Financial Crises
Chapter

Chapter 4. Procyclicality and the Search for Early Warning Indicators

Author(s):
Stijn Claessens, Ayhan Kose, Luc Laeven, and Fabian Valencia
Published Date:
February 2014
Share
  • ShareShare
Show Summary Details
Author(s)
Hyun Song ShinPresented at the IMF conference on “Financial Crises: Causes, Consequences, and Policy Response,” Washington, DC, September 14, 2012. The author thanks Stijn Claessens, Ayhan Kose, Luc Laeven, and Fabian Valencia for comments on an earlier draft, and Laura Yi Zhao for research support.

Finding a set of early warning indicators that can signal vulnerability to financial turmoil has emerged as a policy goal of paramount importance in the aftermath of the 2007–09 global financial crisis. There is a large literature on early warning indicators for crises, described well in Chamon and Crowe (2013). The crises in emerging market economies in the 1990s gave impetus to the work, which has been further developed in the aftermath of the 2007–09 global financial crisis that engulfed both advanced and emerging market economies.

The literature to date could be described as eclectic and pragmatic. It is eclectic in the sense that the exercise involves a wide variety of inputs, covering external, financial, real, and fiscal variables, as well as institutional and political factors and various measures of contagion. In their overview of the literature as of 1998, Kaminsky, Lizondo, and Reinhart (1998) catalogue 105 variables that had been used up to that point.

The literature has also been pragmatic in that it has focused on improving measures of goodness of fit rather than focusing on the underlying theoretical themes that could provide bridges between different crisis episodes.1 For instance, crises in emerging market economies have typically been distinguished from those in advanced economies. Different sets of variables enter into the exercise for each category. Emerging market economy crises focus on capital flow reversals associated with “sudden stops,” for which variables such as external borrowing denominated in foreign currency take center stage; for advanced economies, housing booms and household leverage take on importance. Claessens and others (2010) examine the evidence of the 2007–09 financial crisis on both categories.

The distinction between emerging market and advanced economies is also reflected in the work of the official sector. The IMF has added a new Vulnerability Exercise for Advanced Economies to an existing Vulnerability Exercise for Emerging Market Economies, which both feed into a joint early warning exercise with the Financial Stability Board (IMF, 2010; Chamon and Crowe, 2013).

Although the compartmentalization into emerging market and advanced economies helps improve the goodness of fit, it tends to obscure the common threads that tie together emerging market and advanced economy crises. The capital flow reversals in Spain and Ireland in the European crisis have many of the features of a “sudden stop,” except that the outflow of private sector funds has been compensated for by an inflow of official funds (Merler and Pisani-Ferry, 2012). However, because the euro area crisis is taking place within a common currency area, the traditional classification of emerging market “currency crises,” in which currency movements play a key role, does not fit easily in the empirical exercise.

Given the common threads that run through apparently disparate crises, it can be useful to take a step back from the practical imperative of maximizing goodness of fit and instead consider the conceptual underpinnings of early warning models, which is the purpose of this chapter.

What follows suggests that the procyclicality of the financial system provides an organizing framework for selecting indicators of vulnerability to crises, especially those indicators that are associated with banks and financial intermediaries more generally. More specifically, the chapter examines three broad sets of indicators for early warning purposes, and assesses their relative likelihood of success. The three sets of indicators are

  • indicators based on market prices, such as credit default swap (CDS) spreads, implied volatility, and other price-based measures of default or distress;

  • gap measures of the credit-to-GDP ratio; and

  • banking sector liability aggregates, including monetary aggregates.

To anticipate the conclusions, the first approach (based on market prices) seems most appropriate for obtaining indicators of concurrent market conditions but unlikely to be useful as early warning indicators that provide enough time for meaningful remedial action.

The credit-to-GDP gap measure is a distinct improvement over the first as an early warning indicator, with a good pedigree from the work of economists at the Bank for International Settlements. It has been explored extensively as part of the Basel III bank capital rules. Yet, there are doubts about its usefulness as a real time measure, or as a measure that yields a threshold that can be applied uniformly across countries.

