ESSEC

MSc. in Finance

FINM 32408 /32417

TRADING AND MARKET STRUCTURE

SESSION 1
AN OVERVIEW OF THE TRADING INDUSTRY

Class Handouts

Laurence Lescourret

L Lescourret 1
 An Overview of the Trading Industry

1. The trading industry

a) Which type of markets ?
b) Who are the players ?
c) Which type of orders ?

d) The rules of the game: How are the

markets organized?

e) What type of instruments are traded?

f) Who are the regulators ?

2. Stylized facts

3. Policy issues

L Lescourret 2
 1.) The Trading Industry

The Trading Industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) Rules of the game: How are the markets

organized?

e) What type of instruments are traded?

f) Who are the regulators ?

L Lescourret 3
 1.) The Trading Industry

Markets

Places / Platforms on which traders gather to trade

securities
Electronic systems
Floor markets

We make a distinction between

Primary markets in which a stock is issued and
primarily trades
Secondary markets in which a stock is traded

! The course focuses on the organization of secondary

markets, in particular, cash equity markets.

L Lescourret 4
 1.) The Trading Industry

Roles of Financial Markets/Exchanges

Services related to trading

Listing and trading services

Organizing the trading of securities

Wealth transfer (best intertemporal allocation)
Risk transfer
Allocating scarce resources

Allowing price formation and price discovery

Collection and distribution of market data on quotes,

trades, etc.

Regulation (self-regulatory responsibilities)

e.g., FINRA (Merger of NASD and NYSE)

L Lescourret 5
 1.) The Trading Industry

Some Figures
Largest exchanges by value of share trading in the
Electronic order book in 2015 (Jan-Sept)

%change/ %change/
2015
Exchange Jan/Sep14 Jan/Sep14
January Yeartodate (inUSD) (inlocalcur)
NorthAmerica
BATSGl oba l Ma rkets US 1240661,5 10653413,6 13,4% 13,4%
Na s da qUS 1183196,4 9498061,0 8,7% 8,7%
NYSE 1520414,3 13165372,5 16,5% 16,5%
TMXGroup 120296,7 923260,1 8,4% 6,0%

AsiaPacific
HongKongExcha nges a ndCl ea ri ng 155247,2 1773653,5 62,2% 62,2%
Ja pa nExcha ngeGroup 401511,0 4255426,6 6,1% 24,6%
Sha ngha i StockExchange 1277890,6 17764180,7 508,3% 509,4%
ShenzhenStockExcha nge 800389,1 14511601,3 294,7% 295,8%

EuropeAfricaMiddleEast
BATSChi xEurope 290697,1 2426786,5 27,6% 55,2%
BMESpa ni s hExchanges 94016,4 772142,2 1,3% 20,3%
Deuts cheBoers e 142125,0 1213786,2 10,8% 34,9%
Euronext 183869,6 1614640,5 11,9% 36,2%

Source: WFE 2015 Market Highlights

L Lescourret 6
 1.) The Trading Industry

Some Figures

Source: WFE 2015 Market Highlights

L Lescourret 7
 1.) The Trading Industry

History (in very brief)

1602: Amsterdam Stock Exchange

1724: Paris Bourse

1792: New York Stock Exchange (NYSE)

1971: Nasdaq

1980s: dematerialization of securities, introduction of

the first electronic limit order books ( Big Bang ).

2000s: Demutualization of stock exchanges (for-profit

companies). IPO of these new for-profit companies.
First wave of mergers of exchanges/platforms
examples:
2000 : The PanEuropean Exchange, Euronext
(Amsterdam, Brussels, Paris in 2000, then Lisbonne
2002)
2007: Euronext acquired by NYSE
2013: NYSE-Euronext acquired by ICE

2005: RegNMS in the U.S.

2007: MIFID 1 in Europe
2012: Volcker Rules, Dodd-Franck, French Taxes
2013/14: NYSE-Euronext acquired by ICE / Euronext
to Spin-Off ICE via listing
L Lescourret 8
 1.) The Trading Industry

Examples of U.S Equity Markets

Primary and Secondary Markets

13 Exchanges: around 60%

New York Stock Exchange (ICE group)
Nasdaq (Nasdaq-OMX group)
Other exchange: BATS
US regional exchanges (Pacific Exchange /
Archipelago Acquired by NYSE-Euronext ; Chicago
(formerly Midwestern) Stock Exchange ; Boston Stock
Exchange - Acquired by Nasdaq-OMX ; Philadelphia
Stock Exchange - Acquired by Nasdaq-OMX, etc.)

The Third Market : around 40%

45 dark pools (Liquidnet, POSIT, SigmaX, etc.) : 18 %
Off-exchange trading (22 %): internalization, OTC
trading.

L Lescourret 9
 1.) The Trading Industry

Other Markets

Europe:
Regulated markets: exchanges
London Stock Exchange (LSE)
EuroNext (Paris/Amsterdam/Brussels/Lisbon),
now part of the NYSE group
Deutsche Brse (DB)
Milan Stock Exchange
Swiss Stock Exchange (ou Virt-X)
Stockholm/Copenhagen/Helsinki/Oslo (OMX)

20 dark pools in Europe (10% of mkt share)

Off-exchange trading (30 % of makt share)

Sources: Rosenblatt Securities, Tabb group

L Lescourret 10
 1.) The Trading Industry

Primary Markets

Issuance of new securities on the primary market

Origination and sale of new securities.

Initial Public Offering Mechanisms

Book-building
Auction/tender offer
Fixed price offering

L Lescourret 11
 1.) The Trading Industry

Listing Requirements
Set of conditions imposed by a given stock exchange
upon companies that want to be listed on that
exchange
Conditions on the minimum number of shares
outstanding, on the free float, on a minimum
market capitalization

Conditions on audited financial statements

* Minimum 25% shares distributed to the public

* Three-year track record required (exemption for funds)
* IFRS, US GAAP or recognised accounting standards
EU regulated main
* Listing Agent
Euronext market for large, mid &
small caps
Ongoing obligations
* Audited FY and un/audited half-year accounts

Procedures
Need of a listing sponsor sometimes

Choice of the IPO mechanism

L Lescourret 12
 1.) The Trading Industry

Delisting

Decision to delist
Public-to-Private transaction ( going private )
Example: LBO

Going dark : very costly to maintain a listing

(Sarbanes-Oxley in the U.S.). In the U.S., it is
possible to be traded on OTC markets (OTCBB, or
Pink Sheets).

Obligation to delist
Change of legal status

Impossibility to maintain the listing conditions

required by the exchange.
Example: too low market capitalisation, too
low trading volume, etc

L Lescourret 13
 1.) The Trading Industry

Secondary Markets

Markets in which investors trade securities

previously issued

Different types of markets organisation:

Call or batch vs. continuous markets

Order-driven vs price-driven markets

Consolidated vs fragmented markets

Opaque vs transparent markets

L Lescourret 14
 1.) The Trading Industry

Financial Markets: Regulated vs. OTC

OTC, Over-The-Counter markets

A market with light or no regulatory authority
intervention
Bilateral transaction
Example: the OTC credit market (the corporate
single name bond market, CDS market, US
Treasuries, etc.)

Regulated Markets
A supervised market by a regulatory authority
that regulates the issuance of securities,
transactions, quotations, etc.
Example: NYSE, Nasdaq, BATS, Euronext
Characteristics:
Member Status
Clearing House.
Example: LCH Clearnet for NYSE-Euronext

L Lescourret 15
 1.) The Trading Industry
Financial markets and price formation:
an ideal organization
The Walrasian classical model
Construction of
the aggregate supply
and aggregate demand curves

Determination of a unique and competitive

equilibrium price

However, real-word financial markets show

that price formation is more complex :

Different trading protocols

across markets/platforms

Different impacts on the price formation,

transaction costs and asset prices

A look into the black box

L Lescourret 16
 1.) The Trading Industry
The ideal organization :
the Walrasian mechanism ?
H1. Free and perfect information (available to all
participants)
H2. No transaction costs
H3. Agent are price-takers
H4. Prices and quantities can be infinitely divisible
H5. Atomicity of agents who act independently.

the Walrasian tatonnement

- The auctioneer starts by announcing an indicative price

- The agents register how much they would like to buy or
sell at this price (no incentives to manipulate the
equilibrium price)
- In case of an excess suppy or demand, the auctioneer
announces a new price.
- The process stops when the price is set such that the
supply is equal to the demand.

Is there only one price in reality?

L Lescourret 17
 1.) The Trading Industry
Real-word market organizations:
(1) Demsetz (1968)

The quoted price is not unique, there is a double price:

the sell price (ask : Ai) and the buy price (bid : Bi),
quoted by a market maker

L Lescourret S1 FIN 266 / T3 2009 18

 1.) The Trading Industry
(2) The Nasdaq before 2002
(a dealer market)
On the Nasdaq, more than one market-makers make the
market for on security. Bertrand competition ?
Source: MoneyNet

L Lescourret 19
 1.) The Trading Industry

(3) An ECN called Island

(ii) Snapshot from of the electronic limit order book of

the Yahoo stock! published by the ECN (Electronic
Communication Network) Island, 12/01/01.

L Lescourret S1 FIN 266 / T3 2009 20

 1.) The Trading Industry

(4) Euronext Paris

Source: Euronext

L Lescourret 21
Source: DERIVA B.V.
 1.) The Trading Industry
Consequences: Trading frictions and
transaction costs

Source : Plexus Group

L Lescourret 22
 1.) The Trading Industry

Consequences

Impact of transactions costs on

the liquidity of a firm
the cost of capital
Etc.

cf asset pricing and liquidity, cost of capital

and liquidity, etc.

L Lescourret 23
 1.) The Trading Industry

1. The trading industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) Games rules: How are the markets
organized?
e) What type of instruments are traded?
f) Who are the regulators ?

