Significance Of Roe And Size Of Stock

PERFORMANCE ANALYSIS OF SET50 INDEX PORTFOLIO VERSUS NON SET50 EQUITY PORTFOLIO: MEASUREING THE IMPORTANCE OF ROE AND MARKET CAPITALIZATION IN THAI STOCK MARKET

Panu Chaopricha 20 July 2007

Organization of Presentation Introduction      

Purpose of the study Background of study(& literature review) Theoretical framework Background of SET, SET50 Statement of the problem Significant of the study

Organization of Presentation Methodology     

Research questions and hypothesis Research design Data and time period Data collection Data analysis method

Organization of Presentation Analysis and Results   

Portfolio formation Data analysis Results and findings

Organization of Presentation Summary     

Discussion of findings Confirmation and discovery Alternative explanation for findings Qualifications Recommendations for future research

Introduction

Purpose of the study To establish relationship between  Firm characteristics (1) (2)



The status as a member of SET 50 index, Return on Equity (ROE)

Portfolio returns

Background of Study 

Investors have been seeking to increase wealth via stock market for decades



Large variety of concepts and techniques has been applied to gain the most out of stock market.

Background of Study 

Fundamental analysis involves analyzing the business facts around the company.



Technical analysis deals with forecasting of price trends based on historical price behavior. It is done primarily through the use of chart.

Background of Study 

In an attempt to predict stock returns, researchers have studies relationships between stock returns and each of     

Price-to-Book Value Ratio Price-to-Sales Ratio Price-Earning Ratio Market Capitalization Debt-to-Equity Ratio

Size vs. Stock Return Author(s)

Relationship

Remark

Banz (1981)

Negative

Basu (1983)

Negative

Berk (1995)

Negative

Chan and Chen (1991)

Negative

Chui and Wei (1998)

Negative

Except for Taiwan

Dhatt, Kim and Mukherji (1999)

Positive

Among small stocks universe

Drew, Naughton and Veeraraghavan (2003)

Negative

Risk-adjusted return

Fama (1991)

Negative

Especially in January

Fama and French (1998)

Negative

11 out of 16 markets.

Gerald et al. (1997)

Negative

Only during expansive monetary policy

Keim (1983)

Negative

Loeb (1991)

Negative

Reignamum (1982)

Negative

Small stocks

P/BV Ratio vs. Stock Return Author(s)

Relationship

Remark

Aggarwal et al. (1992)

Negative

Stronger in Jan, June and small firms

Black (1993) and MacKinlay (1995)

Negative

A chance result

Capaul, Rowley and Sharp (1993)

Negative

Risk-adjusted returns

Chan and Chen (1991)

Negative

this variable means risky & distressed

Chan, Hamao and Lakonishok (1991)

Negative

Chen and Zhang (1998)

Negative

Strong in developed market, Weak in growth market

Chui and Wei (1998)

Negative

Only HK, Korea and Malaysia

Daniel and Tittman (1997)

Negative

Debndt and Thaler (1987)

Negative

Dhatt, Kim and Mukherji (1999)

Negative

M2B stronger than PE

Drew, Naughton and Veeraraghavan (2003)

Positive

Risk-adjusted return

Fama and French (1992)

Negative

Strongest variable

P/BV Ratio vs. Stock Return Author(s)

Relationship

Remark

Fama and French (1992, 1996)

Negative

Low P/BV = high risk

Fama and French (1998)

Negative

12 out of 13 major markets

Gerald et al. (1997)

Negative

Only during expansive monetary policy

Gonenc and Karan (2003)

Positive

Haugen (1995)

Negative

Jacob Levy and Reingamum (1988)

Negative

Lakonishok, Shleifer and Vishny (1994)

Negative

Loughran (1997)

Negative

Especially in January

Loughran (1997)

No

Once January is excluded

Malkiel (1995)

Questionable

Roll (1995)

Negative

Rosenberg, Reid and Lanstein (1988)

Negative

Stattman (1980)

Negative

Not statistically significant

Price to Sales Ratio vs. Stock Return

Author(s)

Relationship

Remark

Dhatt, Kim and Mukherji (1999)

Negative

Price-to-Sale Stronger power than P/BV

Barbee, Mukherji and Raines (1996)

Negative

PE Ratio vs. Stock Return Author(s)

Relationship

Aggarwal, Hiraki and Rao (1990)

Negative

Ander (1982)

Negative

Basu (1977)

Negative

Breen (1978)

Negative

Dreman (1980a, 1980b, 1979)

Negative

Goff (1979)

Negative

Reignamum (1981)

Almost absent

Remark

When controlled for firm size

DE Ratio vs. Stock Return Author(s)

Bhandari (1998)

Relationship

Positive

Remark

Theoretical Framework P/BV Price/Sale P/E Market Capitalization

-

+

-

Stock Returns

-

? ?

