Does State Ownership of Banks Matter?

Abstract

This article examines the effects of state ownership and government interventions on lending behaviour and capitalisation of banks over the period 2005–2011. Using data from the highly state-influenced Russian banking sector, it is documented that the relationship between state ownership and lending is nonlinear. While overall loan growth decreased and interest rates rose, it was found that fully state-controlled banks increased lending and charged lower interest rates during the crisis of 2008–2010. Moreover, fully state-owned and state-supported banks demonstrated counter-cyclical lending behaviour during the crisis. However, while state-owned banks were better protected against asset default, there is weak evidence to suggest that government interventions may result in increased riskiness of banks.

Keywords

Bank lending state ownership privatisation financial crisis government interventions

1. Introduction

It is generally accepted that state ownership of commercial banks is inefficient, may have a negative effect on financial performance and is not superior to other forms of ownership (refer e.g., Barth, Caprio, & Levine, 2004; Bonin, Hasan, & Wachtel, 2005a; Boubakri, Cosset, & Guedhami, 2009; Cornett, Guo, Khaksari, & Tehranian, 2010; La Porta, Lopez-de Silanes, & Shleifer, 2002). Consequently, over the past few decades, governments of many developed countries have been rapidly transferring their ownership rights to the market and private investors. At the same time, many emerging countries that still have high levels of state ownership in banking sectors have been actively developing privatisation programmes (Megginson, 2005).

However, since 2008 and the onset of the global financial crisis, instead of massive privatisation, both developed and emerging countries experienced large-scale nationalisation and bailouts of private banks (Brunnermeier, 2009; Erkens, Hung, & Matos, 2012). Such actions were inevitable in most of the cases and were conducted in order to avoid large-scale bankruptcies and meltdowns of financial systems.¹ This reverse transfer of ownership may indicate that state ownership of banks is not necessarily harmful and may even be more desirable in times of financial turmoil.

This article focuses on the effects of government ownership on lending behaviour and capitalisation of commercial banks over the period 2005–2011. Using data from the highly concentrated and state-influenced Russian banking sector, it is documented that the relationship between state ownership and lending behaviour is nonlinear. While the overall banking sector experienced a decline in lending with increased costs of debt, banks that were fully controlled by the government significantly increased lending amounts and charged lower interest rates during the crisis of 2008–2010. Furthermore, state-owned banks appeared to be more capitalised and, therefore, were better protected against asset default than privately owned banks. These empirical findings suggest that high levels of state ownership of banks may be particularly valuable during periods of financial turmoil.

The article also examines the effects of government actions in response to the financial crisis of 2008–2010 on bank lending behaviour and capitalisation. To the best of the author’s knowledge, this is the first attempt to assess the effectiveness of state interventions through the state-owned banks during the crisis period. The results indicate that government support of state-owned institutions seems to fulfil its aim and cause counter-cyclical behaviour of supported banks.

Existing theories suggest that state ownership of commercial enterprises is ineffective because of at least three reasons. First, political interference always conflicts with market objectives and, therefore, would certainly deviate a firm from value maximisation aims (Boycko, Shleifer, & Vishny, 1996; Shleifer & Vishny, 1994). Second, managerial incentives in state-owned enterprises tend to be weaker compared to those in privately owned firms (Shleifer, 1998). Finally, inferior incentives of a government as an owner to implement monitoring efforts may lead to sub-optimal levels of performance (Shirley & Walsh, 2000).

The prior empirical literature on the state ownership of banks in general supports the theory and suggests that commercial banks with state ownership usually underperform those that are privately owned. A large number of studies primarily focus on the non-crisis periods and examine the association between state participation in the banking sector and either financial stability and development or bank performance itself. Many cross-sectional studies find a negative relationship between government ownership of banks and financial development, stability and economic growth (Barth et al., 1999, 2004; Gur, 2012; La Porta et al., 2002). It is also well documented that government participation in the banking sector may affect lending behaviour and be more politically connected (Carvalho, 2014; Dinc, 2005; Infante & Piazza, 2014; Khwaja & Mian, 2005; Micco & Panizza, 2006; Micco, Panizza, & Yanez, 2007; Sapienza, 2004).

Another strand of literature suggests that state ownership of banks is also associated with lower performance, weaker corporate governance and higher levels of risk. Cross-country studies find that state-owned banks are less efficient, experience lower profitability, have worse loan quality and higher interest expenses, lower governance quality and tend to be more risky despite their lower average costs (Bonin et al., 2005b; Borisova, Brockman, Salas, & Zagorchev, 2012; Hawkins & Mihaljek, 2001; Iannotta, Nocera, & Sironi, 2007; Mian, 2003; Shen & Lin, 2012). Several country- or region-specific studies confirm these findings (refer to, e.g., Berger, Clarke, Cull, Klapper, & Udel, 2005; Berger, Hasan, & Zhou, 2009; Cornett et al., 2010). More recently, it has been documented that state-owned banks tend to have higher levels of operating risk (which especially increases in election years) but much lower default risk compared to privately owned banks (Iannotta et al., 2013; Pennathur, Subrahmanyam, & Vishwasrao, 2012). However, Shen, Hasan and Lin (2014) argue that such underperformance of state-owned banks may be due to political aspects of the governmental role in the banking sector. This role requires state-owned banks to be involved in acquisition of distressed banks. Furthermore, Shen et al. (2014) show that there are no differences in performance between state-owned banks that are not involved in mergers and acquisitions (or those that purchase non-distressed banks) and privately held banks.

Nevertheless, according to previous empirical evidence, bank efficiency and profitability as well as financial stability of the whole economy is less likely to be associated with the state’s participation in the banking sector as an owner. Ideally, as proposed by the theory, state ownership in the banking sector is desired to approach zero. However, since the majority of the empirical evidence comes from normal, non-crisis periods, the only conclusion we can make is that state ownership does not create any additional value during the stages of economic expansion. Yet, facing crises and economic downturns, governments are more likely to interfere and serve as a guarantee for the distressed banking sector.² Therefore, state ownership of banks may actually be particularly valuable in harsh times.

There are only a few studies that investigate the behaviour of state-owned banks during crisis periods. One of these exceptions is Brei and Schclarek (2013), who show that government-owned banks in general increased lending during the financial crisis of 2008–2010 by using a sample of 764 banks from 50 countries. They also find that typical state-owned banks do not receive more equity or deposits during the harsh times, and hence, the increase in lending is most likely to be associated with the government response to a crisis situation. Another recent study by Cull and Martinez Peria (2013) examines the impact of bank ownership on credit growth in developing countries around the financial crisis of 2008–2010. They document that government ownership has different effects on lending growth in Eastern Europe and Latin America. In particular, state-owned banks significantly increased lending during the crisis compared to private and foreign banks in Latin America, while state ownership of banks in Eastern European countries did not impose notable lending growth during the credit crunch episode.

Furthermore, De Haas, Korniyenko, Loukoianova and Pivovarsky (2012) argue that state-owned banks can be considered as a relatively stable source of credit compared to foreign bank subsidiaries (which were not part of the Vienna initiative) that are found to decrease their lending even before the financial crisis of 2008–2010 had started in a number of Eastern European countries. More recently, Albertazzi and Bottero (2014) also find that foreign banks restricted credit supply more sharply than their local competitors in the Italian market. De Haas, Korniyenko, Pivovarsky and Tsankova (2015), in turn, find that state-owned banks reduced credit growth less in 2009. Finally, Fungáčová, Herrala and Weill (2013) document that credit supply levels varied across ownership types in Russia during the financial crisis of 2008–2010. Using the stochastic frontier approach, they find that while overall credit supply diminished, the reduction was lower for state-owned banks and higher for foreign banks.

This article contributes to the above literature by examining the direct effects of state ownership on lending behaviour and bank capitalisation, specifically around the time of the global financial crisis of 2008–2010 using a panel of Russian banks. A within-country analysis that is less prone to endogeneity issues compared to large cross-country studies is expected to provide a more detailed view on the effects of state ownership on bank lending behaviour and capital buffer. In addition, the Russian banking sector, which can be characterised as highly state influenced with dense concentration, provides an appropriate environment to examine this issue. The empirical analysis presented in this article may also have important policy implications for the emerging markets with high state influence in the banking sector.

This article is different from the above discussed literature in two aspects. First, according to several recent studies, the relationship between ownership structure and different aspects in bank operations may exhibit nonlinear behaviour (refer to, e.g., Barry, Lepetit, & Tarazi, 2011; Iannotta et al., 2013). And given that none of these studies focus on the potential nonlinearity between state ownership and lending behaviour and capitalisation of banks, this article contributes to the existing literature by examining the exact levels of state ownership in banks’ equity. These data are manually collected from several sources that are discussed in more detail in the data description section. Second, this article aims to contribute to the existing literature on the efficiency of government interventions in response to the financial crisis. Essentially, this is the first attempt to assess whether the government support through the state-owned banks has any effects on lending behaviour and bank capitalisation.

Using a sample of 348 large Russian banks, I find that the relationship between state ownership and lending behaviour is nonlinear and high levels of state ownership may be more useful during financial crises. In particular, the results indicate that the growth of loans is positively associated with high levels of government participation in the ownership stake of banks, especially during the crisis period. At the same time, fully state-owned banks charge lower interest rates on loans. Moreover, I find that state interventions during the crisis of 2008–2010 were rather successful as supported banks increased lending, while charging lower interest rates on their loans. It is also documented that there is a positive relationship between capital ratios and state ownership around the crisis, implying that the government as an owner may provide more protection from asset default. However, there is weak evidence to suggest that state intervention may have forced supported banks to take more risky projects, which negatively affected their capital ratios.

