Chapter 2: Optimal Integration of Macroprudential and Other Policies
- Steven Barnett, and Rodolfo Maino
- Published Date:
- April 2013
Incorporating Macroprudential Instruments into Monetary Policy in Peru
Ninety years ago, in the aftermath of World War I, the Reserve Bank of Peru was created with the purpose of providing “an elastic currency,” meaning that the new institution should avoid sharp fluctuations in credit and activity through timely liquidity injections. This event is representative of the many central banks that originated with the aim of ensuring the stability of the financial system as a whole, an aim that after the 2008–09 global economic and financial crisis is gaining a more prominent role.
This section deals with Peru’s experience with introducing macroprudential issues into the design and implementation of the country’s monetary policy.
General Characteristics of the Macroprudential Framework
The aim of macroprudential policy is to limit systemic financial risks to prevent negative impacts on the real sector and on the financial sector as a whole. Central banks and bank supervisors have a number of different instruments at their disposal to prevent a systemic financial crisis, among them reserve requirements, countercyclical provisions, and limits on loan-to-value ratios. These instruments are oriented first toward preventing financial excesses and second to acting when a systemic crisis arises; they are not designed to replace policies needed to maintain macroeconomic stability or the financial health of individual financial institutions.
The main difficulties with macroprudential policies are related to the lack of a formal institutional framework that has both the mandate to set the policy goal and the proper instruments. Also, there is a great difficulty in implementing models that relate those policy instruments to final objectives. This is uncharted territory, and defining new routes through it will depend on the initial conditions and on the specific situations of different economies. Thus, as the IMF (2011) states, “one-size does not fit all.”
In Peru, the policy tools of the Superintendence of Banks and the central bank, unlike their main operational instruments, have been oriented toward the prevention of a systemic crisis, thus extending the scope of their specific mandates.
Dollarization: The Main Financial Vulnerability in Peru
The concern with macroprudential issues reflects a specific vulnerability in the financial system caused by the importance of the financial dollarization of banking deposits and credits. Dollarization was the consequence of the reaction of private agents looking to protect their wealth in an environment of high inflation in Peru between 1976 and 1986 and hyperinflation between 1987 and 1990, a period in which dollar assets were the only way to hedge against rampant inflation. Since 2002, dollarization has been reversed from its level of 80 percent to 45 percent in 2012, as the result of the attainment of price stability and progress in the development of a securities market in local currency.
Financial dollarization introduced two important macroprudential risks. First, sharp currency depreciation exposes firms to important capital losses, eroding their ability to service their bank credits, which could ultimately produce a credit crunch. This process is known as the balance sheet effect of currency depreciation.
A second negative risk of financial dollarization is associated with a run on bank deposits or a sudden stop of the external funding of banks, which, without an effective lender of last resort to provide foreign currency, would became a systemic risk. In this environment, firms have access only to short-term financing in domestic currency, bringing with it the risk of movements of the interest rate; and to long-term financing in foreign currency, with the accompanying exchange rate risk.
Monetary Policy Framework in Peru: Hybrid Inflation Targeting
Monetary policy in Peru incorporated macroprudential considerations following the contagion of the Russian default in 1998, which resulted in a financial crisis in Peru produced by a combination of a sudden stop of capital flows and sharp currency depreciation. The rapid increase in nonperforming loans ended in a severe credit crunch, with substantial systemic and real effects. The number of firms filing for bankruptcy jumped 400 percent, the number of banks was halved, and economic activity and credit only started to recover four years later.
The monetary policy framework designed after this crisis included conventional inflation targeting, but with the important addition of macrofinancial considerations to ring fence the risks associated with financial dollarization. To limit foreign exchange liquidity risks and to avoid credit booms and crunches, reserve requirements are actively managed and international reserves are accumulated to finance an effective lender of last resort in foreign currency. Also, to limit exchange credit risk, the exchange rate regime is an administered floating system, the objective of which is to reduce the volatility of the exchange rate.
An inflation targeting regime was introduced in 2002, with an inflation target of 2 percent. The short-term interest rate was used as its main policy instrument. This regime reduced the extreme volatility in the money market interest rate associated with the previous use of monetary targets and with the attempts to reduce the volatility of the exchange rate using changes in the liquidity position in the money market. In the inflation targeting regime, the policy interest rate is modified as needed to reach the inflation target, and is used as a countercyclical instrument.
The reduction of volatility in the interbank market has another important implication from a macroprudential perspective. Reducing volatility reduces the possibility of failure in this market, which is a source of liquidity funding, during stress episodes.
Macroprudential Considerations in the Design of Monetary Policy
In an environment of financial dollarization, the conventional channels through which monetary policy decisions affect economic activity and inflation may not be sufficient to attain policy objectives. The conventional policy instrument has to be complemented with unconventional instruments. For example, banking credit in dollars is not within the conventional policy framework. Also, a sharp movement of the exchange rate could alter all the monetary policy transmission mechanisms. For those reasons, the use of reserve requirements, sterilized foreign exchange interventions, and the accumulation of international reserves became part of the policy arsenal of the central bank.
To assess the need to apply the unconventional policies, the Central Reserve Bank of Peru relies on a set of statistics that have proved to be good early indicators of macrofinancial problems. For example, a credit-to-GDP ratio in excess of the range around its tendency is a signal of a possible credit boom or crunch. In that moment, there is room to adjust the unconventional quantitative instruments.
Another important indicator is related to the housing market. An analysis of trends and of ratios of prices to annual rent can detect the possibility of a housing bubble. Also, a deviation of the real exchange rate from its equilibrium level creates an incentive to divert resources to nontraded activities. Financing this process with short-term foreign liabilities causes a potential bubble that can explode with systemic consequences. Thus, real exchange rate misalignments can be another important leading indicator of systemic financial problems. The solution to this misalignment requires far more than just monetary actions.
The identified risks and the proposals to control them are analyzed in the semiannual financial stability reports that the central bank has published since 2006.
Reserve Requirements as a Macroprudential Instrument in a Dollarized Economy
The main unconventional monetary policy instrument in Peru is the reserve requirement ratio, which is applied according to the denomination of banks’ liabilities. The ratio is differentiated between bank liabilities in national currency and those denominated in foreign currency. For bank liabilities in foreign currency, the ratio is further differentiated by the source of the liability, deposits in foreign currency, and foreign credit lines. For example, at the time this section was written the reserve requirement ratio for domestic currency deposits was 25 percent, for foreign currency deposits 55 percent, and for short-term external credit lines 60 percent. There are two special cases in this reserve requirement policy. The first refers to bank liabilities in domestic currency to nonresident investors, for which the required ratio exceeds 100 percent to limit the carry. The second special case is the exemption of foreign credit lines with maturities longer than two years.
In normal times, changes in reserve requirement ratios occur much less frequently than changes in the policy interest rate. However, during periods of financial stress, changes in reserve requirements have been used actively, particularly when there have been significant capital inflows or outflows like those before and after the Lehman Brothers bankruptcy in September 2008.
The contrast in the evolution of the credit-to-GDP ratio in Peru during the Lehman bankruptcy event as compared with the Russian crisis of the late 1990s shows the effectiveness of a policy framework that combines conventional mechanisms with more quantitative tools like reserve requirement ratios. The differentiated structure of reserve requirements has also reduced the importance of short-term liabilities in banks’ balance sheets.
International Reserves Accumulation as a Preventive Macroprudential Tool
The sizable international reserves at the central bank have served as a credible source of liquidity injections during sudden stops of capital inflows. Just as important has been the ability and willingness of the central bank to use the reserves during negative financial events.
Any discussion of the appropriate level of international reserves must consider the different risks confronted by the economy and its level of risk tolerance, and must quantify the possible reductions in the level of reserves during different shock events. It is difficult to mechanically define an optimal level of reserves, considering the different circumstances and vulnerabilities faced by each country. For example, in Peru the actual size of international reserves covers six times the amount of short-term bank liabilities, but still does not cover 100 percent of all bank liabilities with the domestic private sector.
Crisis Management and Liquidity Provision by the Central Bank
The reaction of the Central Reserve Bank of Peru after the shock of the Lehman Brothers crisis illustrates the importance of preventive measures oriented toward accumulating international liquidity and avoiding the ignition of mechanisms that could have ended in a credit crunch (Rossini and Quispe, 2010).
The total injection of liquidity to banks during the last quarter of 2008 was equivalent to 9 percent of annual GDP; the measures included foreign exchange sales to provide liquidity in foreign currency for an equivalent of 6 percent of GDP. The banks’ need for systemic liquidity was covered and complemented with one-year repos and the liberation of funds through reductions of reserve requirements.
Despite the significance of this external shock, banking credit to the private sector continued to grow in Peru in 2008 and in 2009 instead of falling as it did in other countries in Latin America, showing the importance of both the previous preventive accumulation of international liquidity and the proactive and sizable injections of liquidity during the crisis.
The 1998 financial crisis, which had a negative impact on activity, on the financial system as a whole, and on the effectiveness of monetary policy, created a special policy sensibility geared toward avoiding any other credit contraction event. This sensitivity has aided the resilience of the Peruvian economy to the recent global crisis.
