« links for 2011-05-25 | Main | "I Sorta Told You So" »

Thursday, May 26, 2011

"The Eternal Failures of Mechanical Forecasting"

I haven't posted anything on methodology lately, so here's part of a much longer review of Frydman and Goldberg's Beyond Mechanical Markets to -- hopefully anyway -- keep the conversation alive:

Frydman and Goldberg's Beyond Mechanical Markets, by Robert Teitelman: How wrong were we? After several years of commentary on the causes of the financial crisis, we still struggle to plumb the full depths of the event. We have tossed up, like so much confetti, a variety of culprits, both human and systemic, many of which undoubtedly played some role and had some complicity: from executive compensation to lack of transparency to Alan Greenspan to Congress to credit default swaps. Journalists have been excoriated for missing what was apparently obvious; homeowners blamed; Wall Street pilloried; economists accused of intellectual dishonesty. We have chewed over questions of structure, size, capital, leverage, risk. We have put the Zeitgeist (blame the '60s!) in the chair, probed trade imbalances, decried the presence of greed and taken refuge in irrational impulses. And yet, there is a strong sense that we are just swirling pieces of a jigsaw puzzle across the table. There remains a feeling that perhaps we were wrong in some deeper way. ...
Beyond Mechanical Markets ... marshals a powerful argument that's bolstered by empirical reality: the eternal failures of mechanical forecasting; the sheer difficulty of beating the market with consistency; the unforeseeable ways that history unfolds. The belief in precise prediction resembles a kind of utopian project, a tower of economic Babel. At bottom, the pair makes a philosophical point that Knight, Keynes and Hayek (ironic, they comment, given that rational expectations came out of Chicago, where Hayek taught) offered many decades ago: the combination of men and events, particularly in these manmade constructs called markets, certainly improves our ability to price assets (and to forecast) over that of an individual or bureaucracy. But that inclusion of freely determined humanity ... conspires to erode any simple, mechanical or guaranteed relation between past and future. They quote Popper: "Quite apart from the fact that we do not know the future, the future is objectively not fixed. The future is open: objectively opened." At bottom, they're trying to thread the needle in the ancient free will versus determinism argument. ...
Will anything change? Not quickly. As they admit, the power of fully predetermined models may have actually increased because of the crisis. Economic pundits continue to speak with great certainty, and these issues are complex, nuanced and often hidden. Besides, the insurrection Frydman and Goldberg argue for is far greater than just an overthrow of rational expectations; it's an entire economic world view that claims the power to accurately predict, forecast and capture market reality. Generally, the classic response of an orthodoxy (or what Thomas Kuhn famously called a paradigm) is to ignore any threat, not only out of fear of what might be lost (tenure, prizes, careers), but out of incomprehension; to the predetermined model builders, Frydman and Goldberg's argument must literally seem like babble. ...

    Posted by on Thursday, May 26, 2011 at 12:15 AM in Economics, Methodology | Permalink  Comments (18)


    Feed You can follow this conversation by subscribing to the comment feed for this post.