- International Monetary Fund. Monetary and Capital Markets Department
- Published Date:
- September 2011
The difference between the weekly returns of a financial institution’s equity price and the S&P 500 index.
A nontraditional component of investment portfolios that may include commodities, hedge funds, infrastructure, private equity, and real estate.
A form of commercial paper that is collateralized by other financial assets. Typically, ABCP is a short-term instrument that is issued by a bank or other type of financial institution and matures between 1 and 180 days from issuance.
A sharp rise in the price of an asset above its economically fundamental value over a specific period for reasons other than random shocks. In empirical work, the difficulty of classifying an asset price increase as a bubble is related to the complexity of determining what asset price levels are consistent with observable fundamentals.
Funds managed by a financial institution.
A separate institutional unit established by a private or public employer with the express purpose of providing retirement incomes for its employees.
A statistical time-series model in which the current value of a variable depends on its past values.
A set of international banking regulations focused primarily on credit risk, introduced by the Basel Committee on Banking Supervision in 1988. Under Basel I, bank assets were classified into five categories in which the credit risk of the assets were weighted at zero, 10, 20, 50, and 100 percent, respectively. Internationally active banks were required to hold regulatory capital equal to at least 8 percent of the risk-weighted assets. See also Basel III and Capital adequacy ratio.
A comprehensive set of reform measures introduced in the aftermath of the global financial crisis to improve the banking sector’s ability to absorb financial and economic shocks; enhance banks’ risk management and governance; and increase banks’ transparency and disclosure. These measures revise the existing Basel capital adequacy standards and propose, for the first time, minimum liquidity adequacy standards for banks. See also Basel I and Capital adequacy ratio.
The difference in the yields of two financial instruments with similar risk characteristics and similar cash flows. The “basis” also refers to the difference between a futures contract price and the underlying cash instrument or commodity.
A mean-variance portfolio optimization approach, developed by Fisher Black and Robert Litterman (1992; cited in Chapter 2). It enhanced the original approach of Markowitz (1952; cited in Chapter 2) by (i) use of implied market equilibrium returns as the basis for the expected returns used in the optimization; (ii) potential introduction of assumptions about relative returns across assets into expected returns; and (iii) incorporation of uncertainty about the user’s return assumptions. See also Mean-variance approach for portfolio selection.
The ratio of regulatory capital to risk-weighted assets of a financial institution. Regulatory capital is the sum of common equity and additional Tier 1 capital and Tier 2 capital. See also Basel I and Basel III.
A model that combines the Markowitz (1952; cited in Chapter 2) mean-variance approach with market-clearing in asset markets and the existence of a risk-free asset to model the pricing of an individual asset or a portfolio of assets. See also Black-Litterman approach.
A strategy whereby an investor sells an investment in a low-yielding currency (or borrows at low interest rates) and uses the proceeds to purchase an investment in a different currency with a higher yield.
A deposit offered by banks, thrift institutions, and credit unions in the United States, generally at a fixed interest rate for a specified term to maturity.
Transmitting, reconciling, and, in some cases, confirming payment orders or instructions for the transfer of financial instruments before settlement. Clearing includes netting orders and the establishment of final positions before settlement. For futures and options, clearing also entails daily balancing of profits and losses and daily calculation of collateral requirements.
A managed investment company with an unlimited life that rarely issues new shares after its initial offering and does not redeem its own shares from its shareholders.
An option strategy established by holding shares of an underlying stock, purchasing a protective put and writing a covered call on that stock. The put and call contracts are both out-of-the-money, have the same expiration date, and are identical in number. See also Credit Suisse Fear Barometer.
Assets pledged or posted to a counterparty to secure an outstanding exposure, derivative contract, or loan.
An unsecured promissory note with a fixed maturity of 1 to 270 days.
The Value at Risk (VaR) in the financial system conditional on financial institutions being under stress. CoVaR is thus a measure of an institution’s contribution to systemic risk. The contribution to systemic risk is the difference between CoVaR for tail-risk episodes and the CoVaR at the median state. See also tail event.
A class of utility functions that conveniently summarize an investor’s risk preference. The functions assume that the investor’s risk preference is independent of his or her wealth.
A methodology that combines balance sheet data and market prices of traded securities to infer the implicit value of assets and contingent liabilities of a corporation. The method has been extended to study aggregate economic sectors and countries.
A supplementary capital requirement for individual banks aimed at countering the procyclical depletion of capital. The buffer would be triggered by a signal of the buildup of systemwide risk, and would be drawn down when the risk materializes. The CCB would thus act as an automatic countercyclical stabilizer for a bank’s risk. See also Macroprudential capital requirements.
A credit derivative whose payout is triggered by a “credit event,” often a default. CDS settlements can either be “physical”—whereby the protection seller buys a defaulted reference asset from the protection buyer at its face value—or in “cash”—whereby the protection seller pays the protection buyer an amount equal to the difference between the reference asset face value and the price of the defaulted asset. A single-name CDS contract references a single firm or government agency, whereas CDS index contracts reference standardized indices based on baskets of single-name CDS contracts. See also Credit derivative and Derivative.
