The public Limit Order Book of the Korea Exchange: Market capitalization dependent endogenous effects on spreads, volatility and volume

The public Limit Order Book of the Korea Exchange:
Market capitalization dependent endogenous effects on spreads,
volatility and volume*


This version: October 4, 2009

Abstract
In 1999 the Korea Exchange [KE] introduced its public limit order book displaying
volume, price and broker identity to all market participants. Going against the general
trend in market design, the KE event provides important information for exchanges
that are deciding between a transparent market, preferred by most participants, and an
anonymous market accommodating large institutions and their brokers. Using several
alternative metrics for market quality and estimation methodology accounting for fixed
firm effects and endogeneity in explanatory variables, our results contradict previous
research, which mostly supports the drive for opacity. We find that when limit orders
are public, spreads fall, volatility increases and volume increases. The effects differ
across market capitalization segments. The current policy of the Korean Stock
Exchange to publicly display the 10 best orders, including broker identity and trades is
provisionally best practice, as it promotes higher traded volume.

JEL Classification: G10, G15, G18

Keywords: Transparency, Broker ID, Market Quality


1

---

1. Introduction
This paper investigates if the introduction of a public limit order book has an impact on
market quality. Transparency is generally considered to be related to greater fairness,
more efficient information acquisition and better governance. When it comes to the
optimal design of a securities market the impact of transparency becomes more
complicated as important informed market participants may feel exposed in a fully
transparent environment. The only stock exchange in the world that promotes full pre-
trade transparency of the order book to the public is Korea Exchange [KE] (since
October 1999), while a number of other exchanges have altered their market in the
opposite direction.
The bid ask spread is commonly used to measure market quality hence any association
between spreads and transparency is of interest for this study. Foucault, Moinas and
Theissen (2007) propose a model in which an otherwise uninformed limit order
provider gains future private volatility information, but not directional information,
from an informed limit order provider in transparent markets and post wide spreads
when the provider anticipates greater volatility. In opaque markets it is not possible for
this follower to distinguish between informed and uninformed limit order providers
and thus anticipate adverse volatility. An implication is that informed bidders act more
competitively in an anonymous market where average quoted spreads are predicted to
decline significantly.
A peculiar feature of the Foucault et al (2007) model is that astute investors cannot
infer the direction of price movement by observing the order imbalance of informed
limit order participants. Rindi (2008) corrects this deficiency when she models the
effect of transparency on adverse selection costs. The model assumes that liquidity
providers are risk-adverse, all the market participants, both informed and uninformed
2

---

can simultaneously submit limit orders entering and leaving the market at their
convenience and that market participation is endogenous. Under full transparency
informed traders are more aggressive providers of liquidity in that they face less risk
than uninformed traders who can decipher information only by imperfectly observing
informed traders in the limit order book. Moreover, uninformed traders become “quasi-
informed”, incurring reduced adverse selection cost and thus are also ready to offer
liquidity to noise traders as they can better distinguish between endowment shocks and
information. Under opacity that identification is not possible, consequently
transparency increases liquidity. However, when information acquisition is
endogenous, transparency reduces the incentive to acquire costly information and so
reduces the number of informed traders. It follows that the previous results on the
beneficial effect of pre-trade transparency on liquidity can be reversed. Therefore, the
impact of transparency is dependent on what proportion of information acquisition is
endogenous and what proportion remains unaltered by transparency.
Boulatov and George (2008) also allow limit-order liquidity providers to possess
private information. However, their conclusions are quite different to those of Rindi
(2008) in that transparency serves as a coordination device that enables informed
traders and dealers to collude to make uninformed liquidity traders worse off under
transparency. Transparency enables dealers to better forecast trades based on private
information. Informed traders become less aggressive which reduces competition
between these traders. However, as the market becomes larger and more competitive
with the number of informed traders becoming very large the losses of liquidity traders
are minimized, stock price volatility is lower and spreads are narrower (Proposition 6).
Hence, an implication of this model is that for large, highly liquid stocks, the impact of
3

---

transparency on liquidity provision should be less harmful than for smaller stocks with
fewer informed participants.
Of the three models, Rindi (2008) is perhaps the most amenable to formalize expected
hypotheses for the impact of public broker IDs as hers represents a major extension of
Foucault et al. while nesting potentially beneficial and harmful impacts of
transparency, unlike either Foucault et al. or Boulatov and George (2008). Since each
of these models has different implications for the impact of disclosure depending on
size and liquidity, we investigate differential effects of transparency in separate
segments based on market capitalization. To the best of our knowledge, this paper is
the first empirical research focusing on the impact of pre-trade transparency on market
segments with different stock sizes.
For the KE, we find that when broker identities are made public, spreads increase in
small stocks and decrease in large stocks, leading to an improvement in liquidity in the
overall market due to the much greater economic significance of large stocks. It might
be harder and hence more costly to acquire quality information for small stocks, so that
it is more likely information acquisition is endogenous in these segments. It follows
that increased transparency may make market quality worse in smaller market
segments as incentives to buy or obtain costly endogenous information are reduced.
Conversely, it might be easier to get quality information for large stocks as information
might be gained at relative low cost in these segments, where traders may incur lower
or no cost for high quality information that is expressed as exogenous acquisition. In
contrast, traders may have to pay more for access to high quality information of small
stocks. This improvement in liquidity in large stocks is contrary to the model of
Boulatov and George (2008) who predict that in the best case scenario liquidity will
not worsen with the advent of transparency.
4

