Размещено на http://www.allbest.ru/

Bank strategy and its performance



Introduction

This paper aims to discover efficient bank strategies that are able to positively influence its performance. Considering banking activity to be rather challenging and strictly regulated in countries, Russia, in turns, has genuinely outstanding peculiarities of its banking system. Based on this, the paper examines the phenomenon of massive Russian banks failure in the period from 2013 to 2019 in order to identify which strategic decisions of banks regarding asset and liabilities portfolio structures lead a credit organization to perform sufficiently or poorly. The analysis method consisted in application of logistic regression on the panel data containing financial indicators of 896 commercial banks. As a result, the hypothesis that default probability of a bank is affected by its strategic focus of forming asset and liabilities portfolio was statistically confirmed. Moreover, it was found out that concentration of a bank's activity on issuing loans to individuals and to other banks, as well as on raising commercial deposits and interbank loans has a positive effect on bank performance. The prevalence of loans to legal entities and deposits of individuals on a bank's balance sheet, in turns, increases the chances of failure. These new insights for financial and banking spheres have enabled the paper to cover the research gap relating to the absence of critical assessment of various bank strategies.

Banking environment in any country and in any stage of its economic development tends to be challenging. Throughout the whole history of global banking system existence, commercial banks have always been struggling and forced to align with the standards and requirements of Central Banks (hereinafter CB) in order to demonstrate sufficient results to be able to maintain their vitality on financial markets. In accordance with that, the strongest credit organizations survive, and the weakest are liquidated, giving rise to the painful consequences of the liquidation process. Indeed, the phenomenon of bank failure, which is going to be raised and deeply studied in this paper, has a sufficient warranty to be considered a serious occurrence in the economy. Due to its ability to influence direct and related participants of the process immediately, the dynamics of total number of credit organizations defaults may demonstrate the financial and economic climate in a country, not rarely revealing the points of concern (Lanine & Vennet, 2006).

The consequence of being unable to meet financial obligations to federal or state regulator, depositors and creditors, bank failure is the phenomenon that is worth paying attention to in Russia, which banking history in different times was quite unique in many respects (Garvy, 1972). If considering the fact that in 1986 Soviet Russian banking system consisted of just four banks, each of them having a certain function (Gosbank - Central Bank, Stroibank - corporate sector, Vneshtorgbank - international transactions, Sberbank - retail banking and savings) (Johnson, 1994), then after the crucial point of the first commercial bank appearance (1988), the number of commercial banks in the end of 1994 equaled to 2439 (Banki.ru [Electronic source]). However, due to the bank crisis (1995) as well as financial crisis and economic default in 1998, favorable conditions for banking activity slightly came to an end. Because of many small banks insolvency and Central Bank's active measures of toughening the mandatory bank normatives, the first wave of massive bank licenses revocation has started (1995-2000) with the total number of revoked licenses exceeding 1000 (Lanine & Vennet, 2006).

This brief outlook on some vivid peculiarities of the Russian banking sector in the end of 20^th century has generated interest in conduction of the study. When the new chairman of the Central Bank, Elvira Nabiullina, has taken up the position in 2013, the radical changes in the approach of interacting with inefficient banks were implemented, which have become the subject of the second wave of Russian banks failure beginning (Mamonov, 2017). To be more specific, only during 2016, the number of licenses that were revoked by CB was 110, which was 15 per cent from the total number of commercial banks operating in that year. Such a massive license revocations statistics has never been noticed in a new millennium Russia (Figure 1) (Zaitseva, 2016).



Figure 1. 2012-2020 bank number dynamics

phenomenon banking credit

Overall, the banking sector “cleaning” initiated by Central Banks is likely to have a positive effect on economy. With the help of such measures, banking system of a country is being released from unconscientious participants on the market who are able to negatively influence a state's financial environment (Gidadhubli, 2001). However, in any policy mildness and expediency do matter. If the 1997-2004 bank failure wave was rather clear in its premises and reasons, the recent period of numerous defaults genuinely attracts attention (Central Bank of the Russian Federation [Electronic Source]). What is interesting, it is not only the consequence of banks losing their interest in continuing banking activity or mergers and acquisitions happening between commercial banks on the voluntary basis that leads to reduction of incumbents' number. In a great extent, it is because of the Central Bank's severe monitoring actions towards credit organizations which resulted in revealing a plethora of banks either engaged in shady activity or having a discrepancy to the requirements established by the CB for subordinated institutions (Mamonov, 2017).