A financial contract under which an agent buys or sells risk protection against the credit risk associated with a specific reference entity (or specified range of entities). For a periodic fee, the protection seller agrees to make a contingent payment to the buyer on the occurrence of a credit event (usually default in the case of a credit default swap). See also Credit default swap and Derivative.
A measure of the ability of a borrower to meet its financial commitments on a timely basis. Credit ratings are typically expressed as discrete letter grades. For example, Fitch Ratings and Standard & Poor’s use a scale in which AAA represents the highest creditworthiness and D the lowest.
An indicator of investor sentiment maintained by Credit Suisse. The value of the indicator (or the degree of fear) is based on the skew between the upside target and the downside protection embedded in zero-premium three-month collars. The collar is implemented by selling a three-month, 10 percent out-of-the-money S&P 500 call option and using the proceeds to buy a three-month, out-of-the-money S&P 500 put option of equal value.
The deviation of the credit-to-GDP ratio from a long-term trend. Under the modeling assumptions of this GFSR, the gap is estimated as a deviation from a recursive trend based on a Hodrick-Prescott (HP) filter.
The percentage of gross income applied to debt repayment. The debt portion of the ratio may also include taxes, fees, and insurance premiums associated with the debt.
A plan in which an employer pays its employees a predefined lifetime benefit at the time of retirement. The amount of the benefit is usually based on factors including age, earnings, and years of service. See also Defined-contribution pension plan.
A plan in which each participating employee typically has an account to which the employer makes a prespecified contribution. The employee’s benefit is based on the amounts credited to the account (through employer contributions and, if applicable, employee contributions) and any investment earnings on the account. The level of future retirement benefits is not guaranteed. See also Defined-benefit pension plan.
A failure to make a contractual payment on a loan, usually defined as a certain number of days (or more) overdue (e.g., 30, 60, or 90 days).
A financial instrument with a value dependent on the expected future price of its underlying asset, such as a stock or currency. Examples of derivatives include stock options, currency and interest rate swaps, and credit default swaps. See also Credit derivative.
A measure of spillovers across financial institutions developed by Diebold and Yilmaz (2009; cited in Chapter 3). It is based on the matrix of variance decompositions derived from rolling vector autoregressions (VARs) of financial institutions’ weekly credit default swap returns.
A measure of credit risk, associated with the probability that the market value of a firm’s assets will fall below the value of its debt. The equity of the firm is modeled as a call option on the value of its assets; the exercise price is equal to the value of the liabilities, since the firm defaults when its asset value falls below the face value of its debt. In the context of this GFSR, the DD represents the number of banking system standard deviations from default.
A model that seeks to describe the behavior of the aggregate economy by analyzing the interaction of various microeconomic and financial decisions, including consumption, saving, and investment. Agents (or decision makers) in the model are varied and may include households, firms, banks, and governments. DSGE models are dynamic (accounting for various leads and lags in the evolution of the economy) and stochastic (accounting for the impact of random shocks—including technological or policy changes—on the economy).
In an econometric modeling context, endogeneity is linked to the existence of correlation between a variable and the error term of an equation. In broad terms, a variable is endogenous if it is a function of other parameters or variables in the model.
An institution set up by the current 16 euro area countries to preserve financial stability. The EFSF has the ability to issue bonds or other debt instruments in the market to raise the funds needed to provide temporary financial assistance to euro area member states in economic difficulty.
A daily reference rate based on the averaged interest rates at which euro area banks offer to lend unsecured funds to other banks in the euro wholesale money market (or interbank market).
A statistical technique used to assess the short-term impact of an event—such as a crisis—on various macroeconomic or financial variables.
An investment fund traded on stock exchanges, most commonly tracking an index, such as the S&P 500. An ETF may be attractive to investors because of its low cost and tax efficiency.
A probability distribution that exhibits higher kurtosis than the normal distribution, that is, more of the possible outcomes result from infrequent extreme deviations from the mean. See also tail event.
An independent U.S. government agency that provides insurance on deposits in member banks and thrift institutions, currently up to $250,000 per depositor per account ownership category per institution.
An estimate of a country’s financial flows, usually conducted by the country’s monetary authority or central bank.
The act by the lender of taking control of assets, such as real estate, that were pledged or mortgaged to secure a loan, usually with the intention of selling those assets to recover part or all of the amount due from the borrower.
A stock of foreign currency denominated assets held and controlled by monetary authorities to meet balance of payments financing needs, intervene in currency markets to affect the country’s exchange rate, and achieve other related purposes (such as to maintain confidence in the currency and the economy and to support foreign borrowing). Such reserves must be liquid and must constitute claims on nonresidents that are denominated and settled in a convertible foreign currency.
An international group of finance ministries, central banks, and international financial bodies that monitors and makes recommendations about the global financial system.
In this report, the process by which banks issue or assume liabilities associated with assets on their balance sheets.
A panel model has cross-sectional and time dimensions. In a fixed-effects panel model, intercepts are allowed to vary along the cross-sectional dimension (for example, by country), or the time dimension, or both. A country fixed effects model summarizes structural characteristics that are fixed for a certain country over time. A time fixed effects model summarizes, for example, global shocks that affect all countries at a particular time.