Sterilized Intervention to Reduce Volatility and Not to Fix the Level or Trend of the Exchange Rate
Apart from the use of reserve requirement ratios and international reserves accumulation, the central bank in Peru includes sterilized foreign exchange interventions in its policy toolbox to avoid sharp fluctuations in the exchange rate that could ignite a balance sheet effect in the corporate sector. Foreign exchange intervention is not performed according to any rule, and avoids ensuring any target level or any target range. It is far from being some form of fixed exchange rate.
A comparison of the evolution of the nominal exchange rate index in Peru with an average index of exchange rates in Latin American countries that follow inflation targeting regimes illustrates that Peru has been able to eliminate extreme levels without affecting the tendency. The avoidance of extreme variations in the exchange rate has been important to preventing financial disruptions in dollarized firms, and also to discouraging the adoption of open cross-currency speculative positions that could have endangered the stability of the financial system when those positions were sizable and highly leveraged.
An important issue with sterilized foreign exchange interventions is the disruption that can emerge as the size and cost of the central bank’s liabilities grow. In Peru, the succession of fiscal surpluses during periods of favorable tax earnings allowed accumulation of public sector deposits at the central bank as the main source of sterilization equivalent to 10 percent of GDP, reducing the relative importance of placements of central bank securities.
Building a Solid Financial Infrastructure: The Treasury Bond Market
As mentioned, efforts to reduce systemic risks include preventive macroprudential measures and actions to forestall a financial crisis. The development of financial infrastructure is also important to building a more resilient financial system, able to offer more options to diversify risks.
An important development to enhancing long-term financing in domestic currency was the launch in 2001 of long-term treasury bonds in domestic currency, including maturities longer than 30 years. This allowed a benchmarking government debt yield curve to be established as reference for private securities in domestic currency, similar to mortgage loans and corporate bond issuances.
Governance Aspects of the Macroprudential Framework in Peru
The institutional macroprudential framework in Peru is not based on a formal mandate, but occurs in coordination meetings to review risks and actions to control them. The central bank produces and publishes a financial stability report to provide information about risks and possible policy actions. This framework takes into account the independence of the different participants, but raises the possibility of a coordination failure.
The introduction of macroprudential aspects to the design and implementation of the central bank’s monetary policy is accompanied by macroprudential considerations in the microprudential actions of the Superintendence of Banks. For example, in 2008 the superintendence introduced a countercyclical credit provisioning system on the basis of cumulative GDP growth, with the aim of avoiding excesses in banking credit.
In light of the credit risk of currency depreciations, the superintendence has included additional provisioning requirements for banks that extend foreign exchange credits to borrowers without a cash flow in that currency. Also, to reduce exchange risks, limits to open foreign exchange positions of banks are in place. The introduction of an additional capital requirement to prevent excessive loan to value in mortgage loans is also under consideration.
Conclusion: A Work in Progress
The Central Reserve Bank of Peru has a prominent role in the design and implementation of macroprudential policies given its position in the payment system, its access to relevant information to identify systemic financial risks, and its interest in defending the transmission mechanisms for monetary policy.
Conventional monetary policy instruments can be insufficient for preserving macroeconomic stability when financial channels are strong enough to create excesses like credit booms or asset inflation (Borio, 2011a). In these circumstances, the possibility of a systemic crisis would negatively feed back on economic activity and on the stability of the financial system as a whole. Because of the existence of a partially dollarized financial system in Peru, the central bank uses quantitative instruments to respond to early indicators of substantial and autonomous movements of liquidity and credit that may generate systemic risks. The use of reserve requirements, sterilized foreign exchange interventions, and changes in the maturity of its net assets proved to be effective at preventing either a credit boom or a credit crunch during the Lehman event.
However, central banks normally have a limited set of policies at their disposal for maintaining macrofinancial stability. Given that other independent entities have instruments, mainly for microprudential use, there is room to work on formal institutionalization of the macroprudential framework in Peru.
Incorporating Macroprudential Instruments into Monetary Policy: Thailand’s Experience
Macroprudential Policy Formulation at the Bank of Thailand
The role of central banks in preserving price stability and fostering financial stability has evolved over time, and this is true for the Bank of Thailand as well. In the past decade, the bank has gone through many changes, both in legislation and governance and in its monetary policy and supervision framework. With regard to macroprudential measures, the painful lessons from the Asian crisis in 1997 showed that financial imbalances, which can emerge in a period of robust economic expansion, may reveal the vulnerability in the economic system and impede sustainable growth.
Mindful of this, the Bank of Thailand’s Monetary Policy Committee has made “the monitoring of factors contributing to financial imbalances” a part of policy formulation since July 2004. Signs of financial imbalances in seven key areas are monitored. An entire chapter on financial stability conditions and outlook was added to the bank’s inflation report beginning in July 2005. The goal of this surveillance is to ensure constant vigilance against threats to financial stability, even in the absence of clear signs of inflationary pressure or asset-price bubbles.
The inclusion of a financial stability analysis in the inflation report reflects the bank’s conviction that such considerations are important for monetary policy. The Bank of Thailand has long recognized the contribution that monetary policy can make to limiting the buildup of financial imbalances and that price stability alone may not be sufficient to ensure financial stability. In the bank’s experience the issue has never been whether monetary policy should take into account financial stability concerns, but how best to achieve this end. Efforts to develop the analytical framework and to improve data quality to ensure comprehensive assessment of such risks are ongoing.
Illustration of Thailand’s Macroprudential Policy
The Bank of Thailand has made use of macroprudential tools for some time to complement monetary policy in managing potential buildups in financial imbalances. The use of these tools, along with a flexible inflation targeting framework, contributed to the resiliency of the Thai financial system. The macroprudential policies implemented include the following:
tightened regulation of credit card loans and personal loans in 2002, 2004, and 2005;
net foreign exchange position implemented in 2002;
loan-to-value ratio on mortgage loans implemented in 2003, and reduced in 2009, and 2010;
loan-loss provisioning implemented in 2006 and 2007; and
withholding tax imposed in 2010.
Tightened Regulation of Credit Card Loans and Personal Loans
The first example is the use of macroprudential policy to address sectoral imbalances in alignment with monetary policy and microprudential regulation. Against the backdrop of the 1997 Asian financial crisis, weak corporate loan demand (partly caused by the many corporations that were still in the debt-restructuring process) amidst ample liquidity in the banking system put considerable pressure on banks to tap the consumer finance business more aggressively. This led to excessive growth of credit card lending, rising from a growth rate of −3.1 percent in 2000 to 76.7 percent in 2002. During this period the same situation could also be observed in other Asian countries, such as the Republic of Korea, where LG Credit Card company started to show signs of problems.
Against this backdrop, the Bank of Thailand decided to tighten regulation on credit card and personal loans. This decision was implemented incrementally, starting by including nonbank credit card companies in the scope of supervision, followed by minimum requirements for credit card holders, such as maximum credit line and minimum repayment. Similar rules were applied to personal loans in 2005. As a result of these measures, the growth of credit in these markets declined remarkably. This experience shows that macroprudential policy, when implemented properly, can tackle sectoral imbalances directly, and can be used as an alternative to conventional monetary tightening that might have broadly based, but unintentional and undesirable, effects on other economic sectors.
During the early years of the first decade of the 2000s, the Thai economy was still in a fragile stage, not yet fully recovered from the 1997 financial crisis. Therefore, the policymakers of the time decided to adopt more-targeted macroprudential measures to tackle the problem, rather than hiking the policy interest rate, which would have had an adverse impact on the whole economy. The interest rate can be a blunt instrument, particularly when an economy is recovering. While imbalances were growing in credit cards and consumer loans, had the Bank of Thailand raised the policy rate too fast the recovery process might have been stalled.
Net Foreign Exchange Position Implemented
Because of the rising and comparatively high foreign exchange exposures of some financial institutions, the net foreign exchange position, for both individual currencies and on an aggregated basis, has been regulated since 2002. This regulation aims to help contain foreign exchange risk in the Thai banking system. Under this rule, financial institutions are required to maintain a net foreign exchange position of an individual currency not exceeding 15 percent of total capital, and the institution’s aggregate position must not exceed 20 percent of total capital. These measures have been successful; both individual and aggregate currency limits have remained well below the regulatory requirement. In particular, net foreign exchange positions declined significantly during the crisis in 2008. This outcome also reflected the ability of financial institutions to manage foreign exchange risk and adjust their exposures to be appropriate for the changing global environment.
Loan-to-Value Ratio on Mortgage Loans
The use and adjustment of the loan-to-value ratio by the Bank of Thailand demonstrated its preventive nature and, more important, the flexibility to fine tune the policy in response to changing economic circumstance. In 2003, the 70 percent limit for mortgages with a value of at least 10 million Thai baht was first introduced to preempt speculation in the luxury real estate segment. Then, amid the global downturn in 2009, the 70 percent loan-to-value limit was replaced with a more risk-sensitive rule that imposes a higher risk-weighted capital charge for mortgages with loan-to-value ratios greater than 80 percent.