Here the question arises, why do so many banks fail to follow these requirements successfully? Looking at such tendency of bank failures, there is a suggestion that they may be the consequences of insolvent strategies banks choose as a focus of their loan or deposit activity, which is an essential part of any credit organization's assets and liabilities. As an example, the emphasis on crediting mainly legal entities or issuing interbank loans may potentially be a premise for further bank liquidation depending on which strategy this or that credit institution tended to follow. Indeed, namely because of a strategy`s inefficiency, small and medium-size banks may experience customer outflow to large ones, which consequently leads to these niches' cutback and high concentration of bank assets in the biggest and most influential credit institutions (Mironova, 2017; Seryakova, 2017).

Based on these research incentives, the possibility of identifying efficient bank strategies, taking the availability of a bank failure as a central measure, has brought to the key question, the answer to which is aimed to become the essential of the study: Which bank strategies tend to be the most efficient and resilient?

The current research gap consists in lack of studies that critically assess bank strategies. Thus, the objective of the paper is to reveal such strategies that, during the recent bank failure wave in Russia, turned out to be the most efficient, enabling a credit organization to survive on the market. This is expected to give possibility of evaluating bank performance through its loan and deposit portfolio structures, where the occurrence of a consequent bank liquidation will serve a main measure of a strategy`s viability.

To successfully reach the objective, the next tasks should be implemented:

1. To thoroughly analyze the literature on the topic and reveal rationale for the study;

2. To create comprehensive dataset containing official relevant information about Russian commercial banks;

3. To conduct an appropriate data analysis;

4. To correctly interpret the research results and make reliable implications;

5. To provide a clear answer to the research question;

6. To identify the scopes of future research.

The results of the study are intended to bring benefit to experts studying the bank failure phenomenon as well as to external observers (e.g. investors) in order to evaluate banks performance more accurately. By achieving this, the paper is expected to make a significant contribution to the financial and banking research field.

The paper consists of 5 substantive parts: introduction, where the premises for the research are shown and the research question is stated; literature review, in which all the related terms and prior studies are described in order to reveal the research gap; methodology that delineates all specifics of the analysis conduction; results, where the outcomes are defined, as well as the postestimation of the obtained model is shown; discussion and conclusion, in which the areas of results applicability and the direction of future studies are stated. Each of the parts thoroughly explains the logic and consequence of the research. The sources and additional information are presented in References and Appendices sections respectively.



1.Literature review

1.1 Bank failure

1.1.1 Definition of bank failure

Several studies have already attempted to define a bank failure phenomenon. In the study conducted by Lanine and Vennet (2006), the authors in a major extent drew on the Russian legislation formulation of the credit institution insolvency as its inability to comply with the requests of its creditors, as well as effect mandatory payments. In these cases, the Central Bank revokes a bank license in a month after the obligations were due.

Another more specific way to understand bank failure concept was suggested by Karminskiy and Kostrov (2017) who, for the first time, introduced the definition as a “license withdrawal from a bank with negative capital” based on their research results.

However, it is worth clarifying that these two explanations of a bank failure are related to the compulsory kind of bankruptcy, when a credit institution, regardless its will, is deprived of a license according to the Central Bank's decision. The second bankruptcy type represents the voluntary choice of a bank to leave the market mostly often after the thorough consultation with its creditors. (CCRF, 1994).

The last way of calling a bank failed relates to merges and acquisitions (M&A) happening between credit organizations, where usually more powerful and well-performing organization overtakes a less successful one (Styrin, 2005).

The idea of this study lies in using “bank failure” as the most general term for explaining the exit of a credit organization from the market. All the events leading to exclusion of a bank from the list of operating financial institutes in the country (including the bankruptcy occurrence, other CB's decisions to revoke a license, M&A of banks and voluntary self-liquidation of a bank) are going to indicate the insolvency of a particular bank strategy. In other words, all the definitions presented by above mentioned sources will be equally referred to bank failure in the given research.

1.1.2 Key drivers and participants of bank liquidation process

To understand the process of bank failure, it is also crucial to point out the significance of its key drivers and main participants that influence financial and, particularly, banking systems in countries. First and foremost, these are Central Banks that play key role when interacting with commercial banks as the unique power of admitting a bank insolvent and unable to continue its activity belongs namely to them. To identify an insolvent credit organization, Central Banks establish a set of obligatory financial ratios aimed to regulate capital, liquidity and credit risk exposures, known as `prudential norms' (Lanine & Vennet, 2006). These norms vary depending on a country, though always attempting to make them rational for a state's financial and banking systems peculiarities. Nevertheless, the cases of CBs issuing too strict regulations still take place, which not rarely leads to their criticism and to commercial banks implementing various capital mitigation techniques (Lanine & Vennet, 2006).

In spite of these, it is hard to deny that Central Banks strive to create and maintain favorable banking environment in a country. Taking into account that bank license revocation is a final stage of attempting to enhance the efficiency of a bank system, CBs, first, are in a genuine interest of curing credit institutions, which are on the way to insolvency. Such curing measures include attracting investors, establishing temporary administration management and total control under the cure process (Seryakova, 2017). Only after these attempts fail, CBs start making banking system healthier by revoking banks license.

