CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

Joseph Cherian

Professor of Finance (Practice)

Director, Centre for Asset Management Research & Investments (CAMRI)

NUS Business School

Scientific Advisory Board

Orissa Group Inc.

Liquidity Risk data kindly provided by Orissa Group, Inc. (OGI, Inc.)

Liquidity Risk Management for Portfolios

IPARM SEA 2011August 17, 2011

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

2

A few opening claims

Liquidity risk is well-studied but still an “elusive concept”

Recent experience suggests that it is central to asset pricing and risk management

Liquidity risk drives security prices away from fundamentals

In other words, markets are not efficient in pricing liquidity risk and hence it presents trading opportunities if exploited properly

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

3

What we wish to achieve today

Establish a practical understanding of liquidity risk

Introduce extant and new empirical metrics to estimate liquidity risk (or scores) using intraday data

Introduce simple equity trading strategies to exploit liquidity risk

Identify liquidity regimes in markets

Compare US versus Asian equity markets

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

4

Understanding Liquidity and Liquidity Risk

Definition

Liquidity is the ease of trading a security

Liquidity Risk is the uncertainty associated with liquidity

Other Definitions

Ease of availability of financing for very short term maturities

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

5

A few observations

Liquidity is not a fixed property

Liquidity can suddenly dry up

Liquidity influences asset returns

Liquidity is a significant source of risk

Size and trading volume are insufficient proxies of liquidity

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

6

How to measure Liquidity and Liquidity Risk?

First Step: Estimate the cost of liquidating positions (or illiquidity)

Measured as the magnitude of price movements (volume-weighted returns) resulting from order size (dollar volume) – Amihud [2002]

Modeled using intraday trading data for stock “i” over intraday time interval “t” and in month “m”, appropriately normalized:

Second Step: Estimate the uncertainty in the cost

Formulate a time-series model of illiquidity

Estimate liquidity risk as the illiquidity shock – Amihud [2002]

𝐼𝐿𝐿𝐼𝑄𝑡 ,𝑚𝑖 =

|| ,

i

mtr

i

mtV ,

×0CPI

CPI t

ILLIQt = a + b* ILLIQt-1 + t

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

7

Properties of illiquid portfolios

Largest 3000 US stocks by market capitalization

Illiquidity

Portfolio

Next month

ret

Log dollar

trade

volume

Log

market

cap

*ILLIQ

COST

Market

Beta

Size

Beta

Valuation

Beta

1 0.96% 20.75 22.64 0.05% 1.03 0.18 -0.06

2 0.95% 19.15 20.97 0.34% 1.15 0.58 0.04

3 1.06% 18.17 20.16 1.01% 1.19 0.83 0.12

4 1.14% 17.23 19.55 2.26% 1.26 0.86 0.23

5 1.15% 16.25 19.11 4.19% 1.27 0.79 0.42

Table 2: Properties of Illiquidity Portfolios. This table reports the properties of 5 portfolios sorted using

*ILLIQ. Portfolio 1 has the lowest illiquidity and Portfolio 5 has the highest illiquidity. The portfolios are

formed at the end of each month from a universe of 3000 largest US stocks by average market

capitalization for the month. Market, HML and SMB beta are computed using contemporaneous monthly

regressions of excess portfolio returns with Fama-French factors for Market (Rm_minus_Rf), Size (SMB)

and Valuation (HML). All values are reported as monthly averages for the period 1993-2009.