Recognizing its effectiveness, in late 2010 the Bank of Thailand extended the loan-to-value rule to mortgages of less than 10 million baht. Two policy considerations went into the design of the rule. First, the rule prevents excessive risk taking by banks during periods of intense competition in the mortgage market (although there was no obvious sign of a property price bubble in this instance). Second, the rule was purposely made more stringent for high-rise mortgages than low-rise mortgages, going into effect one year earlier, because of the more intense competition in the high-rise segment. The loan-to-value rule for high-rise mortgages went into force at the beginning of 2011; however, the implementation date for low-rise mortgages was postponed a year from the original date of January 1, 2012, in light of the severe flood situation in Thailand. In addition to the loan-to-value measure, the Bank of Thailand also requires banks to report the details of real estate loans valued at more than 100 million baht to make sure that the loans are prudent and not speculative.
Loan-Loss Provisioning in 2006 and 2007
The loan-loss provisioning measure is a macroprudential policy adopted in the context of “leaning against the wind” (or countercyclicality). During the extended periods of profitability in the banking industry in the middle of the first decade of the 2000s, the Bank of Thailand gradually tightened loan loss provisioning rules in accordance with International Accounting Standard 39. The objective of this measure is to have banks build buffers during good times against impaired assets. As a result, Thai commercial banks remain strong and resilient, as reflected by strong profits and rising nonperforming loan coverage ratios.
Imposition of Tax on Foreign Investment in Bonds
The last macroprudential policy is the imposition of withholding tax in 2010. In October 2010, due largely to speculative inflows, the Thai baht appreciated rapidly (approximately 11 percent from the beginning of the year), threatening exports. In response, the government reintroduced a 15 percent withholding tax on interest and capital gains earned by foreign investors on domestic bonds in an effort to curb the currency appreciation. The number of nonresident bond holdings increased notably from less than 2.5 percent in early 2010, to approximately 5.0 percent in October 2010. This measure aims to slow down so-called hot money flows into the bond market.
Practical Challenges: Coordination and the Appropriate Policy Mix
For the future, it is apparent that a combination of monetary policy and prudential policies, micro and macro, is necessary to effectively safeguard financial stability and macroeconomic stability at large. The policy is likely to be better coordinated if regulatory power remains with the central bank. Experience in Thailand has shown that important synergies are to be gained from having the supervision responsibilities within the central bank. Chief among these are the cross-sharing of information and expertise as well as the ability to internalize possible trade-offs between monetary and financial stability concerns in forming the respective policies. This is by no means a fool-proof arrangement, and there is definitely a trade-off, but cost-benefit considerations point to a single authority with a broad mandate.
Although monetary and regulatory authorities are under the same roof in Thailand, attaining effective policy coordination and information sharing have proved challenging at times. Achieving the synergies mentioned above is easier said than done. A key area that is still evolving is the appropriate governance structure. To further support the work between the Monetary Policy Committee and the Financial Institutions Policy Committee, an internal subcommittee on financial stability was recently established to more formally coordinate financial stability analysis from a macroeconomic perspective at the staff level. This subcommittee brings together not only the surveillance capabilities from both the monetary and financial-institution sides of the bank, but also takes input from outside regulatory bodies that oversee institutions that are outside the Bank of Thailand’s mandate but may nevertheless contribute to financial stability. The work of this subcommittee will serve as a key input for the policymaking bodies of the Bank of Thailand.
Choosing the appropriate policy tools at the right time requires careful balancing. Ultimately, the policy interest rate still remains the primary tool for influencing aggregate activity and inflation, and prudential tools are employed to deal with more specific problems in the financial sphere. Weighing the benefits and limitations of each tool is a delicate but necessary challenge. It is important not to exaggerate what macroprudential policy can achieve. Macroprudential measures can be employed to complement monetary policy but are not a substitute for it. In the end, it is the interest rate that sets the price of leverage.
Shared Responsibility, Importance of Price Stability and Independence
Finally, three relevant points must be stressed:
First, the preservation of the financial system is ultimately a shared responsibility. With regard to price stability, the central bank has at its disposal monetary policy instruments to attain price stability. Nevertheless, the central bank by itself does not have sufficient instruments and powers to ensure the stability of the financial system. The stability of the financial system also depends on the action of other regulators as well as on the attitudes and practices of market participants.
Second, notwithstanding the financial stability mandate, the conduct of monetary policy with the aim of achieving price stability over the medium and long terms should remain the main responsibility of any central bank. The performance of financial stability tasks should be consistent with and supportive of the preservation of price stability.
Third, having a larger mandate unavoidably puts central banks in a more complex position because larger segments of the public will have more interest in the banks’ actions. It is likely that the broader the range of responsibilities for financial stability assigned to the central bank, the greater the degree of interaction between the government and the public in general. Central bank independence, therefore, becomes more important to ensuring the ability of central banks to implement unpopular (mostly countercyclical) policies in a timely, independent, and transparent manner.
Macroprudential Policies and Their Effectiveness: Hong Kong SAR
As the various monetary and fiscal programs introduced by the United States and Europe took effect following the collapse of Lehman Brothers in 2008, there were strong flows into the Hong Kong dollar in 2009. The Hong Kong dollar monetary base almost tripled its precrisis level, with significant increases in the aggregate balance of the banking sector and the issuance of additional exchange fund bills to unprecedented levels. One direct impact of these inflows on the banking sector was that the cost of funds for Hong Kong banks dropped to an exceptionally low level, and has since stayed below 0.5 percent. Banks’ lending rates also declined, with their mortgage interest rates remaining low at the 1.0–2.5 percent level.
In the same period, significant upward inflationary pressures occurred, leading to negative real interest rates in Hong Kong SAR. Boosted by low borrowing costs and high inflation expectations, residential property prices in Hong Kong SAR have increased drastically since 2009, after a significant decline during the global economic and financial crisis in 2008–09. The surge in prices has been particularly strong for luxury apartments. One conventional way to assess property price levels is to compare them with their peak levels before the Asian financial crisis of the late 1990s. As shown in Figure 2.1, the current average prices for both the luxury market (flat size at or greater than 100 square meters) and the mass market (flat size less than 100 square meters) are already higher than their historical peaks.
Figure 2.1.Hong Kong SAR: Residential Property Prices and Transaction Volume
Source: Hong Kong Monetary Authority.
Because banks are major providers of funds to property investors, credit growth has been very strong in the Hong Kong banking sector. The year-on-year growth of domestic lending was about 26 percent in March 2011 as shown in Figure 2.2. This credit boom is due to the strong demand for residential mortgages as well as increased lending to Mainland China. With a significant and persistent increase in mortgage lending, the household debt burden, defined as total household loans to annual nominal GDP, also shows an increasing trend, indicating higher financial risk in the household sector.
Figure 2.2.Hong Kong SAR: Credit Growth and Household Leverage
Source: Hong Kong Monetary Authority.
Note: AI = authorized institutions.
The developments have raised concerns about the risk of credit-asset price spirals. Because property-related lending usually accounts for 40–50 percent of total domestic loans, such credit-asset price spirals would have systemic impacts on the banking system. In Hong Kong SAR’s experience, the rapid growth in domestic lending in 1996–97 was followed by a sharp fall of property prices (Figure 2.1) and a significant rise in the nonperforming loan ratio from a low of 2.1 percent to a peak of 10.6 percent at the end of September 1999. The credit loss after rapid credit growth could create a huge burden on banks’ balance sheets for years after the boom.
Given the past evidence and risk of possible credit-asset price spirals, loan-to-value caps become an important macroprudential measure for mitigating the risk in the banking sector. A loan-to-value ratio of 70 percent on all properties was first introduced in 1992. Against a sharp increase in property prices in 1996, the Hong Kong Monetary Authority (HKMA) implemented the first tightening of the loan-to-value cap in January 1997 from 70 percent to 60 percent for luxury properties with a value of more than HK$12 million; the cap remained at 70 percent for properties with lower values.
In the wake of the Asian financial crisis, Hong Kong SAR’s property prices fell by more than 40 percent in one year from September 1997. Notwithstanding the sharp drop in property prices, the subsequent mortgage delinquency ratio never exceeded 1.5 percent, a low level by international standards. Judging from this, the loan-to-value policy is effectively reducing the credit risk that banks face and ensuring the quality of banks’ mortgage loan portfolios. In 2002, the HKMA suspended the tighter loan-to-value cap for luxury properties, reverting to the 70 percent rate for all properties.
There have been four rounds of tightening of the loan-to-value policy in Hong Kong SAR since October 2009 amid strong capital inflows and sharp rises in property prices.1 In view of the risk of an overheated property market and declining housing affordability, in June 2011 the HKMA undertook another round of tightening for residential properties valued at HK$7 million and above. In addition, for cases in which the mortgage payment is not supported by incomes from Hong Kong SAR sources, the loan-to-value caps were lowered by at least another 10 percentage points. Reflecting the impact of the multiple rounds of macroprudential measures so far, the average loan-to-value ratio of new mortgages fell by 4 percentage points from 2010, to 56 percent in the first half of 2011.
