Not all crises are the same

“When the rest of the world are mad we must imitate them in some measure” - John Martin, of Martin’s Bank

Some banking crises have been sudden (panics); some arise as consequences of imbalances: government debt, balance of payments imbalances, stock market crises. And they are costly.

Banks participate in all crises, and they exacerbate them: bad banking and bad policy. Every crisis leads to a call for better prudential regulation.

But cross-country empirical evidence suggests that policy is best directed when focusing on: information disclosed to market participants, market discipline bankers’ behavior, and incentives in the industry (including regulators).

This article studies the post Bretton-Woods era, providing historical background in section 1. Then on section 2 a distinction is made among three intervening facts: mismanagement, government intervention, and macroeconomic shocks. Section 3 reviews the aspects of crises analyzed by theoreticians. Section 4 discusses the costs of crises. Section 5 underscores the importance of prevention and corrective policy. Section 6 posits the extent to which recent crises are recent phenomena. The conclusion reached in the article is that regulation should focus on diminishing both frequency and cost of [inevitable] financial crises.

Early history

Banking crisis: the widespread insolvency of banks leading to closures, mergers, takeovers, or injections of government resources.

Early banking (Europe, XIII century) faced uncertainty not seen by the least developed systems of our time: wars lost, plagues, shortage of coins, ships sunk, defalcation, and monarchs unable/unwilling to repay.

Kindleberger and Aliber (2011): a banking crisis occurs nearly once every 10-year frequency in Europe and North America in the 1800-1945 period. Largely because of isolation and tight regulation, the 1945-1970 period was one of great stability, which ended with the relaxation of those policies.

From the 1970’s on, there is: weakened fiscal discipline, surges in inflation and abandonment of exchange rate (ER) pegs. By 1997, 60% of IMF members had experienced banking crises, and it persisted as of 2013 (year of publication).

Diverse Origins: Management, Government, and Macroeconomics in Recent Crises

Systemic banking crises are inextricably linked to macroeconomic crises, raising the question of which causes which. But management and government interference are also to blame (bad banking and bad policy).

Management and fraud

Large deposits looted by insiders and not registered as liabilities of the bank - a bank of systemic importance - eventually requires central bank loans and ends up destabilizing the macroeconomy. This happened to banks of Venezuela (1994) and Dominican Republic (2003). It also happened to multinational, Luxembourg’s BCCI, in Africa.

There’s also the rogue traders: Nick Leeson (Barings), Jérome Kerviel (Société Générale), Bernie Madoff (Ponzi scheme). These are typically discovered in periods of asset market decline. Some other examples are Crédit Lyonnais and Méridien BIAO (Central Africa), banks that operated under very weak lending policies.

The Tequila crisis (1994) was partly due to mismanagement: poor enforcement of capitalization rules and insider lending (loans to bank officers). Little shareholder equity led to high risks.

Liberalizations have been linked to banking crises, as new risks can be undertaken without the adequate skills to cope. For example, liberalization leads to changes in relative prices, thus altering the creditworthiness of borrowers - but that’s difficult to perceive.

Portfolios tend to skew after liberalization, as banks look for exposure to different risks. But banks are big investors who can shift asset prices and cause an unjustified boom. Also, state-owned banks have employees pursuing careers in politics, not in banking.

The transition from planned economy to free-market economy proved to be breeding ground for banking crises. The high inflation that came after helped relieve the burden on borrowers.

However, banking crises resulted in large costs on governments (thus, on taxpayers). It costed China USD 350 billion between 1998 and 2006.

Government policies

In some countries, as in french-speaking West Africa, governments were too influential in banking operations, resulting in poor managerial practices and credits granted by poorly capitalized banks (directed credits).

Because governments can impose tax-like measures (such as arbitrary fix exchange rates and exchange-control regulation), banks are always dependent on the willingness of states to let them operate profitably. Such dependence makes them fragile in autocracies and populist democracies. That was the case of Argentina (2001) with the forced conversion of foreign currency at an arbitrary (unfavorable) rate.

Calomiris and Haber (2013): politics is an important determinant of crises. In order to cement political support, the constituents enter a “game of bank bargains”, affecting the availability of credit.

Macro boom and bust

Expectations about business prospects dominates the explanations offered to crises: overoptimism driving people nuts. In 1990s, Mexico and East Asia experienced large capital inflows, and the US and Ireland had a mortgage boom in the 2000s. Those tendencies lead businesses and people to over-leverage, and banks to lower provisioning (after all, delinquency rates are low during a boom, which explains why credit expansions are predictors of crises).

