Traditional autocorrelation and variance ratio tests are based on serial uncorrelatedness rather than martingale difference. As such, they do not capture potential nonlinearity-in-mean, which could lead to misleading inferences in favor of the martingale hypothesis. This paper employs various...

Using credit default swap data, we propose a novel empirical framework to identify the structure of credit risk networks across international major financial institutions around the recent global credit crisis. Specifically, we identify three groups of players including prime senders, exchange...

Using daily stock and bond futures data of three developed markets (the U.S., the UK and Germany), this study explores time-varying extreme correlation of stock-bond futures markets. There is evidence of positive extreme correlation between stock and bond futures markets in the U.S. and the UK...