L Lescourret S1 FIN 266 / T3 2009 24

 1.) The Trading Industry

Main players

Buy-side players buy liquidity services, whereas sell-

side players provide them

Investors (buy-side players)

Institutional investors (pension funds, investment
funds, etc...)
Retail investors
Arbitragers
or
Hedgers, liquidity traders
Speculators: informed trader, insider, scalper,
at the extreme: rogue trader

Intermediaries (sell-side players)

Broker : trade on behalf of the clients
Voice-electronic brokers (GFI), Electronic
Communication Networks (ECNs), etc.
Market-makers : trade for their own account
Independent Dealers (KCG Ex Knight)
Dedicated market-makers : specialist/LP
Endogenous market-makers (HFT firms like Virtu
KCG (ex Getco), Hedge Funds (Citadel)
Broker-dealers: do both
Large banks
L Lescourret 25
 1.) The Trading Industry

1. The trading industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) Rules of the game: How are the markets

organized?

e) What type of securities are traded?

f) Who are the regulators ?

L Lescourret 26
 1.) The Trading Industry

Market and limit orders

Orders = traders statement of their intention to trade

Limit Order (sign / size / price / maturity instructions)

Specification of a limit maximum (minimum) price to
pay in case of a buy (to receive in case of a sell) -> Q(P)
Advantage : control the execution price
Risks : (1) risk of not being executed ; (2) risk of being
picked off

Market Order (sign / size)

No price specification
Advantage : Immediate execution
Risk : price risk (price impact)

Limit orders supply liquidity (cheaper), whereas market orders

demand liquidity (but pay the spread)

L Lescourret 27
 1.) The Trading Industry

1. The trading industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) The rules of the game: How are the markets

organized?

e) What type of securities are traded?

f) Who are the regulators ?

L Lescourret 28
 1.) The Trading Industry

The rules of the game

Rules that regulates trading and trading procedures

What is the trading protocol?

Mechanism : (double) discriminant auction,
uniform price auction, etc.
> Role of intermediaries : brokers, brokers-
dealers, dealers,

How do the trading protocols affect

the price formation in stocks,
bonds, derivatives,
and foreign exchange order flow?

L Lescourret 29
 1.) The Trading Industry

Trading protocols

1. Order-driven markets
No intermediary / Direct match between buy and
sell orders
Rules based on order precedence rules that rank
and match orders
1. Auctions
Exogenous date of transaction (CALL AUCTION)
Endogenous date of transaction(CONTINUOUS) /
2. Crossing networks : They use the price derived
from the securities primary markets. They do
not produce price discovery (cannot be used as a
stand-alone system)

1. Quote-driven markets
An intermediary supply liquidity. Dealers
participate in every trade, and may trade with
each other (interdealer-trading)

L Lescourret 30
 1.) The Trading Industry

Call auction markets

Call or batch market # Walrasian market

Transactions occur only when the market is called
(EXOGENOUS DATE)
Multilateral transaction at the same time
Many buy orders, many sell orders -> 1
transaction, 1 execution price (O=D)

All sell orders (buy) with an inferior price than

the equilibrium price are cleared (higher price)
Examples : Gold market (London), Opening and
closing of some markets (NYSE-Euronext, Toronto
or Stockholm), mechanism of equity markets
characterized by with low liquidity (e.g.,
Alternext)
L Lescourret 31
 1.) The Trading Industry

Continuous markets

Process orders as they arrive, and transactions occur

whenever there is a matching
-> ENDOGENOUS date of transaction.

Examples : NSC (ex SuperCAC), CATS, GLOBEX,

AQS, MTS, etc.
(see examples of electronic exchange systems).

Among continuous markets, we make a distiction

between
Quote-driven markets
Order-driven markets

L Lescourret 32
 1.) The Trading Industry

Example: Paris Bourse,1994

Source : Biais, Foucault, Hillion (1997), Microstructure des

Markets financiers, PUF Finance.

L Lescourret 33
 1.) The Trading Industry

Quote-driven Markets

Markets with some designated liquidity suppliers

(Dealers, market-makers)

Characteristics
Counterparty : intermediaries
Role of intermediaries :
Dealers supply all liquidity or immediacy
Dealers trade on their own account. They supply
liquidity using their own inventory account. They
make a market by providing continuously ask and
bid prices
What are the risks ? How to be compensated ?

Examples : OTC Markets in general

NASDAQ before the implementation of
Supermontage in october 2002
OTC Bond Markets: corporate, municipals, U.S.
government, foreign sovereign bonds
Loans, Mortgages
Derivatives Markets (CDS, etc.)
Currencies
Commodities
L Lescourret 34
 1.) The Trading Industry

Order-driven markets

Electronic limit order market

Characteristics
Counterparty: another order. Buy and sell
orders are matched directly, anonymously
Production of liquidity is guaranteed by limit
orders
Examples
Call or batch markets: the opening of Euronext,
NYSE, Tokyo SE, etc...

Continuous electronic limit order book

(snapshot of unmatched limit orders): Toronto
SE, Eurolist dEuronext, Tokyo SE, etc.

L Lescourret 35
 1.) The Trading Industry

Floor markets Before A dinosaur

Pit trading markets

Example : MATIF before1998, CBOE

Markets with a trading floor ( crowd trading, trading

floor
example: NYSE, AMEX

photo: Chicago Board of Trade (CBOT)

L Lescourret 36
 1.) The Trading Industry
Stock markets nowadays
Source: Terry Hendershott

L Lescourret 37
 1.) The Trading Industry

Consolidated vs fragmented markets

Consolidated markets
Formerly, this meant that all trading occurred on one
central computing system
Only one (double) price exists (buy side/sell side) at any
given moment for a given asset
examples
Euronext, TSE (Toronto Stock Exchange),
Stockholm stock exchange
Drawbacks: Monopoly fees
Advantages: Network effect: better liquidity, ease to
find a counterparty

Fragmented Markets
Several prices co-exist
Sources of the fragmentation
Organisation of Markets : NASDAQ, FX
Market practices : internalisation, preferencing,
new rules: MiFID (introduction of Chi-X),
Regulation (regNMS, MiFID)
Drawbacks : Weight of the intermediaries, opacity,
fragmentation of orders and price discovery
Advantages: Competition, innovation -> decrease of fees
L Lescourret 38
 1.) The Trading Industry

Hybrid Organisations

Trend to design hybrid market structures

Mix elements of the various market types

Downstairs markets, centralised by an electronic

limit order, co-existing with an intermediary markets
( upstairs markets)
Example: the NYSE

Centralised order-driven markets introducing

market-makers :
Examples: Alternext, launched in May 2005, Eurolist
for the cross-listed securities (LP)

Fragmented quote-driven market implementing

centralized limit order book
Examples: FX with Reuters Dealing 2000-1 /2 , Nasdaq
with the introduction of SuperMontage in October 2002.

L Lescourret 39
 1.) The Trading Industry

Other trading rules

Price and time priority (FIFO) : Euronext, NYSE.

Price grid (tick): -> Discreteness of prices

Trading halts and circuit breakers

...

L Lescourret 40
 1.) The Trading Industry

Clearing and settlement

Role of the Clearing House:

Counterparty of all transactions
No counterparty risk

Clearing and settlement

Consolidation on the clearing houses
In the US: 1 entity: Depository Trust and Clearing
Corporation (DTCC)
In Europe: LCH.Clearnet (Euroclear) and Clearstream
(Deusche Brse)

L Lescourret 41
 1.) The Trading Industry

1. The trading industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) Rules of the game: How are the markets

organized?

e) What type of securities are traded?

f) Who are the regulators ?

L Lescourret 42
 1.) The Trading Industry

Which instruments can be traded?

Cash Markets
Stocks
Rougly 5,000 actively traded stocks in the US. The largest
1,500 acount for about 90% of the total market
capitalization.

Bonds (Treasury vs Corporate bonds)

Commodities (oil, coffee, )

Currencies

Exchange-traded funds, structured fund

certificats

Derivatives Markets
Options

Forwards, Futures,

Swaps, etc.
L Lescourret 43
 1.) Generalities

1. The trading industry

a) Which type of markets ?

b) Who are the players ?

c) Which type of orders ?

d) Rules of the game: How are the markets

organized?

e) What type of securities are traded?

f) Who are the regulators ?

L Lescourret 44
 1.) The Trading Industry

The Regulators

Aim: designing better markets (fair, stable,

protection of all investors)

Markets are regulated by

Legislators
Regulatory agencies
SRO

L Lescourret 45
 1.) The Trading Industry

The Regulators

Central Banks

Governemental agencies (write and enforce

regulation to implement the law
SEC (Securities and Exchange Commission):
Created in 1934 (Securities Act of 1933 +1934) to
regulate the US market
CFTC (Commodity Futures Trading Commission)
Created in 1975
FCA (Financial Conduct Authority) and PRA
(Prudential Regulation Authority) (ex-FSA, Financial
Services Authority):
Created in 2013 in the U.K.
AMF (Autorit des Marchs Financiers):
Merge of the COB and the CMF (Loi de Scurit
Financire du 17 juillet 2003 )

SRO (Self-regulated Organizations) (regulate their

members) : Exchanges, dealer association. Example:
FINRA (Financial Industry Regulator Authority)
born from the merge of NYSE and NASD.

World Organizations (try to coordinate regulations

across broders): WFE, IOSCO, etc.
L Lescourret 46
 2.) Main stylized facts

Where are we at the moment?

1. An overview of the trading Industry

2. Main stylized facts

3. Policy issues

L Lescourret 47
 2.) Main stylized facts
Some stylized facts and empirical
findings
Existence of a bid/ask spread

U-shapes in stock returns, in trading volume

(weekly, daily...), daily J-shape pattern for
volatility, etc.

