Empirical Examinations of Multiple Crises
Economists today have the benefit of being able to analyze multiple crises over a longer time period. These conclusions have helped to shaped some of the models discussed above, and to provide indications for economic and financial policy.
Babecky et al. (2014) examine crisis episodes over 40 countries between 1970 and 2010, finding that banking and debt crises often precede currency crises, but not vice versa. The authors also find that banking crises are more costly than other types of crises, and take on average six years to recover from. The long-1 asting effects of banking crises have been confirmed by various studies, including Reinhart and Rogoff (2014) and da Rocha and Solomou (2015). Furthermore, countries that have experienced one banking crisis have a higher likelihood of having another banking crisis in the future (Aizenmann and Noy 2013). Qin and Luo
(2014) examine data from G-20 countries between 1989 and 2010, showing that capital controls play an important role in predicting banking crises. Reducing capital account openness somewhat in high-i ncome countries, they find, can reduce the incidence of banking crises; while increasing capital account openness in low-income countries can have the same effect.
Examining financial crises in Organisation for Economic Co-operation and Development (OECD) countries between 1960 and 2008, Furceri and Mourougane (2012) find that financial crises lower potential output by around 1.5-2.4 percent on average, due to a decline in capital. Cohen and Villemot (2015) use data from 97 countries over the period 1970 to 2004 to show that, while exogenous shocks account for most debt crises, self-fulfilling crises occur in 6 to 12 percent of cases, as financial markets panic, leading the debt-to-GDP ratio to stray from its pre-crisis path.
Despite the going wisdom from the most recent crisis, Bretschger et al. (2012) find that banking concentration (“too big to fail”) may lead to financial stability or to financial fragility; there is no evidence that it is one or the other in all cases. Using data from 160 countries between 1970 and 2010, the authors show that market concentration, then, has no direct effect on systemic crises. Still, banking indicators matter. Laina et al.
(2015) find that indicators that best predict systemic banking crises in 11 European Union (EU) countries between 1980 and 2013 include growth rates of loans-to-deposits and house prices. Market power has also been shown to exacerbate a downturn in industries that are more dependent on external finance, based on data from 36 banking crises in 30 countries over the 1980 to 2000 time period (Fernandez et al. 2013). By contrast, interconnectedness to global markets has been shown to reduce the incidence of banking crises (Caballero 2015). This may be because more globally interconnected banking institutions have access to liquidity in the face of negative shocks.
Currency crises became more internationally synchronized over the twentieth century, while banking crises became first more international in the early twentieth century, then less so with the rise of banking regulation after the Great Depression. After the rise of neoliberalism in the 1980s, banking regulation declined, resulting in a rise in banking crisis incidence through the Great Recession (Dungey et al. 2015). Twin crises, or concurrent banking and currency crises, arose. Twin crises have led customers in the crisis-affected country to bank abroad in unaffected countries (Kleimeier et al. 2013).
The impacts of financial crises on exchange rate regimes and vice versa have been studied for many years. Some have argued that the optimal exchange rate regime is a floating rate, while others argue that a pegged exchange rate provides the most credibility. Esaka (2014) and Domac and Martinez Peria (2003) find that consistent pegs, or de facto and de jure pegged regimes, are significantly less likely to experience a currency crisis than countries with other exchange rate policies. The reason for this may be that countries with consistent pegs increase the credibility of their currency regime by clearly committing to a pegged regime. Angkinand and
Willett (2011) show that countries with a soft exchange rate are the most vulnerable to currency crises.
Financial crises result in the deterioration of labor rights practices, due to increases in unemployment as firms cut costs. Blanton et al. (2015) use data from 46 countries between 1985 and 2002 to show that while crises do not impact labor rights laws, they adversely affect labor rights practices, lasting into the early years after the crisis.
The recent global crisis has brought the study of financial crises more into vogue, and new crisis models reflect the latest understanding of how crises arise and are transmitted. Pre-c risis monitoring, particularly of banking systems, to intercept building crises has become an important component of the global financial architecture. Monitoring systems may include some of the indicators found to be relevant in the academic literature. In what follows, we discuss policies to prevent future crises, based on our present knowledge of crisis prevention and containment.