The second key driver of bank failures serves one of the reasons for Central Banks to strengthen banking system. These are the financial and banking crisis not rarely intervening countries' lives and causing negative consequences for the economic, social and political spheres. There is no need to mention the severe influence of 1998 and 2008 deep financial crisis happening in Russia and globally, as well as 1985-1992 banking crisis in the US that led to massive bank failures and significant changes in banking systems of countries. Namely because of crisis' power to invade regular state life, they are considered to be real drivers of historical bank failure waves. (Cole & Lawrence, 2012; Larina, 2017).

Finally, the importance of state participation in dealing with the problems of banks insolvency is also necessary to note. Especially in the times of overcoming banking crises, governments may provide financial support for some banks, take them under state control or eliminate part of banks (Larina, 2017).

All things considered, these drivers of bank failure are globally admitted. In modern economic conditions, it is literally impossible to find a country that have not faced the crisis moments caused by one of these three elements.

1.1.3 Causes of bank failure

Along with the key drivers, several studies have sought to identify the characteristics that cause banks to fail. For instance, Wheelock and Wilson in their article «Why Do Banks Disappear? The Determinants of U.S. Bank Failures and Acquisitions» (2000), discover a tie between geographical markets and banks fortune. For this reason, in order to limit the impact of possible economic shocks, the availability of geographic diversification where banks' brunches are presents, can limit failures.

Furthermore, the authors also pointed out the power of liquidity and capital adequacy to affect bank failure. According to their assertion, banks that possess little capital or assets of insufficient liquidity and low quality are more likely to fail. Lanine and Vennet (2006) further supported this hypothesis with an assertion that the CB is more likely to revoke a bank license not because of its poor solvency condition, but because of liquidity concerns.

As well as liquidity importance, the equity share in a bank's assets is crucial. Banks with higher equity as a percentage of total assets should be less likely to fail. On the contrary, the less equity a bank has, the less protection it has against loan losses or other declines in the value of its assets (Wheelock & Wilson, 2000). Here it is worth mentioning that such decisions of equity amounts allocation may come out from the risk of depositor panics, which is another cause of bank insolvency. If the negative tendency of clients willing to recall their deposits from banks exists, for credit institutions it seems safer to have more equity in order not to be deprived of sufficient capital (Diamond & Rajan, 2005).

Finally, the managerial approach was found out to be a vivid influencer on banks failure stating that inefficient management increases the likelihood of failures occurring. This is especially relevant for financial markets experiencing sufficiently tight competition between banks. The more rigorous the competition and the less efficient bank management performs, the more vulnerable it becomes in terms of solvency retention (Fungacova & Weil, 2013).

1.1.4 Examples of global bank failures

To emphasize the importance and rationale of the given study in terms of examining the recent bank failure in Russia, such massive revocations of credit institution licenses is a phenomenon that has been appearing globally in various countries and time periods. If taking the world's financial capital - the USA - then for the last 35 years it faced genuinely serious bank failures minimum twice. Starting from 1984, the number of commercial banks in country has fallen by one-third, reflecting the first wave of defaults. Such bank failure peak happened in 1985, when banking system was suffering not only from crisis, but was also experiencing high frequency of acquisitions and mergers caused by increased competition altering the market structure of the banking industry (Wheelock & Wilson, 2000). The second period of bank failures escalation, certainly, took place during the world financial crisis, when, according to the FDIC reports, more than 300 smaller depository institutions were closed in 2008-2010 (Cole & Lawrence, 2012).

Moreover, the analysis of the global bank failures 2005-2015 conducted by Larina (2017) stated the top-three countries with the most large-scale bank reduction statistics. What is interesting, the leading position belongs to India with the reduction percentage of 47 (from 296 to 157 credit organizations), Russia took the second place - 45 per cent of failed banks (from 1249 to 681), and the USA rounds out the top three with the percentage of 30 (from 7556 to 5404 banks).

1.1.5 Attractiveness of Russia for the research

In this section the fascinating insights about Russian banking system are illustrated to demonstrate its attractiveness as a country for the research. Despite being on the second place according to the 2005-2015 bank failure statistics, the Russian banking industry is a unique example of an emerging market, which has undergone a large number of bank failures during the last decade, as well as experienced an impressive economic and banking sector growth in recent years (Fungacova & Weil, 2013). However, having faced a number of crisis times for the Russian economy, the banking system, though trying to step onto the path of prosperity, is still distinctive for the frequency of bank failures. For instance, between 2001 and 2007 more than 250 banking licenses were revoked by the Central Bank of Russia mainly because of the liquidity crisis caused by lack of trust that initiated withdrawals of private deposits.

