1

Will the Real Market Failure Please Stand Up?

Chairman Schapiro on September 7 •  The Flash Crash is “clearly a market failure”

An Internet search produces 767,000 references to market failure •  A condition that arises when unrestrained operation of markets yields socially

undesirable results

The story that was (and to some extent still is) being crafted • Market structure itself is the real market failure

Why does the Flash Crash live on? •  It has changed the entire tenor of the regulatory debate concerning market structure

and behavior as conditioned by structure •  The Flash Crash is used as motivation multiple times in an address by Chairman

Schapiro as late as March 23 of this year

2

The Flash Crash of May 28

A bad morning

The market drops sharply in the afternoon, then rebounds

Many stocks decline by over nine percent in less than 12 minutes

Full investigation, but the cause remains undetermined

What did market structure look like?

What was the role of market makers, and how did they react?

How did market and stop-loss orders play into the crash?

Speed, speed and more of the same…

Investor behavior and changes in the impact of normal market mechanisms, with an eye towards fair and orderly markets

Price discovery

3

ETF Volatility

Market  Plunge  on  May  6,  2010:    14:40  -­‐  15:00Average  15  Second  Midquote  Volatility  

0

50

100

150

200

250

14:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

6-­‐May 7-­‐May Avg  5/3-­‐5/5

4

Message Traffic Signals

Market  Plunge  on  May  6,  2010:    14:40-­‐  15:00Average  15  Second  Buy  Limit  Order  Cancellations  /  Modifications  

0

500

1000

1500

2000

2500

14:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

Order  Cancellatio

ns/M

odificatio

ns

Avg  5/3-­‐5/5 6-­‐May 7-­‐May

5

What Order Flow Went Along with the Message Traffic?

Market  Plunge  on  May  6,  2010:  14:40  -­‐  15:00  Average  15  Second    Number  of  Fleeting  Orders

0

2,000

4,000

6,000

8,000

10,000

12,000

14:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

Num

ber  of  Fleeting  Orders

Avg  5/3-­‐5/5 6-­‐May 7-­‐May

6

A Lagging Indicator of Liquidity Provision

Market  Plunge  on  May  6,  2010:  14:40  -­‐  15:00  Average  15  Second  TWA  Spread

0

100

200

300

400

500

600

700

14:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

TWA  Spread  (in  bp

s)

Avg  5/3-­‐5/5 6-­‐May 7-­‐May

7

ETF Depth at the Top of the Book

Market  Plunge  on  May  6,  2010:  14:40  -­‐  15:00    Average  15  Second  Bid  Depth  Sizes

0

10000

20000

30000

40000

50000

60000

14:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

Average  Bid  De

pth  Size  (in  shares)

Avg  5/3-­‐5/5 6-­‐May 7-­‐May

8

ETF Depth Across Ten Price Levels

Market  Plunge  on  May  6,  2010:  14:40  -­‐  15:00  Average  15  Second    TWA  Bid  Depth  Size  Up  to  Level  10

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,00014

:40

14:41

14:42

14:43

14:44

14:45

14:46

14:47

14:48

14:49

14:50

14:51

14:52

14:53

14:54

14:55

14:56

14:57

14:58

14:59

15:00

Bid  De

pth  Size  Up  to  Level  10

Avg  5/3-­‐5/5 40,304 40,305

9

S&P 500 Growth Index Fund

IVW Vs. Portfolio: Bid Depth Level 1

0

0.5

1

1.5

2

2.5

3

3.5

4

14:35 14:37 14:39 14:41 14:43 14:45 14:47 14:49 14:51 14:53 14:55 14:57 14:59 15:01 15:03 15:05

Liqu

idut

y - A

sk_D

epth

_Lev

el 1

Pct

Cha

nge

-80

-70

-60

-50

-40

-30

-20

-10

0

Cumulative Return Pct.

