Example 1: the United Kingdom

12. The U.K. government plans to give the Financial Policy Committee (FPC) control over the CCB. The responsibility of the FPC relates to the identification, monitoring and taking action to remove or reduce systemic risks.²⁰ The U.K. treasury issued a consultation document in September 2012 on the FPC’s powers, including the responsibility for decisions on the CCB.²¹ The CCB’s objective is to build loss absorption capacity of the financial sector, dampen credit cycles and boost medium term GDP growth. The FPC would raise the CCB when threats to financial stability emerge and reduce it when threats to resilience recede (with decision taking effect immediately) taking into account that in periods of acute uncertainty, banks may find it hard to fund themselves at lower capital ratios. The buffer would apply to all U.K. incorporated banks, building societies and large investment firms (broker dealers). The FPC would monitor the extent to which the introduction of the CCB would result in “leakages” to the non-regulated sector and might recommend to the Treasury to expand the set of institutions to which the CCB applies. Moreover, the FPC would be given a power of direction over sectoral capital requirements which would be a more targeted tool than the CCB.

13. The FPC will identify and publish a list of core indicators of systemic risk. While the FPC would review them regularly to provide guidance about the use of the CCB, judgment will play an important role in all FPC decisions given the complexity of the financial system and its tendency to evolve over time. The choice of indicators will evolve over time as the FPC learns from experience and as new research is undertaken. To support its judgment when to use the CCB but also other macroprudential instruments, the FPC will review the core indicators set which will include: (i) measures of balance sheet stretch within the financial system and borrowers; (ii) measures of terms and conditions in financial markets; (iii) indicators of risks concentrated in particular sectors; and (iv) indicators of changing patterns in the distribution of risks across financial institutions, households, and corporations.²² The likelihood of activation of the CCB will depend on whether the indicators convey a homogenous picture of systemic risk. The policy decision and Directions issued to the PRA or the FCA will be published in the quarterly FPC Record while explanation of the background to those decisions including benefits and costs of its actions will be published in the Financial Stability Report.

Example 2: Switzerland

14. Switzerland has recently activated a CCB that focuses on a specific segment of the credit market—the residential mortgage market.²³ The Swiss CCB is a time-varying capital requirement that can be implemented on a broad basis or can be targeted at specific segments of the credit market. The CCB has to be in the form of common equity Tier 1 capital amounting to a maximum of 2.5 percent of the risk weighted assets in Switzerland. The early introduction of the CCB framework in July 2012 was justified by concerns about the risks of imbalances in the mortgage and real estate markets. The decision to activate the CCB in February 2013 requires banks to comply with the additional capital requirement of one percent of risk weighted direct and indirect mortgage-backed positions secured by residential property by end of September 2013 (seven months following the announcement). The authorities expect that the CCB should dampen mortgage lending (and as a consequence house prices) while at the same time build resilience of the banking sector against a correction of the imbalances in the mortgage and real estate markets.

15. The Swiss National Bank plays a key role in the process leading to a CCB decision. It conducts a regular assessment of the mortgage and real estate markets to determine whether the CCB should be activated, adjusted or released based on an approach of “guided discretion.” It proposes the level at which the CCB should be set as well as the implementation period of the CCB. The central bank consults with the banking sector supervisor (FINMA) regarding its view before issuing a proposal to the Federal Council.^{24, 25} The Federal Council takes the final decision. The central bank communicates and motivates its proposal after the Federal Council has announced its decision. FINMA supervises the implementation of the CCB at the individual bank level.

16. The central bank’s decision heavily draws on indicators of domestic mortgage volumes and domestic real estate prices. If the same indicators suggest heterogeneous conclusions more discretion is used. Additional quantitative and qualitative information are used to ensure that the decision is based on a comprehensive view of developments in the domestic mortgage and real estate markets. The level of the CCB and the implementation period (that will vary between three and 12 months) each reflect the degree of imbalances. In addition to key indicators of the real estate and mortgage market, higher-frequency information and judgment will play an important role in taking the decision to deactivate the buffer.

17. To illustrate how the CCB could be used in practice, we consider what a small set of indicators might have signaled to authorities in Ireland and Spain in the period before and during the crisis. These indicators are chosen for illustration purposes. This analysis is not a comprehensive examination of all available information that might be considered. Moreover, it is worth noting that simulation results depend on and can change substantially with different starting dates of the credit gap calculation. In both cases we use 1997 as the starting date.

Example 3: Hypothetical paths of the CCB for Ireland

18. Using the BCBS formula, we calculate the hypothetical path of the CCB (relative to risk weighted assets) for Irish banks in the following way:

• We calculate the credit-to-GDP ratio as the ratio of private sector credit and nominal GDP on a quarterly frequency. Private sector credit includes credit to households, non-financial companies, insurance companies, pension funds and other financial intermediaries. We do not use the broadest measure of credit that would include domestic, international, bank-originated and market based sources of credit (household and non-financial corporate debt) since the data prior 2002 are not publically available.

• We use a one-sided HP filter to extract the trend and calculate the credit gap.

• We assume that the CCB would be activated for exposures to counterparties in Ireland when the credit gap measure breaches the lower threshold set to 2 percent; we set the upper threshold at 10 percent and assume that the CCB increases linearly between zero and 2.5 percent with the value of the credit gap between 2 and 10 percent.

• We use house prices and a principal component of the four biggest banks’ stock prices as two indicators for the release of the CCB; once an exit signal is received the CCB is fully released.²⁶

• The CCB for banks in Ireland is calculated as a weighted average of the buffers applied in the jurisdictions to which Irish banks have exposures.

  • ➢ We assume that the only exposure of Irish banks is to U.K. residents as the majority of the domestic banks’ exposures are to the U.K. (about 90 percent of all foreign credit exposures and about 75 percent of all foreign claims as of June 2012).²⁷ The share of the U.K. credit exposure in overall credit exposure is assumed to be 30 percent to reflect the geographical distribution of domestic banks’ credit exposure as of June 2012.

  • ➢ We take the Bank of England credit gap in order to calculate the CCB for U.K. exposures.

19. Figure 1 shows the hypothetical evolution of the CCB since 1997. If the BCBS proposal had been in place at the time, the additional buffer would have built up to its maximum three years ahead of the financial crisis.²⁸ This suggests that the credit gap measure is a good indicator for the activation of the CCB.²⁹ A simple calculation based on a 2008 Tier 1 capital calculation, shows that the additional buffer would have saved up to a quarter of the fiscal costs of the financial crisis for the Irish authorities.³⁰ Moreover, raising additional capital could have been difficult for some banks, which would then have discouraged them from lending. The likely outcome would have been to lessen the supply of loans which might have mitigated the housing price boom.

View Full Size

Ireland: Simulated Countercyclical Capital Buffer

(In percent of risk weighted assets)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.Note: 1/ The CCB calculated assuming Irish banks do not have any international exposure.2/ The CCB calculated assuming Irish banks have international exposure.

