Scholarly article on topic 'The Third Pillar of the Basel Accord: Evidence of borrower discipline in the Kyrgyz banking system'

The Third Pillar of the Basel Accord: Evidence of borrower discipline in the Kyrgyz banking system Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Eurasian Studies
OECD Field of science
Keywords
{"Basel accord" / "market discipline" / "loan market" / "bank risk" / Kyrgyzstan}

Abstract of research paper on Economics and business, author of scientific article — Edgar Demetrio Tovar-García, Ruslana Kozubekova

Abstract We empirically study the asset side of market discipline in the banking system of the Kyrgyz Republic, examining whether borrowers are willing to pay higher interest rates to high-quality banks. Based on dynamic panel models and a dataset with bank information from 23 banks over the period 2010–2012, our findings suggest the presence of market discipline induced by borrowers. In other words, banks with higher capital ratios and liquidity charge higher interest rates on loans. This result has several implications for the banking policy in Kyrgyzstan, where we can recommend to policymakers a disclosure policy following the Third Pillar of Basel III, because not only can the bank's creditors use bank information to penalize the excessive bank risk, but borrowers can also use this information to discipline their banks.

Academic research paper on topic "The Third Pillar of the Basel Accord: Evidence of borrower discipline in the Kyrgyz banking system"

2222221324506

ARTICLE IN PRESS

journal of Eurasian Studies ■■ (2016) I

Contents lists available at ScienceDirect

Journal of Eurasian Studies

journal homepage: www.elsevier.com/locate/euras

The Third Pillar of the Basel Accord: Evidence of borrower discipline in the Kyrgyz banking system

Edgar Demetrio Tovar-Garcia, Ruslana Kozubekova

National Research University Higher School of Economics, Myasnitskaya Street 20, Moscow 101000, Russia

ARTICLE INFO ABSTRACT

We empirically study the asset side of market discipline in the banking system of the Kyrgyz Republic, examining whether borrowers are willing to pay higher interest rates to high-quality banks. Based on dynamic panel models and a dataset with bank information from 23 banks over the period 2010-2012, our findings suggest the presence of market discipline induced by borrowers. In other words, banks with higher capital ratios and liquidity charge higher interest rates on loans. This result has several implications for the banking policy in Kyrgyzstan, where we can recommend to policymakers a disclosure policy following the Third Pillar of Basel III, because not only can the bank's creditors use bank information to penalize the excessive bank risk, but borrowers can also use this information to discipline their banks.

Copyright © 2016 Production and hosting by Elsevier Ltd on behalf of Asia-Pacific

Research Center, Hanyang University.

Article history:

Received 10 December 2014 Accepted 19 February 2016 Available online

Keywords: Basel accord market discipline loan market bank risk Kyrgyzstan

1. Motivation

Kyrgyzstan is one of the poorest post-Soviet countries, but its commercial banks have been developing rapidly in the last decade, increasing banking competition, yet with several problems of stability (National Bank of the Kyrgyz Republic, 2013a). In 2010, jointly with the global financial crisis, several bank failures provoked significant losses for the Kyrgyz economy. Evidently, the transition to a market economy is a complicated task.

In early April 2010, anti-government political demonstrations occurred in various cities of the Kyrgyz Republic. These protests turned into riots, resulting in loss of life and material damage (Asian Development Bank et al., 2010). The global economic crisis as well as the domestic political crisis significantly reduced the banking solvency. The uncertainty and insecurity associated with these crises led people to increase their cash holdings, negatively impacting banking

Corresponding author: National Research University Higher School of Economics, Myasnitskaya Street 20, Moscow 101000, Russia. E-mail address:etovar@hse.ru (E.D. Tovar-Garcia).

deposits, and credit to the private sector (National Bank of the Kyrgyz Republic, 2011, 2013a, 2013b).

The National Bank of the Kyrgyz Republic (NBKR)1 made attempts to rescue the banking sector, but the failure of one of the largest banks, Asia Universal Bank (AUB), was inevitable, causing a domino effect. Actually, AUB was audited because of suspicious transactions, principally, offenses related to money laundering and corruption. As a result, other three banks (Investbank Issyk-Kul, Manas Bank, and Kyrgyz Credit Bank) were put under temporary closing-down.

In this context, the Soviet past and little experience in the market economy give Kyrgyz banks a particular relevance to study the transition to international standards of accounting, reporting, and regulation (Ruziev & Majidov,

1 This is the official name of the Central Bank. Under perestroika, the monobank system was replaced with a two-tier banking system. In February 1992, the state bank (Gosbank USSR) was renamed the National Bank of the Kyrgyz Republic (NBKR), with responsibilities for safeguarding the payments system, providing liquidity, licensing and supervising second-tier banks (Brown et al., 2009; Ruziev & Majidov, 2013).

http://dx.doi.org/10.1016/j.euras.2016.02.002

1879-3665/Copyright © 2016 Production and hosting by Elsevier Ltd on behalf of Asia-Pacific Research Center, Hanyang University.

ARTICLE IN PRESS

E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■ ■■

2013), where, since 1988, the Basel Committee has played the most important role in banking policy. In 2004, Basel II introduced three basic pillars: (1) Minimum capital requirements, (2) Supervisory review process, and (3) Market discipline (Basel Committee on Banking Supervision, 2006).

The objective of market discipline is to complement the first and second pillars of banking regulation and supervision. Moreover, the new approach in Basel III situates the relevance of market discipline on the same level as governmental supervision, and its main function is to facilitate market punishment for excessive risk. Consequently, the Basel Committee proposes disclosure and transparency policies to equip market participants with all needed information for decision-making, and to combat suspicious financial transactions, money laundering, and the financing of terrorism.

In economic terms, market discipline weakens asymmetric information and moral hazard concerns. It can be understood as the behavior of market participants, and can be easily analyzed using the supply and demand model. For instance, in the deposit market, depositors will respond to riskier behavior of their banks demanding higher interest rates on deposits (price mechanism of market discipline) or/and withdrawing their deposits (quantity mechanism of market discipline). As a result, the banks should moderate their risk-taking (Tovar-Garcia, 2014). Other bank creditors should analogously react to bank risk, mainly subordinated debt holders, because they are the last in line to recover their financial assets in case of bankruptcy (Evanoff, Jagtiani, & Nakata, 2011; Tovar-Garcia, 2015b). The interbank market has been grown considerably during the last three decades, and there is also evidence for peer monitoring, where lending banks monitor and charge higher rates to low-quality banks (Distinguin, Kouassi, & Tarazi, 2013; Tovar-Garcia, 2015a).

