The impact of Black Economic Empowerment transaction

The impact of Black Economic Empowerment transaction
The impact of Black Economic Empowerment transaction
announcements on share price performance of JSE listed
mining companies
LESANG EDMUNDS SENNANYE
Student number: 446387
A research project submitted to the Gordon Institute of Business Science, University of
Pretoria, in partial fulfilment of the requirements for the degree of Master of Business
Administration.
14 January 2015
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
ABSTRACT
The South African government introduced the Black Economic Empowerment (BEE) as
an intervention to resolve economic imbalances.
In furthering inclusivity in the
previously exclusive sectors, like Mining, the BEE legislations and Mining Charter were
introduced to benefit the HDSA. The study addressed a significant gap in BEE
research, which is important within the South African context, as the country currently
reviews progress after the initial 20 years of democratic dispensation.
The research examined the share price performance of mining stocks listed on the JSE
by tracking their share price performance after announcements relating to black
empowerment transactions. The objectives of the research were to, first, determine
whether announcements of BEE transactions lead to better shareholder wealth
creation in the South African mining sector, second, to determine the impact of these
announcements on Old and BEE mining companies that were listed on the JSE post1994, third, to determine whether the early BEE announcements made before the
release of the Mining Charter in September 2010 had a greater positive impact on the
Cumulative Abnormal Returns (CARs) of Mining companies compared to those made
after the amendment to legislation.
The research employed an event study methodology to analyse a sample of 26 mining
companies that made a total of 241 qualifying announcements from January 2000 to
November 2014.
The results of the study showed negative impact on the CARs of the mining
companies. It was noted that the old mining companies that existed before 1994 had
better average abnormal return than the BEE companies. Further, the results showed
that the Average Abnormal Returns (AARs) of the BEE announcements made prior to
the Mining Charter had greater AARs than those made after the implementation. In
sum, the BEE announcements had largely a negative impact on share performance of
the mining companies.
Keywords
Black Economic Empowerment (BEE), Mining Shares, Event Study, Abnormal Returns,
JSE
i|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
DECLARATION
I declare that this research project is my own work. It is submitted in partial fulfilment of
the requirements for the degree of Master of Business Administration at the Gordon
Institute of Business Science, University of Pretoria. It has not been submitted before
for any degree or examination in any other University. I further declare that I have
obtained the necessary authorisation and consent to perform this research.
--------------------------------Lesang Edmunds Sennanye
14 January 2015
iii | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
ACKNOWLEDGEMENTS
I would like to thank my supervisor, Professor Mike Ward, for his time, guidance and
insight; it has been a good leaning. I would also like to thank Chris Muller who
partnered with Mike to assist me with the database and running of their developed
event study model.
I am grateful to GIBS management, the faculty and support services for a great
learning experience. Thanks go to my MBA colleagues whom I learnt with and learnt
from. With a special mention to my buddies, Fikile Holomisa, Lyborn Mashava, Memory
Nyanga, Ntombi Nyaga and Melusi Sigasa. I also thank Anastacia Mamabolo for her
time and guidance. Her passion for research brought light into my research.
I appreciate all the support I got from my friends and family who gave me support
during the pressured times of my studies. I am grateful to the friendship and support I
got from Kabelo Mogajane especially for assisting Chabi with the farming operations.
I am indebted to my wife, Masechaba, for her love, support and allowing me an
opportunity to embark on the MBA journey. I am equally indebted to my son, Leano
who missed a lot of playtime with me during the course of my studies. I thank the two of
you for adding meaning to my life.
iv | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
DEDICATION
This research work is dedicated to my favourite teacher Mrs. Catherine Segwagwa
from Bona-Bona village, one of the teachers who motivated me from an early age and
instilled love for education and learning.
v|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
TABLE OF CONTENTS
ABSTRACT ........................................................................................ I DECLARATION ................................................................................ III ACKNOW LEDGEM ENTS ................................................................... IV DEDICATION .................................................................................... V LIST OF TABLES ........................................................................... VIII LIST OF FIGURES ............................................................................ IX LIST OF EQUATIONS ........................................................................ X CHAPTER 1: INTRODUCTION TO THE RESEARCH PROBLEM ........... 1 1.1 Research title ......................................................................................................... 1 1.2 Introduction ............................................................................................................ 1 1.3 Background of Study.............................................................................................. 2 1.4 Research Problem ................................................................................................. 3 1.5 Research objectives............................................................................................... 4 1.6 Research Scope .................................................................................................... 4 1.7 Outline of Research Study ..................................................................................... 5 CHAPTER 2: LITERATURE REVIEW ................................................. 6 2.1 General background information............................................................................ 6 2.1.1 South Africa’s Black Economic empowerment ......................................................... 6 2.1.2 Broad-Based Black Economic Empowerment Act .................................................... 7 2.2 South African Mining .............................................................................................. 8 2.3 Modes of equity transfer and announcements ..................................................... 11 2.3.1 2.3.2 2.3.3 2.3.4 2.3.5 Impact of BEE announcements on cumulative returns ........................................... 12 Short-term cumulative returns................................................................................. 12 Long-term cumulative returns ................................................................................. 13 Link between the age of the company and its share price performance................. 13 Timing of the BEE announcements ........................................................................ 14 2.4 Review of event study methodology and significant announcements ................. 15 2.4.1 Event study methodology........................................................................................ 15 2.4.2 Some critical assumptions of event studies ............................................................ 16 2.5 Conclusion to the literature review ....................................................................... 17 CHAPTER 3: RESEARCH HYPOTHESES ........................................ 18 Hypothesis 1 ................................................................................................................. 18 Hypothesis 2 ................................................................................................................. 18 Hypothesis 3 ................................................................................................................. 19 Hypothesis 4 ................................................................................................................. 19 CHAPTER 4: RESEARCH M ETHODOLOGY ..................................... 20 4.1 Research Approach ............................................................................................. 20 4.2 Research strategy ................................................................................................ 20 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 An event .................................................................................................................. 21 Event Window ......................................................................................................... 21 Return Estimation ................................................................................................... 22 Expected return....................................................................................................... 23 Abnormal Returns (AR)........................................................................................... 24 vi | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
4.3 4.4 4.5 4.6 4.2.6 4.2.7 4.2.8 4.2.9 Actual return ............................................................................................................ 25 Average Abnormal Return (AAR)............................................................................ 25 Cumulative Average Abnormal Returns.................................................................. 26 Significance test ...................................................................................................... 26 Unit of analysis..................................................................................................... 27 Population ............................................................................................................ 28 Sample size and sampling method ...................................................................... 28 Data collection ..................................................................................................... 29 4.6.1 Identification event date .......................................................................................... 29 4.6.2 Exclusion confounding events ................................................................................ 30 4.6.3 Final events list and share prices............................................................................ 30 4.7 Data analysis ....................................................................................................... 31 4.8 Reliability and Validity .......................................................................................... 31 4.9 Research limitations............................................................................................. 31 CHAPTER 5: RESULTS .................................................................. 33 5.1 Introduction of results........................................................................................... 33 5.2 Description of the sample .................................................................................... 33 5.3 Hypothesis 1: Testing for Average Abnormal Returns ......................................... 36 5.4 Hypothesis 2: Testing for CAR performance ....................................................... 38 5.5 Hypothesis 3: CAR performance based by age of company ............................... 43 5.6 Hypothesis 4: Timing of BEE announcements..................................................... 47 5.7 Conclusion ........................................................................................................... 49 CHAPTER 6: DISCUSSION OF RESULTS ........................................ 50 6.1 Introduction .......................................................................................................... 50 6.2 Hypothesis 1: Testing for Average Abnormal Returns ......................................... 50 6.3 Hypothesis 2: CAR performance ......................................................................... 51 6.4 Hypothesis 3: CAR performance based by age of company ............................... 52 6.5 Hypothesis 4: Timing of BEE announcements..................................................... 53 6.6 Conclusion ........................................................................................................... 54 CHAPTER 7: CONCLUSION ........................................................... 55 7.1 Introduction .......................................................................................................... 55 7.2 Summary of the findings ...................................................................................... 55 7.3 Recommendation for practice .............................................................................. 57 7.4 Recommendations for future research................................................................. 58 7.5 Chapter Summary ................................................................................................ 58 REFERENCES ................................................................................. 59 APPENDICES .................................................................................. 65 vii | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
LIST OF TABLES
Table 4-1: Control Portfolios ......................................................................................... 22 Table 4-2: Keywords used in initial search ................................................................... 30 Table 5-1: Descriptive statistics of the full sample ........................................................ 34 Table 5-2: t-test - positive AARs at 5% level ................................................................ 36 Table 5-3: t-test where AAR is negative at 5% ............................................................. 37 Table 5-4: Negative 10-day CAR (sig. at 5%) ............................................................... 42 Table 5-5: Correlation: Old vs. BEE Miners .................................................................. 46 Table 5-6: Paired t-test: Old vs. BEE Miners ................................................................ 46 Table 5-7: Correlation: pre vs. post amendment of mining charter ............................... 48 Table 5-8: Paired t-test; pre vs. post amendment of mining charter ............................. 49 viii | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
LIST OF FIGURES
Figure 2-1: The contribution of mining to South Africa over the past decade expressed
in 2012 real money terms* ...................................................................................... 8 Figure 2-2: Percentage of mining revenue per commodity* ........................................... 9 Figure 2-3: Market capitalisation of the top-10 mining companies (R ‘billions) ............. 10 Figure 5-1: Histogram for AAR (Equally weighted) ....................................................... 34 Figure 5-2: Chi-squared table – full sample .................................................................. 35 Figure 5-3: Bar graph of AARs for the full sample ........................................................ 37 Figure 5-4: Bar graph of AARs for t-20 to t20 .................................................................. 38 Figure 5-5: Long-term CAR: t-40 to t+240 ......................................................................... 39 Figure 5-6: Short-term, t-20 to t+20: AAR & CAR ............................................................. 40 Figure 5-7: 10day CAR histogram ................................................................................ 43 Figure 5-8: Equally Weighted CAR t-40 to t+240 for New vs. Old Miners ......................... 44 Figure 5-9: Old vs. BEE Miners – AAR & CAR t-20 to t+20 .............................................. 45 Figure 5-10: Pre vs. post amendment of mining charter ............................................... 47 ix | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
LIST OF EQUATIONS
Equation 1: Expected Return ........................................................................................ 24 Equation 2: Abnormal Return (AR) ............................................................................... 24 Equation 3: Actual Return ............................................................................................. 25 Equation 4: Average Abnormal Return (AAR) .............................................................. 25 Equation 5: Cumulative Average Abnormal Return (CAR) ........................................... 26 Equation 6: t-stat for AAR ............................................................................................. 26 Equation 7: t-stat for CAR ............................................................................................. 27 x|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 1:
INTRODUCTION TO THE RESEARCH
PROBLEM
1.1
Research title
The impact of Black Economic Empowerment transaction announcements on share
price performance of JSE Listed Mining Companies.
1.2
Introduction
The South African mining stakeholders committed themselves to achieving a minimum
target of 26% ownership of the South African mining and minerals industry by
Historically Disadvantaged South Africans (HDSA) by 2014. This target was
established in the Mining Charter to enable a change in racial and gender disparities
prevalent in the ownership of South African mining and minerals industry (Department:
Mineral Resources, 2010).
The objective of this research was to assess the impact of Black Economic
Empowerment (BEE) announcements relating to equity ownership by HDSA on share
price performance of South African Mining Companies. Furthermore, this research
assessed whether the introductions of the Mining Charter had an impact on the share
price performance of companies.
The study employed the well-established event study methodology (Kothari & Warner,
2007) that assessed whether investors in the South African mining stocks listed on the
Johannesburg
Stock
Exchange
(JSE)
have
benefited
from
transformational
transactions as guided by the Mining Charter.
In 2004, the South African Government and Mining Industry recognised that one of the
means of ensuring greater participation and benefit for HDSA's in the mining industry
was by encouraging greater ownership of mining industry assets by HDSA's; other
means include holding majority control (50% plus 1 vote) that include management
1|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
control and collective investment, or by using Employee Share Ownership Plans
(ESOPS) and mining dedicated unit trusts. The use of Strategic Joint Ventures (SJVs)
was also proposed as one of the means of achieving ownership and participation of the
HDSA’s in the Mining Sector (Scorecard for the Broad Based Socio-economic
Empowerment Charter for the South African Mining Industry, 2004).
The impending 2014 deadline for achieving the 26% ownership level (Cawood, 2004;
Hamann, Khagram, & Rohan, 2008; Republic of South Africa, 2004) makes this
research relevant, as it demonstrates the degree to which the transformation of the
ownership landscape of South African mining assets has impacted on shareholders.
This study contributes to the review of the success of Black Economic Empowerment,
as South Africa assesses the successes and failures of the first 20 years of the postApartheid era, while seeking guidance on methods to overcome growing inequalities
within the population and economy.
It is hoped that the findings of the study can assist relevant industry stakeholders in
assessing the impact of Black Economic Empowerment on shareholder value.
1.3
Background of Study
Before 1994, the government held South African (SA) mineral rights and few mining
companies dominated the mining industry. After the first South African democratic
elections in 1994, the government embarked on pursuing Black Economic
Empowerment (BEE) initiatives designed as a direct intervention to redistribute assets,
and to create opportunities required to resolve the economic inequalities created by the
Apartheid Government, which had historically favoured white business owners and
multinational corporates (MNCs) rather than benefitting the majority of the black
population (Ribane, 2011). The government promulgated Acts which were intended to
promote economic transformation in South Africa by encouraging meaningful
participation of black people in the economy (Republic of South Africa, 2004).
Between 1994 and 2004, South Africa witnessed the emergence of a handful of
prominent and politically connected black mining entrepreneurs, mainly through the
disproportionate transfer of shares to enrich these few connected individuals. These
2|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
few politically connected individuals have amassed wealth from empowerment
transactions and accompanying directorships (Tangri & Southall, 2008). To further
inclusivity in the ownership of mineral rights and mines, Broad-Based Black Economic
Empowerment (BBBEE) and the Mining Charter legislations were enacted (Republic of
South Africa, 2004). The Charter and the Act were intended to enforce changes in the
way that mining houses operated and were required for these businesses to retain their
licenses to operate.
During 2012, the South African mining sector accounted for 24,7% (R1,8 trillion) of the
JSE’s all-share index, and the industry spent 80% of its R488 billion expenditure within
South Africa. The mining sector is a significant contributor to the South African
economy, the multiplier effect of its fixed investment is estimated at 25% of the
country’s total economy. This sector remains a major contributor to the economy, with
significant contributions to employment numbers, export earnings, attracting foreign
direct investment, creating GDP and contributing to proper, measured and sustained
transformation of the economy (Chamber of mines of South Africa, 2013).
1.4
Research Problem
This research examined the share price performance of mining stocks listed on the
Johannesburg Securities Exchange (JSE) by tracking the stocks’ share price
performance after announcements relating to black empowerment transactions. The
scholars in this field lamented that further studies are required to understand BEE
within mining industry (Fauconnier & Mathur-Helm, 2008; Ribane, 2011; Wolmarans &
Sartorius, 2009). Therefore this study addressed this significant gap in research,
especially within the South African context.
3|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
1.5
Research objectives
The objectives of this study were to:
•
Determine whether announcements of BEE transactions in the long-term and
short-term lead to better shareholder wealth creation in the mining sector.
•
Determine whether BEE announcements have a greater positive impact on the
cumulative abnormal returns of BEE Mining companies that were listed on the
JSE post-1994 compared to their large market cap counterparts.
•
Determine whether the early BEE announcements made before the release of
the Mining Charter in September 2010 have a greater positive impact on the
Cumulative Abnormal Returns of Mining companies compared to those made
after the amendment to legislation.
