Stathis Papaioannou stathisp at gmail.com
Wed Jun 24 15:15:46 UTC 2009

2009/6/25 Dan <dan_ust at yahoo.com>

> I'm not sure what point you're trying to make here.  I believe EMH assumes markets in equilibrium.  Further, real markets are never in equilibrium.  If you agree with that -- viz., real markets are never in equilibrium -- then do you agree this creates problems for EMH?  Or do you believe EMH doesn't rely on an equilibrium assumption?

I don't understand what exactly you mean by "equilibrium" in this
context. A leaf carried by the wind is exactly where the vector sum of
forces acting on it puts it at any particular point in time. You might
say that its position and velocity is in a state of dynamic
equilibrium. Similarly, the market at any point in time is exactly
where the various factors influencing it say it should be. It hardly
ever reaches a static equilibrium.

> Regarding "regress[ing] to the mean," don't you think this requires an underlying assumption of equilibrium?  If markets are never in equilibrium -- whether they're coordinating in a Hayekian/Kirznerian sense, or discoordinating in a Schumpeterian/Schacklean/Lachmannian sense -- then what does it mean to "regress to the mean"?  (No pun intended.)  If you know the mean, then you'd be able to beat traders with regularity, no?  But that'd violate EMH, so why would anyone trade above or below the mean to begin with?  Wouldn't all or most traders find the mean and not violate it?  That they don't do so or that they disagree over just what the price of an asset should be (heck, why would they trade otherwise?*) seems a problem for EMH, no?

By "regressing to the mean" I mean that a trader who historically has
underperformed or outperformed the market will not continue to do so
in future (unless he has special connections, information and
influence, which may be the case for large players), in the same way
that a fair coin will tend to come up heads half the time in the long
run no matter what its history up to a point is.

Stathis Papaioannou

More information about the extropy-chat mailing list