D/E ROE Size + ROE

Earning = Net Profit Equity

Depends on relationship of variable in the ratio

Theoretical Framework ROE Numerator Denominator

Net Profit Equity

Stock Return Capital gain +Dividend gain Purchased price

Return on Investment Return Investment

Background of Stock Exchange of Thailand   



The official market for Thai Equity Instrument First trading on April 30, 1975 Market capitalization was 5,101 billion Baht (US$ 142 Billion) in 2006 518 listed companies categorized under 8 industries and 29 sectors.

Market Capitalization

Graph Unit: Billion USD

Comparison of Market Capitalization

836

358 227

142

129

62 h P i il p p in

d

s e

ia s

n a

re o

g n

ia s y

e n o

il a

d In

h T

la a M

a

o K

p a

re

g n

g in S

o K

o H

Source: Stock Exchange of Thailand

Background of Stock Exchange of Thailand Stock Exchange of Thailand  71% of Thailand GDP  55% foreign, 12% institutional and 34% local investors

Background of SET50 Index SET 50 Index  A market capitalization-weighted price index  The SET has been calculating the SET50 Index since August 16, 1995  In 2006, market capitalization of SET50 companies accounts for 72% of SET market capitalization

Background of SET50 Index Purpose of SET50 Index  to accommodate the issuing of index futures and options (SET50 Index futures was launched in April 2006)  To provide a benchmark of investment in The Stock Exchange of Thailand.

Statement of the problem 



 

What types of stocks investors should be considering? Many preferred large market cap, liquid, wellestablished. => SET50 Is investment in SET50 firms fruitful? What characteristics have strong predictive power for stock return?

Significance of the study     

First evidence on new relationships Prove on perception SET50 index stocks yield superior return Usage of widely-recognized optimization process Frequent rebalancing assumption Unbiased use of data   

 

Realistic adoption period of ROE Realistic period of sector classification Avoidance of survivorship bias

Exclusion of low liquidity stocks Use of powerful model in creating portfolios

Methodology

Research Questions 1.

2.

3.

Is there a difference between returns of stocks that are listed in SET 50 index and those that are not? Is there a relationship between company ROE and stock returns?” If any or both have significant relationship with stock returns, how differently portfolios constructed based on single or both of these variables perform?

Hypothesis RP

denotes Return on portfolio

SET

denotes Consisting of SET50 stocks

NON denotes Consisting of non SET50 and liquid stocks HI

denotes Consisting of stocks with ROE higher than their sectors’ averages

LO

denotes Consisting of stocks with ROE equal to or lower than their sectors’ averages

Hypothesis 1-2 H1: Non SET50 portfolio does not yield higher returns than SET50 portfolio. H0

:

RP [NON] ≤ RP [SET]

HA

:

RP [NON] > RP [SET]

H2: Non SET50 portfolio does not yield lower returns than SET50 high ROE portfolio. H0 HA

: :

RP [NON] ≥ RP [SETHI] RP [NON] < RP [SETHI]

Hypothesis 3-4 H3: Non SET50 portfolio does not yield higher returns than SET50 low ROE portfolio. H0 HA

: :

RP [NON] ≤ RP [SETLO] RP [NON] > RP [SETLO]

H4: Non SET50 high ROE portfolio does not yield higher returns than SET50 portfolio. H0 HA

: :

RP [NONHI] ≤ RP [SET] RP [NONHI] > RP [SET]

Hypothesis 5-6 H5: Non SET50 high ROE portfolio does not yield higher returns than SET50 high ROE portfolio. H0 HA

: :

RP [NONHI] ≤ RP [SETHI] RP [NONHI] > RP [SETHI]

H6: Non SET50 high ROE portfolio does not yield higher returns than SET50 low ROE portfolio. H0 HA

: :

RP [NONHI] ≤ RP [SETLO] RP [NONHI] > RP [SETLO]

Hypothesis 7-8 H7: Among non SET50 stocks, high ROE portfolio does not yield higher returns than low ROE portfolio. H0