1.6
Research Scope
The research scope included the reviews of the performance of mining companies
listed on the JSE. Similar studies were conducted on the JSE companies by Sartorius
and Botha (2008) and Ward and Muller (2010), however these studies covered all the
stocks listed on the JSE. Previous research studied samples of between 72 and 175
JSE listed companies. However, this specific research study only focus on mining
stocks and the study covered 66 companies that are classified as Resources within the
economic grouping and industrial sector of mining on the JSE. Shares listed on the
JSE are categorised into one of the three sectors that are consistent with the South
African (SA) sector categories, namely Resources, Financials and Industrials, based on
their revenue. The SA sector classification is derived from the Industry Classification
Benchmark (JSE, 2014; Sharenet, 2014).
4|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
1.7
Outline of Research Study
CHAPTER 1: This chapter introduced the research problem and exhibited the need for
the research, and stated the research objectives. The chapter has contextualised the
need for the research by including the relevant background and concluded by defining
the scope of the research.
CHAPTER 2: This chapter presents an argument within academic literature that
demonstrates the need for this specific research. Relevant literature has been used to
reveal the intricacies of the topic, by considering various points of argument. It also
covers the review of literature on the theories and application of the measuring
instrument.
CHAPTER 3: In this chapter, the purpose of the research is outlined and the formulated
hypothesis presented.
CHAPTER 4: This chapter outlays the research design and methodology. The details
of the population, sample size and sampling method as well as the research instrument
are discussed. It confirms the data collection methods, and discusses the processing
and analysis of the data. The chapter concludes by emphasising the few limitations of
the research.
CHAPTER 5: This chapter presents a summary of the sample and the findings of the
research by displaying tables and figures with limited commentary.
CHAPTER 6: Chapter analyses the data with the intention of interpreting, discussing
and analysing the findings by connecting the primary findings to the literature review.
CHAPTER 7: The research study concludes the research to satisfy the aims and
objectives of the study. It emphasises the main findings of the research and provides
feasible recommendations for future research.
5|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 2:
LITERATURE REVIEW
2.1
General background information
2.1.1
South Africa’s Black Economic empowerment
Following the successful transition of South Africa to Democracy in 1994, the South
African government and the public at large became increasingly frustrated with the
slow pace of social and economic transformation. The resulting pressure led to the
conceptualisation of Black Economic Empowerment (BEE), and the establishment of
the BEE commission in 1998 (Hamann et al., 2008).
The conclusions of the BEE Commission (report published in 2001) called for the
government to intervene through policies and to facilitate the meaningful participation
of black South Africans in the mainstream economy (Hamann et al., 2008). Following
the BEE Commission report, the mining industry through the Department of mineral
Resources (then, Department of Minerals and Energy) introduced the Mining Charter.
The Mining Charter was released in October 2002 and it outlined fundamental focus
areas and guidance regarding how the mining industry could expand opportunities for
HDSA. The pertinent issues included: ownership of mining assets, Employment and
participation in management, worker and community participation and the sharing of
benefits flowing from the south African mining industry (Cawood, 2004).
As part of the Mining Charter the BEE scorecard was introduced to ensure the
fulfilment of the requirements contained in the Broad Based Socio-Economic
Empowerment Charter for the Mining and Minerals Industry. Its objectives, amongst
others, include: Promotion of equitable access to the country’s mineral resources,
increased participation of HDSA's in mining and the advancement of the social and
economic welfare of mining communities and the major labour sending areas
(Scorecard for the Broad Based Socio-economic Empowerment Charter for the South
African Mining Industry, 2004).
6|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
2.1.2
Broad-Based Black Economic Empowerment Act
The South African government promulgated the Broad-Based Black Economic
Empowerment Act to promote the achievement of the constitutional right to equality, to
increase broad-based and effective participation of black people in the South African
economy, to promote a higher economic growth rate, to increase employment and
opportunities and to increase more equitable income distribution. Other achievements
include the establishment of a national policy on broad-based black economic
empowerment to promote the economic unity of the nation, to protect the common
market, and to promote equal opportunity and equal access to government services
(Republic of South Africa, 2004a).
Black people or the HDSA refers to persons, category of persons or community,
disadvantaged by unfair discrimination before the Constitution of the Republic of South
Africa, 1993 (Act No. 200 of 1993) came into operation (Republic of South Africa,
2004b).
In evaluating the broader impact of BEE in redressing past economic injustices, van
der Berg, Burger, Burger, Louw, and Yu (2006) noted that little has been done in the
area of poverty alleviation other than expanded social grants. With regard to education,
whilst equality in State funding for teachers at all schools exists, teacher skills,
governance and resource availability at black schools remains problematic.
The introduction of BEE has been significant in the mining sector. The forced Joint
Ventures associated with Black Economic Empowerment (BEE) boosted Foreign Direct
Investment (FDI) in the mining sector (Japarov, 2012). South Africa’s net FDI inflows in
2011 were 19% of the country’s total net inflows of R46,7 billion.
The industry invested in expanding the production capacity of platinum and iron ore
mines in anticipation of increased future demand. The declining South African gold
mining sector has led to mergers and acquisitions of domestic companies as these
seek growth in new international destinations. Essentially, this has led to an increase in
outward investments because of a lack of local greenfield opportunities (South African
Reserve Bank, 2012).
7|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
2.2
South African Mining
The South African mining industry is the fifth largest in the world (Chamber of Mines of
South Africa, 2012). With a Citibank-estimate of US$2.5 trillion of mineral resource
base (Antin, 2013), the mining sector is set to play an important role in the future of the
country. In terms of reserves, the country has been classified as the primary producer
of platinum group metals (PGMs), manganese, chromium and gold. Although mining's
contribution to the national GDP has fallen from 21% in 1970 to 6% in 2011, it still
represents approximately 60% of exports (Leon, 2012). Figure 2-1 below depicts the
contribution of the mining industry to the South African economy between 2001 and
2012.
Figure 2-1: The contribution of mining to South Africa over the past decade expressed in
2012 real money terms*
*Source: Chamber of mines of South Africa, 2012
South Africa’s top four mineral commodities in terms of sales and employment have
been coal, platinum group metals (PGMs), gold and iron ore (PricewaterhouseCoopers,
2014). Globally, South Africa is classified in the following positions: Number one in the
production of chrome, manganese, platinum, vanadium and vermiculite; second in the
production of Ilmenite, palladium, rutile and zirconium, and South Africa is the world's
third largest coal exporter and now the fifth largest producer of gold.
8|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Figure 2-2 below highlights the percentage split of how various commodities
contributed to the South African mining revenue for the years 2013 and 2014. Gold
which used to be South Africa’s biggest foreign earner, saw its production output
halved in the decade leading to 2014 (PricewaterhouseCoopers, 2014).
Figure 2-2: Percentage of mining revenue per commodity*
* Source: PricewaterhouseCoopers (2014), Stats SA
Mining-related products accounted for up to 25% of the output of the manufacturing
sector in 2012. Performance of the mining sector therefore has a direct impact of South
African manufacturing sector (South African Reserve Bank, 2012).
The majority of the South African mining sector is privately owned, and the state
currently owns a few mining firms. The Mining Charter calls for 26% full shareholder
rights as the minimum target for effective HDSA ownership. Companies have disclosed
that
this
target
has
been
reached,
and
in
most
cases,
exceeded
(PricewaterhouseCoopers, 2014).
The mining industry has played a critical role in the economic development of South
Africa. Rogerson (2011) cited Crankshaw (2002), Department of Minerals and Energy
(2008) and Mabuza (2009) as having identified that mining assumed the status of key
driver of the South African economy for at least half a century.
The study by PricewaterhouseCoopers (2014) show as per Figure 2-3 that two of the
top ten mining companies by market capitalisation are BEE miners. BEE miners refer
9|Page
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
to HDSA mining companies, thus companies that are owned or controlled by
historically disadvantaged South Africans ( Republic of South Africa, 2004b)
Figure 2-3: Market capitalisation of the top-10 mining companies (R ‘billions)
* Source: PricewaterhouseCoopers (2014), Stats SA
10 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The Fraser Institute’s annual Survey of Mining Companies 2014 highlighted investor
concerns when it placed South Africa at position 64 out of 112 jurisdictions for policy
potential and position 53 for investment attractiveness (Wilson & Cervantes, 2014). The
report covered national policies and mineral resources, like South Africa’s BBE policies
in its assessment of the perceptions of the mining companies surveyed.
2.3
Modes of equity transfer and announcements
Wolmarans and Sartorius (2009) conducted a study on the short-term financial impact
of 125 BEE transactions involving 95 companies. Their study identified three different
types of transactions, namely the sale of equity to a BEE company, the purchase of a
stake in a BEE company and other BEE transactions using Strategic Joint Ventures or
partnerships. Their study concluded that the type of BEE transaction had no impact on
explaining the differences between the performance of shares and the creation of
wealth for shareholders. Their study established that there were differences in the
impact on value creation when the different years of announcements were considered.
BEE transactions between 2002 and 2005 had no significant positive impact on
shareholder value creation, but for 2006 it had a significantly positive impact over both
the three-day and the five-day windows.
None of the studies conducted thus far have specifically focused on mining stocks,
hence the aim of this study was to explore the mining stocks. In the South African
context, this is significant because of historical importance of mining to the South
African economy, its current contribution to GDP, the foreign earnings resulting from
the mining sector and the weighting of the mining sector within the JSE. Furthermore,
most event studies on the JSE have either corrected or noted the effect of mining
sector in the study as being significant (Ward & Muller, 2010; Wolmarans & Sartorius,
2009; Wolmarans, 2012).
A study that focuses on mining shares would be significant since the total market
capitalisation of mining companies listed on the JSE has grown substantially to the
current R2.82-trillion of the JSE’s total R12.19-trillion market capitalisation (Kotze,
2014).
11 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
2.3.1
Impact of BEE announcements on cumulative returns
The research studies that have investigated the impact of the cumulative returns are
considered from both the short-term and the long-term perspectives. The majority of
documents revised during the literature review refer to short-term studies of the market
and are in the range of the three to eleven day window, while those that ran long-term
studies referred to windows of 21 days and more.
The short-term and long-term perspectives are discussed below.
2.3.2
Short-term cumulative returns
An event study of 254 BEE transactions between 1996 and 2006 was performed by
Strydom, Christison and Matias (2009). Their study examined market reactions to BEE
transactions. Their study was not conclusive, as it was found that there was a
statistically insignificant positive market reaction to BEE transactions over the 11 days
event window. However, the authors concluded that there was no evidence of a
negative of a negative market reaction to BEE transactions (Strydom, Christison &
Matias, 2009).
One of the questions addressed by Wolmarans and Sartorius (2009) was whether
announcements of BEE transactions are related to shareholder value creation. Their
results showed shareholder wealth creation over a three-day window for the 125 BEE
transactions that were analysed. Their study found that there were significantly positive
average abnormal returns for the day before and a day after the event, with return of
1.15 percent.
Wolmarans and Sartorius (2009) concluded that South African companies were using
BEE transactions as an important vehicle to give expression to their Corporate Social
Responsibility (CSR) objectives. (Alessandri et al., 2011; Jackson, Alessandri, & Black,
2005) also contended that BEE transactions represent CSR actions, as they created
strategic benefits for the organization by serving both the firm’s business interests and
the interests of salient stakeholders.
12 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
2.3.3
Long-term cumulative returns
A research paper that examined 118 BEE announcements on the JSE found that
companies that made BEE announcements prior to May 2005 performed worse than
those who followed (Ward & Muller, 2010). This study was focused on long-term share
price reaction to BEE and concluded that, generally, the BEE-related stocks had a
positive cumulative abnormal return of 10% after the first year. The positive results
were however confined to smaller companies (market capitalisation of less than
R3,5bn), whilst large companies experienced marginally negative cumulative abnormal
return (Ward & Muller, 2010).
Jackson, Alessandri, and Black (2005) used an event study to measure the impact of
announcements of BEE transactions on share prices on a sample of 20 JSE listed
companies. The authors utilised a market model where they estimated betas over the
200 trading days prior to the announcement. Over a five-day event window they found
significant positive cumulative abnormal returns of 1,8%, suggesting that the market
rewarded such transactions. In the year following the announcements they found that
BEE firms out-performed an equally weighted index by 31%.
Jackson et al. (2005) also noted that BEE transactions were completed with an
average discount of almost 10% to the ruling share price of the relevant company. As
the authors noted however, their research was limited by a small sample size and may
have benefited from a control-portfolio model that eliminated market effects.
2.3.4
Link between the age of the company and its share price
performance
In a paper that outlined the history and development of BEE in South Africa, Ponte,
Roberts, and van Sittert (2007) demonstrated that black control of JSE listed
companies, measured in terms of share of market capitalisation, peaked at 9,6% in
1999 and dropped to 5,8% in 2005. These authors ascribed the decline to poorly
structured empowerment deals with high gearing and over-priced assets.
Ponte, Roberts, and van Sittert (2007) also noted that during the early years of
implementation of the BEE period, the biggest South African multinational companies
like Old Mutual, SAB, Liberty Life, Anglo-American and de Beers relocated their
13 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
headquarters outside of South Africa. This was done presumably to place their major
assets beyond the reach and recall of the post-apartheid South African government.
Ward and Muller (2010) found a sub-sample of large companies that had a marginally
negative cumulative annual return (CAR), while the sub-sample of smaller companies
had a strong positive CAR. They also concluded that large-cap companies on the JSE
were predominantly resource companies. They attributed the performance of the largecap companies to their export-oriented business as these companies sell their
commodities in international markets where they derive little or no benefit from BEE
compliance.
Similarly, Chipeta and Vokwana (2011) found that firm characteristics such as size and
age are important determinants of short-term profitability post the BEE transaction.
2.3.5
Timing of the BEE announcements
The previous studies assessed the effects of announcements at the different points in
time. A study by Chipeta and Vokwana (2011) also assessed the effects of
announcements made in different market cycles; they studied abnormal market returns
under Bull and Bear market conditions. Their study concluded that investors reacted
more positively to BEE announcements during a Bear market and negatively to BEE
announcements during the Bull phase of the market cycle. They made an observation
that the findings of their study contradicted other international studies regarding the
effect of timing equity issues.
A study conducted within a South African context by Ward and Muller (2010) found that
the timing of the BEE transaction had an impact on cumulative abnormal returns of JSE
companies. They concluded that the ‘first-movers’ had no performance advantage over
the deals that were announced from and after May 2005, and that more recent deals
performed comparatively better. Essentially, Ward and Muller (2010) had created two
artificial market cycles, the Early and Late BEE deals.
In a quest to determine the impact of the timing of the BEE announcements on the
share price, Ward and Muller (2010) considered the impact of announcements on
transaction made early (prior to May 2005) and late (post May 2005).
14 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The introduction of the amended Mining Charter in 2010 was aimed at transforming
the mining industry to correct past injustices created by apartheid. The looming
deadline for compliance with the Charter and its precursors has been and remains a
concern for mining companies in South Africa (PricewaterhouseCoopers, 2014). There
is opaque knowledge regarding the impact of the introduction of the mining charter on
the share price performance of JSE listed mining companies.
The mining charter was designed to effect sustainable growth and one of its intentions
was to substantially and meaningfully expand opportunities for HDSA to enter the
mining and minerals industry and to benefit from the exploitation of the South African
mineral resources (Department: Mineral Resources, 2010).
It was thus an expectation that the introduction of the mining charter would be of
benefit to the previously disadvantaged (BEE) miners and might influence the
performance of the share price in a positive way. Thus, the BEE announcements made
post amendment and introduction of the mining charter would have had a better impact
on the cumulative abnormal returns of the BEE miners as compared to the pre
announcement.
2.4
Review
of
event
study
methodology
and
significant announcements
2.4.1
Event study methodology
In reviewing event study methodology, Corrado (2011) established that this method of
study was introduced to a broad audience in 1968 by Ball and Brown. Event study
methodology has since been used extensively and has been widely published.