That leaves the third approach, based on bank liability aggregates, including various components of the money stock. The chapter suggests that this third approach is the most promising because it preserves the advantages of the credit-to-GDP gap measure but also stands a good chance of yielding indicators that can be used in real time.

The downside, however, of the monetary approach is that any measure derived in this way will need to find meaning by reference to specific institutional features of the financial system rather than by being applied in an unthinking way.

In addition, the traditional thinking behind the definitions of monetary aggregates will have to be transcended to make the approach useful. Whereas traditional definitions of monetary aggregates exclude the liabilities between financial intermediaries, such liabilities turn out to be perhaps the most informative of them all.

Price-Based Early Warning Indicators

Figure 4.1 illustrates the credit default swap (CDS) spreads of Bear Stearns and Lehman Brothers. Panel b gives the longer perspective and shows how the spreads increased sharply with the onset of the crisis.

Figure 4.1CDS Spreads for Bear Stearns and Lehman Brothers

Source: Thomson Reuters Datastream.

Note: CDS = credit default swap.

What is remarkable is how tranquil the CDS measure is before the crisis. There is barely a ripple in the series in the period 2004 to 2006 when vulnerability to the financial crisis was building up. Panel a, which plots the CDS series for the precrisis period of January 2004 to January 2007, shows that CDS spreads were actually falling, dipping below 20 basis points at the end of 2006. Other price-based measures, such as value-at-risk, implied volatility, and structural models of default based on equity prices, all painted the same picture.

The failure of price-based measures to succeed as early warning indicators can be traced to their implicit premise that market signals and the decisions guided by those signals always interact in a stabilizing virtuous circle. However, sometimes they go astray and act in concert in an amplifying vicious circle in which market signals and decisions guided by those signals reinforce an existing tendency toward procyclicality. Some of the forces toward procyclicality are described in the 2011 Mundell-Fleming lecture (Shin, 2012).

As an illustration of the outcome of the tendency toward procyclicality, the scatter chart in Figure 4.2 plots how much the change in the size of Barclays’ balance sheet—representative of a typical global bank—is financed through equity and how much through debt. It also shows the change in risk-weighted assets as the balance sheet grows or shrinks.

Figure 4.2Two-Year Changes in Assets, Debt, Equity, and Risk-Weighted Assets of Barclays (1992–2010)

Source: Barclays’ Annual Report.

The fact that risk-weighted assets barely increase in Figure 4.2 even as raw assets are increasing rapidly is indicative of the lowering of measured risks (such as spreads or value-at-risk) during lending booms. Lower measured risks and lending booms thus go together. The causation in the reverse direction will also have been operating: the compression of risk spreads is induced by the rapid increase in credit supply chasing available borrowers. Such two-way causation lays the groundwork for a feedback loop in which greater credit supply and the compression of spreads feed off each other.

The procyclicality evident in Figure 4.2 poses hard challenges for the traditional thinking that places faith in market discipline as an integral part of financial regulation, relying on prices to issue timely warning signals. Indeed, market discipline was one of the three pillars of the Basel II framework for international bank regulation. Economists associated with the Shadow Financial Regulatory Committee were influential in this regard. Calomiris (1999) argues for rules requiring banks to maintain a minimum amount of subordinated debt, the rationale being that banks that take on excessive risk will find it difficult to sell their subordinated debt, and will be forced to shrink their risky assets or to issue new equity to comply with the discipline imposed by private uninsured creditors. However, the experience in the run-up to the 2007–09 crisis showed how market risk premiums erode so as to nullify market discipline.

Larry Summers’s quip that the achievement of finance researchers is to show that “two quart bottles of ketchup invariably sell for twice as much as one quart bottles of ketchup” (Summers, 1985, p. 634) is related to why price-based measures of early warning indicators are likely to fail. Absence of arbitrage means that prices at a point in time are consistent, but they are liable to flip to distress mode (again, fully consistent across assets) with the onset of the crisis. If the task is to give warning of the onset of the crisis, price-based measures have little to say about the transition.