A statistical model in which the variance of the error term is assumed to follow an autoregressive moving average (ARMA) process.
A statistical hypothesis test used to determine whether one time series may be useful in forecasting another, after taking into account various lags of the latter.
The euro area, Japan, the United Kingdom, and the United States.
The Group of Twenty Finance Ministers and Central Bank Governors established in 1999 as a forum for officials from systemically important advanced and emerging economies to discuss key issues related to the global economy. It consists of leaders from the European Union and the following 19 countries: Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Mexico, Russia, Saudi Arabia, South Africa, Republic of Korea, Turkey, the United Kingdom, and the United States.
A discount that a lender applies to the current market value of collateral received as security for a loan. The haircut reflects the risk that, at a later date, if the borrower defaults, the collateral may be worth less or be less easy to sell.
An investment pool, typically organized as a private partnership, that faces few restrictions on its portfolio and transactions. Hence, compared with more regulated financial institutions, hedge funds use a wider variety of investment techniques—including short positions, derivatives transactions, and leverage—in their effort to boost returns and manage risk.
Offsetting existing risk exposures by taking opposite positions in contracts with identical or similar risk—for example, in related derivatives contracts.
A statistical data-smoothing technique for determining the long-term trend in the data by removing short-term fluctuations.
A mortgage in which the monthly payments cover only accrued interest; the principal balance is paid off when the loan matures.
A measure of the joint probability of distress of all institutions in a financial system. The JPoD is constructed as a nonlinear, time-varying measure of “tail dependence” using a multivariate probability distribution of individual institutions’ implied asset value movements.
Indicators that signal well in advance the buildup of financial sector risks.
The proportion of debt to equity (also assets to equity or, in banking, assets to capital). Leverage can be built up by borrowing (on-balance-sheet leverage, commonly measured by debt-to-equity ratios) or by using off-balance-sheet transactions.
The acquisition of a controlling interest in a company’s equity with financing predominantly secured through leverage (borrowing).
An index of the interest rates at which banks offer to lend unsecured funds to other banks in the London wholesale money market. See also LIBOR-OIS spread and Overnight indexed swap.
The difference between the three-month LIBOR rate and the overnight indexed swap rate. The spread is a measure of default risk associated with lending among banks in the LIBOR market. See also London Interbank Offered Rate and Overnight indexed swap.
A statistical model of linear relationships among variables, often under the assumption of normally distributed errors.
The outstanding balance on a loan as a proportion of the value of the asset (for instance, house) pledged as collateral against the loan.
A statistical binary response model in which the response probability follows a logistic distribution and is evaluated as a function of the explanatory variables.
Capital requirements for banks aimed at curbing the probability of a systemic financial crisis. Unlike microprudential capital requirements, which are based on idiosyncratic risk, macroprudential requirements are set to mitigate the buildup of systemic risk (countercyclical capital buffers in the context of Chapter 3) and the interconnectedness of the financial system. See also Macroprudential policy tools.
Tools used by country authorities to limit systemic financial risk and potential disruptions in key financial services that can have a sizable negative effect on the real economy. Such tools seek to (i) dampen the buildup of financial imbalances, (ii) build sufficient buffers to contain the speed and severity of potential downswings, and (iii) identify and address risk concentrations and financial interlinkages that may result in spillover risks and financial system disruptions. See also Macroprudential capital requirements.
The tax rate on the last dollar of income earned. This is different from the average tax rate, which is the total tax paid as a percentage of the amount on which the tax is based. For instance, in a progressive tax system, any additional activity that pushes income above a certain threshold may be taxed at a significantly higher income tax rate than the amount of income below the threshold.
The requirement to record the price or value of a security, portfolio, or account to reflect its current market value rather than its book or acquisition value.
Recording the price or value of a security, portfolio, or account to reflect its current market value rather than its book value.
A measure of goodness of fit in nonlinear estimation models, such as logit and probit, in which traditional R-squared measures can lie outside the [0,1] interval and are generally unsuitable. The McFadden R-squared compares the log likelihoods of a model including only a constant term with a full model with explanatory variables.
The first formal approach to solve an investor’s portfolio allocation problem, developed by Markowitz (1952; cited in Chapter 2). Its name derives from the fact that the only inputs needed are the expected returns of the assets and their covariances.
The process of adjusting the values of model parameters to obtain a model representation that fits the observed data as closely as possible.
An open-ended mutual fund that invests in short-term debt securities such as U.S. Treasury bills and commercial paper.
A condition in which an entity tends to act less carefully than it otherwise would because the consequences of a bad outcome will be largely shifted to another party. For example, financial institutions have an incentive to take excessive risks if they believe that governments will step in and support them in the event of a crisis.
A financial structure through which investors pool their funds to invest in a portfolio of securities, which may be broadly diversified or focused on particular market segments. The equity stakes of the investors consists of shares in the large pool of underlying assets. The selection of assets is made by professional fund managers.