:

RP [NONHI] ≤ RP [NONLO]

HA

:

RP [NONHI] > RP [NONLO]

H8: Non SET50 low ROE portfolio does not yield returns lower than SET50 portfolio. H0 HA

: :

RP [NONLO] ≥ RP [SET] RP [NONLO] < RP [SET]

Hypothesis 9-10 H9: Non SET50 low ROE portfolio does not yield returns lower than SET50 low ROE portfolio. H0 : RP [NONLO] ≥ RP [SETHI] HA : RP [NONLO] < RP [SETHI]

H10: Non SET50 low ROE portfolio does not yield higher returns than SET50 low ROE portfolio. H0 : RP [NONLO] ≤ RP [SETLO] HA : RP [NONLO] > RP [SETLO]

Hypothesis 11-12 H11: Among SET50 stocks, high ROE portfolio does not yield higher returns than low ROE portfolio. H0 : RP [SETHI] ≤ RP [SETLO] HA : RP [SETHI] > RP [SETLO]

H12: Return of portfolio constructed from SET50 stocks is not equal to return of market portfolio. H0 : RP [SET] ≠ RP [MARKET] HA : RP [SET] = RP [MARKET]

Hypothesis 13 H13: Return of non SET50 portfolio is not higher than that of market portfolio. H0 : RP [NON] ≤ RP [MARKET] HA : RP [NON] > RP [MARKET]

Research Design 7 Portfolios are created for return comparisons  SET50 Portfolio  SET50 High ROE Portfolio  SET50 Low ROE Portfolio  Non SET50 Portfolio  Non SET50 High ROE Portfolio  Non SET50 Low ROE Portfolio  Market Portfolio (of which market index is used for return calculation)

Data and Time Period Population includes equity stocks in Stock Exchange of Thailand that existed during 2001 – 2006.

Data collection Data for stock screening  Company’s ROE  Classification of companies under sector  Sector’s ROE and trading value Downloaded from http://www.setsmart.com  SET50 Listings from 2004-2006 Downloaded from http://www.set.or.th/

Example of Setsmart.com

Example of SET.or.th

Data Collection Data for return calculation  Daily trading data (open, high, low, close and volume) from 2001-2006 Purchased from Fundtecon Co., Ltd in form of MetaStock File (right issues, par changes, name/symbol changes are taken care of)  Dividend Downloaded from http://www.siamfn.com

mber

Example of MetaStock File Relative Strength Index (40.7480) 50

MACD (-12.9643)

10 0 -10 -20

89 SET (679.490, 681.140, 672.810, 679.840, -0.51996)

800 790 780 770 760 750 740 730 720 710 700 690 680 670 660 650 640 630 620 610 600 590 580

50000 x100000

2006

February

March

April

May

June

July

August

September October

November

December 2007

February

Example : Data extracted from MetaStock Date Open High Low Close Volume 03/01/2001 36.8 36.8 36.2 36.2 58000 04/01/2001 36.6 37 36.6 36.8 231000 05/01/2001 37 39 37 38 876000 08/01/2001 39.8 40.2 39.8 39.8 1328000 09/01/2001 40.2 41.6 40 41 2245000 10/01/2001 41 41.6 38.8 38.8 938000.1 11/01/2001 38.8 39.6 38.8 39 394000 12/01/2001 40 42.2 40 42 1278000 15/01/2001 42 42.2 41.8 42 794000 16/01/2001 41.6 41.6 40.8 41.2 185000 17/01/2001 41.2 42 41.2 41.8 520000 18/01/2001 42.2 42.8 42 42 55000 19/01/2001 42 43.2 42 43.2 1030000 22/01/2001 43 43 42.4 42.6 69000 23/01/2001 43.2 43.2 42.6 42.6 41000 24/01/2001 42.6 43.2 42.6 43 782000 25/01/2001 43.2 44.2 43 43.6 135000 26/01/2001 44 44.4 43 43.4 637000 29/01/2001 43 44.6 42.4 43.4 413000 30/01/2001 43.4 44.2 43.2 44.2 671000

Example of Siamfn.com

Stock Screening 

Data for 20 quarters are downloaded. Redundant stock names are removed, leaving 517 unique names.



Stocks are screened for liquidity.  

