JOST A MON

There are several benefits to attending a conference. One overhears some interesting gossip. One supposedly learns a thing or two about one's field. One may, if one is overwhelmed by the awesome weather, sleep through a particularly boring presentation with no qualms. And one may also have the opportunity to inveigh against those who have the temerity to not attend the conference after their paper has been accepted.

As you can perceive, I had the opportunity to participate in each of these delicious activities. [As mentioned earlier in the week, I was in Aix-en-Provence, attending the Forecasting Financial Markets 2008 conference.] Chief among them was the gossip. Academics are like old people in this regard. They trade titillation and amusement and raised eyebrows. Unfortunately, I missed out on all these. At my table during lunch, there were, instead, hagiographies of the worthies of the world of financial econometrics. "What a fellow," said one professor admiringly of Mark Taylor.

Mark Taylor, of course, is well-known in our field. He read for a Politics, Philosophy and Economics degree as an undergraduate at Oxford (you can't study any of these in isolation, it appears) and worked in the financial industry for a bit. He then got himself a PhD from Birkbeck College and obtained a faculty position at University College, Oxford. Here, because his doctorate was not Oxon. or Cantab., he was addressed by his colleagues as Mister. The snobbery extended to virtually all his dealings with them, despite his steady progress to the heights of achievement in his field. Eventually, he left them and joined the University of Warwick, where he remains to this day.

Of course, he missed the perks of the high table and the elaborate system designed to support an academic at Oxford for conferences and things. Warwick, in comparison, is a bit of a backwater. But in the interim, he managed - while raising a family, publishing solid research and editing several journals - to obtain an MA in English Literature. And, tiring of paying expert repairers of antique clocks (the family hobby) through the nose, he apprenticed himself to one of the master clockwrights, and learned the trade.

Where the devil does he find the time?

Anyway. The FFM conference is bedevilled by one serious problem - that of presenters not showing up to present their papers after acceptance. For one thing, this completely screws up the scheduling. For another, many of them don't bother to pay the conference fee either, which results in the financial difficulties for the organisers. And finally it is incredibly frustrating for attendees who were looking forward to learning something new on a topic of interest.

I'm not sure what the deal is in the other disciplines, but this lack of attendance appears to be a serious canker among financial conferences.

The main reason for this sad lack of professionalism appears to be that it suffices for a paper to be accepted at a conference to appear on an academic's resume. We mulled over several solutions to this problem. One possibility is not to publish the accepted list of papers until the authors pay up. Another is not to do so unless they present (but why would they present unless they were told their paper was accepted?)

The FFM used to be held in London for years, and provided at one time a good mix of presentations from professionals in the finance industry and academics. These days, though, it is skewed more and more towards the ivory tower. People such as I turn up hoping to learn some new techniques in financial modelling, but some of the academic work is so removed from the day-to-day grind that it would take a person with considerably higher IQ than me to make it applicable to our grotty world of money-making.

I presented a piece of work done jointly with the head of my group. From a statistical standpoint, it was probably not completely rigorous, and I had hoped to obtain some feedback on how to strengthen the work (which I did get). For anyone who is interested, the work dealt with the extraction of currency positioning information by the use of peer-group benchmarks. For currency managers, this is interesting, because it's known that when positions get stretched (i.e., say, there's considerable selling of the US dollar and lots of buying of the Japanese yen), the positions tend to unwind rapidly, and there's usually short-term turmoil in the markets. This happens fairly regularly - so can one forecast it? A peer-group benchmark is an index that tracks the daily aggregated performance of a bunch of currency managers, and there are several published by various sources. If we, by means of our simple approach, can determine that our competitors in other banks are positioned in one particular way, and we have an indication thereby that those positions are stretched, we could take the opposite position, for example, and make a profit when the market snaps back.

Somewhat surprisingly, the simplicity of the approach won us some favourable comment, and the one or two enthusiastic questioners were inclined to suggest more rigorous statistical methods to make our results robust. They also mooted a possible collaboration, and this is something we'll have to take seriously, although I did point out that our market priorities change so often that it was unclear how much time or effort we could spend on further developing the work.

Aix is a lovely little town, and the local council provides some amount of funding for the conference. But the FFM has been held here for four years in a row now, and the organisers are thinking of alternative venues.

Luxembourg, anyone?