Reputation Effects in Trading on the New York Stock Exchange by Andrew Ellul, Robert Jennings, Robert Battalio

Reputation matters. Once again we see that reputation and relationships matter. This papers presents evidence that trading costs on the NYSE are, in part, a function of the interpersonal relationships of floor traders.

SSRN-Reputation Effects in Trading on the New York Stock Exchange by Andrew Ellul, Robert Jennings, Robert Battalio: "reputation plays an important role in the liquidity provision process on the floor of the NYSE."

How cool of a study is this? Ellul, Jennings, and Battalio investigate trading costs on the NYSE following relocation of the specialist's post.

As the authors state, they have identified a "natural experiment". This occurs when specialists are moved but are not followed by the floor brokers who trade with the specialist. If relationships and reputation matter, then trading costs should increase (at least temporarily) following the move.

And the results? Sure enough, following the specialists' moves, trading costs increased. Interestingly, since the parties often knew of a pending move prior to the actual move, the importance of the relationship decreased before the actual move (consider game theory predictions as you get closer to the end of the game).

A few brief "look-ins":

*"we find evidence of statistically increased relative effective spreads for
relocating stocks versus their controls starting around event day -35 and ending after event day +45" p16.

* "With few exceptions, starting 45 days before the switch and continuing 35 days after the switch, the relocating stocks with high adverse selection have relative effective spreads that are higher than their controls"


In most cases the floor brokers did not follow the specialists to the new location. However, in some cases the brokers did move. Using regression analysis the authors find that "moving brokers enjoy significantly lower (statistically and economically) effective spreads than non-moving brokers. This advantage is particularly strong in stocks where the adverse selection problem is more severe." p. 26.

Which again suggests that repuation and relationships do impact trading costs. Why? The most logical explanation is that trust lowers the adverse selection cost of trading.


Very cool!


Cite:
Ellul, Andrew, Jennings, Robert H. and Battalio, Robert H., "Reputation Effects in Trading on the New York Stock Exchange" (March 2005). http://ssrn.com/abstract=684091