• Download Figure
• Download figure as PowerPoint slide

20. While the same indicator performs poorly for the release decision, extreme house price and stock price growth rates could have been used as indicator for the release of the CCB. The fact that the credit-to-GDP ratio remained above its trend during the crisis (Figure 2) is due to the fact that GDP experienced a period of severe contraction and credit growth was lagging behind GDP growth. This illustrates the point discussed above that indicators other than credit-to-GDP should be used for the decision to release the buffer. Extreme (two standard deviation from the mean) house price growth rates and the extreme difference in the principal component of banks’ stock prices signaled the release of the CCB a year before (stock prices, Figure 3) and at a time (house prices, Figure 4) when the crisis hit the Irish economy even as the credit gap continued to rise.

View Full Size

Ireland: Credit-to-GDP Gap

(In percent)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Ireland: Principal Component of Banks’ Stock Prices

(YoY difference)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Ireland: House Prices

(In percent, YoY growth rate)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

Example 4: Hypothetical paths of the CCB for Spain

21. We repeat the same exercise for Spain with the following assumptions:

• The CCB for banks in Spain is calculated as a weighted average of the buffers applied in the jurisdictions to which Spanish banks have exposures.

  • ➢ We assume that the only exposure of Spanish banks is to the U.K. (43 percent), the United States (U.S.) (22 percent), Brazil (19 percent) and Mexico (15 percent) as about 65 percent of banks’ foreign exposures (as of March 2012) correspond to those countries. The share of the foreign exposure in overall banks’ exposure is assumed to be 45 percent to reflect the geographical distribution of domestic banks’ exposure as of March 2012.

• The following credit measures were used in the calculation of the credit-to-GDP ratio:

  • ➢ Spain: the overall debt of nonfinancial firms and household;

  • ➢ Brazil: credit to the private sector;

  • ➢ Mexico: credit to private and public sector;

  • ➢ The U.S.: the overall debt of nonfinancial firms and household; and

  • ➢ The U.K.: the Bank of England credit gap was taken.

• We use nonsubsidized house price and a principal component of the four largest banks’ stock prices as indicators for the CCB release phase. For illustration, we also use spreads on the government bond as a high-frequency indicator (the difference in yields on Spanish and German 10 year government bond).

22. Figure 5 shows the hypothetical evolution of the CCB in Spain. If the CCB framework had been in place the additional buffer would have built up three years ahead of the financial crisis.³¹ The additional capital of 2 percent prior to the crisis would have saved almost all fiscal costs of the financial crisis for the Spanish authorities.³² Moreover, the additional capital is about 70 percent larger than the estimated €24 billion in dynamic loan-loss provisioning.

View Full Size

Spain: Simulated Countercyclical Capital Buffer

(In percent of risk weighted assets)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.Note: 1/ The CCB calculated assuming Spanish banks do not have any international exposure.2/ The CCB calculated assuming Spanish banks have international exposure.

• Download Figure
• Download figure as PowerPoint slide

23. While in the Spanish case the credit gap might be a good indicator for both activating and releasing the CCB, it is expected the authorities look at many other indicators before releasing the buffer. Figures 6-9 suggest that extreme (two standard deviation from the mean) house price, stock price and spread growth rates are useful indicators for the release of the CCB. All three indicators signal an exit decision before or at the time the crisis hit the Spanish economy.

View Full Size

Spain: Credit-to-GDP Gap

(In percent)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Spain: House Prices

(In percent, YoY growth rate)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Spain: Spread on Government Bond

(In percent, YoY difference)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Spain: Principal Component of Banks’ Stock Prices

(YoY difference)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

Box.

How Does the Reciprocity Principle Work?

Under the BCBS agreement, each jurisdiction determines the CCB for credit exposures to counterparties in its country. According to the reciprocity principle, however, the home supervisor ensures that for an internationally active bank domiciled in its jurisdiction the CCB is calculated on a consolidated basis, according to the geographic location of its exposures. In other words, the CCB for internationally active banks will be a weighted average of the CCBs that are being applied in jurisdictions to which the bank has an exposure. The home supervisor is not allowed to impose a buffer requirement for credit exposures to a foreign country that is below the requirement set by the host supervisor. While the home supervisor can set a higher buffer requirement for foreign exposures, the mandatory reciprocity principle would not apply to the amount of the buffer above 2.5 percent. Reciprocity with respect to add-on buffers higher than 2.5 percent would be voluntary.

To illustrate how reciprocity works, assume that Country 1 and Country 2 have set the CCB on domestic exposures of their banks at 2.5 percent and 1 percent, respectively (using the credit gap guide for example). In each country we assume that there are two banks: one with domestic exposures only (Bank 1A in Country 1 and Bank 2A in Country 2) and the other one with exposures to the other country (Bank 1B and Bank 2B). In Country 1 the internationally active bank has 40 percent of its credit exposures to Country 2 counterparties and in Country 2, the share of cross-border exposures of the internationally active bank is 70 percent. According to the reciprocity principle, any loan to Country 1 counterparties, irrespective of the location of the bank making the loan will attract a CCB requirement of 2.5 percent (red arrow, Table 1). While a bank with domestic credit exposures only is subject to the full amount of the CCB determined by their supervisor, the CCB of internationally active banks will reflect the structure of their domestic and foreign exposure. For example, the CCB of Bank 1B would be calculated as: 0.6*2.5 percent + 0.4*1 percent = 1.9 percent.

View Full Size

Example of CCB Calculation

(In percent)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

Table 1.

Number of Countries with Sectoral Macroprudential Tools

Source: IMF staff calculation. Note: Numbers in ( ) shows the proportion of countries with a specific instrument among the sample.

Table 1.

Number of Countries with Sectoral Macroprudential Tools

	Limits on LTV Ratio	Caps on DTI Ratio	Limits on LTV and DTI ratios	Sectoral Capital Requirements	One tool	Any two tools	All three tools
Number of Countries (Total = 46)	24 (52 percent)	14 (30 percent)	14 (30 percent)	23 (50 percent)	36 (78 percent)	18 (39 percent)	7 (15 percent)

Source: IMF staff calculation. Note: Numbers in ( ) shows the proportion of countries with a specific instrument among the sample.

View Table

Table 1.

Number of Countries with Sectoral Macroprudential Tools

	Limits on LTV Ratio	Caps on DTI Ratio	Limits on LTV and DTI ratios	Sectoral Capital Requirements	One tool	Any two tools	All three tools
Number of Countries (Total = 46)	24 (52 percent)	14 (30 percent)	14 (30 percent)	23 (50 percent)	36 (78 percent)	18 (39 percent)	7 (15 percent)

Source: IMF staff calculation. Note: Numbers in ( ) shows the proportion of countries with a specific instrument among the sample.

