A. Monetary Policy

19. Strong complementarities and interactions between monetary and macroprudential policies reinforce the need for a strong macroprudential framework. Complementarities explain why central banks have a strong interest in ensuring the effective pursuit of macroprudential policy and are often at the forefront in the push for the establishment of macroprudential frameworks. Interactions also call for some degree of coordination between monetary and macroprudential policies, while preserving the established independence and credibility of monetary policy.

20. As set out in IMF (2013a), even when monetary policy is set consistent with price stability, the resulting monetary stance may have undesirable side effects for financial stability.

• Where low policy rates are consistent with low inflation, they may still contribute to excessive credit growth and the build-up of asset bubbles and sow the seeds of financial instability.

• In small open economies, increases in interest rates may be necessary in the face of inflationary shocks, but can draw in capital flows that may contribute to excessive financial risks. Conversely, the need for interest cuts to counter subdued domestic demand may lead to large capital outflows that can jeopardize domestic financial stability.

21. Where there is a strong macroprudential policy framework, this can reduce conflicts and create more room for maneuver for monetary policy to pursue price stability.¹⁰ Where macroprudential policy is assigned an appropriate range of tools, it will be better able to address undesired side effects of monetary policy at their source. This can help alleviate conflicts in the pursuit of monetary policy and reduce the burden on monetary policy to ‘lean against’ adverse financial developments, thereby creating greater room for maneuver for the monetary authority to pursue price stability. However, since macroprudential cannot be expected to be fully effective, the conduct of monetary policy needs also to take account of financial stability considerations.¹¹

22. Moreover, to the extent that macroprudential policy reduces systemic risks and creates buffers, this helps the task of monetary policy in the face of adverse financial shocks. It can reduce the risk that monetary policy runs into constraints in the face of adverse financial shocks, such as the zero lower bound—recently hit by many advanced economies—or the risk of strong outflows associated with cuts in interest rates in small open economies.

B. Fiscal and Structural Policies

23. Appropriate fiscal and structural policies are critical to reduce the likelihood of macroeconomic shocks. The build-up of systemic risk can be driven strongly by macroeconomic imbalances—internal or external—and distortions that affect the composition of output.

• Where a consumption boom is fed by capital inflows in the presence of persistent current account deficits, macroprudential policies alone are unlikely to be effective at controlling these underlying forces and prudent fiscal and structural policies are needed to contain these imbalances.

• The crisis also showed that prudent fiscal policies are essential to maintain the safety of sovereign debt and to avoid adverse feedback loops between sovereign risk and the financial system.

While macroprudential policymakers cannot be in control of fiscal and structural policies, they can help analyze the underlying macroeconomic risks and imbalances, draw out systemic implications, and flag their concerns to those policymakers that are in a position to take appropriate action.

24. Existing tax policies can create biases that can contribute to systemic risk (background paper). Macroprudential authorities have an interest to encourage a correction of such biases, since they make it harder for the macroprudential policymaker to achieve its objective.

• First, corporate tax systems generally encourage the use of debt rather than equity finance. Because interest paid is allowed as a deduction in calculating taxable profits, but the return to equity is not, corporate taxes typically create a ‘debt bias.’¹² There is evidence that this effect matters for leverage choices of non-financial companies, as well as financial institutions.¹³ Thus, there is a basic tension between regulatory measures intended to induce banks to hold more capital than they otherwise would and tax incentives to hold less.

• Second, the tax treatment of housing can make households more vulnerable to shocks and exacerbate systemic risk. Many countries do not (or only lightly) tax imputed rent, while providing generous relief for mortgage interest. This can be a source of significant distortion (and revenue loss), as households are encouraged to borrow against housing assets, either to invest in non-housing assets or to finance immediate consumption. High mortgage debt can make households more vulnerable to shocks, thus exacerbating transmission channels in the financial system leading to a crisis.

25. Pigovian taxes and levies can also be used more directly to address systemic externalities. The “Financial Stability Contribution” (FSC) proposed by the Fund is a key example.¹⁴ Such a levy can help discourage leverage and wholesale funding and at the same time be linked to the financing of a credible and effective resolution mechanism. It could be levied at a flat rate, or varied to reflect individual institutions’ contributions to systemic risk, as well as variations in overall risk over time.¹⁵

26. Taxes can also affect asset prices. As future tax liabilities are capitalized, in principle, imposing taxes during a boom can make bubbles less likely; or the announcement of future tax relief on asset returns can support asset prices during a bust. For example, during the crisis countries have used tax measures to bolster house prices by removing stamp duties on housing transactions or extending mortgage interest relief. Stamp duties have more recently also been used in a number of countries to lean against house price appreciation, such as in Singapore and Hong Kong. Such measures may have a role to play in particular when real estate prices are driven by capital inflows that by-pass the domestic financial system. However these measures can also introduce further distortions and may ultimately increase price volatility. Indeed, the practical feasibility of cyclical changes in taxes could be limited due to lags in fiscal decision making and implementation (background paper).

27. Tax measures are usefully complemented by structural measures that affect the supply of housing, such as land use policies, and the functioning of housing markets more generally (IMF, 2011e). While these measures should not be viewed as macroprudential stricto sensu, their effect on house price dynamics may have benefits in reducing systemic risks. Macroprudential policy authorities will therefore often have an interest to influence the design of these policies.

C. Competition Policies

28. Competitive processes within the financial sector can create incentives for excessive risk-taking, leading to tensions between competition and financial stability (background paper).¹⁶ It is well-known that, in general, competition is likely to result in more cost-effective production of goods and services and higher efficiency. However, competitive processes within the financial sector can also contribute to systemic risk. This can create tensions between the objectives of competition authorities and those of the macroprudential policymaker, in particular when the macroprudential authority is concerned about the build-up of risk over time (background paper).

• A substantial body of evidence documents that a relaxation of licensing and branching restrictions can lead to an erosion of lending standards and too much risk taking by financial institutions (e.g., Keeley, 1990).

• Foreign entry, increased access to wholesale funding, and greater availability of “hard” (quantifiable) information on borrowers, while in principle desirable, can lead banks to compete aggressively for market share, reducing margins and creating strong incentives to take too much risk.