Average daily trading over 10 M Baht In 4 quarters before investment quarter

Stock Screening Average daily trading is calculated by Quarterly Trading Value____ Number of trading days in the quarter

Rationale behind 10 M Threshold 2003

2004

2005

2006

Total Market Daily 18,908.22 20,507.75 16,454.88 16,280.91 trading value (Million Baht) No. of stock at year end

407

441

468

488

Average Trading value/stock (Million Baht)

46.46

46.50

35.16

33.36

Investment Quarter

1 Million

5 Million

10 Million

15 million

04Q1

158

107

82

66

04Q2

188

139

106

93

04Q3

184

121

99

92

04Q4

190

124

101

93

05Q1

182

117

94

78

05Q2

184

121

96

79

05Q3

183

119

85

70

05Q4

191

123

89

77

06Q1

197

120

91

69

06Q2

197

120

88

64

06Q3

198

118

82

65

06Q4

205

118

85

66

Average No. of Stocks

188

121

92

76

ROE Screening ROE of stock is classified as follows. 

Higher than sector’s average = HI



Lower than or equal to sector’s average = LO

Size Screening SET50 index listing is used for size screening. H1

Q1

Announce listing

H2

Q2

Q3

Announce listing

Q4

Return Computation Daily return of stock is calculated as required by QuantAnalysis Program for further computation =

P1 – P0 + D1 P0

Where P1 = P0 = D1 =

close price of the day close price of previous trading day Cash dividend and proceeds from selling stock dividends

Note on Dividend 



All dividends are manually adjusted to reflected relevant par changes. Stock dividend is assumed to be sold upon receipt (using close price).

INSTRUMENT QuantAnalysis Program  Upper/lower bound sell limit  Screening: price strength, CV, Return, Standard deviation, average Rf-beta, daily volume, dividend yield  Optimization: MPT, RAMP, MVP, Equal weights  Rebalancing frequency: weekly, biweekly, monthly, bimonthly, quarterly  Investment Period

QuantAnalysis

Investment parameters  



  

Upper/lower sell limit = 100% Screening : manual screening for SET50, non SET50, Hi ROE, Low ROE and average daily trading volume. Optimization : Markowitz’s Modern Portfolio Theory Number of observation : 36 months Optimization Rebalancing Freq.: monthly Investment Rebalancing Freq.: weekly

How it works Returns of last 36 months SET50

Determine investment proportion

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works Returns of last 36 months SET50

Determine investment proportion

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works Returns of last 36 months SET50

Determine investment proportion

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works Returns of last 36 months Determine investment proportion

SET50

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works Returns of last 36 months Determine investment proportion

SET50

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works Returns of last 36 months Determine investment proportion

SET50

Liquidity

Q1

Q2

Q3

Q4

Q1

Q2

ROE

Q3

Q4

Q1

Q2

Q3

Q4

How it works   



Stocks are screened for each quarter. 12 quarters X 6 portfolios = 72 stock pools. QuantAnalysis uses each stock pool to create portfolio during each quarter. Results of all quarters are compiles to form 3 years results of each portfolio.

Data Analysis Method Statistics Used T-test: 2 samples assuming unequal variances to analyze whether mean difference of each pair of portfolio returns is significantly different from 0.

Results and Analysis

QuantAnalysis Output : Parameter Initial Investment Amount

$1,000,000

Portfolio Name

NONHI04Q4

Upper Bound Sell Limit

100.00%

Lower Bound Sell Limit

100.00%

Path and Name of Input Data File (Output of DataGrabber Advance)

C:\Documents and Settings\THINKPAD R60\Desktop\SET\For QuantAnalysis\SET50\NON HI TEXT\NONHI04Q4.xls

Optimization Methods

MPT

# of Observations to Optimize

36

Optimization Rebalancing Freq.

Monthly

Market

S&P 500

Used Provided MSN Betas

No

Risk Free Rate (Rf)

2.50%

Money Market Return (yearly)

2.00%

QuantAnalysis Output : Parameter Screening Methods

----

Price Strength Greater than

1.00

From … Days Back

5

Daily Volume Greater than

10,000

Average Volume over ... Days

30

CV Less than

10.0

Return Greater than

0.020

Standard Deviation Less than

0.100

Dividends ($) Greater than

1.00

(Avg-Rf)/Beta Greater than

0.002

TOP ….

15

Trading Start Date

Month

Year

10

2004

Rebalance Day

Monday

Investment Rebalancing Freq.