---

The realized spread measures how much buyer initiated traders exceed or seller
initiated orders fall short of an estimated post trade value, hence it measures the post
trade price reversal. The realized spread is typically used to measure the compensation
for liquidity providers. Uninformed traders may become “quasi-informed” in a more
transparent market, so they may become more willing to offer liquidity by placing
more aggressive limit orders so as to take the other side of noise trader demands. As a
result, competition in liquidity provision could become stronger in large but weaker in
small market segments.
We find that realized spreads fall for short-term horizons but rise for longer-term
horizons. This indicates that liquidity providers find it harder to earn profits in the
entire market under transparency. Informed traders seem to acquire information
exogenously and accommodate this to alter securities price, leading a rise in the
realized spread with longer-term horizons. However, the higher is competition, the less
profit liquidity providers can earn, therefore making it more difficult for informed
traders to benefit from their private information under transparency. In larger market
segments, competition could be higher due to an increase in liquidity provision as a
result of exogenous information acquisition in transparent markets. Our findings are
consistent with this proposition. Our proxy for compensation for liquidity providers
decreases gradually with increasing market capitalization.
While volatility could in principle either rise or fall with transparency in Rindi (2008),
she considers it more likely to rise. Madhavan (1996) concludes that in large
capitalization markets, transparency reduces volatility and improves market quality but
the opposite could occur in thin markets. Since volatility per se is essentially latent,
several metrics for volatility have been developed. Hence, it is important to measure
the appropriate form of volatility. We expect measures less affected by market
5

---

microstructure noise such as range-based volatility, e.g., see Alizadeh, Brandt and
Diebold (2002), not to be impacted by changes in pre-trade transparency. We expect
measures that include transitory volatility such as realized volatility, see Andersen and
Bollerslev (1997), to be significantly affected by changes in pre-trade transparency.
The Fishman and Hagerty (1995) model can be applied in this context to show that
broker identification anonymity results in increased intraday volatility and decreased
trade volume. In an anonymous market, like Korea before the reform studied here,
informed traders do not have to disclose their intentions pre-trade, hence price
discovery is delayed, causing higher volatility when information is impounded once
informed traders transact. The greater risk of trading with an informed trader makes
uninformed traders cautious, resulting in lower trading volume. The increase in all
measures of volatility is partly attributable to an increase in the number of trades in the
more transparent market for all size categories. Empirical support is found by Duong,
Kalev and Krishnamurti (2009) who find that traders posted more aggressive orders
during the previous regime with public broker IDs as opposed to the current opaque
policy in Australia. In this study of KE we find that all of metrics of volatility are
significantly and positively related to transparency, with a greater impact in the larger
shares by market capitalization.
While the volume of shares traded is not directly addressed in Rindi (2008), greater
trade aggressivity under transparency should result in higher trade volume. Studies
comparing turnover rates and volatility for international markets show that the world’s
largest markets with the highest turnover rates are also the most volatile markets, hence
volatility and volume would also here be expected to be positively correlated. In the
Rindi (2008) model, changes in transparency should impact the interaction of informed
traders with liquidity or noise traders. In a transparent regime informed trading should
6

---

be less profitable leading to a decrease in the number of traders that become informed.
Traders that are informed will be trading more aggressively timing their trades to
periods when there are more liquidity traders present. As a new result in this study, we
find that volume increases with transparency, which is consistent with the lower
spreads and higher volatility. We also find that trading volume significantly and
positively related to transparency, with a greater impact in the smaller shares by market
capitalization.
The remainder of the paper is organized as follows. Section 2 summarizes current
research on transparency with relevance to our study and not covered in the
introduction; Section 3 describes the institutional details; Section 4 provides an outline
of data and methodology while the results are presented in Section 5. Section 6 offers a
conclusion.
2. Literature Review
Other models relevant for the interpretation of our results are reported below. Baruch
(2005) models a specialist’s single price auction market similar to the auction that the
NYSE uses to open the trading day in two different environments: in one environment,
the limit order book is open; in the other, it is closed. The decrease in the transitory
component of the spread offsets the increase in the adverse selection component, so
that overall trading costs are lower and prices are more informative in the open-book
environment. Several empirical studies investigate the impact of changes in limit order
book transparency. Simaan, Weaver and Whitcomb (2003) investigate competition
among market makers on Nasdaq and expect that allowing anonymous quotes could
improve price competition and narrow spreads due to higher competition between
market makers. Bortoli, Frino, Jarnecic and Johnstone (2006) find a decrease in the
best depth when more levels of the limit order book are displayed at Sydney futures
7