Wong and others (2011) conducted a cross-country econometric assessment of the effectiveness of the loan-to-value policy. Among the 13 selected economies in the study, those adopting a loan-to-value policy are found to have a lower sensitivity of mortgage delinquency ratio to property prices, indicating that the loan-to-value policy improves the banking sector’s resilience to property market shocks. The study also discusses whether loan-to-value policy should be adopted as an instrument to stabilize property markets. Only weak evidence suggests that the tightening of loan-to-value caps would be followed by lower property market activities. This conclusion holds for three common indicators of property market activities, including the rise in property prices, the property price gap (from the trend), and property transactions.
The finding seems to suggest that the loan-to-value policy may not be an effective instrument for targeting property prices. In contrast, the study finds strong evidence that a tightening of loan-to-value caps can reduce household debt leverage, which supports the proposition that the loan-to-value policy is a suitable instrument for containing risk in the banking sector. The policy effect is mainly transmitted through reducing household debt leverage, rather than constraining the property market, to contain the risk in the banking sector. The results are consistent with those in Wong and others (2004), which finds that prudent lending practices in Hong Kong SAR, guided by loan-to-value limits, were credited only for pausing the house price boom briefly in 1994. However, their contribution to guarding the system against the fallout from the crash in 1997 was clearly identified.
Other Macroprudential Measures
In addition to the loan-to-value policy, the HKMA introduced a stress-testing element to the debt-servicing ratio limit for residential mortgages. Under this requirement, banks need to calculate unstressed and stressed debt-servicing ratios for borrowers.2 For the unstressed ratio, which assumes the current level of interest rates, borrowers’ debt-servicing ratios are subject to a cap of 50 percent. For the stressed ratio, which assumes a rise in mortgage rates of at least 2 percentage points, borrowers’ debt-servicing ratios are limited to 60 percent.3 This macroprudential measure can alert borrowers and banks that a sudden rise in interest rates could reduce borrowers’ repayment ability and increase default risk significantly.
On a broader scale of containing risk in the banking system, the HKMA also requires banks to hold regulatory reserves against latent credit risk. The main objective of regulatory reserves is to address the insufficiency of accounting standards for banks’ provisioning, stemming from a loss-incurred model and backward looking characteristics.
Under current accounting standards, provisions are set aside only when there is material and objective evidence of loan impairment. The HKMA introduced regulatory reserves in 2005. Banks are required to hold regulatory reserves in excess of individual and collective impairment allowances. The regulatory reserves usually range between 0.5 percent and 1 percent of total loans. Although the setting of the level of regulatory reserves is largely a matter of judgment, banks’ historical loss experiences are a very important reference. Figure 2.3 shows the latest data for the ratios of regulatory reserves to total loans for selected Hong Kong SAR banks. Because strong and persistent credit growth raises concerns about latent credit risk, many banks are required to hold higher regulatory reserves despite their current sound asset quality.
Figure 2.3.Hong Kong SAR: Banks’ Regulatory Reserves
Source: Hong Kong Monetary Authority.
The rising property prices also caused the Hong Kong SAR government to increase the stamp duty in November 2010. The stamp duty is a tax on property transactions. The special stamp duty ranges from 5 percent to 15 percent depending on the holding period of properties.
Effectiveness of the Macroprudential Measures
Although the various macroprudential measures were introduced a relatively short time ago, some of their effects can be observed. First, for the banking sector, the growth of mortgage lending leveled off in 2011 (see Figure 2.4). Households also reduced their debt-servicing burden moderately. Thus, the risk in the banking and household sectors seems to have been further contained.
Figure 2.4.Hong Kong SAR: Mortgage Lending and Debt-Servicing Burden
Source: Hong Kong Monetary Authority.
Regarding the property market, property prices as shown in Figure 2.5 appeared to have declined moderately since September 2011. For property transactions, the blue bar in Figure 2.5 suggests an obvious reduction in the transaction volume after the implementation of the special stamp duty. This stage shows some early signs of intended policy effects on the market. It will take observation over a longer period to assess the overall effectiveness of the measures for containing risks in the banking system.
Figure 2.5.Hong Kong SAR: Property Market Activity
Source: Hong Kong Monetary Authority.
Note: In Hong Kong SAR, an “estate” is a cluster of high-rise residential buildings
Macroprudential Policy Instruments in Singapore
Ng Chuin Hwei
This section sets out some views about macroprudential policy instruments in Singapore and describes Singapore’s recent experience in implementing demand-side and supply-side policy measures for the property market.
Singapore’s use of macroprudential instruments has centered on the housing market, based on the systemic risks it could pose. Systemic risk from the property sector could manifest itself through two channels—the credit and leverage risk channel4 and the asset-price inflation risk channel.5 The policy objectives served by the macroprudential instruments are to promote a stable and sustainable property market in which prices move in line with economic fundamentals, encourage greater financial prudence among property purchasers, and maintain sound lending standards.
A range of macroprudential instruments has been used to address systemic risks posed by the housing market. These instruments operate through different channels and may be implemented by different authorities. The term “macroprudential” is less a defining characteristic of a certain group of instruments than a description of how a policy instrument might be used to target systemic risks. Some of the policy measures commonly used could be seen as microprudential measures wielded in a macroprudential way: they mitigate risks at the individual institution level, but when applied across institutions, serve to mitigate risk at the systemic level. Other policy instruments, even nonprudential measures like transaction taxes and supply-side measures, can have macroprudential objectives and effects.
Targeting Segments and Sectors Causing Systemic Risk
The heterogeneity of macroprudential instruments offers the policymaker versatility in using these instruments in targeted and specific ways. In contrast, the experience with monetary policy suggests that the use of broad policy instruments like exchange rates and interest rates may result in unintended spillover effects and their impacts are often subject to long and variable lags before affecting asset markets. Such broad policy tools may not afford sufficient precision if policymakers want to address systemic risks that originate from specific sectors (such as the housing market) without adverse spillover effects on other developing sectors of the economy. In Singapore’s case, macroprudential policies have been targeted at the housing sector.
Targeting Different Groups of Buyers
This targeted approach is also reflected in the way the measures are calibrated to target demand from three different customer segments: property speculators, investors, and other buyers. For instance, the seller stamp duty, with higher tax rates for resales within a shorter period, aims to discourage short-term speculative purchases. Lowering the loan-to-value ratio for non-individual buyers—and for those with more than one outstanding housing loan—targets property investors without affecting first-time home owners, whereas lowering the loan-to-value ratio for all buyers has a greater impact on overall demand.
Targeting Different Risks
Adopting a targeted approach does not mean that macroprudential instruments should be pigeon-holed into a narrow “one tool–one policy objective” model. Indeed, a single policy tool like loan-to-value ratios could address multiple risks. For instance, loan-to-value ratios can encourage financial prudence among borrowers and provide a buffer for banks should prices correct. This helps address systemic credit and leverage risk concerns in cases in which excessive credit growth could affect banks’ asset quality and high leverage could leave banks with insufficient buffers against sudden turns in the market. Loan-to-value ratios also simultaneously serve to moderate asset-price inflation to the extent that it is driven by credit growth. This helps address asset-price inflation risk when a decline in sentiment could lead to rapid price corrections that could pose potential financial stability risks (e.g., through an adverse impact on the asset quality of property-related loan exposures and the devaluation of property collateral).
Targeted demand-side macroprudential measures like loan-to-value ratios and seller stamp duties have generally been preferred for dealing with exuberance in the short term, given that supply-side measures operate with a greater lag and are not easily reversed. Nonetheless, supply-side measures like land sales are important for mitigating asset-price inflation by addressing any fundamental mismatch between demand and supply. At the same time, demand-side and supply-side tools indirectly address the credit and leverage risk channel by moderating price increases and thereby potential strains on household debt-servicing ability.
In adopting macroprudential policies, Singapore, like other economies in Asia, has taken a targeted approach. Instead of relying on a single instrument, several tools have been used simultaneously. Together, these instruments have proved useful in limiting excessive leverage and promoting prudent borrowing and lending behavior, in turn, mitigating the risk of disruptive price adjustments. The experience so far suggests that macroprudential tools offer policymakers considerable flexibility, and that the trade-offs involved in macroprudential policy may not be as restrictive as in a one tool–one policy objective paradigm.
Incorporating Macroprudential Instruments into Monetary Policy: The Indonesian Experience
A crisis always brings new lessons for central banks. The economic downturn during the Great Depression of the 1930s placed central banking under the control of fiscal authorities for nearly two decades afterward. The Great Inflation of the 1970s had the opposite effect, restoring independence into central banking. High inflation in the early 1980s caused central banks to establish monetary policy frameworks solidly anchored by price-stability mandates and safeguarded by independence. The Asian crisis of the late 1990s also led to Asian central banks, including those of Indonesia, the Republic of Korea, the Philippines, and Thailand, to adopt inflation targeting frameworks after the collapse of exchange rate bands.