The fact of sufficiently big number of banks that are operating on the market, nonetheless, does not guarantee the stability of the system due to its diversity. On the contrary, there are a few dominating large state-controlled banks that account for about 40 percent of the sector's total assets. Such proportion of banks, which are owned directly by the government, central bank or regional authorities, remains quite high in contrast to other countries. This puts the rest of the small and medium banks (which number is rather big) in danger of failure when trying to align with big authorities' performance and CB requirements in order to survive. Therefore, such a vulnerable position of many credit institutions demonstrates the fragility of the whole banking sector (Fungacova & Weil, 2013).

Appealing to the research topic dedicated to bank failures 2013-2019, during the first three years after the management change, Central Bank has withdrawn a license from every third bank in the country (Mamonov, 2017). These failed banks are mostly the private resident banks that faced new requirements issued by the CB. Moreover, a number of insolvent banks were not ready to lose business and money. That is why they did not leave the market once there is a negative difference between bank's assets and liabilities, but started falsifying their financial reports, which is absolutely unthinkable for Europe or the US.

Thus, Russian banking system is believed to be an outstanding platform for the upcoming research. The availability of new circumstances and unanalyzed data have a sufficient potential to bring fascinating outcomes to the field of the study.

1.2 Evaluation of bank performance

Looking at the general picture of historical bank failures the next question arises: is it possible for an external observer to somehow predict the forthcoming bank insolvency, or the failure inevitability comes all of a sudden without leaving a bank a chance to remain on the market? Despite the last option potential possibility, as the global and countries' economies are never fully protected from severe economic shocks, analyzing banks performance and forecasting its future can and should be conducted for the good of financial market environment. However, it seems sufficiently challenging to identify the borders that separate a problem bank from crisis bank or a stable bank from a bank having current insignificant difficulties (Larina, 2017).

The first step undertaken in this direction was made after the record numbers of the US bank failures in the 1980s. Following the desire to deal with the uncertainty of whether a bank will perform good or bad, Congress enacted the Federal Deposit Insurance Corporation (FDIC) Improvement Act of 1991 that obligated annual safety and soundness examinations of all the American credit organizations (Wheelock & Wilson, 1999).



1.2.1 Bank rating as an evaluation instrument

In this section, the discourse is going to be dedicated to special ratings that were elaborated for the aim of revealing the financial condition of commercial banks. In case of the US FDIC and its actions towards bank assessing started in 1991, the CAMELS rating was elaborated which, till nowadays, is considered one of the central techniques in assessing banks performance not particularly in the US, but also globally (Wheelock & Wilson, 1999). The rating consists of six components, which are analyzed according to the scale from 1 to 5, where 1 stands for “strong” and 5 for “critically deficient”. The evaluated components are abbreviated as CAMELS where “C” means “Capital adequacy”, “A” means “Asset quality”, “M” - “Management quality', “E” - “Earnings”, “L” stands for “Liquidity” and “S” for “Sensitivity to market risk”. These components in sum made significant contribution to explaining the failures of banks during 1985-1992 as well as in 2009 when the global financial crisis broke out (Cole & Lawrence, 2012).

Implying such a complex evaluation, CAMELS rating, however, is rather costly to perform (Wheelock & Wilson, 1999). For this reason other countries have elaborated own methods of bank performance assessment that align with peculiarities of their financial system, For instance, Germany uses the method of coefficient analysis named BAKIS, which is based on a system of 47 indicators comprehensively assessing credit risk, market risk, liquidity and profitability. In the UK there is RATE system adopted that is based on measuring the effectiveness of risk assessment instruments. Russia, in turns, is using a group method consisting in relating all credit institutions to the five groups based on the conditions of their financial stability, owned capital, assets, profitability, liquidity, interest rate risk, quality of management and transparency of ownership structure (Zakirova et. all, 2018).

A different way to assess and diagnose the probability of default is to appeal to special credit ratings. They represent an independent opinion of a rating agency about the financial condition and stability level of an evaluated bank. Credit ratings are commonly considered a central method of bank assessment due to their sufficient practical significance both for commercial banks and for their depositors (Claeys & Schoors, 2007). Firstly, credit ratings serve for making banks' credit liabilities analysis easier. Secondly, high rating of a credit organization facilitates its client base enlarging and establishment of tighter relations with business partners. Generally saying, credit ratings have a power to create a bank's creditworthiness portrait by taking into consideration plenty of relevant factors (Zhivaikina & Peresetsky, 2017).

1.2.2 Rationale for additional bank evaluation techniques

In spite of credit ratings sufficient popularity and wide applicability, some factors point out their drawbacks. Appealing to Russia, then its relations with this type of bank evaluation is outstanding. After the annexation of Crimea by Moscow in 2014, the world's three leading rating agencies (Standard & Poor's, Fitch and Moody's) have decreased the credit rating of the whole country. In order to become independent from external political pressure, Russia has switched to appealing to the opinions of national credit ratings only (AKRA, Expert RA etc.), which has given thoughts to investors about their reliability (Thompson Reuters [Electronic source]). This fact has become a premise for several cases taking place in 2015, when licenses were revoked from banks with high agencies rating and thus, seeded the idea of their subjective nature.