Kothari and Warner (2007) conducted a meta-analysis study in which they reported a
conservative figure of 565 articles that were published in five major finance publications
between the years of 1974 and 2000. Furthermore, Kothari and Warner (2007)
provided an overview of event study methods, and concluded that short-horizon
methods are reliable, while the reliability of long-horizon methods has been improving.
15 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Additionally, Wårell (2007) found the basic event study methodology to be relatively
simple and uncomplicated when following the step-by-step procedure for applying the
event study methodology, as meticulously explained by Henderson (1990). The initial
step is to identify the date upon which the market would have received the news of the
transaction being done. The second step estimates the normal returns of the stocks
being studied based on historic price observations before the news of the transaction.
The third step calculates the abnormal return (AR) for each firm by calculating the
difference between observed returns and the estimated normal returns for each firm.
The fourth step is to aggregate the abnormal return (AR) over time to find the
cumulative abnormal return (CAR) over the event window. The Fifth step is to perform
statistical tests to determine whether or not the abnormal returns are significant and, if
so, for how long (Henderson, 1990).
Ward and Muller (2010) used event study methodology to study long-term share price
reactions to Black Economic Empowerment announcements on the JSE. Other authors
have also used event study methodology to study BEE-related transactions on the JSE
(Jackson et al., 2005; Strydom et al., 2009; Wolmarans & Sartorius, 2009). These
studies focused on the effect of announcement of BBBEE ownership transactions on
company performance measured through indicators of market (JSE) performance.
2.4.2
Some critical assumptions of event studies
Event studies are grounded in some assumptions, and follow market efficiency theory
that share prices adjust rapidly to the information. Tests of market efficiency involve the
analysis of the behaviour share prices following a market event (Bowman, 1983).
This particular research study was not focussed on the information content of earnings
announcements, and thus it would have been fruitless to select a previously
unexplored event. Although an event study, similar to what was conducted in this
research project could be used for market efficiency testing (Bowman, 1983), this was
not an explicit objective of the researcher. The research study focused on clarifying and
resolving the conflict presented by anomalous results at the most basic levels.
16 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
2.5
Conclusion to the literature review
A research project focussing on the impact of BBE announcements on the South
African context was necessary because the previous studies (Alessandri et al., 2011;
Strydom et al., 2009; Ward & Muller, 2010; Wolmarans & Sartorius, 2009) did not focus
on this important sector of the economy. Previous studies were concerned about the
JSE as a whole, and some samples of the studies were found to have excluded
resource (mining) shares.
The event study methodology was found to be the most appropriate for the research
project to achieve its objectives. This methodology has also proved its reliability over
time (Alessandri et al., 2011; Corrado, 2011; Henderson, 1990; Kothari & Warner,
2007; Strydom et al., 2009; Ward & Muller, 2010; Wårell, 2007; Wolmarans &
Sartorius, 2009).
17 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 3:
RESEARCH HYPOTHESES
The objective of this research was to examine the impact on shareholder returns
following the announcement of Mining Empowerment deals affecting equity of JSElisted mining companies for the period of 2000 to 2014.
Hypothesis 1
Null (H10): BEE announcements relating to equity issuance made through the Stock
Exchange News Service (SENS) of the JSE result in no Average Abnormal Returns
(AARs) within the event window.
𝐻! : 𝐴𝐴𝑅 = 0
Alternative (H1A): BEE announcements relating to equity issuance made through the
Stock Exchange News Service (SENS) of the JSE show significant AARs within the
event window.
𝐻! : 𝐴𝐴𝑅 ≠ 0
Hypothesis 2
Null (H20): BEE announcements relating to equity issuance made through SENS have
no impact on the Cumulative Abnormal Returns (CARs) of mining companies.
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 = 0
Alternative (H2A): BEE announcements relating to equity issuance made through SENS
have a positive impact on the CARs of mining companies.
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 ≠ 0
18 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Hypothesis 3
Null (H0): The Average Abnormal Return (AARs) of the new (BEE) mining companies
post BEE announcements relating issuance of equity is not greater than the AAR of the
old mining companies.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" ≥ 0
Alternative (H1): The Average Abnormal Return (AARs) of the new (BEE) mining
companies post BEE announcements relating issuance of equity is greater than the
AAR of the old mining companies.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" < 0
Hypothesis 4
Null (H40): The average abnormal returns of the events made before the release of
amended mining charter in September 2010 are not less than the average abnormal
returns of the events made after the amendment of the mining charter.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"#$ ≥ 0
Alternative (H4A): The average abnormal returns of the events made before the release
of amended mining charter in September 2010 are less than the average abnormal
returns of the events made after the amendment of the mining charter.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"#$ < 0
19 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 4:
4.1
RESEARCH METHODOLOGY
Research Approach
The main aim of this study was to determine the impact of the BEE announcements on
share prices as guided by the hypothesis derived from the existing theories. Therefore
the suitable approach for this study was the positivism or the quantitative approach,
which allows hypothesis testing using numerical data (Saunders, Lewis, & Thornhill,
2009).
4.2
Research strategy
This study used the event study methodology. Mitchell and Netter (1994) explained
event study methodology as a statistical technique that estimates the stock price
impact of occurrences such as mergers, earnings and announcements. Mitchell and
Netter posited that the event study methodology would disentangle the effects of two
types of information on stock prices-information that are specific to firms under
investigation (e.g. dividend announcements) and information that is likely to affect stock
prices market wide (e.g. change in interest rate).
As observed by Wårell (2007), the basic event study methodology is said to be
relatively uncomplicated when following the step-by-step procedure for applying the
event study methodology, as detailed by Henderson (1990).
The approach to event study was based on estimating a market-related return for a
company, before and after a specified event. It involves calculating abnormal returns
for a specified period before and after the event that was being studied. These
abnormal returns were assumed to reflect the stock market’s reaction to the arrival of
the new information pertaining to the event (Corrado, 2011; Lyon, Barber, & Tsai, 1999;
Wolmarans & Sartorius, 2009).
This current research project has developed and added to the study performed by
Ward and Muller (2010); while they studied all the JSE stocks, the current study
20 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
focused only on mining stocks. Ward and Muller (2010) used event study methodology
to analyse long-term share price reactions to Black Economic Empowerment
announcements on the JSE.
4.2.1
An event
In respect of this study, an event was identified as an announcement made by
companies listed on the JSE relating to BEE, where ownership and equity is affected.
JSE rules compels listed companies to declare material information that may impact
share prices to all shareholders through its Stock Exchange News Service (SENS)
(Ward & Muller, 2010). These announcements are made through all major stock
exchanges, which is consistent with local regulations, and mandate for material
disclosures to be made (Neuhierl, Scherbina, & Schlusche, 2010).
MacKinlay (1997) found a relationship between the nature of news and the resulting
CAR; he found “bad” news to result in negative CAR by causing the share prices to
decrease, while “good” news results in positive CAR that causes the prices to increase.
Neuhierl et al. (2010) studied the market reaction to various types of news and
confirmed prior findings regarding strong share price responses to financial news. They
also found significant share price reactions to be consistent with news concerning
corporate strategy, customers and partners, products and services, management
changes, and legal developments.
4.2.2
Event Window
The event window is defined as the period where the actual event occurs (Lefebvre,
2007), it is the event day plus and/or minus some period of interest, either days, weeks
or months during which the returns of a sample firms are studied to examine whether
they behave in an unusual way (Henderson, 1990).
It is important to distinguish between the estimation period and the event window since
the estimates (from the estimation period) are used to define the expected or normal
returns for each firm during the event window (Henderson, 1990).
21 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
In this study, JSE SENS announcements by companies relating to their BEE or
empowerment transactions are considered as events. Thus, event windows would be a
defined period around or relative to the announcement.
The event windows for the purpose of the study were classified as follows:
The short-term included the 3day window (t-1 to t+1) and the 11day window (t-5 to t+5)
and 21day window (S. Brown & Warner, 1985; Wolmarans, 2012).
While the long-term was defined as windows beyond the 21day window (Bhana, 2010;
Kothari & Warner, 1997; Ward & Muller, 2010; Wolmarans, 2012). Kothari and Warner
(2007) found the exact definition of “long horizon” (long-term) to be arbitrary and
generally applied to event windows of 1 year or more.
4.2.3
Return Estimation
The study employed the Control Portfolio model to estimate the expected returns for
each share. Although it is well specified and relatively powerful under various
conditions (Brown & Warner, 1980), event studies using the market model have
previously been found to be inadequate (Ward & Muller, 2010). The Control Portfolio
model was preferred to other economic models such as the Capital Asset Pricing
Model and Arbitrage Pricing Model because of their reliance on assumptions that may
influence the results of the event study (MacKinlay, 1997).
Table 4-1: Control Portfolios
Resources
or
non-
Value
or
growth
Company
Control Portfolio
resources company
company
size
SGN
Non-resources
Growth
Small
SGR
Resources
Growth
Small
SVN
Non-resources
Value
Small
SVR
Resources
Value
Small
MGN
Non-resources
Growth
Medium
MGR
Resources
Growth
Medium
MVN
Non-resources
Value
Medium
MVR
Resources
Value
Medium
LGN
Non-resources
Growth
Large
LGR
Resources
Growth
Large
LVN
Non-resources
Value
Large
LVR
Resources
Value
Large
22 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Table 4-1 demonstrates the classification of the twelve control portfolios as developed
by Ward and Muller (2010). They used three main characteristics to compile the
portfolios, which included whether the organisations were resources or non-resources
companies, as well as whether they were value or growth companies, and the final
characteristic related to company size.
The broad JSE sector groupings were used as criteria to decide whether stocks
represented a ‘resource’ share or not. All mining and non-mining resource shares were
classified as resources while the remainder of the market was classified as nonresources (Ward & Muller, 2010).
A company was classified as a growth or a value investment in terms of its price-toearnings ratio. The price-to-earnings ratios were calculated and classified, after which
the median was determined. All companies with price-to-earnings ratios above the
median were classified into the growth portfolio and those below the median were
categorised into the value portfolio (Ward & Muller, 2010).
A company’s market capitalisation was used to categorise companies by size into
either large, medium or small portfolios. All the companies listed on the JSE were
ranked in descending order of market capitalisation.
The top 40 shares with the largest market capitalisation were grouped into the large
capitalisation control portfolio, those with a market capitalisation ranking between 41
and 100 were grouped into the medium capitalisation control portfolio, and the
remaining companies’ shares were grouped into the small capitalisation control
portfolio (Ward & Muller, 2010).
4.2.4
Expected return
Because of the level of criticism against the Capital Asset Pricing Model (CAPM) model
over time (Lyon et al., 1999; Ward & Muller, 2010), this study employed the control
portfolio approach.
The control portfolio model measures the expected return of sharei in periodt as the
sum of the sensitivity of Sharei to the returns on the twelve control portfolios and a
calculated alpha estimate in periodt.
23 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
This was summarised according to the following equation (Ward & Muller, 2010):
Equation 1: Expected Return
E Rit = αi, t + βi, 1SGNt + βi, 2SGRt + βi, 3SVNt + βi, 4SVRt + βi, 5MGNt
+ βi, 6MGRt + βi, 7MVNt + βi, 8MVRt + βi, 9LGNt + βi, 10LGRt
+ βi, 11LVNt + βi, 12LVRt Where:
E Rit =
Expected return on Sharei on periodt;
α i, t =
Alpha intercepts term of Sharei on dayt;
βi, 1 . . . βi, 12 =
Beta coefficient on each control portfolio return;
SGNt . . . SGRt =
Log-function share price return on each of the twelve control
portfolios on day t.
4.2.5
Abnormal Returns (AR)
Abnormal Returns (AR) represented returns earned by the firm after adjusting for the
“normal” or market-related returns. Simply put, it was the difference between the actual
return of a share and the expected return. AR was represented by the following
equation:
Equation 2: Abnormal Return (AR)
𝐴𝑅!" = 𝑅!" − 𝐸(𝑅!" )
Where:
𝐴𝑅!"
Represent the abnormal return of stocki in periodt
𝑅!"
Represent actual return of stocki in periodt
𝐸(𝑅!" ) Represent the expected return of Sharei in dayt
24 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
4.2.6
Actual return
Actual return was the actual gain or loss the investor would receive from the
performance of a share. It was based on the movement in the daily share price.
Equation 3: Actual Return
𝑅!" = ln[𝑃!" / 𝑃!"!! ]
Where:
𝑅!" is the rate of return on share i on day t,
And 𝑃!" is the price of share i at the end of day t.
4.2.7
Average Abnormal Return (AAR)
The Average Abnormal Return (AAR) was calculated by the sum of AR on a specific
event day divided by the number of AR’s.
Equation 4: Average Abnormal Return (AAR)
1
𝐴𝐴𝑅! =
𝑁
!
𝜔𝐴𝑅!"
!!!
𝐴𝐴𝑅! = Average Abnormal Return
𝑁=
Number of sample returns
𝐴𝑅!" = Average Return
𝜔=
Weighting
The Average Abnormal Return was a weighted average; this study used the equally
weighted (EW) average for analysis of the ARs, however the market-capitalisation
weighted (MCW) average on the complete sample was also provided to place the
sample into context when discussing certain results or outcomes of the study. On EW,
all the ARs had the same weighting of one (1), while MCW of the ARs of the events
were weighted using the JSE market-Cap of the company involved in the event
25 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
(transaction announcement); this is similar to the approach followed by Ward and
Muller (2010).
4.2.8
Cumulative Average Abnormal Returns
The Cumulative Average Abnormal Returns (CAR) was calculated for a firm as the sum
of the AAR over the period in question (Binder, 1998; Jackson et al., 2005; Ward &
Muller, 2010).
The performance of the entire sample was evaluated by calculating the cumulative
average abnormal returns of all the shares included in the sample on each day of the
period under investigation.
Equation 5: Cumulative Average Abnormal Return (CAR)
𝐶𝐴𝑅!! ,!! = 𝑒
𝐶𝐴𝑅!! ,!! =
!!
!!!! !!"!
−1
Cumulative Average Abnormal Return for the sample over the time
interval (t1, t2)
𝐴𝐴𝑅! =
Average Abnormal Returns
4.2.9
Significance test
The Chi-Squared test was used to test for goodness-of-fit, to establish normality of the
sample of AARs.
The hypotheses were tested using the t-test, this commonly used parametric test was
found to be adequate for use in event studies (S. J. Brown & Warner, 1980; S. Brown &
Warner, 1985; Ward & Muller, 2010).
The testing of the Null Hypothesis was done as represented in the equation below:
Equation 6: t-stat for AAR
𝑡!!"! = 𝑛
𝐴𝐴𝑅!
𝑆!!"!
26 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Equation 7: t-stat for CAR
𝑡!""# = 𝑛
𝐶𝐴𝐴𝑅
𝑆!""#
Bootstrapping was selected as an appropriate non-parametric test to support the t-test
(Ward & Muller, 2010). Bootstrapping was used because this method is not typically
used in isolation but rather serves as an inspection of the robustness of conclusions
based on parametric tests (Campbell, Cowan, & Salotti, 2010; MacKinlay, 1997).
Bowman (1983) also found that the use of a non-parametric test as a complement
would enhance the perceived validity of the statistical inferences.
Non-parametric tests are more powerful that standard parametric tests (Campbell et
al., 2010), and are motivated by concerns that data, which is assumed to be normally
distributed under the parametric tests, would lead to poor or imprecise inferences
(Corrado, 2011).
Brown and Warner (1980) found that the t-tests were a better approximation of the
theoretical distribution than some non-parametric tests like the Wilcoxon test. Although
the t-test is prone to event-induced volatility, it was found to be well-specified under a
variety of conditions (S. Brown & Warner, 1985).
4.3
Unit of analysis
The research study used a JSE listed mining company that made an unscheduled
announcement in relation to its BEE transaction between January 2000 and November
2014 as the unit of analysis.
27 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
4.4
Population
Population is defined as the elements about which inferences will be made (Cooper &
Schindler, 2014). The population under consideration for the event study comprised of
all mining shares listed on the JSE over the period between 2000 and 2014.