The 2008–09 global economic and financial crisis and its subsequent events also highlight some important issues about the role of the central bank in the macroeconomy and the way monetary policy should be conducted (Taylor, 2009). The financial crisis paradoxically occurred during a time when the global economy managed to achieve its best performance in maintaining price stability and economic growth. Although this golden age of inflation targeting contributed to low inflation, the narrow focus on price stability failed to account adequately for financial sector risks.
The key lesson from the recent crisis is that central banks should explicitly and systematically consider financial sector risks when setting monetary policy. Although monetary policy should focus on price stability, central banks should also have a set of macroprudential instruments for dealing with financial stability. However, because monetary policy and macroprudential policy reinforce each other, policy actions must be coordinated to get an optimal policy result. Furthermore, in some cases, the two policies may be required to be implemented in tandem. Recent Indonesian experience with large and volatile capital flows has been an example of the need for a policy mix between monetary and macroprudential policies.
This section highlights the post–global crisis monetary policy framework in Indonesia, highlighting how to incorporate macroprudential instruments into the monetary policy framework. Bank Indonesia’s recent experience in implementing a mix of monetary and macroprudential policies is also discussed.
Precrisis Monetary Policy Framework
Mishkin (2011) argues that there are nine basic scientific principles, derived from theory and empirical evidence, that guide thinking at almost all central banks: inflation is always and everywhere a monetary phenomenon; price stability has important benefits; there is no long-term trade-off between unemployment and inflation; expectations play a crucial role in the determination of inflation and in the transmission of monetary policy to the economy; real interest rates need to rise with higher inflation (the Taylor Principle); monetary policy is subject to the time-inconsistency problem; central bank independence helps improve the efficiency of monetary policy; commitment to a strong nominal anchor is central to producing good monetary policy outcomes; and financial friction plays an important role in business cycles.
The monetary policy strategy referred to as flexible inflation targeting (Svensson, 1999) was, in fact, built on the first eight principles of the new neoclassical synthesis. It involves a strong, credible commitment by the central bank to stabilize inflation in the long term, but also allows output to stabilize around its natural rate in the short term.
Many central banks have adopted inflation targeting frameworks since 1990. The framework advances two main rules. First, inflation and output gap forecasts should be viewed as summary statistics of the state of the economy. Second, the policy instrument should be fine-tuned so that inflation forecasts are stabilized and output volatility is minimized. However, limiting the information set to inflation forecasts and the output gap can be highly misleading. Output gaps have long been known to be ill defined and subject to a great deal of measurement error.
The Central Bank Act of 1999 gave Bank Indonesia independence and a mandate to pursue currency stability.6 From 2000 to 2004, Indonesia implemented some form of inflation targeting, with a role for output stabilization and base money as an operational target.7 Regimes such as these are sometimes referred to as “inflation targeting lite” because they are eclectic and incorporate some, but not all, of the key features of inflation targeting (Stone, 2003). The amendment of the Central Bank Act in 2004 provided a clearer framework in which the inflation target is set by the government and monetary policy is guided by the target. Since 2005, Bank Indonesia has adopted full-fledged (flexible) inflation targeting and has used Bank Indonesia’s rate as the policy rate.
Empirically, evaluations of the implementation of inflation targeting in Indonesia since 2004 have recorded a number of noteworthy outcomes (Juhro and others, 2009). Despite missing the targets in 2005, 2008, and 2010, primarily as the result of supply shocks (oil shocks and the food price shock), the framework successfully strengthened the monetary policy decision-making process; improved monetary policy signals to influence inflation expectations and thus accelerate the transmission mechanism of monetary policy (Satria and Juhro, 2011); and improved the policy credibility of the central bank as demonstrated by the declining sacrifice ratio, representing an imperfect measure of the trade-off between inflation and aggregate output.
Postcrisis Monetary Policy Framework: Toward Integration of Macroprudential Instruments into the Monetary Framework
Before the 2008–09 global economic and financial crisis, the common wisdom was that achieving price and output stability would promote financial stability. Bernanke and Gertler (2001), for example, argued that monetary policy, which optimally stabilizes inflation and output, is likely to stabilize asset prices, making asset-price bubbles less likely.
The crisis, however, revealed that maintaining low inflation alone, without taking into account financial stability, is insufficient for achieving macroeconomic stability. This lesson is not new for Asian central banks. A number of crises that have occurred in recent decades showed that macroeconomic instability is primarily rooted in financial crises. Financial markets are inherently imperfect and potentially create excessive macroeconomic fluctuations if not well regulated. Therefore, the key to managing macroeconomic stability is not only managing internal and external imbalances, but also imbalances in the financial sector, such as excessive credit growth, asset-price bubbles, and the cycle of risk-taking behavior.
The recent crisis clearly demonstrates that benign economic environments may promote excessive risk taking and may actually make the financial system more fragile (Gambacorta, 2009). Although price and output stability are surely beneficial, a policy focused solely on these objectives may not be enough to produce good economic outcomes. It is clear that during the global crisis, price stability may have encouraged the accumulation of risk in the financial sector, such as excessive credit growth and asset-price bubbles (Blinder, 2010). Stable macroeconomic conditions reflected in a long period of low interest rates created moral hazard among market participants against macroeconomic risks. Investors felt that the macroeconomic risk was already guaranteed by the credible central bank; therefore, they tended to seek higher yields in higher risk assets.
Does inflation targeting remain relevant? Mishkin (2011) argues that none of the lessons from the financial crisis in any way undermines or invalidates the nine basic principles of the science of monetary policy developed before the crisis. The lesson that developments in the financial sector can have a large impact on economic activity does indicate that the ninth principle on the role of financial frictions in creating macroeconomic fluctuation is, of course, valid, but now is even more important than central bankers previously realized. According to this line of thinking, the inflation targeting framework should incorporate macroprudential features to support monetary and financial system stability.
How to Integrate Flexible Inflation Targeting and Macroprudential Policy
The principle behind the integrating macroprudential policy into inflation targeting is that because the financial sectors are inherently procyclical and procyclicality aggravates macroeconomic fluctuations, countercyclical monetary and macroprudential policies should be mutually reinforcing (see Figure 2.6). Macroprudential measures, such as countercyclical capital asset ratios, to reduce excessive credit growth, for example, could support monetary policy by cooling down the overheating economy without excessively tightening monetary policy. This method is particularly needed during periods of large capital flows when monetary policy is facing a dilemma (Agung, 2010).
Figure 2.6.Monetary and Macroprudential Policy to Dampen Procyclicality
Source: Author’s calculations.
For its part, monetary policy can support financial stability by a (mild) reaction of interest rates to excessive risk behavior in the financial sector. The existence of the risk-taking channel of monetary policy clearly supports the view that monetary policy should lean against the build up of financial instability. Support from monetary policy is also necessary given that heavy reliance on macroprudential policy may also foster undesirable outcomes, such as limiting credit availability and causing financial disintermediation (Agénor and da Silva, 2012).
The efficacy of policy coordination depends upon the macroeconomic environment, financial conditions, and the intermediation process, as well as the levels of capital and assets in the banking system. Therefore, it is unrealistic to expect a combination of monetary and macroprudential policy to eliminate the economic cycle in its entirety. The overarching goal of policy integration is to manage the cycle and augment financial system resilience on a macro scale.
Given that this area is still uncharted territory, efforts should be made to improve the understanding of the link between monetary and macroprudential policy. The following agenda should be outlined in central banks:
First, how the monetary and macroprudential policy transmission mechanisms affect economic activity must be understood, which requires a more integrated analysis framework, particularly for calculating the significant role of the financial sector.
Second, accurate measurement of the risks of financial imbalances is needed. Such indicators will also strengthen analysis of the transmission mechanism through the risk-taking channel. For example, Basel III recommends use of the credit-to-GDP gap to measure the financial cycle.
Third, a macroeconomic model that incorporates the financial sector needs to be developed to better understand the link between macroeconomic and financial sector issues and to guide policy decisions. Bank Indonesia is currently developing a core model that includes the financial sector.
Fourth, from an institutional perspective, coordination among the different agents is necessary to make the policy mix effective, particularly if responsibilities for microprudential, macroprudential, and monetary policies are institutionally separated. In Indonesia, a new Financial Services Authority (FSA) Act clearly states that microprudential regulation is under FSA jurisdiction and macroprudential regulation is under the jurisdiction of Bank Indonesia. Furthermore, FSA should coordinate with Bank Indonesia in issuing regulations related to capital, foreign borrowing, derivative products, and systemically important banks.
Managing Capital Flows: Integrating Monetary and Macroprudential Policies
A two-speed global economic recovery since 2009 has created massive capital flows, which have posed a number of challenges for emerging countries like Indonesia. Although recipient countries have benefited from the inflows through financial deepening and wider sources of financing, capital flows have also elicited various challenges in the recipient countries.
Capital flows have put pressures on domestic currency appreciation, accelerated economic overheating, triggered asset-price bubbles, and intensified the risk of financial system instability. Speculative capital inflows could create economic vulnerabilities to changes in investor sentiment, primarily through changes in asset prices and the exchange rate, and maturity mismatches. This is a version of the classic impossible trinity of monetary policy in a small open economy.