These results are broadly consistent with the existing studies that focus on the financial crisis period (refer to, e.g., Brei & Schclarek, 2013; Cull & Martinez Peria, 2013; De Haas et al., 2012, 2015; Fungáčová et al., 2013). Although state ownership has a negative impact on bank performance, efficiency and economic development in normal times (refer e.g., Barth et al., 2004; Bonin et al., 2005a; Cornett et al., 2010; La Porta et al., 2002), the empirical findings of this article demonstrate that government participation in the banking sector may outweigh these disadvantages during crises episodes.

The remainder of this article is organised as follows. Section 2 describes the Russian banking sector and discusses government participation in it. Section 3 introduces the data, while Section 4 presents the methodology. Empirical findings on the effects of government ownership of banks on lending behaviour and capitalisation as well as the effects of state interventions during the crisis are reported in Section 5. Finally, Section 6 concludes the article and discusses potential policy implications.

2. Overview of the Russian Banking Sector

Although the number of commercial banks in Russia has been constantly decreasing over the last decade (from 1,253 in 2005 to 978 in 2011), it is still relatively high if scaled to the size of the economy. For comparison, in the similar emerging market of Brazil, there are less than 150 commercial banks. Hence, the Central Bank of Russia (CBR) still follows the policy of decreasing the number of commercial banks by rising minimum capital requirements and imposing stricter financial standards. Minimum capital requirements, for example, doubled since 1997 from 5 per cent to 10 per cent. Nevertheless, even with such a large number of financial institutions, the Russian banking sector is relatively concentrated both geographically and assets wise.³ The top 200 banks account for more than 94 per cent of total assets of the Russian banking sector, while the top 5 banks hold up to 50 per cent. A similar picture can be observed on the capital side—the top 200 banks hold 93 per cent, while the top 5 banks account for more than 50 per cent of total capital. At the same time, the Herfindahl-Hirschman Index (HHI) being at moderate levels for assets (0.092) and capital (0.101) mainly due to the large number of financial institutions still remains relatively high for deposits (0.225).⁴

While high concentration can potentially cause lower levels of competition and lead to higher interest rates and fees, it may also lead to higher stability in the banking industry through better diversification, higher profitability and the easiness of monitoring of large banks. In fact, it has been argued that high concentration is associated with higher interest margins only for foreign banks in Russia, while state-owned banks, despite their large market share, do not exploit their market power in terms of interest rates (Fungáčová & Poghosyan, 2011).

Table 1.
Main Characteristics of the Russian Banking Sector

Panel A. Russian banking sector 2005 2011 Change, USD Change, %

Total assets, in bill. USD 341.3 1415.3 +1 074.0 +414.7

Total capital, in bill. USD 43.5 178.2 +134.7 +409.7

- % to total assets 12.74 12.59 −0.15

Total loans to non-financial firms and households, in bill. USD 190.9 791.1 +600.2 +414.4

- % to total assets 55.94 55.89 −0.05

Capital adequacy, %* 16.0 14.7 −1.30

Panel B. Ownership structure
2005
2011

Number % of total assets % of total capital Number % of total assets % of total capital

State owned banks ** 32 40.7 33.8 26 50.2 50.8

Foreign banks 51 8.3 9.2 108 16.9 17.6

Large private banks 183 41.2 42.8 132 27.5 24.9

Other private banks 939 9.3 14 656 5.0 6.5

Other credit institutions 48 0.5 0.2 56 0.4 0.2

Total 1253 100 100 978 100 100

Panel A. Russian banking sector	2005	2011	Change, USD	Change, %
Total assets, in bill. USD	341.3	1415.3	+1 074.0	+414.7
Total capital, in bill. USD	43.5	178.2	+134.7	+409.7
- % to total assets	12.74	12.59		−0.15
Total loans to non-financial firms and households, in bill. USD	190.9	791.1	+600.2	+414.4
- % to total assets	55.94	55.89		−0.05
Capital adequacy, %*	16.0	14.7		−1.30

Panel B. Ownership structure	2005	2011
State owned banks **	32	40.7	33.8	26	50.2	50.8
Foreign banks	51	8.3	9.2	108	16.9	17.6
Large private banks	183	41.2	42.8	132	27.5	24.9
Other private banks	939	9.3	14	656	5.0	6.5
Other credit institutions	48	0.5	0.2	56	0.4	0.2
Total	1253	100	100	978	100	100

Source: The data are obtained from the Central Bank of Russia.

Notes: * minimum required ratio is 10.0%.

** numbers according to the CBR’s classification of state ownership—direct holding of more than 50%.

Table 1 presents the main characteristics of the Russian banking sector. As can be noted from Panel A of the table, the Russian banking sector experienced substantial growth over the last decade. Total assets as well as total capital grew by more than four times during 2005–2011. Commercial lending to non-financial firms and households also quadrupled over the same period. It can be also seen from the table that capital and commercial loans were growing proportionally to total assets of the banking sector. Hence, the total capital to assets ratio remained approximately at the same level—12.5 per cent, while the ratio of commercial loans to total assets stayed at around 56 per cent.

Being moderately concentrated in terms of assets and capital, ownership density in the Russian banking sector is even more compact. Panel B of Table 1 characterises the ownership structure of the banking industry. As can be seen from the table, according to the CBR, 50.2 per cent (50.8%) of the industry’s total assets (capital) were controlled by the state-governed banks by the end of 2011. It is important to note that there are no large differences in target markets between state-owned and privately held banks in Russia. Most of the banks apply the universal model of banking and follow the same regulations of the CBR, irrespective of their ownership structure. However, it is acknowledged that some of the state-owned banks may have politically related connections with their customers. These connections may potentially affect banks’ lending behaviour (Barry, Lepetit, & Strobel, 2016).⁵ Nevertheless, Berger, Klapper, Martinez Peria and Zaidi (2008) find only mixed support for the hypothesis that state-owned banks serve any particular kinds of firms. It should also be noted that state ownership of commercial banks in Russia is customary. The largest three banks in Russia have always been state owned. Hence, in contrast to most of the developed markets, the process of bailouts of banks in Russia during the financial crisis of 2008–2010 had only a marginal effect on the distribution of state-controlled banking assets. In particular, the Russian government bailed out only a few relatively small regional banks in 2008–2010, while the reduction of systemic risk in the economy was achieved solely through the existing large state-owned commercial banks.

Although CBR defines a bank as state-owned if more than 50 per cent of equity belongs to the government, certain influence on decision-making may be exerted even when the government does not hold majority stake. Vernikov (2009), for example, finds that the difference between state-owned and state-controlled banks may significantly affect the determinants of market shares. He documents that the gap between state-owned and state-controlled (-influenced) banks is nearly 13 per cent of the industry’s total assets. Thus, the definition of state ownership must be treated with some caution. The CBR’s classification does not necessarily describe the overall influence of the government in the sector also because it does not consider a bank as state owned if state-owned industrial companies own the bank’s equity.

Although foreign banks have doubled their market share since 2005, they still account only for about 17 per cent (17.6%) of the industry’s total assets (capital). Nevertheless, it has been found that foreign banks tend to be more efficient than domestic private banks in Russia (Karas, Schoors, & Weill, 2010). Interestingly, it can be noted from Panel B of Table 1 that large private banks lost substantial amount of market share around the crisis. The total assets held by those banks decreased from 41.2 per cent in 2005 to 27.5 per cent in 2011. Finally, it can be seen from the table that the number of other private banks decreased almost by one-third—from 939 to 656 banks—while the market share held by these banks fell by roughly 4 per cent.

Apart from concentration of the banking sector, all commercial banks in Russia experienced significant difficulties during the financial crisis of 2008–2010. Formal indicators of the crisis first started to appear on the capital side of banks. Expecting and preparing for the possibility of bank runs, many financial institutions preferred to transfer their assets into more liquid instruments, which significantly distressed profitability and consequently negatively affected capital ratios. In addition, financial difficulties of borrowers forced many banks to increase their reserves and loan loss provisions. As a result, some banks ended up with negative capital. But in contrast to the Russian crisis of 1998, the Russian banking sector managed to avoid massive bank runs and bankruptcies largely due to extensive government support. While developed countries had to initiate exceptional monetary policies and force significant nationalisation of financial institutions (Laeven & Valencia, 2010; Lenza, Pill, & Reichlin, 2010), the Russian government stepped in with capital injections, preferential loans on favourable terms and long-term deposits to state-owned banks. To a large extent, these actions allowed the Russian banking sector to avoid a collapse.

3. Data

The empirical analysis in this article is based on data on large Russian banks for fiscal years 2005–2011. The data are mainly obtained from Bureau Van Dijk’s BankScope. Given the peculiarities of the Russian banking sector and its concentration, smaller banks that operate in specific fields or regions are not of great interest to this analysis. Therefore, all banks that had less than USD 100 million in total assets by the end of 2007 (the year prior to the crisis) are omitted from the sample. Further, banks with insufficient financial information are left out. Specialised credit institutions that mainly serve as development banks and governed by distinctive regulations are also excluded from the sample.⁶ Lastly, to avoid any potential bias, banks owned by foreign governments are not included in the analysis. In order to ensure that outliers do not affect the results, the data used in the analysis are winsorised at the 1 per cent and 99 per cent levels.