And there is a herd effect that causes all banks to participate, for managers copy peer-behavior. As Citi’s Chunk Price said the Financial Times in July 2007: “As long as the music is playing, you’ve got to get up and dance.”

Overoptimism waves are too rare in any given country for them to be learned, creating a “disaster myopia”. Sharp falls in property prices reveal the non-recoverability of mortgage loans; currency depreciation creates insolvency among unhedged borrowers; assets sales by distressed investors seeking liquidity puts further pressure on prices.

The East Asian systemic banking crisis of 1997-98 was caused by a currency collapse. And in Chile, 1982, the sudden withdrawal of foreign funds exacerbated an ongoing crisis. What happens in those cases is: expectations result in deposit withdrawals, creating liquidity problems, thus exposing the problems of excessive risk-taking.

The international financial crisis of 2007-2008 seems new in that it involved financial derivatives. However, it exhibits overoptimism leading to extreme leverage, unsound management and weak regulatory responses.

Panic and Contagion: Explaining Sudden and Fast-moving Banking Crises

What is special about banking systems that makes them prone to dramatic collapses?

  1.    Modern banks are highly leveraged:
  Most policy efforts focus on this: capital adequacy regulation. 
  
  2.    The degree of maturity transformation with which they are associated:
  Liquidity problems arise when depositors want to withdraw more than banks expected (that money is a loan to other agents now). Banking panics are possible, because this situation can lead other depositors to withdraw for no reason, other than the possibility of their bank failing. Depositor runs originate mostly in the wholesale market (better informed). Liquidity runs lead to insolvency by forcing the sale of assets at unfavorable prices. This is difficult to distinguish from excessive risk-taking.
  
  3.    The very short-term nature of most of their liabilities:
  First come, first served: those depositors arriving first receive their full deposit and the rest have to bear the capital deficiency. This issue has a positive effect on the behavior of astute depositors: they monitor the performance of bank managers, alert to signs of trouble in order to be firts to withdraw.

  4.    The opaque nature of bank assets:
  Banks are at an informational disadvantage vis-a-vis borrowers, but depositers are also at disadvantage in relation to banks.
  
  5.    The fact that most of their assets and liabilities are denominated in fiat currency.

Some banking crises arise as a consequence of external shocks, but then their failure amplify it. Models of contagion focus on different aspects. One is the reassessment of depositors due to one bank’s failure: one bank’s loss of liquidity turns into a deposit run elsewhere.

Information and fear are transmitted through the same channels: CDS, bank equity prices, rating announcements. Pessimistic opinions are also a channel of contagion: the depth of the 2007-09 credit crunch reflects, indeed, the correlated realization by leading bankers that the risk management paradigm that they all shared had failed.

In models of feedback, multiple equilibria are possible: a good equilibrium has investor’s confidence validated by high asset prices boosting the creditworthiness of borrowers with productive and profitable investment projects; the opposite is true of a bad equlibrium. The occurrence of a crisis can be considered as a coordination failure.

Efforts to prohibit banks from taking risks can backfire. One remarkable reason is the option to go into “non-banking” financial firms.

Costs of crises

Total fiscal costs of crises in developing countries alone since the 1970s exceeds US$1 trillion — a sum far in excess of all development aid provided by the advanced economies.

Two approaches to estimate the cost of banking crises: fiscal (and quasi-fiscal) costs incurred to indemnify depositors; system-wide economic costs of the failure. No estimation has been similar to others, but cross-country estimates come up to similar figures.

The fiscal cost of systemic banking crises, estimated with data on the 39 events that occurred in 1975-2000 is 12.5% (Argentina’s 1982 crisis had costs of 55% of GDP). Because the economic costs (system-wide) incurred include central bank purchase of junk assets, it was too soon to estimate the economic cost of the 2007 crisis as of 2013, though the estimate for the 1975-2000 period was 14.6%.

Neither approach is wholly satisfactory, because the assumptions made to calculate them aren’t constant. For example, in Norway and Sweden, the federal government bought properties to bail out people, but when the price of those properties rose, the cost of the bail out reduced considerably. This approach is good inasmuch as it fills the gap created by loss-making activities; however, it ignores the losses and missed opportunities that are not captured in the fiscal costs.

On the other hand, the economic costs are hard to estimate, as these are sensitive to the counterfactual growth path against which the actual is compared (some crises weren’t followed by an economic downturn, but as a limiting factor on positive growth).