Diagonal effect in limit order books

Negative autocorrelation of price change ;

Tick-by-tick data, irregularly time-spaced

observations

95% of orders are cancelled or modified

Big data: Nasdaq absorbs 1 million of

msg/second, 3 billions of trades per day
L Lescourret 48
 2.) Main stylized facts

Trading volume and stock returns on

the NYSE, (Jain and Joh, 1988)

L Lescourret 49
 2.) Main stylized facts
Trading volume on the Paris Bourse
(Mai and Tchemeni, 1995)

L Lescourret S1 FIN 266 / T3 2009 50

 2.) Main stylized facts
J-curve of the standard deviation of returns on
the NYSE (Wood, McInish and Ord, 1985)

L Lescourret S1 FIN 266 / T3 2009 51

 2.) Main stylized facts
U-curve of the intraday volatility on theParis
Bourse (Hillion and Suominen, 2004)

L Lescourret 52
 2.) Main stylized facts
Trading volume and volatility on the Nasdaq
(Barclay and Hendershott, 2004)

L Lescourret 53
 2.) Main stylized facts

Bond Market - 1 (Fleming, 1997)

L Lescourret 54
 2.) Main stylized facts

Bond Market - 2 (Fleming, 1997)

L Lescourret 55
 2.) Main stylized facts

Bond Market 3 (Fleming, 1997)

L Lescourret 56
 2.) Main stylized facts

CDS Markets 4 (Fulop and Lescourret)

European single-names CDS (5-y, Senior)

0,035 0,12

0,03 Volatility
0,1
Transaction costs (average)

0,025
0,08

0,02

0,06

0,015

0,04
0,01

0,02
0,005

0 0
1 2 3 4 5 6

L Lescourret 57
 3.) Policy issues

Where are we

1. An overview of the trading industry

2. Some stylized facts and empirical

evidence

3. Policy issues

L Lescourret 58
 3.) Policy issues

Market Quality, market performance

Liquidity
Elusive concept : a multi-faceted phenomenon,
referring simultaneously to the availability, cost and
promptness of trading.
Key condition of well-functioning markets (OHara,
1995)
3 dimensions: width (b/a spread) ; depth (number of
shares available at given quote) ; resiliency (speed of
price adjustments after a liquidity shock)

Liquidity and Transaction costs

Minimizing transaction costs allows to attract order
flow, which creates liquidity which, in turn, attracts
liquidity.

Direct transaction costs : commissions, brokerage fees,

trading fees Make/take spreads, etc.

Implicit transaction costs : adverse-selection costs ,

inventory-holding costs, order-processing costs

Fees for clearing and settlement

Taxes

L Lescourret 59
 3.) Policy issues

Market Quality - Contd

Liquidity and Transaction costs

L Lescourret 60
 3.) Policy issues

Market Quality Contd

Price formation and price discovery

(Informational efficiency)
The price should reflect adequately the
supply and the demand which reflect
agents expectations.

Market Stability
A measure : Price Volatility

Risk-Sharing
Optimal allocation of quantities (of
secutrities) among agents having
different risk aversion.

L Lescourret 61
 3.) Policy issues

Market transparency: an issue?

What information do participants have access to?

(a) Ex-ante transparency (before trading)

of the best bid/ask prices (firm price)
(NYSE, Euronext, Toronto, etc.)
Of the available quantity at this price (Euronext,
Nasdaq) ?
How many limits of the order book are available
(Euronext, ECNs), in real time or in 5 to 15 min ?
Of the identity of limit order traders ?
Can we use hidden/iceberg orders?

b) Ex-post Transparency (after trading) :

Can we observe the last trades? (Price, quantity,
identity ) ? What is the ticker delay?
Dash 5 / 6 (Rule 11Ac1-5, Rule 11Ac1-6)

L Lescourret 62
 3.) Policy issues
Current issues in market organizations
and regulation

Should OTC markets (the derivatives

market, the debt market) be more regulated,
more centralized, more transparent?
The TRACE experiment

Impact of low-latency trading

High Frequency trading:
Risks ? FlashCrash 2010 and other episodic
liquidity crashes, Knightmare in aug. 2012
Race competition among platforms
NYSE fined $5 m in sept. 2012 for sharing
date with clients before public (first-ever SEC
financial penalty against an exchange),
Euronext fined by AMF in Nov. 2015 (4 million
Euros)
New business models have emerged (fiber,
colocation, throughput ; news vendors, etc. )
L Lescourret 63
 3.) Policy issues

Current issues Contd

Dark pools and fragmentation

Around 25 dark pools in Europe ; around 50 in the U.S.
Is fragmentation good or bad?
New technologies: smart routers, data aggregators, etc.

Determination of the best execution among the

platforms :
Proliferation of the platforms: when to execute and
why?
Regulation :
In the US (Reg NMS) : the only price criteria (no-trade
trough rule) ; SEC Rules 11Ac5 and 6
In Europe (Mifid) : multi-criteria approach (price, speed of
execution, costs, probability of excution and settlement ,
size and nature of the order)

Governance of Exchanges, platforms

L Lescourret 64
Consolidation of the clearing and settlement systems
 ESSEC MSTF / FEA / MSc. in Finance

FINM 32417/32408
TRADING and MARKET ORGANISATION

SESSION 2 : CALL AUCTION MARKETS

Class Handouts

Laurence Lescourret
 Call Auction Markets

1. Introduction

2. Examples: (1) the gold market ; (2) illiquid

securities (Alternext Paris)

3. Description of the call auction mechanism

4. Attributes of the call

5. Call and market manipulation

6. Limitations

L Lescourret 2
 1.) Introduction
Trading platform overview
M. OHara, 2002
Equity trading system

Trading mechanism

- Single 64%

- M ultiple 33%

- N/ A 2%

Floor system 5%

Electronic system 79%

Floor and electronic system 17%

pure-order system 76%

hybrid system 19%

orden and hybrid system 5%

Trading process

- Continuous 36%

- Call 2%

- Combination 62%

Formal market makers

- Yes 31%

- No 69%

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9

(%) of exchanges (total=42)

17 exchanges reported using a call at either the

open or close, and 7 reported using a call at least
several times during the day. M. OHara.

L Lescourret 3
 1.) Introduction

An overview of the call

Essence: order-driven system

Batch orders together in a single multilateral trade,
How ? Double consolidation of the order-flow
1 place
1 exogenous pre-specified time

At what price?
At 1 single equilibrium price
Buy orders at this price and higher, and sells at this
price and lower, generally execute

L Lescourret 4
 1.) Introduction

Design features
Auctions
Example A: Regulated markets
Use for illiquid securities (mid-, small-caps stocks):
Eurolist at 10h30 and at 16h00, NYSE
Alternext Europe at 15h00);
Arizona Stock Exchange (5 call auctions per day),
Closed in October 2001;
Tawan Stock Exchange, formerly used high-
frequency auctions, now continuous;
Used at the opening, closing, and during trading
halts of some exchanges :
Euronext , Toronto, Tokyo, Nasdaq (Opening
and closing cross implemented in 2005), Xetra
(2 calls/day ; 1 at noon)

Example B: OTC Markets

The reference price of the gold (or silver) market
CDS settlement auctions
Primary market auctions of US Treasury securities

Crossing networks, now belonging to Dark Pools

Examples: POSIT, Liquidnet
Mechanism: cross orders at a number of prespecified
times during a day. Orders are submitted without price
indication, and ranked according to time priority. Price
priority does not exist: the price at which orders are
crossed is derived/imported from the stocks primary
markets. They do not produce price discovery, price
formation, and, thus price clearing. They cannot be
used as a stand-alone system.
L Lescourret 5
 1.) Introduction

Design features

Alternative call market structures

Several structures co-exist:

Call with the intervention of an intermediary
(Auctioneer, Market-Maker) :
Examples: Gold Market OTC at london
(www.kitco.com) ; NYSE (specialist);

Electronic call auction (without any human

intermediation) :
Examples : Arizona SE, opening and closing
or fixing on Euronext, opening on Xetra,
etc.

Design features
The matching algorithm
The level of transparency: dissemination of an
indicative auction price

L Lescourret 6
 2.) Examples

Example 1: Gold Market

Overview of the market: OTC Market - 24/7

OTC trading:
London is by far the largest global centre for OTC
transactions.
Most wholesale OTC trades are cleared through
London, and through the LBMA (London Bullion
Market Association)
Trading on Exchanges:
Limited amounts traded on NYMEX (NY
Mercantile Exchange) and TOCOM (Tokyo
Commodity Exchange).
Gold is traded as a security on the London, NY,
Johannesburg, and Australian SE.

LBMA: established in 1987. Represents the interests of

the participants in the wholesale bullion market (Members:
major international banks, bullion dealers, refiners,
shippers, brokers)

L Lescourret 7
 2.) Examples

Example 1: Gold Market

Sept. 12 1919, the Bank of England -> NM

Rothschild & Sons to organize the Gold fix (the
London Gold Fixing )

Organization of the Gold Fixing :

2 call auctions per day: 10h30 / 15h00
5 Gold fixing market-makers (Members of the LBMA):
Barclays (replaced Rothschild since 2004), SGCIB,
Deutsche Bank, HSBC USA, Bank of Nova Scotia-
ScotiaMocatta
The clearing price is used as a reference price for
settling contracts between members of the LBMA
(producers, market-makers, investors) and for central
banks.

Mechanism:
Physical meeting (at Rothschild and Sons) until 2004,
Since the first meeting occurred (Sept. 12, 1919) in
the , at the start of each call auction, the Chairman
announces an opening price to the other 4 members
who relay this price to their customers ; the auction
will last as long as it is necessary to set a price that
clears the market.
April 14, 2004 : Rothschild and Sons resigned as Chair
of the London Gold Fix withdraw from the market and
was replaced by Barclays. The call auction is now
conducted via conference-call.
January 2014-April: Deutsche Bank tried to sell and
quit the fix in April.
L Lescourret 8
 2.) Examples

Example 1: Gold Market

Now,
BMA Gold Price was launched on the 20th March
2015 to replace the historic London Gold Fix.

ICE Benchmark Administration (IBA) provide the

auction platform;

Auction process: electronic and physically settled,

conducted in dollars, with aggregated and
anonymous bids and offers as well as being
published on-screen and in real-time;

12 participants: Barclays Bank, Bank of China,

China Construction Bank, Goldman Sachs
International, HSBC Bank USA NA, JP Morgan,
Morgan Stanley, Societe Generale, Standard
Chartered, The Bank of Nova Scotia -
ScotiaMocatta, The Toronto Dominion Bank and
UBS

From 1 April 2015 the LBMA Gold Price became a

regulated benchmark

L Lescourret 9
 2.) ExAmples

Buy-side ? and incentives for entering

into the sell-side?