Weekly

Shift Start Day by X weeks

0

Add Extra Day

No

QuantAnalysis Output : Graph Investment vs Market Return

4.00%

2.00%

0.00% S-04

S-04

O-04

O-04

O-04

N-04

N-04

N-04

D-04

D-04

-2.00%

-4.00%

-6.00%

-8.00%

-10.00%

-12.00% NONHI04Q4

MARKET

D-04

J-05

QuantAnalysis Output : Graph DAILY: Investment vs Market Return

4.00%

2.00%

0.00% S-04

S-04

O-04

O-04

O-04

N-04

N-04

-2.00%

-4.00%

-6.00%

-8.00%

-10.00%

-12.00% NONHI04Q4

N-04

D-04

D-04

D-04

J-05

QuantAnalysis Output : Graph 40,000

NONHI04Q4 CUMMULATIVE GAIN/LOSS BREAKDOWN

20,000 00

S-04 -20,000

S-04

O-04

N-04

N-04

D-04

-40,000 -60,000 -80,000 -100,000 -120,000

Money Market Gains

Dividend Gains

Captial Gains/Loss

Total Profit/Loss

J-05

QuantAnalysis Output : Investment Allocation 36

Number of Oberservations:

Monthly

Rebalancing Frequency:

Weekly

Investment Rebalancing Frequency:

MPT

Optimization Method:

13

Number of Investment Periods:

(Markowitz Modern Portfolio Theory)

Recommendations: Portfolio without Short Sales

01/10/2004 Initial Investment Proportions: Company Name Proportion Return of Portfolio Risk of Portfolio

AH 9.73% 5.05% 10.04%

CP7-11 2.39%

DCC 16.20%

FNS 2.39%

HEMRAJ 8.15%

KCE 4.99%

KTC 1.09%

LPN 2.13%

PICNI 7.71%

PLE 8.60%

SCBL 4.60%

SPALI 14.43%

TNITY 11.88%

SCIB 24.43%

SPALI 4.98%

TNITY 4.02%

20% 15% 10% 5% 0% AH

CP7-11

DCC

FNS

HEMRAJ

KCE

KTC

LPN

PICNI

PLE

SCBL

SPALI

TNITY

TRU

TUF

Recommendations: Portfolio without Short Sales

27/12/2004 Final Investment Proportions: Company Name Proportion Return of Portfolio Risk of Portfolio

AH 7.23% 3.42% 8.60%

BH 2.92%

BLAND 3.11%

CP7-11 4.57%

DCC 23.98%

FNS 8.19%

JAS 0.14%

KTC 2.80%

30% 20% 10% 0% AH

BH

BLAND

CP7-11

DCC

FNS

JAS

KTC

MK

PICNI

SCIB

SPALI

TNITY

MK 8.31%

PICNI 5.31%

QuantAnalysis Output : Data Table NONHI04Q4 - CUMMULATIVE DATA TABLE Investment

Start Date

End Date

0 1 2 3 4 5 6 7 8 9 10 11 12 13

-01/10/2004 04/10/2004 11/10/2004 18/10/2004 26/10/2004 01/11/2004 08/11/2004 15/11/2004 22/11/2004 29/11/2004 07/12/2004 13/12/2004 20/12/2004

01/10/2004 04/10/2004 11/10/2004 18/10/2004 26/10/2004 01/11/2004 08/11/2004 15/11/2004 22/11/2004 29/11/2004 07/12/2004 13/12/2004 20/12/2004 27/12/2004

Money Market Gains 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Dividend Gains 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Captial Initial Final Total Gains/Loss Investment Investment Profit/Loss 0.00 24,248.36 4,086.58 -52,004.21 -60,274.45 -100,589.47 -82,595.14 -75,192.42 -71,587.88 -28,461.97 -32,709.70 -45,832.11 -10,239.34 -22,010.53

0 1,000,000 1,024,248 1,004,087 947,996 939,726 899,411 917,405 924,808 928,412 971,538 967,290 954,168 989,761

1,000,000 1,024,248 1,004,087 947,996 939,726 899,411 917,405 924,808 928,412 971,538 967,290 954,168 989,761 977,989

0.00 24,248.36 4,086.58 -52,004.21 -60,274.45 -100,589.47 -82,595.14 -75,192.42 -71,587.88 -28,461.97 -32,709.70 -45,832.11 -10,239.34 -22,010.53