A measure that compares the level of good signals to the level of background noise. In the context of crisis modeling described in this GFSR, the NSR is the ratio of Type II errors (the proportion of times the indicator signaled a crisis and the crisis failed to occur) to legitimate signals (the share of indicator signals of a crisis for which the crisis did occur).
Loans for which the contractual payments are delinquent, usually defined as a certain number of days (or more) overdue (e.g., 30, 60 or 90 days). The NPL ratio is the amount of nonperfforming loans as a percent of gross loans.
Statistical testing in which a model created on the basis of historical data for a sample period of time (“in sample” period) is applied retroactively to a subsequent period of time (“out of sample” period) to determine the quantitative validity of the model and to derive simulated performance results.
e A financial contract whose value derives from underlying securities prices, interest rates, foreign exchange rates, commodity prices, or market or other indices and that is negotiated and traded bilaterally rather than through an organized exchange.
An interest rate swap in which the compounded overnight interest rate in the specified currency is exchanged for some fixed interest rate over a specified term. See also London Interbank Offered Rate and LIBOR-OIS spread.
The probability of observing a value for a test statistic equal to, or more extreme than, the one observed in reality, assuming that the statistic follows a distribution defined by a null hypothesis. The smaller the p-value, the more strongly the hypothesis is rejected. A p-value of less than a chosen significance level rejects the null hypothesis.
A situation in which a currency has a persistent positive differential with respect to the currency to which it is pegged. The differential exists because the market perceives a small probability of a large devaluation.
A statistical binary response model in which the response probability follows a normal distribution and is evaluated as a function of the explanatory variables.
The tendency of changes in asset prices and valuations to move in line with macroeconomic business cycles.
A financial instrument that allows, but does not require, its owner to sell (“put”) an asset to the option seller at a certain price (the strike price) in the future. The option is of value to its owner if the market price of the asset drops below the strike price.
A statistical test used to detect unknown structural breakpoints in a linear relationship by a change in the parameter estimates from a regression equation.
A graphical method for determining the discriminatory power of signaling variables. It summarizes the costs and benefits of choosing various noise-to-signal ratio (NSR) thresholds of an indicator ranging from high to low.
The sale of securities together with an agreement for the seller to buy back the securities at an agreed price at a later date. The securities lender receives cash in return and pledges the legal title of a security as collateral.
Earnings that have not been distributed to stockholders or transferred to a surplus account. Retained earnings are a part of a bank’s capital.
The ratio of a bank’s net income to the average value of its total assets (or equity capital) for a specified period.
The degree to which an investor who, when faced with two investments with the same expected return but different risk, prefers the one with the lower risk. That is, it measures an investor’s aversion to uncertain outcomes or payoffs.
An asset allocation framework that focuses on asset profiles of risk and return rather than on narrowly defined asset classes such as equities and bonds.
Investors’ inclination to look for additional income or other return from their investments by moving into riskier asset classes such as those that are more illiquid, more complex, or contain more duration and credit risks.
The market for the sale of securities or bonds collateralized by the value of mortgage loans.
The creation of securities from a portfolio of existing assets or future receivables that are placed under the legal ownership or control of investors through a special intermediary created for this purpose—a “special purpose vehicle” (SPV) or “special purpose entity” (SPE).
A method of estimating the parameters of a system of equations that accounts for heteroskedasticity and contemporaneous correlation in the errors across equations. Also known as Zellner’s method.
The act that discharges the obligation to transfer funds or securities between two or more parties.
A measure of the asymmetry of the probability distribution of a random variable.
A Directive of the European Union (Solvency II Directive 2009/138/EC) that codifies and harmonizes insurance regulations in the European Economic Area. Among other things, the directive determines the amount of capital that insurance companies must hold to reduce the risk of insolvency.
An investment fund created by a sovereign government for macroeconomic purposes. SWFs hold, manage, or administer assets to achieve financial objectives and employ a set of investment strategies that include investing in foreign financial assets. SWFs are commonly established variously from balance of payments surpluses, official foreign currency operations, the proceeds of privatizations, fiscal surpluses, and receipts resulting from commodity exports.
Usually a subsidiary company with a balance sheet structure and legal status that makes its obligations secure even if the parent company goes bankrupt. See also Securitization.
A data series with a distribution that does not change over time. Consequently, parameters such as the mean and variance also do not change over time. If a data series is not stationary, it should be transformed to a stationary form before some standard econometric techniques can be reliably applied.
A process that evaluates the ability of an individual institution or the aggregate financial system to withstand adverse situations such as negative macroeconomic and financial shocks.
A structural change in a time series that denotes a discontinuity in the relationship among variables. In regression modeling, the occurrence of a structural break is associated with a change in the best fitting coefficients for the periods before and after the break.
A sudden slowdown in private capital inflows from abroad. When the inflows are large in relation to the domestic economy or financial system, as may be the case in emerging market economies, sudden stops are usually followed by a sharp decrease in output, private spending, and credit to the private sector and a real (inflation-adjusted) appreciation of the domestic currency.
An agreement between counterparties to exchange periodic payments based on different reference financial instruments or indices on a predetermined notional amount. See also Swap spread.
The differential between the government bond yield and the fixed rate on an interest-rate swap of the equivalent maturity. See also Swap.