• The proliferation of “non-bank” providers of credit can increase competitive pressures in credit markets that result in risks to financial stability (Bordo and others, 2011). The credit-card boom-busts in Korea in the early 2000’s and more recently in Mexico provide examples.¹⁷

29. In the structural dimension, by contrast, the financial stability perspective often reinforces traditional competition concerns. Here, therefore, the macroprudential authority may favor more forceful action than what is justified purely on competition grounds.

• Mergers between financial intermediaries have the potential to create institutions that are “too big to fail.” Macroprudential authorities therefore have an interest in the control of mergers and takeovers between financial institutions.

• Large and complex financial institutions may be “too difficult to resolve.” This can call for a preventive power to break-up financial institutions or to force changes in their organizational structure, to increase resolvability.¹⁸

30. In general, therefore, competition policy for the financial sector requires a macroprudential perspective.¹⁹ One way of assuring that the interplay between competition and systemic risk is taken into account is to assign traditional powers of competition policy, including licensing, take-over control and break-up powers to the prudential authorities.²⁰ Another is to ensure strong mechanisms of coordination and consultation between the prudential and competition authorities and to introduce financial stability as a secondary objective of the latter. Where there is a distinction between the micro-and macroprudential authorities, the macroprudential body should be consulted on the design and application of these policies.

D. Microprudential Policies

31. In principle, microprudential supervision should work “hand in glove” with macroprudential policy. Revisions made in the wake of the crisis to the Basel Core Principles now place a much stronger emphasis on the need for a macroprudential perspective in supervision.²¹ A key recommendation is that supervisory agencies are tasked not only to ensure the safety and soundness of individual institutions, but also to contribute to the stability of the system as a whole. Shared information, joint analysis of risks and strong dialogue can reinforce the complementarities between microprudential supervision and macroprudential policy. Indeed, strong supervision is essential both to ensure that macroprudential policymakers can draw on supervisory information in risk assessment and to ensure that the macroprudential policy stance adopted is effectively enforced across institutions.

32. Tensions between microprudential and macroprudential perspectives may also arise, especially in “bad times” (Osiński and others, 2013). Microprudential and macroprudential requirements work through similar transmission mechanisms and it is their joint effects that will determine the outcome.²² In “good times,” the microprudential supervisor may often agree that it would be prudent for banks to build up buffers, even though there may not be the same sense of urgency when non-performing loans are small and profits ample. In “bad times” tensions can be stronger, since the macroprudential perspective may call for a relaxation of regulatory requirements that impede the provision of credit to the economy or contribute to fire-sale effects, while the traditional microprudential perspective may seek to retain or tighten these requirements to protect the interest of depositors of individual banks.

• One way of addressing conflicts in “bad times” is to establish in “good times” sufficient prudential buffers. The macroprudential authority may then be in a position to reduce buffers in a manner that respects microprudential objectives.

• Where buffers are not adequate to start with, conflicts can sometimes still be resolved through well-designed prudential action. Microprudential authorities may call for increasing capital ratios in bad times, while macroprudential authorities will be concerned that this leads to excessive deleveraging with adverse effects on the economy. Encouraging increases to be implemented by higher capital levels (privately or publicly provided) avoids that deleveraging and can align microprudential and macroprudential objectives.

• Another way of dealing with these conflicts is to establish institutional mechanisms that allow for their resolution. In particular, where the macroprudential decision-maker is established as a macroprudential committee, this allows for supervisory agencies to participate in macroprudential decision-making, allowing for supervisory perspectives to be brought to bear (Nier and others, 2011).

E. Crisis Management and Resolution Policies

33. Crisis management and resolution policies are complementary to macroprudential policy. The management of crises may require monetary easing and emergency liquidity assistance by the central bank, the effective resolution of failing banks by dedicated resolution or deposit insurance agencies, and potentially public guarantees and capital support provided by the fiscal authorities.

34. Proper design of resolution regimes can support the objectives of macroprudential policy. Effective and credible resolution regimes can strengthen market discipline and reduce incentives to take excessive risks, mitigating the need for macroprudential intervention. By contrast, where difficulties in the resolution of specific financial institutions remain, this will require more forceful macroprudential action, which can in turn lead to greater incentives for circumvention.

35. Crisis management requires close coordination among all financial sector authorities. Macroprudential authorities are likely to continue to provide advice, based on their assessment of the evolution of the level and sources of systemic risk. They may also use instruments in their purview to contain amplification of risks. However, in a systemic crisis, the fiscal authorities are likely to take the lead in coordinating the overall policy response, as taxpayer money would need to be used.

A. Assessing Systemic Risk

37. An effective framework for monitoring systemic risk is key to operationalizing macroprudential policy. It needs to consider (i) the growth in total credit and macroeconomic drivers of imbalances; (ii) financial linkages between the financial and domestic real sectors (household and corporate), and between each sector and the rest of the world; and (iii) the structure of the financial system and linkages within and across key classes of intermediaries and market infrastructures (Figure 3). The assessment will often draw heavily on supervisory and statistical data, and make use of empirical methods, but needs also to make full use of market intelligence and soft supervisory information on trends and market developments.

View Full Size

Systemic Risks in Time and Structural Dimension

Citation: Policy Papers 2013, 082; 10.5089/9781498341707.007.A001

• Download Figure
• Download figure as PowerPoint slide

38. Strong increases in credit can signal a build-up of systemic risk. The empirical literature finds that increases in the ratio of private sector credit to GDP, as measured by the growth in credit relative to GDP or by the deviation from its long-run trend, is the best single indicator of an increase in the probability of a crisis over a horizon of 1 to 3 years (Drehmann and others, 2011; Drehmann and Juselius, 2012; IMF, 2011b; Lund-Jensen, 2012). Use of broad measures of credit, including credit provided by non-banks and cross-border, can improve the indicator (Drehmann and others, 2011; IMF, 2011b; Arregui and others, 2013). However, it is also found that not all credit booms end in a bust, as they may be justified by better fundamentals, and that loan growth can contribute to a healthy financial deepening (Dell’Ariccia and others, 2012).

