St. Petersburg State University
Graduate School of Management
Master in Management Program
EARNINGS MANAGEMENT AROUND CEO
TURNOVER: COMPARISON OF RUSSIAN AND
CHILE
Master’s Thesis by 2nd year student
Concentration - International Business
Juan Luis Leiva Urrea
Research advisor:
Ph.D. Associate Professor,
Tatiana A. Garanina
St. Petersburg
2016
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
ЗАЯВЛЕНИЕ О САМОСТОЯТЕЛЬНОМ ХАРАКТЕРЕ
МАГИСТЕРСКОЙ ДИССЕРТАЦИИ
Я , Хуан Луис Леива Урреа, студент 2 курса магистратуры направления по направлению
38.04.02 «Менеджмент», заявляю, что в моей магистерской диссертации на тему
«Управление прибылью и смена CEO: сравнение России и Чили», представленной в ГАК
для публичной защиты, не содержится элементов плагиата.
Все прямые заимствования из печатных и электронных источников, а также из
защищенных ранее выпускных квалификационных работ, магистерских, кандидатских и
докторских диссертаций имеют соответствующие ссылки.
Я ознакомлен с действующим в Высшей школе менеджмента СПбГУ регламентом
учебного процесса, согласно которому обнаружение плагиата (прямых заимствований из
других источников без соответствующих ссылок) является основанием для выставления за
магистерскую диссертацию оценки «неудовлетворительно».
________________________________________________(Подпись студента)
________________________________________________ (Дата)
PAGE 2
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
STATEMENT ABOUT THE INDEPENDENT CHARACTER
OF THE MASTER THESIS
I, Juan Luis Leiva Urrea second year master student, Management program
38.04.02 «Management», state that my master thesis on the topic «EARNINGS
MANAGEMENT AROUND CEO TURNOVER: COMPARISON OF RUSSIA AND
CHILE », which was presented for the public defense, does not contain any elements of
plagiarism.
All direct borrowings from printed and electronic sources, as well as from master theses,
PhD and doctorate theses which were defended earlier, have appropriate references.
I am familiar with the study process regulations at Graduate School of Management of
Saint-Petersburg State University, according to which finding of plagiarism (direct
borrowings from other sources without appropriate references) can be the reason for
master thesis to be evaluated as «unsatisfactory».
________________________________________________(Student’s signature)
________________________________________________ (Date)
PAGE 3
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Аннотация
ФИО студента
Тема магистерской диссертации
Факультет
Программа
Год окончания
ФИО научного руководителя
Описание целей, задач и основных
результатов исследования
Ключевые слова
Хуан Луис Леива Урреа
Управление прибылью и смена CEO: сравнение России
и Чили
Высшая школа менеджмента
Менеджмент
2016
Гаранина Татьяна Александровна
В данном исследовании изучается вопрос управления
прибылью при смене CEO в таких странах, как Россия
и Чили. Такая постановка является новой для
развивающихся стран. В то же время, как показал
обзор литературы, данный вопрос уже изучался в
развитых странах.
Целью данной магистерской
диссертации является анализ того, существуют ли
различия в управлении прибылью уходящими и
новыми CEO на таких рынках, как Россия и Чили. В
работе также будут выявлены причины и побуждения,
связанные с управлением прибылью. Существуют
различные причины управления прибылью при смене
CEO, о чем свидетельствуют то, что вновь приходящие
CEO чаще склонны к занижению прибыли по
сравнению со своими предшественниками. Данное
поведение можно объяснить тем, что новые CEO
пытаются заявить о том, что их предшественники
неэффективно управляли компанией и именно
благодаря их приходу повысить темп роста прибыли.
Последующее повышение прибыли приведет к более
высокому уровню вознаграждения нового CEO.
В исследовании используется регре ссионное
моделирование для трех разных периодов в Чили и
России. Проведенный анализ позволил сделать вывод о
различиях в полученных результатах для двух рынков.
Модели являются значимыми, что говорит о
возможности их использования для оценки уровня
управления прибылью в Чили и в России.
Управление прибылью; смена CEO; перепроизводство;
издержки производства
PAGE 4
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Abstract
Master Student's Name
Master Thesis Title
Juan Luis Leiva Urrea
Earnings management around CEO turnover:
comparison of Russian and Chile
Faculty
Graduate School of Management
Program
Management
Year
2016
Academic Advisor’s Name
Tatiana Garanina
Description of the goal, tasks and main This study investigates earnings management around
results
CEO turnover in companies in Russia and Chile. This
evidence is new and contributes to the on-going
discussion of earnings management in emerging
countries. This phenomenon is common in other
countries given the literature studied. The task of this
master thesis is intended to explain the importance
whether new CEOs and outgoing CEOs manage
earnings differently from established CEOs. Besides, we
will show whether there exist differences in earnings
management between Russia and Chile -and identify the
connection between earnings management and
incentives. There exist different incentives to study for
Earning Management around CEO turnover that means
how incoming CEOs in general tend to shape earnings
downward relative to the predecessor CEO. This
behavior appears when the new CEO wants to display
that the company was mistreat by their antecessor and
the following periods the company would improve its
earnings growth. Thus, the new chairperson will get all
the credits for improve earnings in the company, and this
will be reflected with higher future bonuses and salary.
The research presents the results of the multiple
regression models for Chile and Russia for three
different periods. It is shown the existence of
significance of Earnings Management for each model
finding different result in both markets. Besides, it is
stated the ability of the different models to explain the
Earnings management for the periods under study for
Chile and Russia.
Key words
Earnings management; CEO turnover; overproduction;
production costs.
PAGE 5
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
INDEX OF CONTENT
INTRODUCTION…………………………………………………………………….....8
Background of the research…………………………………………………….8
Research objectives and research method……………………………………..9
Research Proposal……………………………………………………………….9
Expected Findings and contributions………………………………………….10
Organization of the research…………………………………………………...11
CHAPTER I: STATE OF ART OF EARNINGS MANAGEMENT AROUND CEO’s
TURNOVER…………………………………………………………………………….13
1.1 Earnings Management around CEO’s Turnover…………..…………….13
1.1.1 Earnings Management…………………………………………....13
1.1.2 Earnings Management incentives……………………...…………14
1.1.3 CEO’s Turnover………………………………………………….15
1.2 Literature Evidence…………………………………………………………15
1.2.1 CEO’s Turnover evidence………………………………………..15
1.2.2 Performance evidence in CEO’s turnover ……………………….20
1.2.3 Earnings Management evidence……………………...…………..22
Chapter Summary………………………………..……………………………..27
CHAPTER II: DATA, METHODOLOGY, RESULT ANALYSIS AND
IMPLICATIONS OF EARNINGS MANAGEMENT…………………………..…...28
2.1 Data Selection……………………………………………………………….28
2.1.1 Data sources………………………………………………………..28
2.1.2 Inclusion Dataset…………………………………………………...30
2.2 Methodology………………………………………………………………...31
2.2.1 Accrual-based Earnings Management………………...……………31
2.2.2 Real Earnings Management…………………………………..……32
2.2.3 Variables description.……………………………………………...33
2.3 Data Analysis and Results………………………………………………….37
2.3.1 Sample Statistics…………………………………………………...37
2.3.2 Regression Model…………………………………………………..38
2.3.2.1 Results for Chile.………………………………………...40
2.3.2.2 Results for Russia..……………………………………….43
PAGE 6
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.4 Regression model results discussion……………………………………….46
2.4.1 Analysis for first period of new CEO………………………………46
2.4.2 Analysis for second period of new CEO…………………………...47
2.4.3 Analysis for last period of outgoing CEO………………………….50
2.5 Implications of the results………………………………………………….52
2.6 Current Research Limitations……………………………………………..54
Chapter Summary………...…………………………………….………………55
CONCLUSIONS………………………………………………………………………..58
FURTHER INVESTIGATIONS………………………………………………………62
REFERENCES………………………………………………………………………....63
APPENDIXES…………………………………………………………………………..67
APPENDIX 1: Descriptive Statistics, Russia………………………………….67
APPENDIX 2: Descriptive Statistics, Chile…………………………………...67
APPENDIX 3: Pearson Correlation, Chile……………………………………68
APPENDIX 4: Pearson Correlation, Russia…………………………………..68
APPENDIX 5: Regression summary, Chile…………………………………...69
APPENDIX 6: Regression summary, Russia………………………………….71
PAGE 7
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
INTRODUCTION
Background of the research
This Master Thesis investigates the contrast between Russian and Chilean
companies using earnings management in times of CEO turnover. This issue is rather
complicated, however is very interesting and useful.
At present the CEO has enough independency to manage the earnings of the
company; some authors have shown evidence regarding behavioral patterns in CEOs.
This behavior is formed as different incentives of CEOs, for instance, incentives to
manipulate earnings downward relative to the predecessor CEO. On the other hand, they
could manipulate real business upward by the established CEO.
The importance of the study of this manipulation of real business is aimed at
explaining the differences, mismatches or inconsistencies that exist in the financial report,
when CEOs manipulate the earnings of the company. In this way you can understand
reasons and situations by delivering a statistical tool to check if this effect exists in other
markets.
The problem of this investigation can be described as follows:
The research problem detected is very useful for improvement, where managers
are aware of earnings management and can identify inefficiency of allocation of resources
within the company in the different markets, but there is insufficient empirical evidence
of studies analyzing the Earnings Management around CEO turnover in Emerging
markets. We will contrast Russian companies with Chilean companies where this will be
an important contribution to literature as there is no evidence at all. Finally, we will be
able to conclude which of these CEOs (Russians or Chilean) tend to manipulate more
their real business activities and based on the findings achieved in the research, propose
recommendation and conclude providing more information for the topic of the research to
the market.
In summary, this study is aimed to contribute to the understanding of Earnings
Management around CEO turnover; Analysis of Russian and Chilean companies and
comparison of the obtained results.
PAGE 8
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The overall goal and objectives are defined in the next part, as well as the
methodology use in this research, in order to achieve the results and conclusions
searched.
Research objectives and research method
General Objective
Analyze the existence of earnings management in times of CEO turnover in
Russia and Chile.
Specific Objectives
1.
2.
3.
4.
5.
Make literature review on the topic of earnings management.
Develop hypothesis on the base of literature review.
Collect required data for companies from Russia and Chilean market.
Apply regression analysis for testing the developed hypothesis.
Make managerial application of the obtained results.
In the paper the following issues will be studies in details:
Whether new CEOs and outgoing CEOs manage earnings differently from
established CEOs?
Whether there exist differences between manipulation by Chilean CEOs and
Russian CEOs in real business activities?
To identify the existence of earnings management in Russian and Chilean CEO’s
Turn-over we should first establish some key definitions of the study. Outgoing CEO: it is
the CEO that leaves the company. The incoming CEO’s: are those that are new CEO. The
cut-off date means in theory the very last date where the earnings management can be
manipulated. In the case of accruals earnings management, the date of cut-off is the date
of earnings announcement. This approach is focus on the work of Geertsema, Lont and
Lu (2013) with the paper Earnings Management around CEO Turnovers. The data for
this part of the research will be collected by the Data-Base of GSOM virtually library
(Thomson Reuters).
The first approach will be evaluating what is the best way to analyze real earnings
management and contrast them with our findings, for example Geertsema, Lont and Lu
(2013) established “…two business activities to estimate earnings management; one is the
PAGE 9
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
overproduction and the second one is the delay of discretionary expenditures that inflate
the earnings for a temporally period (or, in the other way around, under production and
front loading discretionary expenditures that temporarily would decrease earnings)”
In case of finding strong evidence of manipulation in earnings business activities,
the next step is to find the specific reasons of both markets (Russian and Chile) and
suggest some actions to avoid this distortion in the accounting process of the companies.
In case of existence of earnings management in CEO’s turn-over, we have some
insights of what reason could motivate CEOs to manage earnings of the company such as
CEO’s salaries/bonuses/option programs, CEO’s education, and previous experience.
With these ideas we can start researching the specific reasons of the markets that we are
studying, Russia and Chile.
Research Proposal
First, our study investigates real management around CEO turnover in Russia and
companies in Chile. This evidence (if it exists) is new and contributes to the on-going
discussion of earnings management in emerging countries. This phenomenon is common
in other countries given the literature studied.