Understanding that different ratings have different approaches to bank performance evaluation, it seems there is no single commonly accepted way to assess the stability of credit institutions, as various indicators may radically influence the final rating result. Hence, it is possible that one bank will have a high rating according to one rating agency and low rating according to another, which makes them not that reliable if considering only them when evaluating bank performance. Even in the study of Ferri, Liu and Majnoni (2000) the authors claim credit ratings as “lagging indicators failing to predict bank failures in East Asian countries”. Moreover, the research of Karminsky and Kostrov (2017) points out that the correlation between the reasons of the Central Bank to revoke a license and the bank ratings given by rating agencies is low. The authors assert that when revoking a license from a bank, the financial regulator is guided by the criteria which are not really coincident with the ones that are used by rating agencies in their methodology.

Based on this, the rationale of using additional evaluation techniques arises. In spite of bank ratings being good helpers when dealing with commercial banks evaluation, they often do not draw the true picture experts may rely on. Therefore, the paper is aimed to reveal new possibilities of identifying banks efficiency based on their adopted strategy, which importance is about to be emphasized in the next section.

1.3 Bank strategy

1.3.1 The role of strategy in bank performance

The efficient and productive management of a bank cannot be maintained based on day-to-day operating processes only. For this reason, well-designed strategy is required in order to mitigate risks of unprofitable performance and, hence, to decrease failure chances (Wheelock & Wilson, 2000). It was already suggested, that strategy elaboration has to become the compulsory rule for any bank currently operating on the market. Only after this requirement is satisfied, credit organizations can be licensed by Central Banks and assessed in terms of their management efficiency. Moreover, having had a glance at the grievous Russian bank insolvency statistics, not only the creation of a strategy has to be in priority, but its successful adoption and alignment with the CB requirements of how a stable and reliable commercial bank should perform (Zhdankin, 2014).

Indeed, drawing on the previous articles that are consistent with the research direction, many bank crises were caused by wrong internal management decisions or inadequate risks dealing. To support the statement, Larina in her article «Banking Crises: Identification Problems and Resolution» (2017) exemplifies the bankruptcy of “Master-bank” in November 2013. While taking the 65th position in total net assets among all Russian credit organizations, the bank was revealed to be engaged in servicing shady economic sectors, illegal turnovers and multiple violations of money laundering legislation.

1.3.2 Identifying strategic potential of a bank

To assess bank performance in terms of its strategy implementation, various techniques are being used. Here the emphasis will be concentrated on such methods that serve to analyze to which extent a credit organization operates favorably on the market and how it interacts with external banking environment. One type of analysis reveals, so called, strategic potential of a bank, which provides experts or management with the picture of internal bank processes and features demonstrating its strong and weak sides. Among these methods there are Profile of Management Efficacy and SNW-analysis. The first one is conducted on the basis of questionnaire surveys given to managers and employees who evaluate the efficiency of major management sections. On the contrary, SNW-analysis helps to assess the internal bank environment regarding various criteria that are individual for each credit organization. However, in order to obtain the full picture of a bank's strategic potential, the more comprehensive method is also used. It is known as Matrix of Strategy Directions where the central role is given to two parameters - competitiveness of a strategic unit and market development perspectives. These parameters are rated on the 10-point scale, which are provided by expert opinions (Zhdankin, 2014).

On the row with strategic potential, there goes strategic climate of a bank which importance should not be underrated. After the internal performance of a bank is assessed, it is also valuable to identify how this performance is actually demonstrated on the market and whether its clients and other incumbents are admitting a bank activity. The identification of a climate is conducted by analyzing external factors able to impact an organization both in a favorable and negative way (revealing opportunities and threats) with the application of several commonly used methods: PEST-analysis, 5 Porter's Forces Model and Weight Matrix (Zhdankin, 2014).

1.3.3 Analysis of banks' various strategic behavior

The given section considers how credit organizations can operate their assets and liabilities, and to which outcomes it may lead.

Several studies provide the paper with such opportunity of general understanding about banks' various strategic decisions. In the paper written by Alves, Dumski and de Paula (2007) it was suggested that bank performance can be not only determined by its own strategy, but simultaneously exposed to bank competitors behavior. For this hypothesis the situation modelling was based on different conditions to information access: the first one stood for complete uncertainty of a bank regarding its borrowers and the prospects of return on loan projects, the other assumed the availability of all the information necessary for individual operating process maintenance.

