Dealers' incentives to reveal their names

• Published date: 01 February 2022
• Author: Arzé Karam
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/fire.12229

DOI: 10.1111/fire.12229

ORIGINAL ARTICLE

Dealers’ incentives to reveal their names

Arzé Karam

Department of Economics and Finance, Durham

University, Durham, UK

Correspondence

ArzéKaram, Department of Economics and

Finance,Durham University, Mill Hill Lane, DH1

3LB,Durham, UK.

Email:arze.karam@durham.ac.uk

Abstract

This research investigates dealers’ motivation to disclose their

names when quoting on the NASDAQ over the years. NASDAQ

enables dealers to quote limit orders either anonymously or with a

feature that reveals their names. Results are consistent with dealers

advertisingby revealing their identities so as to develop and maintain

their reputation for reliable pricing. Dealers strategically choose to

reveal their identities when order flow is profitable. Post-name dis-

closure analysis further suggests that named quotations are likely to

be driven byinformational considerations. This research contributes

to our understanding of the use of non-anonymity in electronic

trading.

KEYWORDS

advertising, dealers’ identity, name disclosure, reputation

JEL CLASSIFICATIONS

G10, G20, L10

1INTRODUCTION

Major dealer-based exchanges offer name disclosure on their limit order books (e.g.,the Toronto Stock Exchange, the

SETSmm system of London Stock Exchange, and the NASDAQ).The study of why market-makers1reveal their names

while quoting on the electronic limit order book interests both practitioners and academics. This paper examines the

use of named quotations by NASDAQmarket-makers, and in particular, investigates whether informed or uninformed

market-makerschoose to reveal their names on the quoting system.

For over 30 years, the NASDAQhas enabled registered market-makers to quote limit orders either anonymously

under the NSDQ feature (previously named SIZE) or with a feature that reveals their names. The identities of market-

makers can beprice relevant if they offer insight into the reasons why market-makers want to trade, and whether they

possess private information about the asset value2. In such instances, name disclosure by dealers can engage mar-

ket participants in a leader–follower type of behavior,resulting in an increasing price impact similar to the signaling

1Theterms “dealers” and “market-makers” are used interchangeably throughout the paper.

2Traditional theory in marketmicrostructure assumes that market-makers are uninformed and only react to potentially informed customer orders, as in

Glosten and Milgrom (1985) among others. More recent models relax this assumption and assume that dealers can be asymmetrically informed (Calcagno

& Lovo, 2006; Foucault,Moinas, & Theissen, 2007), and this theory is supported by many empirical papers (Davies, 2003; Madureira & Underwood, 2008)

amongothers.

Financial Review.2022;57:27–44. wileyonlinelibrary.com/journal/fire c

2020 The Eastern Finance Association 27

28 KARAM

game in Kyle (1985). Despite the signaling risk, market-makersmight be strategically revealing their names in order to

communicate with the market. Market-makers would then be interested in developingand maintaining a reputation

for accurate pricing so that their quotations lead the market, which can increase their payoffs in the long run. In the

industrial organization literature, Milgrom and Roberts (1986) introduce the notion that asymmetric information can

provide such a mechanism in highly competitive markets.This paper tests this rationale using a large sample of stocks

from two periods, namely 24 days from 2004 and 24 days from 2018. The breadth of the cross-section allows for gen-

eral conclusions to be drawn, and recent data are important because significant changes have been made in the U.S.

market structures overthe last 15 years.

This research focuses on three primary economic concerns related to the usage of named quotations by market-

makers for advertising purposes. First, it examines whether name disclosure has been frequently used by market-

makers. It documents that a nonnegligible proportion of quotes are frequently submitted by market-makers under

their names. Second, it investigates the informational content of named quotations relative to their anonymouscoun-

terparts. It shows that named quotations are more informative relative to their anonymous counterparts over the

years. Named quotations are associated with higher effective spreads than anonymousquotations, and they have also

significantly higher realized spreads after 5 min. Finally, it examineswhether named quotations represent a dealer’s

attempt to attract large tradersinterested in automated executions in today’s markets that trade at millisecond speed.

Overall,the results suggest that market-makers use their names strategically to attract profitable order flow, which

is consistent with the advertising hypothesis. Market-makers, at least those who havesurvived in the RegNMS envi-

ronment, seem to still generate higher revenues while revealing their names. They seem to be selectively revealing

themselves when the amount of protected liquidity at the National Bid and Best Offer (NBBO) prices, provided by

other National Market System (NMS) marketcenters, narrows. Named quotations may have gained so much popular-

ity that market-makers use their identity as a screening deviceso that large traders can call them directly, which also

allow them to find liquidity in size in today’s markets, which trade at millisecond speeds. It is, of course, possible that

named quotations havebeen used simultaneously by market-makers in an attempt to control their inventory risk. So, it

is possible that market-makerstactically use their identity to monitor their inventories and keep them from diverging

to arbitrarily large or short positions, as predicted by the theory (Grossman & Miller,1988). If this is the case, named

quotations might be associated with distorted pricing. However,the results show the opposite, that is, named quota-

tions are associated with price improvements. I use the Lo & MacKinlay (1989) variance ratiomeasure defined as the

ratio of long to short price variance to assess the effect of name disclosure on the price process. I find that named quo-

tations are associated with variance ratio closer to 1.0, suggesting that most of the named quotations are used as tools

for advertising purposes.

I test another motive behind the choice of dealers to reveal their names although it is one for which I have no sup-

port. Market-makers might announce their trading motives in advance, a practice called sunshine trading (Admati &

Pfleiderer,1991) to avoid exacerbating any adverse selection. In such instances, dealers would be recognized as unin-

formed because they reveal their lack of information via a sunshine trading announcement. If they are recognized as

such, there would be a decrease in adverse selection costs post-name disclosure. However,I find the opposite. I com-

pute the difference between the realized and effective spreads, known in the market microstructure as a viable mea-

sure of adverse selection costs faced by market-makers.The results suggest that there is an increase in adverse selec-

tion costs post-name disclosure, further corroborating that market-makeridentity conveys information.

This paper contributes to our understanding of the use of non-anonymity in securities trading. Although a num-

ber of studies examine the value of the identity of the intermediary with conflicting results, none has addressed the

use of name disclosure on a dealer-based limit order book. Unlike the fragmented setting I analyze here, Comerton-

Forde, Putniņš, and Tang (2011) examine anonymity in a centralized limit order book with designated market-makers

whose operation is completely different from that of NASDAQ market-makers.Other papers either compare trading

on separate anonymous and nonanonymous platforms (Barclay,Hendershott, & McCormick, 2003), (Reiss & Werner,

2005) or compare trading before and after a regulatory change regarding identity disclosure requirements (Foucault

et al., 2007). I contribute to this literature in severalways. By showing that name disclosure facilitates price discovery,