L Lescourret 10
 2.) Examples

Example 2: Alternext Paris

Illiquid securities

L Lescourret 11
 3.) Description

Description of a call

Phase 1: Period of order accumulation (no trades)

Open book auction orders are electronically displayed.
Investors can observe orders submitted and the
tentative clearing price, which the algorithm computes
continuously
Example: Euronext
Closed book auction: hidden book content
Example: NYSE opening

Phase 2 : Close (often random) of the auction

The algorithm ranks the orders according to price, and
time priority (construction of the cumulated supply and
demand functions):
First in priority are the market orders. Then, buy orders
are ranked by decreasing price limits (demand function) /
the sell orders are ranked by increasing limit price (supply
function)
The algorithm selects the clearing price, according to
the principle of maximization of the number of shares
traded. Other trading rules are required in case of
multiple clearing prices.(*)
A unique price is set :uniform-price auction.

Phase 3: The market clears

According to price time priority, all orders submitted at
price higher (lower) than the market clearing price for
buy (sell) execute ; orders at the marginal price might
instead be partionnaly filled.

(*) If more than one price satisfies the volume maximization, other
price-setting rules come into play: minimizing the difference with
the reference price (for instance, the closing price of the day before)
L Lescourret 12
 Example 1 : A snapshot of the limit order book @ 17.30
(Accumulation of orders between 17.25 and 17.30)
Heure Ord. Qt Achat Vente Qt Ord. Heure
17:25 1 1 500 ATP 63,25 4 000 8 17:39
17:26 2 2 800 63,7 63,2 8 000 2 17:40
17:26 2 1 000 63,35 63,7 3 000 2 17:40
17:27 1 300 63,25 63,35 7 500 2 17:40
17:28 3 1 100 63,2 63,2 50 1 17:40
17:28 2 3 000 63,15 63,15 150

Construction de la fonction de demande Construction de la fonction d'offre

Demande Prix Offre Prix

9 700 63,15 63,15 150
6 700 63,2 63,2 8 200
5 600 63,25 63,25 12 200
5300 63,35 63,35 19 700
4 300 63,7 63,7 22700

63,70

63,35

Prix 63,25

63,20

63,15

0 5 000 10 000 15 000 20 000 25 000

Offre
Nombre de titres
Demande

The trading volume (min(Bp,Sp)) is maximized for

p*=63,20 euros at which 6 700 stocks are executed.
The algorithm finds the price that, scanning across
prices, maximizes the minimum of the buy and sell
aggregate order quantities at each price
L Lescourret 13
 Example 1 : A snapshot of the limit order book @ 17.30
(Accumulation of orders between 17.25 and 17.30)
Heure Ord. Qt Achat Vente Qt Ord. Heure
17:25 1 1 100 ATP 63,25 4 000 8 17:39
17:26 2 2 800 63,7 63,2 5 000 2 17:40
17:26 2 1 000 63,35 63,7 3 000 2 17:40
17:27 1 300 63,25 63,35 7 500 2 17:40
17:28 3 1 100 63,2 63,2 50 1 17:40
17:28 2 3 000 63,15 63,15 150

Construction de la fonction de demande Construction de la fonction d'offre

Demande Prix Offre Prix

9 300 63,15 63,15 150
6 300 63,2 63,20 5200
5 200 63,25 63,25 9200
4900 63,35 63,35 16700
3 900 63,7 63,70 19700

63,70

63,35

Prix 63,25

63,20

63,15

0 5 000 10 000 15 000 20 000 25 000

offre
Nombre de titres Demande

The trading volume is maximized for 2 prices : at

p=63,20 euros p=63,25. Other trading rules are
needed to set a unique price.

L Lescourret 14
 3.) Description

TraderEx

Sparse book
Thick book
 3.) Description
TraderEx: the price-setting
mechanism
When you trade at the open, you must know the rules
of auction price settin in order to determine how best
to price and size your orders in a call auction

Price-setting rule in TraderEx

Price-setting rule #1
Select the price that maximizes the number of shares
that trade.
Price-setting rule #2
If there is more than one price that maximizes trade
quantity, select the price with the smallest imbalance,
between the buying and selling quantities.
Price-setting rule #3
If there is more than one price that has the same trade
quantity and the same buy-sell imbalance, select the
price that conforms with market pressure.
Price-setting rule #4
For more than one price to yield the same maximum
turnover, with equal imbalances and market pressure,
select the price which is closest to the most reference
price.
 4.) Attributes

Remarks

Critical features in calls

Nearly all orders are limit orders ;

The limit price of an order in the call

auction is typically not the price at which
the limit order executes in that facility
price improvement. An example?

L Lescourret 17
 4.) Attributes

Importance of call auction protocols

At the open
Stylized facts: U-shaped trading volume
Paris Bourse: almost 10% of the total daily
volume is traded at the opening
NYSE: around 10%
Tokyo: between1/3 and of the total volume

Aim of implementing a call at the open: Improving

the accuracy of price discovery after the overnight
interruption (numerous orders have accumulated)

Consequence: Nasdaq implemented a centralized

call auction mechanisme in a formerly
decentralized dealers market in 2004.

L Lescourret 18
 4.) Attributes

Importance of call auction protocols

At the close
Stylized facts: U-shaped trading volume
Italian SE : 10% of the daily exchanged volume
Why? The closing price is broadly used as a
reference price (RP)
For fund managers: to calculate the NAV on
which redemptions and purchases are based
For index calculation: Entry/exit of stocks.
Index arbitrageurs will try to unwind their
positions by buying or selling short at closing
prices.
Portfolio performance evaluation: M-T-M of
portfolios, positions, etc.
Derivatives benchmarks: index futures
contracts and other cash-settled derivatives
are usually settled at closing prices.
Consequences: A closing call auction is also
implemented by the Nasdaq in 2004.

L Lescourret 19
 4.) Attributes

Performance

Consolidation of the order flow:

Volume concentration: orders are consolidated

both geographically (in one place) and temporally
(over time)

Positive network externalities:

Good for the illiquid stocks ; in case of market
stress (krach of 1987, see R. Schwartz)
It should reduce transaction costs since the
price-impact is expected to be lower

Better price discovery due to the concentration

of orders?
Pre-opening of SE (Euronext, Toronto, Tokyo,
etc.),
Trading halt

It should ease to stabilize prices

The introduction of a call auction at the closing
is shown to help to reduce the volatility
(Hillion and Suominen, 2004 ; Kandel et al.
2008).

L Lescourret 20
 4.) Attributes

Advocates of a call

Fixed deadlines and manipulative

strategies?
bid shielding
one solution: randomization of the clearing
time
A system which is more difficult to
manipulate ?

A market design response to slow down

high-frequency tradinh arms race ?
Budish, Cramton, and Shim (2015)

L Lescourret 21
5.) Call and Price manipulations

Market manipulations : Definition

The Market Abuse, or Market Manipulation, describes a
deliberate attempt to interfere with the free and fair
operation of the market and create artificial, false or
misleading appearances with respect to the price of a
security (1).
Art. 2 (a) 1 Transaction or order to trade:
Which give, or are likely to give, false or misleading
signals as to the supply of, demand for, or price, of
financial instruments
Which secure, by a person, or persons acting in
collaboration, the price of one or several financial
instruments at an abnormal level

Art 2(b) transactions or orders to trade which

employ fictious devices or any other form of deception
or contrivance.
Art 2(c) dissemination of information through the
media, including the Internet, or by any other means,
which gives, or is likely to give, false or misleading
signals as to financial instruments, including the
dissemination of rumours and false or misleading
news, where the person who made the dissemination
knew, or ought to have known, that the information
was false or misleading.
Directive 2003/6/EC

L Lescourret 22
5.) Call and Price manipulations

Call auction and Price Manipulation

Manipulation of closing prices

More difficult to manipulate: a manipulator needs to
send a very large order to expect to move the clearing
price (the required capital can be prohibitive)

Why manipulate the closing price ? The closing price is

used as a reference price to calculate the NAV of funds,
M-T-M of the trading positions and for index trackers
ONeal (2001) : Price manipulation at the end of the fiscal
year to increase the asset value of the funds
Carhart et al. (2002) observed that the funds managers
manipulate the prices at the end of each quarter by
trading at the last moment and at the last price (last-
minute trading on last-trade prices: painting the tape ,
marking up , portfolio pumping )

Burstoftradingatthe
highestpointinthelast5
min

L Lescourret S2 FINM 266 / FINM 266 23

5.) Call and Price manipulations

Closing Price Manipulation cont-d

Example: on TSX, RT Capital Management

systematically inflated closing prices during 5
consecutive months (Oct, 30 1998 to March, 31 1999)
to improve the performance of a RT Capital
Management pension fund.

Example: Entry in the limit order book of a very

large buy order at a non representative price, few
seconds before the call
SGX: Buy order of 2 million shares, which pushed
the prices up by 30% (the Straits Times Index
increased than by 45 points)
See the example below

L Lescourret 24
5.) Call and Price manipulations

Manipulation of benchmark prices

Gold Fix: a decade of banks manipulation?

The process is unregulated and the five banks can trade
gold and its derivatives throughout the call
Deutsche Bank, internal probe and Bafin
Barclays: 26m fine for attempting manipulation of the
London gold "fix

LIBOR rigging
Libor is an average interest rate calculated through
submissions of interest rates by major banks in London.
BBA involved
June 2012: Barclays fined $200m by the CFTC, $160m
by the United States Department of Justiceand 59.5m
by the Financial Services Authority for attempted
manipulation of the Libor and Euribor rates.
2014: NYSE Euronext took over running the Libor from
the BBA

FX rigging (rigged for five years, from January

2008 to October 2013)
Nov 12, 2014: banks handed 2.6bn in penalties for
market rigging + .
http://www.cftc.gov/ucm/groups/public/@newsroom/docume
nts/file/hsbcmisconduct111114.pdf

L Lescourret 25
5.) Call and Price manipulations

Market manipulations: Consequences

On the short run, price manipulation affects:

Pricing accuracy: price distortion at the close
Liquidity (decrease) and cost of trading
(increase)
Detrimental effect on the allocational role of
market prices

In the long run, it discourages participation and

causes investors to trade in alternative markets

L Lescourret 26
 6.)Limitations

Limitations of call auction markets

Immediacy need is not satisfied:

Waiting costs: news arrive continuously. Speculators
want ot trade on news. Hedging
One of the reasons why hybrid markets start trading
sessions with a call (fixing), and switch to a continuous
market : NYSE, Euronext, etc.

