This paper presents mathematical models for cyber breach probability as function of security spending in protecting a firm's ICT systems. We derive optimal level of security investment as percentage of value-at-risk. We show that the upper bound of optimal investment can be 1/e, 1/√2π or...

Appointed actuaries are responsible for estimating the largest liability on property-casualty insurance companies' balance sheet. Actuarial independence is crucial in safeguarding accurate estimates, where this independence is self-regulated by actuarial professional institutions. However,...

We investigate the risk choices of risk averse CEOs. Following recent theoretical work, we expect CEO risk aversion to be more pronounced in firms with high leverage, or high default probability. We find that the CEOs of these firms reduce firm risk, even in the presence of strong risk taking...

Using daily abnormal currency returns for the universe of countries with flexible exchange rates, we document local currency depreciations ahead of unscheduled, public, sovereign debt downgrade announcements. Consistent with the private information hypothesis, the effect is stronger in lower...

We test for limited attention bias in institutional investor trading. We use the universe of transaction-level data of institutional investors trading in the U.S. corporate bond market. Results show that trading volume abnormally increases in subsamples of uninformative rating actions. We also...

We develop the regime-switching default risk (RSDR) model as a generalization of Merton's default risk (MDR) model. The RSDR model supports an expanded range of asset probability density functions. First, we show using simulation that the RSDR model incorporates sudden changes in asset values...

Rating agencies consult with local government officials several days prior to official announcements of sovereign debt rating changes, making information leakage likely. Using cross-country data from 1988 to 2012, we find evidence of information leakage. In particular, we find statistically and...