There is a similar reciprocity principle in the Recommendation of the European Systemic Risk Board on lending in foreign currencies in the European Union. According to the principle, the home supervisor of financial institutions are recommended to impose measures addressing foreign currency lending at least as stringent as the measures in force in the host jurisdiction where they operate through provision of cross-border services or through branches. This recommendation applies only to foreign currency loans granted to borrowers domiciled in the host jurisdictions.

Liquidity measures—the case of Korea

43. In Korea, banks rapidly increased short-term non-core FX borrowing from 2005, creating sizable FX mismatches.⁵⁵ This surge suddenly stopped, when the international wholesale funding market froze and both domestic banks and foreign banks’ branches were unable to roll over their maturing short-term external liabilities after the Lehman Brothers bankruptcy. The rollover ratio of short-term external debt fell rapidly right after the financial crisis, from 99.8 percent in September 2008 to 39.9 percent in October 2008 (Lee, 2012). The Korean won depreciated rapidly and the CDS premium on government bonds rose to become much higher.

44. Korea adopted the Macroprudential Stability Levy as a macroprudential tool as of August 1, 2011. The levy is imposed in proportion to each bank’s marginal contribution to systemic risk, that is, on banks’ daily average balance of non-deposit foreign currency liabilities of maturity up to a year.⁵⁶ Currently, 19 domestic banks and 38 foreign banks’ branches are subject to the levy.⁵⁷ The rate varies from 2 to 20 basis points, depending on the maturity of the debt instrument, and can be adjusted discretionally to discourage highly volatile short-term foreign currency funding. The proceeds from the levy are accumulated as a part of the Foreign Exchange Stabilization Fund to provide liquidity to banks at times of financial stress, complementing deposit insurance.

45. Korea also implemented caps on the loan-to-deposits ratio (2012) to shift banks’ funding structure away from wholesale funding, and ceilings on banks’ FX derivative positions (2010, 2011, and 2012). Net FX derivative positions were limited to 50 percent of capital for domestic banks and 250 percent for foreign banks’ branches in June 2010, and the limits were lowered to 40 percent and 200 percent in June 2011 and were cut again to 30 percent and 150 percent in December 2012. Combining them with the levy, the authorities target both the source and the costs of the excessive dependence on short-term non-core FX borrowings, and encourage long-term and stable sources of funding.

46. These measures appear to have been effective in curbing banks’ reliance on short-term FX funding and in reducing vulnerabilities from FX mismatches and exchange rate volatility (IMF, 2012c and 2013; Ree and others, 2012; Bruno and Shin, 2012).⁵⁸ Banks’ short-term net external debt, including of foreign banks’ branches, declined steadily from US$65 billion in June 2010 to US$43 billion in June 2012, and the short-term external debt ratio has continuously decreased to 50.7 percent in the second quarter of 2012 after peaking at 72.6 percent of total bank’s external debt in the third quarter of 2008.⁵⁹ The sensitivity of capital inflows to global conditions decreased in the period following the imposition of the levy, relative to a comparison group of countries (Bruno and Shin, 2012). Rollover risks for domestic banks have diminished, as residual maturities of their external debt increased (IMF, 2012c). The sensitivity of exchange rate volatility to changes in the VIX also declined substantially since the financial crisis, reflecting lower FX liquidity mismatches (Ree and others, 2012).

View Full Size

Korea: Effectiveness of Macroprudential Stability Levy

(In percent and US$ billion)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Sources: Bank of Korea and IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

Quantitative liquidity requirements—the case of New Zealand

47. Similar to Korea, New Zealand’s banks depended on short-term cross-border funding to provide credit.⁶⁰ Given low national saving, banks have relied on external debt to fund private sector credit. Gross external debt exceeded 130 percent of GDP in 2009, and while New Zealand’s short-term external debt declined during 2009, it remained high at almost 60 percent of GDP at end-2009.

48. New Zealand banks also experienced some difficulty rolling over their short-term debt after the collapse of Lehman Brothers. The Reserve Bank of New Zealand (RBNZ) introduced broader domestic market liquidity measures and the Term Auction Facility, while the government provided domestic and wholesale funding guarantees to retain access to international markets. Parent banks in Australia also provided funding to their subsidiaries in New Zealand.⁶¹

49. In October 2009 the RBNZ introduced new quantity-based measures to increase banks’ liquidity and reduce reliance on short-term cross-border funding, and the measures became effective from April 2010.⁶² The one-year CFR is core funding divided by total loans and advances, where core funding is defined as all funding with a residual maturity of longer than one year, plus Tier 1 capital and a weighted sum of certain short-term funding instruments.⁶³ A minimum on this ratio aims to ensure that banks hold sufficient retail and long-term wholesale funding, to reduce the vulnerability of the banking sector to a severe liquidity shock. For the initial implementation in April 2010, the minimum CFR was set at 65 percent of total loans and advances, increasing to 70 percent from July 2011 and 75 percent from January 2013 (Figure 12).

View Full Size

New Zealand: Effectiveness of Core Funding Ratios^{1, 2}

(In percent)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Source: Bank of England (2011).Note: ^1/ The minimum CFR was first announced in May 2008.^2/ It was originally set to increase to 75 percent in July 2012, but in November 2011, the increase was deferred by six months to January 2013.

• Download Figure
• Download figure as PowerPoint slide

50. Since the onset of the global financial crisis in 2008, New Zealand banks have almost doubled their holdings of liquid assets and increased retail and long-term wholesale funding. Even though it is not clear whether this was because of the RBNZ’s plans to introduce liquidity requirements, the uncertain and volatile environment, or rating agencies putting pressure on New Zealand banks to reduce their exposure to rollover risk, a comparison with some other countries suggests that the liquidity measure may have played a role. Since end-2007, New Zealand’s short-term external debt declined by about 10 percent of GDP, whereas short-term external debt in percent of GDP rose for many other countries, with the notable exception of Korea and U.K..⁶⁴

51. The impact of the new measure is expected to be stronger during booms, when banks tend to resort to short-term foreign currency funding markets to support credit expansion. The introduction of a minimum CFR will drive banks to either compete for more retail funding that affect their lending margins, or borrow in wholesale markets for terms longer than one year. During periods of rapid credit expansion, New Zealand banks will not have the same ability to expand domestic credit. To satisfy growing credit demand, banks will need to find funding mostly from customer deposits and longer-term markets. As a result, lending rates should move higher, reducing excessive credit growth.⁶⁵

Debt Bias

The issue

52. Corporate income tax systems generally encourage the use of debt rather than equity finance. This is because interest paid is allowed as a deduction in calculating taxable profits, but the return to equity is not. This favorable tax treatment of debt at the corporate level is for the most part not offset by taxes at the personal level, so that corporate taxes typically create a ‘debt bias.’ A large empirical literature suggests that the impact is sizable: it suggests that, on average, a 10 percentage point lower corporate tax rate—a reduction, for example, from 30 to 20 percent—reduces the debt-asset ratio of nonfinancial companies by between 1.7 and 2.8 percentage points. There is some indication that this effect has increased over time (De Mooij, 2011).