Because the onset of a crisis is often accompanied by run-like events, the switch from a benign environment to a hostile one can be precipitous. The global games literature illustrates how the transition into financial distress—the “tipping point”—is associated with self-reinforcing effects between individual constraints and market outcomes, but how the onset of a crisis is triggered by apparently small changes in the underlying fundamentals. Outwardly, the switch has the flavor of a self-fulfilling crisis. Goldstein (2012) discusses how empirical research should take account of such tipping points, and shows how the global games framework (Morris and Shin, 1998, 2001, 2008) can be effectively invoked in the modeling exercise.

To the extent that market prices have been useful for early warning exercises at all, their usefulness comes precisely when the market price of risk is too low rather than too high. Thus, it is when asset prices are too high relative to some benchmark that warning signs are appropriate.

In their paper on the U.S. housing market, Himmelberg, Mayer, and Sinai (2005) argue that a high price-to-rent ratio or a high price-to-income ratio need not be indicators of a housing bubble because discount rates implied by low long-term interest rates had also fallen. But discount rates are prices, so the combination of low discount rates and high housing prices is arguably the kind of point-in-time consistency in prices that Summers (1985) had in mind.

Credit-to-GDP GAP Indicators

Under the Basel III framework, the ratio of credit to GDP takes a central role as the basis for the countercyclical capital buffer. This ratio has been shown to be a practical indicator of the stage of the financial cycle, notably by Borio and Lowe (2002, 2004). To the extent that procyclicality drives financial vulnerability, detecting excessive credit growth is crucial. Normalizing credit to some underlying flow fundamental measure such as GDP and detecting deviations from trend would be one way to use the notion of excessive credit growth.

However, although the existence of a credit boom is clear in hindsight, there are several challenges to using the deviation of the credit-to-GDP ratio from trend as an early warning indicator in real time.

The first challenge is the difficulty of estimating the trend that serves as the benchmark for “excessive” growth. The difficulty is not unique to the credit-to-GDP ratio, but one shared by other macroeconomic time series. Edge and Meisenzahl (2011) find that ex post revisions to the credit-to-GDP ratio gap in real time are sizable for the United States and as large as the gap itself. The source of the ex post revisions is not the revision of the underlying data, but rather the revision of the estimated trend measured in real time.

The second difficulty is that credit growth and GDP dance to somewhat different tunes over the cycle, so the ratio of the two may sometimes issue misleading signals. Bank lending in particular may be influenced by preexisting contractual commitments, such as lines of credit, which are drawn down during the crisis. Ivashina and Scharfstein (2010) document the impact of such lines of credit on credit growth during the 2007–09 crisis. Therefore, lending may continue to increase for some time after the onset of a crisis.

Figure 4.3 is taken from Repullo and Saurina (2011) and shows the credit-to-GDP ratio for the United Kingdom and its Hodrick-Prescott (H-P)-filtered trend (panel a). The H-P filter parameter is set at λ = 40,000 as recommended by the Basel Committee, which effectively means a linear trend. Panel b shows the gap between the credit-to-GDP ratio and the trend.

Figure 4.3Credit-to-GDP Ratio and GDP Growth for the United Kingdom

Source: Repullo and Saurina (2011).

From panel b of Figure 4.3, it can be noted the gap measure is large even as GDP growth is falling very sharply during the crisis. Thus, the ratio of the two gives a misleadingly large credit-to-GDP ratio during the crisis.

Basel III discussions have given a great deal of prominence to the credit-to-GDP gap measure (BCBS, 2009, 2010). To the extent that the Basel rules are expected to be applied uniformly (or at least in a consistent manner) to all Basel Committee member countries, finding common thresholds for the credit-to-GDP ratio would be a basic requirement.

Bank Liability Aggregates, Including Some Monetary Aggregates

Rapid growth of bank lending is mirrored on the liabilities side of the balance sheet by shifts in the composition of bank funding. As intermediaries that borrow to lend, banks must raise funding to lend to their borrowers. When credit is growing faster than the available pool of funds that are usually drawn on by the bank (core liabilities), the bank will turn to other, noncore sources of funding to support its credit growth.