An indicator based on the fraction of financial institutions experiencing both large negative abnormal returns on any given day and negative returns for the two weeks following that day.
A global indicator of liquidity stress that measures the breakdown of arbitrage conditions in major markets.
Financial institutions deemed to be of systemic importance for the financial stability and health of the economy. Prevention of SIFIs’ collapse is important to preserving systemwide financial stability. See also Too important to fail.
In regression estimation, the ratio of the estimated parameter coefficient to its standard error. The t-statistic can be tested against a t distribution to determine the probability that the true value of the coefficient is zero. See also T-test.
Any of various statistical tests based on the t distribution. Commonly used t-tests include tests for difference between two means; tests of whether a population mean is equal to a pre-specified value; and tests of whether the slope of a regression line differs significantly from a certain value (most commonly 0). See also T-statistic.
An event with a small probability but with large (adverse) effects.
Total assets less intangible assets (such as goodwill and deferred tax assets).
Total balance sheet equity less preferred debt less intangible assets.
A regulatory measure of capital supporting the lending activity of a bank. The capital that is most loss-absorbing, it consists primarily of common stock, retained earnings, and perpetual preferred stock. The greater the quantity of highly loss-absorbing capital held by the bank, the greater the protection of uninsured depositors and other bank creditors in an event of a bank failure.
A regulatory measure of the supplementary capital supporting the lending activity of a bank. Less loss-absorbing than Tier 1 capital, it includes undisclosed reserves, general loss reserves, and subordinated term debt.
Institutions believed to be so large, interconnected, or critical to the workings of the wider financial system or economy that their disorderly failure would impose significant costs on third parties or the aggregate financial system. See also Systemically important financial institutions.
A slice of a security—typically an asset-backed security—associated with a specific order for payment and level of risk and that is sold to investors separately. Tranche holders are paid starting with the “senior” tranches (least risky) and then working down through one or more “mezzanine” levels to the “equity” tranche (most risky). If some of the expected cash flows are not received, and after any cash flow buffers are depleted, the payments to the equity tranche are reduced. If the equity tranche is depleted, then payments to the mezzanine tranche holders are reduced, and so on up to the senior tranches.
Inflation-linked bonds issued by the U.S. Treasury that protect the holder from a decrease in purchasing power resulting from an increase in inflation.
An investment company that owns a fixed set of securities for the life of the company. That is, the investment company rarely alters the composition of its portfolio over the life of the company.
An investment made without the use of borrowed money.
An estimate of the monetary loss over a given period that is unlikely to be exceeded statistically at a given probability level (typically the 95 percent confidence level). See also CoVaR.
An econometric model used to capture the evolution of and interdependence among multiple time series, generalizing the univariate autoregressive models.
Chicago Board Options Exchange Volatility Index that measures market expectations of financial volatility over the next 30 days. The VIX is constructed from S&P 500 option prices.
Bank funding—in addition to customer deposits (retail funding)—in private markets to finance bank operations. Wholesale funding sources include debt issuance, short-term instruments such as certificates of deposit and commercial paper, repo transactions, and interbank borrowing.
To recognize in a set of accounts that an asset may be worth less than earlier supposed.
The relationship between the interest rate (or yield) and the time to maturity for debt securities of equivalent credit risk. In this GFSR, the interest (term) spread between the 10-year Treasury bonds and the 3-month Treasury bill.
The number of standard deviations by which an observation is above or below the mean. It is a dimensionless quantity derived by subtracting the mean of the sample from the observation and then dividing the difference by the standard deviation.
ANNEX SUMMING UP BY THE ACTING CHAIR
The Acting Chair made the following remarks at the conclusion of the Executive Board’s discussion of the Global Financial Stability Report on August 31, 2011.
Executive Directors broadly agreed with the staff’s analysis and expressed concern regarding the deterioration in global financial stability, following signs of an economic slowdown and renewed market anxieties about the euro area periphery and the U.S. fiscal and debt challenges. They noted that weaker growth could have a negative impact on both public and private sector balance sheets, with rising financing costs that could aggravate the fiscal position in many advanced economies. Directors observed that emerging market economies are generally in a stronger position, but also face difficult policy challenges, including those related to possible overheating. Directors underscored the importance of strong, coherent policy responses to address rising contagion risks and strengthen financial systems.
Directors noted that the financial crisis has entered a new difficult phase. In this context, they were concerned that political differences within the euro area may have delayed a lasting solution to the sovereign debt crisis. They also noted that difficulties in reaching a political consensus on medium-term fiscal adjustment in some other countries, including the United States, could undermine market confidence. Directors stressed, therefore, that a timely resolution of these difficulties would be critical to limiting risks to financial stability and addressing the fundamental challenges.
Directors observed that the adverse feedback loop between sovereign risk and bank balance sheets has intensified. They noted that risks of a funding disruption would rise if banks were unable to strengthen their balance sheets or sovereigns could not agree on credible medium-term fiscal plans in a timely manner. Directors concurred with the need to address vulnerabilities in the banking sector, including a further strengthening of bank capital buffers where necessary, to minimize risks to the real economy. Directors warned that lack of decisive action in this regard, as well as on the fiscal front, could trigger a negative spiral of rising funding costs and deteriorating debt dynamics.