39. It is important to consider the macroeconomic environment that gives rise to increases in credit and additional indicators of the build-up of systemic risk. Combining the analysis of credit growth with other indicators can help in deciding whether excessive credit growth poses systemic risk. These indicators include:

• the prices of assets that are used as collateral for secured lending and that may contribute to feedback between increases in leverage and asset prices;²³

• the leverage taken by borrowers in those asset markets, on average as well as on new loans, since the latter will be a more timely measure of credit conditions;²⁴

• changes in lending standards, as can be captured by decreases in lending margins and increases in household and corporate leverage;

• measures of balance sheet stretch in the household and corporate sectors, as can be captured by debt-service to income ratios for each sector (Drehmann and Juselius, 2012);

• increases in exposure of the household and corporate sectors to interest rate and currency risks that create vulnerabilities to aggregate shocks; and²⁵

• external imbalances, as reflected in current account deficits and an appreciation of the real exchange rate, that can increase the probability of crises.²⁶

40. Close analysis of the build-up of vulnerabilities within the financial sector is essential. This analysis needs to consider whether an increase in credit to GDP is sustainable, or stretching the system and subjecting it to unsustainable increases in credit, FX and liquidity risks. It should consider:

• the capacity of the financial system to absorb default risk and changes in asset valuations, as can be measured by leverage ratios and measures of profitability, and assessed through various types of stress tests; and

• the extent to which banks increase reliance on wholesale funding to finance credit expansion, thereby exposing the system to liquidity shocks, as measured by the ratio of non-core to core funding or the ratio of credit to deposits.

41. In the structural dimension the analysis needs to consider the extent to which linkages within the financial system pose a threat. The crisis has shown that new products, such as credit default swaps, can lead to rapid changes in risk-concentrations across the financial sector and increased exposure to tail risks for some institutions, which in turn render these institutions “too important to fail.” It also highlighted changes in the structure of the financial system where non-banks, such as money market mutual funds, become important sources of funding for banks, while the liquidity mismatches of such institutions can contribute to endogenous fire sale dynamics that weaken the financial system.²⁷

42. Monitoring and analysis of these interconnections is important to assess the need for a macroprudential response. This needs to recognize that connections within the financial system can act both as shock absorber and as shock transmitter and that an increase in connectivity is not always associated with increases in systemic risk. Recent advances in techniques to assess risks arising from interconnections (such as network analysis), and methods based on market prices (such as contingent claims analysis), can help assess the risk from interconnectedness (Arregui, Scarlata, and others, 2013). However, such analysis is often still hampered by gaps in the data available to the authorities.

43. A range of indicators and methods need to be combined with qualitative judgment to arrive at the decision on when to act. While slow-moving indicators are useful for detecting the build-up of risks, high-frequency and market-based indicators are better in predicting an imminent materialization of systemic risk and can help prepare the authorities to respond to stresses several months in advance (Blancher and others, 2013; Arsov and others, 2013).

44. The systemic risk monitoring framework is work in progress in a number of key dimensions. Tools exist to assess most sectors and levels of aggregation (Annex II). However, these tools provide only partial coverage of potential risks and only tentative signals on the likelihood and impact of systemic risk events. As such, they may not provide sufficient comfort to policymakers. Further progress is needed in the following areas:

• Early warning. The forward-looking properties of systemic risk measures are generally weak, even though some measures appear relatively promising, such as combinations of credit-to-GDP and asset valuation measures.

• Thresholds. Policymakers need clear and reliable signals indicating when to “worry” and when to take action, and allowing them to monitor the impact of such action over time. For example, empirical evidence suggests that when credit and property prices grow beyond certain levels the probability of a financial crisis over 2 to 3 years increases considerably (Annex IV). Despite recent progress, further work is needed in this area.

• System’s behavior. The capacity to model the system’s behavior is limited in several areas, including endogenous responses to the materialization of aggregate shocks, such as feedback and multi-round effects, and nonlinear risk correlations during periods of financial distress.

B. Selecting and Assembling the Macroprudential Toolkit

45. The authorities need to select and assemble a set of macroprudential instruments that can help address the key potential sources and dimensions of systemic risk. Since the manifestations of systemic risk can depend on country-characteristics, no specific set of tools can readily be identified as “best practice.” Equally, since risks can build-up rapidly, and acquiring and implementing new tools takes time, there is merit in considering and introducing a range of tools ex ante. Interlocking use of several tools can help overcome the shortcomings of any one single tool and enable the policymaker to adjust the overall policy response to a range of risk profiles, thereby promoting the effectiveness and efficiency of the policy response.

46. In the time dimension, consideration should be given to three sets of tools (Figure 4): (i) countercyclical capital buffers and provisions, to increase resilience to shocks; (ii) sectoral tools, to contain a build-up of risks in particular sectors; and (iii) liquidity tools, to contain funding risks. These three sets of tools can reinforce and complement each other in addressing the build-up of risks through time.

View Full Size

Mapping Tools to Objectives: Time Dimension

Citation: Policy Papers 2013, 082; 10.5089/9781498341707.007.A001

Source: IMF staff.