53. This debt bias leads, on average, to more highly leveraged banks. The literature on debt bias, including the empirical literature just mentioned, deals almost exclusively with nonfinancial companies. This neglect is unfortunate, since while excess leverage is certainly cause for concern in nonfinancial companies, any excess leverage of financial companies is likely to be especially troubling from the financial stability perspective. There is a basic tension/inconsistency between regulatory measures intended to induce banks to hold more capital than they otherwise would and tax incentives for them to hold less. Banks might certainly be expected to respond to debt bias differently from other types of firms, given the regulatory capital requirements they face. However, banks typically hold buffers of equity beyond the regulatory requirements, which leaves scope for tax effects on leverage. Indeed it turns out that, empirically, the average tax response by banks is about as large as it is for nonfinancial firms: the long-run impact of a 10 percentage point change in the corporate tax rate on the leverage ratio of banks is estimated between 1.5 and 2.7 percent.⁶⁸ This means, for instance, that eliminating the bias to debt finance created by a 25 percent corporate tax rate (which, as will be seen, does not necessarily mean eliminating the corporate tax itself) might increase banks’ capital in the long run by at least 3.75 percent of their assets: an increase of more than 30 percent over current levels (Keen and De Mooji, 2012).

54. The social cost of debt bias in the banking sector is potentially very large. The evidence is that larger banks, which hold smaller equity buffers, are less sensitive to tax. This is important because they hold a very large proportion of all banks’ assets. It does not mean, however, that debt bias is unimportant for the banking sector: even small changes in the leverage of very large banks could have a large impact on the likelihood of their distress or failure, and hence on the likelihood of financial crisis. There is indeed evidence that higher debt bias is associated with significantly higher aggregate bank leverage, and that this in turn is associated with a significantly greater chance of crisis. This implies that tax bias makes crises more likely and, conversely, that the welfare gains from policies to alleviate that bias can be substantial. This is not to say, of course, that debt bias provides anything like a complete explanation of the incidence of financial crises, many of which have occurred in countries with low statutory corporate tax rates and to that extent relatively little debt bias. Nonetheless, the potential impact is sizable. At, for instance, bank leverage ratios as high as they were in some crisis countries in 2008, eliminating debt bias could avoid a loss in expected output of between 1.1 and 11.9 percent of GDP (cumulative over a four-year period) (De Mooij and others, 2013). Looking forward, the significance of the potential gain from reducing debt bias would need to be assessed relative to the strengthened capital standards now envisaged.

Possible policy responses

55. One way to eliminate debt bias—establishing tax neutrality between debt and equity finance—is by adopting a ‘comprehensive business income tax’ (CBIT), which denies interest deductions. The base broadening this implies would also allow the statutory corporate tax rate to be reduced as part of a revenue-neutral reform. However, the CBIT has serious drawbacks: it (i) increases the cost of capital on debt-financed investment (unless compensating measures are taken); (ii) raises significant problems with the taxation of banks (not least in terms of public perception), which would become effectively untaxed on their net interest income; and (iii) significantly distorts international financial transactions unless broadly adopted. Reflecting these difficulties, no country has ever adopted the CBIT. More limited restrictions to the interest deductibility have become more popular recently in the form of thin capitalization rules. These, however, do not fully address debt bias, often do not apply to banks, and create their own problems (including, for instance, in failing to recognize sectors’ and companies’ differing circumstances).

56. More promising is the introduction of an ‘Allowance for Corporate Equity’ (ACE) form of corporate tax.⁶⁹ This would provide a deduction for a notional return on equity, in principle—assuming that firms have complete assurance that they will receive the full value of the deduction—at a risk-free rate of return on capital, but in practice at something like the yield on government bonds. The base to which this rate would be applied is the book value of equity, minus equity participations in other firms (to avoid duplication of tax relief). The ACE is neutral not only with respect to firm’s debt-equity choice, but also with respect to its investment decisions (because tax is ultimately paid only on returns to investment in excess of the normal return). It also makes irrelevant to the firm’s decision both the tax rules for depreciation and the rate of inflation. And there is now ample experience from countries that have or had an ACE or some variant, including in Austria, Belgium, Brazil, Croatia, Italy, and Latvia. These schemes have encountered no particular practical difficulty and, where this has been studied, the available evidence is that they have indeed reduced leverage ratios (see in particular Klemm (2007), who surveys experience in a range of countries, and Princen (2012), who finds that the ACE led to a significant reduction in leverage ratios in Belgium).⁷⁰The Fund has in a number of cases recommended adoption of an ACE in its TA work, and this has also been also supported, for instance, in recent U.K. Article IV reports.

57. One potential concern with adoption of an ACE is its revenue cost—but this can be substantially mitigated. Tentative calculations suggest that an ACE for advanced countries would involve an average budgetary cost of 0.5 percent of GDP (Mooij, 2011). Of course, this could be offset by other tax measures to leave total revenue unchanged; and the losses will be smaller to the extent that the reform induces more investment and thus broadens the overall tax base. In thinking of compensatory fiscal measures, it is important to note that the beneficiaries of the ACE are likely to be not only (or even mainly) the owners of equity, but also employees who see their productivity and wages increase. Importantly, the revenue loss from introducing an ACE can also be mitigated by judicious design. There are two main options. One is to apply the ACE only to new investment: this does not reduce the economic benefits of the reform, since for existing capital the equity deduction is simply a windfall gain. This was the route chosen by Italy in its recent introduction of an ACE. The main challenge with this approach is to minimize abusive transformations of old into new capital, but the Italian experience suggests that this risk can be managed. A second and in some respects more radical option is to restrict the ACE to banks (perhaps again only for new equity), for which the social costs of excessive leverage are expected to be largest. A concern with this is that any sector-specific tax treatment brings its own risks of avoidance. However, banking is a well-defined and intensely regulated sector, where special provisions are common. Though distinctly second best in tax policy terms, it may thus be worth considering an ACE for banks—supporting the buildup of capital—as an intermediate step toward adoption of an ACE more widely.

Financial Sector Taxation

Corrective taxes

58. If there are adverse externalities associated with some forms of debt finance, a case can be made for going beyond tax neutrality to actively discourage their use. Here the potential interplay between tax and regulatory policies comes to the fore. But while there has been some debate as to the relative merits of taxation and regulation in addressing financial market failures, this deep issue remains largely unresolved—intellectually at least (Keen, 2011a and 2011b).