The market discipline effect induced by depositors, subordinated debt holders, and other banks has already been widely examined. This hypothesis has been tested in China (Wu & Bowe, 2012), in ex-socialist countries including Estonia, Latvia, and Lithuania (Hasan, Jackowicz, Kowalewski, & Kozlowski, 2013), several times in Russia (Karas, Pyle, & Schoors, 2010, 2013; Peresetsky, 2008; Semenova, 2007; Tovar-Garcia, 2013; Ungan, Caner, & Ozyildirim, 2008), but it has not been tested in ex-Soviet Asian countries.

Conversely, there is comparatively little attention paid to the asset side of market discipline. There are theoretical and empirical findings suggesting that borrowers also punish their banks because of excessive risk-taking. In a theoretical model, Allen, Carletti, and Marquez (2011, p. 984) state that "when credit markets are competitive, market discipline coming from the asset side induces banks to hold positive levels of capital as a way to commit to monitor and attract borrowers". Thus, banks have incentives to achieve high levels of capital, and in many cases bank capital ratios are higher than the minimal requirement suggested by the Basel Accord.

In the case of Norway, Kim, Kristiansen, and Vale (2005, p. 682) point out that "banks face market discipline induced by borrowers", because they are willing to pay higher rates on loans from high-quality banks for two major reasons:

First, they have a certification motive, where borrowers prefer banks with high-quality loan portfolios to signal their creditworthiness to other stakeholders. Second, borrowers have a refinancing motive, where they choose solvent banks because these banks are able to extend credit lines or new loans in the future. Similarly, in the Mexican case, Tovar-Garcia (2012) found evidence in favor of the certification and refinancing motives, but the largest Mexican banks can avoid this kind of market discipline. In Russia, the interest rates on loans discriminates between high- and low-quality banks, where high-quality banks charge higher interest rates in accordance with the market discipline hypothesis from the asset side (Tovar-Garcia, 2013).

Thus, the Basel Committee proposes market discipline as a key instrument to strengthen the stability and effectiveness of the banking sector. However, in several developing countries the evidence in favor of the market discipline hypothesis is weak, and the stability of the banking system is often the responsibility of monetary authorities, with problems of corruption, lack of experience, and lack of technology (Calomiris, 1999; Tovar-Garcia, 2014). In the case of ex-Soviet countries, arguably, market discipline should not fundamentally work because of deficient markets (Matovnikov, 2012). Consequently, the Kyrgyz Republic is timidly following the recommendations of Basel III, including the disclosure policy to support market discipline (Barth, Caprio, & Levine, 2013).

Given this, the objective of this research is to test the existence of market discipline from the asset side in the Kyrgyz banking system and to analyze its implications for banking policy. Accordingly, this research is focused on the following question: In Kyrgyzstan, do borrowers pay higher interest rates to high-quality banks? Based on previous studies, we hypothesize that Kyrgyz borrowers pay higher interest rates on loans to high-quality banks (low-risk banks) because they prefer solvent banks with high capital ratios and low loan losses; that is, borrowers have a refinancing motive (to ensure future credit lines and new loans) and a certification motive (to signal their creditworthiness to other stakeholders).

We test these hypotheses using a sample of 23 commercial banks of the Kyrgyz Republic and quarterly data from January 2010 to December 2012. Based on dynamic panel models, our findings suggest the presence of market discipline induced by borrowers. This has several implications for banking policy in Kyrgyzstan, where we can recommend a disclosure policy following the Third Pillar of Basel III. The rest of the paper proceeds as follows. Section 2 describes the Kyrgyz banking system, and the data used in this study. Section 3 outlines the empirical strategy and presents the results. Section 4 concludes.

2. The Kyrgyz banking system

In the 1990s and in the beginning of the 2000s, several reforms supported the establishment of a two-tier banking sector, foreign bank entry, financial liberalization, and privatization, intended to increase the size, stability, and efficiency of the Kyrgyz banking sector. The result was a reduction in interest rates on loans and deposits (to support economic

ARTICLE IN PRESS

E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■-■■ 3

100,000.0 80,000.0 60,000.0 40,000.0 20,000.0

222222222222

22.00%

20.00%

ITotal assets = Total loans ■ Capital

-Specific reserve for loan losses

mmmmmmmmmmmm

Fig. 1. Kyrgyz banking system. Major indicators 2010-2012 (%, million soms).

growth),2 although there were no reductions in intermediation spreads (Brown, Maurer, Pak, & Tynaev, 2009).

Several Kyrgyz state and private banks failed before and after the Russian crisis in 1998. The number of banks has varied from 19 to 23 during the period 1997-2008, with three foreign-owned banks in 1997, and 10 in 2008 (Kazakh banks acquired several Kyrgyz banks). The asset share of state-owned banks has diminished, with the maximum share of 25.8% in 1999, and the minimum of 3.4% in 2006 (Ruziev & Majidov, 2013).

In comparison to other countries of the region, the effects of the global financial crisis on Kyrgyzstan were stronger (Ruziev & Majidov, 2013). In addition, in 2010 the widest political events in the modern history of the country occurred (the Peacock Revolution),3 which were accompanied by mass protests against the authoritarian regime (Murzaeva & Ak^ali, 2013; Zabortseva, 2012). As a result, the president and government were removed, and many bankers suffered because they were related to the regime.

In 2010, the main financial indicators of the largest bank, Asia Universal Bank (AUB), announced its failure. This caused negative impacts on other banks, borrowers, depositors, and other debt holders. In general, all Kyrgyz banks suffered enormous losses and lack of liquidity, seriously affecting investments and trade (Asian Development Bank et al., 2010).

Thus, the main impact of the global financial crisis and the domestic political events in 2010 in the banking sector was the bankruptcy of AUB, on October 27. The monetary authorities unsuccessfully explored options for a bank bailout, but the results show that there are banks too-big-to-save (Demirgu^-Kunt & Huizinga, 2013).

Fortunately, for small depositors, since August 2008 the Kyrgyz Republic has a deposit insurance system, provided by the Deposit Insurance Agency in cooperation with the Kyrgyz banks. Within 15 calendar days, the Deposit Insur-

2 In the 1990s the average real GDP growth rate was -3.4% (with a minimal value in 1994, -20.1%), and in the 2000s the average was 4.3% (with a maximum in 2007, 8.5%). In 2010, 2011, and 2012 the rates were -0.5%, 5.9%, and -0.9%, respectively (data from World Development Indicators, World Bank). Kyrgyzstan surpassed its pre-transition GDP level in 2008 (Ruziev & Majidov, 2013).