• Download Figure
• Download figure as PowerPoint slide

47. Countercyclical capital buffers and provisions. The countercyclical capital buffer (CCB) will be available in many countries through implementation of Basel III. ²⁸ The advantage of the CCB is that it can be activated to increase the resilience of the system to a whole range of shocks. The release of the buffer is expected to avoid credit crunch effects by reducing the pressure on banks to deleverage in times of financial stress.²⁹ While experience with these tools is still to be gained, simulation exercises suggest that activation of CCBs would have been triggered by increases in credit to GDP well ahead of the crises in Ireland and Spain, providing additional resilience and complementing dynamic provisions in the case of Spain (see Figure 7 in Annex III and background note). However, the CCB is a blunt tool, applied uniformly to all exposures and is likely to be slow to react to the build-up of risks in particular segments of the credit market. In addition, where banks hold voluntary buffers above the minimum, or can easily generate capital through strong earnings, the activation of the buffer may not markedly slow down overall credit growth.³⁰

48. Sectoral capital requirements. Increases in risk weights for lending to particular segments of the credit market can complement the CCB. In principle, a targeted increase in risk weights can be applied to any category of loans for which strong credit growth gives cause for concern. This can include mortgage lending, unsecured consumer credit, or specific segments of such credit, as in Brazil and Turkey, and corporate lending or specific corporate segments, such as lending to commercial property, as proposed in the U.K. (background paper). An increase in risk weights is expected to increase loan interest rates for lending to the targeted sector and can also increase the resilience of lenders to a deterioration in credit quality.³¹

49. LTV and DTI ratios. The international experience justifies a particular emphasis on tools that can contain vulnerabilities in residential housing markets. A number of countries have found that the pass-through of an increase in capital requirements on mortgage loans to loan growth can be limited when strong increases in asset prices and credit feed each other (e.g., Israel, Korea). This suggests the use of additional tools that act on the demand for credit and directly increase the resilience of borrowers to shocks.

• An LTV ratio introduces a cap on the size of a mortgage loan relative to the value of a property, thereby imposing a minimum down payment.

• A DTI ratio restricts the size of a mortgage loan to a fixed multiple of household income, thereby containing unaffordable and unsustainable increases in household debt.

The available research (surveyed in the background paper) suggests that these tools can reduce feedback between credit and prices in upswing, as well as improve resilience to shocks, thereby reducing default rates and boosting recovery values when the housing market turns. However, they can also be seen as more intrusive and calibration can seek to soften their impact, e.g., by exempting first-time buyers (as in Korea, background paper). Moreover, the evidence suggests that LTV ratios can have a relatively strong effect on house prices and aggregate demand (IMF 2013b), which can justify a gradual approach to tightening of such ratios (as in Canada and the Netherlands).

50. Liquidity tools. The crisis has highlighted the systemic externalities associated with funding liquidity risk (Perotti and Suarez, 2010; Shin, 2010a; Huang and Ratnovski, 2011) and sparked a greater emphasis on liquidity tools that reduce vulnerabilities from a system-wide increase in wholesale, short-term and FX funding. These tools can be quantity-based or price-based constraints that aim to reduce reliance on vulnerable non-core funding. While Basel III includes prudential measures to reduce funding risks (Liquidity Coverage Ratio and Net Stable Funding Ratios (NSFRs)), additional tools can be adapted to local conditions.³² Examples are the Macroprudential Stability Levy introduced in Korea and the Core Funding Ratio (CFR) introduced in New Zealand, as well as the use of (marginal) reserves requirements for macroprudential purposes in a number of emerging markets. An additional benefit of liquidity tools is their effect on loan growth. Since core funding, such as retail deposits, grows slowly, credit booms will often be funded by increases in wholesale funding. By constraining such funding, liquidity tools can then also contribute to a slowing of overall credit growth, complementing the effects of the countercyclical capital buffer or provisions and sectoral tools.

51. A range of complementary tools can contain structural risks from interconnectedness and contagion within the financial system (Figure 5). First, to improve resilience and resolvability of those institutions whose failure poses systemic risks to the system, prudential requirements can be tightened on those firms. Capital surcharges for global and domestic systemically important institutions are the key example. Second, to reduce the contagious effect of the failure of such institutions, prudential tools can be used to discourage exposures to such institutions or excessively large exposures within the financial system more generally (Arregui, Scarlata, and others, 2013).

View Full Size

Mapping Tools to Objectives: Structural Dimension

Citation: Policy Papers 2013, 082; 10.5089/9781498341707.007.A001

Source: IMF staff.

• Download Figure
• Download figure as PowerPoint slide

• This can take the form of increased sectoral capital requirements for exposures within the financial system or specific types of exposures that are growing rapidly, as proposed for the U.K. Financial Policy Committee (FPC).

• It can also take the form of quantitative limits on the size of such exposures relative to capital, extending existing large exposure regimes to exposures within the financial system, as is currently being discussed by the Basel Committee.

• Since non-core funding is often raised in wholesale financial markets, liquidity instruments can contribute to a reduction of domestic or cross-border exposures among financial institutions (Shin, 2010a and 2010b).

52. Changes can also be made to the market infrastructure, including payment, settlement and clearing arrangements, to reduce the build-up of credit exposures arising from transactions within the financial system. Growing counterparty credit risks from derivatives transactions have been the main concern in the aftermath of the crisis and spurred efforts to introduce central counterparties for the clearing of derivatives as part of the G-20 led reform initiative. These entities need to be designed prudently and supervised closely, however, since while they reduce interconnectedness they also concentrate systemic risk. Finally, since contagious effects are often strengthened by a lack of information, mandating increased transparency over exposures can be a particularly useful measure in the structural dimension.

C. Calibrating Macroprudential Tools

53. Use of macroprudential tools needs to trade off their benefits and costs. It needs to bring together the analysis of changes in the sources and level of systemic risk; an understanding of the transmission of the available macroprudential policy tools; and an assessment of the costs and distortions arising from macroprudential policy action (CGFS, 2012; Arregui and others, 2013). While macroprudential action in the structural dimension will often take the form of a sustained initiative that imposes new constraints, action in the time dimension can involve dynamic changes to evolving conditions, with tightening of constraints followed by a relaxation as risks abate or crystallize.

54. The transmission mechanism of macroprudential action is subject to considerable uncertainty. While it is possible to map the channels of transmission of macroprudential tools conceptually, the strength of these effects is still uncertain. A growing literature—reviewed in detail in the background paper—documents that macroprudential actions transmit to resilience, as well as the growth in credit and asset prices (Lim and others, 2011; Vandenbussche and others 2012; CGFS, 2012; IMF 2013b; Arregui and others, 2013). An emerging literature also assesses the short-run costs of variation in macroprudential tools on output (CGFS, 2012; IMF, 2013b; Arregui and others, 2013).³³ However, the relative strength of these effects across tools and their precise quantitative impacts are difficult to measure. Moreover, while progress is being made, the benefits of macroprudential action for the ultimate objective of reducing the probability and depth of future crises, remains difficult to quantify (Annex IV). Further research is essential to improve the policymaker’s understanding of the benefits and costs of each tool. In addition, the authorities need to monitor the transmission of macroprudential action on key indicators of vulnerability in real time, so as to adjust the calibration in light of emerging indications of their effects.

