Sunday, April 24, 2011

Algorithmic trading gone wrong

Much research effort, by academics and practitioners alike, has been devoted to algorithmic trading in recent years, and at least one person I know blames the 2008 crisis for this surge in emphasis. Talking about blame, it appears almost certain that algorithmic trading played an important role in the flash crash of 2010.

But this is by far the most amusing example of algorithmic pricing gone astray: the ask price of a single book reaching millions and millions of dollars on Amazon before anyone noticed.

Thursday, April 7, 2011

Princeton visit

I have just ended a 5-day visit to Princeton as part of a bit of road trip through the American Northeast. Even during such a short stay, I could get a feeling for the intensity of the place: apart from my own seminar talk on Minsky and bubbles at ORFE, I attended two good quality talks in the graduate student workshop  at the Bendheim center, followed by Jim Ma's colloquium on BSDEs, then Carl Graham's seminar talk on opinion dynamics and finally a very stimulating talk by Andrew Lo back at Bendheim on the origins of behavior !

If you feel overwhelmed by all the links above, try to imagine being there...

Tuesday, April 5, 2011

Quantitative Finance Seminars - March edition

As usual on the last Wednesday of the month, we had two talks last week as part of the Quantitative Finance seminar series at Fields. The first talk was by Rafael Mendonza-Arriaga, who spoke about hybrid credit-equity models using time-changed Levy processes, a fruitful topic that attracted a lot of attention in one of the industrial academic forums that we had during the thematic program last year.

The second talk was by Alfred Lehar, who spoke about a general way to allocate capital requirements for systemic risk. His key message is that capital requirements themselves change the risk profile of a bank and its contribution to the overall risk in the system, so that whichever way one uses to measure systemic risk, the final allocation must be a fixed point of an iterative scheme.

Alfred then visited McMaster the next day where he gave a talk at the De Groote School of Business on the uses of market information for bank regulation.