Directors noted the staff’s analysis of the potential impact of sovereign stress on bank balance sheets in Europe. A number of Directors welcomed this analysis as important to an understanding of contagion risks, but a number of others expressed concerns about the underlying methodological and data choices and the challenges in communicating it to the public. Many Directors saw scope for further work in this area.
Directors agreed that significant measures would be needed to restore the financial system to health. In the United States, policymakers still need to address the legacy of the financial crisis as it persists in household balance sheets. In the euro area, Directors called for swift implementation of the decisions taken at the Euro Area Summit in July 2011.
Directors noted that low policy interest rates, while necessary in many countries under current conditions, could pose long-term financial stability risks. The “search for yield” could push some market segments into excessive leverage, exposing them to higher risk. Directors concurred that a prompt completion of balance sheet repair in financial institutions could help contain these risks. A few Directors noted, however, that such risks could be overstated.
Directors shared the view that, while the recent surge of capital inflows to emerging markets has had beneficial effects, these inflows have fueled liquidity and credit expansion that could be destabilizing in case of a global growth slowdown, a rapid increase in funding costs, or a sudden reversal of the flows. Directors agreed that a variety of measures, combining macroeconomic and macroprudential tools, could help limit the buildup of financial vulnerabilities, and stressed the importance of closely monitoring credit growth.
Overall, Directors agreed that the key near-term policy challenges should center around implementing coherent solutions to reduce sovereign risks, introducing credible efforts to rebuild the strength of the financial system in advanced economies, and controlling the risks of overheating and asset bubbles in emerging market economies. They also stressed that the international financial reform agenda needs to be completed as soon as possible. A few Directors were mindful of the possible adverse impact on bank lending and growth of an accelerated compliance with Basel III capital standards. A number of Directors noted that following up on the implementation of the previous advice and more specific policy recommendations by staff would help focus the discussion and more clearly define policy options.
Directors welcomed the analysis in Chapter 2 of the forces driving global asset allocation by long-term investors. They noted that public and private pension funds, insurance companies, and asset managers have altered their behavior since the crisis by focusing more on market, liquidity, and sovereign risks. They agreed that the generalized move to safer, more liquid securities may limit the stabilizing role that long-horizon investors can play in global markets.
Directors observed that the main factors underlying the long-term trend toward emerging market assets are strong domestic economic growth prospects and lower perceived country risk. They noted the risk of a reversal of these flows, especially in the current times of heightened global risk aversion, but agreed that the longer-term trend favoring emerging markets is likely to continue. To mitigate the chances of a sudden reversal, Directors called on policymakers in emerging markets to focus their attention on maintaining strong and stable growth, and financial system resiliency.
Directors welcomed the analysis in Chapter 3, which takes a step forward in the design and operation of macroprudential policy frameworks. In particular, they appreciated the discussion of the macro-financial linkages and concurred that understanding the sources of shocks would help the design of policy instruments. Directors concluded that establishing macroprudential frameworks would require careful consideration, taking into account country-specific circumstances.
Note: This chapter was written by Srobona Mitra (team leader), Jaromír Beneš, Silvia Iorgova, Kasper Lund-Jensen, Christian Schmieder, and Tiago Severo. Research support was provided by Ivailo Arsov and Oksana Khadarina.
Systemic risk is the risk of disruptions to financial services that is caused by an impairment of all or parts of the financial system, and can have serious negative consequences for the real economy (IMF-BIS-FSB, 2009; IMF, 2011b). Systemic risk is driven by economic and financial cycles over time, as well as by the degree of interconnectedness of financial institutions and markets.
For the precrisis literature, see Enoch and Ötker-Robe (2007) and references therein. Some recent studies include Mendoza and Terrones (2008); Barajas, Dell’Ariccia, and Levchenko (forthcoming); De Nicoló and Lucchetta (2010); Claessens, Kose, and Terrones (2011a and 2011b); Kannan, Rabanal, and Scott (2009a and 2009b); Borio and Drehmann (2009); Drehmann, Borio, and Tsatsaronis (forthcoming).
The IMF and major central banks have developed one or more versions of these DSGE macroeconomic models to study the effectiveness and desirability of different macroeconomic policies (Roger and Vlček, 2011 and forthcoming). More recently, DSGE models have also been used for forecasting purposes. For example, Smets and Wouters (2007) show an application of Bayesian techniques for the estimation of DSGE models that yields good forecasting properties.
Bordo and others (2001), Reinhart and Rogoff (2009), and Mendoza and Terrones (2008) have compiled vast amounts of evidence about various drivers of boom-and-bust cycles across numerous countries over time. Moreover, Borio and Drehmann (2009), Drehmann, Borio, and Tsatsaronis (forthcoming), and Ng (2011) study the performance of alternative indicators of financial crisis; those studies show that some variables, including measures of excessive credit growth, could forecast crises occurring one to three years ahead. De Nicoló and Lucchetta (2010) explore the links between credit growth and GDP growth with a dynamic factor model using the concept of tail risk (the risk of negative shocks of low probability but high impact).