55. Efficient calibration requires a degree of judgment, to enable a response to evolving risks. A key advantage of a static or rules-based calibration is the reduced need to overcome political opposition to the discretionary variation of macroprudential tools. However, to provide sufficiently strong defenses against a build-up of systemic risk, a static calibration may need to be inefficiently tight at all times, distorting financial activity and creating incentives for circumvention (Goodhart, 2008). One way of balancing these considerations is to introduce “guided discretion”, based on key indicators, but complemented by judgment that takes account of all available information (Swiss National Bank 2012; Bank of England (BoE), 2013; background paper). Another is to complement tools that work as automatic stabilizers (such as dynamic provisions) with a range of other tools that can be targeted and adjusted to evolving risks.

56. Access to multiple tools can improve policy trade-offs. The marginal benefit of tightening any one tool is likely to be decreasing as aggressive use of any tool is likely to create distortions and incentives for circumvention (CGFS, 2012). Interlocking use of complementary tools can reduce such costs and can be more closely tailored to the prevailing profile of risks, thereby reaping more efficiently the desired benefits of macroprudential action. For instance, a targeted tightening of sectoral tools can suffice when risks are confined to particular segment, while broader impact tools can address risks from a more pervasive growth of credit (Lim and others, 2011).

57. Use of macroprudential policy tools needs to be supported by clear communication. Communication with markets and the public can foster an understanding of the benefits of specific macroprudential tools. It can also be used to steer market expectations on how and under what circumstances any given policy tool will be used, for instance by mapping ‘presumptive indicators’ for the use of each tool (Goodhart, 2011).³⁴ An impact on expectations can in turn condition behavior and reduce costs associated with variation in macroprudential tools (CGFS, 2012, Giese and others 2013). However, communication of a policy strategy needs to balance the need for flexibility in the deployment of macroprudential tools.

58. Trade-offs in the calibration of macroprudential tools may be particularly stark in stressed conditions. In periods of financial stress, the macroprudential policymaker will want to relax those macroprudential constraints that impede the provision of credit to the economy, lead to fire sale dynamics and exacerbate a vicious feedback between deteriorating economic and financial conditions. At the same time, relaxation of macroprudential constraints needs to maintain confidence and ensure an appropriate degree of resilience against future shocks.

59. These trade-offs may differ across macroprudential tools. For instance, using tools that relax liquidity constraints can help sustain credit in periods of stress (IMF, 2013b). It can also accommodate investors’ preference to shorten the maturity of funding provided to banks in stressed conditions and need therefore not entail a loss of confidence. A relaxation of sectoral constraints that is confined to new lending can strike a balance between the need to maintain resilience and the desire to promote the provision of credit to the economy. Finally, dynamic provisioning regimes act as automatic stabilizers, since the draw-down of the provisioning pool is built into the design of the regime.

60. Trade-offs can be improved when macroprudential policy action is complemented by policy action in other areas. For instance, robust resolution frameworks and disclosure of stress test results for individual banks can allow the authorities to credibly deal with individual weak institutions, thereby allowing for a relaxation of system-wide capital buffers in a manner that maintains confidence (Osiński and others, 2013). Similarly, liquidity requirements can be more readily relaxed when there is a strong commitment on the part of the central bank to provide emergency liquidity.

D. Monitoring and Closing Regulatory Gaps

61. Authorities need to monitor the migration of activity outside of the reach of macroprudential tools and close regulatory gaps. The authorities need to monitor and assess the extent of such migration and respond, by expanding the scope of macroprudential action as necessary. The extension of prudential tools to non-bank and market activity needs to be guided by economic function, rather than legal form, and be proportionate to the risks to financial stability posed by such activity (FSB 2012).

62. Macroprudential policy intervention needs to be aware of ‘boundary problems.’ If regulation is effective but its reach is confined to certain types of regulated institutions, it will constrain the regulated firms from achieving their preferred position and lower their return on equity (Goodhart, 2008). This will tend to lead to a switch of business activity to less constrained intermediaries that are outside the scope of the regulation. It can also lead to a shift toward market-based financing and efforts by regulated firms to open up connected operations in the less regulated sector. This problem is generic, and applies to all financial regulation. ³⁵

63. In principle, a macroprudential approach is well-suited to address the boundary problem. It is less focused on protecting bank depositors, but takes a system perspective that encompasses all individually systemic institutions and all providers of financial services that are collectively systemically important (IMF 2011a; Nier, 2011). A macroprudential intervention or an extension of the regulatory perimeter that is motivated by macroprudential concerns, rather than the desire to protect investors, need therefore not imply an extension of the “safety net.”

64. A macroprudential approach calls for an extension of tools to all provision of credit that is ultimately dependent on short-term funding and leverage cycles in markets.

• In line with recommendations by the FSB, equivalent prudential intervention should be extended to “bank-like” intermediaries that do not take deposits, but whose economic function is the provision of credit to the economy and who depend on short term debt funding.

• Regulation should also extend to providers of leverage in markets that are dependent on short term funding or secured funding (e.g., broker dealers).

• A macroprudential approach calls for the regulation of credit products traded in financial markets. This can include the regulation of securitization activity in a manner that creates transparency and ‘skin in the game’ across the intermediation chain.

• A macroprudential approach finally calls for the regulation of haircuts in securities lending and repo markets, as well as margin requirements in derivatives markets, to avoid margin spirals that contribute to excess leverage and procyclicality (Geanakoplos, 2010; Longworth, 2010; Hanson and others, 2011; FSB 2012).

65. While some of these approaches are new, there is growing practical experience, as countries have encountered the need to respond to arbitrage and the growing provision of credit by non-banks and in markets when deploying macroprudential tools. Annex V highlights the experience in Croatia, Korea, New Zealand and the U.S.