3 In March 2005, other political turmoil (the Tulip Revolution) affected the Kyrgyz economy. In that year, the real GDP growth rate was -0.2%, with a fast recovery of the economy in the following years.

ance Agency reviews the bank default, and the payment of compensation should begin no later than a period of 60 calendar days after the bank's default (each depositor's compensation pays no more than 100,000 soms).4 In Kyr-gyzstan, the participation in the deposit insurance system is mandatory for all commercial banks operating in the country (Jogorku Kenesh of the Kyrgyz Republic, 2008).

Later on, as shown in Figure 1, the banking sector recovered the path of growth. In the fourth quarter of 2012, the total assets of the banking system amounted to 81.8 billion of soms, an increase of 22.7% on the same period of the previous year. The main share of the banks' assets refers to loans and interbank accounts, 45.2%.

This growth of the industry was accompanied by bank concentration. In 2010, four of the largest banks accounted for 39.1% of the total assets in the banking sector, and they accounted for 55.7% in December 2012. In this year, considering the structure of the resource base, these four banks held 37.1% of the capital in the banking sector.

In the fourth quarter of 2012, the total loan portfolio of the banking system was 37.02 billion soms, which increased from the beginning of the third quarter of 2012 by 2.69%, presenting a positive trend during the period 20102012. The relative increase of the share of loans in total assets is stable, and there are no sharp fluctuations.

This growth of loans should not happen at the expense of reducing the requirements for borrowers, which could lead to an imbalance of the market, and increase the risks of non-repayment of loans. Because excessive credit growth in the economy is systemically important, this rapid credit growth should be accompanied by quality in the credit portfolio. In the Kyrgyz case, it seems that this loan growth is a response to lower interest rates on loans, which present a negative trend during the period 2010-2012. The weighted average interest rate on loans5 decreased from 2010, although in December 2012 the average interest rate was

4 Som is the national currency of the Kyrgyz Republic, introduced on 10 May 1993. On 1 March 2014 the exchange rate to US dollar was 53.9.

5 The weight interest rate is used in determining the total cost of credits,

and takes into account not only the quantity of credit accounts, but also their volumes. It is calculated as the sum of multiplying the interest rate for each credit on the credit amount, and then dividing by the total amount of all credits.

ARTICLE IN PRESS

E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■ ■■

22.9%, an increase of 1.12 percentage points on the previous quarter.

2.1. Data and summary statistics

In 2010, there were 22 commercial banks operating in the Kyrgyz Republic, for the current situation there are 23 commercial banks. Among them 4 banks are considered large, 11 banks are small, and the other 8 are medium-sized, according to the National Bank of the Kyrgyz Republic (NBKR).

To test the asset side of market discipline, the database of this research includes financial balance sheets and reports on profits and losses from Kyrgyz commercial banks covering the period from January 2010 to December 2012 on quarterly basis. Because of data limitations, we cannot compare the pre-crisis with the post-crisis period. Our sample includes data from 23 commercial banks (see Table 1), but due to data limitations a maximum of 22 banks are included in the regression analysis (the observations of AUB are excluded). Our dataset is built using two main sources: banks' web sites and reports and web-site of the NBKR.

Tables 1 and 2 present descriptive statistics on bank variables, where it is possible to see differences and variability of the main variables, intra bank and between banks. The above-mentioned weighted average interest rate on loans becomes the dependent variable of the regression analysis in the next section. There are a few banks presenting a high variability (standard deviation) on interest rates (EcoIslamicBank, Kazkommertsbank Kyrgyzstan, and National Bank of Pakistan Bishkek branch) and other few banks with a low variability (Dos-Credobank, Halyk Bank Kyrgyz-stan, Tolubay, and Zalkar). Thus, from these tables, it is easy to note that there are relevant differences in interest rates across banks during the period of analysis.

The independent variables are: the ratio capital to total assets (CAPITALR), specific reserves for loan losses divided by nonperforming loans (RESERVE), doubtful loans divided by total loans (DOUBTFUL), the logarithm of total assets (SIZE), net income divided by total assets (ROA), net income divided by capital (ROE), and high liquid assets divided by the short-term liabilities (LIQUIDITY).6 These kinds of indicators are known as bank fundamentals, and they are widely used to approach bank risk.

As shown in Table 1, there are relevant differences in bank fundamentals across banks. For instance, the capital ratio is very low for Zalkar bank (0.02); moreover, this bank also presents serious concerns in indicators about profitability, with losses during the period of analysis (ROA equals -0.001 and ROE equals -0.77), suggesting that this is a low quality bank. However, Zalkar is a medium-sized bank and has high reserves for loan losses (0.54), and a good level of liquidity (1.25), making this bank attractive for some market participants. Other interesting example, with different characteristics, is Kyrgyz Credit Bank, which presents the highest

6 High liquid assets are cash, interbank deposits and credits, and securities of Central banks and the Government of the Kyrgyz Republic, and short-term liabilities are the liabilities with the payment within 30 days.

capital ratio (0.76) and high liquidity (6.13), but with a large variability (standard deviation equals 4.37). However, this is a small bank, with very low levels of profitability (ROA equals 0.004 and ROE equals 0.001).

As shown in Table 2, the overall mean of CAPITALR equals 0.29, the minimum is -0.46, and the maximum is 0.99. The negative value is presented by AUB, which failed at the end of the third quarter of the year 2010. Note that during April and June 2010 the political and economic crises led to significant reductions of total assets and capital, and this bank was recognized to be insolvent in accordance with the court decisions.7

The overall mean of RESERVE is 0.44 points, the minimum is 0.001 and the maximum is 0.96. This ratio shows the extent of the risk in credit operations. High ratios reflect an adequate level of reserves for possible loan losses, and better protection of the bank from damages. Higher values of RESERVE indicate lower bank risk.

The mean of DOUBTFUL is 0.05, and the maximum is 0.61. Higher levels of DOUBTFUL indicate higher credit risks. In other words, it indicates a higher probability of financial losses due to default on its obligations under the loan. The highest value (0.61) belongs to Manas Bank, in the second quarter of the year 2011. This value is a result of a criminal case (Sytenkova, 2013).

The minimal value of return on assets (ROA) is -0.97, and the maximum is 0.06. The negative value corresponds to the third quarter of the year 2010, presented by AUB. The minimal value of return on capital (ROE) is -7.61, and the maximum is 2.50. These indicators show profitability on one unit of assets and capital, respectively. If the level of the ratio is high, the financial position of the bank is better.