Thus, the ratio of noncore to core liabilities serves as a signal of the degree of risk taking by the bank and hence of the stage of the financial cycle. Hahm, Shin, and Shin (2012) conduct a cross-country panel probit study and find that the ratio of noncore to core liabilities (especially noncore liabilities to foreign creditors) consistently emerges to be the most robust predictor of a currency crisis or credit crisis.

The distinction between core and noncore bank liabilities has similarity to monetary aggregates. Traditionally, the importance of monetary analysis for the real economy rested on a stable money demand relationship that underpinned the link between money and macroeconomic variables. Money demand is seen to be the result of portfolio decisions of economic agents choosing between liquid and illiquid claims, regardless of whether based on an inventory holding of money for transactions purposes. For this reason, the traditional classifications of monetary aggregates focus on the transactions role of money as a medium of exchange.

However, unlike commodity money, monetary aggregates are the liabilities of banks and therefore have an asset-side counterpart. Recognition of the asset-side counterpart of money and of the determinants of bank lending focuses attention on the supply of money by banks. Indeed, rather than speaking of the demand for money by savers, the relationship could be turned on its head to speak of the supply of funding by savers. Similarly, by speaking of the supply of money as the demand for funding, the shift in language serves to focus attention on the banking sector and its balance sheet management over the cycle.

But monetary aggregates are traditionally measured by netting out claims between banks. For financial stability purposes, however, the claims between banks—especially when they are cross-border—take on great significance.

Figure 4.4 plots the four-quarter growth in cross-border assets and liabilities of euro area banks in euros. The destination of euro-denominated lending reached outside the euro area as euro area banks expanded into Central and Eastern Europe. However, the cross-border euro-denominated liabilities series in Figure 4.4 can be seen as noncore liabilities generated through capital inflows. From 1999:Q1 to 2008:Q3, cross-border liabilities rose almost 3.5-fold from 1.56 trillion euro to 5.4 trillion euro. This rapid spurt translates into a constant quarterly growth rate of 3.33 percent, which when annualized is close to 14 percent.

Figure 4.4Cross-Border Euro-Denominated Assets and Liabilities of Euro Area Banks

Source: Bank for International Settlements, Locational Statistics, Table 5A.

Core and Noncore Liabilities in China

However, what counts as core or noncore will depend on the financial system and the institutions. For economies with banks operating in developed, open capital markets, noncore funding will typically take the form of wholesale funding of the bank from capital markets, sometimes denominated in foreign currency. However, if the economy has a closed capital account, and if banks are prevented from accessing capital market funding from abroad, what counts as noncore funding could be quite different.

Compare the Republic of Korea and China. Figure 4.5 plots the monthly growth rates of various banking sector liability aggregates for Korea (panel a) and for China (panel b). The growth rates have been filtered through an H-P filter at business cycle frequency. Note that the H-P filter is used here with hindsight to highlight differences in time series patterns, not the real-time, trend-finding exercise of Basel III.

Figure 4.5Monthly Growth Rates of H-P-Filtered Bank Liability Aggregates for Korea and China

Source: Author’s compilation.

Note: H-P = Hodrick-Prescott; for this analysis, the H-P filter is set to 14,400.

In Korea, banks have access to capital markets, either directly or through the branches of foreign banks operating there. For this reason, the most procyclical components of the bank liability aggregates are those associated with wholesale funding, especially the series for the foreign exchange–denominated liabilities of the banking sector.2 Before the 1997 Asian financial crisis and the 2007–09 global crisis, noncore liabilities grew rapidly, only to crash with the onset of the crises. In contrast, the growth of broad money (M2), reflecting household and corporate deposits, is much less variable over the cycle.

However, panel b of Figure 4.5 demonstrates that in China, the subcomponents of M2 show considerable variation in their time series properties, with corporate deposits displaying the telltale procyclical patterns as compared with household deposits.

For an economy such as China, where banks are prevented from accessing international capital markets in the way that Korean banks do, applying the liability classifications from Korea of core and noncore would be inappropriate.

Instead, more thought is needed on how financial conditions are transmitted across the border into China. Just as water finds cracks to flow through, even a closed financial system is not entirely immune to global financial conditions, especially a highly trade-dependent economy such as China. If banks are prevented from accessing international capital markets, nonfinancial firms will be the conduit for the transmission of financial conditions.