E. Closing Data and Information Gaps

66. Effectiveness of macroprudential policy will benefit from a sustained effort to close information gaps. Data and information gaps can hinder the early detection of systemic risk and increase uncertainty on the need for a response to identified concerns, impede the design and enforcement of macroprudential instruments, and complicate the policing of the regulatory perimeter. While the most pressing data and information gaps can vary across financial systems, the crisis has thrown such gaps into sharp relief, in both advanced and emerging markets.³⁶

67. Information gaps can impede the assessment and mitigation of risks in the household and corporate sector. Data on the household sector is scarce in many countries, and greater granularity is needed to assess and mitigate risks. For instance, to implement and enforce effective LTV and DTI ratios, information on real estate prices, preferably at a regional level is needed.³⁷ In

addition, credit registers are required to provide information on pre-existing senior loans, enabling the overall ratios to be computed. Similarly, to assess the need for and enforce measures that contain FX exposures on the part of the corporate sector, granular data is needed on cross-border credit flows to such institutions. More generally, loan-level data that would allow computation of indicators of leverage and vulnerabilities to asset price and interest rate shocks across borrower and asset classes are often not available.³⁸

68. Gaps also hamper the assessment of interconnectedness and the mitigation of risks in the structural dimension. Information is often missing on domestic and cross-border exposures between financial institutions, including between those that are potentially systemically important. Detailed information on counterparties and maturities in funding markets is often not collected. In addition, information on exposures arising in over-the-counter derivatives markets is often missing, reducing the ability of the authorities to assess risk concentrations and design and enforce regulatory constraints, such as on potentially systemically important institutions.

69. Information gaps often also hinder the policing of regulatory boundaries. For instance, there is a lack of information on “Other financial corporations” (OFCs), where typically only statistical data is available and this information only captures very basic indicators, such as the total assets of such institutions. ³⁹ Information is missing in particular on money market mutual funds, where little is known about investment portfolios and maturities (Heath, 2013).

70. Closing information gaps requires improvements in both statistical and supervisory data. This requires not just new data, but also improvements in the granularity, frequency, and timeliness of existing data. In closing information gaps, it is important to ensure homogeneity and comparability of data at the international level and to leverage existing official databases, such as those maintained by the IMF and BIS. Consideration needs also to be given to the costs of information collection for both the financial industry and the official sector, which calls for well-targeted improvements in national and international data. The IMF/FSB/G20 Data Gap initiative and special data dissemination standard (SDDS) Plus represent important advances in this regard (see further Annex VI).

A. Macroprudential Powers

72. Where financial systems evolve dynamically, limiting systemic risk requires powers to foster the ability to act. Where linkages with the real sector and between financial institutions change with time, a macroprudential policymaker requires powers to adjust its approach accordingly. Powers are needed to ensure the policymaker can obtain the necessary information; influence the strength of regulatory constraints placed on the financial system; designate individual institutions as systemically important; and initiate changes in the regulatory perimeter to expand the reach of macroprudential policies to collectively important providers of credit and liquidity (IMF, 2011a).

73. The strength of such powers can vary and be

• “hard” (direct), enabling the policymaker to have direct control over the calibration of specific macroprudential tools,

• “semi-hard,” enabling the policymaker to make formal recommendations, coupled with a ‘comply or explain’ mechanism, or

• “soft,” enabling the policymaker to express an opinion, or a recommendation that is not subject to comply or explain.

Each type of power can be useful and the effectiveness of the policy frameworks can benefit from a combination of these powers.

74. Hard powers are usefully assigned over the calibration of a well-defined set of macroprudential tools.⁴¹ Such powers can avoid delay and other frictions in implementation that arise when there is a need for cooperation by other policymakers. They will therefore often be considered in particular for tools that control the rapid build-up of risks in the time dimension. Direct powers can also increase effectiveness of policy since they enable the policymakers to communicate credibly with the financial markets. Direct powers over a well-defined set of tools provide the macroprudential policymaker with a “stick” that she can credibly threaten to use.

75. The advantage of a power to recommend actions, coupled with a ‘comply or explain’ mechanism, is that it is broad. Recommendations can be used to influence the whole range of regulatory actions that can be taken by other supervisory and regulatory agencies. In particular such recommendations may be used to address the structural component of systemic risk, where macroprudential interventions may be less frequent, or where implementation requires further judgment by the supervisory agency. The ‘comply or explain’ mechanism is important for effectiveness since it increases the chance of compliance and ensures transparency and public accountability as regards cooperation by other agencies. When separate supervisory agencies can point to the recommendation by the macroprudential authority this can also strengthen their hand and help overcome industry opposition or political pressure.

76. Soft powers are useful to extend the influence of the macroprudential policymaker beyond existing prudential tools. A soft recommendation is appropriate when the macroprudential policymaker addresses the legislature to initiate the establishment of new macroprudential tools, or changes in the legal framework to extend the regulatory perimeter. Soft tools, such as ‘opinions’ can also be appropriate when the macroprudential policymaker is concerned that the build-up of systemic risk is fed by broader macroeconomic imbalances. They can then be used by the macroprudential authority to urge policy action by the government to contain such imbalances. However, soft powers alone are unlikely to be sufficient to ensure the effectiveness of the overall policy framework.

77. Information collection powers need to complement powers over policy tools. Information collection powers are needed to close information gaps. In order to avoid duplicative costs on the financial industry, the macroprudential authority should seek to obtain information that is available to other agencies and legal impediments to such exchange of information will often need to be reviewed. Since financial activity can migrate in response to regulation in unintended ways, the policymaker needs to have the power to collect information beyond the regulatory perimeter. It can therefore be useful to establish a broad back-up power that enables the authority to collect information directly from financial firms, such as provided to the Office for Financial Research in the United States.

B. Objectives and Accountability

78. The exercise of macroprudential powers needs to be guided and constrained by a well-defined objective. This can form the basis for a framework to hold the policymaker accountable for achieving the objective. A well-defined objective can also guard against the risk of abuse of macroprudential policy, and its use as a substitute to escape more difficult policy choices in other policy areas, such as fiscal and structural policy.