Companies listed on the JSE are required to make announcements to shareholders of
any material issues that may impact share prices through the Stock Exchange. JSE
Stock Exchange News Service (SENS) is an electronic notice board and information
system designed to ensure that investors and analysts can receive price-sensitive
announcements timeously and simultaneously.
4.5
Sample size and sampling method
The sampling method is the process of selecting some elements from a population to
represent that population while the sample size is the number of the elements from
which the inferences will be made (Cooper & Schindler, 2014). The study had a sample
of 26 mining companies listed on the JSE that made a total of 241 qualifying
announcements through the JSE’s SENS for the period from January 2000 to
November 2014. The list of companies included in the sample is in Appendix 1: List of
mining companies in the sample.
Non-probability sampling was conducted to gather samples from the database. An
array of keywords was used to filter events specific to the mining industry and that were
related to BEE announcements that affected shareholding or equity. The researcher
used knowledge and professional judgment to eliminate announcements that were
initially included in the sample. Information from company websites, industry reports
and past research papers were accessed to triangulate and validate the selected
announcements.
The sampling method used in this study best fits the definition of purposive sampling
(Saunders et al., 2009).
28 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
In purposive sampling, the researcher’s objective is to produce a sample that can be
logically assumed to be representative of the population and would be appropriate for
the study. One of the disadvantages of purposive sampling is that the results obtained
from a sample are subject to some degree of bias (Saunders et al., 2009).
Lyon et al. (1999) noted that the analysis of long-term abnormal returns is
“treacherous” Therefore, an important consideration for event studies, and particularly
for long-term studies, is the selection of a benchmark against which abnormal returns
are estimated. As such, the research study followed an approach similar to that taken
by Ward and Muller (2010) that required four years of share price data prior to the
announcement date for the estimation of betas and a further 250 trading days after the
announcement for the analysis of the abnormal returns. Furthermore data were
collected by removing outliers, and many thinly traded small company shares with
market capitalisation of less than R100m at the event date were removed.
4.6
Data collection
4.6.1
Identification event date
Using a database containing all SENS announcements, a content search was
conducted for all BEE-related announcements for a period extending from 2000 to
2014; from these events a sample of BEE-related transactions was compiled for
analysis.
Data for this study was obtained from the Sharenet database. This was also the
database that housed JSE SENS announcements that was used by Ward and Muller
(2010) for their research on long-term performance of JSE stocks. To access the data
from the database, a search using the keywords that are provided in Table 4-2 below
was conducted.
29 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Table 4-2: Keywords used in initial search
black economic
empowerment
empowerment
transaction
broad-based BEE
b-bbee
BEE
BEE deal
mining
strategic joint venture
joint venture
resources
partners
share price
equity
company name
sector
The initial search generated an extensive list because the keywords used created
many correlations. The researcher then used judgment to narrow down the core
keywords and reduced this list further to include the following keywords: BEE, Black
Economic Empowerment and Joint Venture.
The resulting list of announcements from the database was manipulated by utilising
pivot tables to condense the list to include only mining and resources stocks.
4.6.2
Exclusion confounding events
Follow-up announcements on the SENS that provided updates to an already
announced BEE transaction were excluded from the analysis. Only announcements
that were deemed to be the first to break the news relating to the BEE transaction were
considered.
4.6.3
Final events list and share prices
The final events list was compiled having excluded all compounding events and
announcements that were classified as not significant. The listed events met the
criterion of being the first announcement relating to a specific BEE transaction.
The event list consisted of JSE mining companies, their share code (ticker) and the
date on which the announcement was made. These events where used to retrieve
share prices from the Sharenet database and used to execute the event study.
30 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
4.7
Data analysis
To measure the impact of BEE announcements on JSE mining shares, the research
used a well-established event study. The analysis followed a step-by-step procedure
for applying the event study methodology as detailed by Henderson (1990).
The calculation of the AR was computed using the event study model developed and
maintained by Ward and Muller (2010). The computation followed once the relevant
events were identified and the corresponding share prices of mining companies were
assimilated into the model.
4.8
Reliability and Validity
Validity is the characteristic of measurement concerned with the extent of measures
that measures what the researcher wishes to measure; and that differences found with
a measurement tool reflect true differences among participants drawn from a
population (Cooper & Schindler, 2014: 668). To ensure validity the researcher used
other sources of information to confirm events. Other sources of information that
published BEE announcements like websites (Moneyweb) and specific company web
pages were accessed to ensure the events were interpreted correctly.
Reliability is a characteristic measurement concerned with accuracy, precision and
consistency (Cooper & Schindler, 2014: 664). In order to ensure reliability the
researcher performed the event study through an event study engine (Ward & Muller,
2010) that was previously tested over a long period of time, and used through various
studies.
4.9
Research limitations
Owing to time constraints, the study did not test and contrast the performance of the
types of transactions. By dividing the sample into the three BEE transaction types, as
31 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
identified by Wolmarans and Sartorius (2009), it would have been possible to ascertain
whether the mining shares performed the same irrespective of whether the transaction
was a sale or purchase of equity or a joint venture.
The study was industry specific (Mining & Resources Sector) and thus the selected
companies resulted in a non-probability sample. This type of sampling generally has
bias.
The size of the sample as well as the number of events and the long-term view of the
study made it possible that the sample was impacted upon by confounding events.
Some of companies made announcements within a couple of months of each other,
thus repeat announcements relating to different deals.
The Department of Mineral Resources (DMR) is understood to be insisting that mining
companies should repeatedly enter into new BEE transactions every time an existing
BEE partner exits, so that they maintain their BEE credit. On the other side companies
whose Black Economic Empowerment (BEE) partners have chosen to exit argue that
deals from the past should continue to count towards empowerment credits
(PricewaterhouseCoopers, 2014).
32 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 5:
5.1
RESULTS
Introduction of results
This study employed an event study methodology to calculate the CARs associated
with the public announcement through SENS of BEE transactions by mining companies
from January 2000 to November 2014. The approach to this event study was based on
estimating a market-related return for a specific company, and then calculated the
abnormal returns (ARs) for a certain number of days before and after the event that
was studied. These abnormal returns were assumed to reflect the stock market’s
reaction to the arrival of the new information pertaining to the event.
The CAR was calculated by accumulating the daily average abnormal returns (AARs).
5.2
Description of the sample
The sample covered 26 JSE-listed mining companies, jointly involved in 241 BEE
transactions over a period spanning almost 14 years. The size of the available sample
in this study was adequate for the AARs to tend towards a normal distribution,
consistent with the Central Limit Theorem. The distribution of a large sample is likely to
be normal.
However, running descriptive statistics and plotting the histogram of the AAR confirmed
that the distribution might not have been normal when overlaid with a normal
distribution curve as evidenced Figure 5-1. This necessitated a test for normality using
statistical tools.
33 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Figure 5-1: Histogram for AAR (Equally weighted)
As show in Figure 5-1, the distribution of the AARs was slightly flatter than a normal
distribution with a wider peak, and the values are spread wider around the mean.
Table 5-1: Descriptive statistics of the full sample
EW Full Sample AAR*
Count
Mean
Skewness
Skewness Standard Error
Mean LCL
Kurtosis
Mean UCL
Kurtosis Standard Error
Variance
Coefficient of Variation
Standard Deviation
Mean Deviation
Mean Standard Error
Median
Geometric Mean
Harmonic Mean
* Alpha value (for confidence interval) = 5%
The descriptive statistics showed the sample was skewed to the left; it had skewness
of -0,241 as shown in Table 5-1and has kurtoses of 3,197. The sample kurtoses were
slightly above 3, meaning that it is a Leptokurtic distribution with values concentrated
around the mean and thicker tails. The descriptive statistics were not adequate to
conclude the normality of the sample, thus a Chi-squared test was conducted.
34 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The Chi-Squared test was run to test for goodness-of-fit, to establish whether the
sample of AARs was normally distributed. Figure 5-1 (histogram of the frequency data)
shows that the data was unimodal and was slightly to moderately skewed to the left.
For the Chi-Squared test, the null hypothesis stated that the calculated AARs of the
sample fits the normal probability distribution with the mean of 0,03% and a standard
deviation of 0,24%.
Figure 5-2: Chi-squared table – full sample
Chi-Squared,Calculation
Intervals
Observed,
frequency
Normal,
probability
Normal, Expected,
probability, frequenc
value
y
Chisquared
mean
Standard,
x,-,mean Deviation
z
NORMSDIST(z)
Below,-0,47%
21 P(Below,-0,47%)
,,,,,,,,,,0,047
,,,13,174
,,,,,,,,,4,648
-0,030% -0,440%
0,242%
-1,82 ,,,,,,,,,,,,,0,034
-0,47%,to,-0,19%
19 P(-0,47%,to,-0,19%) ,,,,,,,,,,0,073
,,,20,505
,,,,,,,,,0,111
-0,030% -0,300%
0,242%
-1,241 ,,,,,,,,,,,,,0,107
-0,33%,to,-0,05%
60 P(-0,33%,to,-0,05%) ,,,,,,,,,,0,147
,,,41,250
,,,,,,,,,8,522
-0,030% -0,160%
0,242%
-0,661 ,,,,,,,,,,,,,0,254
-0,19%,to,0,09%
65 P(-0,19%,to,0,09%) ,,,,,,,,,,0,213
,,,59,874
,,,,,,,,,0,439
-0,030% -0,020%
0,242%
-0,082 ,,,,,,,,,,,,,0,467
-0,05%,to,0,23%
51 P(-0,05%,to,0,23%) ,,,,,,,,,,0,223
,,,62,712
,,,,,,,,,2,187
-0,030%
0,120%
0,242%
0,497 ,,,,,,,,,,,,,0,690
0,09%,to,0,37%
46 P(0,09%,to,0,37%) ,,,,,,,,,,0,169
,,,47,400
,,,,,,,,,0,041
-0,030%
0,260%
0,242%
1,076 ,,,,,,,,,,,,,0,859
0,23%,to,0,51%
14 P(0,23%,to,0,51%) ,,,,,,,,,,0,092
,,,25,851
,,,,,,,,,5,433
-0,030%
0,400%
0,242%
1,655 ,,,,,,,,,,,,,0,951
,,,10,171
,,,,,,,,,2,629
-0,030%
0,540%
0,242%
Above,0,51%
5 P(Above,0,51%)
,,,,,,,,,,0,036
Summation:
,,,,,,,,,,,,,,,,,8,000
,,,,,,,,,24,01 Chi-Stat df,=,8,-,1,,=,7
,,,,,281,000
,,,,,,,,,,1,000
,,,280,94
0,11% p-value
Alpha,=,1%
2,235 ,,,,,,,,,,,,,0,987
Chi-crit,=,18,475
p-value,<,alpha
The test returned a p-value of 0,11% (X2 of 24, df=7), which is less than the rejection
level of 1% significance. Thus the null hypothesis was rejected in favour of the
alternative hypothesis. The alternative hypothesis was that the AARs do not fit a
normal distribution is therefore probably true based on the evidence of the test.
It can be concluded with 99% confidence that the AARs do not follow a normal
probability distribution with a mean of 0,03% and a standard deviation of 0,24%.
Having established the nature of the sample, the researcher performed the various
tests to answer the research questions.
35 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
5.3
H ypothesis 1: Testing for Average Abnormal
Returns
The first hypothesis was that BEE announcements relating to equity made through the
Stock Exchange News Service (SENS) of the JSE resulted in no average abnormal
returns (AARs) within the event window. Thus, the AAR should be zero.
The computed ARs for the individual events (announcements) were used to calculate
AARs for each day. The resultant AARs were tested for significance using the t-test,
this was done to examine the significance in distance from zero (0). The important
summaries of the t-test are presented in this chapter, while full outcomes of the results
are presented in Appendix 2: AAR t-stat test for significance.
The outcome of the t-test was that, of the equally weighted the returns for the long-term
window, 11 out of 281 days were found to be significantly positive or negative at the
5% level.
Table 5-2: t-test - positive AARs at 5% level
Two of these were positive, as shown in Table 5-2 while the majority were negative as
evident from Table 5-3. It is worth noting that only one statistic (day 171) after the event
returned positive AAR that was significant at 5%, while the majority (9 days) returned
negative AARs that were significant at 5%.
36 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Table 5-3: t-test where AAR is negative at 5%
Within the short-term window, there are no significant events at 5% levels, except for
day -19 (nineteen days before the event) as shown in Table 5-3.
As shown by Figure 5-3, most AARs plot around the zero percent line, however there
are a few that lie away from zero.
Figure 5-3: Bar graph of AARs for the full sample
A visual examination of the 41-day window in Figure 5-4 shows that up to 50% of the
ARR are above 0,25% and -0,25%.
37 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Figure 5-4: Bar graph of AARs for t-20 to t20
Although 96% (271/281 days) of the AARs were within the +0,5% and -0,5% range, it
does not mean they were insignificant.
There was sufficient statistical evidence to not support the null hypothesis in favour of
the alternative. Therefore, the null hypothesis 1, which tested the impact of BEE
announcements on the cumulative abnormal returns of mining companies in the longterm, is not supported.
The conclusion of the study is that the BEE announcements relating to equity issuance
made through the Stock Exchange News Service (SENS) of the JSE show significant
AARs within the event window.
5.4
H ypothesis 2: Testing for CAR performance
The second hypothesis was that BEE announcements relating equity made through the
Stock Exchange News Service (SENS) of the JSE have no impact on the cumulative
abnormal returns of mining companies.
38 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
To calculate CAR, the AARs were cumulated progressively from day zero (0), going
both to the positive side and the negative side using the exponential summation. The
CAR was cumulated exponentially because the ARs were calculated using the log
returns.
Figure 5-5: Long-term CAR: t-40 to t+240
!
!
!
!
!
!
!
!
!
!
!
!
!
Figure 5-5 plots CAR’s of the complete list of 241 transactions, using the controlportfolio method of calculating CAR. The equally weighted and market-capitalisation
weighted graphs present similar performance from the event day (t0), in that both trend
negatively. From day 70 (t+70), the two graphs show a divergence to t+240, although still
in negative CAR territory the market-capitalisation weighted graph starts to trend
positively while the equally weighted graph continues trending deeper into the negative.
The market-capitalisation weighted CAR accelerated much more quickly in the
negative in the first 10 days, it’s CAR stabilised at an average of -2,5% from t+10 to t80.
For the corresponding period of t+10 to t80, the equally weighted CAR of the whole
sample decreased by 3%.
Beyond day 80, the market-capitalisation weighted graph shows a trend towards
positive cumulative abnormal returns while the equally weighted shows the deepening
of negative cumulative abnormal returns (Figure 5-5: Long-term CAR: t-40 to t+240). The
39 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
market-cap weighted CAR had an average improvement of 2% between day 80 (t80)
and day-240 (t240), while peaking at a positive 0,5% on three occasions between day180(t180) and day-210 (t210).
The equally weighted CAR decreased by another 4% between day 80 (t80) and day-240
(t240), cumulatively recording a loss of approximately 7,5% from the event day (t0).
Figure 5-6: Short-term, t-20 to t+20: AAR & CAR
Figure 5-6 displays the impact of the only significant AAR (at 5%) within the short-term
(day 3), as it results in the peak CAR performance for the study in day-2.
There was positive reaction to the BEE announcement by the market in the 3-day
window, all three days (t-1 to t+1) had positive AARs and was highest on the event day.
The 11-day window was noted to generally have negative AARs, except for the 3-day
window around the event. This could be used to conclude that the market liked the
news of BEE announcements, and that once the information had filtered through the
market, the shares reverted back to the natural trajectory for the period.
The CAR performance over the 21-day window was generally positive; it as negative
only on three days (days 3,4 & 10). It is worth noting that the CAR performance had a
40 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
negative slope, thus trending down before the 21-day window, it then held positive for
this window and thereafter it continued on a downward trend into negative CAR. Almost
60% (12/21) of the days within the 21-day window returned negative AARs. These
points were shown in Figure 5-6: Short-term, t-20 to t+20: AAR & CAR.