Ind IVW Bid_1 Ind Portfolio Bid_1 IVW CumRet Portfolio CumRet

10

S&P 500 Growth Index Fund

IVW Vs. Portfolio: Bid Depth Level 1-10

0

0.2

0.4

0.6

0.8

1

1.2

1.4

14:35 14:37 14:39 14:41 14:43 14:45 14:47 14:49 14:51 14:53 14:55 14:57 14:59 15:01 15:03 15:05

Liqu

idut

y - A

sk_D

epth

_Lev

el 1

Inde

x Ch

ange

-80

-70

-60

-50

-40

-30

-20

-10

0

Cumulative Return Pct.

Ind IVW Bid_10 Ind Portfolio Bid_10 IVW CumRet Portfolio CumRet

11

ETFs with Broken Trades

Affected ETF's Vs. Aggregated Portfolio: Ask Depth Level 1-10

0

0.2

0.4

0.6

0.8

1

1.2

1.4

14:35 14:37 14:39 14:41 14:43 14:45 14:47 14:49 14:51 14:53 14:55 14:57 14:59 15:01 15:03 15:05

Liqu

idity

- Bi

d_De

pth_

Leve

l 1-1

0 In

dex

Cha

nge

-70

-60

-50

-40

-30

-20

-10

0

Cumulative Return Pct.

Indx-AffectedETF's-Ask Depth10 Indx-Affected Portfolio-AskDepth10AffectedETF's-CumRet Affected Portfolio-CumRet

12

ETFs without Trade Breaks

Unaffected ETF's Vs. Aggregated Portfolio: Ask Depth Level 1-10

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

14:35 14:37 14:39 14:41 14:43 14:45 14:47 14:49 14:51 14:53 14:55 14:57 14:59 15:01 15:03 15:05

Liqu

idity

- Bi

d_De

pth_

Leve

l 1-1

0 In

dex

Cha

nge

-12

-10

-8

-6

-4

-2

0

Cumulative Return Pct.

Indx-GoodETF-Ask Depth10 Indx-UnAffected Portfolio-AskDepth10 Good ETF-CumRet UnAffected Portfolio-CumRet

13

Was the Liquidity Issue Limited to ETFs?

The bulk of the final SEC report is summarized by a few quotes

•  “the largest and most erratic price moves were caused by withdrawals of liquidity and the subsequent execution of trades at stub quotes”

•  “liquidity had virtually evaporated” (repeated in different contexts)

•  “drops in price become increasingly more severe with ever-larger drops in liquidity”

•  All in all, “what happened is best described in terms of two liquidity crises—one at the broad index level in the E-Mini, the other with respect to individual stocks”

From market structure to behavior, and onwards to algorithms

14

The New Algorithmic Trading Debate

2:32 ET on May 6

An unpriced order was submitted to a volume participation algorithm

35,000 contracts sold as the contract went from 1130 to 1056

Five second pause by the CME

Futures go back to previous level

40,000 more contracts sold

Total elapsed time: 20 minutes

15

Volume Participation, Volatility, and Size

0.1

0.5

0.9

1.3

1.7

0.0

2.4

4.8

7.2

9.6

Intraday volatility (%)Relative size (%)

Volume Participation strategy

Strategy-specific cost (bps)

Strategy conditional impact vs. intraday volatility(relative order size set to the mean)

0.1 0.4 0.7 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3.1 3.4 3.7 4.0 4.3 4.6 4.9 5.2

Intraday volatility (%)

Relat

ive co

nditio

nal im

pact

(bps)

ActiveDarkISVWAPVP

16

Participation Algorithms for Small Orders

17

Participation Algorithms for Large Orders

18

If We Have Time for the Regulators…

Regulation of trading strategies Stub Quotes Regulatory Constraints on Brokers Enhanced Data Transparency Circuit Breakers Automated Review Policies

Disclaimer

These materials are for informational purposes only, and are not intended to be used for trading or investment purposes. The information contained herein has been taken from trade and statistical services and other sources we deem reliable but we do not represent that such information is accurate or complete and it should not be relied upon as such. No guarantee or warranty is made as to the reasonableness of the assumptions or the accuracy of the models or market data used by ITG. These materials do not provide any form of advice (investment, tax or legal).