The average value of LIQUIDITY is 3.25, the minimal value is 0.29, and the maximum is 148.22. The minimal value belongs to AUB, in the third quarter of the year 2010. This ratio shows the ability to fulfill all obligations to clients in time. Higher ratios indicate higher levels of solvency.

Table 3 presents a correlation matrix with the major variables used in this research. We can see strong correlations between several pairs: loans and deposits (0.72), total assets and loans (0.85), total assets and deposits (0.94), loans and capital (0.73), total assets and capital (0.73), and between deposits and capital (0.57).

The high correlation between loans and deposits is explained by the fact that high volumes of deposits provide the financial resource base for the bank lending. The high correlation between loans and total assets reflects the obvious participation of loans in the total assets. The correlation between total assets and deposits appears because the bank is attracting deposits, to increase the volume of active operations, such as lending and operations with securities. The correlation between loans and capital is explained by the fact that general provisions are created for nonperforming loans, which are included in the total regulatory capital. Operational functions of the bank capital provide an adequate basis for the growth of bank assets; this explains the high correlation between total assets and

7 The observations with negative values were not included in the regression analysis, to avoid outliers.

Table 1

Kyrgyz banks and major banking indicators.

CAPlTALRb RESERVE1 DOUBTFULb ROAb ROEb LIQUIDITY15

(692232) 0.33 (0.05) 0.56(0.12) 0.002 (0.001) 0.02 (0.009) 0.06 (0.03) 1.18(0.29)

(37021) 0.90 (0.07) 0.008 (.01) n.a. -0.08 (0.09) -0.10(0.11) 45.8 (54.4)

(446212) 0.19(0.02) 0.35(0.10) 0.02 0.02(0.01) 0.08 (0.06) 0.90 (0.23)

(0.01)

(1.22e+07) -0.21 (0.28) 0.32(0.13) 0.09 -0.56 (0.50) 1.54(1.33) 0.51 (0.31)

(0.07)

(211278) 0.22 (0.03) 0.26 (0.05) 0.04 0.02 (0.02) 0.09 (0.08) 1.04(0.15)

(0.03)

(248839) 0.30 (0.06) 0.19 (.08) 0.006 (0.003) 0.02 (0.01) 0.08 (0.05) 1.21 (0.18)

(282682) 0.34(0.01) 0.24(0.01) 0.07 0.01 (0.008) 0.04 (0.02) 1.89 (0.69)

(0.006)

(864303) 0.14(0.02) 0.52 (0.09) 0.02 0.02(0.01) 0.14 (0.08) 0.57 (0.07)

(0.005)

(1865523) 0.11 (0.02) 0.53(0.17) 0.006 (0.004) 0.01 (0.009) 0.13 (0.09) 0.53 (0.07)

(302639) 0.29 (0.07) 0.18(0.12) 0.03 0.01 (0.01) 0.05 (0.06) 0.92 (0.09)

(0.07)

(646132) 0.18(0.05) 0.34 (0.04) 0.04 0.003 (0.01) 0.02 (0.06) 0.92 (0.16)

(0.02)

(250463) 0.21 (0.03) 0.26 (0.05) 0.01 -0.006 (0.02) -0.03 (0.10) 0.67(0.11)

(0.007)

(288420) 0.37 (0.05) 0.35 (0.04) 0.11 0.02 (0.007) 0.04 (0.02) 1.01 (0.20)

(0.02)

(326748) 0.47(0.16) 0.73 (0.25) 0.11 -0.08 (0.18) -0.22 (0.52) 1.51 (1.45)

(0.05)

(244410) 0.38 (0.05) 0.86 (0.07) 0.01 0.02 (0.009) 0.04 (0.02) 1.05 (0.15)

(0.01)

(2351631) 0.22 (0.02) 0.29 (0.05) 0.01 0.02(0.01) 0.08 (0.06) 0.90 (0.07)

(0.005)

(240580) 0.76(0.19) 0.63 (0.08) 0.06 0.004(0.03) 0.001 (0.05) 6.13 (4.37)

(0.05)

(554439) 0.20 (0.04) 0.47 (0.08) 0.42 -0.04 (0.04) -0.17(0.19) 0.55 (0.22)

(0.21)

(2666159) 0.15(0.05) 0.55(0.10) 0.005 (0.004) 0.01 (0.008) 0.06 (0.05) 0.80 (0.17)

(158249) 0.20(0.01) 0.64(0.19) 0.0003 (0.0004) 0.01 (0.009) 0.07 (0.04) 1.31 (0.27)

(2316684) 0.15(0.03) 0.47 (0.08) 0.03 0.03 (0.02) 0.21 (0.11) 0.63 (0.10)

(0.01)

(360477) 0.28 (0.05) 0.46(0.14) 0.01 0.02(0.01) 0.06 (0.03) 0.85 (0.06)

(0.02)

(203893) 0.02 (0.02) 0.54(0.14) 0.08 -0.001 (0.006) -0.77 (2.60) 1.25 (0.43)

(0.02)

fi g o356

Interest rate on loans (%)

Deposits3

Loan growthb

Capital3

Total assets3

£ 3 e

1 m ff g u

£a c5]65l a §366

) 23 kg>

401 403

■404

408 s409

Aiyl Bank 19.1 (1.88)

Akyl 24.2 (0.43)

Amanbank 27.2 (0.76)

Asia Universal 23.1 (0.37)

Bank (AUB)*

Bakai Bank 21.4(1.28)

Bank Asia 22.8 (2.62)

BTA Bank 24.5 (1.06)

Commercial 25.6 (2.62)

Kyrgyzstan

DKIB 22.8 (0.91)

Dos-Credobank 25.4 (0.36)

EcolslamicBank 29.2 (7.1)

FCB KAB 28.9 (3.13)

Halyk Bank 21.1 (0.30)

Kyrgyzstan

Investbank Issyk- 23.6 (1.02)

Kazkommertsbank 8.8

Kyrgyzstan (5.1)

KICB 23.0(2.71)

Kyrgyz Credit 20.6 (0.88)

Manas Bank 19.8 (0.88)

SSC Bank 19.4 (3.7)

Tolubay 27.4(0.35)

UnicreditBank 23.6 (0.54)

National Bank of 14.7 (7.6)

Pakistan

Bishkek branch

Zalkar 22.6 (0.31)

3001238(343497) 28213(26083)

1089841(262345)

3435015(334611)

495827(109459)

383536(113108) 1846737(118584)

484177(387018) 4888 (5716) 1089747(241629)

8849537(8251432)

943334(172466)

385251(127023) 314156(123561)

1.02 (0.05) 0.67 (0.26)

1.06 (0.06)

0.91 (0.00)