59. Bank taxes can help correct failures in financial markets, acting as a Pigovian tax.⁷¹ The IMF 2010 study for the G20 proposed to address systemic externalities associated with excessive risk taking by banks with a Financial Stability Contribution (FSC), linked to and financing a credible and effective resolution mechanism. A key purpose of the FSC would be to pay for the direct fiscal cost of any future government support to the sector and, in particular, for the resolution of weak institutions. In its simplest form, the base of the FSC would be uninsured bank borrowing and the charge would be paid by all financial institutions, initially at a flat rate; in broad terms, such a charge can be seen as partly offsetting the debt bias discussed above. The suggestion was that the FSC be subsequently refined to reflect individual institutions’ riskiness and contributions to systemic risk—such as those related to size, interconnectedness, and substitutability—and variations in overall risk over time.

60. Around a dozen European countries—including France, Germany, and the U.K. —have adopted bank taxes that resemble an FSC as a way to support financial stability. These vary quite widely in rate, base, and in whether the proceeds feed a fund or go to general revenue. Assessment of these charges is only now beginning, but some preliminary work suggests a (perhaps surprisingly) large consequent increase in capital ratios (Devereux, Johannesen and Vella, 2013), and it is notable too that the U.K. levy has consistently raised less than expected, perhaps suggesting larger behavioral impact than anticipated. These taxes are to a large extent a work in progress, and a variety of technical issues have arisen, for instance, in terms of novel measures needed to avoid double taxation. It remains to be seen whether divergences between these taxes will create sufficiently large distortions to warrant closer convergence in their design.

Financial sector taxes less directly related to financial stability

61. There is a case for special taxes to offset distortions associated with the VAT exemption of margin-based financial services.⁷² The first-best response to this exemption would be to levy the VAT on financial services—and it is now understood how this could, in principle, be done.⁷³ Progress in this area is slow, however, and the Financial Activities Tax (FAT) that is also proposed in the IMF’s report to the G20 is a second-best way to alleviate the distortion that exemption creates. It does this by imposing a charge on the sum of financial institutions’ profits and remuneration—which is its value added.⁷⁴ Thus, to some degree, it offsets the distortions caused by the exclusion of most financial services from the VAT.⁷⁵ Iceland, Israel, and Quebec have taxes of this type, and France and Denmark have imperfect variants. The FAT is inferior to full taxation under the VAT in that, without some special arrangements, business use of financial services would continue to be subject to tax; issues also arise over the treatment of exported financial services which, in principle, should be removed from the tax base. Importantly for present purposes, however, the primary purpose of the FAT is not to address financial stability issues, but rather to mitigate a pre-existing tax distortion that impedes the smooth operation of the financial sector.

62. The Financial Transaction Tax (FTT) has many advocates as a way to address some financial market imperfections, but few claim it addresses basic issues of financial stability. Several emerging economies (Colombia, India, Peru, Poland, Ukraine) and advanced countries (Belgium, Finland, Singapore, Sweden, Switzerland, and the U.K.) have forms of FTTs, including stamp duties that apply to a fairly narrow range of transactions. In the EU, there are plans for 11 member states to adopt a much more broad-based form of FTT, under enhanced cooperation, at the start of 2014. France and Italy have adopted forms of FTT in recent months, and have also adopted particular taxes on high frequency trade.

63. FTT’s are generally regarded as inferior to the FSC and the FAT in addressing core market failures in, or raising additional revenue from, financial markets. For instance, advocates of the FTT argue that discouraging short-term trading is desirable, but there is little evidence that short-term trading is a significant source of financial risk or that it was a major factor in the crisis.⁷⁶ Nor is there convincing evidence that FTTs reduce short-term price volatility; in fact, high transaction costs are likely to increase it. Even at a modest rate, an FTT can have significant social costs due to cascading effects (tax levied on tax), increasing costs of capital, encouraging avoidance schemes, and potentially impeding socially worthwhile transactions.⁷⁷ And if revenue is the objective, a basic principle of public finance is that taxing intermediate transactions is a bad way to do this—which is why the world has moved toward a VAT, not to turnover taxes on all sales.

Taxation and Asset Prices

64. Taxes can powerfully affect asset prices. Through their capitalization in asset prices, higher future taxes of many kinds (such as capital gains taxes, stamp duties, or recurrent property taxes) can reduce current asset prices; and the announcement of future tax relief can support asset prices (Jeanne and Korinek, 2010). Countries have indeed used tax measures during the crisis to bolster house prices, for instance, by removing or reducing stamp duties on housing transactions (as in Ireland. The Netherlands and Singapore (as noted in paragraph 36) or extending mortgage interest relief (Ireland). Conversely, Hong Kong and Singapore have recently raised real property taxes in an effort to cool down sharply rising house prices.

65. The effects of tax measures on asset prices can be complex and hard to predict. Lowering taxes on asset returns, for example, will have no effect on equity prices if the marginal investor is tax exempt. Likewise, if higher property taxes finance public expenditures of equal value, house prices should remain unchanged. The price effect of stamp duties or real property taxes will also depend on supply and demand responses and the frequency of market transactions. Evidence on the price effects of tax changes is mixed (and plagued by identification problems, as tax measures are generally introduced along with regulatory changes). For example, historical evidence for the U.S. suggests a significant negative correlation between asset prices and the tax burden on equity securities (Sialm, 2009). Yet, a recent study using Swiss data finds no effect of transaction taxes or capital gains taxes on housing price dynamics (Aregger and others, 2013). The European Central Bank (ECB) has recently noted that transaction taxes appear to have had only a minor impact on asset price bubbles in Europe (ECB, 2003). In Japan, by contrast, increased land taxation immediately preceded (which of course does not mean that it caused) the collapse of prices in the early 1990s.

66. Tax changes might not only affect price levels, but also their rate of change and their volatility. Macroprudential policymakers may care not only about the level of asset prices but also their rate of growth and volatility. The impact of taxation on these aspects of asset pricing are complicated and incompletely understood. For example, if an asset yields its return only as capital gains, increasing the tax on those gains means that its price must rise more rapidly if it is to yield the same after-tax return as other assets. Moreover, transaction taxes can discourage speculative transactions and reduce volatility in prices, but can also lead to lock-in effects, thinner markets, and more price volatility.

67. Favorable tax incentives for owner-occupied housing can make households more vulnerable to shocks. Many countries do not tax (or tax only lightly) imputed rent and capital gains from housing transactions, while providing generous relief for mortgage interest. This can be a source of significant distortion (and revenue loss), as productive capital shifts into the housing sector, and as households are encouraged to borrow against housing assets, either to invest in non-housing wealth or to finance immediate consumption. This can make them more vulnerable to shocks, exacerbating transmission channels within the financial system that can trigger crisis.⁷⁸ Large and unexpected reforms in the income tax treatment of housing can have large implications for home values and the financial positions of households. While this renders reform difficult, the U.K. experience—phasing out mortgage interest relief over many years—shows that it can be done.

