A fourth issue is that most existing studies use relatively simple empirical techniques. for example, Reinhart and Rogoff (2009a) focus on the peak-to-trough fall in annual real GDP per capita around the start of crises. Similarly, other studies, such as Bordo et al. (2001), compare the severity of recessions accompanied by financial crisis with those not experiencing financial difficulties. while simple summary statistics are often helpful and illuminating, they may sometimes lead to questionable conclusions.Looking at the peak-to-trough decline in output around a financial crisis may understate the impact of a crisis in a country with very strong trend growth.