1.05 (0.13)

1.08 (0.07)

1.01 (0.03)

1174400(67726) 94316(38407)

388027(53512)

-380792(2609656)

305810(35547)

228731 (37257) 1150118(82867)

2237901 (478009) 2629290(676925) 1.06(0.07) 562691 (77936)

1624869(728902) 395579 (181776)

848148 (215627)

605437(155653)

1293644(198567)

207951(73093)

312581(184381)

3848416(1140936)

280724(110747)

369072(151844)

2644570 (897730) 586212 (107272) 5800953 (933750)

196388(19096)

(1462732) (213172) 1.13 (0.06) 1.10(0.15)

! (511774) 1.06 (0.10)

'(211497) 1.02 (0.18)

(47767) 0.97 (0.06)

'(111717) 0.93 (0.08)

(59503) 0.95 (0.23)

(1492569) 1.07 (0.07)

(94401) 0.90(0.03)

1(231754) 0.91 (0.09)

1(1491476) (145816) (1638254) 1.13 (0.14) 1.05 (0.06) 1.04 (0.04)

(320663) 1.01 (0.08)

(80274) 0.97 (0.15)

731799(154994) 296314(19824)

386403 (24087)

270678(18731)

875632(21706)

161181(85853)

634608(11261)

1600192(403205)

409820(18360)

178804(104915)

1207503(279581) 195062 (34533) 1467477 (239037)

359678 (55767)

Mean of the variable over the period 2010q1-2012q4. * For AUB, mean from 2010q1 to 2010q3. The standard deviations are in parentheses. aMillion soms.

bRatio, n.a. = not available.Source: Authors' calculations.

ARTICLE IN PRESS

6 E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■ ■■

Table 2

Summary statistics.

Variable Obs Mean Std. Dev. Minimum Maximum

Interest rate on loans (%) 252 22.91 4.90 1.64 40.00

Loansa 264 1367.81 1482.72 0.30 6873.97

Deposits3 264 1508.02 2015.36 0.21 18400.00

Loan growthb 241 1.01 0.15 0.26 1.42

Capitala 264 573.87 530.83 -2168.94 2613.81

Total assetsa 264 3023.01 3249.53 75.863 26300.00

CAPITALRb 263 0.29 0.22 -0.46 0.99

RESERVEb 249 0.44 0.22 0.001 0.96

DOUBTFULb 264 0.05 0.10 0.00005 0.61

ROAb 264 -0.003 0.09 -0.97 0.06

ROEb 264 0.03 0.54 -7.61 2.50

LIQUIDITY15 264 3.25 14.58 0.29 148.22

aMillion soms.

bRatio.Source: Authors' calculations.

capital. Finally, the high correlation between deposits and capital is explained by the protection function of the capital, as a buffer. In the case of bankruptcy, it becomes a source of payments to creditors and depositors; that is, the banks increase the capital in proportion to attracted deposits (Lavrushin, 2009).

It is interesting to note that there are no high correlations among bank fundamentals. Consequently, we can exclude multicollinearity concerns in the regression analysis, but theoretically we can expect correlation between RESERVE and DOUBTFUL because both variables approach asset quality, and between ROA and ROE, both indicate bank profitability. Therefore, in the regression analysis we include one or the other, to ensure the absence of multicollinear-ity, and to check robustness to different indicators.

3. Empirical strategy

In the empirical literature is possible to identify different econometric methods to test the market discipline hypothesis. It is important to recognize that the variables under study present endogeneity concerns (reverse causality), affecting the estimations of least squares, if we are not using instrumental variables. Accordingly, Tovar-Garcia (2012) proposes dynamic panel models using the

generalized method of moments (GMM), based on Blundell and Bond (1998), known as the SYS GMM estimator.

In comparison with least squares, the GMM estimator, developed by Lars Peter Hansen, is more flexible, because it requires more simple assumptions about moment conditions; consequently, this method became popular in the empirical literature (Chaussé, 2005).

In presence of endogeneity, and autoregressive qualities of the dependent variables, as our case, Arellano and Bond (1991) developed a method to be employed with panel data, known as the DIF GMM estimator. It transforms the variables in first-differences, uses lags of the independent variables as instruments, and allows lags of the dependent variable to be entered as regressors, assuming the absence of serial correlation in the errors.

Blundell and Bond (1998) argue that the first-differenced standard GMM estimator may cause biases in finite-sample, which provide weak instruments. Consequently, they use an extension of the DIF GMM model, employing lagged differences and lagged levels, to be used as instrumental variables in the equation in first-differences. The validity of the instruments can be tested using the Sargan test of over-identifying restrictions, and in absence of serial correlation (particularly of second order), the estimator ensures consistency and efficiency.

Table 3

Correlation matrix.

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

Interest rate on loans (1) 1

Loans(2) -0.03 1

Deposits (3) 0.004 0.72 1

Loan growth (4) 0.14 0.23 0.29 1

Capital (5) -0.20 0.73 0.57 0.22 1

Total assets (6) -0.09 0.85 0.94 0.29 0.73 1

CAPITALR (7) -0.15 -0.40 -0.44 -0.49 -0.08 -0.44 1

RESERVE (8) -0.38 -0.06 -0.04 -0.16 -0.01 -0.01 0.16 1

DOUBTFUL (9) -0. 12 -0.16 -0.16 -0.18 -0.21 -0.18 -0.05 0.06 1

ROA (10) -0.02 0.05 0.02 0.17 0.47 0.08 0.08 0.06 -0.17 1

ROE(11) 0.02 0.15 0.15 0.09 -0.01 0.11 -0.09 -0.05 -0.08 -0.14 1

LIQUIDITY (12) 0.04 -0.15 -0.13 -0.49 -0.13 -0.14 0.49 0.06 -0.07 -0.02 -0.02

Source: Authors' calculations.

555444213

555444435

555444657

ARTICLE IN PRESS

E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) I

Table 4

Regression results.