Performance:
Ambiguous result on the volatility : positive effect in
Paris, Nasdaq, but mixed in LSE and in the Australian
SE
Reasons
Algorithms + or efficient (see Comerton-Forde)
Traders modify their order submission strategies
(changes in the level of impatience : a call is shown to
reduce the aggressiveness of orders submission strategies,
which reduces the price volatility)

Failure of the Arizona Stock Exchange

Open book during preopening period and traders

communication
non-binding communication, last-mover advantage?
Pre-opening of Paris SE
Pre-opening of Nasdaq
L Lescourret 27
 ESSEC MSTF / FEA / MSc. in Finance

FINM 32417/32408
TRADING and MARKET ORGANIZATIONS

SESSION 3 : QUOTE-DRIVEN MARKETS

Class Handouts

Laurence Lescourret

L Lescourret 1
 Quote-driven markets

1. Introduction
2. Description
3. Role of financial intermediaries
4. Attributes
5. Example

L Lescourret 2
 Reminders Dealers Market

Continuous trading

Endogenous transaction time

Why trading continuously?

News arrive continuously. Speculators trade on
news.
If hedgers cannot trade continuously, their risk
increases.
Index membership: only stocks that trade
continuously can belong the index.

Among continuous markets, we make a distinction:

1. Dealer markets (quote-driven markets): OTC markets,
in general

2. Order book markets (order-driven markets) : Regulated

markets, in general, : NYSE-Euronext
(Archipelago/NSC), Tokyo Stock Exchange, Toronto
Stock Exchange, etc.

3. Floor markets: NYSE, AMEX, CME

L Lescourret 3
 1.) Introduction Dealers Market

Introduction

Mechanism: Counterparty = intermediaries

Role of intermediaries :
Dealers supply all liquidity.
Brokers choose which dealer they trade with.

Source: R. Schwartz

Examples : Over-the-Counter markets

Foreign Exchange (FX)
The debt markets: US governement and corporate
bonds, CDS, loans
SEAQ (London SE)

L Lescourret 4
 2.) Description Dealers Market

Main features

Dispersed network of dealers linked by automated

quotation systems, computers and telephones

Price competition among market-makers (minimum #

required on the Nasdaq was 3)
No price or time priorities enforced: customer limit
order are not well handled (and, thus, rarely used)
Trading between dealers: they lay off positions
(against other dealers) in an inter-dealer market

L Lescourret 5
 2.) Description Dealers Market

Main features

Even if dealers quotes are widely disseminated, bids

and asks are often indicative, and brokers/clients
need to phone to get quotes
The quotes are often oral and good only a time they
are made ( take it or leave it offer )
The dealer/client or dealer/broker relationship is
sustained by reputation.

Interdealer markets
A dealer making a market will build up a
suboptimal position over time (too long or too
short positions)
Dealers lay off or build position in the inter-dealer
market

Interdealer features
Not visible to non-dealers
3 mechanisms: non-anonymous bilateral
negotiation (dealer to dealer) ; anonymous trade
intermediated by a voice broker or interdealer
LOB

L Lescourret 6
 3.) Intermediaries Dealers Market

Participants: (1) Market-Makers/Dealers

Market-makers (dealer, broker-dealer)

They make a two-sided market: they quote a bid
and ask price, which may be firm (for specified
sizes). In that case, dealers must honor them.

Examples: SG Cowen, Spear, Leads & Kellog,

Goldman Sachs, etc

Roles
Liquidity / immediacy : A dealer supplies
liquidity on his own inventory
Immediacy for a small size order (200 shares) is not a
problem.
What about the immediacy of a large order (300,000
shares, or more) ?

Price Discovery

Other Services
Liquidity providers around corporate events (IPOs,
SEOs, Repurchase of cequity, etc.)
Research, clearing and settlement services, etc...

L Lescourret 7
 3.) Intermediaries Dealers Market

Dealers compensation

Sources of profit for dealers

Bid/ask spread:
Access to private information about : (i) the order
flow (temporary order imbalances, etc.) (ii) the type
of client (retail vs individuals)
timing the order flow
Lack of price-competition among dealers?

BID PRICE ASK PRICE

POSITION TO BUY POSITION TO SELL
MM1.25 25.25....MM4
MM2.24.875 25.5..MM5
MM3.24.875 25.5..MM3
MM524.875 25.5..MM2
MM4.24.875 25.5..MM1

Make-take spread: explicit trading fees

Sources of risk and therefore of costs

Adverse-selection cost
inventory holding cost

L Lescourret 8
 3.) Intermediaries Dealers Market

Participants contd: (2) Brokers

Brokers :
Trade on behalf of or for the account of (agency
trading)
no obligation to quote a price or to supply liquidity
obligation of best execution
Examples: Newedge, GFI, Lime Brokerage

Electronic brokers, Electronic Communication

Networks (ECNs)
Give access to an electronic order book
Examples: Inet (Instinet+Island), Chi-X, Virt-X,
etc
Success of ECNs
Systems based on the latest technologies and
algorithms (latency)
Commissions : in 1998, to execute 100 shares, the
brokers can charge up to 50$, Island charged
between 2.5$ and 5$, and paid if the investors
supplied liquidity.

L Lescourret 9
 3.) Intermediaries Dealers Market

Brokers and Best Execution Duty?

b/a spread/market spread, sometimes only indicative

BID PRICE ASK PRICE

MM1.25 25.25....MM4
MM2.24.875 25.5..ECN1
MM3.24.875 25.5..MM3
ECN124.875 25.5..MM2
MM4.24.875 25.5..MM1

Inside spread =
25 25.25

If indicative, search costs of the best price?

L Lescourret 10
 4.) Attributes Dealers Market

Search of the best price

Example: U.S. muni market

Source: Schrhoff, (2013)

L Lescourret 11
 4.) Attributes Dealers Market

Attributes

Advantages
No trading halts
Liquidity supplied even in the case of market
stress
Downward market,
Expiry of derivatives contracts,
Market opening and closing,
Block trading

Drawbacks
Fragmentation of the order flow (search costs of
the best price).
Transactions are usually the result of bilateral
negotiations over the phone

Quasi-absence of limit orders, which limits

competition among liquidity suppliers

Too opaque and too lightly-regulated markets?

The design of the optimal level of transparency is
an issue: collusion vs. Predatory trading.

Tacit collusion among dealers easier?

SEC investigation in september 2010 on the CDS
corporate market
L Lescourret
Nasdaq in 1994 12
 5.) Example 1: the Nasdaq Dealers Market

Opacity and collusion : NASDAQ in 1990s

Source : Christie et Schultz (1994), Journal of Finance. Why do mm avoid odd

eight?

1998: US Dpt of Justice et SEC fined 30 dealers ($1

billion)
L Lescourret 13
 5.) Example 2: Credit market Dealers Market

More Transparency? The U.S. corporate bond

market

2002 : TRACE implementation

FINRA required that bond dealers report all
transactions in publicly issued corporate bonds in
its TRACE system
Gradual implementation: from very large bonds in
2002 to all bonds in 2005 and from 75
minutes delay in 2002 to a 15-minute delay to
report. in 2005

Result: Transaction costs reduction: spreads reduced

(5 to 10 cents lesss, (Edwards, Harris et Piwowar,
2005).

Unintended consequence: [T]oo much price

transparency reduces dealers willingness to commit
capital.? (Bessembinder et Maxwell, 2008).

L Lescourret 14
 5.) Example 2: Credit market Dealers Market

More Transparency? The U.S. corporate bond

market (contd)
Unintended consequence:

L Lescourret 15
 5.) Example 3: Credit market Dealers Market

The CDS Market

Plain vanilla CDS : Genesis in the mid-1990s

(Blythe Masters, JP Morgan)
An instrument to transfer and manage credit risk, or to
take a position on the underlying asset.
a buyer pays a seller a premium, which is generally
expressed in basis points against a notional value (for
example, $10 million), in exchange for the right to sell the
underlying bonds at par in the event of default, or to
receive the cash equivalent.
A default is defined within the terms of a CDS contract,
and typically includes bankruptcy, failure to pay interest
or a material restructuring of the issuers obligations.
Traded in the OTC market:
Terms of the contracts privately negotiated ; absence of
transparency requirements ; lightly-regulated
One of fastest growing derivatives markets before the
2008-9 crisis:
Outstanding notional value increasing from $100 billion
in 1998 to $1 trillion in 2000 to $62 trillion in 2007.
Notional value declined below $30 trillion in 2009 as
industry initiatives netted gross notional outstanding
through portfolio compression and clearing.

Concentrated market : 14 (now 16) dealers (G14/G16)

hold 82% of the total notional amount:
Bank of America-Merrill Lynch, Barclays Capital, BNP
Paribas, Citi, Credit Suisse, Deutsche Bank, Goldman Sachs,
HSBC, JP Morgan, Morgan Stanley, RBS, Societe Generale,
UBS and Wells Fargo Bank.
L Lescourret 16
 5.) Example 3: Credit market Dealers Market
An opaque and lightly-regulated market until
the 2008-9 crisis

3 segments:
corporate single-name, sovereign single-name, and
index
Trades cluster
around a few typical sizes ($5 million for corporate
single-name for instance) and around 5-year maturity

Trade frequency in single-name ref. entities is

relatively low:
a majority of single-name trades less than once a day,
whereas the most active trades over 20 times per day.

60% of all transactions are trades between G16

dealers
L Lescourret 17
5.) Example 3 : Credit market Dealers Market

An opaque and lightly-regulated market

Opacity, financial crisis and counterparty risk:

Lehman Brothers bankruptcy ($400 billion of CDS
were presented for settlement) ; bailout of AIG
(huge exposure to CDS)
Risk of domino effects? Systemic risks?

Consequences of the financial crisis:

effort to strengthen regulation of global financial
markets (Dodd-Franck, EMIR)
Implementation of CPP
Implementation of SEF
> Futurization of CDX

Other usual suspects of opacity: insider trading in

the U.S.
very strong pattern of CDS spreads gapping out in
advance of debt issuance by large corporates.
difficult for SEC to prosecute insider trading in
the CDS market, since CDSs arent securities
give the SEC formal jurisdiction over single-name
CDS?

