DeMarzo, P. M., Vayanos, D. & Zwiebel, J. (2001). Social networks and financial markets. Working paper, Massachusetts Institute of Technology and Stanford University.
Not read. No copy of the working paper was found on 2026-10-29 (see the status note).
Why this is here
The triage of 2026-10-29 put it in the File tier for CLAIM-297 and QUESTION-117. The authors’ persuasion-bias paper (LIT-1134, NOTE-1392) holds listening weights fixed and shows that position alone can give an agent influence out of proportion to its accuracy. Its conclusion points to this companion, in which the weights move: whom an agent listens to is chosen by what listening earns. As the 2003 paper describes it, that is CLAIM-297’s loop, authority updated by performance, ending in influence that perpetuates itself whatever its accuracy, and it would be a predecessor QUESTION-117 has to meet: an established model with evolving influence that already gives capture without fidelity.
Whether it states the loop is not settled by that description. In CLAIM-297 an operator’s authority is scored against the interpretation it helped make dominant. In the companion, as described, listening is scored by trading profits, which depend on prices and on what the asset pays, so the score need not be self-referential. Which it is, and whether the stable structures are reached from a small initial advantage, can only be read from the paper.