Dependent variable: Weighted average interest rate on loans (RATE)

PredSign SYS GMM

DIFGMM

Lagged dependent as regressor

CAPITALR

RESERVE

DOUBTFUL

LIQUIDITY Year 2011 (dummy) Year 2012 (dummy) Period

Observations N x T

Sargan test (p-value) First order serial correlation test (p-value)

Second order serial correlation test (p-value)

0.47*** 0.14*** 0.004

0.09*** 0.25

0.52*** 0.16***

0.09* -0.13

0.04*** 0.03*** -0.01* -0.01

-0.01*

2010Q1-2012Q4

218 22 x 11 7.47 (1.00) -1.03 (0.30)

1.07 (0.28)

215 21 x 11 8.64 (0.99) -1.23 (0.22)

1.09 (0.27)

0.50*** 0.13*** 0.01

0.09***

0.03 0.04*** -0.01 -0.01

218 22 x 11 6.63 (1.00) -1.09 (0.27)

1.08 (0.27)

0.50*** 0.14***

-0.0003 0.09***

0.02 0.04*** -0.01 -0.01**

215 21 x11 9.79 (0.99) -1.15 (0.25)

1.09 (0.27)

0.16*** 0.04*** 0.01***

0.002 0.17**

0.003 -0.05*** -0.06***

21 x10 19.33 (0.37) -0.91 (0.36)

1.31 (0.19)

0.17*** -0.01

-0.02*** -0.03*** 0.16***

-0.05*** -0.05***

192 20 x 10 17.08 (0.51) -0.90 (0.37)

1.36 (0.17)

0.16*** 0.03*** 0.01***

0.05*** 0.003 -0.05*** -0.06***

21 x10 19.11 (0.38) -0.92 (0.36)

1.32 (0.18)

0.17*** 0.01

-0.02*** -0.04***

0.03*** 0.01*

-0.05*** -0.05***

192 20 x 10 16.08 (0.57) -0.92 (0.36)

1.38 (0.17)

* and *** indicate statistical significance at the 10%, 5% and 1% levels.

3.1. Test of market discipline from the asset side

To investigate whether borrowers pay higher interest rates to high-quality banks, we estimate the parameters of the baseline equation (1), using the SYS GMM estimator.

LnRATEit = a1LnRATEit-1 + p1LnCAPITALRit-1

+ p2LnAssetQualityit-1 + p3SIZEit-1 (1)

+ p4Profitabilityit-1 + p5LnLIQUIDITYit-1 + Tt't + uit

where, RATE is the weighted average interest rate on loans. LnAssetQuality includes the logarithm of RESERVE or the logarithm of DOUBTFUL, and Profitability includes ROA or ROE (in that way, we obtain four different specifications). CAPITALR, RESERVE, DOUBTFUL, SIZE, ROA, ROE and LIQUIDITY have been defined in the former section. T is a time dummy variable for years, controlling effects of unspecified macroeconomic and financial market conditions, which are assumed constant across banks.8 Note that the independent variables in the econometric model and the panel data provide fixed effects controlling for monotonous variables. In other words, the empirical strategy deals with differences across banks which are time-invariant, or variables that are slowly changing, as we can expect is the case of competition, currency structure of loans, and riskiness of customers.

We test the effect of bank fundamentals on interest rates on loans, as previous empirical tests suggest (Kim et al., 2005; Tovar-Garcia, 2012). We use a logarithmic transformation of the dependent and explanatory variables (excluding ROA and ROE, because they have some

8 It is not possible to include dummy variables for quarters, because of multicollinearity concerns.

negative values). Consequently, the coefficients indicate elasticity (and the model achieves linearity). Additionally, we enter the explanatory variables with one lag to account for the delay in the information used by borrowers to make decisions. We assume that potential borrowers, principally those well financially educated, will recognize changes in bank fundamentals, from one quarter to another. In addition, we assume a substantial volume of new lending each quarter, which is consistent with the increase of total loans over the years of analysis (see Figure 1).

The statistical hypothesis is that RATE depends positively upon the level of CAPITALR, RESERVE, SIZE, ROA, and LIQUIDITY, and depends negatively upon the level of DOUBTFUL, which is interpreted as evidence in favor of the market discipline hypothesis from the asset side. In particular, we expect a positive relationship between CAPITALR and RATE (in favor of the refinancing motive), and we expect a positive relationship between RESERVE and RATE, and a negative relationship between DOUBTFUL and RATE (in favor of the certification motive). The variables SIZE, ROA, ROE, and LIQUIDITY are principally working as control variables, but we also expect a significant effect on the interest rate, suggesting that borrowers also may take into account other bank fundamentals.

Table 4, columns (1-4), presents regression results. The dynamic model is well justified, because the dependent variable as regressor enters significantly. All regressions pass the Sargan test (validating the employed instrumental variables) and the Arellano-Bond tests for serial correlation of first and second order.

In the four specifications, the coefficients of CAPITALR are positive and statistically significant at the 1% level. That is, capital ratios are positively linked to interest rates on loans, as the refinancing motive predicts, and in favor of the asset side market discipline effect.

ARTICLE IN PRESS

E.D. Tovar-Garcia, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■-■■

The coefficients of RESERVE and DOUBTFUL are not statistically significant; that is, these bank characteristics are not important in determining the interest rate on loans. This result contradicts the certification motive and the market discipline hypothesis.

On the contrary, bank size presents significant and positive coefficients, suggesting that larger banks charge higher interest rates on loans, arguably, borrowers prefer large banks, and they are willing to pay higher rates to these banks. Similarly, the coefficients of LIQUIDITY are positive and significant. Therefore, bank size and liquidity may be approaching the certification and refinancing motives.

ROA and ROE are not statistically significant, and some coefficients related to time dummies are negative and significant, as we can expect, because of the general negative trend of the interest rate on loans during the period of analysis (see Figure 1).

It is possible to recognize other reasons for higher interest rates on loans. It could be that risky borrowers, willing to pay higher interest rates, systematically prefer high-quality banks. However, note that our panel data, through time series, account for this possibility. Furthermore, probably solvent banks operate in areas with less competition, working on sophisticated and specific financial services, lending more in domestic currency than in foreign currency, and so on. In other words, the interest rate on loans might be determined by the supply side, by the bank's internal capital supply. We are controlling for this possibility using the time dummy variables, because we do not have enough data to try to estimate both the demand and supply curves; therefore, we used the reduced form equation proposed by Park (1995).

As additional robustness checks, we also estimated the four basic specifications using the DIF GMM estimator (see Table 4, columns 5-8). These regressions also pass the correlation tests and the Sargan test. In general, we can say that the findings remain qualitatively the same with respect to the effects of capital ratio on interest rate on loans. Note that only in two of the four regressions CAPITALR has positive and significant coefficients, in favor of the working hypothesis. In the other two regressions, its coefficients are not statistically significant, but there is no evidence against the working hypothesis. Similarly, the findings in the DIF GMM regressions support the effects of LIQUIDITY; subsequently, banks with higher liquidity charge higher interest rates.

It is interesting to note that under the DIF GMM model, the indicators of asset quality (RESERVE and DOUBTFUL) present significant coefficients with the expected sign, in favor of the certification motive, and ROA and ROE are positive and statistically significant, suggesting that banks with higher profitability charge higher interest rates on loans. Nevertheless, these results are not robust, although they are good news for future studies.

