This post is inspired by this post in Noahpinon's blog.
Paul Krugman:
True for macroeconomics and the DSGE models, but also true for quant finance?Don’t get me wrong: I like mathematical modeling. Mathematical modeling is a friend of mine. Math can be a powerful clarifying tool...But it’s really important to step away from the math and drop the jargon every once in a while, and not just as a public service. Trying to explain what you’re doing intuitively isn’t just for the proles; it’s an important way to check on yourself, to be sure that your story is at least halfway plausible...
Think about more complicated models that can explain, but also hide, more? About models, thats additional information get lost in the numerical jungle?
I recall the Financial Modeler's Manifesto. Models shall be understandable and computational, in short?
THE speculative swap again
5-Feb-14. Andreas gave the public lecture We valuate THE swap, for close to hundred people with different background, knowledge and interest from the citizen to the distinguished quant finance experts. Andreas managed perfectly to explain the complex concept to non experts whilst explaining where the complexity comes from to the experts.
What I did not mention yet: in the spirited panel discussion there were contributions of participants that were obviously involved (maybe, as advocate, …). The offense was: the models, Andreas used, are too simple.
But, without anticipating the results here, if a simpler model used for scenario analytics shows a danger that is not "balanced" by a chance, you should be alerted. A more complicated model may refine the values and risk spectra.
On the contrary, if a simulation with a complex model looks fine, use other, also simpler ones, to check.
Computational and understandable!
The next episode on THE swap will be posted by Andreas tomorrow.
Picture from sehfelder