Such gains could result from one or more developments, including new technologies, new resources, and institutional improvements.
The model uses the concept of financial friction (see Bernanke, Gertler, and Gilchrist, 1999), in which limited enforcement of loan covenants gives the borrower an incentive to default and allows the lender to seize the collateral. The aggregate risk in the model arises from procyclicality in the system; the model does not take into account the systemic risk arising from interconnected-ness in the financial system.
No distinction is made between various types of assets—productive real capital, real estate, claims to investments, equity shares, and so on.
The analysis assumes “irrational” bubbles—investors’ and traders’ sentiments and expectations are driven by extraneous or nonfundamental factors such as fads, fashions, rumors, and informational “noise,” which can disrupt and destabilize asset markets and generate excessive volatility in asset prices (Kindle-berger, 1989). A “rational” bubble, on the other hand, reflects the presence of self-fulfilling (rational) expectations about future increases in the asset price raising the possibility of deviation of the asset price from the fundamental value (Blanchard, 1979; Blanchard and Watson, 1982; Froot and Obstfeld, 1991; and Evans, 1991). In a rational bubble, stock price growth contains occasional corrections when investors realize the price is not increasing as expected, as opposed to diverging continually as in the “irrational” case.
Dell’Ariccia and Marquez (2006) show that, as more and more customers apply for bank loans, banks weaken their lending standards and collateral requirements to raise market share by undercutting their competitors.
A bubble scenario could arise if the actual productivity gains are less than expected.
Baseline parameterization drives the impulse responses that are used to construct Figure 3.2. Different parameterizations of the model are analyzed in Annex 3.1 and Beneš, Kumhof, and Vávra (2010). Impulse-response functions represent the deviations of macroeconomic variables from their regular path as a consequence of a disturbance, keeping all other elements constant. They compare the performance of the economy over time after a shock relative to a nonshock scenario. The length of the shocks is approximated using information about the time shocks tend to last in previous cases in a set of representative countries.
Only four indicators have been shown in the figure for analytical purposes, but there are many other indicators that could be shown. Also see notes to Figure 3.2.
This is not the loan-to-value (LTV) ratio imposed by banks, but literally the observed amount of credit for a given level of asset value.
The ratio of credit to asset value actually declines slightly with the onset of an asset bubble because the bubble increases the value of assets that collateralize loans before lending increases enough to boost the ratio.
Additional indicators are based on Shin (2010), Sun (2011), and IMF (2009). Ideally, also included would be the capital adequacy ratio, shown above to be informative; however, for the entire time period, it is available for only a few countries.
The FSI is a monthly indicator of national financial system strain. See Cardarelli, Elekdag, and Lall (2011) for advanced economies; and Balakrishnan and others (2009) for emerging economies. This index—not to be confused with the Financial Soundness Indicators—relies on price movements relative to past levels or trends. For advanced economies, the index is the sum of seven variables, each of which is normalized to have a zero mean and a standard deviation of one: (i) the banking-sector beta (a measure of the correlation of bank equity returns with overall equity market returns); (ii) the TED spread (the difference between the three-month Treasury bill rate and the Eurodollar rate); (iii) term spreads (the difference between short- and long-term government bonds); (iv) stock market returns; (v) stock market volatility; (vi) sovereign debt spreads; and (vii) exchange market volatility. For emerging economies, the FSI comprises five variables (it excludes the TED and term spreads from the preceding list of seven and uses an index of exchange market pressure instead of exchange market volatility). See IMF (2008) for more details and Box 3.2 for details on the methodology. The average 5th percentile value of the FSI was 7.4 at the beginning of the 2007—09 financial crisis and 9.7 at its peak.
The broader credit measure includes private-sector credit from banks (derived from monetary statistics) and cross-border loans to domestic nonbanks (derived from “other investment, liabilities” of international investment position statistics). The number of countries in the sample falls considerably when the broader measure is included.
The credit-to-GDP gap (Borio and Drehmann, 2009; and Drehmann, Borio, and Tsatsaronis, forthcoming) and change in the credit-to-GDP ratio are prime candidates for comparison. The former is the deviation of the credit-to-GDP ratio from a recursive Hodrick-Prescott filter trend. The advantage of the gap measure is that it is cumulative and takes into account the country-specific trend. Its disadvantage is that a gap of zero could still reflect a very high rate of credit growth, which is the core concern for financial stability. In the same vein, the indicator is less convenient for policy purposes, and ultimately macroprudential policies will have to target credit growth as such (that is, the gap has to be translated back into credit growth). The advantage of the ratio measure is that it readily focuses on the pace of credit growth. Its main disadvantage is that it omits cumulative aspects.
In this context, foreign liabilities refer only to loans and deposit liabilities of the private sector and are taken from balance of payments statistics (changes in the international investment position for banks and nonbanks under “other investment, liabilities”). Instead of focusing on the current account deficit, only the above set of capital inflows are emphasized here, since countries reliant on such flows have been more prone to the recent crisis, at least in emerging Europe (Cihak and Mitra, 2009).