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

8

Liquidation cost (Market Illiquidity Level – MIL)

Largest 3000 US stocks by market capitalization

Cost of trading a USD 10 Million position in a day

US Median Liquidation Cost

0%

1%

2%

3%

4%

5%

6%

Jan-9

3

Jan-9

4

Jan-9

5

Jan-9

6

Jan-9

7

Jan-9

8

Jan-9

9

Jan-0

0

Jan-0

1

Jan-0

2

Jan-0

3

Jan-0

4

Jan-0

5

Jan-0

6

Jan-0

7

Jan-0

8

Jan-0

9

Jan-1

0

CreditLTCM

(MIL)

Note: MIL and other acronyms / variables defined in Appendix

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

9

Early warning indicator: liquidity deterioration

Increase in Liquidation Cost (Capital Markets vs. Market Median) Jan-06 - Jun 07: 22% vs. -23%

Jun 07 - Jul 08: 253% vs. 132%

Market Universe: US Largest 3000 stocks by market capitalization

Liquidation Cost: Capital Market vs. Market Average

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

Jan-0

6

Apr-

06

Jul-06

Oct-

06

Jan-0

7

Apr-

07

Jul-07

Oct-

07

Jan-0

8

Apr-

08

Jul-08

Oct-

08

Jan-0

9

Apr-

09

Jul-09

Oct-

09

Jan-1

0

Capital Markets Market Aggregate

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

10

Liquidity premium – Market Illiquidity Factor

Market Illiquidity Factor (MIF) measures how liquidity risk is priced by market participants.

0

20

40

60

80

100

120

140

160

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

Market Illiquidity Factor (MIF)

LTCM Crisis, Aug 1998

Lehman Bros Bankruptcy, Sep 2008

Subprime Crisis,

Cumulative return of illiquid securities relative to liquid securities

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

11

Liquidation regimes - concentrated during crisis

1.0%

1.1%

1.2%

1.3%

1.4%

1.5%

1.6%

90

100

110

120

130

140

150

160A

pr-

2009

Ma

y-2

009

Jun-2

009

Jul-2

009

Aug-2

009

Sep-2

009

MIL

MIF

Benign Liquidity Regime

MIF MIL

1.2%

1.4%

1.6%

1.8%

2.0%

2.2%

2.4%

2.6%

90

92

94

96

98

100

102

104

106

108

110

Apr-

1998

Ma

y-1

998

Jun-1

998

Jul-1

998

Aug-1

998

Sep-1

998

MIL

MIF

Liquidity Crisis Regime

MIF MIL

0.9%

1.0%

1.1%

1.2%

1.3%

1.4%

1.5%

1.6%

90

92

94

96

98

100

102

104

106

108

110

Jul-2

008

Aug-2

008

Sep-2

008

MIL

MIF

De-Leveraging Regime

MIF MIL

1.1%

1.2%

1.3%

1.4%

1.5%

1.6%

1.7%

1.8%

1.9%

2.0%

85

87

89

91

93

95

97

99

De

c-20

08

Jan-2

009

Feb

-2009

MIL

MIF

Liquidity Correction Regime

MIF MIL

(Flight-to-liquidity)

(Fight-for-liquidity)

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

US Market Illiquidity FactorTM

(MIF)

125

130

135

140

145

150

155

160

165

170

175

180

185

Se

p-0

7

No

v-0

7

Ja

n-0

8

Ma

r-0

8

Ma

y-0

8

Ju

l-0

8

Se

p-0

8

No

v-0

8

Ja

n-0

9

Ma

r-0

9

Ma

y-0

9

Ju

l-0

9

Se

p-0

9

No

v-0

9

Ja

n-1

0

Iilliq

uid

s O

utp

erf

orm

Liquidity deteriorates

Illiquids outperform

("deleveraging" regime)

Liquidity improves

Illiquids underperform

("liquidity-correction" regime)

Liquidity improves

Illiquids outperform

("benign" regime)

Liquidity deteriorates

Illiquids underperform

("flight-for-liquidity" regime)

US Median Liquidation Cost

0%

1%

2%

3%

4%

5%

6%

Se

p-0

7

No

v-0

7

Ja

n-0

8

Ma

r-0

8

Ma

y-0

8

Ju

l-0

8

Se

p-0

8

No

v-0

8

Ja

n-0

9

Ma

r-0

9

Ma

y-0

9

Ju

l-0

9

Se

p-0

9

No

v-0

9

Ja

n-1

0

Liq

uid

ity d

ete

rio

rate

s

12

Liquidity regimes

Liquidity Premium

Liq

uid

ity

Leve

l

Illiquids Underperform

Level Deteriorating

"flight-for-liquidity"

