Dating financial bubbles via online multiple testing procedures

The paper addresses the problem of real-time detection of financial bubbles which, after the latest global crises, have become an issue of economic concern. It shows how some online procedures, called LORD, implemented using an appropriate p-value calibration, represent new potential tools for economists to monitor and respond to bubbles in real time. This has substantial implications for financial stability and crisis prevention. LORD algorithms are applied to some well recognized global stock indexes to test their reliability across a variety of financial scenarios. A comparison with standard offline approaches confirms their effectiveness.