Second, it is important to establish whether new CEOs and outgoing CEOs
manage earnings differently from established CEOs
Third, we want to find whether there exist differences in earnings management
between Russia and Chile. Finally, we want to demonstrate the importance of controlling
firm performing, regarding general measure of earnings management. Summarizing we
want to answer the following questions about the markets of Russian and Chile:
Does the manipulation in real business CEOs turnover exist?
Do new CEO and outgoing CEO manage earnings differently?
Does the earnings management differ between Russia and Chile?
Are there measures to avoid this manipulation?
Given the research questions, it is important that in this stage we aim to investigate
their truth through this study, sustained in studies and backgrounds of researchers that
have solved this problem in the international market. Both departing and new CEOs may
PAGE 10
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
have strong incentives to manage earnings. Thus, we can state a hypothesis that will help
us to lead the research in this topic. The hypothesis for this study is:
“The fact of change of CEO does not influence the level of earnings management in a
company”
Expected Findings and Contributions
Our study benefits from a much larger and more diverse information of CEO
change firms in the Chilean market and Russian market. This new evidence will give us a
better understanding of both markets, contrast them and learn of the differences. Besides,
with these new findings, we will be able to advice companies how to control perverse
movement of new and outgoing CEO, avoiding consequences of manipulation of real
business operations because of selfish interests.
Companies will be beneficiated of our findings, being aware about the incentives
of CEOs to manage earnings and how these practices could be prevented in order to avoid
consequences.
Another point that is very important to highlight is that Russia and Chile are
emerging countries with different factors that lead the way that companies behave, in case
we compare with the literature studied (Mainly developed countries). The factors that
could make the difference are such as the Russian/Chile culture, the principles of doing
business in Russia/Chile and the earnings management when there is CEO turnover.
PAGE 11
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Organization of the research
This thesis is organized into three chapters described below:
The first chapter is intended to literature overview of the topic. Presentation of the
main concepts involved in the research problem. Followed by the incentives of the CEO’s
to manage earnings in different companies, illustrate performance of the CEO turnover
and the earnings management evidence in different markets.
The second chapter sets out the methodology used to achieve the objectives of the
empirical research based in the literature review of the first chapter. Besides, it is
described the variables involved in the research and the Regression model. Thus, get the
output of the model for analysis.
In the last chapter, we discuss the results obtained and the interpretation of the
data. Finally, the conclusions and the recommendations found in this investigation are set
and showed in the last chapter.
PAGE 12
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
CHAPTER I: STATE OF ART OF EARNINGS
MANAGEMENT AROUND CEO’s TURNOVER
1.1 EARNINGS MANAGEMENT AROUND CEO TURNOVER
1.1.1 Earnings Management
Managers of the companies have the possibility to choose the accounting policy of
the firm. Given, this accounting policy could arise two situations; first of all, the policy
can maximize the benefit of the shareholders, but in the other case could maximize the
benefit of the manager at the expenses of the owners (Wartfiled et al., 1995). Thus, if we
expect that the manager act in a rational way, it is natural to expect that they choose the
policy that maximizes their own benefit, objective that could crash with the maximization
of the enterprise value that the shareholders demand. The situation described above is
called Earnings Management that is the election of the accounting policy made by the
managers, in order to achieve certain specific goals (Wartfiled et al., 1995).
There are many definitions of Earnings Management. One of the more famous
definitions of Earnings Management is the intervention in the process of elaboration of
the financial and accountable information, of course that the main goal is obtain some
personal benefit (Schipper, 1989). Another definition is how the directors use their own
criteria in a discretional way for the elaboration of the financial statements with the
purpose to influence in the perception of the investors or creditors over the subjacent
result or to influence in the results based in accounting numbers (Healy and Whalen,
1999).
The information related to Earnings Management and the wide number of articles
about the topic, the researchers have seen that all the effort are limited because of the
difficulty of measuring the incentives of managers, the process of taking decisions and
phenomenon that are not observables (García et al., 2005). Moreover, the process of
researching this issue, it is particularly difficult also because managers have more
information; they have incentives to dissimulate any adjustment that could benefit them
in the future.
There is a concept that is called accrual adjustment; this means the calculation
between the difference of the accounting profit and cash flow of the operations. The key
here is separate the observable variable accrual adjusted into two components not
PAGE 13
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
observables that is the discretional and non-discretional parts. The changes of the
discretional part are an attempt to alter the results (Poveda et al., 2001).
1.1.2 Earnings Management Incentives
The problem to be addressed in this investigation is that there is not enough
evidence from studies that focus on CEO turnover, occasion which creates an
environment with strong stimulus of earnings management for both outgoing and
incoming CEOs. Earnings Management around CEO turnover means how incoming
CEOs in general tend to shape real business activities (but not accounting accruals) to
manage earnings downward relative to the predecessor CEO. This behavior appears when
the new CEO wants to display that the company was mistreat by their antecessor and the
following periods the company would improve its earnings growth. Thus, the new
chairperson will get all the credits for improve earnings in the company, and this will be
reflected with higher future bonuses and salary (Income) (Reitenga and Tearney, 2003).
Later in this chapter, it will be described in more details how managers could manage
earnings in companies. In the case of the CEO that is departing, he/she would want to do
the opposite as new CEO. Established CEO would trick real business activities upward to
leave the charge and get the highest possible bonus. (Graham et al., 2005)
In this way we want to know if there is a real effect in Russian companies and
contrast them with companies from Chile. This will be an important contribution to
diminish the literature gap, in accordance with the existence of little researches in Russia
and Chile. We will be able to conclude which of these CEOs (Russians or Chilean) tend
to manage earnings in companies.
To develop the theoretical framework, it will be describe how different authors
addressed this problem to determine the Earnings Management around CEO turnover and
the different variables around this topic for a better understanding of the problem.
There are different studies that try to investigate how change the relation during
different periods of CEO turnover, for instance, studying turnover of CEOs, for board
driven (internal) and through takeover and bankruptcy (external).
PAGE 14
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
1.1.3 CEO’s Turnover
Replacing a leader by other is a source of controversy and also of real conflict
within the firm. In the enterprise, the Chief selection is one of the most important
decisions that the board could take, but this controversial situation could be ended in the
continuity or cracking of part of the company. This figure not only has a huge impact on
the performance of the company, but the process of subsequence, on the one hand,
profoundly influences the image of leadership for employees, investors, analysts,
shareholders and other constituencies that have the organization. Secondly, there is a
considerable pressure on the entire management team, most notably if the candidate
comes from outside.
Generally, the CEO is aware that the last service he can make to the company is
that someone better than him arrives. As we shall see, their cooperation in this task is,
however, complicated in its design and in its implementation, since many circumstances
come into play: the progress of the company's prospective strategies, the Chief himself,
the board administration or similar frame in family businesses, the applicants and the
management lineup. Thus, the consequences of success or failure in the election of the
CEO, and its subsequent implementation are affecting the whole company.
1.2 LITERATURE EVIDENCE
1.2.1 CEO’s Turnover evidence
In the research How has CEO Turnover Change of Kaplan and Minton (2012), the
motivation of the authors to study CEO turnover was study a topic where directors are
commonly criticized for paying exaggerated rents to CEOs and increasing the wages and
bonuses of CEOs through the time. The main criticism is the role of the board of directors
increasing CEO payment and auditing pay practices. While there was a lot of attention on
this topic, changes in CEO turnover have been neglected receiving just little attention. In
this point, the authors examine how the performance of different companies changes
given CEO turnover.
Kaplan and Minton (2012) observed from a sample of large companies of U.S.A
from 1992 to 2007, that the annual CEO turnover is higher in recent periods than earlier
PAGE 15
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
periods estimated in previous investigations. This finding suggests that the average
administration of a CEO is less than seven years for the total period of study, but in the
more recent period, since 2000 this average tenure CEO turnover fall down to less than
six years of administration. The research also establish that internal turnover is somewhat
more related to three items of firm stock performance – performance related to industry,
industry related to the market, and performance of the entire stock market -.
It was noticed that internal turnover over the period 2000 is more delicate related
to the three factors named above of stock performance than the former periods. And the
most interesting finding is that the board use to get more attentive to the stock
performance when they have been reprobated in certain period.
This sensitive issue that drives turnovers could have some very important
implications in the firm. First of all, CEO performance through the time has become more
and more deficient. In the case of external takeovers that are included in the research,
Chief remains less time in the firm than in the past reported in previous studies. Second
point to highlight, the turnovers that are considered as “unforced”, in fact, are not
voluntary (Kaplan and Minton, 2012).
The third point, include the role of the board described in Kaplan and Minton,
(2012). There is a discovery that shows that the board operates well to the different
insufficient performance of the industry and external annexing. Fourth, the higher the
CEO pays the shorter CEO tenure in the firm and sensitiveness to stock performance. As
last point, the higher rent of CEO the higher the incentive to earnings management inside
the firm.
Other interesting finding about CEOs turnover is the understanding how the
independence of a board affects the CEO’s competence to earn rents from the firm. This
investigation assumes that the CEO has enclosed information about the company , where
the board would have to take the decision whether to remove him/her or not. But, if the
board is more active replacing underperforming CEOs, the Chief Executive Officer is
better able to use the private information to get better results that would be translated in
higher rent (Laux, 2007).
The model that Laux (2007) implemented shows a significant correlation between
the board independence and CEO turnover. Thus, the more independent the board, the
PAGE 16
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
more recurrent will be the CEO turnover. This paper takes a common market opinion
about those boards that are controlled by external directors are more efficient controlling
CEO’s performance than boards that are monitored by internal directors. It is very
interesting that the authors put special attention on cost/benefits of independent boards, in
order to understand more variables (Empirical) involved in the CEO turnover concept.
The first point for the board to discover is; the capabilities of the CEO of learning,
understanding and whether the CEO had enough time to adapt, so with this information
take the decision whether he/she should be replaced or not. The previous point is
fundamental for the relation between Board and Shareholders, because the fact to expel a
CEO is very expensive for the company (Severance pay) and the effort for educating an
expert is eroded. If the board is considered that have fired in an inappropriate way, the
communication link within the company is harmed. On the other hand, fire to low
performance CEO has positive impact (Laux, 2007).
The previous paragraph defines dependency of the board in the case when the
board has stimulus to keep the Chief as expense of the company. It is developed when the
CEO’s success is associated to the director’s success. The consequence of this fact is that
the board become passive and takes the role of an inactive player in the company. In
contrast, an independent board would try to maximize the benefits of the shareholders.
The research conclude that dependent boards fires less poor performing CEOs than
independent boards, given this point the shareholders benefits from the dependency of the
board, protecting shareholder interests within the firm (Laux, 2007). The topic helps to
understand better some factors of CEO’s turnover; nevertheless our research does not
focus on the board independence variable.
Regarding to audit fees in turnover CEO companies, CEO Turnover and Audit
Pricing article (Huang et al., 2014) has been found that the fees paid for CEOs that were
force to leave were importantly higher than companies that were not force to remove the
CEO. Besides, they did not find differences in audit fees between firms with voluntary
turnover and firms without turnover. This topic helps to understand in a better way the
topic, however the audit fees in CEO’s turnover is not a variable for our research.
The paper wants to extend the literature of the link bounded by CEO’s turnover
and the audit fees. It is important to take in consideration that CEOs that are force to
leave the managerial position are considered situations with higher risk than companies
PAGE 17
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
with voluntary CEO turnover. The audit would claim for higher fees to forced turnovers
because this required better performance management (higher risk), instead voluntary
turnover that is recognized as a pure transition procedure in the company. The
substitution of the CEO is a matter of risk because generates some uncertainty about the
new leader and his/her capabilities within the firm as effectiveness and future strategy; all
of this, it is interpreted as risk, given the responsibility of auditors to select the right
candidate. Given this uncertainty the audit companies increase the audit fees as extra
costs (Huang et al., 2014).
In order to give some statistical information about CEO’s turnover and the trend
of this concept, a continuation it is shown a breakdown of number of CEO’s turnover in
the different regions of the world and proportions as an aggregated analysis (Figure 1).