Amusingly, these two situations imply completely different approaches to strategy creation. In case of the first one, the `hang together' mentality expressed in imitating other banks' loan expansion behavior (intensive enlargement of loan operations aiming to obtain additional profits) would be a really useful and safe step to undertake, as it guarantees both market-share and institutional reputation. However, in the situation of complete absence of any information lacks, following a benchmark is considered irrational due to the possibility of banks to determine highly competitive deposit and loan volumes. Thus, the bank's market share may become dominant due to individual marketing abilities, attractive deposit rates and number of branches.

Another factor that forces bank to adopt different strategic approaches is the attitude to risk facing. On the one hand, there are banks that target their credit expansion regardless the extent to which they may face liquidity risk in order to reach the goal. This implies credit institutions managing their liabilities by setting lending targets before knowing their deposit volume will cover their loan commitments. On the other - so called “safety first” type of banks that initially determine acceptable liquidity-risk levels and on this basis conduct their loan-market activity.

This article significantly clarified how banks can determine and manage their deposit and loan strategic behavior. Supposing that banks in general have two components for their strategy identification - how aggressively to market deposits, and how aggressively to make loans, it becomes quite crucial to determine upon these extents in order not to experience illiquidity in case of shock and put themselves in danger of a default. For this reason, a mild strategy is believed to be optimal for adoption that implies not too high (or too low) interest rates (Berardi & Tedeschi, 2017).

To sum up, strategy do influence on the development of banking system. When setting up goals and elaborating plan of their successful achievement, without formulating bank's mission and conducting thorough analysis of its external and internal environment it will be impossible for bank management to choose a right strategic direction.

1.4 Prior researches on the topic

Having considered the key concepts leading the study, the analysis of prior researches in which such terms also were the subject of scientific interest is undoubtedly needed.

The study of Wheelock and Wilson «Why Do Banks Disappear? The Determinants of U.S. Bank Failures and Acquisitions» (2000) aims to discover the probability of bank failure. The authors took management quality as a measure of banks productivity based on the earlier hypothesis that “poorly managed banks are likely targets for acquisition” (Hannan & Rhoades, 1987). In order to identify whether a credit organization is productive or not, the paper concentrates on assessing management quality by analyzing financial reports in the general framework of CAMELS rating. Overall, six major indicators were chosen for further management quality measurement: A1 = total loans/total assets; A2 = real estate loans/total loans; A3 = commercial and industrial loans/total loans; A4 = other real estate owned/total assets; A5 = income earned, but not collected on loans/ total assets; A6 = nonperforming loans/total assets. Taking into account that on the row with “total loans” indicator the ones that differ according to their type of loan concentration (real estate loans, commercial and industrial loans etc.) were also considered, the similarity of approaches is visible. As namely this difference in loan and deposit concentration is intended to be used in the study, the key findings of whether concentration in either category affects the probability of failure will enable to identify the most efficient and resilient strategies.

Another paper of the same research direction studying the US bank failures “Dйjа Vu All Over Again: The Causes of U.S. Commercial Bank Failures This Time Around” (Cole & Lawrence, 2012) was also attempting to identify the failure determinants. Having applied the logistic regression on the 2009 financial crises data, the authors claimed the CAMELS rating to be highly efficient when evaluating banks that lead them to a conclusion of putting limits on commercial real estate loans.

Furthermore, Lanine and Vennet (2006) were examining bank failures in Russia in order to build a model aimed to assess the risk of failure of Russian commercial banks. Being interested in sufficiently high number of failures in the last decade in a country as well, the authors attempted to develop Early Warning System for Russian credit institutions by applying logit and trait recognition models, which would enable to identify problem banks and avoid their future bankruptcies. The explanatory variables were determined as indicators from banks' financial reports, as the availability of highly relevant data (on a monthly basis) is believed to facilitate the obtaining of accurate and reliable outcomes due to the possibility of observing the shifts in the financial position of banks.

In addition, the research conducted by Mamonov (2017) is dedicated to the same phenomenon of recent massive bank liquidations in Russia. The outstanding peculiarity of the study is its comprehensive analysis of bank license revocation cases from 2013 to 2016 that were initiated by Central Bank because of revealing their severe violation of CB requirements, namely the balance sheet falsifications. This analysis in consequence enabled the author to formulate the new hypothesis of bank balance falsification, high assets turnover and low margin of the banking business and offer the indicators that could properly test them.

Mamonov used the official data extracted from “Vestniki Banka Rossii” that included 106 bankrupt banks. After applying the robust regression, the factors leading to “holes” in the capital and consequent balance falsification were found out. What is more, the similar study of Karminsky and Kostrov (2017) also aimed to reveal factors influencing on capital “holes” appearance, but with an application of the logit regression.

Couple of more studies were dedicated to bank performance examination. Fungacova and Weill (2013) investigated the influence of market power on bank failure occurrence and revealed that banking environment is highly competitive which, indeed, can serve a driver of credit organizations bankruptcy.