• The objective can articulate the scope of responsibilities of the macroprudential policymaker in both the time and structural dimension. For instance, it might specify that the policymaker should (i) ensure the overall resilience of the system; (ii) contain risks from unsustainable increases in credit, leverage and asset prices; and (iii) contain structural risks from inter-linkages within the financial system.

• To help ensure that the macroprudential policymaker recognizes trade-offs in the pursuit of financial stability, it can be appropriate to specify secondary objectives, such as the need to maintain the contribution of the financial system to the long-run growth of the economy, or the need to protect the interests of depositors.

79. In addition to a well-defined objective, an accountability framework can include a range of communication tools. Such tools can help the public to establish whether the authority is taking appropriate action to achieve its objective. They can also influence the conduct of the macroprudential policymaker in ways that foster the effective pursuit of the objective.

• Publication of a policy strategy. The framework can encourage the development and publication of a policy strategy that the decision-maker intends to follow in the deployment of specific macroprudential tools over which it has direct control. Such a strategy can generate a degree of commitment by setting out under what conditions these tools would be employed.

• Record of meetings. Where policy decisions are made by a macroprudential committee, the framework can specify the publication of a record of the meetings of the committee that establishes transparency on issues discussed and clarity as regards the votes cast by members on policy decisions.

• Periodic reports. A basic mechanism is a requirement to publish periodic reports on the activities of the macroprudential policymaker, including an assessment of risks and policy actions taken to mitigate the risks.

C. Assignment of the Mandate

80. To strengthen ‘willingness to act,’ it is important that the macroprudential mandate is assigned to someone, a body or a committee (IMF, 2011a). Where a clear assignment is lacking, collective action problems lead to underinvestment in systemic risk identification and mitigation across agencies and reduce accountability, since in the end no one is fully responsible for the crisis outcome.

81. It is desirable for the central bank to play an important role in macroprudential policy (IMF, 2011a; Nier and others, 2011; IMF, 2013a; Viñals, 2011). This can harness the expertise of the central bank in systemic risk identification and its incentives to ensure macroprudential policy is pursued effectively. It can also help shield macroprudential policymaking from political interference that can slow the deployment of tools or bias their use toward other objectives.

82. In practice, these two basic principles lead to the increasing prevalence of three models for macroprudential policymaking:

• Model 1: The macroprudential mandate is assigned to the central bank, with macroprudential decisions ultimately made by its Board (as in the Czech Republic).

• Model 2: The macroprudential mandate is assigned to a dedicated committee within the central bank structure (as in the U.K.).

• Model 3: The macroprudential mandate is assigned to a committee outside the central bank, with the central bank participating on the macroprudential committee (as in Australia, France and the U.S.).

83. The choice of model in any given country will often be driven strongly by traditions as well as political economy considerations. This includes importantly the perceived need for checks and balances in the conduct of macroprudential policy. In addition, the existing arrangements for monetary and supervisory policy as well as legal (constitutional) constraints are likely to play a strong role in shaping the arrangements. The pros and cons of each model, and mechanisms to address the drawbacks of each model are examined in more detail in Nier and others (2011).

84. Model 1 is a natural choice in highly integrated arrangements where the central bank already concentrates the relevant regulatory and supervisory powers. Where supervisory and regulatory agencies are established outside the central bank the assignment of the mandate to the central bank is usefully complemented by coordination mechanisms, such as a coordination committee chaired by the central bank.

85. Model 2 can help counter the risk of dual mandates for the central bank, by creating dedicated decision-making structures for monetary and macroprudential policy even as both functions are under the roof of the central bank. It also allows for participation of separate supervisory agencies and external experts on the decision-making committee. This can foster an open discussion of trade-offs that brings to bear a range of perspectives and helps disciplining the powers being assigned to the central bank.

86. Model 3 can more easily accommodate a desire for a strong role of the Ministry of Finance (MoF). Participation of the MoF can be useful when changes in legislation are needed to expand the macroprudential toolkit or the regulatory perimeter. However, a dominant role of the MoF risks delaying macroprudential action and can compromise the independence of participating agencies, including the central bank and separate supervisory agencies (Nier and others, 2011). Some of these risks can be countered by assigning the central bank the chairmanship (as in Australia), a strong voice (as in Mexico) or a veto over policy decisions (as in Germany). They can also be countered by establishing only soft powers for the decision-making committee.

87. More generally ‘willingness to act’ can be driven by the governance of the decision-making committee, including its voting arrangements. In principle, it is desirable for a macroprudential committee to seek a consensus. However, a formal requirement for unanimity assigns a veto to each of the members of the committee and risks paralyzing macroprudential policy. Simple or qualified majority arrangements can strike a better balance between the need to avoid delay and the need to ensure that the committee has taken full account of trade-offs and different perspectives.

A. Lack of Action

90. Lack of forceful macroprudential action in one country can increase the likelihood of crisis and imposes negative externalities on other countries. There has been a rapid increase in trade and financial linkages across countries over the past few decades (Annex VII). This means that when a financial crisis erupts in one country, negative trade and particularly financial spillovers are felt across regions or even globally. A lack of preventive macroprudential action to contain the build-up of systemic risk can then impose substantial costs on other countries.

91. Strong institutional frameworks at the national level are essential to address biases in favor of inaction. Macroprudential policy is subject to a strong bias in favor of inaction or insufficiently timely and forceful action as risks are building up (IMF 2011a). These can be countered by establishing appropriate powers, strong accountability and the appropriate assignment of the macroprudential mandate. Since strong national mandates can contribute to both domestic and global stability, the IMF is promoting the establishment of such mandates across its membership, and is providing advice through FSAPs and technical assistance (e.g., IMF, 2012a).

92. National frameworks can be buttressed by international guidance and surveillance (IMF 2011a). For instance, the Basel Committee has issued guidance for countries to monitor increases in the ratio of credit-to-GDP and to prepare to increase countercyclical capital buffers in response. The Fund is in a unique position, through its existing programs of surveillance, FSAPs, and technical assistance, to help countries conduct an in-depth assessment of systemic risks, and to advise on preventive macroprudential action in the light of this assessment. Such surveillance can strengthen the hands of national macroprudential authorities in the face of opposition and increase their resolve to take potentially unpopular risk mitigating action.