68. The purposive use of tax policy to affect asset prices is problematic, but tax-induced distortions should concern macroprudential authorities. One potentially attractive feature of tax measures in relation to asset prices, not usually associated with discretionary fiscal measures, is that implementation lags may be short, in that tax changes are likely to be capitalized as soon as they are announced or at least are perceived as credible. The risk remains, however, that delays may result in measures becoming credible only once the immediate need has passed: in Japan, for instance, the increased land tax came only after stock prices had begun to tumble. Gaps between announcement and implementation can also create distortions: delaying transactions, for instance, in the expectation of reduced transactions taxes. The use of taxes to affect asset prices can also entrench distortions that, from a wider perspective, would be better off withdrawn: increasing transactions taxes to dampen house prices, for instance, can also worsen the performance of labor markets by reducing mobility. Given the difficulty of distinguishing bubbles from asset price movements reflecting fundamentals, the natural focus for tax policy is to ensure neutrality in the treatment of differing assets and forms of income. A macroprudential authority might usefully flag instances where tax non-neutralities risk distorting asset price levels and movements to an extent that impairs financial stability.

An Intermediate Degree of Bank Competition Is Optimal for Financial Stability

69. Bank competition policy has been a focus of much research and policy debate. The reason for this is the special nature of banks. In the non-financial sector, competition policy mainly focuses on efficiency (competitive pricing).⁸⁰ Yet for banks and the financial sector more generally there is another relevant dimension: systemic risk. When the degree of competition adversely affects banks’ risk-taking incentives, this should be internalized by competition policy. Put differently, bank competition policy should have a macroprudential component.

70. The theoretical predictions and empirical results on the link between bank competition, risk-taking, and stability are somewhat ambiguous. But on net they suggest that an intermediate degree of bank competition is optimal, i.e. no excess restrictions but no unbridled competition either.

Literature ⁸¹

71. Much of the theoretical literature warns that competition may increase bank risk-taking. Competition lowers margins and charter value (the discounted stream of profits) of banks, making them more willing to gamble and less able to withstand negative shocks (Marcus, 1984; Chan and others, 1986; Keeley, 1990; Hellman and others, 2000; Matutes and Vives, 2000; Repullo, 2004). Another channel is that competition may force banks to focus on maintaining market share instead of screening existing borrowers (Dell’Ariccia and Marquez, 2006).

72. Yet some papers highlight opposite effects where bank competition lowers interest rates in the economy, making borrowers safer and reducing risk (Boyd and De Nicolo, 2005). The effects can be reconciled in models that show an inverse U-shaped relationship between bank competition and stability. There, introducing competition in monopolistic systems initially increases stability as borrowers become safer, but high competition becomes destabilizing due to the charter value effect (Martinez-Miera and Repullo, 2010).

73. The empirical literature also offers ample evidence that too much bank competition may reduce stability. (In the papers, competition is measured through margins or concentration, and stability is captured as loan quality or probability of failure; see Keeley, 1990; Dick, 2006; Jimemez and others, 2007; Boyd and Runkle, 1993; Beck and others, 2006a and 2006b). There is evidence that competition distorted credit quality also during the recent crisis (Dell’Ariccia and others, 2012). Interestingly, some papers show that the stability effects of lower competition stem not from charter values, but from diversification benefits in large banks (Paroush, 1995; Benston and others, 1995; Craig and Santos, 1997; Beck and others, 2006a and 2006b).

74. Yet as theory predicts, too little competition may compromise bank stability. This is predominantly driven by distortions in large banks. As banks get larger and more diversified, they may increase the risks of their portfolios, or strategically choose to operate at a closer distance to default (Chong, 1991; Hughes and Mester, 1998; De Nicolo, 2000; Boyd and others, 2006). Larger banks also become subject to internal inefficiencies and increased operational risk (Beck and others, 2006; Laeven and Levine, 2007; Cetorelli and others, 2007).⁸² Taken together, the two effects may— similar to the theoretical predictions—lead to an inverse U-shaped effect of bank competition on stability, as suggested in recent papers (Fernandez and Maudos, 2011; Carbó-Valverde and others, 2013).

75. Of course, beyond the ambiguous effect of risk, higher competition is beneficial—in the static sense—as it lowers costs and increases access to finance, benefitting most firms (especially financially dependent ones, see Petersen and Rajan, 1995) and households. But there are two caveats. First, competition may make banking services more arm’s length and hence disadvantage information-sensitive borrowers (Berger and others, 2004; Carow and others, 2004; Karceski and others, 2005; Sapienza, 2002; Degryse and others, 2005). Second, when banks do not have good knowledge of customers, they may restrict credit during downturns (Petersen and Rajan, 1994; Bae and others, 2002; Bolton and others, 2012), increasing procyclicality.

Policy tools

76. The literature therefore suggests that an intermediate degree of bank competition is optimal. Competition should not be unbridled in order to allow banks to accumulate charter value that offsets their risk-taking incentives. But too little competition may make banks inefficient and possibly also lead to the TBTF problem.

77. The intermediate level of bank competition may be achieved by policies that focus on market structure (i.e. concentration):

• Entry/exit rules (for domestic and foreign banks);

• Consolidation policy (which may be particularly relevant around crises, when authorities can direct bank mergers; see Perotti and Suarez, 2003); and

• Restrictions on activities (non-lending activities of banks, and bank-like activities of non-banks such as insurance companies).

And by policies that affect contestability in banking services (competition given market structure):

• Establishing credit registries;

• Providing equal access to infrastructure, such as payment systems; and

• Other measures that enable easier switching of banks by customers.

Evidence from the crisis

78. The evidence from the crisis on whether certain market structures have enhanced the stability of banking systems is somewhat weak. Figure 13 provides an illustration. On the one hand, the evidence is consistent with some inverse U-shaped relationship between concentration—a measure of market structure—and performance of banks during the crisis (as predicted by the standard arguments). This is driven primarily by positions of Canada and Australia, the countries least affected by the crisis.

79. Yet the robustness of this relationship is unclear. Banks from countries with similar bank concentration were vastly differently affected by the crisis (compare Canada, France, and Ireland). Other factors—notably non-core exposures funded in wholesale markets (e.g., in Germany and U.K.) along with high leverage (e.g., in Ireland or Germany)—have certainly played a larger role in explaining bank performance.

View Full Size

Bank Concentration and Performance during the 2008 Crisis

(In percent)

Citation: Policy Papers 2013, 080; 10.5089/9781498341714.007.A001

Sources: World Bank Financial Structure Database (Beck and Demirguc-Kunt, 2009) and IMF staff calculations (Huang and Ratnovski, 2009).

• Download Figure
• Download figure as PowerPoint slide

80. One reason why the link between bank concentration and financial stability is weak might relate to recent evolution of financial services provision. Progress in information technology increased the availability of “hard” (quantifiable, verifiable) information on borrowers, particularly in advanced economies. This reduced the grip that banks had over their customers thanks to “soft” (proprietary) information accumulated in existing bank-customer relationships.⁸³ As a result, today, banks in advanced economies may act competitively even when market structure is highly concentrated. High competition means low profits and structurally (permanently) high incentives to take risk.

81. Another reason might be that competition in the provision in financial services has become more international and cross-sectoral. The last decades saw significant international expansion of banks. This makes the structure of domestic banking industry just one factor in defining the competitive environment. There is (anecdotal evidence) that foreign bank entry might have affected bank profits and risk-taking incentives in countries such as Spain and Ireland. Also, non-banks (e.g., finance companies and independent mortgage originators) have entered some lending markets (especially, prior to the crisis, mortgage markets) to compete with banks.

82. The final reason is that even when the relationship between bank market structure and stability is present, its properties are highly country-specific. Optimal bank market structure depends on financial development, quality and stringency of regulation, etc (Beck and others, 2013). This makes it hard to use a certain degree of bank concentration as a universal policy objective.

83. The fact that the link between market structure and bank risk-taking became weak has been long recognized in the literature (Claessens and Laeven, 2004; Demirguc-Kunt and others, 2004), but insufficiently reflected so far in policy debates. To sum up, while the correct market structure of banking is beneficial, it is not a panacea. Other aspects of prudential regulation should also be appropriate, especially so in modern financial systems.

Additional Lessons for Bank Competition Policy

84. The crisis experience highlighted several additional ways in which bank competition policy can support financial stability.

Addressing the TBTF problem

85. The TBTF problem is widely recognized as a major prudential concern. Banks that are complex and interconnected cannot be easily wound down, and have high incentives to take risk (O’Hara and Shaw, 1990; Flannery, 2009; Ueda and Weder di Mauro, 2012).

86. Currently, policy aims to address the TBTF problem through Basel III capital surcharges for systemically important banks (SIBs). The current measures might be insufficient to fully deal with the problem (see Haldane, 2012 and 2013). A notable reason is that the Basel SIFI capital surcharge is relatively small (up to 2.5 percent of risk-weighted assets), and might not give banks sufficient incentives to shrink or cut risks. There is then both a need and much scope to use in addition competition policy to address the TBTF problem. There are two approaches by which this can be done, a quantity- and a price-based approach.

87. The quantity-based approach is to use competition policy tools to directly restrict bank size. This can be achieved by limiting mergers, forcing spin-offs, etc. The challenge for this approach is to find a correct rationale for using such tools. Modern banks can act competitively and be efficient despite their size. Moreover, large banks may demonstrate spurious economies of scale, i.e. seem more efficient than small ones, thanks to: (i) lower cost of funds due to the TBTF subsidy, (ii) less borrower screening because they can afford more risk in lending operations, and (iii) more market-based activities, which are scalable but risky. It may therefore be hard to restrict bank size on pure competitive grounds. There are two ways around this. One way is to let competition policy adopt an explicit macroprudential objective, which would enable it to deal with large banks on the grounds of the welfare costs of possible crises. Another way, maybe more easily acceptable, is to focus on sub-segments of bank operations where bank size may indeed be a detriment to competition. For example, large banks are generally poor in dealing with opaque customers, such as small business borrowers or other customers intensive in “local” information. Ensuring access to finance on competitive terms by such customers may offer a rationale to restrict bank size or facilitate bank entry on pure competitive grounds.

88. The price-based approach is to use competition policy to correct competitive distortions created by TBTF. Interestingly, this may reduce the underlying TBTF problem as well. A key distortion is the uneven playing field that arises because large banks have access to cheaper funding than small banks. The cheaper funding is a result of an implicit TBTF guarantee to large banks’ creditors (that they will be bailed out in a crisis); the size of the funding advantage can be as high as 80 basis points a year (Ueda and Weder di Mauro, 2012). This distortion affects static competition, is a barrier to entry, and creates socially inefficient incentives for banks to grow. Correcting the uneven playing field is a natural goal for competition policy. The most direct tool to do so is to impose taxes or fines on large banks, to extract their unfair competitive advantage. (Think of a tax on wholesale funding of banks, with a rate that is increasing in bank size). From the perspective of competition policy, this would ensure a level playing field. From the prudential perspective, such taxes or fines would reduce the excess incentives of banks to grow, reducing the TBTF problem and enhancing financial stability.

Interaction with structural policies

89. The complexity of today’s financial system creates scope for structural policies. Structural policies, as defined here, are prudential measures that restrict bank or non-bank activities that contribute to systemic risk (Haldane, 2012; Gambacorta and van Rixtel, 2013). Many bank failures during the recent crisis were a consequence of non-core, market-based bank activities: investing in securitized credit, wholesale loan origination (originate-to-distribute), carry trade, proprietary trading (see Acharya and others, 2010; Boot and Ratnovski, 2012). At the same time, at the aggregate level, bubbles in housing markets were amplified by the lending of under-regulated non-banks (Dagher and Fu, 2012). Recent structural policy initiatives—which aim to reduce the risks described above – affect the competitive environment, and therefore have important interactions with competition policy.

90. A number of recent proposals suggest restricting market-based activities of banks: prohibition of proprietary trading in the Volcker Rule in the Dodd-Frank Act in the U.S. or a segregation of a wide range of non-core activities in the Vickers and Liikanen proposals in U.K. and for EC. Non-core activities are risky and have low margins, so restricting them might boost charter value and increase the stability of banks. Restrictions on bank activities may ease competition policy. By separating less contestable (core) from more contestable (non-core, including international) activities, it would allow competition policy to be more targeted—that is, able to use different approaches to the two sub-sectors, instead of focusing on banking as a whole.⁸⁴ For the core bank activities, it may mean more clarity on the competitive environment, and more precise competition policy.

91. Another possible structural measure is restricting competition in lending from non-banks. While historically lending was largely a bank-based activity, today non-banks, such as finance companies or independent mortgage originators, compete directly with banks. In the run-up to the crisis, lending by non-banks was riskier than that by banks (Dagher and Fu, 2012; Demyanyuk and Loutskina, 2012). This likely put pressures on banks to follow suit in order to maintain market share. Overall, non-bank lending amplified the mortgage bubble and led to riskier bank balance sheets. Non-bank lending was riskier in part because non-banks faced lax regulatory standards. Besides leading to more risk, the lax regulatory standards gave non-banks an artificial competitive advantage (of being less constrained by capital requirements or other prudential rules). Competition policy may help prudential policy by insisting on level playing field in lending, where non-banks can only compete directly with banks if they are subject to the same or similar prudential standards.