The main variable of interest is state ownership. Many previous studies on state ownership of banks use a dummy variable that takes the value of 1 if the government holds a certain percentage of equity (refer to, e.g., Brei & Schclarek, 2013; Iannotta et al., 2007; Jackowicz, Kowalewski, & Kozlowski, 2013; Shen & Lin, 2012). However, quite often, the obtained dummies are rather static over time and account only for exceptional changes in ownership, depending on the definition of the variable. Moreover, dummy variables cannot capture the nonlinear relationship and, therefore, may lead to incorrect inferences. In contrast to previous studies, this article uses a different approach that accounts for possible dynamic changes in state ownership and for potential nonlinearity. Thus, instead of a dummy, this article uses a continuous ownership variable defined as the percentage held by the government in the bank’s equity.

There are, however, two potential problems with this approach. First is the problem of lack of historic data. For certain banks, usually smaller ones, BankScope does not contain all historical information on shareholders. If that was the case, the information on owners was hand collected from different publicly available sources.⁷ The second problem arises from the complex ownership structures. As pointed out by Vernikov (2009), official classification of state ownership by the CBR is very narrow and congregates only on direct participation of federal or local government in more than 50 per cent of equity stakes. In practice, governments may exert certain influence on the bank’s governance mechanisms even without holding controlling stakes. Moreover, governments may affect the bank’s behaviour indirectly, through sophisticated pyramid-type ownership structures. Chernykh (2008), for instance, shows that the Russian government exercises its control over some financial and industrial companies using extensive ownership pyramids. Therefore, to account for hidden state control, it is essential to determine the final ultimate owner. For this purpose, BankScope ownership data on financial institutions is augmented with Bureau van Dijk’s Orbis database that also contains information on the ultimate owners of industrial firms.

The ownership structure was manually ascertained up to the third level of major stakeholders. For example, if a bank’s equity was mainly held by an industrial company that in turn was controlled by the federal or local government, this bank was defined as state owned and the percentage of equity belonging to the government was calculated proportional to the stakes held in the industrial company. Fortunately, these pyramid ownership structures mostly prevail only in larger banks in Russia.

The final sample used in the empirical analysis consists of an unbalanced panel of 2,196 bank-year observations on 348 individual banks for the fiscal years 2005–2011. This article does not consider data from the later periods because of the higher levels of political uncertainty and recent geopolitical conflicts.⁸ The number of observations varies across variables due to lack of financial data. The sample used covers about 90 per cent of the country’s banking assets and can thereby be considered representative of the whole Russian banking sector.

Table 2 reports summary statistics of the sample. As can be noted from the table, the sample is quite heterogeneous in terms of bank size. Total assets vary from USD 2.12 to 336,534 million with a mean of about USD 2.4 billion. The average growth rate of banks’ total assets is fairly high—around 30 per cent per annum, while the mean growth rate of loans is about 35 per cent. Russian banks also make an average reserve of 8.2 per cent of gross loans to cover their potential losses, while equity covers about 16 per cent of assets and 34 per cent of net loans. Customers’ deposits on an average finance 56 per cent of bank assets. The table also shows that the average Russian bank is rather profitable. While the mean of return on assets is only 1.44 per cent, the average earnings before taxes to total assets are about 2 per cent. Average net interest revenue to assets is more than 5 per cent and cost to income ratio of about 75 per cent implies that Russian banks are relatively efficient and exert utility from attracting cheaper funds and maintaining low running costs.

4. Methodology

The empirical analysis begins with the examination of the relationship between bank ownership and the growth of loans’ supply with the following regression specification:

\begin{matrix} Δ L_{i, t} = α_{0} + β_{1} G O V_{i, t} + β_{2} G O V ​_{i, t}^{2} + β_{3} S I Z E_{i, t - 1} \\ + β_{4} D E P_{i, t - 1} + β_{5} P R O F_{i, t - 1} + β_{6} I N C_{i, t - 1} \\ + β_{7} C T I_{i, t - 1} + β_{8} C A P_{i, t - 1} + β_{9} L L R_{i, t - 1} \\ + \sum_{k = 1}^{n - 1} α_{k} B A N K_{i}^{k} + ​ \sum_{t = 2006}^{2011} ω_{t} Y E A R_{i}^{t} + ε_{i, t} \end{matrix}

(1)

where ΔL_i,t is the growth rate of total loans by bank i at time t. Following Micco and Panizza (2006) and Ferri, Kalmi and Kerola (2014), the growth rate of loans is defined as the difference between log loans at time t and log loans at time t−1. GOV is the percentage of equity stake directly or indirectly held by the government and GOV² is a squared term of state ownership. Following recent literature on the ownership structure and banking (refer to, e.g., Brei & Schclarek, 2013; Cull & Martinez Peria, 2013), control variables for bank size, sources of financing, profitability, income diversification, cost efficiency, capitalisation and riskiness are included. SIZE is the logarithm of total assets, while DEP is the logarithm of deposits to assets ratio. PROF denotes profit before taxes scaled to total assets, INC is a proxy for income diversification calculated as the net interest revenue divided by average assets, CTI is the logarithm of cost to income ratio measured by operating costs divided by operating income, CAP is the logarithm of capital ratio calculated as the equity to assets ratio and LLR is the logarithm of loan loss reserves to gross loans ratio. BANK and YEAR denote bank and year fixed effects. All the bank-specific control variables are lagged by 1 year.

Table 2.

Summary Statistics

Variable	No. of Obs.	Mean	Median	Standard Deviation	Min	Max
Asset structure
Total assets (mill. USD)	2 196	2 496	276.8	153.3	2.12	336 534
Loans growth	1 829	35.07	17.54	122.0	0.66	285.5
Assets growth	2 007	29.82	23.05	38.24	−49.92	231.7
Loans loss reserves/Gross loans	2 148	8.21	5.83	8.57	0.00	108.9
Capital structure
Equity/Assets	2 152	15.85	12.84	9.93	4.78	72.92
Equity/Net loans	2 140	33.81	22.58	40.03	8.02	423.9
Deposits/Assets	2 147	56.39	57.62	18.74	5.58	88.93
Income structure
Return on average assets	2 152	1.44	1.20	1.56	−6.59	8.51
Profit before taxes/Assets	2 152	1.96	1.68	1.74	−6.95	11.33
Net interest revenue/Average assets	2 152	5.19	4.97	2.03	0.86	13.98
Cost to income ratio	2 145	75.70	80.63	20.96	24.37	121.7
Ownership structure
State share	2 196	4.50	0.00	18.14	0.00	100.0
Foreign share	2 196	10.0	0.00	28.77	0.00	100.0
Listed	2 196	0.16	0.00	0.37	0.00	1.00
H-quarters	2 196	0.56	1.00	0.49	0.00	1.00

Source: Compiled by the author.

In addition to the above two-way fixed effects’ panel regression specification, further analysis uses a crisis dummy variable CRISIS that equals one in fiscal years 2008, 2009 and 2010 and the interactions of GOV and GOV ² with the crisis dummy. This approach allows capturing the effect of state ownership on lending behaviour relative to private banks during the crisis period.

Given relatively the high concentration of the Russian banking sector and differences in funding sources, it is important to control for the geographical location of banks as well as for the easiness of access to external sources of financing. In this respect, two additional dummy variables are included in some specifications. Listed denotes whether a bank was listed on stock exchange during the sample period, while H-quarters takes the value of 1 if the bank’s headquarters were located in Moscow.

As the next step of the analysis, to examine the relationship between bank ownership and interest rates on loans, the following regression specification is estimated:

\begin{matrix} A I R_{i, t} = α_{0} + β_{1} G O V_{i, t} + β_{2} G O V_{i, t}^{​ 2} + β_{3} S I Z E_{i, t} + β_{4} D E P_{i, t} \\ + β_{5} P R O F_{i, t} + β_{6} I N C_{i, t} + β_{7} C T I_{i, t} + β_{8} C A P_{i, t} \\ + β_{9} L L R_{i, t} + β_{10} G R O W T H_{i, t} + β_{11} A R D_{i, t} \\ + \sum_{k = 1}^{n - 1} α_{k} B A N K_{i}^{k} + ​ \sum_{t = 2006}^{2011} ω_{t} Y E A R_{i}^{t} + ε_{i, t} \end{matrix}

(2)

where AIR is the average interest rate on loans measured as interest income on loans divided by average gross loans. GOV, GOV ², SIZE, DEP, PROF, INC, CTI, CAP, LLR, BANK and YEAR are defined as in Equation (1). Besides these variables, two additional control variables are included in these regressions. According to the previous literature, cyclicality of the financial sector may have an effect on economic conditions, including the level of interest rates on bank loans (refer e.g., Cardarelli, Elekdag, & Lall, 2011). Therefore, GROWTH—the growth rate of total assets—is included to control for the cyclicality of the financial sector. It is measured as a percentage of the present to the previous year’s total assets. To control for the relative levels of interest rates on the deposits market, ARD is the average interest rate on deposits measured as interest expenses on customer deposits divided by average customer deposits are also included in these regressions. It is important to note that the bank-specific control variables are not lagged in these regression specifications. As described by Lainela and Ponomarenko (2012), the high volatility of interest rates in Russia is customary and has been especially substantial in periods of markets stress. Given the high volatility of interest rates in Russia, it is reasonable to use contemporaneous control variables since lagged values are less likely to capture associated effects on the current levels of interest rates.