Abnormal

Illiquids Outperform

Level Deteriorating

"deleveraging"

Illiquids Outperform

Level Improving

"benign"

Abnormal

Illiquids Underperform

Level Improving

"liquidity-correction"

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

13

Applications in active portfolio management: Liquidity

analysis presents alpha generation opportunities

When Market Illiquidity Level increases (i.e., as liquidity deteriorates)

Investors favor liquid securities over illiquid securities

US: Russell 2000 (proxy for illiquid securities) underperforms DJ Industrial Average (proxy for liquid securities)

When Market Illiquidity Level decreases (i.e., as liquidity improves)

Market participants favor illiquid securities over liquid securities

US: Russell 2000 outperforms DJ Industrial Average

Analysis period: Jan 1, 1993 through Nov 20, 2009

Trailing Liquidity

# of

weeks

Russell 2000

(RUT) DJIA (DJI)

Russell 2000

(RUT) DJIA (DJI)

Return Return Std Dev Std Dev

Deteriorating 384 -3.6% 7.6% 25.2% 19.8%Improving 484 17.6% 9.0% 17.8% 15.2%

Return Std Dev

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

14

Applications in active portfolio management: Liquidity

Analysis presents alpha generation opportunities

When Market Liquidity deteriorates

Short Russell 2000 (RUT)

Long Dow Jones Industrial Average (DJI)

When Market Liquidity improves

Long Russell 2000 (RUT)

Short Dow Jones Industrial Average (DJI)

Weekly rebalancing 0%

2%

4%

6%

8%

10%

12%

Lo

ng

DJI

Sh

ort

Ru

ssell 2

000

exclu

de

cu

rr w

eek

exclu

de last

2 w

eeks

exclu

de last

3 w

eeks

exclu

de last

4 w

eeks

exclu

de last

5 w

eeks

exclu

de last

6 w

eeks

We

ek

ly R

etu

rn (

An

nu

alize

d) Outperforms

Naïve Long DJI / Short RUT strategy

HFR Equity Market Neutral strategy

The liquidity based trading signal is persistent

Results does not consider transaction costs-100%

-50%

0%

50%

100%

150%

200%

Jan-9

3

Jan-9

4

Jan-9

5

Jan-9

6

Jan-9

7

Jan-9

8

Jan-9

9

Jan-0

0

Jan-0

1

Jan-0

2

Jan-0

3

Jan-0

4

Jan-0

5

Jan-0

6

Jan-0

7

Jan-0

8

Jan-0

9

Cu

mu

lati

ve

Re

turn

%

Trading Strategy Long RUT/Short DJI

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

15

Applications in passive portfolio management:

Determining optimal holding periods

Notes

1. Liquidity Risk is expressed using the Stock Liquidity Rating (SLR) scheme: (AAA, AA, A, BBB, BB, B, CCC, CC, C and D) with AAA having lowest risk and D having highest risk

2. Cumulative Return is the equally weighted return for given SLR portfolio, adjusted for round trip market impact cost. A SLR portfolio is defined as all stocks with a given SLR selected from the universe of largest 3000 U.S. stocks. The average size of a portfolio is $300 million. The portfolio is held constant throughout the holding period.