In the DIF GMM regressions, bank size present some negative and significant coefficients, contradicting the effects of bank size in the SYS GMM regressions. Therefore, the effect of bank size is uncertain.

To sum up, our findings imply a positive evidence of market discipline from the asset side, principally in favor of the refinancing motive, where banks with higher capital ratios and higher levels of liquidity charge higher interest

rates on loans. In other words, borrowers are willing to pay higher interest rates to banks with high levels in these bank fundamentals, which indicate the banks' ability to provide credits in the future.

4. Conclusions

Market discipline plays an important role for reliability of banking systems. The new Basel Accord, theoretical frameworks, and empirical evidence supporting the existence of market discipline (from the liability and asset sides) provide the basis for testing market discipline in the Kyrgyz Republic.

In this paper, we verified whether borrowers pay higher interest rates to high-quality Kyrgyz banks, as the market discipline hypothesis from the asset side predicts. Similar tests were developed in Norway, Mexico, and Russia where, in general, the findings suggest that banks confront market discipline induced by borrowers (Kim et al., 2005; Tovar-Garcia, 2012, 2013).

In Kyrgystan, our empirical findings suggest that higher levels of capital to asset ratios and liquidity are positively linked to higher interest rates on loans, as the refinancing motive predicts. In other words, borrowers are willing to pay higher interest rates to solvent banks to facilitate future credit lines and new loans. In that way, borrowers discipline bank risk-taking.

The theory also highlights a certification motive, where borrowers are willing to pay higher interest rates to banks with high levels of asset quality (quality of bank loans). In Kyrgystan, our findings on this respect are not robust. Our indicators, reserves for loan losses/nonperforming loans and doubtful loans/total loans, show the expected effects in DIF GMM regressions, but this result is not supported by the SYS GMM regressions (our baseline model). Similarly, we did not find robust effects of bank profitability and bank size on loan rate.

Our results only suggest that borrowers, to select a commercial bank, are concerned about major sources of funds, such as capital ratios and liquidity. This result is statistically significant and robust to different methods and control variables. We can interpret this as evidence in favor of the market discipline hypothesis induced by borrowers in the Kyrgyz banking system.

Barth et al. (2013) show that Kyrgyzstan, in comparison with the rest of the world, has very low levels of information disclosure requirements. In their index on the private monitoring of banks (incentives and the ability of private investors to monitor and exert effective governance over banks), Kyrgyzstan ranks fairly low. Nonetheless, following the economic and political crisis, Kyrgyzstan increased private monitoring and official supervisory powers, and tightened bank capital regulations.

In general, in Kyrgyzstan the implementation of the Basel Accords is lacking and lagging, there are only draft regulations, and they are not published. It is not clear if there are plans to implement the Third Pillar (Financial Stability Institute BIS, 2013). Bankers argue that a disclosure policy is costly because the exposure of risk management strategies will negatively affect bank profitability, and can generate banking panic. However, the Third Pillar (market discipline) requires disclosure and transparency to be

ARTICLE IN PRESS

E.D. Tovar-García, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■ ■■ 9

enhanced. This will lead to higher levels of capitalization and the reduction of bank risk, with useful social implications.

Subsequently, we found evidence to support the Third Pillar of Basel III, and we can recommend to policymakers to work according to the disclosure policy, because borrowers also can use bank information to regulate the risky behavior of their banks.

In other words, there is empirical evidence of market discipline, which can be reinforced by bank information disclosures, strengthening bank transparency as proposed by Basel II in 2004 (Basel Committee on Banking Supervision, 2006), and reaffirmed in 2010 in Basel III, which also proposed the detailed disclosure of the capital base (Basel Committee on Banking Supervision, 2011, 2013). This will complement supervision activities, because bank creditors and debtors monitor and influence bank risk-taking, allowing supervisors to react to market signals, as well as regular practices.

This work presents important empirical findings, which should encourage policymakers to continue working on market discipline in Kyrgyzstan, and suggesting that other post-Soviet countries of Central Asia should explore the market discipline hypothesis. Nevertheless, it is important to recognize data limitations, our period of analysis covers only a few years, and we mainly used accounting information. Future research for Kyrgyzstan should develop surveys, and empirical tests, directly employing relevant information from banking agreements.

Acknowledgment

The article was prepared within the framework of the Basic Research Program at the National Research University Higher School of Economics (HSE) and supported within the framework of a subsidy granted to the HSE by the Government of the Russian Federation for the implementation of the Global Competitiveness Program.

References

Allen, F., Carletti, E., & Marquez, R. (2011). Credit market competition and capital regulation. Review of Financial Studies, 24(4), 983-1018. doi:10.1093/rfs/hhp089. Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(2), 277-297. Available from <http://restud.oxfordjournals.org/content/58/2/277 .short>.

Asian Development Bank, International Monetary Fund, & The World Bank. The Kyrgyz Republic. Joint economic assessment: Reconciliation, recovery and reconstruction. (2010). <http://www.imf.org/external/np/ pp/eng/2010/072110.pdf>. Barth, J. R., Caprio, G., & Levine, R. (2013). Bank regulation and supervision in 180 countries from 1999 to 2011. Journal of Financial Economic Policy, 5(2), 111-219. doi:10.1108/17576381311329661. Basel Committee on Banking Supervision. International convergence of capital measurement and capital standards: A revised framework comprehensive version. (2006). <http://www.bis.org>. Basel Committee on Banking Supervision. Basel III: A global regulatory framework for more resilient banks and banking systems (revised version). (2011). <http://www.bis.org>. Basel Committee on Banking Supervision. Consultative document: Revised Basel III leverage ratio framework and disclosure requirements. (2013). <http://www.bis.org>.

Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115-143. doi:10.1016/S0304-4076(98)00009-8.

Brown, M., Maurer, M. R., Pak, T., & Tynaev, N. (2009). The impact of banking sector reform in a transition economy: Evidence from Kyrgyzstan. Journal of Banking & Finance, 33(9), 1677-1687. doi:10.1016/j.jbankfin .2009.04.002.

Calomiris, C. W. (1999). Building an incentive-compatible safety net. Journal of Banking & Finance, 23(10), 1499-1519. doi:10.1016/S0378-4266(99)00028-X.

Chaussé, P. (2005). Computing generalized method of moments and generalized empirical likelihood with R. Journal of Statistical Software, 34(11), 1-35.