L Lescourret 18
 ESSEC MSTF / FEA / MSc. in Finance

FINM 32417/32408
TRADING and MARKET ORGANIZATIONS

SESSION 4: ORDER-DRIVEN MARKETS

Class Handouts

Laurence Lescourret

1
 Order-Driven Markets

1. Introduction

2. Mechanism

3. Stylized facts

4. Order books and algorithmic trading

5. Advantages

6. Limitations

L Lescourret 2
 1.) Introduction

Order-driven markets

1. Periodic Markets :
Example: Call/batch market (Fixing) :
Collection of orders and multilateral
transactions at a single price.

2. Continuous Markets directed by orders :

Trading mechanism used in most of the
worlds stock, futures, and options markets
Easy to set up
Cheap to operate
Intuitive to understand

L Lescourret 3
 2.) Mechanism

The basic order book market

No intermediary: direct matching between buy

and sell orders ;

Traders submit orders

The book is a collection of unexecuted/pending

orders
Buy orders (bids) and sell orders (ask) go into
separate collections.
Orders are sorted by price-time priority:
On the bid (buy_side), a higher bid has
priority over lower bids ; on the offer side,
a lower ask has priority over higher asks
(price priority)
When two orders have the same direction
and price, the order arrived first is first
executed (time priority).

Examples :
Euronext 100 (NSC), Deutsche Brse (Xetra),
London Stock Exchange (SETS), Inet (Nasdaq),
Tokyo SE, etc...
L Lescourret 4
 2.) Mechanism

Example

L Lescourret 5
 2.) Mechanism

Orders

Orders are instruction to trade that participants give

to brokers/exchange

Intructions are usually standardized to save time and

avoid mistakes

Orders always specify

Item to be traded
Sign : buy / sell
Size
They may also specify: Expiration, Counterparts,
Timing, etc.

L Lescourret 6
 2.) Mechanism

Limit Order (LO)

This order specifies a limit price condition:

Do not buy at a price above the prespecified limit
price
Do not sell at a price below the prespecified limit
price

Advantages
control of the execution price

Risks :
Execution uncertainty : might not be executed or
only partially;
Risk of being picked off (the Winners Curse
problem: the winner will tend to overpay) : poor
execution when prices move towards and through
the price limit, front-running and information
leakage.

Limit orders supply liquidity

Limit orders earn the b/a spread, when filled

Importance: 85%-95% of the orders passed on

LBO are LO.

L Lescourret 7
 2.) Mechanism

Market(able) Order

These orders demand liquidity in the form of

immediacy

They usually pay the b/a spread

Advantage :
Immediately and completely filled

Risks :
Execution price uncertainty (price impact)

L Lescourret 8
 2.) Mechanism

Other types of orders

Marketable order at the best limit

Trade at the best price currently available. Any
unfilled part of the order becomes a limit order
with a price limit equal to the price of the last trade
(limit order).

Hidden / iceberg order

Do not display the total quantity in the book

Stop Order
Price-contengent orders: they activate when their price
contingency is met
Almost always market orders

Kill or fill, etc.

L Lescourret 9
 2.) Mechanism

Order book and aggressive buy limit

order
Achat Vente Dernires Transactions
Ord. Qt Achat Vente Qt Ord. Qt cours
1 150 63,50 63,55 40 8 50 63,45
2 280 63,40 63,60 150 2 300 63,50
2 100 63,35 63,65 300 2 150 63,45
1 300 63,25 63,70 750 2 27 63,50
3 1 100 63,20 63,80 50 1 50 63,55
2 3 000 63,15 63,85 150 1 270 63,50

Sumission of a buy limit order of

500 shares at 63,65
Achat Vente Dernires Transactions
Ord. Qt Achat Vente Qt Ord. Qt cours
1 10 63,65 40 63,55
1 150 63,50 150 63,60
2 280 63,40 300 63,65
2 100 63,35 63,70 750 2 50 63,45
1 300 63,25 63,80 50 1 300 63,50
3 1 100 63,20 63,85 150 1 150 63,45

New limit order book snapshot

Achat Vente Dernires Transactions
Ord. Qt Achat Vente Qt Ord. Qt cours
1 10 63,65 63,70 750 2 40 63,55
1 150 63,50 63,80 50 1 150 63,60
2 280 63,40 63,85 150 1 300 63,65
2 100 63,35 63,90 500 2 50 63,45
1 300 63,25 64,00 800 5 300 63,50
3 1 100 63,20 64,10 1500 8 150 63,45

L Lescourret 10
 2.) Mechanism

Order book and market(able) limir order

Achat Vente Dernires Transactions

Ord. Qt Achat Vente Qt Ord. Qt cours
1 150 63,50 63,55 40 8 50 63,45
2 280 63,40 63,60 150 2 300 63,50
2 100 63,35 63,65 300 2 150 63,45
1 300 63,25 63,70 750 2 27 63,50
3 1 100 63,20 63,80 50 1 50 63,55
2 3 000 63,15 63,85 150 1 270 63,50

Submission of a market
order for 500 shares
Achat Vente Dernires Transactions
Ord. Qt Achat Vente Qt Ord. Qt cours
1 150 63,50 40 63,55
2 280 63,40 150 63,60
2 100 63,35 300 63,65
1 300 63,25 63,70 740 2 10 63,70
3 1 100 63,20 63,80 50 1 50 63,55
2 3 000 63,15 63,85 150 1 270 63,50

New limit order book snapshot

Achat Vente Dernires Transactions

Ord. Qt Achat Vente Qt Ord. Qt cours
1 150 63,50 63,70 740 2 40 63,55
2 280 63,40 63,80 50 1 150 63,60
2 100 63,35 63,85 150 1 300 63,65
1 300 63,25 63,90 500 2 10 63,70
3 1 100 63,20 64,00 800 5 50 63,55
2 3 000 63,15 64,10 1500 8 270 63,50

L Lescourret 11
 2.) Mechanism

Orders: Least to Most Aggressive

We classify orders by their position relative to current

market prices

The market is the range between the best (highest)

bid and the best (lowest) offer to sell;

Orders placed
Beyond the best bid or offer are away from the market
At the best bid or ask are at the market
Between the best bid and the best ask are in the
market

Examples:
Market(able) Order : Totally executed: the sell order
(resp. buy) walk the book
Marketable order at the best limit: complete fill if the
quantity available at the best limit is larger or equal.
Aggressive limit order :
Example: A buy order specifying a price higher than
that the best ask price in the book.

L Lescourret 12
 2.) Mechanism

Ecology of an order book

The offer of liquidity is produced by non-agressive

limit orders

With no limit orders, no book (remind that a book

order is simply a collection of limit orders that are
waiting to be executed.)

What are the risks ?

Not to be executed at all or being executed
against a better informed agent.
Limit orders : A free option ?

Advantages:
Win the spread + other compensations :
execution fees are less, even a payment is
repayed by the platform for every limit order
executed. (see graph 19)

Competition between the suppliers of liquidity

(perfect or strategic competition).

L Lescourret 13
 Order-Driven Markets

1. Introduction

2. Mechanism

3. Stylized facts

4. Order books and algorithmic trading

5. Advantages

6. Limitations

L Lescourret 14
 3.) Stylized facts

Clustering of orders and transactions

Biais et al. (1995)

L Lescourret 15
 3.) Stylized facts

Diagonal effect - Biais et al. (1995)

Limit order at or away from the quotes are

particularly frequent after limit orders on the same
side of the book
Competition to supply liquidity

large (small) trades on one side of the market are

most frequent after large (small) trades on the same
side of the market (positive autocorrelation of the
orders flow)
Strategic order splitting to reduce market impact,
or imitation ? Follow on strategy

Cancellations are particularly frequent after

cancellations on the same side of the book
Concern about adverse selection ?

Empirical findings on: Euronext, OMX, NYSE

(clustering of the activity vs no activity)

L Lescourret 16
 3.) Stylized facts

Price Formation - Biais et al. (1995)

Link between the type of orders submitted and

the state of the book

Determinants of the order flow and the

spread :
Aggressive orders are more frequent
when the spread is small, and LO are
inside the spread more frequently when
the spread is large.
Determinants of the order flow and the
depth :
LO inside the spread more frequently if
the depth at the best limit is high
enough. If not, LO queue at the best
limit if the depth is weak-> trade-off
between the execution price and the risk
of non-execution.
The orders are placed in a way to consume
liquidity when the book if full (thick book)
and to supply liquidity when the book is
drained (sparse book)
-> Resiliency of the order book

L Lescourret 17
 3.) Stylized facts
Price Formation (contd.)
Source: Biais et al. (1995)

L Lescourret 18
 Order-Driven Markets

1. Introduction

2. Mechanism

3. Stylized facts

4. Order books and algorithmic trading

5. Advantages

6. Limitations

L Lescourret 19
 4.) Order books and A.T.
Electronic trading, algorithmic trading
and high-frequency trading
Special care shoud be taken to distinguish:
1. Electronic trading refers to the ability to transmit
the orders electronically as opposed to telephone,
mail or in person

2. Algorithmic trading is more complex than electronic

trading and can refer to a variety of algorithms
spanning order-execution processes as well as high-
frequency portfolio allocation decisions

3. Algorithmic trading may

or may not be high-
frequency. Much of this
session is devoted to high-
frequency algorithms.

L Lescourret 20
 4.) Order books and A.T.

Algorithmic trading

Reasons to use algorithmic trading

L Lescourret 21
 4.) Order books and A.T.

Algorithmic algorithm

By asset classes

L Lescourret 22
 4.) Order books and A.T.

AT trend

23
 4.) Order books and A.T.

Definition and Characteristics

Algorithmic trading: automated, computer-based

execution of orders via DMA channels

Sometimes referred as rules-based trading:

if this, that and perhaps some other conditions, then I
want to submit this order, cancel that order, and so
on. (Schwarz, Francioni, and Weber)
Example: TraderExs machine-resident market-
makers: algorithmic traders who follow simple rules
based on their current inventory positions, relative to
their position limits, direction of the last trade, and the
quote changes of the other dealers. The rules specify
when they will match the inside bid or ask, initiate a
trade with another dealer

Algorithm determines the timing, price and quantity

of orders, by using a mix of active and passive
strategies (limit ordres vs marketable order): slice
and dice algorithm (ordres are separated by
thousands of a second)
Originally developed for use by the buy-side to manage
orders and to reduce market impact by optimising
trade execution

24
 4.) Order books and A.T.