B. Need for Coordination—Action in the Time Dimension

93. National policies in the time-dimension that are designed to contain risks from a rapid build-up of domestic credit can lead to an increase in the share of credit provided across borders. Cross-border arbitrage can occur through direct cross-border lending and lending by foreign branches, and through a “rebooking” of loans, whereby credit is originated by subsidiaries, but then booked on the balance sheet of the parent institution. These “leakage” effects are well documented empirically (Aiyar and others, 2012) and have complicated macroprudential policies in a number of countries ahead of the crisis (e.g., Bulgaria and Croatia).

94. Addressing cross-border arbitrage calls for “reciprocity” in the imposition of macroprudential action. This ensures that the same constraint is imposed on all relevant credit exposures to borrowers in a given country, whether credit is provided by domestic or foreign entities. Reciprocity is enshrined in the Basel III agreement on the CCB and calls for full cooperation by the home authorities, such that when the CCB is activated in any given country, all countries are meant to apply the buffer to exposures into that country (Annex VIII explains this in more detail).

95. While reciprocity for the countercyclical capital buffer is an important advance, further work may be necessary to expand the scope of reciprocity. In particular, where countries are concerned about the build-up of risks in particular segments of the credit market, such as mortgages or corporate credit, they may want to impose more targeted sectoral measures, such as increases in capital requirements (risk weights) for lending to such segments. However, there is currently no international agreement that would call for reciprocity with regard to sectoral measures, or risks taken in securities markets, while the strength of implementation of reciprocity for the CCB remains untested.

96. In the absence of reciprocity, there is a risk that countries resort to unilateral imposition of capital controls to increase effectiveness of macroprudential actions. However, these measures come with their own distortions. Moreover, since the riskiness of cross-border banking flows increases in the absence of reciprocity, it is in the interests of the home authorities to reciprocate the macroprudential action. International agreements are therefore usefully supplemented by bilateral and regional arrangements. This can help ensure that reciprocity applies with full force and across the range of tools that are needed to contain the build-up of risks in the time dimension.⁴²

C. Need for Coordination—Action in the Structural Dimension

97. Policies to strengthen the resilience of systemic institutions in one country can cause their activities to migrate to other countries. Authorities in each country may want to impose measures to increase the resilience of systemically important institutions, such as capital surcharges for such institutions. However, each jurisdiction may fear that these institutions relocate or that they move their activities to other countries, reducing jobs and tax revenue. This can lead to a race to the bottom in the application of such measures and greater concentration of risky activities in less strictly regulated jurisdictions.⁴³

98. In principle, this issue can be addressed by international agreements and guidance. An important advance is the determination of capital surcharges for global systemically important financial institutions (G-SIFIs) by the FSB. In addition, the Basel Committee has issued guidance for national authorities to assess capital surcharges for domestic systemically important banks (D-SIBs), in an attempt to achieve some degree of consistency of approach.

99. However, such guidance leaves considerable room for national discretion in the application of a framework for dealing with risks from systemically important institutions. Guidance issued by standard setters is usefully complemented by surveillance that can help the authorities assess risks and the need for mitigating action. Such surveillance can also help internalize the costs of failure that are borne in other countries, such as when financial institutions have cross-border subsidiaries and branches that are systemic in the host country.

D. Need for Coordination—Home-Host Issues

100. Where financial institutions have affiliates in multiple jurisdictions, this can lead to conflicts between home and host authorities (FSB, IMF, and WB, 2011).

• Macroprudential actions taken by the home country to improve resilience, such as the imposition of strict liquidity requirements or loan-to deposit ratios, can lead to undesirable deleveraging in the host country.

• Lack of action by home authorities can lead to spillovers when affiliates are systemic in the host country but this is not taken into account in the supervisory approach taken by the home authority.

• Macroprudential actions taken by the host country to constrain systemic risk of foreign branches or subsidiaries can impose costs and a loss of efficiency at the level of the banking group, such as when capital and liquidity are required to be held in and “ring-fenced” by the host country.⁴⁴

101. Where financial institutions operate across multiple borders this complicates the assessment of systemic risk for the home and the host authorities, as well as globally. In countries with a high share of foreign banks and in the absence of information on parent institutions and their exposures, it is difficult to assess domestic systemic risk. Foreign branches, in particular, can become “shadow banks” for the host supervisor.

102. These issues require bilateral and multilateral coordination and consultation. Supervisory colleges can facilitate information exchange among regulatory authorities. They can also foster recognition and understanding of home-host interdependencies. However, they may not overcome the present impediment that national authorities often do not have incentives or even a legal mandate to take full account of the effects of their actions on financial stability in other countries.⁴⁵

E. Multilateral, Regional and Bilateral Solutions

103. Although the case for a cross-border perspective in the application of macroprudential policies appears clear, major challenges still lie ahead. The way forward may be a combination of multilateral, regional and bilateral approaches.

• Multilateral approaches can be based on international agreements or standards, such as the Basel framework. However, the development of international standards is a slow process, and it cannot take full account of local conditions. It therefore needs to be supplemented by surveillance that promotes a faster response to the evolving nature of systemic risk.

• Regional coordination is essential for countries belonging to highly integrated regional financial systems. Where cross-border financial activity is promoted, as part of a single market, or required, as in a currency union, this requires dedicated institutional mechanisms to deal with the associated cross-border challenges.⁴⁶ Examples of such regional coordination mechanisms are the ESRB⁴⁷ and the Nordic-Baltic Macroprudential Forum.⁴⁸ There can also be more ad-hoc structures organized to deal with specific problems, such as the “Vienna initiative.”⁴⁹

• Bilateral approaches require strengthening home-host cooperation in the macroprudential field. Forms of this cooperation may differ, ranging from the cooperation in relation of individual institutions (as in supervisory colleges), to bilateral agreements that coordinate macroprudential measures of the countries involved.