Doomers, I have a really bad feeling arising out of this recent ZeroHedge comment.
.
.
.
.
.
… But there is a potentially nasty side-effect to this strategy: it seems to make the market essentially isomorphic among all asset classes. At this juncture, it does not seem too big of a stretch to assume that there are algorithms A1, …, An trading against each other. Each one has a (probabilistic) trading strategy, say S1, …, Sn. Eventually, they should end up in equilibrium, according to elementary game theory (Nash eq.). … by induction we’ll end up with only 1 algorithm, being the MM of everything, everywhere (even coupling Europe/Asia in the game, just look at the recent correlations between DAX and DOW during overlapping trading hours; it’s uncanny!) …
There are a lot of assumptions. Such as, retail investors will not invest in the market. and that all these models across asset class has the same input.
surely we can merge all the input and have a meta model but the construction of model this scale is too expensive and time consuming to be realistic.
Fear not. You can sleep soundly, John. No business actually uses game theory.
[W]e assembled a panel of 30 respected game theorists around the world, and we sent them a survey asking, “Can you think of any examples of real, live companies that have consciously applied game-theoretical concepts to a real business problem?”
The response was . . . a deafening chorus of head scratching.
“The short answer is, I don’t know,” said David Levine of UCLA. “Let me think about this,” replied MIT’s Muhamet Yildiz.
Others on our expert panel, while not offering up any actual, you know, examples, were willing to speculate on why they couldn’t. Traditional game theory “prescribes a lot of advice that does not actually seem to work,” admitted Paul Bartha of the University of British Columbia.
http://www.fastcompany.com/magazine/91/debunk.html
malthus,
I’m old enough so that for me Helen Bannerman was still a “children’s author.” The concept “speed kills” brings up a rather funky image.
It seems to me that the program traders are acting so fast that they have entered a different domain of reality, sort of like where relativity effects become important in physics. I’m willing to entertain the commenter’s implicit assumption that the algorithms the program traders are running may actually have substituted things like the Nash equations for conventional thought and common sense.