Crisis Response and Prevention

An Ounce of Prevention

“Prevention would be easier if the onset of crises could be predicted, but models are better at showing fragility than predicting timing”. (Demirgüç-Kunt and Detragiache, 2005).

Basel Committee on Bank Supervision: standard setter for bank regulation and supervision. 3 pillars: Capital, Supervision, Disclosure.

  • A minimum amount of capital to face the risks of banking;
  • A supervisory regime to ensure compliance with minimum capital and to discourage excessive risk-taking;
  • Disclosure of relevant accounting information.

However, Basel’s approach has an intrinsic flaw: herding. It isn’t sound to have banks modeling risks in the same way. Indeed one study by Barth, Caprio, and Levine (2006) compiles indexes on the three Basel pillars (capital, supervision, disclosure), relating them to the development, efficiency, vulnerability, integrity (lack of corruption), and governance of the banking system, controlling for the determinants of these variables and endogeneity. In the study, it is suggested that an approach to regulation that tries to work with market forces would work better

There was a key finding regarding vulnerability: none of the three pillars explained the probability of a banking crisis. Instead, authorities should encourage banks to diversify both their activities and their geographic and sectoral exposure. Lack of such diversification helps explain the large number of failures in the US (roughly 15,000 bank failures in the period 1920–1933), compared with Canada (just 1 in the period).

A Pound of Cure

Walter Bagehot’s advice on LOLRs: lend freely, but at a penalty, only to solvent institutions (collateralized), acting before the run takes off. Also, only use the LOLR rarely, to avoid moral-hazard.

So, it is important that intervention be comprehensive. For example, in the 1997 bank restructuring package in Indonesia, failure can be attributed to it being less than comprehensive. Depositors ran from private banks to public banks; then, the central bank extended the package to include private banks, but it was too late, as they had already purchased foreign assets, exacerbating the decline of the currency.

Bagehot’s advice works almost always, with a few exceptions: Guinea, where six of the seven banks that accounted for 98% of banking assets went bankrupt (the other bank failed a few years later); Iceland, where the three national banks failed in 2008 and LOLR couldn’t work as their liabilities denominated in foreign exchange were 10X GDP.

The Crisis of 2007 and the Euro Crisis: Some Things Old and New

The crisis of 2007

“Securitization. Source: https://medium.com/corda/tokenized-securitization-62aceeab1621”

“Originate and distribute” model: knowing that the loans they originated would be sold to others reduced the incentive to make careful credit assessment. Thanks to securitization, US banks had retained only part of the mortgage risk, passing much of it to European and other banks and investment funds.

At each stage of the securitization process, large bonuses were taken from the fees, even by the ratings agencies. Thus, in banks such as Lehman Brothers and UBS, audits found that bonuses were rewarding return, with conscious manipulation or evasion by some agents of risk management mechanisms.

In Ireland, Iceland, and the UK, the crisis was homegrown, and not due to securitization. However, both crises share these features: the boom in leverage, the availability of loans with no requirements for information, soaring compensation in the sector, and the financial alchemy of manufacturing safe securities from doubtful loans.

Distinguishing elements of this crisis: speed of spread; “too big to save” banks in Ireland, Iceland and Cyprus (banks that exceeded their country’s GDP).

The Eurocrisis of 2010

This crisis was the result of interrelated banking and fiscal and competitiveness weaknesses in many euro area countries. The key feature is excessive loans to governments, and that’s an old issue of banking (as old as banking).

Though the crisis originated in Italy, Spain, Greece, Portugal and Ireland, it threatened the sustainability of institutions in Europe, as even stronger countries experienced banking losses.

Crisis management evolved slowly in the early days of the 2007 crisis: first in a case-by-case basis, until the collapse of Lehman Brothers; then there was government intervention (Citigroup in the US and Royal Bank of Scotland in the UK) and coordinated international action. But that shift came too late, as risk aversion was huge by then. The “credit crunch” had begun to depress economic activity worldwide.

The fallout:

  • The case of Cyprus, in 2013, shows there was no fixed template for dealing with banking crises after 6 years of crisis;
  • Slow recovery probably stemming from the high leveraging of the boom years;
  • Also, slow recovery due to the concurrence of a credit crisis and housing prices collapse.

Toward Dynamic Regulation

  1. low interest rates and bad projects;
  2. financial safety nets (boom in deposit insurance);
  3. current account imbalances eroded by slowdown in surplus party or lower creditworthiness in deficit party.