Subscribe to DSC Newsletter

What am I doing with my blogging? A Theory Forge for Quantitative Risk Management!

Is it a vanity project? Am I writing a story with a beginning a middle and an end? In a way Yes! I really started blogging here on Analytic Bridge, Vincent Granville encouraged me, back in the day when Analytic Bridge had few members. I thought Analytic Bridge was an optimal place to blog references to banking supervision papers which had a quantitative orientation generally and I think I was correct in that. Now I have to blog corporately, I have to support the company in which I am a partner. My corporate blogs are here;-

But I still believe that Analytic Bridge is the optimal locus to get a message with an essentially quantitative analytic subject heard.

Here in this post I want to set out the point to whch my blogging has come, what my current objective is;-

Back in the day of the Monetarist v Keynesian debate, which in the US really fired when Reagan was elected, the Keynesians (e.g. Cowles Commission) could always attack monetarists by saying that monetarists had no theory. The general monetarist response (Chicago School and Berkeley) was that the Keynesians did not understand science and particularly did not understand 'Logical Positivism'.

The real champions of monetarist Logical Positivism were though in the UK (where Thatcher had put the monetarist ideas into practice before Reagan); these two champions were Friedrich August von Hayek and Sir Patrick Minford, even though Friedman and Anna Schwarz had invented monetarism at Chicago and it was alive and well at Harvard and Yale in the presences of Barro and Lucas.

This debate is now alive and well again with the election of President Obama. The new administration is essentially Keynesian as you know but a key aspect of the economics debate is that Keynesian modelling methodologies are structural whereas monetarist are reduced form; the former are focussed upon specifying how real values like unemployment or capital investment respond to government spending whereas the latter simply want to accurately estimate the manner in which the price level is impacted by money supply and related financial dynamics.

For risk management, modelling will remain essentially monetarist, essentially statistical, essentially positivist, all about the Financial Sector, which most Keynsian models do not even consider, the best structural models have modules analysing every economic relationship you could think of but they do not model the Banking Sector. The theoretical reasoning is lost in smoke filled rooms just after WW2 but its something to do with the invisibility of equilibrum (sounds like Harry Potter or more closely Terry Pratchett or Philip Pullmann, right?)

Therefore the best way to support modelling in risk management is to create a theory-forge which is prior to the code. That is give the positivism a theoretical back-bone, or rather create a locus for that theoretical reference manual to live in. That theory forge would consist of the jewels in the crown of research papers from central banks and universities which are not waffle Keynesian theory but are applied risk management 'technology' (used in the American sense of the term). By identifying the best and most important of those positivist papers (to classify them as such a key criterion is they have actually used S+, Matlab or R to evidence their conclusion). By doing this one can point the world risk management community at the best ideas, the correct deployments of VAR, Copula, PCA, Dickey-Fuller, whatever.

I have already started it in the references page on my corporate site, two examples below;-

As anyone reading this blog will realise I update my personal theory forge here on Analytic Bridge, I am posting in the low teens per month on average. It's all a bit haphazard right now (its not an easy process per se) and more to the point, funding for the academics is pouring out of the central banks in Europe right now and new (but good, applied) papers are appearing almost daily. I see my job as to blog them when I see them but its hard to keep up and I do not think I am omniscient. I intend now to consolidate a top-twenty at month end every month. I also have a terrible selection problem in that I can never decide what to post on Analytic Bridge and what is appropriate for Asymptotix, the older I get (measured in hours) I see the only relevant material as quantitative, all the rest is pontification!

But inevitably what you could call my back catalogue gets dropped off in this process, I notice that the discerning readers of Analytic Bridge find my old pre-credit crunch White Paper (published by SAP) relatively more interesting than alot of my new stuff, well they are right, SAP made sure through their rigorous QA process that paper was thoroughly referenced and checked. I have to thank here the two guys who supported the development of that paper, Nick Illingworth of SAP UK and Professor Alex McNeil. Good people of Analytic Bridge - Go on reading it!

Well that's my explanation of what I am doing and what I am working towards, I hope any reader finds it interesting, I think it has been important for me to state the real purpose of my activity.

Views: 453

Tags: Analytic Bridge, Quantitative Risk Management, REvolution Computing, asymptotix


You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

On Data Science Central

© 2020 is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service