b, “Shadow Banking Around the Globe: How Large and How Risky?” Global Financial Stability Report, October 2014 (Washington).
  - 2015, “Housing Recoveries: Cluster Report on Denmark, Ireland, Kingdom of The Netherlands—The Netherlands, and Spain,” IMF Country Report No. 15/1. (Washington).
  - 2016a, “Fiscal Monitor: Debt: Use it Wisely,” October (Washington).
  - 2016b, “The Insurance Sector – Trends and Systemic Risk Implications,” Global Financial Stability Report, April 2016 (Washington).
- OECD, 2011, “Housing and the Economy: Policies for Renovation,” Chapter from Going for Growth-2011 (Paris).
- OECD, 2012, OECD Banking Statistics.
- OECD, 2015, Base Erosion and Profit Shifting, Action 4: Limiting Base Erosion Involving Interest Deductions and Other Financial Payments, Final Report.
- OECD, 2016, OECD Tax Database.
- UK Treasury, 2015, “Summer Budget”.

*Source: References (content unit _100716 - References).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2016/_100716.pdf_