Figure 4.6 depicts the activities of a Chinese nonfinancial firm with operations outside China, which borrows in U.S. dollars from an international bank in Hong Kong SAR, China, and posts renminbi deposits as collateral. The transaction would be akin to a currency swap, except that the settlement price is not chosen at the outset. The transactions instead resemble the operation of the old London Eurodollar market in the 1960s and 1970s. For the Chinese corporate, the purpose of having U.S. dollar liabilities and holding the proceeds in renminbi may be to hedge its export receivables, or simply to speculate on renminbi appreciation.

Figure 4.6Structure of Borrowing Relationships for Nonfinancial Corporations in China

Source: Author’s representation.

Note: RMB = renminbi.

Figure 4.7 provides evidence for the transactions depicted in Figure 4.6. Figure 4.7 plots the claims and liabilities of Hong Kong banks in foreign currency to customers in China. Foreign currency, in this case, would be mainly U.S. dollars for the assets and mainly renminbi for the liabilities. Both have risen dramatically in recent years, reflecting the rapidly increasing U.S. dollar funding of nonfinancial corporates.

Figure 4.7Hong Kong Banks’ Claims and Liabilities to Nonbank Customers in China in Foreign Currency

Source: Hong Kong Monetary Authority.

The procyclical pattern in corporate deposits in panel b of Figure 4.5 may be caused by such activities of nonfinancial corporates.

In addition, such activities of nonfinancial corporates may explain why China experienced dollar shortages in 2011 with the deterioration of global funding markets caused by the crisis in Europe. During this period the renminbi came under pressure, depreciating against the U.S. dollar. Although China’s banking system is largely closed, the global activities of its nonfinancial firms will be reflected in the corporate deposits within M2 when those firms hold the proceeds of dollar liabilities in their accounts in China.

Figure 4.8 illustrates the growth in the component of the money stock that is due to the deposits of corporates rather than households. Panel a shows the time trend in personal deposits and corporate deposits; panel b shows the ratio of corporate to personal deposits. The proportion of corporate deposits has increased in recent years, consistent with the operations of Chinese corporates as shown in Figure 4.6.

Figure 4.8Components of China’s Monetary Aggregates

Source: People’s Bank of China.

The excess liquidity generated by the activity of nonfinancial corporates in China is an important element of the lending boom in China, and is reminiscent of the lending boom in Japan in the 1980s following financial liberalization that allowed Japanese companies to access global capital markets. Both in Japan in the 1980s and in China more recently, monetary aggregates, especially corporate deposits, played the role of noncore liabilities in the way that foreign exchange borrowing by Korean banks plays the role of noncore liabilities in Korea.

The similarity between the foreign exchange liabilities in Korea and the corporate deposits in China is that both are liability components of banks. If the demarcation between core and noncore liabilities is correctly defined, the same method of tracking the ratio of noncore to core liabilities can serve as an early warning indicator of financial vulnerability.

References

    Basel Committee on Banking Supervision (BCBS)2009Strengthening the Resilience of the Banking Sector (Basel: Bank for International Settlements). http://www.bis.org/publ/bcbs164.pdf.

    Basel Committee on Banking Supervision (BCBS)2010International Regulatory Framework for Banks (Basel III)” (Basel: Bank for International Settlements). http://www.bis.org/bcbs/basel3.htm.

    BergAndrew and CatherinePattillo1999Are Currency Crises Predictable? A TestIMF Staff Papers Vol. 46 No. 2 pp. 10738.

    BorioClaudio and PhilipLowe2002Asset Prices, Financial and Monetary Stability: Exploring the NexusBIS Working Paper No. 114 (Basel: Bank for International Settlements).

    BorioClaudio and PhilipLowe2004Securing Sustainable Price Stability: Should Credit Come Back from the Wilderness?BIS Working Paper No. 157 (Basel: Bank for International Settlements).

    CalomirisCharles1999. “Building an Incentive-Compatible Safety NetJournal of Banking & Finance Vol. 23 No. 10 pp. 1499519.