Source: Booz and Company
Figure 1 Regional CEO Breakdown
PAGE 18
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
It can be seen clearly that the tendency of the market is converging across the
globe in different industries. We can see that the proportion of Chief Executive Officers
those are changed through the years in Europe, have been stabilizing to similar levels to
the position of United State market and the Japanese market. In the specific case of the
rest of Asia the tendency also is going to reach levels as Japan and U.S.
PAGE 19
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
In the Figure 2 (U.S Turnover) is illustrated the tendency of CEO’s Turnover
through six consecutive years. This is an empirical example to have a better
understanding of how CEO turnover have behaved in these years in the U.S market.
Source: Challenger, Gray & Christmas, Inc.
Figure 2 CEO departure by month, Year in U.S.- based companies
The January of 2004 achieved the highest number of CEO that left the position.
The percentage of turnover have increased 32,3% compared with the largest level of
turnover in the previous four years. The number of CEOs that have departed was 131 in
January 2014, corresponding to a report elaborated by the company Challenger, Gray &
Christmas, Inc. Thus, with this information it is possible to show graphically the concept
of CEO turnover in the US market.
It is very difficult to find information of this topic in emerging markets, in
accordance to this the lack of information supports the idea of literature gap that our
research is intended to study.
PAGE 20
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
1.2.2 Performance evidence in CEO’s turnover
Performance evidence in CEO’s turnover literature have found a very strong
support in the idea that in the US market, companies with not habitual CEO turnover
have better performance than those enterprises without a CEO change. In this case, a
question that has received less attention is how CEO turnover impacts the performance of
the company in the next periods. The CEO turnover is because of unsatisfactory
performance of the incumbent, in this way it is important to understand how this
adjustment in the management will impact the futures quarters in the U.S.A. market in the
period of 1996-2004, whether they will improve or not. The other question that they
wanted to address is identifying the different types of turnover, (voluntary or not
voluntary) and studies the differences of how they impact the performance of the firms
(He et al., 2001).
The contribution of this study relies on:
This investigation was the first research that used Efficient Frontier analysis in
corporate governance studies. This approach has the advantage to incorporate to the
analysis not only public companies, but private (Not only stocks for public firms)
companies as well using measures as Cost Efficiency and Revenue Efficiency. Other
advantage of this research is that they might measure performance through
multidimensional enterprises’ production process. The methodology was pioneer in this
paper, showing advantages for subsequent investigations.
Secondly, the study has helped to get a better understanding in an industry that has
had little research of the relationship between CEO turnover and the firm performance in
the insurance industry. Besides, it helps to study turnover performances post the period of
change of CEOs in insurer enterprises.
The main conclusion of the paper is that firms with not regular turnover show
greatly better performance than those companies without CEO turnover. The positive
impact of the performance of CEOs during the turnover is clear and statistical significant
in this research. A deeper understanding of different competitive industries is needed for
futures investigation (He et al., 2001).
PAGE 21
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Extending other aspects that stimulate CEO Turnover is the family firm’s factor
that accounts roughly one third of the S&P 500 market (Anderson and Reeb, 2003).
Splitting ownership and control of the company would help to family firms to deal with
some conflict of interest. In the research of Chen, Cheng and Dai (2013), they analyze
how family ownership and control of the company affect CEO turnover decisions. They
focus on CEO turnover decisions because this is one of the most important managerial
decisions with long term implications, as operations, investments and financing decisions
(Huson, Parrino and Starks, 2001). The findings of the paper are: in professional family
member companies was found a higher sensitivity of Chief turnover-performance
comparing to nonfamily firms. On the other hand, family CEO firms tend to protect the
unsatisfactory performing CEO family member to be expelled from the company, thus
showing a lower sensitivity of CEO turnover within the company.
The authors highlighted two situations that were very interesting for them prior to
start the analysis. Agency Conflicts could be the main reason of the enterprise failing to
discard poor performing CEO that at the same time it would imply a very costly agency
conflict issue. On the other hand, the scholars summarized that Agency Conflicts are
reflected by the low sensitivity CEO turnover performance. This implies that the bigger
the conflict of the agent, the harder would be change the poor performing CEO (Chen et
al., 2013).
The two main things to study are how family ownership and control are related
with agency conflicts. Family participant in the company have long-term interest in the
firm (Firm value). Consequently, this leads a stricter monitoring of managers, increasing
the chances to remove underperforming CEOs, contrasting with nonfamily enterprises.
Family shareholders have enough influence in the company generally because of
concentrated equity of the company. In this case, family members could tend to run with
their own interests, do not paying that much attention on the other shareholders. These
two effects entail different things; the sensitiveness around CEO turnover depends
whether there is a CEO family member or not. Whether the CEO is a family participant,
the incentives to squeeze higher rents and stay longer as a CEO are much more attractive
than the scenario where there is not a family member involved.
To solve this question, the study considers family CEO firms and Professional
family CEO firms with data of U.S.A. companies during the period 1996-2005. It is
PAGE 22
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
summarized the importance of separating professional CEO family firms from family
CEO firms for a better and clearer understanding of the topic. The results suggest that
family firms are less transparent at the moment to give managerial forecasting than nonfamily firms (Chen et al., 2013).
1.2.3 Earnings Management evidence
Identifying the possible motivations of CEOs to manage earnings is a very
important point to understand earnings management phenomenon. It is required an
analysis to investigate whether the motivational factors apply for Russian and Chilean
companies. In the investigation of Andersson & Lilja (2013), the authors present the
following CEO motivations to earnings management in the Swedish market “…factors
that have larger impact of the use behind earnings management. Suggestions for such
additionally control variables are: CEO’s salaries/bonuses/option programs, CEO’s
education, and previous experience as well as previous industry belonging to see whether
different backgrounds and biases would be a reason”. (Fredrik Andersson & Viktor Lilja,
2013)
An entity called SOX (2002) was created given the earnings management
practices used by CEOs. The Sarbanes Oxley (SOX) acts in the United States and was
developed to supervise public firms, with the purpose to prevent arguable behavior. The
main goal of this entity is assure the investor interests using their capabilities to foreseen
fraud and risk of bankruptcy.
In our study we want to find the effect in Chilean and Russian companies, thus we
can find some previous evidence that have been shown that CEO’s use two methods to
manage earnings in companies around the world. The methods to manage earnings are the
accruals management of the company and operational activities in the companies. The
objective of the method is to achieve the required earnings targets, nevertheless, the
earnings management has increased through the time, and in the meanwhile accruals
earnings management has been decreasing. This tendency could be possible because real
earnings management is harder to identify than accruals earnings management that was
described in the research Earnings Management around CEO Turnovers (Geertsema
Geertsema, Lont and Lu 2013).
PAGE 23
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The incentives of the CEO to manage earnings showed in the paper described
above, says that the new CEO would manage earnings downward, in order to show that
the company was mishandled and start a clear path of good performance. In contrast, we
can observe the incentives for the departing CEO is exactly the opposite. The incumbent
would want to manage earnings upward to get as much bonus as possible (or hide
deficient performance).
The earnings management takes place in the period of transition quarter of both
CEOs. The information was obtained from Audit Analytics. They compare earnings
management in a quarter where there is a change of CEO and a normal quarter. The main
finding is that the new CEO does not use accrual-based methods, but manages real
activities, in order to decrease the company earnings for the previous period. This
anomaly arises as soon as the transition period (quarter) takes place.
Analyzing nonroutine CEO turnover firms and routine CEO companies, it can be
concluded that nonroutine CEO enterprises are smaller in size and smaller in the ratio
ROA (Return on Assets) than companies those are already experienced on turnover.
Moreover, nonroutine CEO turnover firms use to be much more traditional in financial
reporting than the routine CEO turnover. Other important conclusion of the paper is that
the new CEO manages earnings more actively than the departing CEO through unusual
production cost and discretionary expenditures (Geertsema, Lont, Lu, 2013).
In order to reach a higher level of statistical power for the test, large number of
data was needed in this investigation. Higher statistical power is very necessary, because
earnings management is noisy data (Dechow et al. 2010).
The methodology that Geertsema, Lont and Lu (2013) use in their article
Earnings Management around CEO Turnovers to determinate earnings management of
real business activities is:
Overproduction: The overproduction decrease the cost of the products that are
ready to be sold (per unit basis), this effect would be translated as increase in
earnings for that specific quarter. However, the storage of the goods is limited and
the following periods this effect would be translated in underproduction for the
PAGE 24
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
next period. On the other hand, it is possible to under-produce in order to have the
exact opposite effect for the next period.
Delaying discretionary expenditures: The variable includes Research and
Development, Advertising and Sell, General and Administrative Expenditures.
The idea of delay discretionary expenditures would boost earnings for the
following quarter (Geertsema, Lont and Lu, 2013).
In order two have a clearer and deeper understanding of what we are looking for,
we will analyze two more cases about Earning Management and CEO turnover. We are
going to present two cases, the evidence of Korea and the evidence of Indonesia.
Korea’s case study placed four scenarios of CEO turnover, force or voluntary
departure of the CEO, and whether the incoming incumbent is promoted or hire from
outside. One interesting finding in this research is that the outgoing CEO in a “force”
departing scenario, managed upward the company earnings, but only when the incoming
CEO comes from inside of the company. In this specific case the new CEO also manage
downward earnings through discretionary accrual and real activities. Another conclusion
is that the external incoming CEO manages earnings upward in a voluntary CEO turnover
(Choi, Kwak and Choe, 2014).
One of the objectives of this study is to add an extra variable in the research,
differentiating whether the incoming CEO is from outside or developed his/her career
within the firm. The new variables provide to the research a better insight of the earnings
management around the CEO turnover. In a context where a routine departure of CEO is
held, the incoming professional (Promoted) has less incentives to manage earnings
downward to show previous poor performances in order to shine more in the firm. Also,
the fact that whether the departing Chief Executive Officer will be on the board
diminishes the probabilities for the new CEO to influence real business (Choi, Kwak and
Choe, 2014).
For the four types of CEO turnover, the authors studied the link between Earnings
Management and Turnover. The sample data collected includes 403 CEO turnover
enterprises with data from 2001to 2010 listed in the Korean Stock Exchange Market.
Summarizing this paper, the main conclusions in the Korean market are the
following; in a forced scenario the outgoing CEO manage earnings upward, but only is
PAGE 25
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
significant when the new CEO have been promoted. The Earnings Management in this
specific case is through discretionary accruals and not earnings management of real
business activities. Second, the authors could not find enough evidence of earnings
management in a context of peaceable departure of the CEOs. Again, in a “force”
departure scenario, the promoted CEO use earnings management downward, but there is
not enough evidence of earnings management in a “no forced” scenario. Finally, the
external new CEO uses earnings management upward in a context of peaceful departure.
This paper helps to understand the earnings management around CEO turnover in a
country where many players of the market think that firm’s accounting is not transparent
enough, blaming this issue to the crisis of Korea at the end of the 1990’s. Given the
previous problem described, new reforms were developed in Korea after the crisis,
focusing mainly in improving corporate governance (Choi, Kwak and Choe, 2014).
Evidence of Indonesia CEO turnover: The topic of this paper study accounting
performance and earnings CEO turnover in an Asian country that have not been examined
previously, Indonesia. The article uses data of Indonesian companies from 1998 to 2006.
The incentives for the board to change the CEOs are in order to reach better company’s
performance. For example, Indonesian National bank had passed through some obstacles
for three consecutive years of low revenues and also they were increasing the ratio Total
Debt to Total Asset. The firm overcame the situation as soon as the decision of CEO
turnover happened in the company. This is a very illustrative example where the company
succeeded after a change of CEO (Lindrianasari and Hartono, 2011).
Coughlan and Schmidt (1985) study the relationship between accounting
variables and CEO turnover. This approach will encourage to the board to put more
attention in the financial reports at the moment to take managerial decisions. The case
proves that accounting information and the CEO turnover are linked, the board will be
more confident to take decisions about the data of revenues made by accounting than
other information (Coughlan and Schmidt, 1985). Given the approach of Coughlan and
Schmidt (1985), the main result of the Indonesian Market’s study is provide information
about the main accounting ratios as Total Assets, Total Sales, ROA, ROE and Earnings,
and show how they have summarized that the accounting information is logical with the
turnover of the Indonesian firm decisions (Lindrianasari and Hartono, 2011).
PAGE 26
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Given the literature presented above, we can have a clear overview of the
importance of the main variables of our research (Figure 3). The Earnings Management
around CEO turnover in Russian and Chilean companies will give us an understanding of
how those companies is managed and whether the incoming and outgoing CEO manages
earnings in different ways.
Earning
Management
Incentives of
CEO
CEO Turnover
Research
Emerging
Markets
Performance
Source: Own elaboration.
Figure 3 Main aspects involved in the research
In companies, the CEO’s could have different incentives for Earnings
Management, it would depend whether the CEO is departing (They want higher bonuses
or dissimulate low performance) from the company or is the new manager taking change
in the company (Show a new beginning for the company with better numbers). Thus, we
can state a hypothesis for the following question: Does there exist earnings management
in a company at the moment of a change of CEO?
H1: The fact of change of CEO does not influence the level of earnings
management in a company in Russia.
H2: The fact of change of CEO does not influence the level of earnings
management in a company in Chile.
PAGE 27
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
As we are studying the markets of Russian and Chile, and our goal is compare
both markets in the context whether CEO’s in Chile manage differently the earnings of
the firms than in the Russian market.
CHAPTER SUMMARY
In this chapter the state of art of the investigation of Earnings Management around
CEO Turnover is described. The definitions of the main concepts involved in the research
and literature analysis are represented. Besides, it is shown some tendencies of CEO’s
turnover breakdown in the world and it is illustrated the specific example of North
American companies.
It is possible to identify a literature gap of Earnings Management and CEO
Turnover in Russia and Chile. The Research Proposal is devoted to extend this topic to
the Russian and Chilean market to find the importance of Earnings Management in these
two locations and the effects and consequences in Chile and Russian companies (First
Latin-American country under study and Russia).
The evidence illustrated in this chapter provides an empirical and practical
overview of the Earnings Management and CEO Turnover concepts. It is very important
to highlight that Russia and Chile are emerging countries with different factors that lead
the way that companies behave, in case we compare with the literature studied (Mainly
developed countries). The factors that could make the difference are such as the
Russian/Chile culture, the principles of doing business in Russia/Chile and the earnings
management incentives where there is CEO turnover.
Finally, the background presented of previous studies in this chapter is the first
approach defining the expected findings of our paper.
PAGE 28
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
CHAPTER II: DATA, METHODOLOGY, RESULT
ANALYSIS AND IMPLICATIONS OF EARNINGS
MANAGEMENT
This research is going to be conducted as a Quantitative Research. Thus, the
empirical investigation is going to look for the answer of Earnings Management around
CEO turnover via statistical technique. In contrast, it exist the Qualitative Research where
the research asks broad questions and the nature of the collecting information itself is
word data. The main objective at this stage is to develop a Regression Model to estimate
the existence and impact (if there is any) of companies under this topic.
The idea of developing a Regression Model is to connect the empirical
information taken from the financial statements and mathematical expressions
(Regression Models) searching for the relation among the variables involved in Earnings
Management.
2.1 DATA SELECTION
2.1.1 Data sources
International Financial Reporting Standards (IFRS) are standards of financial
accounting character, governing how to prepare and present information on the
accounting facts of firms to interact with their environment. The adoption of this financial
standard took place in Russia in the year 2012. IFRS are part of the Russian legislative
framework (Federal Law 208-FZ) to consolidate the financial statements of all the
companies that are listed in The Moscow Interbank Currency Exchange or MICEX.
MICEX is one of the largest entities of universal values in the Russian Federation and
Eastern Europe. MICEX opened in 1992 and is the leading stock exchange in Russia.
The Chilean market had adopted the International Financial Reporting Standards
since 2008. In this sense and for the analysis of our research we are going to take the data
from 2008, where all the companies from Chile are shown with IFRS standards.
Regarding the Russian market we also are going to take data from 2008 to set the CEO
PAGE 29
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
turnover, but in the case, the companies that report before 2012 (date when was adopted
IFRS in Russia) are going to be consider only those ones that report with IFRS standards.
Thus, in our study we want to be consistent with the data, so we took as for
Chilean and Russian companies within the same period of time for turnover selection data
(2008 – 2014). Besides, the companies that are part of the financial industries are
excluded from the studied sample.
Given the nature of the research the type of data required in the investigation is
described as Secondary Data. The data for the investigations is going to be collected
from the Database of GSOM Virtual Library. The data required for measure Accrualbased and real Earnings Management is collected from the financial statements and
annual reports from Datastream database of Thomson Reuters. Later in order to
determinate the regression model, we use the financial statement prepared by the
companies (IFRS standard) for the regulators of the different countries.
The number of companies to analyze whether they had turnover in the studied
period is 484 firms for the Russian market and 433 companies for the Chilean market.
Table 1 Russian companies’ selection
Initial sample
484 companies
(-) Financial firms are excluded
28 companies
Final sample
456 companies
Table 2 Chilean companies selection
Initial sample
433 companies
(-) Financial firms are excluded
38 companies
Final sample
395 companies
PAGE 30
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.1.2 Inclusion of Dataset
The second part of the data selection is identifying the CEO turnover within the
studied period for the Russian and Chilean markets. This information is collected
manually from the Database Thomson One in the Officers & Directors field. We consider
all the companies described in the previous section and identify the exact dates of CEO
turnover and the names of the new manager.
Companies with CEO turnover events are considered in our dataset only whether
they are not in the following scenarios:
CEO with less than 12 months in the company
Financial institution are not considered
Two (or more) consecutive turnovers are not considered
The summary of the number of companies considered per country as a sample are
the following:
Table 3 Summary of Russian turnover selection
Initial sample (w/o financial firms)
456 companies
(-) companies without turnover in the selected period or information
370 companies
Final sample
76 companies
Table 4 Summary of Chilean turnover selection
Initial sample (w/o financial firms)
395 companies
(-) companies without turnover in the selected period or information
311 companies
Final sample
74 companies
PAGE 31
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.2 METHODOLOGY
For measuring Earnings Management the literature reviewed presented three main
m e t h o d s : Accrual-based Earnings Management, Real Management through
Overproduction and Real Management through Delaying discretionary expenditures. The
three options will be described below; it is going to be explained each one and afterward
selected one of the options for our investigation.
2.2.1 Accrual-based Earnings Management
The Accrual-based Earnings Management is one of the three ways we are going to
study Earnings Management. Discretional accruals is how could be managed the earnings
in the company as Accrual-based.
We are going to define the Discretional accrual concept as the difference between
the total accrual of the company in the required period, and the normal level of accrual.
For this process we have to define how to calculate the normal level of accruals and we
are going to follow the methodology of Dechow et al. (1995).
The model for each industry-quarter is the following one (Cross-sectional model):
Total Accrualt
∆ St
PPE t
1
=α 0 +α 1
+α 2
+α 3
+ε {1 }
A t −1
A t −1
A t−1
A t −1 t
The Total Accrualst is earnings before extraordinary items and discontinued
operations minus the operating cash flow in the period t. In the denominator we have A t-1
that is the Total Assets in the previous quarter (t-1). ΔSt reflect the change of the revenue
of both periods and finally the PPEt is the gross property, plant and equipment of the
company.
In order to calculate the normal level of accruals we can use the following
regression:
Norm accrualt =α^ 0 + α^ 1
∆ St −∆ ARt
PPE t
1
+ α^ 2
+ α^ 3
{2 }
A t−1
A t −1
A t −1
PAGE 32
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
In this part ΔARt is showing the change in account receivables and the DAt reflect
the change between total accruals and the normal level of accruals.
Geertsema, Lont and Lu (2013) found a contrast between accruals earnings
management and real earnings management. In the first one, it is more pronounced for
non-routine CEO turnovers. In contrast, real earnings management is significant
following both routine and non-routine CEO turnovers. Thus, we decided to focus better
in real earnings management, instead of accrual earnings management.
In other words, the earnings management has increased through the time, and in the
meanwhile accruals earnings management has been decreasing (Geertsema, Lont and Lu
2013).
2.2.2 “Real” Earnings Management
The methodology described a continuation is devoted to determinate the CEO
turnover and earnings management using the methods of Geertsema, Lont and Lu (2013)
in the article Earnings Management around CEO Turnovers. The variables to measure
real activities are:
Overproduction: The overproduction decrease the cost of the products that are
ready to be sold (per unit basis), this effect would be translated as increase in
earnings for that specific quarter. However, the storage of the goods is limited and
the following periods this effect would be translated in underproduction for the
next period. On the other hand, it is possible to under-produce in order to have the
exact opposite effect for the next period.
The normal level of production cost could be defined as the following regression
model (establish relation between the variables to find the normal level):
∏ ¿t =α +α
A t −1
0
1
S
∆ St
∆ S t−1
1
+α 2 t +α 3
+α 4
+ε t { 3 }
A t −1
A t−1
A t −1
A t−1
¿
The production costs can be described as PRODit = COGSit + ΔINVit, where
COGS is the cost of goods sold and ΔINVit is the changes in inventory in two
period of time.
St Is the revenue in the period t and
PAGE 33
At
is the amount of
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Total Assets in t. Thus, we are able now to find the normal level of production
costs.
Delaying discretionary expenditures: The variable includes Research and
Development, Advertising and Sell, General and Administrative Expenditures.
The idea of delay discretionary expenditures would boost earnings for the
following quarter (Geertsema, Lont and Lu, 2013). The equation is the following:
DiscExp t
S
1
=α 0 +α 1
+α 2 t +ε t { 4 }
A t −1
A t −1
A t −1
Discretionary expenditure (DE) in quarter t could be defined as the sum of R&D
and SG&A. Abnormal discretionary expenditure in t (Ab_DiscExpt) is the
regression residual from the equation (difference of actual value and normal
value), with an negative sign for make easier the interpretation. Thus, a low level
of abnormal discretionary expenditure means that the CEO manages downward
earnings through an increase in discretionary expenditure for that period.
Now analyzing a bit further these two options for real earnings management,
Roychowdhury (2006) found a problem in modeling the current discretionary expense.
Current sales leads a mechanical problem if companies manage sales upwards to increase
earnings in a certain period, resulting in an important lower residuals (the difference
between the variables observed and the predicted ones) from running the regression
model even though when they do not decrease discretionary expenses as the methodology
says (Roychowdhury, 2006).
Finally, the methodology selected to identify earnings management around CEO
turnover in Chile and Russia is Real Earnings Management through overproduction.
2.2.3 Variables description
As in our investigation we decided to study Earnings Management through
Overproduction, we are going to describe all the variables involved to develop our
Regression Analysis.
PAGE 34
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The dates that we are aware for our research are the Cut-off date, this date means
that is the last date that it is possible for managers to manage earnings in the company.
For real earnings management, it will be the cut-off date in the balance sheet.
The first part is identify the normal level of production where is involved COGS
and change in Inventory;
PRODit = COGSit + ΔINVit
{5 }
COGS is the cost of goods sold and ΔINVit is the changes in inventory in two
period of time.
In the literature investigated, Geertsema, Lont and Lu (2013) and Hazarika et al.
(2012) as in the majority of the literature reviewed stated that exist other factors that
could influence the earnings management and that are correlated with the measure error
in Earnings Management. Given the previous reason, we study some control variables
and check if they fit in our model. These variables are the following; ROA, Leverage,
Revenues and ROE.
Return on Assets (ROA) is a measure of performance that could be obtained from
the ratio of earnings before interest and taxes to total asset. It will be interesting to
analyze when the ROA ratio is higher, whether prior change in CEO turnover or after. In
this direction we can also identify whether managers try to capitalize more to show a
better performance. However, this variable is already included in our production model,
thus the variable cannot be included as a control variable for the correlation with the
model.
ROA=
Operating Income
{6 }
Total Assets
Revenues, as an alternative measure of size has the same problem that the control
variable ROA. For the previous reason was considered exclude it from the analysis of our
research.
The third control variable analyzed is Return on Equity (ROE), this variable is a
performance indicator that it is an assessment for the success of the company in its
management for equity restoration. The calculation of the value of ROE is obtained from
PAGE 35
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
the ratio of company earnings divided by total equity. This variable is believed to have a
negative relationship with turnovers, in that the failure in the return of capital would
contribute as one reason for a turnover in the company (Lindrianasari and Hartono 2011).
We include this variable in our model and it was significant.
Finally, we analyzed the variable Leverage, Jensen (1986) stated in his research
that high level of leverage reduces the chances of opportunistic behavior of earnings
management of CEOs. On the other hand, some other empirical studies as Press and
Weintrop (1990), Sweeney (1994) and DeFond and Jiambalvo (1994) imply that
companies with high level of leverage are more incline to manage companies’ earnings.
We consider this variable in our model, but it did not explain it properly, it was decided
that the leverage variable is excluded from our research.
In the next table we are going to summarize the variables involved in this research
and give the definition of each one.
Table 5 Summary of variables
Variable
PRODit
Definition
Production Cost is a dependent variable in the regressions specified as
the sum of Cost of Goods sold (COGS) and the change of the inventory
(ΔINVit) in two consecutive periods.
At
Dependent variable in the regression model that shows the Total Assets
St
of the companies
Dependent variable in the regression model that shows Revenues of the
ROE
companies
Control variable that is obtained from the ratio of company earnings
divided by total equity
After the selection of all these variables, it is important to define the periods of the
study in more detail. The moment of the CEO turnover will be denoted by t = 0, this
means the first year where the outgoing CEO has not power to influence in the company
financial statements anymore. At the same time, it is the year that the incoming CEO take
control over the financial statements of the company. The last year of the outgoing CEO
in charge it is going to be denoted by t = - 1.
PAGE 36
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
For the earnings management metrics, we will examine three periods. First, we
study the last year period of the outgoing CEO in charge in the company (t = - 1).
Second, the first year of the incoming CEO (t = 0). Finally, we are going to include the
following period of the new CEO (t = + 1), in order to analyze if the earnings
management is concentrated in the turnover period per se or if the earnings management
is persisting through the next year.
Consolidating the Database of our research, all the companies that have no shown
the necessary information of the investigation have been excluded of the dataset. For
example, the turnovers that occurred in 2014 are not considered, because there are no two
periods of the new CEO for analysis. Other reason for excluding some data was that some
income statements did not show some key data for some companies.
The final number of companies considered per country as a final sample is the
following:
Table 6 Summary of Russian final sample selection
Previous sample
76 companies
(-) companies without necessary information
36 companies
Final sample
40 companies
Table 7 Summary of Chilean final sample selection
Previous sample
74 companies
(-) companies without necessary information
24 companies
Final sample
50 companies
PAGE 37
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.3 DATA ANALYSIS AND RESULTS
2.3.1 Sample statistics
The statistics of Chile and Russia are described in the section, in order to compare
characteristics of both, the Russian and Chilean markets.
As a size measure we can see that the average value of Total Assets of Russian
companies is about ten times the size of the Chilean companies with an average of 15,390
and 1,595 million US dollars respectively. The case of Revenues as could be intuitive
follow the same direction as Total Assets (9,216 million dollars for Russia and 845
million dollars for Chile). If we put attention on ROA ratio, we can see that Russian and
Chilean markets have similar returns, being slightly higher in Chile (5,9% in Chile and
5,2% in Russia).
As a indicator of performance (Return on Equity), the Russian companies showed
to be more efficient with an average of 11,15% in the meanwhile Chile showed an
average of 7,46%. A significant difference of almost 3,5% of efficiency of ROE.
Regarding the level of Leverage of both markets, Chile has 45,7% of average
leverage and in the Russian market has a level of leverage located in 51,6%.
The following table is showing a comparison of Chile and Russia on the
dimensions we were describing above.
Table 8 Sample Statistics
Total Assets
(Million USD)
Revenue
(Million USD)
ROA (%)
ROE (%)
LEVERAGE (%)
Country
Chile
Russia
Chile
Russia
Chile
Russi
a
Mean
1595
15391
845
9217
6%
5%
7%
11%
46%
52%
20
0
10
-25%
-11%
-188%
-20%
7%
18%
240372
10804
108992
103%
22%
144%
43%
94%
88%
Minimum
1
Maximu
m
16040
PAGE 38
Chile
Russia
Chile
Russia
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The full descriptive analysis of Russia and Chile is presented in the appendix 3
and 4.
Correlation tables are presented in the Annexes 5 – 6. This table presents the
correlation between variables for the Russian and the Chilean sample. As we excluded
ROA and LEVERAGE from the model we do not consider it in our analysis of
correlations. There is a strong correlation between Revenues and Total Assets (TA) as
could it be intuitive. In general all the rest of the variables have little correlation.
2.3.2 Regression Model
The model of earnings management through overproduction shows how a decrease
(or vice versa) in cost of products would be translated into earnings management for a
certain period. The model selected for this analysis is multiple regression model, this is a
statistical process that explain the relation among the variables. Multiple Regression
Model explains how a dependent variable (Production costs) is related with different
independent variables, included the control variable.
The equation is showed as followed:
∏ ¿t =α +α
A t −1
0
1
S
∆ St
∆ S t−1
1
+α 2 t +α 3
+α 4
+α 5 ROE+ε t {7 }
A t −1
A t−1
A t −1
A t−1
¿
This equation was run in each company in Russia and in Chile of the sample for the
period t. The model presented by Geertsema, Lont and Lu (2013) shows the performance
of the manager within the company. On the other hand the control variable ROE is
intended to show the performance of the shareholders within the company, in accordance
that for them is important earnings management within the company as well.
The multiple regression model was run on SPSS statistic software for the three
periods of analysis in the Chilean market and Russia market, where t = 0 is the first year
of new CEO in charge, t = -1 is the last period of the outgoing CEO in and finally t = +1
is the second year of the incoming CEO in the company. In this way we were be able to
PAGE 39
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
analyze the Earnings Management in these markets for several periods in the process of
CEO turnover.
These different models explains for different periods of times how is managed the
earnings within the company through over production (Production costs).
The results of ANOVA, Significance of the model and Model summary of each
country for each period is presented in Annexes 7 – 8.
PAGE 40
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.3.2.1 Results for Chile
We present in the next table, the result of the model of Chile for the first period of the
new CEO in the company.
Table 9 Regression Model Chile t = 0
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
1
Std. Error
(Constant)
,047
,017
1/At-1
,254
,035
St/At-1
,079
∆
St1/At
∆
St-
t
P- Valueb
Beta
2,695
,010 (***)
,656
7,343
,000(***)
,023
,383
3,503
,001(***)
,180
,074
,263
2,417
,020(**)
-,014
,066
-,020
-,204
,839
-,107
,043
-,248
-2,491
,017(**)
-1
1/At-1
ROE
a. Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
The model is statistically significant at a level of 1%, explaining 66,2% of the
variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis.
The dependent variables have a positive relation with the independent variable
and the variables are statistically significant, with the exception of one variable that is not
significant in this model. The theory says that incoming CEO is willing to decrease cost
production through over production to show that the company was mishandled. The
results are consistent with the theoretical background. (Reitenga and Tearney, 2003)
Regarding the control variable ROE has a negative relation with the production
costs and is statically significant at 5%. The correlation of this control variable has a
positive magnitude with the dependent variable (Production Costs).
The hypothesis of our research established “The fact of change of CEO does not
influence the level of earnings management in a company in Chile” Thus, for this model
the hypothesis is rejected given the significance of the model where it exists Earnings
PAGE 41
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Management for the first period of the incoming CEO in the direction that the theory
stated.
The model illustrates the result of Chile for the period of the departing CEO is
presented in the next table.
Table 10 Regression Model Chile t = -1
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
1
Std. Error
(Constant)
,043
,021
1/At-1
,168
,043
St/At-1
,077
St1/
St-
∆
t
P- Valueb
Beta
2,057
,046 (**)
,474
3,918
,000 (***)
,035
,437
2,238
,030 (**)
-,063
,061
-,139
-1,041
,303
-,023
,071
-,062
-,319
,751
-,004
,040
-,013
-,097
,923
At-1
∆
1/At-1
ROE
a. Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
The model is statistically significant at a level of 1%, explaining 40,0% of the
variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis.
All the variables that are statistically significant are positively related NOT being
consistent with the theory presented in the second chapter. As Graham et al,. (2005)
presented in his research, the CEO that is departing is inclined to do the opposite as new
CEO. Departing CEO would trick real business activities upward to leave the charge and
get the highest possible bonus. The analysis of the control variable is not applicable in
this model for a lack of significance.
Finally, for the hypothesis we can say that it is rejected for the significance of the
model and the Earnings Management is not aligned with the background theory.
PAGE 42
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
PAGE 43
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The following regression model illustrates the result of Chile for the second
period of the new CEO.
Table 11 Regression Model Chile t = +1
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
1
Std. Error
(Constant)
,078
,023
1/At-1
,218
,036
St/At-1
,032
St1/
St-
∆
t
P- Valueb
Beta
3,334
,002 (***)
,653
6,043
,000 (***)
,029
,352
1,102
,277
-,057
,032
-,628
-1,795
,080 (*)
-,111
,066
-,271
-1,688
,098 (*)
,011
,013
,090
,820
,417
At-1
∆
1/At-1
ROE
a. Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
The model is statistically significant at a level of 1%, explaining 49,3% of the
variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis.
There are two variables statistically significant with positive relation with the
dependent variable and other two as well significant, but with negative relation.
For the hypothesis of the investigation, we can say that it is rejected for the
significance of the model and the existence of Earnings Management is present in this
period for the Chilean market.
There is no conclusion about the control variable again, given that it is not
statistically significant in this model.
PAGE 44
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.3.2.2 Results for Russia
In this section we present the result for the Russian market. In the next table, the
result of the model of Russia for the first period of the new CEO in the company is
showed.
Table 12 Regression Model Russia t = 0
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
(Constant)
,069
,146
-19,410
16,506
,187
St1/
St-
1/At-1
St/At-1
1
∆
Std. Error
t
P- Valueb
Beta
,472
,640
-,434
-1,176
,248
,065
,765
2,879
,007 (***)
-,011
,266
-,011
-,041
,968
,080
,318
,051
,251
,804
-,010
,662
-,003
-,016
,988
At-1
∆
1/At-1
ROE
Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
a.
The model is NOT statistically significant (Significance of 10,5%), it means that the
result is likely due randomness. It is not possible to verify the hypothesis, where the
CEOs do not influence the level of earnings management in a company.
The model is not able to determinate the existence of Earnings Management for this
period in the Russian market.
PAGE 45
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
In the next table, the result of the model of Russia for the last period of the departing
CEO in the company is presented
.
Table 13 Regression Model Russia t = -1
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
(Constant)
,167
,107
-41,737
18,541
,253
St1/
St-
1/At-1
St/At-1
1
∆
Std. Error
t
P- Valueb
Beta
1,564
,127
-,771
-2,251
,031 (**)
,069
1,061
3,676
,001 (***)
,176
,263
,112
,669
,508
-,136
,087
-,310
-1,558
,128
-1,082
,272
-,491
-3,983
,000 (***)
At-1
∆
1/At-1
ROE
Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
a.
The model is statistically significant at a level of 1%, explaining 51,1% of the
variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis.
Regarding the independent variables, there are only two variables that are
statistically significant and with mixed relationships (Positive and Negative). The
hypothesis of the model is rejected given the significance of the model where there exists
Earnings Management for the first period of the incoming CEO.
The control variable ROE has a negative relation with the production costs and it
is statically significant at 1%.
PAGE 46
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
As a final model we present in the following table the results of Russia for the
second period of the new CEO.
Table 14 Regression Model Russia t = +1
Coefficientsa
Model
Unstandardized
Standardized
Coefficients
Coefficients
B
(Constant)
1/At-1
,052
,118
-17,521
22,145
,203
∆
St1/At
∆
St-
P- Valueb
Beta
,438
,664
-,283
-,791
,434
,076
,745
2,678
,011 (**)
-,100
,218
-,090
-,456
,651
,013
,283
,010
,045
,964
-,273
,480
-,101
-,569
,573
St/At-1
1
Std. Error
t
-1
1/At-1
ROE
a. Dependent Variable: Prod/At-1
b. * - significant at 10%, ** - significant at 5%, *** - significant at 1%
The model is statistically significant at a level of 5%, explaining 26,8% of the
variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis.
There is only one variable statistically significant with positive relation with the
dependent variable. There is no conclusion about the control variable (ROE), given that it
is not statistically significant.
For the hypothesis of the investigation, we can say that is rejected for the significance
of the model and the existence of Earnings Management is present in this period for
Russia.
PAGE 47
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.4 REGRESSION MODEL RESULT DISCUSSION
In the next sections are presented the result of the Chilean and Russian markets, in
order to interpret and compare the results obtained.
2.4.1 Analysis for first period of new CEO
The following table presents a comparison of the results for the period t = 0, first
period of the new CEO.
Table 15 Comparative results t = 0
Model t = 0
Chile
B
Model significance
P- Valuea
B
P- Valuea
,662
,000(***)
,226
,105
(Constant)
,047
,010 (***)
,069
,640
1/At-1
,254
,000(***)
-19,410
,248
St/At-1
∆
∆
ROE
a.
Russia
,079
,001(***)
,187
,007 (***)
St1/At-1
,180
,020(**)
-,011
,968
St-1/At-1
-,014
,839
,080
,804
-,107
,017(**)
-,010
,988
* - significant at 10%, ** - significant at 5%, *** - significant at 1%
As a first analysis we can see that there are mixed results for the model
significance. For the Russian sample, the model is NOT statistically significant (P-Value >
10%), it means that the result is likely due randomness. The model developed is not able
to verify the hypothesis, where the CEOs do or do not influence the level of earnings
management in a company. In other words, the model is not able to determine whether
the new CEO in Russia does Earnings Management in his/her first period in charge.
The model for Chile is statistically significant at a level of 1 %, explaining in
66,2% the variance of Production Costs, that means that the hypothesis of the research
stated that Earnings Management exist and reported that the amount of Production Costs
is positively related for the first period of the new CEO. Rephrasing, Chilean managers
PAGE 48
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
do Earnings management around CEO turnover through over production for the first
period of the incoming CEOs.
The second part of the analysis is to study the relation of the variables between
each other, in order to analyze the direction of the Earnings Management and to check the
consistency of the results with the theory. The analysis for this period is going to be
possible only for the sample of Chile.
The Chilean dependent variable has a positive relation with the independent
variable. The literature analysis proves that the new CEO in charge is inclined to do
underproduction, in order to increase cost of production to show that the company was
mishandled. The results are consistent with the theoretical background (Reitenga and
Tearney, 2003).
The Chilean market does underproduction practices to increase the cost of goods
sold to show earnings downward. The manager wants to be perceived that he received the
company in not very good and healthy stage, and it will be his duty to show that the
company will be manage better in the next periods. The idea is supported by existing
studies that find downward earnings management by the new CEO (Dechow and Sloan,
1991; Pourciau, 1993; Brickley et al. 1999; Reitenga and Tearney, 2003; Conyon and
Florou, 2004).
Unfortunately, it is not possible to compare the market under study. We know that
in the Chilean market exists Earnings Management in line with the theory, but the model
proposed for Russia in this period was not able to explain whether exist or not Earnings
Management. Comparison between the countries is not possible for this period.
2.4.2 Analysis for second period of new CEO
The results presented in this section are devoted to the second period of the new
CEO for both markets, the Chilean and Russian. The reason for studying this period is
because we wanted to analyze whether earnings management are concentrated in the
turnover period per se or the earnings management is persisting through the next year.
This idea was presented in the research of Choi, Kwak and Choe 2014. Nevertheless, for
overproduction earning management the storage of the goods is limited and the following
PAGE 49
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
periods the earnings management effect would be translated in underproduction for the
next period.
The following table presents the comparison of the two samples for the second
period after the turnover.
Table 16 Comparative results t = + 1
Model t = +1
Chile
B
Model significance
P- Valueb
B
P- Valueb
,493
,000 (***)
,268
,050 (*)
(Constant)
,078
,002 (***)
,052
,664
1/At-1
,218
,000 (***)
-17,521
,434
St/At-1
,032
,277
,203
,011 (**)
St1/At-1
-,057
,080 (*)
-,100
,651
St-1/At-1
-,111
,098 (*)
,013
,964
,011
,417
-,273
,573
∆
∆
ROE
b.
Russia
* - significant at 10%, ** - significant at 5%, *** - significant at 1%
The model for Chile is statistically significant at a level of 1%, explaining 49,3%
of the variance of the Production Costs (Dependent Variable) around its mean, through R
Square analysis. On the other hand we have the model of Russia that is also statistically
significant, but at a level of 5%, explaining 26,8% of the variance of the Production
Costs. In other words, the hypothesis of the research is rejected and managers of both
markets do Earnings management around CEO turnover through over production in the
second period of the new CEOs.
The control variable is not significant in both samples, thus we could not conclude
neither analyze the results of this aspect of the models.
The model of Chile has two variables statistically significant with positive relation
with the dependent variable and other two as well significant, but with negative relation .
Taking into account that the independent variables are the result of rearranging the model
variables that predict the unit production cost for predicting earnings management
through real activities, the independent variables are going to be analyzed as a whole.
PAGE 50
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Assuming that the numerical values (value that join to the coefficients of the models) are
the same (for comparison purposes), the sum of the coefficient magnitudes will lead the
direction of the model. In this specific case (Chile), the sum of positive coefficients is
higher that the absolute value of the negative coefficients. The result leads to the
conclusion that in the Chilean market for the second year of the new CEO, managers tend
to manage earnings downwards.
Analyzing the persistency of earnings management after the turnover transition
period, we can see that earnings management for this period persists over, intensifying
the idea of blame poor performance of the previous CEO.
The model of Russia has only one variable statistically significant with positive
relation with the dependent variable. Thus, the model implies that managers do Earnings
Management downwards. Managers go through underproduction, increasing the cost of
unit goods sold managing the earnings downwards. The result in this model is contrary to
what the literature stated. Managers are inclining to show poor performance in the second
period, instead to show resiliency.
This result could be explained by some assumptions. In the first period of the new
CEO in charge our model was unable to determine the relation between the variables in
the model. In case of existing earnings management in the first period the contrary effect
happens in the second one. Overproduction earnings management has the limitation of
inventory where storage of the goods is limited and the contrary effect in the next periods
the earnings management effect would be translated in underproduction for the next
period.
As second possible reason, we will see in the next section the last period of the
outgoing CEO, the CEOs are managing earnings upwards, and as the limitation of
inventory of goods the contrary effect is expected for the next period or periods, it would
depends how intensive is the earnings management in this period. As we cannot conclude
about the period t = 0, conclusions are only under scenarios. The new CEOs are worried
about audit companies that could lag the decision of earnings management for the next
periods.
PAGE 51
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.4.3 Analysis for last period of outgoing CEO
Finally, we analyzed the last period of the outgoing CEO. In the next table is
presented the result of the Chilean and Russian markets, in order to interpret and compare
the results obtained.
Table 17 Comparative results t = - 1
Model t = -1
Chile
B
Model significance
P- Valuec
B
P- Valuec
,400
,000 (***)
,511
,000 (***)
(Constant)
,043
,046 (**)
,167
,127
1/At-1
,168
,000 (***)
-41,737
,031 (**)
St/At-1
,077
,030 (**)
,253
,001 (***)
St1/At-1
-,063
,303
,176
,508
St-1/At-1
-,023
,751
-,136
,128
-,004
,923
-1,082
,000 (***)
∆
∆
ROE
c.
Russia
* - significant at 10%, ** - significant at 5%, *** - significant at 1%
The models in this final analysis are both statistically significant at a level of 1%,
explaining 40,0% and 51,1% of the variance of the Dependent Variable for Chile and
Russia respectively. As in the previous section, managers of both markets do Earnings
management around CEO turnover through over production in the last period of the
outgoing CEOs.
In the analysis of the Chilean model, all the variables that are statistically
significant are positively related with the dependent variable. The result of this model is
not consistent with the theory presented in the second chapter. Departing CEO would
trick real business activities upward to leave the charge and get the highest possible bonus
as Graham et al,. (2005) stated. The lack of earnings management upward in this scenario
could be explained by a turnover in peaceful terms and the departing CEO’s desire to
meet the expectations of the board and shareholders. Inconsistency with the theory was
also founded in Geertsema, Lont and Lu (2013).
PAGE 52
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Regarding the independent variables for the Russian model, there are only two
variables that are statistically significant and those have different directions (Positive and
Negative). Following the same logic of the second period in the Chilean market, the sum
of the significant coefficient magnitudes will lead the direction of the model. In this case
(Russia), the sum of positive coefficients is lower that the absolute value of the negative
coefficients. The result leads to the conclusion that in the Russian market for the last year
of the outgoing CEO, managers tend to manage earnings upwards. The results are
translated into upward earnings management for this period (significant). The result is
aligned with the theory, where CEOs want to leave the company with good results or
disguising bad previous performance.
PAGE 53
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.5 IMPLICATION OF THE RESULTS
The research results derive in useful implications that can be used by different
stakeholders of the companies.
Earnings management around CEO turnover is a field not very explored and with
lack of evidence. This topic helps stakeholders to be aware about these earnings
management practices in Russian and Chilean companies. Besides, they could understand
the incentives that outgoing / incoming could have to manage earnings within the
company. The existence of earnings management in a company could imply the
inefficient allocation of resources (Overproduction/Underproduction affecting inventory
capacity). Stakeholders that are aware of these practices can take decisions for
diminishing or avoid completely the earnings management by overproduction
(underproduction). For example, increasing transparency of the financial statement
(increase periodicity of financial statements as an internal purpose or incorporate a
specific part of production costs as a key variable), give incentives for keep production
cost indicators as COGS or Inventory, internal audit in the specific periods of transitional
turnover, etc.
More specifically, the existence of managerial opportunism in Russia and in Chile
could damage the image and attractiveness of investors. The earnings management has
several implications of how investors see the company.
The case of Chile earnings management is present in three years around the
change of CEOs. Investors and analysts have to be more meticulous analyzing the
company. Three consecutive years of earnings management could imply a very wrong
perception of the company which could show numbers out of the reality. A bad or bias
investment decision could lead to a loss of money. For this particular case, we
recommend to stakeholders not only pay attention on financial statements, but also in
overproduction/underproduction practices through production costs.
For example, for the first period of the new CEO in Chile, those investors and
analysts that are aware of the overproduction (under) techniques will put attention of the
increasing unit production cost (Downward overproduction). As soon as they see any
increase in amount of the production costs, it could be a signal of earnings management
PAGE 54
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
practice. For the reason described above, stakeholders should not put attention only in
financial statements, in order to have a better picture of the company.
In the case of Russia we also recommend to analyze the same indicators described
in the paragraph above. Even though, we did not find Earnings Management techniques
for the first period of the new CEO, we found these practices in the second period and the
last period of the outgoing CEO. The directions are exactly the opposite as the Chilean
ones, but the way to alert investors is through the same variable (Production unit cost).
Companies that are able to avoid this practices, will experience more stable performance,
less volatility in their results, indicators as inventory not stressed (normal levels) and all
of these factors will lead a company with less risk. Thus, the value of the company
increases.
The findings of existence of overproduction (under) as earnings management
practices contribute to understand how should be the treatment of the production cost
when there is a CEO turnover. Besides, the role of motivation that CEOs have when they
manage earnings within the company, for all different periods. Accounting practices in
Chile and Russia are similar. Both countries use IFRS to develop the financial statements,
however the findings were different. We conclude that incentives for earnings
management are not defined by accounting standards, but by contextual factors as for
example, peaceful CEO turnover or force turnover. Identification of these contextual
factors will lead to a fair treatment of production cost in the companies.
PAGE 55
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
2.6 CURRENT RESEARCH LIMITATIONS
The main objective of this paper was to understand Earning Management in two
markets and compare them for different periods around CEO’s turnover. Nevertheless, the
contribution of Earning Management in these markets is not free of limitations.
First, the “reasons” of CEO’s turnover is a variable unattended in this
investigation. For instance, the scenarios of a routine CEO turnover would be less incline
to intervene earnings upward in the company compared with a manager that was removed
by, for example -poor performance, candidates that succeed their predecessors (Peaceful
turnover), CEOs that remain in the board, death, retirement, etc-. The limitation of our
study is that we did not consider these variables into consideration for a sharper analysis
and results.
Second, in our research was used firm-annual data for identifying earnings
management. We know that it is possible to reduce the misclassification of annual
financial data for departing and incoming CEOs by using quarterly data. In other words,
we could find incoming CEO’s managing real operations downward as soon as the
turnover occurs. The idea of quarterly data is supported by Geertsema, Lont and Lu
(2013) that found that earnings bath happen as soon as the transition quarter. This would
allow to future researchers to make a sharper distinction among types of CEOs (incoming
and departing) increasing the power of the statistical tests.
Third, the research conducted did not study the institutions in charge of
surveillance earnings management in the two samples. The strength and transparency of
the institutions could incentive more or less earnings management through the turnover
transit period. A better understanding of the different institutions would lead a better
comparison of the obtained results between Russia and Chile.
The previous limitations described in this section are not diminishing the
reliability of this work. The results can be used for different stakeholders, being aware
about the incentives of CEOs to manage real business earnings and how these practices
could be prevented in order to avoid consequences. As we found earnings management in
Russia and Chile, our results and analysis are useful for stakeholders of both markets.
PAGE 56
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
CHAPTER SUMMARY
In this chapter the data and methodology of the investigation are presented. It is
presented the selection of companies, the criteria for selection, the period selection and all
the factors involved for the Dataset selection of the investigation in both markets. The
final result gave as a sample of 90 companies for analysis (40 Russian companies and 50
Chilean companies).
On the other hand, different types of methodologies were considered for the study
of Earnings Management around CEO turnover. Nevertheless, only one of the
methodologies was selected, as the best one that fit for the research. The methodology
selected was the Over-Production through the analysis of the dependent variable of
Production Cost. The Over-production methodology was applied for three specific
periods, the last year in the company for the outgoing CEO (t = -1) and the first and
second year of the new CEO in charge (t = 0 & t = +1 respectively).
Besides, in the majority of the literature reviewed stated that exist other factors
that could influence the earnings management and that are correlated with the measure
error in Earnings Management. Given this reason it was decided to include ROE as a
control variable in the research.
The part devoted to Data Analysis and Results was presented. The sample
statistics of both markets Russian and Chilean, in order to compare by size, performance
and leverage. Finally, it is presented the results of the multiple regression models for
Chile and Russia for the three different periods. It was presented the relation of the
independent variables with the dependent variable and the control variable (ROE),
determining the existence of significance of Earnings Management for each model and
rejecting or not the hypothesis suggested in this investigation. Besides, it was stated the
ability of the different models to explain the Earnings management for the periods under
study for Chile and Russia.
The next part of this chapter is presenting the analysis of the results and its
implications of the investigation. The data was presented in comparison tables for better
understanding and analysis. The analysis was made separately by the different stages of
CEO’s turnover.
PAGE 57
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The information showed in the tables allows us to identify the existence and the
directions of the earnings management of Chile and Russia.
Based on the analysis of the three periods for Russia and Chile, it can be
concluded that both countries are acquainted with Earnings Management practices in all
the periods for Chile and for two out of three periods in Russia. The incentives differ of
depending the country and the period under study. The differences of size of the markets
(Russia is about 10x bigger than Chile) could be the reason of higher earnings
management practices in Chile than Russia. The bigger the size of the companies, the
more probably is the supervision of audit entities. (Miller, 1991; Cannella and Lubatkin,
1993; Zajac and Westphal, 1996; Finkelstein and Hambrick, 1996)
The results of Chile showed that departing CEOs manage earnings differently as
the theory stated. It could be explained by a turnover in peaceful terms and the departing
CEO’s desire to meet the expectations of the board and shareholders. Downward earnings
management was found in the first period of the incoming CEO. The incentives in this
case are different, for example, they would like to show that the company was
mishandled and now they have a brand new beginning to start to perform. In the process
of analysis of persistency of earnings management over the transitional period, we found
that earnings management downwards persists intensifying the idea of blame poor
performance of the previous CEO.
In case of Russia, an analysis is more complicated, given that there is no
significant model for the first period of the incoming CEO. Upward earnings
management is significant for the last year of outgoing CEO. The result is aligned with
the theory. Nevertheless it was presented in the previous sections some scenarios of this
possible behavior.
Even though, the research was conducted meticulously and the results are useful
for the different stakeholders, we identified some improvements and limitations that
could be considered for next researches. For instance, use quarterly data instead annual
data for sharper results; consider the reasons of the CEO turnover for a better
understanding of the incentives of earnings management; and a study of the strength of
audit companies that could impact earnings management decisions.
PAGE 58
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
As implication the existence of earnings management in a company could imply
the inefficient allocation of resources (Overproduction/Underproduction affecting
inventory capacity). In the case of underproduction practices, the following periods could
be forecasted high demand, and the speed of supply could not fit the demand of the
customers. The impact of this scenario would lead to a loss of money in cost opportunity,
loss of image of customers that could lead in decreasing number of contracts.
Stakeholders that are aware of these practices can take decisions for diminishing or avoid
completely the earnings management by overproduction (underproduction).
The second implication is about accounting practices in Chile and Russia. Both
countries use IFRS to develop the financial statements. We conclude that incentives for
earnings management are not defined by accounting standards, but by contextual factors
as for example, peaceful CEO turnover or force turnover. Identification of these
contextual factors will lead to a fair treatment of production cost in companies.
PAGE 59
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
CONCLUSIONS
Different authors study the concepts of Earnings Management and CEO Turnover
in different parts of the worlds, but just a few merge this two concepts. As we studied the
markets of Russian and Chile, our goal is compare both markets in the context whether
CEO’s in Chile manage differently the earnings of the firms than in the Russian market.
We presented the following hypothesis: “The fact of change of CEO does not influence
the level of earnings management in a company”.
It is possible identify a literature gap of Earnings Management and CEO Turnover
in Russia and Chile. The research extended the topic to the Russian and Chilean market to
find the importance of Earnings Management in these two locations and the effects and
consequences in Chile and Russian companies. The research is devoted to extend the
literature in the first Latin-American country under study of this topic and Russia (Both
emerging markets).
Different types of methodologies were considered for the study of Earnings
Management around CEO turnover. Nevertheless, only one of the methodologies was
selected, as the best one that fit for the research. The methodology selected was the OverProduction through the analysis of the dependent variable of Unit Production Cost. The
Over-production methodology was applied for three specific periods, the last year in the
company for the outgoing CEO (t = -1) and the first and second year of the new CEO in
charge (t = 0 & t = +1 respectively). The two first periods (t = -1 & t = 0) selected were
mainly devoted to compare practices of outgoing and incoming CEOs. The second period
of the new CEO (t = +1) was intended to show whether there was any persistency in
earnings management after the transitional period of turnover.
Besides, in the majority of the literature reviewed stated that exist other factors
that could influence the earnings management and that are correlated with the measure
error in Earnings Management. Given this reason was decided to include ROE in the
models as a control variable in the research.
Based on the analysis of the three periods for Russia and Chile, it can be
concluded that both countries are acquainted with Earnings Management practices in all
PAGE 60
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
the periods for Chile and for two out of three periods in Russia. The incentives differ of
depending the country and the period under study.
The results of Chile showed that departing CEOs manage earnings differently as
the theory stated. It could be explained by a turnover in peaceful terms and the departing
CEO’s desire to meet the expectations of the board and shareholders. Downward earnings
management was found in the first period of the incoming CEO. The incentives in this
case are different, for example, they would like to show that the company was
mishandled and now they have a brand new beginning to start to perform. In the process
of analysis of persistency of earnings management over the transitional period, we found
that earnings management downwards persists intensifying the idea of blame poor
performance of the previous CEO.
In case of Russia, an analysis is more complicated, given that there is no
significant model for the first period of the incoming CEO. Upward earnings
management is significant for the last year of outgoing CEO. The result is aligned with
the theory. As we do not have information of the next period, it is difficult to link the
reasons why in the second period of new CEO exist downward earnings management. We
as well can conclude that Chile and Russia manage earnings differently, even though both
countries are developing markets, but with very different sizes.
Even though, the research was conducted meticulously and the results are useful
for the different stakeholders, we identified some improvements that could be considered
for next researches. For instance, use quarterly data instead annual data for sharper
results. Other limitation of our work was that we did not consider the reasons of the CEO
turnover. A better understanding of the incentives of earnings management, for example
considering the variables of a peaceful turnover or a force turnover would make our
analysis sharper. Studying how strength are audit companies could impact the
understanding of how strong (Whether exist or not these practices) earnings management
is in the companies.
The investigation suggests to next researchers focus mainly on what we consider
the more important unattended variable -the incentives of earnings management as next
step of the study for both markets (Chile and Russia)-.
PAGE 61
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
The research results derive useful implications that can be used by different
stakeholders of the companies. Stakeholders could understand the incentives that
outgoing / incoming could have to manage earnings within the company. The existence of
earnings management in a company could imply the inefficient allocation of resources
(Overproduction/Underproduction affecting inventory capacity). The decision taken
against earnings management through real activities would increase efficiency in the
operations. For example, in the case of underproduction practices, the following periods
could be forecasted high demand, and the speed of supply could not fit the demand of the
customers. The impact of this scenario would lead to a loss of money in cost opportunity.
The actions that stakeholders could use for diminishing or avoid completely the
earnings management by overproduction (underproduction) are: transparency of specific
part of production costs as a key variable, give incentives to CEOs for keep production
cost indicators as COGS or Inventory in normal levels, internal audit in the specific
periods of transitional turnover, etc.
More specifically, the existence of managerial opportunism in Russia and in Chile
could damage the image and attractiveness of investors. The earnings management has
several implications of how investors asses companies.
Investors and analysts have to be more meticulous analyzing the company. Three
consecutive years of earnings management (Chile) could imply a very wrong perception
of the company which could show unreal financial numbers. Consequently, bad or bias
investment decision could be held impacting directly in the economic situation of the
company. For this particular case, we recommend to stakeholders not only pay attention
to financial statements, but also in overproduction/underproduction practices through
production costs. For example, for the first period of the new CEO in Chile, those
investors and analysts that are aware of the overproduction (under) techniques will put
attention of the increasing unit production cost (Downward overproduction). As soon as
they see any increase in amount of the production costs, it could be a signal of earnings
management practice.
In the case of Russia we also recommend to analyze the same indicators described
in the paragraph above. Even though, we did not find Earnings Management techniques
for the first period of the new CEO, but for the rest, we did.
PAGE 62
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Companies that are able to avoid earnings management practices, will experience
more stable performance, less volatility in their results, indicators as inventory not
stressed (normal levels) and all of these factors will lead a company with less risk. Thus,
the value of the company increases.
Accounting practices in Chile and Russia are similar. Both countries use IFRS to
develop the financial statements, however the findings were different. We conclude that
incentives for earnings management are not defined by accounting standards, but by
contextual factors as for example, peaceful CEO turnover or force turnover. Identification
of these contextual factors will lead to a fair treatment of production cost in the
companies.
PAGE 63
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
FURTHER INVESTIGATIONS
The investigation suggests that next researchers should focus on the incentives of
earnings management as next step of the study for both markets (Chile and Russia).
Define and classify types of turnover by peaceful, forced, routine and non-routine is
important for a sharper analysis. These contextual factors impact the decisions of
managers for earnings management and we believe that will strength the statistical tests.
PAGE 64
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
REFERENCES
Anderson and Reeb (2003). “Founding-Family Ownership and Firm Performance:
Evidence from the S&P 500”. The Journal of Finance Volume 58, Issue 3, pages 1301–
1327, June 2003.
Andersson and Lilja (2013). “Earning Management in Times of CEO Turnover”.
Uppsala University Department of Business Studies.
Cannella, Albert, and Lubatkin (1993).
“Succession as a sociopolitical process:
Internal impediments to outsider selection”. Academy of Management Journal 36(4):
763-793.
Chen,Cheng and Dai (2013). “Family Ownership and CEO Turnover. Contemporary
Accounting Research Vol. 30 No. 3 (Fall 2013) pp. 1166–1190
Choi, Kwak and Choe (2014). “Earnings Management Surrounding CEO Turnover:
Evidence from Korea”. ABACUS, Vol. 50, No. 1, 2014
Coughlan, and Schmidt. (1985). “Executive compensation, managerial turnover, and
firm performance: An empirical investigation”. Journal of Accounting and Economics 7:
43-66.
Dechow, Ge, and Schrand (2010). “Understanding earnings quality: A review of the
proxies, their determinants and their consequences”. Journal of Accounting and
Economics 50 (2–3):344-401.
DeFond and Jiambalvo (1994). "Debt covenant violation and manipulation of accruals."
Journal of Accounting & Economics 17(1/2): 145176.
He, Sommer, Xie (2001). “The Impact of CEO Turnover on Property–Liability Insurer
Performance”. The Journal of Risk and Insurance, 2011, Vol. 78, No. 3, 583-608
PAGE 65
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Finkelstein & Hambrick (1996). “Strategic leadership”. New York: West.
Geertsema, Lont and Lu (2013). “Earnings Management around CEO Turnovers”.
Journal of Economic Literature.
Graham, Harvey and Rajgopal (2005). “The economic implications of corporate
financial Reporting”. Journal of accounting and Economics 40 (1):3-73.
Hazarika et al. (2012). “How Effective is Board Independence to the Monitoring of
Earnings Manipulation?”, International Accounting and Business Conference 2015,
IABC 2015
Healy and Wahlen (1999). “A review of the earnings management literature and its
implications for standard setting”, Accounting Horizons, vol.13, núm.4: 365-383.
Huang et al., (2014). “CEO Turnover and Audit Pricing”. Accounting Horizons
American Accounting Association Vol. 28, No. 2 DOI: 10.2308/acch-50706 2014 pp.
297–312
Huson, Parrino and Starks (2001). “Internal Monitoring Mechanisms and CEO
Turnover: A Long-Term Perspective”. The Journal of Finance
Vol. 56, No. 6 (Dec., 2001), pp. 2265-2297
Jensen (1986). "Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers."
American Economic Review 76(2): 323
Kaplan and Minton (2012). “How Has CEO Turnover Changed?” International Review
of Finance, 12:1, 2012: pp. 57–87
Laux (2007). “Board Independence and CEOTurnover”. Journal of Accounting
Research Vol. 46 No. 1
PAGE 66
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Lindrianasari and Hartono (2011). “The relationship between Accounting Performance
and CEO Turnover: Evidence from Indonesia”. Accounting & Taxation, Vol. 3, No. 2, pp.
97-107, 2011
Miller (1991). ”Stale in the saddle: CEO tenure and the match between organization and
environment”. Management Science 37: 34-52.
Opler et al. (1999). “The determinants and implications of corporate cash holdings”.
Journal of Financial Economics, 52: 3-46.
Parrino (1996). “CEO turnover and outside succession a cross-sectional analysis”.
Journal of Financial Economics Volume 46, Issue 2, November 1997, Pages 165–197
Poveda (2001). “Cuestiones estadísticas sobre modelos y contrastes de ajustes por
devengo anormales”, Working Paper of Instituto Valenciano de Investigaciones
Económicas (IVIE), Serie EC. 2001-16.
Press and Weintrop (1990). "Accounting Based Constraints in Public and Private Debt
Agreements." Journal of Accounting & Economics 12(13): 6595.
Reitenga and Tearney (2003). “Mandatory CEO retirements, discretionary accruals, and
corporate governance mechanisms”. Journal of Accounting, Auditing & Finance 18
(2):255-280.
Roychowdhury (2006). “Earnings management through real activities manipulation.”
Journal of Accounting and Economics 42 (3):335-370.
Schipper (1989). “Commentary on earnings management, Accounting Horizons”. Journal
of Accounting and Economics 3: 91-102.
Sweeney (1994). "Debt covenant violations and managers' accounting responses."
Journal of Accounting & Economics 17(3): 281308.
PAGE 67
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Wartfiled, Wild, Wild (1995). “Managerial ownership, accounting choices, and
informativeness of earnings”, Journal of Accounting and Economics, vol.20, núm.1: 6191.
Zajac and Westphal (1996). “Who shall succeed? How CEO/board preferences and
power affect the choice of new CEOs”. Academy of Management Journal 39: 64-90.
PAGE 68
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
APPENDIXES
APPENDIX 1:
Descriptive Statistics, Russia
Country
Total Assets
(Million USD)
Revenue
(Million
USD)
ROA
ROE
LEVERAGE
Russia
Russia
Russia
Russia
Russia
Mean
15390,90908 9216,73238 0,0521004 0,1115106 0,5166021
Standard Error
6892,429631
3536,7941
Median
Mode
Standard Deviation
Sample Viarance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
1727,206667
#N/A
43591,55249
1900223449
20,20034865
4,375222388
240351,7917
19,70833333
240371,5
615636,3633
40
1747,64
#N/A
22368,6499
500356500
15,0000293
3,87905672
108981,917
10,4833333
108992,4
368669,295
40
0,0546365
#N/A
0,0601908
0,0036229
0,9074177
-0,0309
0,3248973
-0,108005
0,2168922
2,0840173
40
0,1271387
#N/A
0,1287504
0,0165767
0,6437001
-0,345271
0,6320314
-0,202266
0,4297657
4,4604255
40
0,5168246
#N/A
0,1676865
0,0281188
-0,275532
0,0127153
0,6964951
0,1837575
0,8802526
20,664082
40
Total Assets
(Million USD)
Revenue
(Million
USD)
ROA
ROE
LEVERAGE
Chile
Chile
Chile
Chile
Chile
0,009517 0,0203572 0,0265136
APPENDIX 2:
Descriptive Statistics, Chile
Country
Mean
1594,605727
845,21562 0,0593543 0,0746361 0,4576285
Standard Error
488,1272038 280,013863 0,0224327 0,0518785 0,0317149
Median
Mode
Standard Deviation
Sample Viarance
Kurtosis
Skewness
Range
Minimum
Maximum
Sum
Count
343,18
#N/A
3451,580559
11913408,36
10,59284759
3,300106456
16039,13667
0,53
16039,66667
79730,28633
50
166,058333
#N/A
1979,99701
3920388,17
15,8613408
3,83952646
10803,9533
0,04666667
10804
42260,781
50
0,0428048
#N/A
0,1586229
0,0251612
29,91816
4,7326119
1,2786636
-0,24689
1,0317732
2,9677149
50
APPENDIX 3:
PAGE 69
0,0783046
#N/A
0,3668362
0,1345688
20,254124
-2,047731
3,3275075
-1,884359
1,4431482
3,7318069
50
0,4065848
#N/A
0,2242581
0,0502917
-0,641099
0,425254
0,8703142
0,072473
0,9427872
22,881427
50
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Pearson Correlation, Chile
Correlations Chile
Pearson
Correlation
Prod
Prod
1,000
TA
Rev
ROE
ROA
TA
-,003
1,000
Rev
,054
,952
1,000
ROE
,162
,037
,026
1,000
ROA
,086
-,038
-,042
,769
1,000
LEV
,086
,294
,311
-,208
-,221
LEV
1,000
APPENDIX 4:
Pearson Correlation, Russia
Correlations Russia
Prod
Pearson
Correlation
TA
Rev
ROE
ROA
Prod
1,000
TA
-,131
1,000
Rev
-,122
,934
1,000
ROE
-,136
,116
,196
1,000
ROA
,013
,197
,284
,883
1,000
LEV
,341
-,159
-,118
,140
-,049
PAGE 70
LEV
1,000
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
APPENDIX 5:
Regression summary, Chile (t = 0)
Model Summary
Model
R
1
R Square
,813
a
Adjusted R
Std. Error of the
Square
Estimate
,662
,623
a. Predictors: (Constant), ROE, 1/At-1,
∆
,078150986235770
St1/At-1,
∆
S-t/At-1, St/At-
1
ANOVAa
Model
1
Sum of Squares
df
Mean Square
Regression
,526
5
,105
Residual
,269
44
,006
Total
,794
49
F
17,213
Sig.
,000b
a. Dependent Variable: Prod/At-1
∆
b. Predictors: (Constant), ROE, 1/At-1,
St1/At-1,
∆
S-t/At-1, St/At-1
Regression summary, Chile (t = -1)
Model Summary
Model
R
1
R Square
,632
a
Adjusted R
Std. Error of the
Square
Estimate
,400
,332
a. Predictors: (Constant), ROE, 1/At-1,
∆
,086964866949280
St1/At-1,
∆
S-t/At-1,
St/At-1
ANOVAa
Model
1
Sum of Squares
df
Mean Square
Regression
,222
5
,044
Residual
,333
44
,008
Total
,555
49
a. Dependent Variable: Prod/At-1
b. Predictors: (Constant), ROE, 1/At-1,
∆
St1/At-1,
∆
PAGE 71
S-t/At-1, St/At-1
F
5,866
Sig.
,000b
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Regression summary, Chile (t = +1)
Model Summary
Model
R
1
,702a
R Square
Adjusted R
Std. Error of the
Square
Estimate
,493
,435
a. Predictors: (Constant), ROE, 1/At-1,
∆
,101059928071775
St1/At-1,
∆
S-t/At-1,
St/At-1
ANOVAa
Model
1
Sum of Squares
df
Mean Square
Regression
,437
5
,087
Residual
,449
44
,010
Total
,886
49
a. Dependent Variable: Prod/At-1
b. Predictors: (Constant), ROE, 1/At-1,
∆
St1/At-1,
∆
PAGE 72
S-t/At-1, St/At-1
F
8,549
Sig.
,000b
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
APPENDIX 6:
Regression summary, Russia (t = 0)
Model Summary
Model
R
1
R Square
,476a
Adjusted R
Std. Error of the
Square
Estimate
,226
,113
∆
a. Predictors: (Constant), ROE, 1/At-1,
,588560822009331
St1/At-1,
∆
S-t/At-1,
St/At-1
ANOVAa
Model
Sum of Squares
Regression
1
df
Mean Square
3,445
5
,689
Residual
11,778
34
,346
Total
15,223
39
F
1,989
Sig.
,105b
a. Dependent Variable: Prod/At-1
∆
b. Predictors: (Constant), ROE, 1/At-1,
St1/At-1,
∆
S-t/At-1, St/At-1
Regression summary, Russia (t = -1)
Model Summary
Model
R
1
R Square
,715
a
Adjusted R
Std. Error of the
Square
Estimate
,511
,439
a. Predictors: (Constant), ROE, 1/At-1,
∆
,475670668001979
St1/At-1,
∆
S-t/At-1, St/At-1
ANOVAa
Model
1
Sum of Squares
df
Mean Square
Regression
8,035
5
1,607
Residual
7,693
34
,226
15,728
39
Total
a. Dependent Variable: Prod/At-1
b. Predictors: (Constant), ROE, 1/At-1,
∆
St1/At-1,
∆
PAGE 73
S-t/At-1, St/At-1
F
7,103
Sig.
,000b
EARNINGS MANAGEMENT AROUND CEO TURNOVER
COMPARISON OF RUSSIA AND CHILE
Regression summary, Russia (t = +1)
Model Summary
Model
R
1
R Square
,517
a
Adjusted R
Std. Error of the
Square
Estimate
,268
,160
a. Predictors: (Constant), ROE, 1/At-1,
∆
,529325717360735
St1/At-1,
∆
S-t/At-1, St/At-
1
ANOVAa
Model
1
Sum of Squares
df
Mean Square
Regression
3,482
5
,696
Residual
9,526
34
,280
13,008
39
Total
a. Dependent Variable: Prod/At-1
b. Predictors: (Constant), ROE, 1/At-1,
∆
St1/At-1,
∆
PAGE 74
S-t/At-1, St/At-1
F
2,486
Sig.
,050b
Отзывы:
Авторизуйтесь, чтобы оставить отзыв