High-Frequency trading: definition

and characteristics
High-frequency trading (HFT) is a subset of AT

HFT strategies are characterized by a high

number of trades (usually fairly small in size)
which are sent into the market at high speed,
with round-trip execution measured in
microseconds and a low average gain per trade.

Empirical evidence reveals that the average US

stocks is held for 22 seconds.

Absence of overnight positions to do away with

Overnight risk
Overnight carry rate

L Lescourret 25
 4.) Order books and A.T.

High-Frequency trading: Players

HF traders are mainly proprietary traders (use their

own capital for trading)

Most HF firms are based in New York, Connecticut,

London, Singapore, and Chicago.

The largest HF names are: Virtu, Optiver, IMC, DE

Sha, Worldquant, Renaissance Technonogies. Most
of HF firms are hedge funds or other proprietary
investment vehicles

L Lescourret 26
 4.) Order books and A.T.

How fast are high frequency traders?

It takes you 500,000 microseconds just to click a

mouse. But if youre a Wall Street algorithm and
youre 5 microseconds behind, youre a loser.
Kevin Slavin, TED Talk, Jul 2011
http://www.ted.com/talks/kevin_slavin_how_algorith
ms_shape_our_world.html

Spread Network (US Telecom Provider): Chicago-

New York roundtrip latency: of 12.98 milliseconds
 4.) Order books and A.T.

How being fast?

 4.) Order books and A.T.

High-Frequency trading: definition

and characteristics contd
HF trading strategies : Act both as market-makers and
as short-term speculators (SEC, 2010): seek to
benefit from market liquidity imbalances or other
short-term pricing inefficiencies

Liquidity-providing strategies: mimic the traditional

role of market makers aiming at profiting by
earning the b/a spread. Facilitated by maker/taker
pricing models

Statistical arbitrage strategies: based on correlated

prices between securities and seek to profit from
imbalances in those correlations.
Example: arbitrage between cross-border or domestic
marketplace to arbitrage between tradable index
(fsuture/basket of underlying stocks), cross-asset pairs
trading (arbitrage between a derivative and its
underlying)

Thousand other strategies based on technical

analysis, fundamental analysis, etc.

L Lescourret 29
 4.) Order books and A.T.

HFT trend

30
 4.) Order books and A.T.

HFT Revenues and Profitability

Electronic market-making
make / take spreads: rebates form a major
component of the revenue of electronic market-
makers
The rise of the machine: in 1997, Island: paid
liquidity providers 0.1 cents per share , while
charging those that took liquidity 0.25 cents par
share

Winner-takes-all market structure:

Revenues persistently and disproportionally
accumulates to the top performing HFTs

L Lescourret 31
 4.) Order books and A.T.

Impact ? : Number of orders and

algorithmic trading

Impact of slice and dice strategies

Block trading ?

Order / trade size ?

More odd-lots (<100): 4.5 %, contributes significantly to
price discovery (38%)

B/A spread: 90 bp in 1995, 3 bp in 2014

Duration of arbitrage opportunities: from 97

milliseconds on 2005 to 7 milliseconds in 2011
(no change in profitability)
L Lescourret 32
 4.) Order books and A.T.

Impact of HFT?

Nb of cancelled orders?
NASDAQ: trading frequency increased from
microseconds to nanoseconds the order
cancellation/execution ratio increased
dramatically from 26:1 to 32:1. ->no impact on
liquidity, price efficiency and trading volume,
but found evidence consistent with quote
stuffing
95% of orders cancelled?

Fleeting Orders
On August 30, 2011, about three-million orders were submitted to
the NASDAQ exchange to trade the stock SPDR S&P 500 Trust
(ticker symbol SPY). This image shows that 18.3 percent of the
orders were cancelled within one millisecond, and 42.5 percent of
orders had a lifespan of less than 50 milliseconds, less time than it
takes to transfer a signal between New York and California. More
than 40 percent of orders, in other words, disappeared before a
trader in California could react.

Source : Mao Ye

33
 4.) Order books and A.T.

Impact of HFT?

Impact on academic and regulator analysis?

Players in the trading game are superfast
computers. To study them you need the same
power.: raw data of a day can be as large as
ten gigabytes.
SECs Quantitative Analytics Unit (launched
in 2012): staffed by computer trading and
math experts (PhDs and extensive
backgrounds in mathematics, physics, and
computer science) who help gather data,
identify risks and target examinations of the
most sophisticated algorithmic trading firms
and other investment advisers.

34
 4.) Order books and A.T.

High-frequency trading: good? Or bad?

L Lescourret 35
 4.) Order books and A.T.

High-frequency trading: good? Or bad?

New risks? Risk of a codage error:

An infinite loop and the Facebook IPO (the 3rd
largest IPO in US industry) on the NASDAQ
exchange
After technical difficulties delayed the opening (30
minutes), a huge influx of orders to buy, sell and cancel
overwhelmed NASDAQs software, causing a 17-second
blackout in trading, and persistence of problems for hours
after opening it cost traders $100 million

Another infinite loop and the cancelled IPO

of BATSs
software bug affecting ticker symbols from A to
BFZZZ.
Another bug software and Knight-mare
Erroneous orders sent in the market during 30
minutes.
Error position that has to be sorted through the balance of
the day
$457.6 million loss for the company. It never recovered
and in December 2012 was acquired by GETCO.

Do HFT make markets more fragile?

Episodic crisis of illiquidity?

L Lescourret 36
 4.) Order books and A.T.

Manipulations

quote stuffing ?
it involves submitting an abnormally large number of
orders followed immediately by a cancellation (within
0.001 seconds or less) with the aim to generate order
congestion.

Layering:
remplir le ct dun carnet par couches (layers) pour
attirer les investisseurs sur ce mme ct et les surprendre
en excutant du ct oppos lorsque cest possible tout en
annulant les ordres qui ont aid remplir le carnet

Front-running?
HFT buy early access to public data, jumping in between
buyers and sellers who would have found each other
anyways in a few milliseconds
More adverse selection: HFT crowd out other traders (slow-
traders and also HFTs)

A new form of insider trading ?

new algorithms referred to as news aggregation search
the internet, news sites and social media (twitter) for
selected keywords, and fire off orders in milliseconds.
Trades are so quick, often before the information is widely
disseminated, that authorities are debating whether they
violate insider trading rules.
37
 4.) Order books and A.T.

Regulation and litigation

Sept 2012: SEC fined NYSE (NYSE had

delivered stock-price quotes and other data
to subscribers of two so-called proprietary
data feeds as much as several seconds before
the information was transmitted to the
broader market.)

How to curb erroneous orders, codage risk,

manipulation from HFTs?
New rules (15c3-5: Market Access Rule in the
U.S.)
HFT tax (France)
Speedbumps (IEX)
Fill rate (NASDAQ OMX)

38
 4.) Order books and A.T.

Flash Crash on May 6, 2010

DJIA: the biggest one-day point decline (998.5

points) of its entire history
E-mini S&P 500 futures price collapsed by 5% between
2:30 andd 2:45
Apple traded at $100,000 per share while Accenture traded
at a penny a share
Two simultaneous liquidity crises: one of the index
level in the E-Mini, the other with respect to
individual stocks

An extremely illiquid day, despite high volumes ->

high trading volume is not necessarily a reliable
indication of market liquidity - CFTC-SEC 39
 4.) Order books and A.T.

Hash Crash on April 23, 2013

Source: TABB Group

40
 4.) Order books and A.T.
Perspectives and Regulatory
aspects
New rules governing HFT servers?
Conclusions drawn by regulators following
the flash crash?
EU: the commission is reviewing the MiFID
framework directive: it intends to subject HFT
to MiFID requirements (risk management
obligations, capital requirements, market-
maker obligation such as order persistence and
tick sizes) and to supervision by a competent
authority
US: Dodd Frank Wall Street Reform and
Consumer Protection Act

New rules governing co-location of HFT

servers?
EU commission proposes that co-location
facilitites need to be offered on a non-
discriminatory basis

41
 4.) Order books and A.T.
Technology investment and
social cost: a waste of money?

42
 4.) Order books and A.T.

Proposed policy?

Changing the definition of designated market-maker

to include HFT market-makers
Advantage: two-sided quotes, etc.
Disadvantage: increase the cost of intermediaries,
which might lead to reduction of supply liquidity due to
the exit of some of them

Minimum quote life ? 50 milliseconds, suggested

several sources.
Quotes can currently be updated in the low-millisecond
range.
Problem, less limit orders? That would ultimately hurt
liquidity and widen spreads?

Tobin tax? (11 members of the EU, opposed to 15

members)
Will decrease HFT trading
but also will decrease overall market liquidity and
impair dynamic hedging strategies / migration of
trading activity to venues without tax (London)

43
 Order-Driven Markets

1. Introduction

2. Mechanism

3. Stylized facts

4. Order books and algorithmic trading

5. Advantages

6. Limitations

L Lescourret 44
 5. ) Advantages

Performance

Price/time priorities enforced

Electronic double discriminatory auction

(continuous): lower order-processing costs,
greater ability to disseminate information on
quotes/trades, level playing field, less adverse-
selection

A continuous trading protocol close to the

Walrassian ideal model ? (Biais, Foucault,
Salani , 1998) , fairness ?

Consolidation of the order flow (centralisation /

fragmentation)

Transparency ?

Price Discovery ?

L Lescourret 45
 6.) Limitations

Limitations

Block trades

Inactively traded stocks

There seems to be a critical mass of
trading interest required for a viable boo

Volatility, and trading halts

L Lescourret 46
 6.) Limitations
Inactively traded stocks: The book is
empty?

Carrefour (02/21/2006,
at 13h30)
on Euronext, and on Virt-X

L Lescourret 47
 6.) Limitations

Role of DMM

How to ensure a continuous narrow spread in a

book?

A designated market-maker
also called, liquidity provider on Euronext
formerly specialist on the NYSE

The DMM is in charge of

supporting the markets reputation as a place which is
liquid.
posting a (narrow) bid and ask, and other
responsibilities depending on the Exchange

How are they compensated?