In order to test the performance of the CAR, a sample distribution of 10-day CARs was
computed. The t-test was then used to test the significance of the 10-day CAR at a
level of 5%.
The AARs were used to calculate 10day CARs per day. The resultant 10day CARs
were tested for significance using the t-test, this was done to test how significantly they
are form zero (0). The null hypothesis of this 10-day t-test is that the 10-day CAR is
equal to zero.
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 = 0
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 ≠ 0
The important summaries of the t-test are presented in this section while full outcomes
of the results are presented in the appendices. The outcome of the t-test was that, 25
out of 231 days were found to be significantly away from zero at 5% level. The 10day
CAR was computed from day 10 to day 240 (end of window).
Table 5-4: Positive 10-day CAR (sig. at 5%)
EW&Full&Sample& t3stat3
p3value3
103day&CAR&*
10day&CAR 10day&CAR
Event&day
51
0,556%
2,170
3,10%
65
0,992%
3,871
0,01%
80
1,188%
4,635
0,00%
91
0,555%
2,164
3,14%
92
0,918%
3,583
0,04%
162
0,732%
2,858
0,46%
174
0,580%
2,262
2,46%
222
0,892%
3,482
0,06%
233
0,590%
2,303
2,21%
Significant5at55%
Table 5 4: Positive 10-day CAR (sig. at 5%) shows a summary of the results were the
test statistic of 10day CAR was positive at 5% level. It is worth noting that the found no
41 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
positive 10day CAR with the short-term windows of 3, 11 and 21 days. All of the
significant 10day CARs fall in the long-term.
Table 5-5: Negative 10-day CAR (sig. at 5%)
EW&Full&Sample& t3stat3
p3value3
103day&CAR&*
10day&CAR 10day&CAR
36
#0,584%
#2,280
2,35%
56
#0,970%
#3,786
0,02%
66
#0,833%
#3,252
0,13%
71
#0,891%
#3,478
0,06%
74
#0,686%
#2,675
0,80%
81
#0,765%
#2,986
0,31%
89
#0,645%
#2,518
1,24%
106
#0,659%
#2,572
1,07%
143
#0,595%
#2,321
2,11%
165
#0,839%
#3,273
0,12%
178
#0,568%
#2,215
2,77%
180
#0,762%
#2,975
0,32%
186
#0,519%
#2,024
4,41%
188
#0,822%
#3,207
0,15%
213
#0,623%
#2,433
1,57%
215
#0,594%
#2,320
2,12%
Significant6at65%
Event&day
Table 5-5 shows a summary of the results were the test statistic of 10day CAR was
negative at 5% level. Similar to the positive 10day CARs, there was no negative 10day
CAR within the short-term windows of 3, 11 and 21 days. All of the significant 10day
CARs fall in the long-term view.
42 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Figure 5-7: 10day CAR histogram
Figure 5-7 abover shows the distribution of the 10day CARs, with 0,005 shown in
black.
There was sufficient statistical evidence to not support the null hypothesis in favour of
the
alternative.
Therefore,
null
hypothesis
2,
which
postulated
that
BEE
announcements relating equity made through the Stock Exchange News Service
(SENS) of the JSE have no impact on the cumulative abnormal returns of mining
companies is not supported.
It can be concluded from the presented performance of the CAR that BEE
announcements had a negative impact on cumulative abnormal returns of mining
companies. The alternative hypothesis is therefore accepted. The established impact is
however negative, resulting in losses for the shareholders.
5.5
H ypothesis 3: CAR performance based by
age of company
Hypothesis 4 postulated that the impact on CARs of mining companies due to BEE
announcements relating to issuance of equity is not affected by the age of the
company, thus there should be no difference in the CAR performance between the old
43 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
and the new mining companies. In testing hypothesis 4, cumulative abnormal returns of
the whole sample were divided into two groups, one represented a sample of old
mining companies that have operated and existed pre-1994 and the other was a
sample that represented new companies that were formed or reconstituted to benefit
from the BEE laws of the democratic South Africa. The new mining companies (new
miners) can and are being classified in this regard as the beneficiaries of BEE.
For these two groups (sub-samples), their AARs were used to calculate their respective
CARs. The CARs were then plotted to examine how they performed when compared to
each other. Of the 241 events, 105 announcements related to old miners and 136
announcements were classified for BEE miners.
Figure 5-8: Equally Weighted CAR t-40 to t+240 for New vs. Old Miners
&
! &
&
&
&
&
"&
&
&
&
!
"
# ! " # Figure 5-8 shows that both old and BEE miners experienced negative CAR, however,
the old miners tend to fair better from day 60 onwards.
44 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Figure 5-9: Old vs. BEE Miners – AAR & CAR t-20 to t+20
!#!
!
"
"
"
"
!
Figure 5-9 illustrates the CAR performance of the old miners versus the BEE miners in
the short-term. The BEE miners have a better CAR performance meaning that the
market viewed the announcement as positive for the miners.
Conversely, the market did not perceive the announcement as being good for the old
miners; the CAR performance had a negative trend. It should be noted that the
negative trend of the old miners’ CAR started before the event day and continued
thereafter.
A paired t-test was performed to compare the two sub-samples and to ascertain
whether they significantly differ from each other. A test was also run to determine the
correlation between the performances of old miners when compared to BEE miners.
Both tests were performed using SPSS (predictive analytics software by IBM) to
determine whether there is a difference in the means of the two categories of the
mining companies. The correlation was performed first to determine whether the
variance between the two populations is the same, then later the t-test was conducted.
45 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Table 5-6: Correlation: Old vs. BEE Miners
a
Bootstrap for Correlation
95% Confidence
Interval
Std. Error
Lower
Upper
N
Correlation
Sig.
Bias
Old Miners AAR &
281
,031
60,28%
0,020%
BEE Miners AAR
a. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples
5,823%
-7,703%
14,694%
Table 5-6 above indicates that there was a weak positive correlation of 0,031 between
the old miners and the BEE miners. This means the two samples are not correlated as
this figure is very close to zero (0).
This was followed by a more rigorous paired t-test, that was conducted together with
bootstrapping. The null hypothesis of the paired t-test is that the difference between the
AARs of the Old miners and the AARs of the BEE miners is greater or equal to zero
(0).
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" ≥ 0
The alternative hypothesis for this test is that difference between the AARs of the Old
miners and the AARs of the BEE miners is less than zero (0).
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" < 0
The results are shown in Table 5-7: Paired t-test: Old vs. BEE Miners:
Table 5-7: Paired t-test: Old vs. BEE Miners
Paired Differences
Mean
Old Miners AAR BEE Miners AAR
0,0306%
95% Confidence Interval of
the Difference
Std. Deviation Std. Error Mean
Lower
Upper
0,4394%
0,0262%
-0,0210%
0,0822%
t
1,167
df
Sig. (2-tailed)
280
24,4%
Bootstrap a
Mean
Bias
Std. Error
Sig. (2-tailed)
Old Miners AAR 0,0306%
-0,0007%
0,0264%
24,9%
BEE Miners AAR
a. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples
95% Confidence Interval
Lower
Upper
-0,0206%
0,0823%
NB: Variable 1 is new miners and variable 2 old miners.
The results indicates that the t-stat of 1,167; both the p-values for the standard t-test
and the bootstrap were greater than the significance level of 5%, which they were
46 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
tested at. Therefore the null hypothesis is supported at the 5% level of significance and
it can be concluded that the null hypothesis is probably true. The statistical evidence
supported the view that the AARs of the old miners are greater than the AARs of the
new miners.
5.6
H ypothesis 4: Timing of BEE announcements
Hypothesis 5 was designed to test whether the BEE deals that were announced before
the implementation of the Amended Mining Charter that was introduced in September
2010 performed better or worse than those announced after the milestone date.
The sample was divided into two groups, one that represented announcements made
before September 2010 and the second group represented the period thereafter. The
computed CARs of the two sub-samples where then compared to each other.
Assumedly as shown by Figure 5-10, the pre-September 2010 events trended very
closely to the full sample of weighted CAR because those events account for 78% of
the total sample (204 of 262 events).
Figure 5-10: Pre vs. post amendment of mining charter
$
$
$
$
$
$
$
$
$
$
$
! ! 47 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
As evident in Figure 5-10, the CAR performance of events post-September 2010
turned negative from day-1. The CAR of events than were announced after the
amendment of the mining charter lost 7% in the 20 days preceding the event, while the
CAR of the events before September 2010 remained relatively flat, averaging at 0,26%.
After the event day (t0), the post-September 2010 announcements demonstrated a
sharp loss of 5% up to day-20. It recovered slightly between day-20 and day-35 before
plummeting again to reach – 9,4% on day-80. The CAR stabilised to an average of
8,24% between day-80 and day-215, and thereafter saw a late recovery of 4% to the
end of the study window.
Similarly, from day-20, the pre-September 2010 announcements returned negative
CAR to the end of the study window. However the losses were not as volatile and
erratic as the post-September ones. They trend was gently and consistently towards
the negative over time, losing 8,5%.
However the short-term CAR (t0) to day-20 (t20) of the announcements made preSeptember showed positive performance, which meant that those were well received
by the market.
Table 5-8: Correlation: pre vs. post amendment of mining charter
a
Bootstrap for Correlation
95%
N
Correlation
Sig.
Bias
Pre-Charter AAR &
281
-,057
34,33%
0,398%
Post-Charter AAR
a. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples
Std. Error
6,337%
Confidence
Interval
Lower
Upper
-17,644%
7,975%
Table 5-8 above indicates that there was weak negative correlation of -0,057 between
pre- and post- amendment of the Mining Charter in September 2010. This means that
the correlation between the two samples is very weak and has an inverse relationship.
This was followed by a more rigorous paired t-test, which was generated by the
addition of bootstrapping. The null hypothesis of the paired t-test is that the difference
between the AARs of the announcements made pre- compared to those made postSeptember 2010 is greater or equal to zero (0).
𝐻! : 𝜇! ≥ 0
48 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The alternative hypothesis for this test is that the difference between the AARs of the
announcements made pre- when compared to those made post-September 2010 is
less than zero (0).
𝐻! : 𝜇! < 0
The results are shown in Table 5-9: Paired t-test; pre vs. post amendment of mining
charter:
Table 5-9: Paired t-test; pre vs. post amendment of mining charter
Paired Differences
Mean
Pre-Charter AAR Post-Charter AAR
0,0115%
95% Confidence Interval of
the Difference
Std. Deviation Std. Error Mean
Lower
Upper
0,5663%
0,0338%
-0,0550%
0,0780%
t
df
0,340
Sig. (2-tailed)
280
73,4%
Bootstrap a
Mean
Bias
Std. Error
Sig. (2-tailed)
Pre-Charter AAR 0,0115%
0,0006%
0,0342%
76,8%
Post-Charter AAR
a. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples
95% Confidence Interval
Lower
Upper
-0,0547%
0,0745%
The results of the t-test according to Table 5-9 indicated that the t-stat is 0,34; the pvalues for the standard t-test and the bootstrap were greater than the significance level
of 5%, at which they were tested. Therefore the null hypothesis is supported at the 5%
level of significance and it can be confirmed that the null hypothesis is probably true.
The statistical evidence supports the view that the AARs of the events of preSeptember are greater than the AARs of the post-September 2010 events.
However, it should be noted that both sets of announced transactions resulted in
negative CAR in the long-term.
5.7
Conclusion
While the results of the research study were comprehensively discussed in this
chapter, the proceeding Chapter 6 profoundly analyses the results in terms of reflecting
on the information garnered from the literature review, as well as by determining the
effect on the South African mining industry at large.
49 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 6:
6.1
DISCUSSION OF RESULTS
Introduction
This Chapter discusses the results of hypotheses that were tested. The data was
gathered quantitatively from a sample of 26 mining companies that made a total of 241
qualifying announcements. Each of the four hypotheses is discussed and contrasted
against the current literature findings on BEE announcements.
6.2
Hypothesis 1: Testing for Average Abnormal
Returns
The hypothesis postulated the existence of average abnormal returns resulting from
BEE announcements; this was successfully tested.
𝐻! : 𝐴𝐴𝑅 = 0
𝐻! : 𝐴𝐴𝑅 ≠ 0
The t-test was conducted and confirmed the existence of significant AARs within the
event window. The majority of the observed AARs were negative, signalling negative
share price reactions to BEE announcements that lead to the destruction of
shareholder value.
The outcome of the test demonstrated that the results were consistent with that of
Ward and Muller's (2010) previous study. The majority of the AARs were negative and
were intermingled with some positive AARs. Although Ward and Muller (2010) found a
distinction of positive and negative AARs in the short-term when compared to the longterm, this study found no such obvious distinction between the short-term and the longterm.
The results of the study were also found to be different that those of Jackson et al.
(2005); their study returned positive AARs from a sample of 20 JSE companies and
50 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
they found no evidence of negative post-announcement share price performance for
BEE transactions. This confirms that, since mining companies returned negative AARs,
they could have a different response to BEE announcement than the general JSE
market. It is also possible that the study by Jackson et al. (2005) did not note this
phenomenon since their sample of 20 companies potentially had no mining shares
included; instead it had four broad industry categories of financial (six transactions),
consumer services (six transactions), manufacturing (four transactions), and other (four
transactions).
6.3
Hypothesis 2: CAR performance
Null hypothesis 2 that was test, postulated that BEE announcements relating to equity
issuance made through SENS had no impact on the Cumulative Abnormal Returns
(CARs) of mining companies. The hypotheses were tested as follows.
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 = 0
𝐻! : 10𝑑𝑎𝑦 𝐶𝐴𝑅 ≠ 0
The results from testing the second hypothesis affirmed that BEE announcements
related to equity through the SENS have negative impact on the cumulative abnormal
returns of mining companies in the long-term. This is in contrast to the findings of the
study by Ward and Muller (2010), who found the long-term CAR to be positive.
The announcement of BEE deals in mining could be taken as “bad” news, since these
announcements resulted in negative CAR. Bad news generally causes shares to
decrease while good news causes shares to increase (MacKinlay, 1997).
The testing for short-term CAR performance as summarised in Table 5-2: t-test positive AARs at 5% level, Table 5-3: t-test where AAR is negative at 5% and Figure
5-3: Bar graph of AARs for the full sample, confirmed that the results of short term CAR
present enough statistical evidence that BEE announcements had negative impacts on
cumulative abnormal returns of mining companies in the short-term, except for the 3day window around the event.
51 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The findings of this test are contrary to those of Strydom et al. (2009), where the
authors found a statistically insignificant positive market reaction to BEE transactions
over the 11-day event window. However, Strydom et al. (2009) concluded that there
was no evidence of a negative market reaction to BEE transactions. Wolmarans and
Sartorius (2009) also found that there was a significantly positive average abnormal
return for the day before and a day after the event.
Ward and Muller (2010) found one significantly negative 10-day CAR that ended on
day 15 after the announcement, and attributed this to the negative market reactions
following the announcement. They however found the CAR performance to be
generally positive in the short-term, which is not consistent with the findings of this
current research study.
6.4
Hypothesis 3: CAR performance based by
age of company
Hypothesis 3 was designed to test whether there was a difference in the CAR
performance between the old and the new mining companies, thus to test whether the
age of company had an impact on the cumulative abnormal returns. These are the
hypotheses that were tested.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" ≥ 0
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"" < 0
In testing Hypothesis 4, as summarised in Figure 5-5: Long-term CAR: t-40 to t+240 and
Table 5-9: Paired t-test; pre vs. post amendment of mining charter, significant
differences were found between CAR performance of the old and the new mining
companies. This lead to a finding that the age of mining company had an impact on the
cumulative abnormal returns, and those old mining companies that had operated and
existed pre-1994 had a better CAR performance than the newer companies which are
BEE beneficiaries.