Finally, the paper “Estimating bankruptcy probability of credit organizations” (Zakirova et. all, 2018) examines the modern Russian banking system by applying logit regression model of binary choice. The data is based on the 2017 banks bankruptcy statistics, where the financial indicators “Share of deposits of individuals”, “Share of loans to the real sector”, “Share of long-term debts in the loan portfolio” serve as key variables. Apart from practical value of the obtained model that can be consequently used by banks to assess the risk of their default and, if relevant, to elaborate a strategy to avoid it, the research has revealed the deterioration of financial stability of credit institutions.

These research papers significantly helped to understand the crucial concepts of the study, as well as its incentives. What is more, the specifics of these researches and techniques used in their conduction will be taken into account during the methodology stage of the paper.



1.5 Research gap

The comprehensive literature review revealed the possibility of identifying the research gap currently existing in the area of the study. Despite the availability of sufficient number of prior researches, none of them completely corresponds with the topic of the paper. For instance, the studies of Mamonov (2017), and Karminsky and Kostrov (2017) although similar to this one in terms of data frame, sample and methodology have different research intentions. Both of them were focusing on identifying the factors of bank capital “holes” appearance. Wheelock and Wilson (2000), despite using financial data to measure management quality as well, were still operating majorly in scopes of CAMELS rating.

Thus, the research field is currently experiencing a lack of studies directed to the investigation of strategy influence on bank performance. Taking into account that the number of papers dedicated to banks strategic behavior in terms of its financial activity is rather small, the critical assessment of their efficiency is absent at all.

The given study has an outstanding incentive to assess bank activity through the lens of asset and leabilities portfolio structures. The occurrence of a consequent bank liquidation will serve a main measure of a strategy`s efficiency and resilience. In other words, this procedure is believed to enable the forthcoming research results to define the most efficient strategies that various banks adopt when deciding on concentration of asset and liabilities portfolio. What is more, this research is intended to bring some interesting outcomes as it is conducted based on the peculiarities of the Russian credit organizations market and considers the new data about massive bank failures initiated by Central Bank as a fascinating quasi-experiment.



2.Methodology

2.1 Research hypothesis

Having analyzed the existing literature on the topic and revealed the peculiarities of the modern Russian banking system, the preliminary hypothesis was formulated in relation to the research question of the study:

H1: Strategic focus of forming asset and liabilities portfolio of a bank has an effect on its default probability.

Based on the further paper sections describing data gathering and its analysis processes it is expected to obtain statistical premises for rejecting or not rejecting the null hypothesis on the 95 per cent confidence interval.

H0: Performance of a bank is not affected by its strategic choices of forming asset and liabilities portfolio.

Moreover, new important findings about the efficiency of various strategies in terms of structuring banks' asset and liabilities portfolio are anticipated to be obtained.

2.2 Data sample

The process of dataset creation included collection of comprehensive information about the Russian banking system relevant for the time of massive bank failures being studied. This information itself represents the list of all Russian commercial banks which performance is demonstrated from January 01, 2012 to December 01, 2019. Although bank failure wave began in 2013, the year 2012 was included in order to have a look at the banking system as it was one year before the dramatic changes in banking regulation have started. Overall, there are 896 credit institutions presented, the information about which was available, and which fitted in the research scopes. All of them are commercial banks that either have left the financial market due to the Central Bank's decision, merger or acquisition or financial insolvency. All the data is structured as a panel dataset where all the indicators included in the forthcoming analysis are captured on a monthly basis. For each bank the number of periods is numerated from 1 to 96, where the first period is January 2012 and the 96^th is December 2019. This enables to conveniently follow banks number dynamics within the period the Central Bank was implementing its radical policy in relation to inefficient banks and revoked licenses. In other words, if the cells containing some data in a previous month have become blank in a consequent month, this will mean that bank has defaulted (Appendix 1).

2.2.1 Dependent variable

As the study is aimed to reveal strategies that are able to influence on either a bank is likely to successfully remain on the market or to fail, namely the occurrence of the last event will be considered determinant for measuring strategy's efficiency. Proceeding from this, for the upcoming data analysis the dependent variable “Bank Failure” was chosen. Its nature is binary which implies that it can only be expressed in two ways: either “1” if a bank failed during the 2012-2019 time period or “0” if it survived.

2.2.2 Independent variables

Aiming to reflect banks' performance, the dataset consists of sufficient number of indicators that will serve either independent or control regressors. Such current uncertainty about roles is explained by complicacy to identify levels of significance of this or that indicator in influencing bank's performance before making data analysis. Overall, there are two groups of variables differing in their nature and way of collection.

The first group consists of 11 financial indicators revealing banks' state of their parts of balance and income sheets (Table 1).