Finally, the impact of state ownership on bank capitalisation is examined with the following regression specification:

\begin{matrix} C A P_{i, t} = α_{0} + β_{1} G O V_{i, t} + β_{2} G O V_{i, t}^{​ 2} + β_{3} S I Z E_{i, t - 1} \\ + β_{4} D E P_{i, t - 1} + β_{5} P R O F_{i, t - 1} + β_{6} I N C_{i, t - 1} \\ + β_{7} C T I_{i, t - 1} + β_{8} L L R_{i, t - 1} + β_{9} G R O W T H_{i, t - 1} \\ + \sum_{k = 1}^{n - 1} α_{k} B A N K_{i}^{k} + ​ \sum_{t = 2006}^{2011} ω_{t} Y E A R_{i}^{t} + ε_{i, t} \end{matrix}

(3)

where CAP is the logarithm of capital ratio of bank i at time t, measured by total equity divided by total assets. GOV, GOV ², SIZE, DEP, PROF, INC, CTI, LLR, GROWTH, BANK and YEAR are defined as in Equations (1) and (2). All control variables in this regression specification are 1-year lagged.

5. Empirical Results

5.1 Univariate Tests

Table 3 reports the comparison of the main variables between state-owned and privately owned banks. As can be noted from Panel A of the table, the average growth rate of loans seems to be on the same level for state- and privately owned banks during the whole sample period 2005–2011. Loan portfolios of these banks were growing on an average by 35 per cent per annum. The average capital ratio measured by the equity to total assets ratio is about 16 per cent for all ownership types and the difference between them is not statistically significant. However, as can be noted from the table, there is a highly statistically significant difference in average interest rates on loans between banks. State-owned banks charged on average about 13 per cent on their commercial loans, while private banks charged 134 basis points higher interest rates on their loans.

Next, the means of the same variables are compared exclusively during the crisis period—2008–2010. Panel B of Table 3 reports the comparison of means and medians of lending behaviour and bank capitalisation variables between state-owned and privately owned banks year by year. As can be noted from the table, growth rates of loans are significantly lower in periods of markets stress for both types of commercial banks. The average growth rate of loans for state-owned banks varies between roughly 8 per cent and 11.5 per cent, while the corresponding number for private banks is much more volatile and varies from −4.9 per cent to 24.3 per cent. However, the statistical significant difference between the two types of banks can be observed only in 2008 in median and in 2009 in mean loans growth rate. Median privately owned banks shrank its lending by more than 4.5 per cent in the beginning of the crisis, while the credit portfolio of median state-owned banks grew by more than 7 per cent in 2008. This difference is significant at the 10 per cent level. Moreover, in the peak year of the crisis, in 2009, the average credit portfolio of state-owned banks experienced faster growth than in 2008 and rose up to 11.5 per cent, while lending by privately owned banks continued shrinking and decreased by 4.9 per cent. This difference is statistically significant at the 5 per cent level. Nevertheless, in 2010, both types of banks experienced growth in lending but the difference is not statistically significant.

As can be seen from the table, average capital ratios of private and state-owned banks increased in 2008 and 2009 but slightly decreased in 2010 for state-owned banks, while private banks experienced a decline of about 1.3 per cent from the previous year. However, the difference between the two samples is not statistically significant in any of the years. At the same time, the difference in interest rates on loans is consistently significant at the 1 per cent and 5 per cent levels throughout the whole crisis period. Thus, in 2008, state-owned banks charged 167 basis points lower interest than their private rivals. However, as can be noted from the table, average interest rates on loans slightly increased in 2009 but state-owned banks still were charging 149 basis points lower interest rate. Furthermore, interest rates decreased to pre-crisis levels in 2010, while the interest rate spread between state-owned and privately owned banks increased even more. Government-owned banks charged 211 basis points lower interest rate on their loans than privately owned banks. Overall, Panel B of Table 3 indicates that there are some significant differences in the lending behaviour between state-owned and private banks during the crisis period.

Table 3.
Univariate Tests

Panel A. Full sample period 2005–2011

State Private

Loans growth

Mean 35.00% 34.92%

Median 21.17% 17.32%

No. of obs. 130 1 701

Capital ratio

Mean 15.61% 15.87%

Median 12.21% 12.92%

No. of obs. 147 2 005

Average interest rate on loans

Mean 12.98%* 14.32%

Median 12.92%* 14.13%

No. of obs. 144 1 759

Panel B. Crisis period

2008
2009
2010

State Private State Private State Private

Loans growth

Mean 7.94% −1.04% 11.57%** −4.92% 11.18% 24.34%

Median 7.29%* −4.53% −3.07% −6.38% 12.00% 16.27%

No. of obs. 20 314 22 290 25 287

Capital ratio

Mean 13.19% 16.35% 18.61% 17.02% 18.43% 15.76%

Median 10.94% 13.65% 14.17% 14.18% 12.80% 12.79%

No. of obs. 19 310 22 291 24 285

Average interest rate on loans

Mean 13.41% 15.08% 14.10% 15.59% 12.08%* 14.19%

Median 13.16% 14.71% 13.96%* 15.51% 11.55%* 13.18%

No. of obs. 20 294 22 279 25 276

Panel A. Full sample period 2005–2011
Loans growth
Mean	35.00%	34.92%
Median	21.17%	17.32%
No. of obs.	130	1 701
Capital ratio
Mean	15.61%	15.87%
Median	12.21%	12.92%
No. of obs.	147	2 005
Average interest rate on loans
Mean	12.98%***	14.32%
Median	12.92%***	14.13%
No. of obs.	144	1 759

Panel B. Crisis period
Loans growth
Mean	7.94%	−1.04%	11.57%**	−4.92%	11.18%	24.34%
Median	7.29%*	−4.53%	−3.07%	−6.38%	12.00%	16.27%
No. of obs.	20	314	22	290	25	287
Capital ratio
Mean	13.19%	16.35%	18.61%	17.02%	18.43%	15.76%
Median	10.94%	13.65%	14.17%	14.18%	12.80%	12.79%
No. of obs.	19	310	22	291	24	285
Average interest rate on loans
Mean	13.41%**	15.08%	14.10%**	15.59%	12.08%***	14.19%
Median	13.16%**	14.71%	13.96%***	15.51%	11.55%***	13.18%
No. of obs.	20	294	22	279	25	276

Source: Author’s calculations.

Note: ***, ** and * denotes significance at the 1 per cent, 5 per cent and 10 per cent levels, respectively.

5.2 State Ownership and Bank Lending Behaviour

Table 4 presents the estimation results of Equation (1) with the growth of total loans as the dependent variable. Models 1–3 are estimated for the whole sample period 2005–2011, while Models 4–5 specifically examine the crisis period 2008–2010. Model 1 includes dummy variables for banks that are listed and located in the capital city but excludes bank fixed effects, whereas Model 2 is a two-way fixed effects regression. Model 3, in turn, includes a dummy variable that proxies for the crisis period that equals one in fiscal years of 2008, 2009 and 2010. In addition, interactions of the state ownership and squared state ownership with the crisis dummies are included into the model. Model 4 is estimated in a similar manner as Model 2 but for the crisis period only, whereas Model 5 uses the variable (GOV-High) with high levels of state ownership (at least 95 per cent of equity) for robustness check of the nonlinear relationship between growth of loans and state ownership of banks. As can be noted from the table, all models have a good explanatory power with adjusted R-squares varying between 25 per cent and 42 per cent.

The first four models in Table 4 suggest that state ownership in general is negatively related to loan growth rates. However, higher levels of state participation in banks’ equity have a reverse effect and render a statistically significant positive effect on the growth rates of loans, implying that the relationship between state ownership and lending growth is nonlinear. The observed nonlinearity in state ownership and loan growth is illustrated in Figure 1(a). The figure plots the annual loan growth rates for different levels of state stakes in banks’ equity during the crisis of 2008–2010. As shown in the figure, the loan growth was around zero for an average bank where the government held up to 50 per cent of equity. But those banks where the government kept greater equity stakes were able to sustain higher loan growth rates. The average growth rate of loans for fully state-owned banks during the crisis was about 20 per cent per annum.

Table 4.