To confront this issue, Bank Indonesia’s policy has been to strike an optimal balance among the three objectives by adopting a policy mix that integrates monetary and macroprudential policies (see Figure 2.7).
Figure 2.7.Managing the Trilemma in Indonesia: Incorporating Macroprudential Measures into the Monetary Framework
First, Bank Indonesia has allowed some room for nominal exchange rate appreciation in response to the wave of inflows. Although the appreciation has helped reduce inflation stemming from import prices, foreign exchange intervention is also needed to avoid a large misalignment and excessive volatility in the exchange rate. Considering that the inflows are highly vulnerable to the risk of reversal, the intervention also allows the central bank to build up foreign reserves during good times and use them during bad times (the possible reversal).
Second, to contain excessive short-term capital flows, the central bank has issued some macroprudential regulations to mitigate excessive volatile flows by “throwing sand in the wheels.” These regulations include requiring investors to hold Bank Indonesia Certificates (SBI, for Sertifikat Bank Indonesia) for six months, reinstating limits on short-term offshore bank borrowing, and increasing the foreign exchange reserve requirement ratio. These policies have encouraged the diversification of foreign portfolio capital flows, and reduced risks of excessive foreign borrowing by banks.
Third, the loss of monetary policy freedom as a result of rapid inflows can partly be resolved by adopting a macroprudential policy to absorb excess liquidity in the economy to mitigate the risks of an asset-price bubble. The actions in this policy include an increase in the statutory reserve requirement and the maintenance of financial stability through a macroprudential policy encouraging banks to manage credit and liquidity risks through a loan-to-deposit ratio–linked reserve requirement. This type of macroprudential policy is effective if banks play a major role in intermediating capital flows. However, if the capital flows emanate directly from unregulated sectors, such as direct loans from the private sector, measures to control capital inflows, for example, by limiting private loans, are another option.
In the current post–global crisis era, the monetary policy framework in Indonesia needs to be strengthened. The core element should still be a focus on price stability. However, the new framework should incorporate macroprudential policy for two reasons. First, highly procyclical movements in financial sectors warrant an integration of monetary and macroprudential tools for countercyclical policy. Second, in a small open economy, like Indonesia, monetary authorities often face dilemmas, especially during periods of large and volatile capital flows, thus support from macroprudential policy can be beneficial in maintaining both price stability and financial stability, as evidenced in recent years.
Efforts should be made to ensure the new framework works well. These efforts should include attaining a better understanding of the transmission mechanism of macroprudential policy. At the institution level, coordination among authorities, in particular, monetary and macroprudential authorities and the microprudential regulator, should be strengthened, although in practice this may be easier said than done.
See Table 2.1 for a summary of Indonesia’s macroprudential measures.
|Minimum holding period on Bank of Israel (BI) bills (one month in 2010, changed to six months in 2011)|
|Shift BI bills to term deposits as of June 2010|
|Reinstate limits (maximum of 30 percent of capital) on the short-term offshore borrowing of banks, effective end-January 2011|
|Increase foreign exchange reserve requirements of the banks from 1 percent of foreign exchange deposits to 5 percent effective March 2011 and 8 percent effective June 2011|
|Increase rupiah reserve requirement from 5 percent to 8 percent, effective November 2010|
|Lengthen (from weekly to monthly) auctions and offer longer-maturity (3, 6, and 9 months) BI bills as of June 2010|
Putting the “Macro” into Macroprudential: The Israeli Experience8
Macroprudential is messy. Every discussion of “macroprudential” quickly runs into difficulties of definition, delineation, and categorization. It does not fit neatly into categories—theoretical, organizational, or operational. It is hard to quantify, lacks benchmarks, and cuts across numerous areas of responsibility. Models are hard to come by and unsatisfactory. It is extremely country specific. Success is frequently unobservable whereas failure is immediately and painfully obvious. It has few tools that are exclusively its own, but borrows them from other areas—and even its name is a hybrid borrowed from more clearly defined aspects of economic policy.
This ambiguity is perhaps most succinctly summarized in the question posed at the very beginning of one recent paper on macroprudential policy, which can be paraphrased as “is the goal of macro-prudential to protect the banks from the economy or to protect the economy from the banks?” (Bank of England, 2009).
However, it is this very complexity, along with its importance, that makes the successful design and implementation of macroprudential policy so challenging, so interesting, and in many ways the toughest test a central bank will face—a comprehensive “final exam” for central bankers. To succeed, the central bank must meet multiple goals using multiple tools in a dynamic and risky environment. Macroprudential means tying together areas and operations both within and outside of the central bank, that are not used to working together, to meet common goals in a holistic fashion.
This is because the “macro” in macroprudential stands for macroeconomic as well as macroprudential. The initial innovation in macroprudential policy was moving from microprudential—ensuring the stability of a single institution—to ensuring the stability of the financial system as a whole. That is clearly a major and vital advance. But macroprudential policy should also aim at ensuring stability of the economy as a whole, including price stability. Specifically, monetary policy and macroprudential policy are complementary, in a triangular relationship:
Macroprudential should ensure systemic financial stability.
Macroprudential should support price stability by assisting monetary policy.
Monetary policy should support systemic stability through integration with macroprudential.
The first point is obvious, and the second point is gaining acceptance. However, the third point is less clear, and possibly even somewhat controversial, so this section focuses on that, and uses experience in Israel to illustrate.
Example 1. Macroprudential Measures in the Mortgage Market—An Evolving Policy
The share of owner-occupied housing in Israel is high (69 percent, with mortgage loans comprising 70 percent of household debt and 25 percent of all bank credit; all figures for year-end 2011). All mortgages are full recourse, and are usually extended by the five largest banks in a very concentrated banking market. Historically, default rates have been extremely low.
In the wake of record low interest rates reached in response to the 2008–09 financial crisis, along with low returns on alternative investments and their higher perceived risks, the volume of new mortgages increased significantly, from an average yearly growth rate of 5 percent earlier in the decade to 13.2 percent in 2010 (Figure 2.8, panel a). Housing price increases accelerated, and concerns of a possible bubble in the housing market surfaced. Significantly, the composition of mortgages changed dramatically, with the share of long-term fixed-rate9 mortgages falling from more than 50 percent of new mortgages in 2004 to less than 10 percent in 2009, while the share of non-indexed floating-interest-rate mortgages rose rapidly from less than 10 percent of mortgage origination in 2004 to a peak of more than 75 percent in mid-2009 (Figure 2.8, panel b). In August 2009, the Bank of Israel began raising interest rates from a record low of 0.50 percent and the share of floating rate mortgages began to decline, but remained about half of all new mortgages from the end of 2009 until mid-2011 (Figure 2.8, panel c).
Figure 2.8.Israel: Confronting Financial Risks
Source: Bank of Israel.
Note: NIS = new Israeli sheqel.
The increased share of floating-rate mortgages raised a number of serious and interrelated concerns in the central bank. The first issue was a traditional microprudential issue regarding the stability of the banks that were extending floating-rate mortgages. Although, as noted, mortgage lending was traditionally considered very low risk, these were a new form of mortgage that included for the first time substantial interest rate exposure for the borrowers. Both lenders and borrowers lacked experience and historical data, so it was not clear that the reliance on past low default rates was justified. Moreover, a floating-rate mortgage was most attractive, and most easily marketable, when rates were low but there was a high probability of interest rate increases over the lifetime of the mortgage, bringing much higher monthly repayments and payment-to-income ratios. Consumer protection and conduct of business were also concerns.10
For a number of reasons, concern soon increased that the extremely rapid growth in floating-rate mortgages could pose systemic risks as well. Obviously, the extremely rapid growth of a new instrument with new and possibly underappreciated risks in almost all banks at the same time, constituting a growing portion of all bank credit, meant that it was inherently a system-wide issue. Moreover, the new and primary risk factor—interest rate exposure—was common across the sector, and difficult for the system to hedge or diversify. A period of rising interest rate risks might bring higher default rates, which would be highly correlated across institutions given that they shared a common principle risk factor. Rapid growth in this new type of mortgage and its share in total bank credit were peaking just as interest rates were at a historical and cyclical low.
Moreover, the risk in mortgage loans could be highly correlated with another major part of banks’ credit portfolios—construction loans. Residential construction in Israel is funded to a significant degree by builders selling units before they are actually built. Builders are essentially funded through mortgages taken by buyers, in addition to the loans they take directly. Thus, a rapid increase in interest rates would lead to more expensive mortgages, a decrease in demand for housing, and an inability of builders to fund themselves through the pre-sale of units, all of which would increase their risk. Because banks also had large exposures to construction firms through direct lending, two major parts of the credit portfolio—residential mortgages and construction loans—would be highly correlated.11
Another concern was that the banks would face increasing default risk on floating-rate mortgages due to a rising interest rate environment just as funding difficulties were increasing generally. Given the inherent mismatch between the long-term asset in mortgages and their short-term interest rate exposure, this was a special concern in the case of floating-rate mortgages.