This measure could be interpreted as a measure of noncore/ core liabilities, which tend to grow with assets. See Shin and Shin (2011).
Equity prices are a part of the FSI indicator and hence tend to be contemporaneous with distress window peaks. For this reason, equity prices were not included in the event study.
The noise-to-signal ratio is defined as the proportion of Type II errors (cases with indicator signaling a crisis as a fraction of cases in which crisis did not occur) divided by the proportion of legitimate signals (cases with indicator signaling a crisis as a fraction of cases in which crisis did occur). See Kaminsky, Lizondo, and Reinhart (1998); Berg and others (2000); and Box 3.2.
The Laeven-Valencia index of episodes is a broad, coincident indicator for full-blown financial crises that uses government intervention in the financial sector to date the episodes. On the other hand, the FSI used in the previous section is an indicator of financial stress that might not materialize into a full-blown crisis. The advantage of the LV index is that it covers 169 countries rather than the 40 countries covered by the FSI, but a considerable drawback is its annual frequency and the scarcity of crisis occurrences—at most one crisis per country for most countries and 109 overall.
The sample is reduced for different indicators based on data availability. Results are similar for a one-year lag.
Borio and Drehmann (2009), who advocate this measure, consider a small set of advanced economies only.
The stock of cross-border loans is derived from other investment liabilities data from the balance of payments of the IMF’s International Financial Statistics (IFS). The latter source of data was chosen to maintain consistency with data on credit, which comes from the monetary statistics of the IFS. However, the number of countries fall dramatically both because of data availability and coverage of the Laeven-Valencia index.
If the predictive power of an indicator is very low, then it is hard to choose meaningful thresholds for it.
A probit (unbalanced panel data) model with country fixed effects is estimated across 94 countries (with advanced, emerging, and low-income economies) over 1975—2010 using annual data. The fixed effects of a country denote the time-invariant characteristics that affect the crisis probability; a country with very high fixed effects (80th percentile) is termed “high risk.” Using the Laeven and Valencia (2010) definition of crisis in the form of a crisis dummy (1 for crisis and 0 otherwise), the estimation evaluates the ability of the different indicators to explain the probability of crises at three different forecast horizons—one, two, and three years.
The estimation of the multivariate probit model is based on a smaller dataset because of data gaps for equity prices. The dataset shrunk further when other variables were included. Even so, indicators like the growth in foreign liabilities and the level of the loan-to-deposit ratio were tested and found to increase the marginal effect of credit aggregates on the probability of crisis.
The CoVaR is the Value at Risk of the financial system conditional on institutions being under distress. An institution’s contribution to systemic risk is the difference between the CoVaR for tail-risk episodes and the CoVaR at the median state. The time-varying CoVaR is estimated by quantile regressions of the returns of the financial system on the returns of an institution and other state variables. The latter includes the yield curve (the difference between interest rates on long-term Treasury bonds and short-term Treasury bills) and the spread between the London Interbank Offered Rate (LIBOR) and the overnight indexed swap (OIS).
See IMF (2011b).
See IMF (2011b). Monetary policy, with a separate objective and policy tool, is characterized by a simple inflation-targeting rule in a flexible exchange rate regime. Banks are subject to fixed microprudential capital requirements to address idiosyncratic credit risk. The macroprudential policy requirements are added due to concerns about banks’ exposure to aggregate risk. Even though the risk could be addressed by containing the cycles of financial risk and addressing the interconnectedness of financial institutions, only the former is taken up in this section, as interconnectedness has not yet been introduced in the structural model.
Banks do not expand credit as much during the boom phase because they fear they might not be able to satisfy the higher requirements when they are confronted with a future reversal. Hence, leverage is endogenously less procyclical in the model.
It can be argued that although the two policies, monetary and macroprudential, have different objectives and use different tools, their eventual impact on credit aggregates and on real economic cycles can be very similar, potentially reinforcing or offsetting each other. See Kannan, Rabanal, and Scott (2009a) on how welfare improves when a credit aggregate is included in the monetary policy rule; and Jacome and others (forthcoming) on institutional arrangements for macroprudential policies.
See IMF (2010c).
Prepared by Jaromír Beneš.
In that case, the bank can pay a collection cost to make the defaulted borrower repay the loan in full; the probability that the bank succeeds is set to a number arbitrarily close to 1.
The terms related to a situation in which the household member succeeds in walking away from the loan are dropped; the probability of such an outcome is set to a numerically negligible value.
Such a policy is not termed a managed exchange rate regime because it is typically implemented through sterilized interventions.
Prepared by Kasper Lund-Jensen.
This model specification corresponds to xi, t—h = (ΔCtGi, t—h, Δln(equity price)i, t—) and θ = (θ1, θ2)T in equation (1).
The change in the CtG ratio has a significant impact on the crisis probability at both a one- and two-year forecast horizon (Table 3.5). To incorporate information from both lags, the change in the CtG ratio was defined in the forecasting exercise as ΔCtGt = (CtGt - CtGt–2)/2.
Prepared by Srobona Mitra.
LIBOR is the London Interbank Offered Rate, and OIS is the overnight indexed swap rate.