---

markets. Comerton-Forde, Frino and Mollica (2005) investigate Euronext Paris, the
Tokyo Stock Exchange, which stopped displaying broker IDs in 2003, and the KE also
investigated in this study. They find a decrease in relative bid-ask spreads and effective
spreads at Euronext Paris, little change at the TSE and the reverse effect at the KE.
Comerton-Forde and Tang (2009) examine the effects of the removal of broker
identifiers from the central limit order book of the Australian Securities Exchange.
They find that spreads and order aggressiveness decline, and order book depth
increases, with the introduction of anonymous trading. All of these studies are
potentially affected by endogeneity in the controls that are utilized. They conclude that
some of the observed improvements in liquidity may be significant only if the stocks
are sufficiently large and liquid. Some recent studies find positive effects of
transparency. Boehmer, Saar and Yu (2005) investigate pre-trade transparency of order
level information, showing that effective spreads of trades decreased and liquidity
improved following the introduction of Open Book at the NYSE (which provided any
willing subscriber with order size and price information previously only available to
specialists). Hendershott and Jones (2005) show that price discovery declined when the
ECN Island stopped providing order book information for exchange traded funds.
Eom, Ok and Park (2007) correct for endogeneity in market quality metrics used as
dependent or control variables and find that two Korean increases in the level of order
book information displayed to the public has positive effects on market quality.
3. Institutional details
The KE is a typical order-driven market, where buy and sell orders compete for best
prices. The whole trading procedure - from order placement to trade confirmation - is
conducted in an electronic order driven system. Throughout the trading hours, orders
8

---

are matched according to price and time priority. The opening and closing prices,
however, are determined by call auctions.
In KE, every stock has a daily price variation limit set at ±15% of the previous day
closing price. Orders outside this limit are rejected. If the price limit is reached, trading
is not halted and will continue when/if the price moves back to within this ±15% daily
price variation limit.
The KE market is open from 9:00 a.m. to 3:00 p.m. during weekdays. Investors can
submit their orders from 8:00 a.m.2, one hour before the market opening. Orders
delivered to the market during the period from 8:00 a.m. to 9:00 a.m. are queued in the
order book and matched in a call auction at 9:00 a.m. to determine opening prices.
After opening prices are determined, the trades are conducted by continuous auction
until 2:50 p.m., 10 minutes before the market closing. During the last 10 minutes,
orders are pooled again and executed by call auction to determine closing prices of the
day. Lunchtime breaks were abolished in May 2000. From 3:10 p.m. to 4:00 p.m. the
KE operates an after-hours session for 50 minutes. During the after-hours sessions,
orders are matched at the closing prices of the day. In addition to limit and market
orders, the KE allows another type of orders, the limit-or-market-on-close orders. This
is a limit order that automatically converts to a market order at the market closing to
participate in the call auction for closing price determination. The trading unit is 10
shares for stocks. Orders with sizes smaller than the trading units (“odd-lots”) are
traded at the after-hours session or on the OTC market. The tick sizes vary according to
the price levels.

2 Since December 2003, the pre-hours session newly introduced has lasted from 7:30 – 8:30 am with the
closing prices of the previous day applied for orders. Orders delivered to the market from 8:30 – 9:00 are
queued in the order book and matched by call auction method to determined opening prices.
9

---

The KE has gradually increased the transparency of order and trade information. The
exchange provides various means of information dissemination. Market information on
price and trading volumes such as current price, highest/lowest prices, opening/closing
prices, trading volume and value, is available on a real time basis through information
terminals distributed by the KOSCOM (the Korea Securities Computer Corporation),
commercial telecommunications networks, websites of the KE and securities firms, etc.
The order book information is open to the public. During call auctions (08:00-09:00
and 14:50-15:00), all investors can get information about expected matching prices
(opening and closing prices), expected quantities to be matched at the expected
matching prices, the prices and order quantities of expected best bid and ask
quotations. During the continuous auction (09:00-14:50), order information on the five
best bid and ask quotations for all listed stocks is disclosed to the public on a real time
basis along with the aggregate order quantity of each side. From January 2002, the
scope of the bid/ask information disclosed was extended to the ten best bid and ask
quotations. This change was designed to prevent any attempt to mislead investors by
placing unreasonably large orders (fake orders) at the prices that are unlikely to be
matched, i.e., intentionally increasing the aggregate order quantity of a certain issue.
Statistics identifying of the five most active brokers has also been disseminated to
public since 25 October 1999.
4. Data and methodology
4.1 Data

Our data consists of the top 49 stocks by trading activity that were traded continuously
at the KE during the investigated time period 1 May 1999 to 31 March 2000. As a
result, the data includes only the stocks for which we obtain intraday trade and quote
10

---