STD Dev

1.37% 1.90% 1.94% 2.00% 1.68% 2.10% 1.88% 1.36% 1.15% 1.59% 1.31% 1.27% 0.60%

SP500 668.29 679.13 677.93 646.51 648.38 626.96 629.2 647.56 644.95 657.25 655.83 645.75 675.71 663.86

Investmen t 0.00% 2.42% 0.41% -5.20% -6.03% -10.06% -8.26% -7.52% -7.16% -2.85% -3.27% -4.58% -1.02% -2.20%

SP500 0.00% 1.62% 1.44% -3.26% -2.98% -6.18% -5.85% -3.10% -3.49% -1.65% -1.86% -3.37% 1.11% -0.66%

Results SET50 Capital Gain -967,113.33 Dividend Gain 682,101.23 Total Profit -285,012.09 Returns -0.19% Std Dev(%) 5.76%

SETHI -2,345,567.40 1,070,490.95 -1,275,076.45 -0.86% 6.38%

SETLO NON 1,047,728.68 -3,526,632.35 653,929.37 388,582.32 1,701,658.05 -3,138,050.03 1.15% -2.12% 6.89% 8.62%

Average Return SET LO NON HI MKT SETLO 1.15% 0.2335 0.0950 NON HI 0.58% 0.2906 MKT 0.18% SET 50 -0.19% SET HI -0.86% NON -2.12% NON LO -4.88%

NONHI -407,431.48 1,263,186.63 855,755.15 0.58% 6.61%

SET 50 0.0351 0.1428 0.2884

SET HI 0.0048 0.0288 0.0704 0.1724

NONLO MARKET -7,417,194.81 193,562.07 -7,223,632.74 -4.88% 0.18% 9.29% 5.74%

NON NON LO 0.0002 0.0000 0.0014 0.0000 0.0037 0.0000 0.0123 0.0000 0.0772 0.0000 0.0042

Conclusion No Null Hypothesis .

Results

Significant at α = 0.05

1

RP [NON] ≤ RP [SET]

Do Not Reject

Sig.

2

RP [NON] ≥ RP [SETHI]

Do Not Reject

Not Sig.

3

RP [NON] ≤ RP [SETLO]

Do Not Reject

Sig.

4

RP [NONHI] ≤ RP [SET]

Do Not Reject

Not Sig.

Conclusion No Null Hypothesis .

Results

Significant at α = 0.05

5

RP [NONHI] ≤ RP [SETHI]

Reject

Sig.

6

RP [NONHI] ≤ RP [SETLO]

Do Not Reject

Not Sig.

7

RP [NONHI] ≤ RP [NONLO]

Reject

Sig.

8

RP [NONLO] ≥ RP [SET]

Reject

Sig.

Conclusion No Null Hypothesis .

Results

Significant at α = 0.05

9

RP [NONLO] ≥ RP [SETHI]

Reject

Sig.

10

RP [NONLO] ≤ RP [SETLO]

Do Not Reject

Sig.

11

RP [SETHI] ≤ RP [SETLO]

Do Not Reject

Sig.

12

RP [SET] ≠ RP [MARKET]

Reject

Not Sig.

13

RP [NON] ≤ RP [MARKET]

Do Not Reject

Sig.

Conclusion 





SET 50 portfolio yields higher return than non SET50 portfolio Between 2 SET50 portfolios, Low ROE one performs better than High ROE one. Between 2 Non SET50 portfolio High ROE one performs better than LOW ROE one..

Conclusion 



SET 50 portfolio does not perform differently from Market. Non SET50 Low ROE is the worst performer

Confirmation and Discovery 



Contrary to most of previous findings, size has positive relationship with stock returns. Confirm the finding of Dhatt, Kim and Mukherji (1999) who found positive relationship in small stock universe.

Confirmation and Discovery 





For SET50, ROE might mean risk; higher risk => higher return For Non SET50, ROE might mean growth potential; high ROE => high return High ROE portfolios yields significantly higher dividend gain than low ROE portfolios. (Both SET50 and non SET50)

Alternative Explanation 



If SET50 stocks are regarded as index stocks, in stead of large market cap stocks, it is consistent with the perception that index funds usually outperform other types of fund. ROE might not be relevant variable since because of inconsistent results

Qualifications 

There was a major surge in stock index during 2002-2003. This might distort the results.