State Ownership and Growth of Total Loans

	Dependent Variable: Loans Growth
	1	2	3	4	5
GOV_t	−0.003**	−0.005**	−0.008**	−0.016***
	[2.28]	[1.95]	[2.14]	[2.71]
GOV_t²	0.0001***	0.0001*	0.0001**	0.0003***
	[2.79]	[1.64]	[2.03]	[2.61]
GOV-High_t					0.015***
					[5.11]
Crisis			−0.27***
			[4.33]
Crisis × GOV_t			0.004*
			[1.71]
Crisis × GOV_t²			−0.00
			[1.26]
Assets_t-1	0.01	−0.37***	−0.37***	−0.27	−0.24
	[0.50]	[7.08]	[13.3]	[1.31]	[1.25]
Deposits_t-1/Assets_t-1	0.03*	0.06***	0.06***	0.08*	0.07*
Deposits_t-1/Assets_t-1	[1.75]	[3.06]	[3.91]	[1.89]	[1.93]
Profit BT_t-1/Assets_t-1	0.02*	0.03***	0.02**	0.02	0.02*
Profit BT_t-1/Assets_t-1	[1.63]	[4.12]	[2.38]	[1.43]	[1.65]
Net interest revenue_t-1/Average assets_t-1	−0.00	−0.02***	−0.03**	−0.02***	−0.02***
Net interest revenue_t-1/Average assets_t-1	[0.70]	[2.92]	[2.22]	[7.97]	[5.67]
Cost to income ratio_t-1	0.001*	0.002***	0.002	0.003***	0.003***
Cost to income ratio_t-1	[1.91]	[3.05]	[1.30]	[5.63]	[6.06]
Capital ratio_t-1	0.03	−0.01	−0.03	0.24***	0.25***
	[1.14]	[0.81]	[0.38]	[5.99]	[6.39]
Loan loss reserves_t-1/Gross loans_t-1	−0.02	0.01	0.04	0.06	0.06
Loan loss reserves_t-1/Gross loans_t-1	[1.47]	[0.31]	[1.04]	[0.54]	[0.57]
Listed_t	−0.01
	[1.15]
H-quarters_t	0.01
	[0.31]
Constant	0.08	4.68***	4.88***	3.56	3.27
Constant	[0.41]	[6.62]	[10.6]	[1.26]	[1.21]
No. of obs.	1 681	1 681	1 681	889	889
Adj. R-squared	0.28	0.42	0.37	0.25	0.25
Bank fixed effects	No	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	No	Yes	Yes

Source: Author’s calculations.

Note: ***, ** and * indicate the significance at the 1 per cent, 5 per cent and 10 per cent levels, respectively.

Model 3 in Table 4, in turn, implies that the crisis of 2008–2010 had a significant negative effect on loan growth. However, the interaction of state ownership and the crisis dummy, Crisis × GOV, is positive and significant at the 10 per cent level. An unreported F-test indicates that the hypothesis that the sum of the coefficients on GOV and Crisis × GOV is equal to zero cannot be rejected, suggesting that the general negative effect of state ownership on loan growth may have been cancelled out during the crisis period. Magnitudes of these estimates suggest that state-owned banks decreased their lending by a lesser amount than did private banks in 2008–2010.

Yet, the estimates of Model 5 suggest that banks that were fully owned by the government had a positive impact on loan growth particularly during the crisis period. The magnitude of the estimated coefficients indicates that loans on average grew by one and a half percentage points faster in fully state-owned banks during the episode of financial stress. This argument is further supported in Figure 1(b), which plots mean loan growth rates for different ownership types through the whole sample period. As shown in the figure, the average growth rate of loans significantly decreased in 2008. However, while the growth rates of loans remained close to zero for private banks and state-owned banks maintained a rather nominal growth of about 10 per cent, fully state-owned banks, in turn, substantially increased lending by almost 60 per cent in 2009. Moreover, as loan growth rates began to recover in 2010, fully state-owned banks decreased lending to the market averages. This correlation indicates that fully state-owned banks may actually act counter-cyclically during the crisis episodes as the result of government response to market stresses.

Figure 1.

Source: Author’s calculations.

The estimated coefficients of Model 5 in Table 4 also suggest that smaller, less efficient but more profitable banks with larger deposit base and diverse income structure are positively associated with loan growth rates. Interestingly, Models 4 and 5 suggest that capital ratios of banks matter for the growth rates of loans only during the crisis period. This result is consistent with prior evidence from the developed markets that suggests that the relationship between capital ratios and bank lending is significant only during the financial crisis (refer to, e.g., Carlson, Shan, & Warusawitharanaa, 2013).

Next, the relationship between state ownership of banks and average interest rates charged on loans is examined. The estimation results of Equation (2) with the interest income on loans to average gross loans ratio as the dependent variable are reported in Table 5. Models are specified in a similar manner as in Table 4.

The estimates of Model 1 in Table 5 suggest that average interest rates on loans are negatively associated with the state ownership. Yet, the model also indicates that this relationship may be nonlinear. The estimated coefficients for GOV and GOV ² are not statistically significant in Models 2 and 3. Furthermore, Model 3 provides strong evidence to suggest that average interest rates on loans increased by more than 1.3 percentage points during the crisis period. However, state-owned banks increased rates by a lesser amount than the private banks. Figure 2(a) plots the relationship between state participation in bank’s equity and average interest rates on loans during the crisis of 2008–2010. The figure shows that the difference in interest charged on loans is about 1.25 per cent between partly and fully state-owned banks. While partly state-owned banks charged, on average, about 13.25 per cent, fully state-owned banks required about 12 per cent on their loans during the crisis period.

Table 5.

State Ownership and Interest Rates on Loans

	Dependent Variable: Average interest rate on loans
	1	2	3	4	5
GOV_t	−0.07***	0.01	0.00	0.02
	[6.82]	[0.13]	[0.03]	[1.63]
GOV_t²	0.001***	−0.00	−0.00	−0.001***
	[6.65]	[0.49]	[0.41]	[4.79]
GOV-High_t					−0.06***
					[15.2]
Crisis			1.33***
			[4.09]
Crisis × GOV_t			−0.01*
			[1.75]
Crisis × GOV_t²			0.00
			[0.85]
Assets_t	−0.11**	0.06	−0.25	0.39	0.33
	[1.99]	[0.17]	[1.54]	[0.57]	[0.47]
Deposits_t/Assets_t	0.19	−0.03	0.10	−0.26	−0.27
	[1.38]	[0.14]	[0.40]	[0.71]	[0.81]
Profit BT_t/Assets_t	0.10	0.19**	0.14	0.06	0.04
	[1.32]	[2.07]	[1.46]	[0.98]	[0.60]
Net interest revenue_t/Average assets_t	0.66***	0.72***	0.71***	0.75***	0.76***
Net interest revenue_t/Average assets_t	[8.41]	[15.6]	[30.2]	[14.5]	[15.1]
Cost to income ratio_t	0.01	−0.00	−0.01**	−0.01	−0.01
Cost to income ratio_t	[1.10]	[0.02]	[2.01]	[0.74]	[0.97]
Capital ratio_t	−1.16***	−1.45***	−1.35***	−0.73	−0.73
	[3.66]	[3.96]	[3.03]	[1.15]	[1.13]
Loan loss reserves_t/Gross loans_t	0.09	0.30*	0.35*	0.66***	0.61***
Loan loss reserves_t/Gross loans_t	[1.16]	[1.85]	[1.77]	[9.47]	[8.14]
Growth_t	0.00	−0.00	−0.00	−0.00	−0.00
	[0.44]	[0.74]	[0.63]	[0.99]	[0.92]
Average rate on deposits_t	0.01*	0.01*	0.01*	0.03***	0.03***
Average rate on deposits_t	[1.72]	[1.91]	[1.89]	[3.07]	[2.96]
Listed_t	0.19
	[1.16]
H-quarters_t	−1.42***
	[5.76]
Constant	9.40***	5.86	9.81***	2.99	4.12
	[8.50]	[1.36]	[5.14]	[0.36]	[0.49]
No. of obs.	1 587	1 587	1 587	751	751
Adj. R-squared	0.29	0.64	0.62	0.70	0.71
Bank fixed effects	No	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	No	Yes	Yes

Source: Author’s calculations.

Note: ***, ** and * indicate significance at the 1 per cent, 5 per cent and 10 per cent levels, respectively.

Model 4 in Table 5 confirms these findings and implies that fully state-owned banks charged lower interest rates on their loans during the financial crisis of 2008–2010 as the estimated coefficient for GOV ² is negative and statistically significant at the 1 per cent level. Model 5 supports these findings as the estimated coefficient for GOV-High is negative and significant, implying that the conditional average interest rate on loans for fully state-owned banks is 6 percentage points lower, assuming that bank-specific control variables are fixed. Figure 2(b) illustrates mean interest rates on loans across ownership types. As shown in the figure, private banks increased their interest rates on loans already in 2007 as the response to an average increase in volatility in the global financial markets. As a result, the interest rate on loans by these banks rose from about 14 per cent in 2007 to almost 16 per cent in 2009. State-owned banks, in turn, had lower volatility in interest rates than private banks during the crisis but also marginally increased in 2009. What is more remarkable is that fully state-owned banks were consistently decreasing their interest rates on loans even amidst the financial crisis of 2008–2010. Hence, this evidence suggests a counter-cyclical behaviour of fully state-owned banks not only in lending amounts but also in relative costs of loans.

Figure 2.

Source: Author’s calculations.

As for the control variables in Table 5, as one would expect, with the change in market rates, average rates on loans and rates on deposits are positively related across all models. Similarly, net interest revenue to average asset ratio is positively associated with interest rates on loans. In the full sample period estimations, interest income on loans to average gross loans ratio is negatively associated with the capital ratio, while this significance, however, vanished in the crisis period. Finally, the positive relationship between interest rates on loans and loan loss reserves to gross loans ratio is observed in most of the models, implying that banks charge higher interest rates if the quality of their loan portfolio is worsening. All models have a very good explanatory power with R-squares varying between 29 per cent and 71 per cent.

5.3 State Ownership and Bank Capitalisation

As the next step of the analysis, the association between state ownership and bank capitalisation is examined. Table 6 presents estimation results of Equation (3). The dependent variable now is the logarithm of capital ratio, measured as total equity divided by total assets.

Similar to Tables 4 and 5, Model 1 in Table 6 uses two additional dummy variables, Listed and H-quarters, that capture whether a bank is listed on a stock exchange and whether the headquarters of the bank is located in the capital city. The estimated coefficient for Listed is negative and statistically significant at the 10 per cent level, implying that publicly listed banks may be more risky than non-listed banks in terms of the bank’s capital. In contrast, the estimated coefficient for H-quarters is positive and highly statistically significant, suggesting that banks located in Moscow are better capitalised. The model also shows that the squared term of state ownership is positively related to the capital ratio. However, Models 2, 3 and 4, which also include bank fixed effects, do not support this relationship and indicate that capital ratios are not associated with the state ownership of banks. These results also do not provide strong evidence on the nonlinear relationship between state ownership and bank capital ratios.

Consistent with the univariate tests in Table 3, Model 3 in Table 6 shows that the average capital ratio increased during the financial crisis of 2008–2010 as the estimated coefficient for Crisis is positive and statistically highly significant. Given that the relationship between state ownership and bank capitalisation appears to be linear, Model 5 estimates Equation (3) using a dummy variable that takes the value of 1 if the government held more than 20 per cent of a bank’s equity. The estimates of this specification indicate that state-owned banks, on average, were better capitalised during the crisis period, as the coefficient for the dummy variable is positive and statistically significant at the 5 per cent level. The explanatory power of this model is very high which is 83 per cent, while in Models 1–4, the R-squares vary between 23 per cent and 83 per cent.

Table 6.

State Ownership and Bank Capitalisation

	Dependent Variable: Capital ratio
	1	2	3	4	5
GOV_t	0.00	0.00	0.00	0.01
	[0.64]	[0.71]	[0.85]	[1.57]
GOV_t²	0.00*	0.00	0.00	0.00
	[1.67]	[0.91]	[1.14]	[1.40]
GOV_Dum					0.18**
					[2.32]
Crisis			0.08***
			[4.13]
Crisis × GOV_t			0.00
			[0.38]
Crisis × GOV_t²			0.00
			[0.84]
Assets_t−1	−0.06***	−0.08**	−0.08***	0.02	0.02
	[7.97]	[2.38]	[4.67]	[0.28]	[0.23]
Deposits_t−1/Assets_t−1	−0.12***	−0.02	−0.02	0.04**	0.04***
Deposits_t−1/Assets_t−1	[8.38]	[1.19]	[0.65]	[2.27]	[3.21]
Profit BT_t−1/Assets_t−1	0.06***	0.00	0.01**	0.00	0.00
Profit BT_t−1/Assets_t−1	[7.36]	[0.88]	[1.97]	[0.36]	[0.32]
Net interest revenue_t−1/Average assets_t−1	0.02***	0.00	0.00	0.00	0.00
Net interest revenue_t−1/Average assets_t−1	[4.17]	[0.02]	[0.84]	[0.61]	[0.70]
Cost to income ratio_t−1	−0.002***	−0.001**	−0.001*	−0.001**	−0.001***
Cost to income ratio_t−1	[3.92]	[2.49]	[1.68]	[2.48]	[2.76]
Loan loss reserves_t−1/Gross loans_t−1	0.02	0.06***	0.03***	0.05***	0.05***
Loan loss reserves_t−1/Gross loans_t−1	[1.50]	[6.29]	[2.59]	[3.26]	[3.28]
Growth_t−1	−0.002***	−0.001***	−0.001***	−0.00	−0.00
	[5.63]	[2.85]	[3.88]	[0.99]	[0.95]
Listed_t	−0.02*
	[1.91]
H-quarters_t	0.17***
	[6.10]
Constant	−1.40***	−0.94**	−0.94***	−2.25**	−2.21**
	[40.4]	[2.08]	[4.08]	[2.24]	[2.21]
No. of obs.	1 514	1 514	1 514	842	842
Adj. R-squared	0.23	0.77	0.76	0.83	0.83
Bank fixed effects	No	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	No	Yes	Yes

Source: Author’s calculations.

Note: ***, ** and * indicate significance at the 1 per cent, 5 per cent and 10 per cent levels, respectively.

Regarding the control variables, SIZE, CTI and GROWTH seem to be negatively related to the capital ratio, while profitability and loan loss reserve to gross loans ratio seem to have positive effects on banks’ capitalisation. The coefficients for these variables are statistically significant at the conventional levels in most of the specifications. Finally, being negatively related to the capital ratio in the whole sample period (as suggested by Model 1), the deposits to assets ratio seems to switch its relationship to positive during the crisis period.

5.4 The Effect of State Intervention During the Financial Crisis

During the recent financial crisis, state interventions into the banking sector have increased around the world. States with low levels of state ownership of commercial banks had to nationalise or bailout not only large systemic banks but also smaller ones in order to avoid the collapse of the financial system. Countries with substantial state influence in the banking sector, in contrast, used their ownership stakes in banks for liquidity injections and recapitalisation. While the evidence on the contribution of bailout programmes like Troubled Asset Relief Program (TARP) is rather mixed and does not provide a definitive conclusion (refer to e.g., Bayazitova & Shivdasani, 2012; Black & Hazelwood, 2013; Harris, Huerta, & Ngo, 2014; Veronesi & Zingales, 2010), the effects of government intervention through the state-owned banks remain unknown. This section aims to fill this gap by examining the effect of state support on lending behaviour and bank capitalisation in the highly government-influenced Russian banking sector.

During the financial crisis of 2008–2010 besides quantitative easing, the Russian government provided additional liquidity to the banking sector through preferential subordinated loans. It allocated about RUB 700 billion (about USD 25 billion) to 18 commercial banks and roughly 93 per cent of this sum went to state-owned institutions. Hence, it is reasonable to assume that government response to the crisis in the banking sector has been mainly realised through the state-owned financial institutions. Table 7 presents the estimation results of the effects of state intervention on bank lending behaviour and capitalisation.

Using a dummy variable GOV-SUP as the main explanatory variable which takes the value of 1 if a bank received monetary support from the government during the crisis episode, Models 1–3 of Table 7 re-estimate Equations (1–3) for the sub-sample period of 2008–2011. As can be noted from Model 1, state support had a significant positive effect on loan growth rates. The financial institution that received such support increased lending by about 13 per cent. Model 2, in turn, indicates that these banks also charged significantly lower interest rates on their loans. The estimated coefficient suggests that banks, which received state support, decreased their interest rates on loans by 1.4 percentage points.

Interestingly, the results of Model 3 in Table 7 suggest that supported by the government banks had roughly 7 per cent lower capital ratio. Given that capital ratio may be used as the proxy for bank risk-taking, these results indicate that state-supported banks were enforced to be involved in riskier operations during and right after the crisis period. Assuming that government response to the crisis is rather counter-cyclical, increased lending is most likely to be associated with riskier strategies that affect the level of reserves and consequently the capital ratio. Although the estimated coefficient is only marginally significant at the 10 per cent level, these results are broadly consistent with the recent evidence on the state ownership of banks and risk-taking (Angkinand & Wihlborg, 2010; Dong, Meng, Firth, & Hou, 2014). Political and social aspects of state support of financial institution may force state-controlled banks to take more risk in order to serve as a counter-cyclical tool in harsh times.

Table 7.

The Effects of State Intervention on Bank Lending Behaviour and Capitalisation

	Dependent Variables
	Loans growth	Average interest rate on loans	Capital ratio
	1	2	3
GOV-SUP	0.13***	−1.42***	−0.07*
	[6.40]	[8.70]	[1.78]
Assets	0.01	−0.31***	−0.05***
	[1.34]	[9.23]	[5.02]
Deposits/Assets	0.04*	−0.02	−0.09***
	[1.69]	[0.13]	[9.77]
Profit BT/Assets	0.03***	0.11	0.06***
	[3.07]	[1.46]	[4.56]
Net interest revenue/Average assets	−0.01	0.55***	0.03***
Net interest revenue/Average assets	[0.71]	[5.64]	[9.06]
Cost to income ratio	0.002***	−0.02***	−0.002**
	[2.57]	[3.52]	[2.27]
Capital ratio	0.02	−1.48***
	[0.43]	[3.38]
Loan loss reserves/Gross loans	−0.01	0.18**	0.01
Loan loss reserves/Gross loans	[0.88]	[2.08]	[0.90]
Growth		0.00	−0.002***
		[0.26]	[4.78]
Average rate on deposits		0.01**
Average rate on deposits		[2.48]
Constant	−0.27***	13.9***	−1.41***
	[2.65]	[12.4]	[8.62]
No. of obs.	1 163	989	1 113
Adj. R-squared	0.12	0.22	0.20
Bank fixed effects	No	No	No
Year fixed effects	Yes	Yes	Yes

Source: Author’s calculations.

Note: ***, ** and * indicate significance at the 1 per cent, 5 per cent and 10 per cent levels, respectively.

5.5 Robustness Checks

To ensure the robustness of the results, several additional tests are performed.⁹ First, the main specifications of Equations (1), (2) and (3) are re-estimated using a dummy variable instead of ownership percentages. Given that governments may exert certain influence on the bank’s governance mechanisms even without holding controlling stakes, this variable takes the value of 1 if the government holds more than 20 per cent of the bank’s equity. Similar tests are also performed using a 50 per cent ownership definition. To account for nonlinearity in Equations (1) and (2), this dummy variable was split into two variables for high and low levels of state influence. Specifically, the high level of state influence is defined if the government holds at least 95 per cent of equity, while the low level is defined if less than 50 per cent belongs to the government. All results with dummy variables confirm the main conclusions of the article.

Second, the possible differences between private foreign and domestic banks may drive some of the estimated coefficients for state-owned banks. Therefore, the additional variable for foreign ownership of banks is included in all regression specifications of all models. Similar to the state ownership variable, foreign ownership varies from 20 per cent to 100 per cent. Conclusions made in this article remain unchanged after accounting for the foreign ownership of banks.

Finally, exchange listed banks may potentially have better access to funds which may affect the results. Hence, in addition to the analysis with the dummy variable Listed, which takes the value of 1 if the bank was listed on a stock exchange, Equations (1–3) are re-estimated in the two sub-samples of privately owned and publicly traded banks. There are 54 banks that were listed on stock exchanges during the sample period. The results of these tests indicate that the observed relationship between state ownership and lending behaviour and capitalisation holds for both privately held and publicly traded banks.

6. Conclusions and Policy Implications

While the issue of privatisation of state-owned financial institutions is an on-going concern in many emerging markets, the global financial crisis of 2008–2010 reemphasised the importance of government support of the financial sector in both developed and emerging countries. This article examines the association between different levels of state ownership and lending behaviour and capitalisation of commercial banks. In particular, using data on 348 large banks from the highly concentrated and state-influenced Russian banking sector over the period 2005–2011, the article investigates whether the government ownership of banks has any implications for loan growth rates, interest rates charged on loans and capital ratios.

The empirical findings reported in this article contribute to the existing literature that focuses on the association of state ownership and lending behaviour of banks around the crisis episodes (refer to, e.g., Bertay, Demirgüç-Kunt, & Huizinga, 2015; Brei & Schclarek, 2013, 2015; Cull & Martinez Peria, 2013; Fungáčová et al., 2013) by documenting nonlinearity in the relationship between state ownership and the lending behaviour of banks. In addition, this article is the first to analyse the effect of state intervention through state-owned banks on bank lending behaviour and capitalisation. While commercial lending decreased and average interest rates rose during the crisis of 2008–2010, it is found that state-owned banks decreased lending and increased interest rates by a lesser amount than private banks. Moreover, the results show that fully state-owned banks, in fact, enhanced lending and charged lower interest rates. These findings indicate that fully state-owned banks may serve as a counter-cyclical power during the crisis episodes. Furthermore, state-owned banks were better protected against asset default by having higher capital ratios during the crisis. However, there is a weak evidence to suggest that state intervention forced supported banks to engage in riskier lending which negatively affected their capital ratios. Nevertheless, government support of state-owned banks seemed to fulfil its aim and caused an increase in lending, while it drove the interest rates on loans downwards despite increased riskiness during the crisis. These findings, in turn, provide further evidence on the importance of state ownership of banks during periods of markets stress.

The observed large-scale nationalisation of banks in many countries during the global financial crisis brought the argument of a negative effect of state ownership of banks to a substantial controversy. The results of this article lead to policy implications, which may be especially relevant in times of markets stress. First, they suggest that state-owned banks may serve as a stabilising and counter-cyclical power on the commercial loans market. Instead of mass bailouts of private banks in times of crisis, governments may exploit their participation in equity stakes and make direct liquidity injections through capital increases and preferential loans. Such actions, if implemented properly, may lead to the general stabilisation of the banking sector as a whole.

Second, during crises, uninsured depositors are more likely to run rather than monitor or verify leaked negative information on commercial banks. Even with the sufficient deposit insurance system, privately owned banks are more susceptible to bank runs amidst financial turmoil. The recent example of the Northern Rock—a British bank that suffered a severe bank run and was fully nationalised during the crisis of 2008–2010—supports this statement. In contrast, as a bank’s shareholder, the government may act as an additional guarantee to depositors and thus prevent them from a funds withdrawal.

Third, the existing evidence on the benefits of bank dependence during the crisis periods and the superiority of bank debt to other sources of corporate debt for industrial firms (refer to, e.g., Allen & Paligorova, 2015; Davydov & Vähämaa, 2013) imply that state ownership of banks may have important implications for the real sector as well. If governments act as a balancing source of credit in crises times, firms that have relationships with state-owned banks would potentially have better options for refinancing and renegotiations in case of financial distress.

Finally, these arguments lead to a conclusion that government-owned banks should not necessarily be completely privatised, even though they are less efficient and profitable. Instead, governments may focus on the enhancement of the quality of corporate governance mechanisms and improvement of efficiency of these banks by decreasing political influence. Several recent studies provide empirical evidence to suggest that privatisation of state-owned banks may in fact have harmful effects on financial stability and development and argue that governments should not hurry to privatise their financial sectors (refer to, e.g., Andrianova, 2012; Andrianova, Demetriades, & Shortland, 2008; Karas et al., 2010; Körner & Schnabel, 2011). On the contrary, state ownership of banks may be associated with higher long-run economic growth rates (Andrianova et al., 2012). The empirical findings of this article suggest that the state ownership of banks may be particularly valuable in periods of financial turmoil, when governments exert their interventions through state-controlled banks and provide stability to the whole financial system.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: Financial support from the Finnish Foundation for Economic Education.

Footnotes

Acknowledgements

I am thankful to Andrei Vernikov, Sami Vähämaa, Stanley D. Smith, John Sedunov, Karolin Kirschenmann, Panu Kalmi, Zuzana Fungáčová and participants at the BOFIT research workshop, the Fifth Annual CInST Banking Workshop, the 2015 Southern Finance Association Conference, the 2015 Finnish Economic Association meetings and the 2014 Eastern Finance Association Conference for valuable comments on the previous versions of this article.

Notes

References

Albertazzi

, & Bottero

(2014). Foreign bank lending: Evidence from the global financial crisis. Journal of International Economics, 92(S1), S22–S35.

Allen

, & Paligorova

(2015). Bank loans for private and public firms in a liquidity crunch. Journal of Financial Stability, 18, 106–116.

Andrianova

(2012). Public banks and financial stability. Economics Letters, 116(1), 86–88.

Andrianova

, Demetriades

, & Shortland

(2008). Government ownership of banks, in situations, and financial development. Journal of Development Economics, 85(1–2), 218–252.

Andrianova

, Demetriades

, & Shortland

(2012). Government ownership of banks, institutions, and economic growth. Economica, 79(315), 449–469.

Angkinand

, & Wihlborg

(2010). Deposit insurance coverage, ownership, and banks’ risk-taking in emerging markets. Journal of International Money and Finance, 29(2), 252–274.

Anzoategui

, Martinez Peria

, & Melecky

(2012). Bank competition in Russia: An examination at different levels of aggregation. Emerging Market Review, 13(1), 42–57.

Barry

, Lepetit

, & Strobel

(2016). Bank ownership structure, lending corruption and the regulatory environment. Journal of Comparative Economics, 44(3), 732–751.

Barry

, Lepetit

, & Tarazi

(2011). Ownership structure and risk in publicly held and privately owned banks. Journal of Banking and Finance, 35(5), 1327–1340.

10.

Barth

, Caprio

, & Levine

(1999). Banking systems around the globe: Do regulation and ownership affect performance and stability? (World Bank Policy Research Working Paper, 2325). Washington, DC: The World Bank.

11.

Barth

, Caprio

, & Levine

(2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation, 13(2), 205–248.

12.

Bayazitova

, & Shivdasani

(2012). Assessing TARP. Review of Financial Studies, 25(2), 377–407.

13.

Berger

, Clarke

, Cull

, Klapper

, & Udel

(2005). Corporate governance and bank performance: A joint analysis of the static, selection, and dynamic effects of domestic, foreign, and state ownership. Journal of Banking and Finance, 29(8–9), 2179–2221.

14.

Berger

, Klapper

, Martinez Peria

, & Zaidi

(2008). Bank ownership type and banking relationships. Journal of Financial Intermediation, 17(1), 37–62.

15.

Berger

, Hasan

, & Zhou

(2009). Bank ownership and efficiency in China: What will happen in the world’s largest nation? Journal of Banking and Finance, 33(1), 113–130.

16.

Bertay

, Demirgüç-Kunt

, & Huizinga

(2015). Bank ownership and credit over the business cycle: Is lending by state banks less procyclical? Journal of Banking and Finance, 50, 326–339.

17.

Black

, & Hazelwood

(2013). The effect of TARP on bank risk-taking. Journal of Financial Stability, 9(4), 790–803.

18.

Bonin

, Hasan

, & Wachtel

(2005a). Bank performance, efficiency and ownership in transition countries. Journal of Banking and Finance, 29(1), 31–53.

19.

Bonin

, Hasan

, & Wachtel

(2005b). Privatization matters: Bank efficiency in transition countries. Journal of Banking and Finance, 29(8–9), 2155–2178.

20.

Borisova

, Brockman

, Salas

, & Zagorchev

(2012). Government ownership and corporate governance: Evidence from the EU. Journal of Banking and Finance, 36(11), 2917–2934.

21.

Boubakri

, Cosset

J-C.

, & Guedhami

(2009). From state to private ownership: Issues from strategic industries. Journal of Banking and Finance, 33(2), 367–379.

22.

Boycko

, Shleifer

, & Vishny

(1996). A theory of privatization. Economic Journal, 106(435), 309–319.

23.

Brei

, & Schclarek

(2013). Public bank lending in times of crisis. Journal of Financial Stability, 9(4), 820–830.

24.

Brei

, & Schclarek

(2015). A theoretical model of bank lending: Does ownership matter in times of crisis? Journal of Banking and Finance, 50, 298–307.

25.

Brunnermeier

(2009). Deciphering the liquidity and credit crunch 2007–2008. Journal of Economic Perspectives, 23(1), 77–100.

26.

Cardarelli

, Elekdag

, & Lall

(2011). Financial stress and economic contractions. Journal of Financial Stability, 7(2), 78–97.

27.

Carlson

, Shan

, & Warusawitharanaa

(2013). Capital ratios and bank lending: A matched bank approach. Journal of Financial Intermediation, 22(4), 663–687.

28.

Carvalho

(2014). The real effects of government-owned banks: Evidence from an emerging market. The Journal of Finance, 69(2), 577–609.

29.

Chernykh

(2008). Ultimate ownership and control in Russia. Journal of Financial Economics, 88(1), 169–192.

30.

Cornett

, Guo

, Khaksari

, & Tehranian

(2010). The impact of state ownership on performance differences in privately-owned versus state-owned banks: An international comparison. Journal of Financial Intermediation, 19(1), 74–94.

31.

Cull

, & Martinez Peria

(2013). Bank ownership and lending patterns during the 2008–2009 financial crisis: Evidence from Latin America and Eastern Europe. Journal of Banking and Finance, 37(12), 4861–4878.

32.

Davydov

, & Vähämaa

(2013). Debt source choices and stock market performance of Russian firms during the financial crisis. Emerging Markets Review, 15, 148–159.

33.

De Haas

, Korniyenko

, Loukoianova

, & Pivovarsky

(2012). Foreign banks and the Vienna initiative: Turning sinners into saints (European Bank for Reconstruction and Development (EBRD) Working Paper, 143). London: European Bank for Reconstruction and Development.

34.

De Haas

, Korniyenko

, Pivovarsky

, & Tsankova

(2015). Taming the herd? Foreign banks, Vienna initiative and crisis transmission. Journal of Financial Intermediation, 24(3), 325–355.

35.

Dinc

(2005). Politicians and banks: Political influences on government-owned banks in emerging markets. Journal of Financial Economics, 77(2), 453–479.

36.

Dong

, Meng

, Firth

, & Hou

(2014). Ownership structure and risk-taking: Comparative evidence from private and state-controlled banks in China. International Review of Financial Analysis, 36, 120–130.

37.

Erkens

, Hung

, & Matos

(2012). Corporate governance in the 2007–2008 financial crisis: Evidence from financial institutions worldwide. Journal of Corporate Finance, 18(2), 389–411.

38.

Ferri

, Kalmi

, & Kerola

(2014). Does bank ownership affect lending behavior? Evidence from the Euro area. Journal of Banking and Finance, 48, 194–209.

39.

Fungáčová

, & Poghosyan

(2011). Determinants of bank interest margins in Russia: Does bank ownership matter? Economic Systems, 35(4), 481–495.

40.

Fungáčová

, Solanko

, & Weill

(2010). Market power in the Russian banking industry. International Economics, 124, 127–145.

41.

Fungáčová

, Herrala

, & Weill

(2013). The influence of bank ownership on credit supply: Evidence from the recent financial crisis. Emerging Markets Review, 15, 136–147.

42.

Gur

(2012). Government ownership of banks, job creation opportunities and employment growth. Economics Letters, 117(2), 509–512.

43.

Harris

, Huerta

, & Ngo

(2014). The impact of TARP on bank efficiency. Journal of International Financial Markets, Institutions and Money, 24, 85–104.

44.

Hawkins

, & Mihaljek

(2001). The banking industry in the emerging market economics: Competition, consolidation and systemic stability: An overview (Bank for International Settlements (BIS) Working Paper, 4, pp. 1–44). Basel: BIS.

45.

Iannotta

, Nocera

, & Sironi

(2007). Ownership structure, risk and performance in the European banking industry. Journal of Banking and Finance, 31(7), 2127–2149.

46.

Iannotta

, Nocera

, & Sironi

(2013). The impact of government ownership on bank risk. Journal of Financial Intermediation, 22(2), 152–176.

47.

Infante

, & Piazza

(2014). Political connections and preferential lending at local level: Some evidence from the Italian credit market. Journal of Corporate Finance, 29, 246–262.

48.

Jackowicz

, Kowalewski

, & Kozlowski

(2013). The influence of political factors on commercial banks in Central European countries. Journal of Financial Stability, 9(4), 759–777.

49.

Karas

, Schoors

, & Weill

(2010). Are private banks more efficient than public banks? Evidence from Russia. Economics of Transition, 18(1), 209–244.

50.

Khwaja

, & Mian

(2005). Do lenders favor politically connected firms? Rent provision in an emerging financial market. Quarterly Journal of Economics, 120(4), 1371–1411.

51.

Körner

, & Schnabel

(2011). Public ownership of banks and economic growth. The impact of country heterogeneity. Economics of Transition, 19(3), 407–441.

52.

La Porta

, Lopez-de Silanes

, & Shleifer

(2002). Government ownership of banks. The Journal of Finance, 57(1), 265–301.

53.

Laeven

, & Valencia

(2010). Resolution of banking crises: The good, the bad, and the ugly (IMF Working Paper (10/146)). Washington, DC: International Monetary Fund.

54.

Lainela

, & Ponomarenko

(2012). Russian financial markets and monetary policy instruments (Bank of Finland Institute for Economies in Transition (BOFIT) Discussion Paper 3/2012). Helsinki: Bank of Finland.

55.

Lenza

, Pill

, & Reichlin

(2010). Monetary policy in exceptional times. Economic Policy, 25(62), 295–339.

56.

Megginson

(2005). The economics of bank privatization. Journal of Banking and Finance, 29(8), 1931–1980.

57.

Mian

(2003). Foreign, private domestic, and government banks: New evidence from emerging markets (University of Chicago Working Paper). Chicago, IL: Graduate School of Business.

58.

Micco

, & Panizza

(2006). Bank ownership and lending behavior. Economics Letters, 93(2), 248–254.

59.

Micco

, Panizza

, & Yanez

(2007). Bank ownership and performance. Does politics matter? Journal of Banking and Finance, 31(1), 219–241.

60.

Pennathur

, Subrahmanyam

, & Vishwasrao

(2012). Income diversification and risk: Does ownership matter? An empirical examination of Indian banks. Journal of Banking and Finance, 36(8), 2203–2215.

61.

Sapienza

(2004). The effects of government ownership on bank lending. Journal of Financial Economics, 72(2), 357–384.

62.

Shen

, & Lin

(2012). Why government banks underperform: A political interference view. Journal of Financial Intermediation, 21(2), 181–202.

63.

Shen

, Hasan

, & Lin

(2014). The governments’ role in government-owned banks. Journal of Financial Services Research, 45(3), 307–340.

64.

Shirley

, & Walsh

(2000). Public versus private ownership: The current state of debate (World Bank Policy Research Working Paper, 2420). Washington, DC: The World Bank.

65.

Shleifer

(1998). State versus private ownership. Journal of Economic Perspectives, 12(4), 133–150.

66.

Shleifer

, & Vishny

(1994). Politicians and firms. Quarterly Journal of Economics, 109(4), 995–1025.

67.

The World Bank (2012). Global financial development report 2013: Rethinking the role of the state in finance. Washington, DC: The World Bank.

68.

Vernikov

(2009). Russian banking: The state makes a comeback? (Bank of Finland Institute for Economies in Transition (BOFIT) Discussion Paper 24/09). Helsinki: Bank of Finland.

69.

Veronesi

, & Zingales

(2010). Paulson’s gift. Journal of Financial Economics, 97(3), 339–368.

Panel A. Full sample period 2005–2011
	State	Private
Loans growth
Mean	35.00%	34.92%
Median	21.17%	17.32%
No. of obs.	130	1 701
Capital ratio
Mean	15.61%	15.87%
Median	12.21%	12.92%
No. of obs.	147	2 005
Average interest rate on loans
Mean	12.98%***	14.32%
Median	12.92%***	14.13%
No. of obs.	144	1 759

Panel B. Crisis period
	2008		2009		2010
	State	Private	State	Private	State	Private
Loans growth
Mean	7.94%	−1.04%	11.57%**	−4.92%	11.18%	24.34%
Median	7.29%*	−4.53%	−3.07%	−6.38%	12.00%	16.27%
No. of obs.	20	314	22	290	25	287
Capital ratio
Mean	13.19%	16.35%	18.61%	17.02%	18.43%	15.76%
Median	10.94%	13.65%	14.17%	14.18%	12.80%	12.79%
No. of obs.	19	310	22	291	24	285
Average interest rate on loans
Mean	13.41%**	15.08%	14.10%**	15.59%	12.08%***	14.19%
Median	13.16%**	14.71%	13.96%***	15.51%	11.55%***	13.18%
No. of obs.	20	294	22	279	25	276