Demirgûç-Kunt, A., & Huizinga, H. (2013). Are banks too big to fail ortoo big to save? International evidence from equity prices and CDS spreads. Journal of Banking & Finance, 37(3), 875-894. doi:10.1016/j.jbankfin .2012.10.010.

Distinguin, I., Kouassi, T., & Tarazi, A. (2013). Interbank deposits and market discipline: Evidence from Central and Eastern Europe. Journal of Comparative Economics, 41(2), 544-560. doi:10.1016/j.jce.2012.07.005.

Evanoff, D. D.,Jagtiani,J. A., & Nakata, T. (2011). Enhancing market discipline in banking: The role of subordinated debt in financial regulatory reform. Journal of Economics and Business, 63(1), 1-22. doi:10.1016/ j.jeconbus.2010.07.001.

Financial Stability Institute BIS. FSI Survey Basel II, 2.5 and III implementation. (2013). <http://www.bis.org/index.htm>.

Hasan, I., Jackowicz, K., Kowalewski, O., & Kozlowski, t. (2013). Market discipline during crisis: Evidence from bank depositors in transition countries. Journal of Banking & Finance, 37(12), 5436-5451. doi:10.1016/ j.jbankfin.2013.06.007.

Jogorku Kenesh of the Kyrgyz Republic. Zakon Kyrzyzskoi Respubliki o zashite bankovskih vkladov (The law of the Kyrgyz Republic on banking deposit protection), The Kyrgyz Republic. (2008). <http://www .nbkr.kg/index1.jsp?item=1200>.

Karas, A., Pyle, W., & Schoors, K. (2010). How do Russian depositors discipline their banks? Evidence of a backward bending deposit supply function. Oxford Economic Papers, 62(1), 36-61. doi:10.1093/oep/ gpp006.

Karas, A., Pyle, W., & Schoors, K (2013). Deposit insurance, banking crises, and market discipline: Evidence from a natural experiment on deposit flows and rates. Journal of Money, Credit, andBanking, 45(1), 179-200. doi:10.1111/j.1538-4616.2012.00566.x.

Kim, M., Kristiansen, E. G., & Vale, B. (2005). Endogenous product differentiation in credit markets: What do borrowers pay for? Journal of Banking & Finance, 29(3), 681-699. doi:10.1016/j.jbankfin .2004.05.024.

Lavrushin, O.I. Bankovskoe delo: Ekspress kurs (Banking: Express course). Moscow: Knorus. (2009). <http://nashol.com/2011042554649/ bankovskoe-delo-ekspress-kurs-lavrushin-o-i.html>.

Matovnikov, M. Y. (2012). Novatsii v regulirovanii: Zlo ili blago? (Innovations in regulation: Evil or goodness?). Dengi I Kredit, (5), 3034.

Murzaeva, D., & Akçali, P. (2013). General perceptions of a good political leader in Kyrgyzstan. Journal of Eurasian Studies, 4(1), 47-54. doi:10.1016/j.euras.2012.07.001.

National Bank of the Kyrgyz Republic. Annual Report of the National Bank of the Kyrgyz Republic for 2010. Bishkek. (2011). <http://www.nbkr.kg>.

National Bank of the Kyrgyz Republic. Annual Report of the National Bank of the Kyrgyz Republic for 2012. Bishkek. (2013a). <http://www.nbkr.kg>.

National Bank of the Kyrgyz Republic. The financial sector stability report of the Kyrgyz Republic. Bishkek. (2013b). http://www.nbkr.kg.

Park, S. (1995). Market discipline by depositors: Evidence from reduced-form equations. The Quarterly Review of Economics and Finance, 35, 497-514. doi:10.1016/1062-9769(95)90051-9.

Peresetsky, A. Market discipline and deposit insurance in Russia (No. 14/2008). Helsinki. (2008). <http://www.bof.fi/bofit>.

Ruziev, K., & Majidov, T. (2013). Differing effects of the global financial crisis on the Central Asian Countries: Kazakhstan, the Kyrgyz Republic and Uzbekistan. Europe-Asia Studies, 65(4), 682-716. doi:10.1080/ 09668136.2013.766044.

Semenova, M. How depositors discipline banks: The case of Russia (No. 07/02). (2007). <http://ideas.repec.org/p/eer/wpalle/07-02e.html>.

Sytenkova, D. Mejdunarodnyi arbitraj po "Manas banku" eshe ne zakonchen (The international arbitration of "Manas Bank" has not finished yet). (2013). <http://www.eng.24.kg/economics/167796-dzhoomart -otorbaev-mezhdunarodnyj-arbitrazh-po.html> Accessed 24.06.14.

Tovar-García, E. D. (2012). Market discipline in Mexican banks: Evidence from the asset side. Cuadernos de Economía: Spanish Journal ofEconomics and Finance, 35(99), 172-181. doi:10.1016/S0210-0266(12)70033-X.

ARTICLE IN PRESS

10 E.D. Tovar-García, R. Kozubekova / Journal of Eurasian Studies ■■ (2016) ■■ ■■

820 Tovar-García, E. D. (2013). Distsipliniruyut li rossiyskie zaemshchiki svoi Ungan, E., Caner, S., & Ozyildirim, S. (2008). Depositors' assessment of bank 832

821 banki? (Do Russian borrowers discipline their banks?). Vestnik riskiness in the Russian Federation. Journal of Financial Services Research,

822 Moskovskogo Universiteta Seriya 6. Economica, (4), 69-79. Available from 33(2), 77-100. doi:10.1007/s10693-007-0025-0. 834 <http://publications.hse.ru/articles/110402222>. Wu, Y., & Bowe, M. (2012). Information disclosure and depositor discipline

824 Tovar-García, E. D. (2014). Market discipline: A review of the Mexican in the Chinese banking sector. Journal of International Financial Markets,

825 deposit market. Latin American Economic Review, 23(6), 1-33. Institutions and Money, 22(4), 855-878. doi:10.1016/j.intfin.2012.05.004.

826 doi:10.1007/s40503-014-0006-2. Zabortseva, Y. N. (2012). From the "forgotten region" to the "great game" 838

827 Tovar-García, E. D. (2015a). Exposure to interbank market and risk-taking region: On the development of geopolitics in Central Asia. Journal of

828 by Mexican banks. Cuadernos de Economía: Spanish Journal of Economics Eurasian Studies, 3(2), 168-176. doi:10.1016/j.euras.2012.03.007.

829 andFinance, doi:10.1016/j.cesjef.2015.11.002.

830 Tovar-García, E. D. (2015b). Market discipline through subordinated debt

831 in Mexican banks. Revista de Economia Aplicada, 23(68), 61-80.