Pierre Aury on all the new tech and systems attempting to make shipping more efficient.
The Geneva Dry 2024 conference was the occasion to hear from a number of panels about how new technologies and systems – all cloud-based, and all, of course, harnessing the endless power of AI and/or big data – could help shipping become more efficient.
And to be honest there is indeed a good case that some of these systems are certainly adding value in the at trade and post-trade parts of the trade cycle. Some of these systems can speed up and render error free some tedious and human error prone tasks.
Where the case is less convincing is for the pre-trade part of the trade cycle. One must forever be a student of one’s market. This is a good way to continuously improve the quality of trading decisions. With today’s slick, intuitive and powerful tools there is no excuse for not being such a student. But these tools should come with health warnings. These cloud-based tools are highly addictive. They tend to create a false sense of comfort being so slick and, for some of them, so intuitive that it is easy to spend endless hours indulging in data torture. The problem being, of course, as said by the 1991 Nobel prize winner in economy, Ronald H. Coase: “If you torture the data long enough, it will confess to anything.”
Coase received his prize for a theorem since called the Coase theorem. By the way, in order to be applicable, the Coase theorem needs the market to be efficient and competitive with transactions done at zero costs which is never the case in real life. This is reminiscent of another Nobel prize in economy, the one awarded in 1997 to Robert Merton and Myron Scholes for the development of a model to value options. The so-called Black Scholes (Fischer Black assisted both Merton and Scholes in their work) has the same problem: its underlying assumptions are never met.
These new tools can create a false sense of comfort leading to blind trust
Back to torturing data. Coase has a point and forgetting it can be very costly. A number of years back some traders started to arbitrage product freight routes TC2 versus TC4 whenever they deviated away from their normal good correlation only to lose millions the day they did not go back to normal. The same mistake was made in the coal freight market arbitraging synthetic freight (API2-API4) versus real freight (C4) with again millions being lost.
These new tools can create a false sense of comfort leading to blind trust but they can lead as well to analysis paralysis: the trading case is never robust enough leading to endless data torturing with no action in the end.
A final point needs to be mentioned and it is related to job protection. There is less risk booking blindly the trade recommended by one or a combination of these new tools as if it goes wrong the trader can always blame the computer whereas questioning the finds of the system and putting on a different trade in the book will see the responsibility of the trader being invoked if the trade ends up being a bad one. In order not to put his or her job at risk the trader becomes like a banker, i.e. not somebody making money in a unique way but somebody losing money the same way as the others.
Finally a word on at trade and post-trade systems which are being deployed more and more in shipping. They are now so enmeshed in customer’s organisations that it creates a huge exit barrier for the users. Maybe it is the price to pay to become more efficient.