Table 1. Description of financial indicators

The data was extracted from Banki.ru [Electronic source] - an official website containing all the relevant information about Russian credit and non-credit organizations. Its reliability is additionally supported by usage of this source in related papers (Fungacova & Weil, 2013; Karminskiy & Kostrov, 2017; Zakirova et. all, 2018).

These indicators are strongly important as, appealing to the research hypothesis, they will help to identify general strategy of credit organizations in terms of structuring their asset and liabilities portfolio. For this aim, six additional variables are created that represent shares of each loan and deposit type in bank's related portfolio:

1. Share of Individual Loans = Individual Loans / Total Assets (IL_TO_TA)

2. Share of Commercial Loans = Loans to Legal Entities / Total Assets (CL_TO_TA)

3. Share of Interbank Loans = Issued Interbank Loans / Total Assets (IB_TO_TA)

4. Share of Individual Deposits = Individual Deposits / Total Liabilities (ID_TO_TL)

5. Share of Commercial Deposits = Deposits of Legal Entities / Total Liabilities (CD_TO_TL)

6. Share of Interbank Loans Raised = Interbank Loans Raised / Total Liabilities (IB_TO_TL)

These shares will become bank's strategy identifiers as well as the key regressors in the forthcoming data analysis.

The second group of variables consists of seven mandatory bank normatives. They are enacted by the Central Bank and are compulsory for all commercial banks to follow. The decision to include these normatives in data analysis was made based on their outstanding role in evaluating a credit institution performance that is being used by the CB. It is anticipated that the usage of them as additional control regressors would increase the general quality of the model and make it more exponential. The meanings of each normative are presented below (Table 2).



Table 2. Description of bank normatives

The indicators were gathered from The Central Bank of the Russian Federation website [Electronic source] which provides the possibility to compare these values with the minimum or maximum numbers that are strictly required to follow for normal and legal bank functioning. The data shows whether a bank succeeded to fit in the requirements or not, and how well it was performing (Central Bank of the Russian Federation [Electronic Source], 2012).

Summing up, all the chosen variables comply with the research topic. Due to high availability of the data, it is believed to obtain reliable implications for the study that may provide new explanations of massive bank license revocations.



2.3 Data analysis method

In the given section the methods of data processing are described in details. The panel type of structuring the dataset implies conduction of thorough quantitative research to correctly examine the dependency of bank failure on key financial indicators representing its strategy. For this aim the regression modeling is required.

On the back of prior researches and intention to assess a statistical probability of credit organization insolvency arising from improper strategies adopted by it, the panel logistic regression with random effects will be applied. The choice of the regression is fully explained by dichotomous nature of the dependent variable “Bank Failure”, which widely took place in the studies of similar direction (Arena, 2008; Fungacova & Weil, 2013; Lanine & Vennet, 2006; Zakirova et al., 2018; Zhivaikina & Peresetsky, 2017).

For generating the equation of the regression model intended to be used in the study, initially, the basic formula (1) for the multiple linear regression is presented below:

y = b₁X₁ + b₂Х₂+ ...+ b_nX_n + a, where (1)

X₁, X₂, X_n - the values of the independent variables,

b₁, b₂, b_n - the coefficients aimed to be obtained after applying the regression, a - constant (the intercept point of y).

However, despite attempting to reveal the specifics of relationship between the dependent variable of a binary choice and independent variables of other types (ordinal, nominal, interval etc.), this equation may provide outcomes that are greater than one or less than zero by default. For this research this is unallowable, as the probability of an outcome can only be expressed in the diapason from 0 to 1. If p belongs to the interval from 0 to 0.5, then it will mean the failure is not likely to happen if adopting a strategy. On the contrary, the values from 0.5 up to 1 will stand for the higher failure chances (Neuhaus, 1992).

Thus, in order to treat the dependent variable as dichotomous, the equation will be modified in accordance with the goal of the econometric modelling for this study. The new formula (2) for the regression is the following:

p = , where (2)

p - the probability of a bank failure,

e - the base of the natural logarithm equaling to 2,71828,

y - the standard regression equation (1).

Such modification is known as logit, which measures how the log odds of a credit organization default will be affected if particular independent variable changes by one unit (Neuhaus et al., 1991). To be clearer, when running the panel logistic regression reporting coefficients, the sought-for values will be calculated as a ratio of the probability of bank failure to the probability of its survival. Afterwards, these ratio know as odds will be taken its natural logarithm, hence, obtaining the log odds values. This step is undertaken in order to reveal the non-linear relationship between the chosen predictors and the dependent variable. This provides the possibility of the regression results to be more representative in terms of their interpretation: if the obtained coefficient is less than 0, that will mean the failure is not likely to happen if adopting a strategy. On the contrary, the positive values will stand for the higher failure chances.

...