1) The document discusses how to extract the best liquidity from dark pools by understanding the different types of participants and separating them, measuring performance across dark pools and participants, and optimally allocating orders. 2) It emphasizes the importance of measuring total implementation shortfall rather than just fill rates, as higher adverse selection can negate good execution prices. 3) ITG's POSIT Marketplace and Liquidity Filter technology aims to separate "toxic" from "non-toxic" participants, measure dark pool performance, and allocate orders to the highest quality pools and participants to minimize total implementation shortfall.

1. Rising to the Top:
Finding the Best Liquidity in Dark Pools
Hitesh Mittal, Managing Director
Investment Technology Group
© 2009 Investment Technology Group, Inc. All rights reserved. Not to be reproduced or retransmitted without permission. Broker­dealer
products and services offered by ITG Inc., member FINRA, SIPC. POSIT Marketplace accesses, among other destinations, the POSIT Now and
POSIT Match ATSs. 041709­40099

2. Dark Pools Are a ZERO SUM Game
• Traditional dark pools invited like­minded participants
• The objective was to minimize impact
• Newer entrants tend to invite a wider variety of participants with
different objectives:
• Institutions – minimize impact
• DMA flow – reduce cost
• Market makers, prop desks, and high frequency funds – earn short­
term alpha
• Dark pools are a ‘zero sum game’ by construction
• The objectives of newer entrants make them less attractive to
institutions
1

3. How to Extract Quality Liquidity
1 2 3 4
Optimally allocate to
Understand Separate Measure performance
minimize
contra­types in contra­types in and contra­types
Implementation
dark pools dark pools in dark pools
Shortfall
2

4. Understanding Contra­Types
Minimizes
Institutional
Impact
Block Order
May Support Stocks Institutional
Market Bound
No Spread Savings High Frequency
Adverse Selection Market Makers
Gaming Gamers
*Based on internal ITG research: “Are You Playing in a Toxic Dark Pool? A Guide to Preventing Information Leakage,”
Hitesh Mittal, Journal of Trading, Summer 2008.
3

5. Separating Contra­Types
ITG’s Liquidity Filter technology uses this logic to interact
intelligently against each type of liquidity
Uses minimum size on orders
Uses order type Peg Passive without min size
Large Client
Order
Shuts down IOIs by sending small aggressive
orders
Avoids gamers by not allowing orders to
get pinged
4

6. Understanding Implementation Shortfall
Example: Buy 100k ABC
Executed shares: 35k
Unexecuted shares: 65k
­7 bps Opportunity Cost
IS of Executed Opportunity Total
Shares Cost Implementation
Shortfall
­3bps ­7bps 5.6bps
­3bps IS of executed shares
Total Implementation Shortfall = execution cost vs. arrival +
potential cost of executing the residual order based on the current
stock price
5

7. How to Measure Performance of a Dark Pool
• Implementation Shortfall of executed shares
• Fill rate
• Adverse selection
• Opportunity Cost: the cost of trading unexecuted shares
• Implementation Shortfall + Opportunity Cost
6

8. Looking at Fill Rates Alone is Not Enough
15 bps Example: 6 Buy Orders
10 bps § Fill rate: 50% Great!!
§ IS cost on fills: Negative Great!!
5 bps
Price § Opportunity Cost: High Not Great
Time
§ Total IS cost:
­5 bps
(IS Cost + Opportunity Cost) High Not Great
­10 bps
­15 bps
7

9. Looking at Fill Rates Alone is Not Enough
• Fill rate: 25% OK
• IS cost on fills: 0 OK
Price
Time • Opportunity Cost: 0 Great!!
• Total IS Cost:
(IS Cost + Opportunity Cost) 0 Great!!
Higher adverse selection = Higher opportunity cost
8

10. 9
­140
­120
­100
­80
­60
­40
­20
0
20
40
60
­5
0
5
­30
­25
­20
­15
­10
10
15
ATS 1 ATS 1
ATS 2 ATS 2
ATS 3 ATS 3
ATS 4
ATS 4
ATS 5
ATS 5
ATS 6 ATS 6
ATS 7 ATS 7
ATS 8 ATS 8
ATS 9 ATS 9
ATS 10
ATS 10
ATS 11
ATS 11
ATS 12
ATS 12
ATS 13
ATS 13
ATS 14 ATS 14
ATS 15 ATS 15
ATS 16 ATS 16
ATS 17
ATS 17
Adverse Selection End Time +1 min
ATS 18 ATS 18
Adverse Selection End Time +15 min
ATS 19 ATS 19
ATS 20
Adverse Selection Across Dark Pools
ATS 20
ATS 21 ATS 21
ATS 22 ATS 22
ATS 23
ATS 23
ATS 24
ATS 24
ATS 25
ATS 25

11. Optimally Allocate for Best Performance
• Spreading orders across multiple pools leads to more information
leakage
• Higher information leakage = Higher IS cost
• Allocation should be adjusted based on
• Adverse selection
• Average impact in each pool
• Average fill rate in each pool
10

12. Introducing POSIT Marketplace
SM
11

13. ITG’s Liquidity Filter
SM
At the Core of POSIT Marketplace
Separate toxic and Measure the
non­toxic contras performance of each
in dark pools dark pool and contra
Measure total IS of Allocate first to
orders and adjust highest quality dark
allocation pools and contras
12

14. Disclaimer
These materials do not provide any form of advice (investment, tax or legal). Neither Investment
Technology Group, Inc. nor ITG Inc. is a registered investment adviser and they do not provide investment
advice or recommendations to buy or sell securities, to hire any investment adviser or to pursue any
investment or trading strategy.
The screen shots provided in this presentation represent examples of certain products available from
ITG. They contain sample data and are provided for informational purposes only. All functionality and
delivery estimates described herein are subject to change without notice.
These materials are highly confidential and are not to be copied, displayed or transmitted in any form
without the prior written permission of ITG.
13