52 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
However, the BEE miners portrayed better performance in the short-term. The results
demonstrated that although both new and old miners experienced a decline is their
share performance post the announcement of a BEE transaction, the old miners were
able to recover in the long-term although their CAR remained slightly negative. These
results are not fully comparable with those of Ward and Muller (2010), as the authors
found a sub-sample of large companies to have a marginally negative CAR, while the
sub-sample of smaller companies had a strong positive CAR.
The results of the two studies can be compared because generally, the large-capital
companies are predominantly resource companies (Ward & Muller, 2010). The
performance of the large-capital companies was attributed to their export-oriented
business as these businesses sell their commodities in international markets where
they derive little or no benefit from BEE compliance.
6.5
Hypothesis 4: Timing of BEE announcements
Null hypothesis 4 postulated that the average abnormal returns of the events made
before the release of amended mining charter in September 2010 are not less than the
average abnormal returns of the events made after the amendment of the mining
charter. These are the hypotheses that were tested.
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"#$ ≥ 0
𝐻! : 𝐴𝐴𝑅!"# − 𝐴𝐴𝑅!"#$ < 0
In testing Hypothesis 4, the results revealed a clear distinction between early BEE
announcements made before the release of the Mining Charter in September 2010 and
those made thereafter. The long-term CARs of the sub-sample of announcements
made before the amendment of the mining charter are less negative while that of the
announcements made after the introduction of the Amended Mining Charter are more
negative.
The outcome of this test is different when compared to previous studies. When Ward
and Muller (2010) divided their sample in terms of early versus later BEE
53 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
announcements, they found that CARs for the ‘early’ sub-sample were negative until
approximately day 140, and thereafter it became positive. In their study, the CARs of
announcements made later in the period were initially positive and consistently
exceeded 12% from day 180 onwards (Ward & Muller, 2010).
6.6
Conclusion
This Chapter presented the discussions of the results that were analysed. The results
of the hypotheses were discussed with reference to the reviewed literature. In sum, the
BEE announcements had largely a negative impact on share performance of the
mining companies. Further conclusions and recommendations for practice and future
research are discussed in Chapter 7.
54 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
CHAPTER 7:
7.1
CONCLUSION
Introduction
This research examined the share price performance of mining stocks listed on the
JSE by tracking their share price performance after announcements relating to black
empowerment transactions. The study addressed a significant gap in BEE research,
which is important within the South African context, as the country currently reviews
progress after the initial 20 years of democratic dispensation. This Chapter concludes
the study by presenting the summary of the findings and recommendations.
7.2
Summary of the findings
The study found mixed reactions to the announcements of BEE transactions in the
long-term and short-term.
In the short-term, the 11-day window was assessed and it was found that there are
generally negative AARs, except for the 3-day window around the event. The 3-day
window (t-1 to t+1) had positive AARs and was highest on the event day. It is concluded
that the market enjoyed the news of BEE announcements when these were publicised.
This is consistent with the idea of the efficient market, as once the info had filtered
through; the market reverted back to its natural trajectory for the period.
The CAR performance of the total sample had a negative slope, thus being
downwardly trending over the entire study window. There were a few instances of
recoveries were the trend plateaued; most significantly this occurred around the 21-day
window. CAR performance held positive for this window and thereafter it continued on
a downward trend into negative CAR.
The conclusion from the presented performance of the CAR and AARs is that BEE
announcements had a negative impact on cumulative abnormal returns of mining
companies in the short term.
55 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
A paired t-test was run to determine whether BEE announcements have a greater
positive impact on the cumulative abnormal returns of old mining companies that were
listed on the JSE pre-1994 compared to their BEE counterparts.
The test revealed there were significant differences between CAR performance of the
old and the new mining companies. This lead to a finding that the age of the mining
company had an impact on its cumulative abnormal returns, and those old mining
companies that had operated and existed pre-1994 had a better CAR performance
than the newer companies that are BEE beneficiaries.
The BEE miners showed better performance in the short-term while the old miners
showed better performance in the long-term. The results exposed that although both
new and old miners experienced a decline is their share performance after the
announcement of a BEE transaction, the old miners were able to recover in the longterm.
The short-term versus long-term performance of old and BEE miners is not
reconcilable with the study performed by Ward and Muller (2010). Their study had
concluded that larger market-cap companies (similar to old miners) had marginally
negative CAR, while the smaller companies (similar to BEE miners) had a strong
positive CAR.
This study however demonstrated that generally the old miners (blue-chip shares)
performed better in the long-term. The results of the two studies can be compared
because in general, the large-cap companies are predominantly resource companies
(Ward & Muller, 2010).
A paired t-test was also generated to determine whether the early BEE announcements
made before the release of the Mining Charter in September 2010 had a greater
positive impact on the cumulative abnormal returns of mining companies compared to
those made after the amendment to legislation.
Although the results exposed negative CAR performance of mining shares to both preand post-September 2010, a clear distinction was established between early BEE
announcements made before the release of the Mining Charter in September 2010 and
those made thereafter.
56 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
The earlier events announced before September 2010 had better performance than the
later ones. This was evidenced by long-term CARs of the sub-sample of
announcements made before the Amendment of the Mining Charter being slightly
negative while those made after were significantly negative.
This study has returned different results when compared to the study of Ward and
Muller (2010), where they concluded that BEE transactions generally yielded positive
CAR performance, except for the early transactions that were negative until
approximately day140 and thereafter became positive.
7.3
Recommendation for practice
This study found that the BEE announcements resulted in a negative share price
performance, therefore South African government should determine the reasons of the
negative view to qualify why mining investors are reacting negatively. Addressing
investor concerns could contribute to the increment of foreign direct investment into the
country.
In addition to using BEE as an act of CSR like most South African companies
(Alessandri et al., 2009, 2011; Jackson et al., 2005; Wolmarans & Sartorius, 2009),
mining companies also have to comply with BEE legislation to ensure they attain or
maintain their licence to operate (Department: Mineral Resources, 2010). It is therefore
imperative for the mining sector to find a balance between BEE as a business
sustainability measure and shareholder value creation, since the study found negative
reaction by the market to BEE announcements.
57 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
7.4
Recommendations for future research
The study makes the following recommendations for future research:
First, it was noted from the study that some of the companies might not be complying
with the mining charter; therefore future studies should separate the mining companies
as per mining charter compliance.
Second, a qualitative approach should be conducted to explore the perceptions of the
investors with regard to the BEE announcements.
Third, future studies should focus on the different sectors to determine the sector
impacted the most by BEE announcements.
Fourth, future studies should contrast the performance of the share price by the types
of transactions.
7.5
Chapter Summary
This research was successful in examining the share price performance of mining
stocks listed on the JSE by tracking their share price performance after
announcements relating to black empowerment transactions. Summary of findings
were presented from the event study and the results showed negative impact on the
CARs of the mining companies. Recommendations to industry were made were above
relating to concerns flowing from the outcome of the study.
58 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
REFERENCES
Alessandri, T. M., Black, S. S., & Jackson, W. E. (2009). Black Economic
Empowerment In South Africa: Corporate Social Responsibility That Creates
Value? In Academy of Management Annual Meeting Proceedings (8, 1–6).
doi:10.5465/AMBPP.2009.44244332
Alessandri, T. M., Black, S. S., & Jackson, W. E. (2011). Black Economic
Empowerment Transactions in South Africa: Understanding When Corporate
Social Responsibility May Create or Destroy Value. Long Range Planning, 44(4),
229–249.
Antin, D. (2013). The South African Mining Sector  : An Industry at a Crossroads.
Johannesburg. Retrieved from http://www.hss.de/no_cache/home.html
Bhana, N. (2010). The stock market reaction to criticism of corporate governance
practices of companies listed on the JSE. Investment Analysts Journal, 72, 1–12.
Retrieved
from
http://www.scopus.com/inward/record.url?eid=2-s2.0-
78649801872&partnerID=tZOtx3y1
Binder, J. (1998). The event study methodology since 1969. Review of Quantitative
Finance
and
Accounting.
Retrieved
from
http://link.springer.com/article/10.1023/A:1008295500105
Bowman, R. G. (1983). Understanding and Conducting Event Studies. Journal of
Business
Finance
&
Accounting,
10(4),
561–584.
Retrieved
from
http://doi.wiley.com/10.1111/j.1468-5957.1983.tb00453.x
Brown, S. J., & Warner, J. B. (1980). Measuring security price performance. Journal of
Financial Economics, 8(3), 205–258. doi:10.1016/0304-405X(80)90002-1
Brown, S., & Warner, J. (1985). Using daily stock returns: The case of event studies.
Journal
of
Financial
Economics.
Retrieved
from
http://www.sciencedirect.com/science/article/pii/0304405X8590042X
59 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Campbell, C. J., Cowan, A. R., & Salotti, V. (2010). Multi-country event-study methods.
Journal
of
Banking
&
Finance,
34(12),
3078–3090.
doi:10.1016/j.jbankfin.2010.07.016
Cawood, F. T. (2004). The Mineral and Petroleum Resources Development Act of
2002: A paradigm shift in mineral policy in South Africa. Journal of the South
African Institute of Mining and Metallurgy, 104(1), 53–64.
Chamber of Mines of South Africa. (2012). Facts And Figures 2012. Johannesburg.
Retrieved
from
https://commondatastorage.googleapis.com/comsa/facts-and-
figures-2013.pdf
Chamber of mines of South Africa. (2013). Chamber of mines of South Africa Annual
Report
2012/2013.
Johannesburg.
Retrieved
from
https://commondatastorage.googleapis.com/comsa/annual-report-2012-2013.pdf
Chipeta, C., & Vokwana, T. C. Z. (2011). The impact of Black Economic Empowerment
transactions on shareholder wealth and firm profitability  : evidence from the JSE.
African
Finance
Journal,
13,
71–91.
Retrieved
from
http://reference.sabinet.co.za/webx/access/electronic_journals/finj/finj_v13_si1_a4
.pdf
Cooper, D. R., & Schindler, P. S. (2014). Business Research Methods (12th ed.). New
York: McGraw-Hill Education.
Corrado, C. J. (2011). Event studies: A methodology review. Accounting and Finance,
51, 207–234. doi:10.1111/j.1467-629X.2010.00375.x
Department: Mineral Resources. (2010). Amendment of the broad-based socioeconomic empowerment charter for the South African mining and minerals
industry,
September
2010.
Republic
of
South
Africa.
Retrieved
from
http://www.dmr.gov.za
Fauconnier, A., & Mathur-Helm, B. (2008). Black economic empowerment in the South
African mining industry  : A case study of Exxaro Limited. South African Journal of
Business Management, 39(4), 1–15.
60 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Hamann, R., Khagram, S., & Rohan, S. (2008). South Africa’s Charter Approach to
Post-Apartheid Economic Transformation: Collaborative Governance or Hardball
Bargaining?*. Journal of Southern African Studies, 34(1), 21–37.
Henderson, G. V. (1990). Problems and solutions in conducting event studies. Journal
of Risk and Insurance, 57, 282–306.
Jackson, W. E., Alessandri, T. M., & Black, S. S. (2005). The Price of Corporate Social
Responsibility: The Case of Black Economic Empowerment Transactions in South
Africa. Working Paper Series (Federal Reserve Bank of Atlanta), (29), 1–30.
JSE. (2014). JSE: SA Sector Categories. Retrieved January 07, 2015, from
https://www.jse.co.za/services/market-data/indices/ftse-jse-africa-index-series/sasector
Kothari, S. P., & Warner, J. B. (1997). Measuring long-horizon security price
performance. Journal of Financial Economics, 43, 301–339. doi:10.1016/S0304405X(96)00899-9
Kothari, S. P., & Warner, J. B. (2007). Econometrics of Event Studies. In B. E. Eckbo
(Ed.), Handbook of Corporate Finance: Empirical Corporate Finance (Volume 1.,
pp. 3 – 36). Hanover, New Hampshire, USA: North-Holland.
Kotze, C. (2014). Number of JSE-listed miners decrease, while market capitalisation in
resource
sector
increases.
Retrieved
January
07,
2015,
from
http://www.miningweekly.com/article/number-of-jse-listed-miners-decrease-whilemarket-capitalisation-in-resource-sector-increases-2014-08-08
Lefebvre, J. (2007). The bootstrap in event study methodology. Université de Montréal.
Retrieved
from
https://papyrus.bib.umontreal.ca/xmlui/bitstream/handle/1866/1503/a1.1g1077.pdf
?sequence=1
Leon, P. (2012). Whither the South African mining industry? Journal of Energy and
Natural Resources Law, 30(1), 1–24.
Lyon, J. D., Barber, B. M., & Tsai, C. (1999). Improved Methods for Tests of Long-Run
Abnormal Stock Returns. The Journal of Finance, 54(1), 165–201.
61 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
MacKinlay, A. C. (1997). Event Studies in Economics and Finance. Journal of
Economic Literature, 35(1), 13–39. doi:10.2307/2729691
Mitchell, M., & Netter, J. (1994). The Role of financial economics in securities fraud
cases: applications at the Securities and Exchange Commission. The Business
Lawyer, 556–557. Retrieved from http://www.jstor.org/stable/40687469
Neuhierl, A., Scherbina, A. D., & Schlusche, B. (2010). Market Reaction to Corporate
Press Releases. SSRN Electronic Journal. doi:10.2139/ssrn.1556532
Ponte, S., Roberts, S., & van Sittert, L. (2007). “Black Economic Empowerment”,
Business and the State in South Africa. Development and Change, 38(5), 933–
955. doi:10.1111/j.1467-7660.2007.00440.x
PricewaterhouseCoopers. (2014). SA Mine 2014.
Republic of South Africa. (2004a). Broad-Based Black Economic Empowerment Act,
2003 (2004). Republic of South Africa: Department of Trade and Industry.
Retrieved from http://www.dti.gov.za
Republic of South Africa. (2004b). Scorecard for the Broad Based Socio-economic
Empowerment Charter for the South African Mining Industry. Republic of South
Africa. Retrieved from http://www.gov.za/sites/www.gov.za/files/26661.pdf
Ribane, K. P. (2011). Obstacles facing BEE mining entrepreneurs during the growth
stage. Gordon Institute of Business Science, University of Pretoria. Retrieved from
http://upetd.up.ac.za/thesis/submitted/etd-06032012100150/unrestricted/dissertation.pdf
Rogerson, C. M. (2011). Mining enterprise, regulatory frameworks and local economic
development in South Africa. African Journal of Business Management.
doi:10.5897/AJBMX11.013
Sartorius, K., & Botha, G. (2008). Black economic empowerment ownership initiatives:
a Johannesburg Stock Exchange perspective. Development Southern Africa,
25(4), 437–453. doi:10.1080/03768350802318530
62 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Saunders, M., Lewis, P., & Thornhill, A. (2009). Research methods for business
students (5th ed.). England: Pearson Education Limited.
Sharenet. (2014). FTSE/JSE Sector Layout. Retrieved January 07, 2015, from
http://www.sharenet.co.za/v3/sector_codes.php
South African Reserve Bank. (2012). Annual Economic Report 2012. Johannesburg.
Retrieved
from
https://www.resbank.co.za/Lists/News
and
Publications/Attachments/5831/AER2013.pdf
Strydom, B., Christison, A., & Matias, J. (2009). An empirical investigation of the effect
of black economic empowerment transactions on share prices: 1996 to 2006.
African Journal of Accounting, Economics, Finance and Banking Research, 5(5).
Retrieved from http://ssrn.com/abstract=1534214
Tangri, R., & Southall, R. (2008). The Politics of Black Economic Empowerment in
South Africa. Journal of Southern African Studies, 34(3), 699–716. Retrieved from
http://www.tandfonline.com/doi/abs/10.1080/03057070802295856
Van der Berg, S., Burger, R., Burger, R., Louw, M., & Yu, D. (2006). Trends in poverty
and
inequality
since
the
political
transition
(No.
06).