Waive fees
Trading advantages

L Lescourret 48
 ESSEC MS

FINM 32408
Trading et Organisation de march

SEANCE 5: LIQUIDITY AND TRANSACTION COSTS

Laurence Lescourret

1
L Lescourret
 Session 5 Plan

1. Definition

2. Measures of Liquidity

3. Make-take spreads

4. Implementation shortfall

L Lescourret 2
 1.) Definition

Liquidity - Definition

Elusive concept: little agreement about its definition.

a market is liquid if traders can quickly buy or sell large
numbers of shares without large price effects.

4 dimensions
Width: bid-ask spread for a given number of shares and
commissions and fees to be paid per share.
Depth: number of shares that can be traded at given bid
and ask prices.
Immediacy: how quickly trades of a given size can be done
at a given cost.
Resiliency: how fast prices revert to former levels after
they changed in response to large order flow imbalances
initiated by uninformed traders.

It is clear that these different dimensions do not stand

independently on their own, but may interact with each
other.

L Lescourret 3
 1.) Definition

Recent Trends

L Lescourret 4
 1.) Definition

Transaction costs

Many methods: Direct methods or econometric

methods.

Methods using a benchmark price

The average of the bid and ask quotes (bid-ask
midpoint, BAM) prevailing at the time the decision to
buy/sell was made.
The bid-ask midpoint subsequent to the trades
The closing price of the previous session
The VWAP

2 kinds of liquidity measures: ex-ante measures (quoted

spread) and ex-post measures of liquidity (effective spread,
price impact, etc.).
The ex ante mesures give a commited or displayed
measure,
The ex post mesures, based on the prices of actual
transaction,take into consideration the hidden liquidity

L Lescourret 5
 2.) Measures of Liquidity

Remarks

Let us denote:
At= ask price at t
Bt = bid price at t
Pt = Transaction Price at t
Mt = midpoint at t

min Ai ,t max Bi ,t
Midpoint t M t i i
2

Data: to estimate liquidity and transaction costs, we need

data:
High-frequency data : intra-daily tick-by-tick data (TAQ,
NASTRAQ, Bloomberg, etc.)
Low-frequency data: daily data (CRSP, datastream)

L Lescourret 6
 2.) Measures of Liquidity

Ex-ante measures: the quoted spread

The Quoted spread : cost of a rount-trip transaction

usually expressed in percentage (RQS)

Quoted spreadt QSt Best Ask t - Best Bidt min Ai ,t max Bi ,t

i i
min Ai ,t max Bi ,t
Relative Quoted spreadt RQSt i i
Mt

In order-driven markets:
Difference between the best buy limit order and the
best sell limit order in the book

In quote-driven markets:
Difference between the best ask (min Ai) quoted by (at
least) 1 dealer, and the best bid price (max Bi) quoted
by another dealer.

Quoted spread and the size of the tick:

For the very liquid markets: quoted spread = minimum
tick size (the tick size is binding)

Drawbacks:
It may vary over the day (U-shaped)
Valid for small volumes only
L Lescourret 7
 2.) Measures of Liquidity

Example

Quoted
Spread

Market order to buy 1 000 shares of

XYZ
Best Ask: 65.22
Best Bid: 65.18
Quoted Spread:
0.04
6 bp

Source: I. Werner
L Lescourret 8
 2.) Measures of Liquidity

Examples:

The debt markets

Single-name CDS market: the cost of a round-trip for
an investment in a European 5 yrs CDS for a face value
of 100 $ is, in average, 32 cents (Fulop et Lescourret,
2007)
Euro-denominated corporate bond market: average
quoted spread = 30 cents (Biais and Declerck, 2007)
US corporate bond market : average quoted spread: 20
cents (Edwards et al., )

Euronext
For the most liquid stocks (ABN Amro, ING Groep Dm,
Fortis, Dexia, or Alcatel): quoted spread=1,2 cents () in
2007.

L Lescourret 9
 2.) Measures of Liquidity

Ex ante Measures : The depth

Depth

Volume(a t ) Volume(b t )
Depth t
2

(a )Volume(a t ) (b t ) Volume(b t )
Depth t at /$ t

2
where

Arbitrage between the tick and the size of spread :

the role of the tick size.
Decimalization in U.S. has reduced spreads and
reduced depth (i.e., increased price impact).
Good for retail investors, bad for institutional investors.
Consequences: Need to break-up orders across
platforms: Average trade size on NYSE 160, down from
522 shares (2003) and 785 shares (2002).

L Lescourret 10
 2.) Measures of Liquidity

Ex post Measures : Effective Spread

Effective Costs :
Effective Costs Qt Dt (Pt - Mt)
with Dt 1 in case of a buy order and Dt -1 in case of a sell

Effective Spread :

Effective Spread t ESt 2 Dt ( Pt - M t )

Dt ( Pt - M t )
Relative Effective Spread RES t 2
Mt
with Dt 1 in case of a buy order and Dt 1 in case of a sale ordre de vente.

Trading costs for investors

L Lescourret 11
 2.) Measures of Liquidity

Example

Execution Price

The small size orders can be executed at better

prices than the best quoted price (here, best ask):
non-displayed liquidity
Price improvement: 0.01
Effective spread: 2*(Pt-Mt)=0.02 or RES= 3bp

Source: I. Werner 12
L Lescourret
 2.) Measures of Liquidity
Ex post Measures ex post: Realized
Spread

Realized costs

Realized Costs Qt Dt (Pt - Mt )

Realized spread

RS t 2 Dt ( Pt M t )
with t=5, 10, 30, 60 min, and Dt=+1 in case of a buy order and
-1 in case of a sale order.

The realized spread is smaller than the quoted spread.

This is a post-trade benchmark

This cost can be interpreted as the profit realized by the LS
that was the (contra) side of the trade, assuming that she
could lay off the position at the new BAM.

L Lescourret 13
 2.) Measures of Liquidity

Ex post Measures : Price impact

Measure of the informational content of the order

flow
Effective Spread t Realized Spread t
Price Impact t
2
or
Price Impact t Dt M t M t

Measure the Amihud s illiquidity (2002) :

Daily price response associated with 1$ of trading
volume: rough measure of price impact.
1 R yd
ILLIQ
D y VolDvyd

L Lescourret 14
 2.) Measures of Liquidity

Resiliency

How quickly prices revert to former level after they

change in response to large order flow imbalances
initiated by uninformed traders Harris, 1990.

Depend on (Foucault, Kadan, Kandel, 2005):

Tick size (the smaller, the stronger the resilience)
The market capitalization (the bigger the market cap,
the stronger the resilience)

L Lescourret 15
 2.) Measures of Liquidity

Examples

Euronext (Degryse et al, 2005): 20 messages (order

submission/modification/cancellation) are needed to
revert to the former level

LSE (Large, 2007) : The book replenishes fully in less

than 20 seconds but only with 40% of the so-called
liquid stocks

Flash crash, May 6 2010

L Lescourret 16
 2.) Measures of Liquidity

Costs of Transaction and Size of Orders

Estimation of the costs of transactionfor the small

size orders
Easy to measure, to interpret and to predict
Measures: Commissions, quoted spread, effective
spread

Estimation of the costs of transactionfor the big size

orders
Hard to measure, to interpret and to predict : non-
displayed liquidity.
Mesures: ?

More than 70% of the trading in developped markets is

realised by institutionals who send block orders >10K
shares.
The quoted depth cannot satisfy their needs of
immediate liquidity.
Displaying large order will result in adverse price
impact. How do you tease out non-displayed liquidity
without incurring price impact?

L Lescourret 17
 2.) Measures of Liquidity

Example: Price impact of a large order

Large buy order for 600,000 shares of XYZ

Walk up the book using displayed liquidity
Buy 1600 @ 65.22, 2200 @ 65.23, etc.
Expensive: price impact
Solutions: Access to non-displayed liquidity and
work order over time
shop the order
split the order
send to dark pools
Send to brokers to work over the day 18
L Lescourret
 2.) Measures of Liquidity

The problem of the blocs

Latent demand problem

Order exposure problem
Price discrimination problem

Asymmetric information problem

How convince a counterpart that the order is
uninformed ? Sunshine trading (Admati and Pfleiderer,
1994)
Price reaction to block trades: the empirical literature
shows that block trades move prices:

L Lescourret 19
 2.) Measures of Liquidity
Reaction of institutional investors to
changing landscape

L Lescourret 20
 2.) Measures of Liquidity

Blocks and market organizations

Theoretical models and common beliefs suggest that

large trades should pay a higher transaction cost:
Inventory risks are larger
Adverse selection risks ? (price reaction to block trades
vs algorithmic trading)
This is indeed the case for NYSE data, Paris Bourse
data.
However,
large trades in London get significant price
improvement (Reiss and Werner, 1996), which are
based on reputation/relationships (Bernardt et al. 2003)
Transaction costs signficantly decline with trade size in
the US corporate bond market (01/03-01/05, Edwards et
al.,2005)

L Lescourret 21
 2.) Measures of Liquidity

Direct Costs : The pricing lists, or

Make/take spread

http://haimbodek.com/electronic_markets.html

L Lescourret 22
 2.) Measures of Liquidity

Stakes of the pricing list

The difference between the direct costs of suppliers

vs seekers of liquidity, the make/take spread is
not neutral, because :

1. It affects the strategies of the book surveillance

2. At the end, it affects the traded volume.

More precisely, Foucault et al. show that the

make/take spread increases with

i. The size of the tick

ii. The ratio between the suppliers and seekers of

liquidity

iii. The ratio of the book surveillance costs between

the suppliers and seekers of liquidity.

L Lescourret 23
 2.) Measures of Liquidity

Stake?: Regulation

Determination of the best execution among the

platforms :
Proliferation of the platforms: when to execute and
why?
Regulation :
In the US (Reg NMS) : the only price criteria (no-
trade trough rule) ; SEC Rules 11Ac5 and 6
In Europe (Mifid) : multi-criteria approach (price,
speed of execution, costs, probability of excution
and settlement , size and nature of the order)

L Lescourret 24
 2.) Measures of Liquidity
Importance of the measure of the costs
of transaction

Plexus Group estimate that of the research and the

portfolio management are lost in the costs of
transactions

Implementation shortfall: developped by Plexus

Group

L Lescourret 25
 2.) Measures of Liquidity

Implementation shortfall

Prix benchmark

L Lescourret 26