Clearly, the rapid growth in this sector raised a number of prudential concerns. Significantly, it also posed serious challenges for monetary policy. One concern was that with a large portion of new mortgages having floating interest rates, the impact of changes in the monetary policy rate would be considerably amplified. Even though this could have advantages, in certain situations it could present the central bank with difficult dilemmas. For example, consumer price inflation and overall economic activity might suggest room for lower interest rates—yet concern over high housing costs and the possibility of feeding an asset boom could restrain the Bank of Israel’s policy choices. Conversely, the bank’s options might be limited if otherwise justified rate increases might pose the danger of a sudden stop in mortgage origination and rapid increase in defaults, potentially damaging financial stability as well as a vital sector of the economy.
Another side effect of the increased sensitivity of the housing market to short-term interest rates was the impact on inflation as measured by the consumer price index (CPI). Although the CPI is based on the imputed cost of housing services, the variable relationship between rental housing costs (which affect the CPI) and the price of housing could introduce significant volatility in the inflation rate, complicating monetary policy.12
Beginning in 2010, the Bank of Israel took a number of regulatory steps to reduce risks in the mortgage market, including
requiring a supplemental loan loss provision of 0.75 percent on mortgages with a loan-to-value ratio exceeding 60 percent;
applying a 100 percent risk weighting on mortgages with loan-to-value ratios of 60 percent or more and a floating rate portion on 25 percent of the total loan;13
intensifying reporting requirements; and
enhancing on-site and off-site reporting.
These steps served a clear purpose in reducing risks at both the institutional level and system wide. However, their impact on the actual volume of floating-interest-rate loans was limited once they were priced in by the market. Moreover, the strong link between monetary policy steps and developments in the housing market remained, albeit at a lower level. Although the traditional purpose of macroprudential policy was partially achieved—reducing risks to the financial system—the additional goal (supporting monetary policy) was not. The need for additional action was apparent.
In May 2011, the Bank of Israel imposed a ceiling limiting the floating rate component of mortgages to one-third of the total loan. The impact was clear: the share of floating-rate mortgages dropped significantly while interest rates remained stable, and remained at lower levels even after the Bank of Israel resumed lowering interest rates in the second half of 2011 (Figure 2.8 panel c).
Although it is still early, the impact of the May 2011 ceiling appears to be a weakening of the automatic and mechanical link between the policy interest rate and mortgage rates, which could in certain situations limit the Bank of Israel’s ability to implement monetary policy, or lead to potentially unacceptable side effects of that policy. So far, the appropriate macroprudential steps have not only met their direct goal of strengthening systemic stability, but have also supported and strengthened monetary policy.
Example 2. Supporting Exchange Rate Policy
Beginning in 2007, the exchange rate of the Israeli sheqel appreciated rapidly. In March 2008, the Bank of Israel began to intervene in the foreign exchange market, deviating from a 10-year policy of nonintervention. By the beginning of 2011, the bank had purchased US$44 billion, increasing reserves from US$29 billion in 2008 to US$70 billion at the end of 2010. Because further intervention was becoming ever more difficult and costly, the bank sought supporting measures. The first measure implemented was the imposition of a reserve requirement on foreign exchange swap and forward transactions.
Data routinely reported to the Bank of Israel showed that a considerable portion of purchases of Israel sheqalim against foreign currency were by means of forward or swap contracts, rather than outright spot purchases. Using a derivative contract had numerous advantages for those wishing to obtain a financial exposure to the exchange rate without necessarily having any “real” need for or exposure to the Israeli currency, economy, or financial system. Moreover, the leverage inherent in these transactions allowed market participants and financial intermediaries (primarily Israeli commercial banks) to take larger positions than they might have otherwise. The rapid increase in the volume and exposure to these instruments in itself raised prudential concerns.
In January 2011, the Bank of Israel imposed a 10 percent reserve requirement on all foreign exchange derivative transactions of nonresidents with an Israeli bank. Because the reserve requirement is unremunerated, this was estimated to be equivalent to an increase in the cost of transactions of close to10 basis points.
The impact was significant. Outstanding foreign exchange derivative transactions by nonresidents contracted rapidly (Figure 2.8, panel d). In addition to the effect of the direct increase in costs, the contraction was attributed to a strong signaling effect—the market saw that the central bank had, and would use, other tools at its disposal to achieve its policy objectives. Because of the other factors involved, it is difficult if not impossible to isolate the impact of a single step, but the direction was clear.
It was apparent from the beginning that the reserve requirement would be limited to a supporting role. First, it was evident that the impact of the step would diminish over time, because the market found ways to circumvent the reserve requirement, either by transacting offshore or by transacting with nonbanks financial institutions not subject to reserve requirements. Moreover, the possibility of contributing to disintermediation of the banking system, and even to the growth of shadow banking, was seen as a significant potential drawback, and was one of the factors in limiting the size and extent of the reserve requirement.
In summary, it appears that the reserve requirement played a small but positive role in reducing undesirable pressure on the exchange rate as well as lessening potential risks to financial market stability, thus, it was a useful addition to the central bank’s macroprudential toolkit.
Example 3. Supporting the Exchange Rate—Further Steps
As part of its efforts to alleviate unwarranted appreciation pressure on the Israeli sheqel, the Bank of Israel sought ways of introducing some friction into the financial markets, to “throw some sand into the wheels.” Here, too, the goals were twofold: to dampen pressure on the exchange rate while reinforcing financial market stability. Unequal conditions and underpriced risks could contribute to the buildup of destabilizing flows and positions. Developments in the central bank bill (Makam) market were a case in point.
For many reasons, Makam became the instrument of choice for nonresidents who were able to take a cash position in the domestic currency. Along with natural advantages (good liquidity, free of credit risk) there was an aberration—nonresidents were exempt from taxes on Makam and government bonds. (The exemption remained from when government policy was aimed at attracting foreign investment into Israel and expanding undeveloped financial markets.) By 2011, with Israel facing large capital inflows complicating monetary policy, it was difficult to justify maintaining that policy.14 Because tax policy was the responsibility of the ministry of finance, the Bank of Israel urged the government to remove the tax benefit granted to nonresidents, creating a more level playing field. The ministry agreed and announced its intention to introduce the necessary legislation.15
Because the tax changes would take time to be implemented, in January 2011 the Bank of Israel announced its intention to impose a reporting requirement for all Makam transactions involving nonresidents. After a relatively short consultative process with market participants, the bank implemented the reporting requirement. Although the reporting requirements were not particularly onerous, and there were no restrictions, prohibitions, or explicit costs (other than those involved in compiling the reports), they had a distinct impact on the market for Makam and indirectly on the exchange rate (Figure 2.8 panel e).16 One factor was the regulatory and compliance burden involved. Beyond that, there was a significant signaling effect—the market perceived the move as signaling the central bank’s determination and ability (and perhaps preparations) to take further, unconventional actions to meet its policy goals.
Example 4. Possible Financial Stability Implications of Low Interest Rates
The debate concerning the role of low interest rates in the recent crisis, or in contributing to financial instability, asset bubbles, or misallocation of resources, is far from resolved and highly contentious. There is good reason to consider seriously the possibility outlined in the following:
It would be a delusion to expect that a macroprudential framework on its own could ensure financial stability … monetary policy is key … Monetary policy sets the universal price of leverage in a given currency area, and as such is harder to circumvent. It operates precisely by influencing credit and asset prices and as such is more likely to act as an effective speed limit. And, as increasing evidence suggests, it can influence risk perceptions and attitudes—the price of risk—and complement macroprudential tools. (Borio, 2011b, p. 17)
Given that no definitive conclusion has been reached, it might be wise for policy to err on the side of caution, with a reasonable assumption being that if there are negative side effects of low interest rates, they will become more and more pronounced as rates approach the zero bound.
Financial stability issues were never a decisive factor in interest rate decisions in Israel, but they were raised as concerns. Israel has seen very rapid growth in the corporate bond market, from a very low level to the highest share of total credit in the developed world (Figure 2.8 panel f). Since the beginning of 2009 there has been a rapid increase in the level of distressed debt (Figure 2.9 panel a) and the number of reschedulings and debt workouts (Figure 2.9 panel b).
Figure 2.9.Israel: Credit and Corporate Bonds
Source: Bank of Israel.
1 There were no reschedulings before 2008.
2 Through February 2012.
Concerns are high as the result of some rather unusual features of corporate structures in Israel, specifically a high level of ownership concentration, and the prominent role of “pyramid companies”—extremely highly leveraged holding companies. The possibility has been raised that low interest rates allowed the buildup of very high levels of leverage, which resulted in increased risks to the economy and financial sector. Therefore, it could be argued that prolonged excessively low interest rates in Israel should be avoided not only because they might threaten price stability, but also because they might endanger financial stability—a case of monetary policy supporting financial stability through its macroprudential impact.
Numerous difficulties arise in assessing the impact of the macroprudential steps Israel has taken so far. The steps were only recently enacted, and in some cases the counterfactual cannot be ignored. The problem of isolating the effect of individual measures is especially troubling because measures were frequently bundled together. It is also impossible to separate the effect of other external events occurring at the same time. The clearest example is in trying to assess the impact of measures taken to dampen capital inflows: in the same period, regional turbulence in the Middle East, the European debt crisis, geopolitical concerns, and gyrating global risk appetites were influencing the exchange rate as well.