Retrieved
from
http://www.tips.org.za/files/forum/2006/papers/VanderbergTrends_in_Poverty.pdf
Ward, M., & Muller, C. (2010). The long-term share price reaction to Black Economic
Empowerment announcements on the JSE. Investment Analysts Journal, (71),
27–36.
Wårell, L. (2007). A horizontal merger in the iron ore industry: An event study
approach.
Resources
Policy,
32(4),
191–204.
Retrieved
from
http://www.sciencedirect.com/science/article/pii/S030142070700058X
Wilson, A., & Cervantes, M. (2014). Survey of Mining Companies. Retrieved from
http://www.mining-security.org/files/Survey-of-Mining-Companies-2013.pdf
Wolmarans, H. (2012). Corporate social responsibility: The performance of black
economic empowerment (BEE) companies in a developing country after the global
financial crisis. African Journal Of Business Management, 6(14), 4974–4981.
63 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Wolmarans, H., & Sartorius, K. (2009). Corporate Social Responsibility: The Financial
Impact of Black Economic Empowerment Transactions in South Africa. South
African Journal of Economic and Management Sciences, 12(2), 180–193.
64 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
APPENDICES
Appendix 1: List of mining companies in the sample
Code
Name
Market Cap (ZAR)
BIL
BHP Billiton PLC
730 062 740 759
AGL
Anglo American PLC
394 233 729 120
AMS
Anglo American Platinum
134 840 943 000
ANG
AngloGold Ashanti
76 555 892 159
IMP
Impala Platinum
74 854 170 278
ASR
Assore ltd
56 760 017 990
ARI
African Rainbow Minerals
42 838 765 020
GFI
GoldFields
34 163 575 493
LON
Lonmin Platinum
28 638 080 695
NHM
Northam Platinum Limited
16 301 029 690
HAR
Harmony
15 154 017 311
ATL
Atlatsa Resources
2 300 297 159
RSG
Resource Generation Limited
1 976 693 149
AQP
Aquarius Platinum
1 928 008 639
DRD
DRD Gold
1 290 156 944
CZA
Coal of Africa
744 341 715
EPS
Eastern Platinum Limited
584 758 318
BDM
Buildmax Limited
480 447 223
VIL
Village
416 278 990
GBG
Great Basin Gold
386 705 505
TAW
Tawana
375 413 713
CRD
Central Rand Gold Limited
210 506 262
FCR
Ferrum Crescent Limited
186 495 361
JBL
Jubilee Platinum Plc
178 267 080
BAU
Bauba Platinum Limited
82 589 922
CMO
Chrometco Limited
24 591 442
DMR
Diamond Core Resources
#N/A
KMB
Kumba Resources Limited
#N/A
PGL
Pallinghurst Resources
#N/A
65 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Appendix 2: AAR t-stat test for significance
Day
day&340
day&339
day&338
day&337
day&336
day&335
day&334
day&333
day&332
day&331
day&330
day&329
day&328
day&327
day&326
day&325
day&324
day&323
day&322
day&321
day&320
day&319
day&318
day&317
day&316
day&315
day&314
day&313
day&312
day&311
day&310
day&39
day&38
day&37
day&36
day&35
day&34
day&33
day&32
day&31
day&0
day&1
day&2
day&3
day&4
day&5
day&6
day&7
day&8
day&9
day&10
day&11
day&12
day&13
day&14
day&15
day&16
day&17
day&18
day&19
day&20
day&21
day&22
day&23
day&24
day&25
day&26
day&27
day&28
day&29
day&30
EW&Full&
Sample&
STDEV.S3 Std.Error3
AAR
AAR
AAR
!0,407%
3,978%
0,256%
0,335%
3,978%
0,256%
0,298%
3,978%
0,256%
0,042%
3,978%
0,256%
0,130%
3,978%
0,256%
!0,179%
3,978%
0,256%
0,131%
3,978%
0,256%
!0,069%
3,978%
0,256%
!0,525%
3,978%
0,256%
!0,318%
3,978%
0,256%
!0,056%
3,978%
0,256%
0,111%
3,978%
0,256%
0,023%
3,978%
0,256%
0,064%
3,978%
0,256%
0,026%
3,978%
0,256%
0,025%
3,978%
0,256%
0,593%
3,978%
0,256%
!0,110%
3,978%
0,256%
!0,425%
3,978%
0,256%
0,102%
3,978%
0,256%
0,286%
3,978%
0,256%
!0,712%
3,978%
0,256%
0,070%
3,978%
0,256%
!0,482%
3,978%
0,256%
0,154%
3,978%
0,256%
!0,479%
3,978%
0,256%
!0,037%
3,978%
0,256%
0,000%
3,978%
0,256%
0,048%
3,978%
0,256%
0,018%
3,978%
0,256%
0,114%
3,978%
0,256%
0,439%
3,978%
0,256%
!0,273%
3,978%
0,256%
0,271%
3,978%
0,256%
!0,210%
3,978%
0,256%
!0,018%
3,978%
0,256%
!0,055%
3,978%
0,256%
!0,186%
3,978%
0,256%
!0,134%
3,978%
0,256%
0,184%
3,978%
0,256%
0,333%
3,978%
0,256%
0,163%
3,978%
0,256%
!0,121%
3,978%
0,256%
!0,427%
3,978%
0,256%
!0,118%
3,978%
0,256%
0,215%
3,978%
0,256%
0,144%
3,978%
0,256%
0,181%
3,978%
0,256%
!0,168%
3,978%
0,256%
!0,198%
3,978%
0,256%
!0,335%
3,978%
0,256%
!0,131%
3,978%
0,256%
!0,021%
3,978%
0,256%
!0,166%
3,978%
0,256%
0,168%
3,978%
0,256%
!0,218%
3,978%
0,256%
0,088%
3,978%
0,256%
!0,074%
3,978%
0,256%
0,212%
3,978%
0,256%
!0,241%
3,978%
0,256%
0,009%
3,978%
0,256%
0,253%
3,978%
0,256%
!0,350%
3,978%
0,256%
!0,141%
3,978%
0,256%
0,118%
3,978%
0,256%
!0,028%
3,978%
0,256%
!0,211%
3,978%
0,256%
0,119%
3,978%
0,256%
0,067%
3,978%
0,256%
!0,295%
3,978%
0,256%
0,289%
3,978%
0,256%
t3stat3AAR
000000(1,590)
000000001,306
000000001,163
000000000,164
000000000,506
000000(0,700)
000000000,511
000000(0,268)
000000(2,050)
000000(1,243)
000000(0,218)
000000000,432
000000000,090
000000000,248
000000000,103
000000000,096
000000002,315
000000(0,431)
000000(1,658)
000000000,397
000000001,116
000000(2,778)
000000000,275
000000(1,880)
000000000,601
000000(1,867)
000000(0,145)
000000(0,002)
000000000,187
000000000,068
000000000,446
000000001,714
000000(1,067)
000000001,059
000000(0,820)
000000(0,071)
000000(0,216)
000000(0,724)
000000(0,522)
000000000,717
000000001,298
000000000,636
000000(0,471)
000000(1,665)
000000(0,460)
000000000,840
000000000,560
000000000,705
000000(0,654)
000000(0,772)
000000(1,307)
000000(0,512)
000000(0,083)
000000(0,648)
000000000,656
000000(0,850)
000000000,342
000000(0,290)
000000000,828
000000(0,939)
000000000,037
000000000,989
000000(1,364)
000000(0,551)
000000000,459
000000(0,108)
000000(0,824)
000000000,464
000000000,263
000000(1,151)
000000001,128
p3value3
AAR
000000000,113
000000000,193
000000000,246
000000000,870
000000000,613
000000000,484
000000000,610
000000000,789
000000000,041
000000000,215
000000000,827
000000000,666
000000000,929
000000000,804
000000000,918
000000000,924
000000000,021
000000000,667
000000000,099
000000000,692
000000000,265
000000000,006
000000000,784
000000000,061
000000000,548
000000000,063
000000000,885
000000000,998
000000000,852
000000000,946
000000000,656
000000000,088
000000000,287
000000000,291
000000000,413
000000000,944
000000000,829
000000000,470
000000000,602
000000000,474
000000000,196
000000000,525
000000000,638
000000000,097
000000000,646
000000000,402
000000000,576
000000000,482
000000000,514
000000000,441
000000000,192
000000000,609
000000000,934
000000000,517
000000000,513
000000000,396
000000000,733
000000000,772
000000000,409
000000000,348
000000000,971
000000000,324
000000000,174
000000000,582
000000000,647
000000000,914
000000000,411
000000000,643
000000000,793
000000000,251
000000000,260
Day
day&31
day&32
day&33
day&34
day&35
day&36
day&37
day&38
day&39
day&40
day&41
day&42
day&43
day&44
day&45
day&46
day&47
day&48
day&49
day&50
day&51
day&52
day&53
day&54
day&55
day&56
day&57
day&58
day&59
day&60
day&61
day&62
day&63
day&64
day&65
day&66
day&67
day&68
day&69
day&70
day&71
day&72
day&73
day&74
day&75
day&76
day&77
day&78
day&79
day&80
day&81
day&82
day&83
day&84
day&85
day&86
day&87
day&88
day&89
day&90
day&91
day&92
day&93
day&94
day&95
day&96
day&97
day&98
day&99
day&100
day&101
EW&Full&
Sample&
STDEV.S3 Std.Error3
AAR
AAR
AAR
!0,155%
3,978%
0,256%
0,024%
3,978%
0,256%
!0,065%
3,978%
0,256%
!0,151%
3,978%
0,256%
!0,055%
3,978%
0,256%
!0,466%
3,978%
0,256%
!0,218%
3,978%
0,256%
0,134%
3,978%
0,256%
!0,167%
3,978%
0,256%
0,189%
3,978%
0,256%
!0,312%
3,978%
0,256%
!0,451%
3,978%
0,256%
!0,403%
3,978%
0,256%
!0,028%
3,978%
0,256%
!0,127%
3,978%
0,256%
!0,056%
3,978%
0,256%
0,418%
3,978%
0,256%
0,180%
3,978%
0,256%
0,236%
3,978%
0,256%
0,004%
3,978%
0,256%
0,103%
3,978%
0,256%
!0,091%
3,978%
0,256%
!0,132%
3,978%
0,256%
!0,175%
3,978%
0,256%
!0,161%
3,978%
0,256%
!0,557%
3,978%
0,256%
0,374%
3,978%
0,256%
0,042%
3,978%
0,256%
!0,225%
3,978%
0,256%
!0,048%
3,978%
0,256%
0,077%
3,978%
0,256%
0,145%
3,978%
0,256%
0,019%
3,978%
0,256%
!0,319%
3,978%
0,256%
0,430%
3,978%
0,256%
!0,462%
3,978%
0,256%
0,319%
3,978%
0,256%
!0,015%
3,978%
0,256%
0,138%
3,978%
0,256%
0,044%
3,978%
0,256%
!0,748%
3,978%
0,256%
0,398%
3,978%
0,256%
!0,488%
3,978%
0,256%
!0,259%
3,978%
0,256%
!0,071%
3,978%
0,256%
0,202%
3,978%
0,256%
!0,046%
3,978%
0,256%
!0,073%
3,978%
0,256%
0,245%
3,978%
0,256%
0,431%
3,978%
0,256%
!0,370%
3,978%
0,256%
!0,308%
3,978%
0,256%
!0,424%
3,978%
0,256%
!0,032%
3,978%
0,256%
!0,225%
3,978%
0,256%
0,180%
3,978%
0,256%
!0,209%
3,978%
0,256%
0,028%
3,978%
0,256%
!0,217%
3,978%
0,256%
!0,260%
3,978%
0,256%
0,245%
3,978%
0,256%
0,491%
3,978%
0,256%
!0,027%
3,978%
0,256%
0,175%
3,978%
0,256%
!0,181%
3,978%
0,256%
!0,386%
3,978%
0,256%
0,217%
3,978%
0,256%
!0,047%
3,978%
0,256%
!0,223%
3,978%
0,256%
0,071%
3,978%
0,256%
0,218%
3,978%
0,256%
t3stat3AAR
000000(0,604)
000000000,093
000000(0,254)
000000(0,590)
000000(0,214)
000000(1,819)
000000(0,849)
000000000,521
000000(0,651)
000000000,736
000000(1,216)
000000(1,759)
000000(1,574)
000000(0,110)
000000(0,495)
000000(0,218)
000000001,630
000000000,702
000000000,923
000000000,017
000000000,401
000000(0,356)
000000(0,513)
000000(0,682)
000000(0,626)
000000(2,172)
000000001,461
000000000,165
000000(0,879)
000000(0,189)
000000000,300
000000000,565
000000000,075
000000(1,244)
000000001,677
000000(1,803)
000000001,244
000000(0,060)
000000000,537
000000000,173
000000(2,918)
000000001,555
000000(1,904)
000000(1,009)
000000(0,276)
000000000,789
000000(0,181)
000000(0,287)
000000000,957
000000001,683
000000(1,443)
000000(1,203)
000000(1,653)
000000(0,124)
000000(0,878)
000000000,703
000000(0,814)
000000000,110
000000(0,846)
000000(1,015)
000000000,955
000000001,915
000000(0,104)
000000000,683
000000(0,707)
000000(1,506)
000000000,847
000000(0,182)
000000(0,869)
000000000,278
000000000,850
p3value3
AAR
000000000,547
000000000,926
000000000,800
000000000,556
000000000,831
000000000,070
000000000,397
000000000,603
000000000,516
000000000,463
000000000,225
000000000,080
000000000,117
000000000,913
000000000,621
000000000,828
000000000,104
000000000,483
000000000,357
000000000,986
000000000,689
000000000,722
000000000,608
