Sunday, 1 March 2009

Andrew Lo's Matrix

Andrew Lo is the director of MIT's Laboratory for Financial Engineering and he made a quite remarkable suggestion recently during an interview on a proposed tool which would provide for improved regulation of the capital markets.
He is quoted as follows in regard to a new global monitoring facility to track the cobweb of financial nodes in the global system

My science-fiction vision for this is a high-tech war room with hundreds of screens where the one thing you'll see is this map of linkages between every single financial institution. You'll see lots of nodes and arrows of different thicknesses that measure how important certain connections are. That's the one tool we need to build for regulators. The second set of tools will be analytics that pop up for each particular participant in the financial industry [showing] what kind of assets they're holding, how much leverage are they using, and what the likelihood is that they will fail over a three-, four-, or five-year horizon.

There are two major problems with this technocratic solution as I see it.
Firstly such a highly transparent monitoring of the network of nodes that underpin global inter-market linkages moves in completely the opposite direction to the way that the actual market players would want to do business. Macro asset-allocators and hedge funds have already shown a predilection for "dark pools of liquidity" and even more obscurantist algorithmic trading capabilities to conceal their trading footprints.
The second problem is even more severe and reveals the real limitations of the overly quantitative mindset that underlies the academic approach to risk management technology.
This mindset believes that regulators just have to get better at precise quantification and that by capturing and calibrating the complexity of networks it should be possible to craft more robust safety nets. This is what I call the econophysicist fallacy of seeing market metrics as clearly locatable points in a conceptual phase space.
To be more down to earth I would call this the fallacy of modeling capital markets in a similar fashion to modeling an electric power grid or telecommunications network. Admittedly markets can be construed as networks and are subject to malfunctions. But the reasons why power stations go offline and need repair are of a categorially different order to the reasons why asset managers can decide to see the glass as half full in one second only to see it as half empty the next. That is the enigma of the markets and that is what would remain an enigma even with Andrew Lo's high tech matrix scenario.

No comments:

Post a Comment