92. The third structural problem highlighted by the crisis is excess competition for retail deposits. Retail deposits are scarce (their supply is capped by the part of household saving that goes to banks), yet valuable to banks because insured deposits are the most stable source of funds (Huang and Ratnovski, 2009 and 2011). There were various types of excess competition for deposits in the run-up to the crisis. One was competition from non-banks (such as insurance companies, asset managers, money market funds) for household savings—as in Australia or the Nordics. Another was competition from local savings banks with implicit public guarantees that historically accumulated a large share of depositor base—notably in Germany. In any case, commercial banks were left with too little household deposits and had to rely on wholesale funding, which was a major source of vulnerability during the crisis.⁸⁵ Resolving the scarcity of deposits is a key structural challenge for policy. Competition policy might assist by establishing a level playing field in access to household savings—including by dealing with implicit guarantees and lax regulation of non-banks.

Competition policy and crisis management

93. The crisis put into sharp relief possible conflicts between bank competition policy and crisis management (Vickers, 2010; Hasan and Marinč, 2013). Normally, competition policy advocates limited government involvement in banks in order to maintain the level playing field between government-affiliated and other banks. Yet, crisis management might exceptionally require governments to take ownership in banks or offer banks guarantees in order to maintain financial stability and the capacity to lend. Also, governments might need to exercise control over banks to direct their restructuring. (These measures are not the only or most efficient paths for crisis management, but may be necessary under some conditions; Laeven and Valencia, 2012). In such exceptional circumstances, competition policy should acknowledge the trade-off between the need for effective bank resolution versus preserving the level playing field, and aim to strike a balance. Also, competition policy might need to temporarily allow higher banking system concentration, when that is necessary to allow banks to rebuild charter values or to facilitate the shrinking of a previously over-expanded banking system.

Conclusions

94. This note offers two main lessons as to how competition policy might support financial stability. The first lesson is that, because the degree of competition affects bank risk-taking, bank competition policy should have a macroprudential component. Sometimes there is a need to sacrifice some competition to ensure more financial stability. Therefore, bank competition policy may require rules that are distinct from those for non-financial firms. (This is similar to bank resolution laws that have different procedures than those in regular bankruptcy codes; Marinc and Vlahu, 2012). Regardless, the competition authority needs to cooperate with the prudential regulator.⁸⁶ And there is a strong rationale to have financial stability as one of the objectives of the financial competition authority. Sometimes, there might be rationale to transfer some responsibility for bank competition (inasmuch as it affects financial stability) to the prudential regulator. The second lesson is that bank competition policy can also help respond to specific prudential concerns. It can help dealing with the TBTF problem, help address some structural (related to the scope of activities) sources of financial instability, and—by temporarily allowing higher concentration or government control of banks—may help facilitate effective crisis resolution.

The Context

105. At the height of the global financial crisis in 2008, the Bank of Israel (BOI) slashed policy interest rates from 4.25 percent in September 2008 to 0.5 percent by April 2009. These reductions, along with other measures to improve liquidity, enabled output to recover quickly; signs of recovery emerged by mid-2009. The sharp interest rate reductions also spurred housing market activity, as lower interest rates encouraged households to invest in real estate.⁹¹ With parts of the corporate sector damaged by the crisis, banks were also eager to diversify their portfolios by lending to households. Land supply in Israel is limited, and building regulations are cumbersome, constraining the speed at which land is approved for use and new building can start. With supply restricted, the boost in demand from lower interest rates fed an increase in prices that eventually came to cause concern.

The Policy Response to Rising Risks from the Housing Market

106. The BOI faced a dual challenge stemming from its responsibilities both for conducting monetary policy and as the supervisor of banks. With the economy among the first to recover from the global crisis, policy interest rates were gradually raised starting in September 2009, to reach 3.25 percent in June 2011. During this period, the housing market remained buoyant and banks expanded mortgage lending aggressively. Between 2008Q4 and 2011Q1, real house prices surged by 38 percent. In mid 2009, the BOI called for stronger risk management in banks, and starting in 2010 it adopted macroprudential policy measures to address the building risks to financial stability. The measures included:

• requesting banks to tighten their risk management, scrutinize the mortgage loans to households, and enhance disclosure, particularly with respect to loans carrying floating interest rates that were extended to households (August 2009);

• instituting a supplementary reserve requirement of 0.75 percent for all outstanding mortgages with a LTV ratio that exceeds 60 percent (July 2010);

• imposing a capital charge of 100 percent (instead of 35 percent) on all loans over a threshold of NIS 800,000 that have a floating interest rate component that applies to more than 25 percent of the principal of the loan and involves a LTV ratio that exceeds 60 percent (October 2010); and

• requiring that the adjustable interest rate component of a mortgage loan apply to no more than one-third of the principal amount of the loan, and that banks notify customers whose mortgage loans carry a floating interest rate component that applies to one-third or more of the loan (May 2011).

107. Real house price growth moderated from 2011Q1, but picked up again in 2012, and the BOI adopted further macroprudential policies starting in July 2012. By this time, however, the economic environment had worsened and, with inflation on the decline, the BOI gradually lowered policy interest rates to support economic activity. At the same time, it tightened financial policies further. Measures included:

• imposing a 100 percent capital charge on groups of borrowers that collectively bought property for residential projects and engaged third parties to execute the construction and development of the property (July 2012)

• Imposing a limit of 70 percent on LTV for housing loans—excluding first-time buyers, for which a maximum LTV of 75 percent would apply. In addition, LTV for mortgage loans for investment purposes could not exceed 50 percent (November 2012)

• In February 2013, further measures were proposed, including increased capital requirements (though graduated risk weights by LTV levels), and higher provisioning for housing loans.

Policy Challenges

108. Central banks also responsible for bank supervision need to be prepared for the potential for tension between the objectives of price stability and financial stability. In Israel, this tension was not present in the recovery episode as both the economy and the housing market picked up; the BOI’s interest rate increases and its adoption of macroprudential policies went in the same direction. By contrast, in 2011, global economic conditions were volatile and particularly uncertain, threatening Israel’s recovery, and inflation subdued, while house prices looked to be picking up strongly again. Price stability and financial stability objectives were pointing to policy actions in opposing directions: loosen monetary and tighten financial. The BOI implemented looser monetary policy to address price stability issues and tightened macroprudential policy to contain financial stability risk.

Other Challenges

109. Communications and accountability. Communications of policy objectives and intentions, always tricky, can be particularly challenging in circumstances when policy actions seem to go in opposing directions. The BOI has taken care to explain its decisions carefully, signaling its concern on the housing market and linking this to macroprudential policy actions taken in this context. At the same time, it made publicly clear that monetary policy decisions were separate from macroprudential policy decisions. The Monetary Committee was responsible (and accountable) for the interest rate decision, while the decision on LTV was the responsibility of the Supervisor of Banks.

110. Social and political sensitivities. Housing is a politically sensitive issue; in 2011, large public protests took place, including to call for more affordable housing. The BOI’s actions to restrain house price inflation have sometimes generated some negative commentary in the press, particularly with respect to reducing access to mortgage loans, when LTV limits were tightened.