000000000,496
000000000,532
000000000,031
000000000,145
000000000,869
000000000,381
000000000,851
000000000,765
000000000,573
000000000,940
000000000,215
000000000,095
000000000,073
000000000,215
000000000,952
000000000,592
000000000,863
000000000,004
000000000,121
000000000,058
000000000,314
000000000,783
000000000,431
000000000,856
000000000,775
000000000,340
000000000,094
000000000,150
000000000,230
000000000,100
000000000,901
000000000,381
000000000,483
000000000,416
000000000,913
000000000,398
000000000,311
000000000,341
000000000,057
000000000,918
000000000,496
000000000,480
000000000,133
000000000,398
000000000,856
000000000,386
000000000,781
000000000,396
66 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Day
day&101
day&102
day&103
day&104
day&105
day&106
day&107
day&108
day&109
day&110
day&111
day&112
day&113
day&114
day&115
day&116
day&117
day&118
day&119
day&120
day&121
day&122
day&123
day&124
day&125
day&126
day&127
day&128
day&129
day&130
day&131
day&132
day&133
day&134
day&135
day&136
day&137
day&138
day&139
day&140
day&141
day&142
day&143
day&144
day&145
day&146
day&147
day&148
day&149
day&150
day&151
day&152
day&153
day&154
day&155
day&156
day&157
day&158
day&159
day&160
day&161
day&162
day&163
day&164
day&165
day&166
day&167
day&168
day&169
day&170
EW&Full&
STDEV.S3
Std.Error3
Sample&AAR AAR
AAR
0,218%
3,978%
0,256%
.0,104%
3,978%
0,256%
0,137%
3,978%
0,256%
.0,130%
3,978%
0,256%
.0,022%
3,978%
0,256%
.0,444%
3,978%
0,256%
.0,077%
3,978%
0,256%
0,208%
3,978%
0,256%
0,228%
3,978%
0,256%
0,257%
3,978%
0,256%
.0,346%
3,978%
0,256%
0,466%
3,978%
0,256%
.0,145%
3,978%
0,256%
.0,131%
3,978%
0,256%
.0,036%
3,978%
0,256%
.0,118%
3,978%
0,256%
0,294%
3,978%
0,256%
.0,031%
3,978%
0,256%
.0,153%
3,978%
0,256%
.0,150%
3,978%
0,256%
0,279%
3,978%
0,256%
.0,157%
3,978%
0,256%
0,105%
3,978%
0,256%
.0,064%
3,978%
0,256%
0,082%
3,978%
0,256%
0,030%
3,978%
0,256%
.0,176%
3,978%
0,256%
.0,177%
3,978%
0,256%
0,072%
3,978%
0,256%
0,142%
3,978%
0,256%
0,212%
3,978%
0,256%
.0,352%
3,978%
0,256%
.0,116%
3,978%
0,256%
0,482%
3,978%
0,256%
.0,094%
3,978%
0,256%
.0,070%
3,978%
0,256%
.0,077%
3,978%
0,256%
.0,006%
3,978%
0,256%
0,107%
3,978%
0,256%
.0,124%
3,978%
0,256%
.0,362%
3,978%
0,256%
0,215%
3,978%
0,256%
.0,115%
3,978%
0,256%
.0,244%
3,978%
0,256%
.0,029%
3,978%
0,256%
.0,020%
3,978%
0,256%
.0,196%
3,978%
0,256%
.0,134%
3,978%
0,256%
.0,219%
3,978%
0,256%
.0,010%
3,978%
0,256%
0,334%
3,978%
0,256%
0,124%
3,978%
0,256%
.0,556%
3,978%
0,256%
0,044%
3,978%
0,256%
.0,245%
3,978%
0,256%
0,133%
3,978%
0,256%
.0,233%
3,978%
0,256%
0,238%
3,978%
0,256%
0,248%
3,978%
0,256%
.0,148%
3,978%
0,256%
.0,024%
3,978%
0,256%
0,172%
3,978%
0,256%
0,062%
3,978%
0,256%
.0,245%
3,978%
0,256%
.0,706%
3,978%
0,256%
.0,158%
3,978%
0,256%
0,381%
3,978%
0,256%
.0,044%
3,978%
0,256%
.0,105%
3,978%
0,256%
.0,123%
3,978%
0,256%
t3stat3AAR
,,,,,,,,,,,0,850
,,,,,,,,,,(0,404)
,,,,,,,,,,,0,535
,,,,,,,,,,(0,508)
,,,,,,,,,,(0,087)
,,,,,,,,,,(1,731)
,,,,,,,,,,(0,299)
,,,,,,,,,,,0,813
,,,,,,,,,,,0,890
,,,,,,,,,,,1,002
,,,,,,,,,,(1,352)
,,,,,,,,,,,1,817
,,,,,,,,,,(0,566)
,,,,,,,,,,(0,511)
,,,,,,,,,,(0,140)
,,,,,,,,,,(0,462)
,,,,,,,,,,,1,147
,,,,,,,,,,(0,123)
,,,,,,,,,,(0,597)
,,,,,,,,,,(0,586)
,,,,,,,,,,,1,088
,,,,,,,,,,(0,611)
,,,,,,,,,,,0,409
,,,,,,,,,,(0,249)
,,,,,,,,,,,0,318
,,,,,,,,,,,0,116
,,,,,,,,,,(0,686)
,,,,,,,,,,(0,690)
,,,,,,,,,,,0,281
,,,,,,,,,,,0,555
,,,,,,,,,,,0,829
,,,,,,,,,,(1,376)
,,,,,,,,,,(0,453)
,,,,,,,,,,,1,882
,,,,,,,,,,(0,367)
,,,,,,,,,,(0,274)
,,,,,,,,,,(0,299)
,,,,,,,,,,(0,022)
,,,,,,,,,,,0,418
,,,,,,,,,,(0,484)
,,,,,,,,,,(1,413)
,,,,,,,,,,,0,838
,,,,,,,,,,(0,450)
,,,,,,,,,,(0,952)
,,,,,,,,,,(0,111)
,,,,,,,,,,(0,078)
,,,,,,,,,,(0,764)
,,,,,,,,,,(0,523)
,,,,,,,,,,(0,856)
,,,,,,,,,,(0,041)
,,,,,,,,,,,1,305
,,,,,,,,,,,0,486
,,,,,,,,,,(2,171)
,,,,,,,,,,,0,171
,,,,,,,,,,(0,957)
,,,,,,,,,,,0,521
,,,,,,,,,,(0,907)
,,,,,,,,,,,0,930
,,,,,,,,,,,0,968
,,,,,,,,,,(0,576)
,,,,,,,,,,(0,096)
,,,,,,,,,,,0,671
,,,,,,,,,,,0,241
,,,,,,,,,,(0,957)
,,,,,,,,,,(2,756)
,,,,,,,,,,(0,615)
,,,,,,,,,,,1,488
,,,,,,,,,,(0,170)
,,,,,,,,,,(0,411)
,,,,,,,,,,(0,479)
p3value3AAR
,,,,,,,,,,,0,396
,,,,,,,,,,,0,687
,,,,,,,,,,,0,593
,,,,,,,,,,,0,612
,,,,,,,,,,,0,931
,,,,,,,,,,,0,085
,,,,,,,,,,,0,765
,,,,,,,,,,,0,417
,,,,,,,,,,,0,374
,,,,,,,,,,,0,318
,,,,,,,,,,,0,178
,,,,,,,,,,,0,070
,,,,,,,,,,,0,572
,,,,,,,,,,,0,610
,,,,,,,,,,,0,889
,,,,,,,,,,,0,644
,,,,,,,,,,,0,253
,,,,,,,,,,,0,902
,,,,,,,,,,,0,551
,,,,,,,,,,,0,559
,,,,,,,,,,,0,277
,,,,,,,,,,,0,542
,,,,,,,,,,,0,683
,,,,,,,,,,,0,803
,,,,,,,,,,,0,751
,,,,,,,,,,,0,908
,,,,,,,,,,,0,493
,,,,,,,,,,,0,491
,,,,,,,,,,,0,779
,,,,,,,,,,,0,579
,,,,,,,,,,,0,408
,,,,,,,,,,,0,170
,,,,,,,,,,,0,651
,,,,,,,,,,,0,061
,,,,,,,,,,,0,714
,,,,,,,,,,,0,785
,,,,,,,,,,,0,765
,,,,,,,,,,,0,982
,,,,,,,,,,,0,677
,,,,,,,,,,,0,629
,,,,,,,,,,,0,159
,,,,,,,,,,,0,403
,,,,,,,,,,,0,653
,,,,,,,,,,,0,342
,,,,,,,,,,,0,912
,,,,,,,,,,,0,938
,,,,,,,,,,,0,446
,,,,,,,,,,,0,601
,,,,,,,,,,,0,393
,,,,,,,,,,,0,967
,,,,,,,,,,,0,193
,,,,,,,,,,,0,628
,,,,,,,,,,,0,031
,,,,,,,,,,,0,865
,,,,,,,,,,,0,339
,,,,,,,,,,,0,603
,,,,,,,,,,,0,365
,,,,,,,,,,,0,353
,,,,,,,,,,,0,334
,,,,,,,,,,,0,565
,,,,,,,,,,,0,924
,,,,,,,,,,,0,503
,,,,,,,,,,,0,810
,,,,,,,,,,,0,339
,,,,,,,,,,,0,006
,,,,,,,,,,,0,539
,,,,,,,,,,,0,138
,,,,,,,,,,,0,865
,,,,,,,,,,,0,681
,,,,,,,,,,,0,633
Day
day&171
day&172
day&173
day&174
day&175
day&176
day&177
day&178
day&179
day&180
day&181
day&182
day&183
day&184
day&185
day&186
day&187
day&188
day&189
day&190
day&191
day&192
day&193
day&194
day&195
day&196
day&197
day&198
day&199
day&200
day&201
day&202
day&203
day&204
day&205
day&206
day&207
day&208
day&209
day&210
day&211
day&212
day&213
day&214
day&215
day&216
day&217
day&218
day&219
day&220
day&221
day&222
day&223
day&224
day&225
day&226
day&227
day&228
day&229
day&230
day&231
day&232
day&233
day&234
day&235
day&236
day&237
day&238
day&239
day&240
EW&Full&
STDEV.S3
Std.Error3
Sample&AAR AAR
AAR
0,593%
3,978%
0,256%
0,079%
3,978%
0,256%
0,074%
3,978%
0,256%
.0,131%
3,978%
0,256%
.0,147%
3,978%
0,256%
0,363%
3,978%
0,256%
0,434%
3,978%
0,256%
.0,672%
3,978%
0,256%
0,221%
3,978%
0,256%
.0,174%
3,978%
0,256%
.0,112%
3,978%
0,256%
.0,206%
3,978%
0,256%
0,135%
3,978%
0,256%
0,249%
3,978%
0,256%
0,090%
3,978%
0,256%
.0,087%
3,978%
0,256%
.0,233%
3,978%
0,256%
.0,603%
3,978%
0,256%
.0,142%
3,978%
0,256%
.0,092%
3,978%
0,256%
.0,069%
3,978%
0,256%
.0,064%
3,978%
0,256%
.0,069%
3,978%
0,256%
0,091%
3,978%
0,256%
0,136%
3,978%
0,256%
0,035%
3,978%
0,256%
.0,264%
3,978%
0,256%
0,167%
3,978%
0,256%
.0,300%
3,978%
0,256%
.0,168%
3,978%
0,256%
.0,019%
3,978%
0,256%
.0,147%
3,978%
0,256%
.0,077%
3,978%
0,256%
.0,019%
3,978%
0,256%
0,188%
3,978%
0,256%
0,227%
3,978%
0,256%
0,134%
3,978%
0,256%
.0,211%
3,978%
0,256%
0,116%
3,978%
0,256%
.0,076%
3,978%
0,256%
0,018%
3,978%
0,256%
.0,125%
3,978%
0,256%
.0,643%
3,978%
0,256%
0,220%
3,978%
0,256%
.0,369%
3,978%
0,256%
.0,218%
3,978%
0,256%
0,231%
3,978%
0,256%
0,175%
3,978%
0,256%
0,237%
3,978%
0,256%
.0,012%
3,978%
0,256%
0,009%
3,978%
0,256%
0,244%
3,978%
0,256%
0,052%
3,978%
0,256%
.0,357%
3,978%
0,256%
0,083%
3,978%
0,256%
.0,009%
3,978%
0,256%
.0,036%
3,978%
0,256%
.0,214%
3,978%
0,256%
0,310%
3,978%
0,256%
0,012%
3,978%
0,256%
0,077%
3,978%
0,256%
.0,311%
3,978%
0,256%
0,231%
3,978%
0,256%
0,278%
3,978%
0,256%
0,173%
3,978%
0,256%
0,203%
3,978%
0,256%
.0,038%
3,978%
0,256%
.0,071%
3,978%
0,256%
.0,162%
3,978%
0,256%
0,208%
3,978%
0,256%
t3stat3AAR
,,,,,,,,,,,2,314
,,,,,,,,,,,0,309
,,,,,,,,,,,0,289
,,,,,,,,,,(0,511)
,,,,,,,,,,(0,572)
,,,,,,,,,,,1,417
,,,,,,,,,,,1,694
,,,,,,,,,,(2,624)
,,,,,,,,,,,0,861
,,,,,,,,,,(0,678)
,,,,,,,,,,(0,439)
,,,,,,,,,,(0,805)
,,,,,,,,,,,0,528
,,,,,,,,,,,0,973
,,,,,,,,,,,0,351
,,,,,,,,,,(0,338)
,,,,,,,,,,(0,909)
,,,,,,,,,,(2,354)
,,,,,,,,,,(0,554)
,,,,,,,,,,(0,358)
,,,,,,,,,,(0,270)
,,,,,,,,,,(0,248)
,,,,,,,,,,(0,269)
,,,,,,,,,,,0,355
,,,,,,,,,,,0,532
,,,,,,,,,,,0,138
,,,,,,,,,,(1,029)
,,,,,,,,,,,0,651
,,,,,,,,,,(1,171)
,,,,,,,,,,(0,657)
,,,,,,,,,,(0,072)
,,,,,,,,,,(0,575)
,,,,,,,,,,(0,302)
,,,,,,,,,,(0,076)
,,,,,,,,,,,0,734
,,,,,,,,,,,0,884
,,,,,,,,,,,0,525
,,,,,,,,,,(0,825)
,,,,,,,,,,,0,453
,,,,,,,,,,(0,298)
,,,,,,,,,,,0,070
,,,,,,,,,,(0,488)
,,,,,,,,,,(2,508)
,,,,,,,,,,,0,860
,,,,,,,,,,(1,441)
,,,,,,,,,,(0,850)
,,,,,,,,,,,0,902
,,,,,,,,,,,0,681
,,,,,,,,,,,0,926
,,,,,,,,,,(0,046)
,,,,,,,,,,,0,034
,,,,,,,,,,,0,951
,,,,,,,,,,,0,201
,,,,,,,,,,(1,395)
,,,,,,,,,,,0,325
,,,,,,,,,,(0,036)
,,,,,,,,,,(0,139)
,,,,,,,,,,(0,834)
,,,,,,,,,,,1,210
,,,,,,,,,,,0,045
,,,,,,,,,,,0,301
,,,,,,,,,,(1,214)
,,,,,,,,,,,0,900
,,,,,,,,,,,1,085
,,,,,,,,,,,0,677
,,,,,,,,,,,0,793
,,,,,,,,,,(0,147)
,,,,,,,,,,(0,275)
,,,,,,,,,,(0,632)
,,,,,,,,,,,0,813
p3value3AAR
,,,,,,,,,,,0,022
,,,,,,,,,,,0,757
,,,,,,,,,,,0,773
,,,,,,,,,,,0,610
,,,,,,,,,,,0,568
,,,,,,,,,,,0,158
,,,,,,,,,,,0,092
,,,,,,,,,,,0,009
,,,,,,,,,,,0,390
,,,,,,,,,,,0,498
,,,,,,,,,,,0,661
,,,,,,,,,,,0,422
,,,,,,,,,,,0,598
,,,,,,,,,,,0,332
,,,,,,,,,,,0,726
,,,,,,,,,,,0,735
,,,,,,,,,,,0,364
,,,,,,,,,,,0,019
,,,,,,,,,,,0,580
,,,,,,,,,,,0,721
,,,,,,,,,,,0,787
,,,,,,,,,,,0,804
,,,,,,,,,,,0,788
,,,,,,,,,,,0,723
,,,,,,,,,,,0,595
,,,,,,,,,,,0,890
,,,,,,,,,,,0,305
,,,,,,,,,,,0,516
,,,,,,,,,,,0,243
,,,,,,,,,,,0,512
,,,,,,,,,,,0,942
,,,,,,,,,,,0,566
,,,,,,,,,,,0,763
,,,,,,,,,,,0,940
,,,,,,,,,,,0,464
,,,,,,,,,,,0,378
,,,,,,,,,,,0,600
,,,,,,,,,,,0,410
,,,,,,,,,,,0,651
,,,,,,,,,,,0,766
,,,,,,,,,,,0,945
,,,,,,,,,,,0,626
,,,,,,,,,,,0,013
,,,,,,,,,,,0,391
,,,,,,,,,,,0,151
,,,,,,,,,,,0,396
,,,,,,,,,,,0,368
,,,,,,,,,,,0,497
,,,,,,,,,,,0,356
,,,,,,,,,,,0,963
,,,,,,,,,,,0,973
,,,,,,,,,,,0,342
,,,,,,,,,,,0,841
,,,,,,,,,,,0,164
,,,,,,,,,,,0,746
,,,,,,,,,,,0,972
,,,,,,,,,,,0,889
,,,,,,,,,,,0,405
,,,,,,,,,,,0,227
,,,,,,,,,,,0,964
,,,,,,,,,,,0,763
,,,,,,,,,,,0,226
,,,,,,,,,,,0,369
,,,,,,,,,,,0,279
,,,,,,,,,,,0,499
,,,,,,,,,,,0,428
,,,,,,,,,,,0,883
,,,,,,,,,,,0,783
,,,,,,,,,,,0,528
,,,,,,,,,,,0,417
67 | P a g e
© 2014 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement