Размещено на http://www.allbest.ru/

ФЕДЕРАЛЬНОЕ ГОСУДАРСТВЕННОЕ АВТОНОМНОЕ

ОБРАЗОВАТЕЛЬНОЕ УЧРЕЖДЕНИЕ

ВЫСШЕГО ОБРАЗОВАНИЯ

«НАЦИОНАЛЬНЫЙ ИССЛЕДОВАТЕЛЬСКИЙ УНИВЕРСИТЕТ

«ВЫСШАЯ ШКОЛА ЭКОНОМИКИ»

Международный институт экономики и финансов

Курсовая работа/Выпускная квалификационная работа - БАКАЛАВРСКАЯ РАБОТА

по направлению подготовки 38.03.01 «Экономика»

образовательная программа «Программа двух дипломов по экономике НИУ ВШЭ и Лондонского университета»

The analysis of introduction of ICS in Russia

Давыдов Олег Алексеевич

Научный руководитель

И.Ю. Верем

Москва 2019



Аннотация

Система Внутреннего Контроля (СВК) - система, включающая в совокупности мероприятия внутреннего контроля в обществе, которые позволяют обеспечить достоверность финансовой отчетности, эффективность деятельности и соблюдение законов. СВК включает в себя выявление, оценку и управление рисками, что, в свою очередь, должно положительно влиять на финансовые показатели компании. В данной работе исследуется влияние Зрелой СВК на индикатор стоимости компании в виде коэффициента Тобина и индикатор рентабельности активов. При этом, индикатором Зрелой СВК является участие Компании в Налоговом Мониторинге ФНС России. Проанализированы различные показатели 20 компаний, 10 из которых участвуют в Налоговом Мониторинге, за период с 2013 по 2018 год ежеквартально. По результатам оценки, коэффициент переменной СВК оказался значим для обеих зависимых переменных.

financial value control



Abstract

Internal Control System (ICS) is a system that includes actions made to provide reliable financial reporting, effective performance and compliance with laws. ICS includes revealing, evaluation and management of the risks, which in turn should have a positive influence on the financial performance of the company. In this paper, I study the influence of the Matured Internal Control System on the firm's value represented by Tobin's Q and the influence on Return on Assets. Herewith, the indicator of the Matured Internal Control System is the participation of the company in Tax Monitoring by FTS of Russia. I have analyzed different financial indicators of 20 companies, 10 of which participate in Tax Monitoring, the period used is 2013-2018 quarterly. The results show that the ICS coefficient is significant for both dependent variables.



1. Introduction

Today, the principles of corporate governance become more and more sophisticated and difficult to execute due to the conditions of rising control inside and outside the entities, so the “price” of errors and penalties for them grow significantly. The corporate governance itself is created to deal with organizational management, ethics, qualitative financial reporting and lack of fraud; and from the beginning of any commercial relationships, entities try to establish such principles and rules that will struggle possible problems inside the company, minimize internal and external risks for the company.

By the initiative of the private business, in 1985 The Committee of Sponsoring Organizations of the Treadway Commission (or COSO) was created whoseinitial purpose was to create and develop recommendations for the issues of fraud in the financial reporting. In 1992 COSO introduced the conceptual framework of the internal control (COSO 4), which had defined the internal control and its model (with which other companies may compare their own systems of internal control). With time, the majority of companies have started following COSO ICS principles; however, the question is how much it is effective for the company's performance, are key financial indicators being affected, including equity.

Speaking about introducing COSO in Russia, there is no concrete data when the first company started following principles of internal control, however, according to CFO Magazine survey in 2006 , 82% of Russian companies follow or try to follow COSO principles. Until 2013, the issue of creating an internal control service was the right of organizations (except for credit organizations for which this is an obligation).From January 1, 2013, according to Art. 19 of the Federal Law dated December 6, 2011 No. 402-FZ “About Accounting” all economic entities must exercise internal control over the committed facts of economic life, and those that are subject to mandatory audit - also exercise internal control of accounting and preparation of financial statements (with the exception of cases when the manager took over the responsibility of accounting for himself). The decision on the introduction of accounting by him can be taken by the head of a small and medium-sized business entity, respectively, in such an organization it is necessary to carry out only internal control of the committed facts of economic life. That is, from 2013 any public entity of significant size must have the COSO ICS established, however, the key question is how effective ICS is.

In this paper I will study the effects of COSO ICS on the financial performance of the company, measured by Tobin's Q indicating for the firm's value and ROA, assuming the ICS in the company is developed enough. The indicator of well-developed ICS will be the participation of the entity in the Tax Monitoring Regime introduced by the Federal Tax Service of Russia in 2011. This regime assumes that by the specific ways of measure from 1 to 5, the Internal Control System will have a “mark” not less than “3” which indicates for a satisfactory status of the Internal Control System. Those companies will be compared with analogical companies within the industry that do not participate in tax monitoring. Before applying the data to the models, I will discuss the nature of Internal Control System to make the reader of the paper understand, why it is important, and then analyze the ways and methods of how the influence of ICS on financial results was studied by authors before, which models did these authors approach and why these models may be useful for my paper.



2. Theoretical Framework

2.1 The nature of Internal Control System

InternalControlSystemrepresentsaltogether entity's policies, procedures and organizational actions, which are provided by directors, managers and other staff of the entity and which goal is to ensure reasonable confidence in both achieving goals and aims of the entity and minimization of risks, reliable financial reporting and compliance with law. Thesubjectsofinternalcontrolareallstructural divisions of an entity, which provide control procedures according to their power, including divisions of internal control.

Accordingto COSO ICS (COSO 2017 Internal Control - Integrated Framework) and rules of Federal Tax Service of Russia (Order dated 16.06.2017 №MMV-7-15/509@ “About the Requirements relating to the organization of the Internal Control System” for participation in Tax Monitoring), Internal Control System of the entity should consist of five mutually connected components, which are functioning as the whole integrated system:

· Control environment - a set of standards, processes, and structuresdetermining the understanding and requirements for internal control andwhich are the basis for internal control throughout the entity;

· Risk valuation - a process of revealing and analyzing the risks, which obstruct achievingthe goals by the entity; prioritization of risks and development of such actions and controls which will assist in risk management;

· Control procedures - actions which should provide control and risk minimization;

· Information systems- a set of IT-systems, information and actions which will be the basis for automatized or semi-automatized controls and risk management;

· Monitoring - systematic and periodic actions for the purpose of valuation of Internal Control System efficiency

A control environment component requires such things as ethical development of the entity, principles of decency and corporate governance. Inagoodcontrolenvironment, theboardofdirectors will be independent of management and provide their own efforts for maintaining the Internal Control System. Management will function under the principle of segregation of duties, determining the corporate structure of the entity. Besides, a controlenvironmentshouldprovideregulations of hiring staff, statutory principles of the company, rules of behavior and responsibility.

In the process of risk valuation, the entity should have clear goals for identifying and valuation of such risks, which obstruct these goals, i.e. for every corporate task and its elements the entity should identify risks that are connected to these tasks. One of the principles under risk valuation is that the entity should count the probability of fraud while evaluating the risks. Everyriskfortheentity, includingrisks for Internal Control System, should be analyzed and managed.

The entity should choose and develop such control procedures, which will minimize the risks obstructing the achievement of an entity's goals. Control procedures may be automatized, semi-automatized and manual. These procedures not only struggle with risks but also develop the discipline inside the entity and its divisions. Control procedures are realized by regulating the policies and actions described in these policies.

Information systems represent a mix of systems used for risk management, accounting, internal control and audit, monitoring and development of ICS. As long as automatized processes are, generally, more efficient, the entity should possess information technologies in a maximum amount of elements of the business.

Monitoring and development should be a systematic process of seeking the deficiencies in the Internal Control System and managing them. This process includes the regular evaluation of the Internal Control System and reporting to the management and directors about limitations found. Evaluation should consist of self-evaluation (by the owner of control procedure) and independent valuation (by the department of internal audit and external experts).

Having all these in sum, the Internal Control System must provide the following actions:

· Analyze the current financial situation;

· Ensure in the normal work of all the divisions within the company by the internal audit;

· Provide information about any violations of internal and external laws and policies;

· Provide information about the risks;

· Document the results of the audit and provide the opinion;

· Provide and develop recommendations about violations and limitations;

· Overwatch and control the execution of these recommendations.

There are also principles, under which the Internal Control System is organized and operated:

· Integrity - internal control in the entity is systematic and integrated into strategical and operative management on all the levels of entity, divisions, and business processes;

· Common methodology - internal control is realized on the base of common approaches and standards of the methodology of an entity;

· Systematic functioning - internal control operates systematically and constantly;

· Complexity - internal control covers all the branches of operating and business processes;

· Risk-orientation - Internal Control System cooperates with processes of risk management, by which actions of risk management are integrated well-timed and effectively;

· Responsibility - all the subjects of the Internal Control System are responsible for the operating of internal control, rights and obligations of subjects are defined in local acts;

· Segregation of duties - no responsibilities for authorization and applying operations in accounting should be secured only under one person;

· Documentation of operations - all the controls and its results are defined in local acts by owners and executors of control procedures;

· Reasonable sufficiency - the volume and economic efficiency of internal control procedures should be sufficient for achieving the goals of the Internal Control System of an entity;

· Timeliness of information - information about the execution of controls and found limitations should be shared with management, directors, and department of internal control/audit in time;

· Provision of functioning - control functions of internal control subjects are provided with authorities and technical base;

· Constant development and enhancement - procedures of internal control of the entity are being permanently developed in the purpose of effective management of internal control processes;

· Interaction and coordination - internal control is carried out on the basis of clear and effective interaction between all the subjects of internal control;

· Independent valuation - subjects of internal control are independent of others to provide an objective and significant valuation of internal control efficiency;

· Countermeasures to corruption and fraud - internal control assists in the development and applying steps orientated on prevention and countermeasures to corruption.

What are the objects of control under the Internal Control System? These are the assets of an entity and obligations to its employees and third parties which are reflected in the financial reporting of an entity, in other words, business processes and results that compose financial and operating activities.

2.2 Tax Monitoring and Internal Control System

The Regime of Tax Monitoring introduced by the Federal Tax Service of Russia in 2011 assumes the presence of the Internal Control System within the entity. In fact, practically all the requirements for the Internal Control System are copied from those in COSO ICS, except for some moments about tax control. To enter Tax Monitoring Regime, the company should prepare documents, including self-evaluation of the Internal Control System and regulating documentation. After submission of these documents, the FTS examines the application and the current state of ICS.

For the components of the Internal Control System, there is a quite similar (as in the COSO ICS) and, at the same time, a complicated system of marking, or Maturity levels. The Internal Control System of an organization must have maturity levels, such that each is characterized by specific goals, objectives, strategy, organizational structure, criteria, et al.The maturity level of the Internal Control System of an organization must be assessed by the organization by awarding points for each criterion and assigning a specific maturity level to the Internal Control System of the organization (according to Order dated 16.06.2017 №MMV-7-15/509):

· Initial level (1) - at this level processes have not been defined and the result depends on individual efforts made by participants in the Internal Control System of the organization. There are no unified standards and principles governing the Internal Control System of the organization.

· Defined level (2) - the basic principles and processes of the Internal Control System of the organization have been defined at the organization. The main processes, used for core business processes, have been formally documented.

· Monitored level (3) - systems are in place at the organization which are used by employees, and the management is focused on developing the organization's Internal Control System. The organization defines and formalizes all processes associated with the functioning of the organization's Internal Control System.

· Managed level (4) - the organization's Internal Control System processes and standards have been integrated with other business processes and information systems of the organization. Decisions are made by the organization on the basis of a system of analytical indicators/

· Optimizing level (5) - characterized by continual improvement of the organization's Internal Control System processes. Data required for risk identification are automatically collected, analyzed and evaluated for the purpose of optimizing the organization's risk management system.

The Maturity level, at first, is being counted for every subcomponent of Internal Control System (as it is given at the Order dated 16.06.2017 №MMV-7-15/509), and then for the whole component. More precisely, a control environment consists of four subcomponents - principles of ethics, the participation of the director, organizational structure, competence, and staff development. Risk valuation includes subcomponents of revealing, valuation and documentation of the risks. The component “Control procedures” consists of detailing the controls, documentation of controls, valuation of efficiency of controls, presence of automatized control procedures and proportion of automatized and manual control within the entity. Speaking about information systems, the maturities of the following components should be valued - internal/external audit of the information systems, the existence of protection against unsanctioned access to the data, information system for accounting and tax purposes, information system used for the control for accuracy and correctness of tax calculations and “online”-controls. Finally, the component of monitoring consists of monitoring itself, actions made for ICS development and external/internal audit. Maturity of every subcomponent depends on how developed and well organized it is. After the valuation of subcomponents maturity, the average maturity for the component is calculated and then the average of component scores is applied to the whole entity. The firm is able to join tax monitoring regime if the Maturity Level of the entity is equal or higher than 3 out of 5.

2.3 LiteratureReview

Overviewing the nature of Internal Control System, we may conclude that all the established principles and regulations should provide more qualified and fair financial reporting and improve the financial performance of the company due to the control environment. Since we are talking about matured ICS, which follows Tax Monitoring rules, the question is how do the following COSO 4 principles improve the aspects of the firm's valuation: as long as the ICS is a part of Enterprise Risk Management System, the risks within the company should be monitored, analyzed and managed under control, which must improve the capital of the company and thus its portfolio.

One of the most convincing and conclusive studies of this effect was performed by Mark Farrell and Ronan Gallagher in 2015 (Mark Farrell and Ronan Gallagher. The Journal of Risk and Insurance. 82, No. 3, 625-657 (2015), that is why this article will be some kind of a “basis” for my own research. Authors assumed that entities with a matureERM level may have a higher value of the firm, which is measured by Tobin's Q coefficient and proved it via econometric model. That is, we may perform the alike approach to find out how much is developed ICS(recall that COSO ICS is the part of ERM) influence financial performance of the company, assuming the fact that the entity joined Tax Monitoring regime.

Authors argue the elder thoughts that “risk management activities will not create value until any irrelevance of an organization's capital structure exists” (Farrell and Gallagher (2015), pp.628-629) by the fact that markets are imperfect with limitations in terms of agency costs (with reference to Jensen and Meckling (1976), bankruptcy costs and taxes. Since imperfections and frictions exist, the risk management approach to reduce the idiosyncratic variance within the company may be useful enough.

The connection between ERM and Firm Value is described by the authors in the following way - as long as ERM assumes risk aggregation, monitoring and examination, risk management system must improve decision making and thus capital allocation. Besides, authors share the view that ERM “goes beyond focusing on just risk avoidance activities in recognition of the value to be gained from exposure to risks for which the firm has a strategic competitive advantage. This is partly in recognition of the fact that the desire for risk avoidance may actually increase the volatility and fragility of financial markets as a whole via certain investment products” (Farrell and Gallagher (2015), pp. 629). What is more, they argue that the existence of ERM itself and its effects may improve the rating of the organization(for example, credit rating agencies scores). If we try to connect this argument with Russian companies, the reflected certain of the existence of strong ICS within the company (which may be noted in the auditors' opinion) may both indicate the stronger credit rating and the fact ofassured following the local legislation (No. 402-FZ “About Accounting”).

The mathematical connection or the model itself was based on the dependent variable of log Tobin's Q, which is the log of total equity of entity plus debt and divided by total asset. This variable was used as practical and usable on alike studies of Smithson and Simkins (2005), Hoyt and Liebenberg (2011) and others. The key dependent variable was the dummy variable of matured ERM (Farrell and Gallagher (2015), pp.642) (d=1 if ERM is 3 points from 5 or higher and d=0 otherwise). The levels of maturity of companies were taken via analyzing the auditors' conclusions and Standard &Poor's ratings. The other variables were controls, like size, leverage, sales growth and so on to avoid omitted variables problem. Authors achieved significant results that dummy of mature ERM affects the Tobins'Q (thus the firm's value). Besides, they regressed the components of ERM and found out that all the components except Risk Appetite Management and Business Resilience and Sustainability are significant with the ERM.

A quite similar survey was done by Robert E. Hoyt and Andre P. Liebenberg (The Journal of Risk and Insurance, Vol. 78, No. 4 (December 2011), pp. 795-822) with data of insurance companies. They think that “if the entity establishes ERM, managers, and staffwill better understand the aggregate risk inherent in different business activities, which should cause a better allocation of resources, capital and, thus, financial performance” (Hoyt and Lindenberg (2011), pp. 798).What is more, ERM is a structure that combines into one framework all regulating activities and provides rules for revealing, valuation and monitoring.

The authors also claimed, that in fact, collecting data about whether the company participates in ERM is quite difficult, since entities are not required to report whether they are engaged in ERM and the data collection meant “a detailed search of financial reports, newswires, and other media for evidence of ERM” (Hoyt and Liebenberg (2011). The Journal of Risk and Insurance, Vol. 78, No. 4 (December 2011), pp. 799)

Speaking about the model, they used Tobin's Q too; but authors emphasized that the disadvantage of using Q is that “it ignores potential selectivity bias that arises due to the likely endogeneity of ERM choice. In other words, some of the factors that are correlated with the firm's choice to adopt ERM may also be correlated with observed differences in Q”(Hoyt and Lindenberg (2011), pp.801). To avoid endogeneity bias, authors performed a two-system equation that looked like (Hoyt and Lindenberg (2011), pp.802):

Where ERM is dummy (equal 1 if it is established by the company and 0 otherwise), X are controls, is a vector of ERM components. The authors achieved significant results and initial on the value relevance of ERM for insurance companies.

Even the system was used to struggle endogeneity, the problem of such research in comparison with my idea is that no maturity of ERM is counted, which is obligatory for the idea of tax monitoring. Besides, the data came only from insurance companies, no other industries were researched, which may bias the effect of ERM/ICS on a particular company. That is, there are practically no financial entities (like banks, insurance companies, funds, etc.) of high echelon without a “repeatable” (3 points from 5 by ERM regulation) ERM or something close to it.

What is interesting in Hoyt and Liebenberg (2011) is that this research acts as a basis for other ones in particular countries. For example, there is a paper by Chen, Chuang, Huang and Shih (The value of implementing enterprise risk management: Evidence from Taiwan's financial industry. The North American Journal of Economics and Finance, Feb. 2019), where methods of sampling and model of Hoyt and Liebenberg (2011) are involved, with a hypothesis that “ERM program adoption is positively associated with firm value”. There is nothing new in this research in comparison with Hoyt and Liebenberg (2011), except for the certain that the sample is much less than in original study and not only insurance companies were used. There were 615 observations in total and still 1% significant results for both Q and ERM were achieved. That is, the study of ERM may be still robust and significant even for relatively small samples.

A more simple and in the same time wide research (in terms of variables) was made by Yilin Wang and Lei Ruan(Filomat, Vol. 30, No. 15, Quantitative Economics and Its Development (2016), pp. 4223-4234 ), the authors tried to find out “whether internal control and cash dividend will have an interactive synergetic effect on enterprise performance or not is a study of both theoretical value and practical significance” (Wang and Ruan (2016), pp. 4223) Wang and Ruan accentuate that internal managerial factors play more and more role in influence on financial performance, including internal control and corporate jurisdiction. The authors paid more attention to COSO regulations (2013). They mix the idea of testing internal control with testing cash dividends, as, in their opinion, dividend policy is necessary for corporate governance - dividends meet the demand of investors and thus enterprise performance may be enhanced.

The research assumes the hypothesizes that internal control positively correlates with both short-term and long-term enterprise performance, and cash dividends may both positively and negatively correlate with types of performance. Given that, “Internal control and cash dividend have an interactive synergetic effect on short-term and long-term enterprise performance”(Wang and Ruan (2016), pp.4225-4226). This paper is strong enough due to the big amount of observations - 11,507 from Chinese A-share motherboard.

According to the hypothesizes, authors regress return on assets (ROA) on different explanatory variables (ICS, Controls and, optionally DPS) and then regress Tobin's Q on the same variables. While regression with ROA measures short-term performance, Tobin's Q represents long-term performance. Controls are simple like size, leverage, etc. The results (Wang and Ruan (2016), pp.4228-4233) showed that:

· Internal control and business performance over both the short and long term has a significant positive correlation;

· There is a significant positive correlation between the company's cash dividend payout and the level of short-term and long-term performance;

· There is a strong correlation between internal controls and cash dividends.

These results mean that cash dividend and internal control are bothnecessary to improve enterprise performance and therefore necessary factors to be considered in the overall success of a company.

Recall that ICS as a part of ERM does not exist only for improving shareholders' wealth and financial results, but it should reduce fraud, human errors and thus material misstatements in the financial reporting. Misstatements are regulated by different corporate acts, one of which is the Sarbanes-Oxley Act (SOX 404), which links misstatements with underlying control weaknesses.

The research on whether the possessing of Internal Control System is indeed linked with reported misstatements and weaknesses was made by Sarah C. Rice and David P. Weber (Journal of Accounting Research, Vol. 50, No. 3 (June 2012), pp. 811-843). They had 488 total observations, cleared from missing information and from observations, where material misstatements were not connected with a lack of effective internal control.

The model is interesting enough, since the dependent variable is whether misstatement is reported under the terms of inefficient ICS, while dependent variables are basic financial controls, and different explanatory variables as dummy of change in auditor, audit fees, having the auditor from the big 4 of accounting companies (EY, KPMG, PWC and Deloitte) and years passed from SOX 404 (this method of counting years is useful for my own research since different entities are following Tax Monitoring now for a different period of time).

In the results (Rice and Weber (2012),pp. 825-831), authors achieved obvious results for their controls (for instance, size of the company is negatively correlated with a chance to report material misstatement), but what is more important, there are significant results for explanatory variables - indeed, the companies which auditors are BIG4, there likelihood of reporting material weakness decreases; coefficient of time passed from starting following SOX 404 law is significant too. This method of study will be useful for my own research and represents the “audit” side of ICS question - whether the controls system is able to influence relevance, faithful representation, and the existence of material misstatements in financial reporting.

Another article to discuss is “Voluntary reporting on Internal Control Systems and Governance Characteristics: An Analysis of Large U.S. Companies” by Stephen Owusu-Ansah and Gouranga Ganguli (Journal of Managerial Issues, Vol. 22, No. 3 (Fall 2010), pp. 383-408) which tries to solve an issue of whether any reports on internal controls should be made (as no mandatory requirements for reporting do exist). Except the hypothesizes about independent boards and committees, in which I am not interested in my paper, authors issue the hypothesis that “company having an audit committee that meets frequently is more likely to voluntarily report on its internal control”. Since the ISA standards require from auditors in the board to be independent, qualified in accounting and finance, the critiques inside the board and decisions may influence the voluntary reporting on ICS.

Authors estimated Probit model with the dependent variable of voluntary reporting on Internal Control System in the prior year and different financial and audit-type regressors (the sample was taken from the 2002 Fortune 500 list). There are also test variables of auditor (company from previous BIG 5), frequency of meetings and so on. According to the authors, “this study finds that there are positive, significant relationships between management voluntary reporting on Internal Control Systems and frequent audit committee meetings”.

All in all, I think that these articles are the best in the methodology used to describe possible effects of ICS on both financial performance and lack of material misstatements. The authors have different approaches, but there are some ways I may follow in my own research, like dummy of joining the Tax Monitoring regime and Tobin's Q as a dependent variable.



3. Empirics

3.1 Sample design

The initial idea was to take 24 firms which participate in the tax monitoring regime from different industries and fill the sample to 48 firms by adding similar firms, respectively to the industries; then use quartal financial and corporate data within the model. The period is from the 1-st of January 2013 to the 31-st of December 2018, since the tax monitoring regime started to work in 2013. According to the information of the FTS of Russiaprovided to the RBC (Economics, 19 Dec 2018) , there are 44 companies within the regime, however, there is no public information about all the 44 companies, besides, some companies joined the regime in 2019, so counting them will not be effective and indicative. That is why I found information about only 24 firms from different industries (RBC, Economics, 19 Dec 2018). These firms are represented in Table 1.

Table 1

Industry	Companies
Electric, Gas, Metallurgy	Lukoil, Gazpromneft, Rosneft, Novatek, SMP-Neftegaz, Severstal, NorNickel, Rosenegoatom, InterRAO
Telecommunications, Media, Technologies	NTV, Google, Megafon, MTS, Tele2
Wholesale	Unilever, Russol, Komatsu CIS, Jaguar Land Rover, Rostekh
Finance	National Clearing Centre, HSBC, VTB
Transport	Aeroflot, RZD

Some problems appeared during the analysis of financial reporting of these companies, organizational structure and legal form of entity. Starting from SMP-Nefregaz, according to its capital structure provided by the company, there are no public shareholders. All the shares of JSC Rosenergoatom belong to State Corporation Rosatom and Atomenergoprom, besides, there is no any financial data in Capital IQ or other services. The history repeats with RZDand Rostekh, since it is the corporation which fully belongs to the State. Besides, all the branches of foreign companies in Russia are represented as limited liability companies, which exclude the opportunity of using indicators referring to exchange (Google, Unilever, Komatsu, Jaguar Land Rover, HSBC).NTV and Russol are limited liability companies too. National Clearing Centre and VTB are banks, which imply strong internal control. Finally, there is a problem with Aeroflot, the information about the company is sufficient, however, there are no big companies to make a parallel, some companies lack a significant amount of financial data, some companies are state-ruled, some companies have its capital belonged to a limited amount of individuals or entities.

That is, ten firms are left, and I need to find ten similar firms respectively to every firm left in my Tax Monitoring sample. Using Capital IQ, Bloomberg and open sources, I found the firms, which are publicly traded and have the required quartal data for the period 2013-2018. These firms, and, respectively, total sample are represented in Table 2. Besides, I have indicated at what stock exchange these firms are traded, despite the fact, that it may differ, and some firms are registered not in Russia, the main operating activity happens in Russia. What is more, the question appears - how do I identify, when the firm did join the tax monitoring regime? I have analyzed news from the different economic and financial journal, websites and other sources, besides, I have read news of Federal Tax Service of Russia, where different agendas for participants or commentaries were described. In this way, I know the years of starting to follow the Tax Monitoring regime.

Table 2

Tax Monitoring	Analogical Firms
PJSC Lukoil (MISX) 2016	PJSC Tatneft(MISX)
PJSC Gazpromneft (MISX) 2016	PJSC Surgutneftegas (MISX)
PJSC Rosneft Oil Company (MISX) 2016	PJSC Magnitogorsk Iron Steel Works (MISX)
PAO Novatek (MISX) 2016	PJSC RusHydro (MISX)
PAO Sevestal (MISX) 2018	PJSC Unipro (MISX)
PJSC Norilsk Nickel (MISX) 2016	UC Rusal (MISX)
PJSC InterRAO UES (MISX) 2018	PJSC Rosseti (MISX)
PJCS Mobile TeleSystems (MISX) 2016	PJSC Tattelecom (MISX)
PJSC Megafon (MISX) 2016	PJSC Rostelecom(MISX)
Tele2 AB (SS) 2018	MailRu Group Ltd (LSE)*

* No applicable company from telecommunication sector was found due to the lack of data or the legal form of the entity.

Allthe financial data is taken from Capital IQ and Bloomberg, except that which lacks betas or some quartal financials - these indicators were found either on companies or investing websites. Besides, some lacking betas were calculated for the annual periods by the CAPM formula (using returns). The currency is rubles. All the series are stationary (see 7.1. of Appendix)

Besides, there is a need to notice, that I assume that companies that do not participate in the tax monitoring regime have no mark for maturity higher than 3. My assumption is based on the best practices of EY and PWC, whose services include assistance in preparation for tax monitoring regime -in average, the projects for preparation last for year, as it also mentioned in rules for this preparation (Order dated 21.04.2017 №MMV-7-15/323@ “About the Requirements relating to the organization of the Internal Control System”), during which the entity should prepare the documents and improve its Internal Control System. That is, I assume that Internal Control Systems of the analogical companies exists but does not have a strong maturity score.

3.2 Models

In my study, I base my two hypothesizes on works of Farrell and Gallagher (2014) and Wang and Ruan (2016). Whereas Farrell and Gallagher (2014) studied the influence of matured ERM on the firm's value (measured by Tobin's Q), Wang and Ruan (2016) studied how the simple presence of Internal Control System affects both Tobin's Q indicator and return on assets. Therefore, I mainly based my hypothesizes on these two studies, putting aside the hypothesis of cash dividends effect by Wang and Ruan (2016). In fine, I have the following ones:

H1a: Matured Internal Control System, under the assumption of entering Tax Monitoring regime, has a positive influence onthe firm's value, measured by Tobin's Q (long-term performance);

H1b: Matured Internal Control System, under the assumption of entering Tax Monitoring regime, has a negative influenceonthe firm's value, measured by Tobin's Q;

H2a: Matured Internal Control System, under the assumption of entering Tax Monitoring regime, has a positive influence on the firm's return on assets (short-term performance);

H2b:Matured Internal Control System, under the assumption of entering Tax Monitoring regime, has a negative influence on the firm's return on assets.

In my two models, I will have a dependent variable for each and eight explanatory variables, eight of which are controls. Here is the description:

1) Q_it(for model 1) - natural logarithm of Tobin's Q, which is calculated by the method of Chung and Pruitt (1994) (Kee H. Chung and Stephen W. Pruitt. Financial Management.Vol. 23, No. 3, Venture Capital Special Issue (Autumn, 1994), pp. 70-74):

where

MVE = Market value of equity;

PS = Total value of preferred stock capital;

D = Total debt, which is Current Liabilities less Current Assets plus Long-Term Debt; note that it may appear that debt is less than zero, then it assumed to be zero for Tobin's Q.

TA = Total assets.

Tobin's Q is a way of representing the firm's value, which is also, as I have mentioned, was widely used in studies related to the ERM and ICS. Besides, the relation to the total assets indicates how strongly do other factors influence the company's performance. According to Wang and Ruan (2016), Tobin coefficient more likely indicates long-term performance.

2) ROA_it(for model 2) -return on assets. ROA is the indicator of how the company generates profits relative to its total assets and allows managers, financial analysts and investors to understand whether assets are being managed effectively to generate the profits. According to Wang and Ruan (2016), ROA more likely indicates short-term performance.

3) ICSDummy_it- dummy variable, indicating whether the company has a matured Internal Control System, under the assumption of joining tax monitoring regime. That is:

ICSDummy = 1, if ICS maturity has a monitored, managed or optimizing level (i.e. levels 3,4 or 5);

ICSDummy = 0, if ICS maturity has initial or defined level (i.e. levels 1 or 2).

4) Size_it- control variable, which represents the natural logarithm of the book value of total assets. Total consists of current assets like cash, inventories, and trade receivables, and non-current assets like land, plants, machinery and goodwill. Total assets are being reviewed by the investors before making a decision to invest - whether the entity has enough existing value.

5) Leverage_it- control variable, which represents the value of total liabilities divided by the market value of equity. That is, we are speaking about financial leverage or debt-to-equity ratio. It helps the company's corporate participants (shareholders, managers, lenders) in understanding the risk of capital structure.

6) SalesGrowth_it- control variablewhich characterizesthe growth of sales of the company. Sales growth may influence net profit for the year, which in turn may improve the value of the company. I have calculated sales growth as , where t - quartal period.

7) ROE_it- return on equity for a given period of time, which is .

8) Beta_it- control variable representing beta of the particular company in a particular period. Beta measures the risk of the asset relative to the market. Practically all the betas were taken from Bloomberg, however, some betas were lacked. That is why I have calculated lacking betas for the whole period by the CAPM formula:

where

E(r_i) - expected required return on asset i;

R_f - risk-free rate of return;

R_m - average market return.

9) IndDiv_it- dummy of industry diversification factor. Industry diversification may indicate both risk-avoidance and good investment decision making, that is, a firm's value may be affected. It was originally created by Thompson Reuters and represented whether company's products belong to more than one industry. Since I have no access to this Thompson Reuters data, I have analyzed the product structure provided by Capital IQ to understand whether the company is industrially diversified. That is:

IndDiv = 1 if the company is industrially diversified;

IndDiv = 0 otherwise.

10) IntDiv_it- dummy of international diversification. As 7), it originally belongs to Thompson Reuters and it also may emphasize risk-avoidance and decision making. I estimatedthis factor by myself - if the company sells its goods on both domestic and international markets, I assume that the company is internationally diversified. Some companies have a non-stable value of it due to either sanctions introduced (i.e. UC Rusal in 2018) or entering the international market during the revisedperiod of time (Rostelecom introduced first international products at the end of 2014). The data is achieved due to the analysis of the information about the company's participation in the international market and news about the introduction of international products. That is:

IntDiv = 1 if the company is internationally diversified;

IndDiv = 0 otherwise.

That is, I have two panel models in my studies,where I regress variables (3-9) on the log of Tobin's Q (1) - log-linear model; and return on assets (2). Mathematically, the models are following (see next page)

Model 1:

Model 2:



4. Results

4.1 Descriptive statistics

All the results are my own estimations made in the EViews. Table 3 represents descriptive statistics for both dependent variables and all the explanatory variables except dummies (as they are equal to 0 or 1). The average indicator of a long-term performance (log of Tobin's Q) is -0.24472, which means that Tobin's Q itself is, on average, less than 1 - that is, Russian companies are likely to be underestimated; minimum and maximum are 1.1431 and -2.2113 respectively. The average indicator of a short-term performance (ROA) is 0.0194, with a minimum of -0.1763 and maximum of 0.3431. The negative value is caused by the presence of net losses of some companies in the sample. The same explanation is actual for the negative sign of a minimum of ROE. Another notable fact is that there exist two quite big ranges between maximum and minimum of both leverage and sales growth, which are respectively 13.5338 and 6.2345.

Table 3

Source: My own estimations



4.2 Model 1 results

Table 4 represents the estimation results for Model 1. At first, I made the Hausman test (see Table 14 in Appendix) which rejected the null hypothesis that the random effect model is applicable, that is, the fixed effect model should be used. However, there was a problem of near singular matrix appeared. That is, I estimated my model as Pooled Panel, via list squares estimations. The P-value of F-statistics is zero, which indicates that the model itself is significant. As shown in Table 4, the coefficient between the dummy of Matured Internal Control System (ICSDUMMY) and log of Tobin's Q (TOBIN_S_Q) is 0.2416 and significant at 0.01 level, which means the verification of H1a. All the controls except growth of sales (SALES_GROWTH) and indicator of international diversification (INTDIVERS) are significant at 0.01 too, with INTDIVERS being significant at 0.1 level. The explanation for the insignificant growth of sales is that it behaves quite randomly for the whole sample during the given period.

The leverage (LEVERAGE) has a negative coefficient of -0.1553, which seems logical. It is notable, that size of the company behaves as negative factors which is consistent with Farrell and Gallagher (2015) results, the same may be said about ROE variable. The situation with ROE may be explained by the presence of negative values. Beta has a coefficient of 0.6516, this happens as this version of Tobin's Q includes the long-term in the nominator, so there is a positive correlation between the presence of debt in the company and the riskiness of it. Indicator of industrial diversification (INDDIVERS) is negatively correlated with the firm's value while international diversification is positively correlated. The significant problem with this estimation is the DW-statistics of 0.881575; if DW is less than 2, the positive autocorrelation appears (the error terms tend not to be independent).



Table 4

Source: My own estimations

However, I was disturbed about the error of singular matrix appeared and found out that it happened due to collinearity between ICSDUMMMY, INDDIVERS, and INTDIVERS. That is, I ran out the fixed effect model, which is applicable by the Hausman Test, without last two dummies. The results are represented in Table 5 (on the next page).



Table 5

Source: My own estimations

As you can see, the signs for coefficients remained the same, as did their significance status, expect ROE being significant only at 0.05 level. What is more, due to the fixed effect, the intercept became significant due to its dummy status. Though, the effects specification performs better results for the adjusted fixed-effect model, providing higher R-squared value, better Durbin-Watson and a lower AIC (which exists for comparison of the several models). Overall, the adjusted Model 1 without INTDIVERS and INDDIVERS provides better results due to the possibility of applying fixed-effect model estimation.

4.3 Model 2 results

Speaking about Model 2, the same situation has repeated as with the Model 1: according to Hausman test (see Table 15 in Appendix), the fixed-effect estimation was required,and it was impossible due to singular matrix problem. That is, I got rid of indicators of industrial and international diversification and estimated the fixed-effect model. The results are represented in Table 6.

Table 6.

Source: My own estimations



As it is shown in Table 6, the model itself is significant according to the zero p-value of F-statistics. All variables except leverage (LEVERAGE) and growth of sales (SALES_GROWH) are significant at 0.05 level. The coefficient between ICSDUMMY and ROA is 0.0105 which verifies H2a. DW-statistics shows the value of 1.5386 which indicates for small positive autocorrelation. Speaking about significant variables, the coefficient of SIZE is negative again. However, ROE now performs a positive influenceonthe dependent variable, due to the presence of net income in both indicators. Beta has a coefficient of -0.02 which stands for negative influence on ROA, which means that ceteris paribus, increase in beta by 1 affects the ROA by -0.025, in other words, the riskier is the company, the less is the return on assets.

4.4 Summary of results

As it appears, both hypothesizes were verified by fixed-effect estimations, in other words, under the assumption that those companies that entered Tax Monitoring Regime have the Matured Internal Control System, while analogical companies do not, the Matured Internal Control System has a positive sign for both log of Tobin's Q (which represents firm's value and long-term performance) and Return on Assets (which represents short-term performance). I left in the paper the results of Pooled estimation for the Model 1 to show that dummies of industrial and international indicators became unrequired during the study due to causing the collinearity with the ICS Dummy. Both models are significant, but the Model 1 (under fixed-effect estimation) includes a higher number of significant coefficients for control variables with a higher positive autocorrelation, while the Model 2 has a lower number of significant coefficients for control variables with a lower positive autocorrelation.



Conclusion

All in all, this study is built on a personal interest in how the Matured Internal Control System affects the value of the firm and on the methods of research for ERM/ICS influence on different financial indicators presented by different authors. In this paper, we discussed the nature of the Internal Control System and how it appears in Tax Monitoring Regime and reviewed the most interesting articles with ERM/ICS research. The base for research is interesting enough, since the presence of ICS may influence not only financial performance but the corporate governance and quality of audit reports. For my own research, I have used a sample of 20 companies, half of which entered Tax Monitoring regime (as the indicator of Matured ICS), mainly from oil/energy sphere, over the period of 2013-2018 (quarterly). My results show significant evidence of ICS influence on both firm's value (Tobin's Q) and return on assets, which respectively indicate for long-term performance and short-term performance. Besides, for applying a fixed-effect panel model, I got rid of two unrequired dummy indicators of international and industrial diversification. Results also show that practically all control variables are significant on either 0.05 or 0.01 level.