    ChamonMarcos and ChristopherCrowe2013‘Predictive’ Indicators of Financial Crisesin Handbook in Financial Globalization: The Evidence and Impact of Financial Globalization ed. by GerardCaprio (London: Elsevier) pp. 499505.

    ClaessensStijnGiovanniDell’AricciaDenizIgan and LucLaeven2010Cross-Country Experiences and Policy Implications from the Global Financial CrisisEconomic Policy Vol. 62 pp. 26793.

    EdgeRochelle M. and Ralf R. Meisenzahl2011The Unreliability of Credit-to-GDP Ratio Gaps in Real Time: Implications for Countercyclical Capital BuffersInternational Journal of Central Banking Vol. 7 No. 4 pp. 26198.

    GoldsteinItay2012Empirical Literature on Financial Crises: Fundamentals vs. Panicin The Evidence and Impact of Financial Globalization ed. by G. Caprio (Waltham, Massachusetts: Academic Press) pp. 52334.

    HahmJoon-HoHyunSong Shin and KwanhoShin2012Non-Core Bank Liabilities and Financial VulnerabilityNBER Working Paper No. 18428 (Cambridge, Massachusetts: National Bureau of Economic Research).

    HimmelbergCharlesChristopherMayer and ToddSinai2005Assessing High House Prices: Bubbles, Fundamentals, and MisperceptionsNBER Working Paper No. 11643 (Cambridge, Massachusetts: National Bureau of Economic Research).

    International Monetary Fund2010The IMF-FSB Early Warning Exercise: Design and Methodological Toolkit” (Washington: International Monetary Fund). www.imf.org/external/np/pp/eng/2010/090110.pdf.

    IvashinaVictoria and DavidScharfstein2010Bank Lending during the Financial Crisis of 2008Journal of Financial Economics Vol. 97 No. 3 pp. 31938.

    KaminskyGracielaSaulLizondo and CarmenReinhart1998Leading Indicators of Currency CrisisIMF Staff Papers Vol. 45 No. 1 pp. 148.

    MerlerSilvia and JeanPisani-Ferry2012Sudden Stops in the Euro AreaReview of Economics and Institutions Vol. 3 No. 3.

    MorrisStephen and HyunSong Shin1998Unique Equilibrium in a Model of Self-Fulfilling Currency AttacksAmerican Economic Review Vol. 88 No. 3 pp. 58797.

    MorrisStephen and HyunSong Shin2001Rethinking Multiple Equilibria in Macroeconomic ModelingNBER Macroeconomics Annual 2000 Vol. 15 (Cambridge, Massachusetts: MIT Press) pp. 13961.

    MorrisStephen and HyunSong Shin2008Financial Regulation in a System ContextBrookings Papers on Economic Activity Vol. 38 No. 2 pp. 22974.

    RepulloRafael and JesúsSaurina2010The Countercyclical Capital Buffer of Basel III: A Critical Assessmentin The Crisis Aftermath: New Regulatory Paradigms ed. by MathiasDewatripont and XavierFreixas (London: Centre for Economic Policy Research). http://www.voxeu.org/sites/default/files/file/Crisis_Aftermath.pdf.

    ShinHyun Song2012Global Banking Glut and Loan Risk PremiumIMF Economic Review Vol. 60 No. 2 pp. 155192.

    ShinHyun Song and KwanhoShin2010Procyclicality and Monetary AggregatesNBER Working Paper No. 16836 (Cambridge, Massachusetts: National Bureau of Economic Research).

    SummersLawrence H.1985On Economics and FinanceJournal of Finance Vol. 40 No. 3 pp. 63335.

The pragmatic focus has also meant that traditional regression techniques, such as the probit model used in Berg and Pattillo (1999), have increasingly given way to nonparametric techniques that minimize the signal-to-noise ratio as in Kaminsky, Lizondo, and Reinhart (1998). Nonparametric techniques fare better when there are a large number of explanatory variables.

The other noncore liabilities are bank debentures, repos, and other nondeposit items such as promissory notes (Shin and Shin, 2010).

    Other Resources Citing This Publication