3. Period analyzed: Jan 1993 – Jun 2008

Cumulative Return by Liquidity Risk (SLR)

-5%

0%

5%

10%

15%

20%

25%

1 2 3 4 8

Holding Period (Quarters)

Cu

mu

lati

ve

Re

turn

AAA

A

BB

CCC

C

The U.S. evidence suggests

for shorter holding periods (less than 3 quarters) liquid securities provide a higher return on investments

for longer holding periods (more than 3 quarters) illiquid securities provide a higher return on investments

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

16

Applications in Asia: Hong Kong equities versus

US equities

Period Analyzed: Jan – Dec 2007

Hong Kong Equity market has significantly more liquidity risk as indicated by two liquidity metrics

Liquidity VaR: This is the cost incurred for liquidating an (equal-weighted) market portfolio. Higher liquidity VaR indicates more liquidity risk.

Turnover: A lower turnover number (as % of market capitalization) indicates higher liquidity risk

Equity Market

# of

securities

Total Market Cap

(Millions)

Monthly Turnover

(Millions)

Avg Daily Turnover as

% of Total Market Cap

Liquidity VaR

(bps)

Hong Kong (Main Board) 1021 17218587 HK$ 1355482 HK$ 0.38% 766

NYSE 2698 15644242 USD 2757078 USD 0.77% 185

Hong Kong equity market has four times more liquidity risk, and half the average daily turnover, compared to U.S. equity market

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

17

Applications in Asia: Hong Kong equities more

vulnerable to liquidity crisis

The Market Illiquidity Level (MIL) is the barometer of liquidity conditions for an Equities Market. An increase in this level indicates deteriorating liquidity conditions.

US equities’ illiquidity peaked during the last week of November 2007. However, Hong Kong equities illiquidity peaked during the last week of January 2008

OGI's Market Illiquidity Level (MIL)

0

50

100

150

200

250

300

350

11/2

6/2

006

1/2

6/2

007

3/2

6/2

007

5/2

6/2

007

7/2

6/2

007

9/2

6/2

007

11/2

6/2

007

1/2

6/2

008

3/2

6/2

008

MIL (HK Equities)

MIL (U.S. Equities)

Incre

asin

g Liq

uid

ity R

isk

c

Source: OGI, Inc.

Hong Kong equities’ illiquidity deteriorated considerably more compared to US Equities

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

18

OGI Composite India Fund Sample Performance Report

(Source: OGI, Inc.)

Applications in Asia: India

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

19

Applications in Asia: India (Cont’d)

OGI Composite India Fund – Top Vs. Bottom

(Source: OGI, Inc.)

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

20

Summary

Liquidity Risk an important source of risk and still being understood

We introduced empirical metrics to estimate liquidity risk using intraday data that have predictive ability, both in US and Asian markets

Introduce practical applications to manage and exploit liquidity risk

CENTER FOR

ASSET

MANAGEMENT

RESEARCH &

INVESTMENTS

21

Appendix: Definitions

Market Illiquidity Level (MIL) is the median illiquidity level for stocks, as captured by the Stock Illiquidity Level (SIL), for the entire market of stocks selected from a universe of 3000 largest public U.S. equities by market capitalization, as determined at the beginning of the quarter. The weekly SIL for each stock is determined using intra-day trading data (ILLIQt). The median SIL across the universe is denoted as MIL. The. The MIL is based on an initial value of 100 registered on Jan 8, 1993. An increase in MIL indicates deteriorating liquidity conditions. When MIL declines, illiquid securities can be expected to outperform liquid securities. When MIL increases, illiquid securities can be expected to underperform liquid securities.

Stock Liquidity Rating (SLR) measures a stock's liquidity risk, given by the uncertainty associated with the cost of liquidating a position (εt). SLR categorizes a stock into one of ten liquidity risk buckets (AAA, AA, A, BBB, BB, B, CCC, CC, C, D), with AAA having the least risk and D the greatest risk

Market Illiquidity Factor (MIF) measures how liquidity risk is priced by market participants. It measures the cumulative return of illiquid securities relative to liquid securities as ranked by the stock-level liquidity rating system (SLR). The MIF for the U.S Equities Market is created through analysis of the 3,000 largest U.S. stocks. The MIF is based on an initial value of 100 registered on April 1, 1993.