Market Capitalization and Index : Stock Exchange of Thailand

Market Capitalization SET index mai index

Market Capitalization and Index

Billion Baht

Points

(USD 142) 6,000

1,400

1,280.81

5,101

1,200

5,000

1,000

4,000 772.15

800

831.57

600

668.1 713.73 679.84

3,000

481.92 372.69

400

269.19

355.81

303.85

356.48

2,000 193.45

200 0

1,000 0

1995

1996

1997

Source : SETSMART as of 29 December 2006 Remarks: SET and MAI

1998

1999

2000

2001

2002

2003

2004

2005

2006

Source: Stock Exchange of Thailand

Qualifications 



There was a major surge in stock index during 2002-2003. This might distort the results. Methodology of this research is different from others in many aspects, possibly causing different results.

Qualifications 





Stock Exchange of Thailand as an emerging market which might not be 100% efficient. Many parts of this research are conducted manually, e.g. data input, cut and paste. It cannot be guaranteed of error free in spite of several rechecks. Trading commission is not taken into account, resulting in overstated results.

Qualifications 



Earning as reported in financial statement can be manipulated. Therefore, one needs to be careful in using accounting data. Dividend data retrieved from www.siamfn.com (by Atkinson Plc.) need verification with SET yearly fact book.

Recommendation for future research  



Covering longer period of study Studying index stocks vs. non index stocks in other markets to see if index stocks has the same meaning of large market cap. Varying rebalancing policies to find which frequency is optimal. This research tried monthly rebalancing and found majority of relationships not significant but in same direction as reported here (weekly).

Recommendation for future research  

Varying liquidity criteria as market grows Trying different optimization processes, e.g. equal weights, minimum variance portfolio, Risk-Adjusted Model-Driven Portfolio.

THANK YOU VERY MUCH

Steps in Screening ROE Step 1 creating ROE table by company by quarter

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Stock 1

XXX XXX XXX XXX XXX XXX XXX XXX

Stock 2

XXX XXX XXX XXX XXX XXX XXX XXX

.…

XXX XXX XXXCompany XXXROE XXX XXX XXX XXX

Stock 517 XXX XXX XXX XXX XXX XXX XXX XXX

Steps in Screening ROE Step 2 creating Sector table by company by quarter

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Stock 1

AGRI AGRI AGRI AGRI AGRI AGRI AGRI AGRI

Stock 2

TECH TECH TECH TECH TECH TECH ICT

…

XXX XXX XXXSector Classification XXX XXX XXX XXX XXX

ICT

Stock 517 XXX XXX XXX XXX XXX XXX XXX XXX

Steps in Screening ROE Step 3 creating ROE table by sector by quarter

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Agri

XXX XXX XXX XXX XXX XXX XXX XXX

Food

XXX XXX XXX XXX XXX XXX XXX XXX

Fashion

XXX XXX XXX XXX XXX XXX XXX XXX

Home

XXX XXX XXX XXX XXX XXX XXX XXX

…

XXX XXX XXX XXX XXX XXX XXX XXX

Tech

XXX XXX XXX XXX XXX XXX XXX XXX

Sector’s Average ROE

Steps in Screening ROE Step 4 using tables created in Step 2 and Step 3 to create Sector’s average table by company by quarter

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Stock 1

XXX XXX XXX XXX XXX XXX XXX XXX

Stock 2

XXX XXX XXX XXX XXX XXX XXX XXX

…

XXX XXX XXXSector’sXXX XXX XXX XXX XXX Average ROE

Stock 517 XXX XXX XXX XXX XXX XXX XXX XXX

Steps in Screening ROE Step 5 minus all values in table from Step 4 from all values in table from Step 1

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Stock 1

XXX XXX XXX XXX XXX XXX XXX XXX

Stock 2

XXX XXX XXX XXX XXX XXX XXX XXX

…

XXX XXX XXXCompany XXX XXXAverage XXX ROE - Sector’s ROE XXX XXX

Stock 517 XXX XXX XXX XXX XXX XXX XXX XXX

Steps in Screening ROE Step 6 consider for use in the other quarter

03Q3 03Q4 04Q1 04Q1 …… ……. 06Q1 06Q2 Stock 1

XXX XXX XXX XXX XXX XXX XXX XXX

Stock 2

XXX XXX XXX XXX XXX XXX XXX XXX

…

XXX XXX XXXPositive XXX or Negative XXX XXX XXX XXX

Stock 517 XXX XXX XXX XXX XXX XXX XXX XXX