Nevertheless, the interim conclusion is clear: Israel’s experience with macroprudential policy has proven beneficial. Measures taken regarding mortgages appear to have reduced systemic risks to the banking system and to important sectors of the economy, and at the same time were supportive of monetary policy. Innovative reserve requirements on derivative transactions along with heightened reporting requirements and rationalized tax policies helped dampen volatile capital flows and their potentially deleterious impact on financial stability, market function, and macroeconomic variables. Awareness of the mutual interactions between macroprudential policy and monetary policy can only improve the design and implementation of each, with beneficial results for both.
Another observation is that the impact of macroprudential measures frequently stems from their signaling effect—they are strong indicators of the availability of a wide range of tools that the central bank can use, and equally important, of the central bank’s willingness to use them to reach its goals. Credibility is strengthened.
There was a commonly held view during the “great moderation” of the 1980s that there could be a neat separation between the monetary and the financial stability functions of the central bank. “Monetary policy would take care of price stability while regulation and supervision would take care of financial stability” (Borio, 2011a, pp. 2–3). This is no longer believed to be the case.
Macroprudential policy should be viewed in a broad context, tying together the two key goals of a central bank: price stability and financial stability. Macroprudential policy can and should support monetary policy; in certain circumstances monetary policy can and should support financial stability, and at the very least not endanger it.
In difficult environments central banks need to achieve multiple aims, and to do this they need multiple tools. Macroprudential measures can provide those tools. The use of unconventional measures along with traditional tools can enhance the effectiveness of overall policy and increase credibility. The international context is vital to learn from other experiences, to maintain cooperation, and to observe best practice. The appropriate mix of steps can also lessen unwanted side effects and increase the flexibility of central bank policies.
A central bank should not only have the ability to implement macroprudential policy; it should also be independent in decision making and implementation. The argument is similar to that of monetary independence: time inconsistency. Research has shown that the duration of financial cycles is longer than that of real economic cycles, ranging from 6 up to 18 quarters, whereas business cycles typically last four to five quarters. Only the central bank will usually have the independence and time horizon needed to deal adequately with the challenges posed by financial cycles.
AgénorPierre-Richard and Luiz A. Pereirada Silva2011 “Macroeconomic Stability, Financial Stability, and Monetary Policy Rules,” International Finance Vol. 15 No. 2 pp. 205–24.
AgungJuda2010 “Integrating Monetary and Macroprudential Policy: Towards a New Paradigm of Monetary Policy Indonesia Post Global Crisis,” (unpublished: Jakarta: Bank Indonesia).
AlamsyahH.C.JosephJ.Agung and D.Zulverdy2001 “Towards Implementation of Inflation Targeting in Indonesia,” Bulletin of Indonesian Economic Studies Vol. 37 No. 3 pp. 309–24.
Bank of England2009 “The Role of Macroprudential Policy,” Discussion Paper (London).
BernankeBen S. and MarkGertler2001 “Should Central Banks Respond to Movements in Asset Prices?” American Economic Review Vol. 91 No. 2 pp. 253–57.
BlinderAllan S.2010 “How Central Should the Central Bank Be?” CEPS Working Paper No. 198 (Princeton, New Jersey: Princeton University, Center for Economic Policy Studies).
BorioClaudio2011a “Central Banking Post-Crisis: What Compass for Uncharted Waters?” BIS Working Paper No. 353 (Basel: Bank for International Settlements).
BorioClaudio2011b “Implementing a Macroprudential Framework: Blending Boldness and Realism,” Capitalism and Society Vol. 6 No. 1 p. 17.
GambacortaLeonardo2009 “Monetary Policy and the Risk-Taking Channel,” BIS Quarterly ReviewDecember pp. 43–53.
Hong Kong Monetary Authority (HKMA)2011 “Assessing the Effectiveness of the Limit on Debt Servicing Ratios under Interest Rate and Income Shocks,” in “Half-Yearly Monetary and Financial Stability Report (March)” (Hong Kong SAR) pp. 72–76.
International Monetary Fund2011 “Macroprudential Policy: An Organizing Framework,” Monetary and Capital Markets Department (Washington).
JuhroSolikin M. and others2009 “Review of the Application of ITF in Indonesia,” Directorate of Economic Research and Monetary Policy (Jakarta: Bank Indonesia).
MishkinFrederic S.2011 “Monetary Policy Strategy,” NBER Working Paper No. 16755 (Cambridge, Massachusetts: National Bureau of Economic Research).
RossiniRenzo and ZenonQuispe2010 “Monetary Policy During the Global Financial Crisis of 2007–09: The Case of Peru,” BIS Working Paper No. 54 (Basel: Bank for International Settlements).
SatriaDoni and Solikin M.Juhro2011 “Risk Behaviour and Monetary Policy Transmission Mechanism in Indonesia,” Bulletin of Monetary Economics and BankingJanuary pp. 243–70.
StoneMark R.2003 “Inflation Targeting Lite,” IMF Working Paper 03/12 (Washington: International Monetary Fund).
SvenssonLars E.O.1999 “Price-Level Targeting Versus Inflation Targeting: A Free Lunch?” Journal of Money Credit and Banking Vol. 31 No. 3 pp. 277–95.
TaylorJohn B.2009Getting Off Track: How Government Actions and Interventions Caused Prolonged and Worsened the Financial Crisis (Stanford, California: Hoover Institution Press).
WongE.T.FongK.Li and H.Choi2011 “Loan-to-Value Ratio as a Macroprudential Tool: Hong Kong’s Experience and Cross-Country Evidence,” Hong Kong Monetary Authority Working Paper No. 01/2011 (Hong Kong SAR). http://www.hkma.gov.hk/media/eng/publication-and-research/research/working-papers/HKMAWP11_01_full.pdf.
WongJ.L.FungT.Fong and A.Sze2004 “Residential Mortgage Default Risk and Loan-to-Value Ratio,” Hong Kong Monetary Authority Quarterly Bulletin No. 41December pp. 35–45. http://www.hkma.gov.hk/media/eng/publication-and-research/quarterly-bulletin/qb200412/fa3.pdf.
For details, see relevant press releases and circulars, which are available on the HKMA website.
Debt-servicing ratios have been used for supervisory purposes in Hong Kong SAR since the 1990s.
For details, see HKMA (2011) and the HKMA circular “Prudential measures for residential mortgage loans,” issued on August 13, 2010, available on the HKMA website.
If property values fall because of a downturn in the property market, households would suffer a negative wealth effect. Slowdowns in the housing market are usually accompanied by a general economic slowdown, wages fall or growth decelerates and unemployment rises. These effects, combined with the negative wealth effect, would reduce households’ debt-servicing ability, thus increasing the credit risk to financial institutions. Faced with declining asset quality, banks would reduce credit to households, which in turn would reduce aggregate demand for housing. This feedback loop would exacerbate the property market downturn, the negative effect on households, and ultimately the impact on financial institutions.
The diminished value of property collateral in a housing market downturn would mean a decline in recovery values from borrowers who default. In addition, a large decline in the value of collateral would discourage financial institutions from extending credit to new borrowers. Reduced credit to housing market participants, in turn, would drag down the housing market, which together with reduced demand from households, could potentially contribute to a downward spiral in property prices.
Currency stability refers to both price stability and exchange rate stability. But, in practice, Bank Indonesia puts more weight on price stability (controlling inflation).
See Alamsyah and others (2001) for the early implementation of inflation targeting in Indonesia.
The author would like to thank staff of the Market Operations Department, Research Department, and Supervisor of the Banks for assistance in preparing data and for their excellent work in macroprudential policy.
Until 2009, almost all mortgages in Israel (and many contracts) were indexed to the consumer price index (CPI). Therefore, a fixed-rate mortgage refers to a fixed (real) interest rate but linked to the CPI, whereas a floating-interest-rate mortgage refers to an unlinked mortgage for which the nominal interest rate is derived from the Bank of Israel policy interest rate.
The Supervisor of Banks in the Bank of Israel took a number of steps to mitigate the risk of borrowers assuming unsustainable mortgage burdens: requiring banks to detail to the borrower the increase in payments under various scenarios, increased reporting requirements and inspections, and the like. These had little overall impact; it appears that the attraction of lower monthly payments upfront more than offset the risk of much higher payments in the future.
Together they reached close to 40 percent of total bank credit.
Israel has an inflation targeting regime; the Bank of Israel is mandated to keep inflation within a 1–3 percent range for a horizon not exceeding two years.
This applied only to mortgages above 800,000 Israeli new sheqalim.
Another concern was the possible impact on the Makam market (and its role as a monetary tool) of large holdings by nonresident, less-stable holders of the bills.
Israel has double taxation treaties with most relevant countries, so most foreigners would not, in fact, have incurred any final tax liabilities in Israel. However, the reporting and filing complications involved in obtaining exemptions proved to be as much of an obstacle as the actual tax.
The announcement explicitly stated that the reporting was to be used only for data gathering and analytical purposes, and to assist in monitoring market developments.