http://www.elitetrader.com/vB/showthread.php?t=188697

Dear colleague,

Thank you very much!

I have three points.

1) What exactly do you mean by "systematic adverse selection"? In what follows I just assume that this term relates to adverse selection occurring in a repeated and systematic way. If that is not correct, my points below may be wrong.

2) I don't think that systematic adverse selection cannot occur with rational agents. Adverse selction is a consequence of informational asymmetries, and in fact most theoretical models of adverse selction do assume that agents are fully rational.

3) A good example from financial markets is the presence of informed traders in equity (and other asset) markets. These traders impose adverse selection costs on market makers. These market makers, in turn, are fully rational and price the losses to informed traders into their bid and ask prices. This gives rise to the adverse selection component of the bid-ask spread. The classical reference here is of course Glosten and Milgrom (JFE 1985). Alternatively, you may look at the corresponding chapter in Maureen O'Hara's maket microstructure theory textbook.

Dear colleague, 

Thank you very much for the answer. 

Indeed, the meaning is right, but, indeed, i did not found many references for the subject defined in this way.

I agree that rational agents should not make systematic adverse selection. However, they can be affected by other agents decisions. For instance, let's assume 80% from agents are taking systematically a wrong decision. Which will be these effects if the majority is not making their decisions rationally?