Page 1 of 34
Universiteit van Amsterdam
Competition Level in the National Banking Sector and the Response to the
Unconventional Monetary Policy in The Euro Area
Empirical Study in selected Euro Area Countries Author: Tony, OnKit Yeung 114200565
Supervisor: Dr. Mihnea Constantinescu Abstract
As the aftermath of the serval financial crisis in late 2000 to 2010s, the ECB massively adopted the unconventional monetary policy to stimulate the economy. The relationship between the monetary transmission mechanism and market structure plays a significant role in financial stability. This paper uses a random-effect model with country-level data drawn from 12 selected countries within the Euro area, with a period of 2007 to 2017. The theoretical model rests on the oligopolistic version of the Monti-Klein model (Freixas and Rochet, 1997), intended to discover the relationship of the Herfindahl- Hirschman Index (HHI), Unconventional Monetary Policy (UMP) the interaction of HHI and UMP to the lending margins. The preliminary empirical results suggest that HHI and UMP are partially significant under different market segmentation. At the same time, the evidence of UMP is relatively weaker, and no evidence is found to support the relationship of the interaction variables.
Keywords: Unconventional monetary policy, Monetary Transmission Mechanism, Interest rate pass-through, bank risk-taking channel, Herfindahl–Hirschman Index, Monti-Klein model
Page 2 of 34
Statement of Originality
This document is written by Student Tony, OnKit Yeung, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of the completion of the work, not for the contents.
Page 3 of 34
Table of Contents
1. Introduction 4
2. Literature review 5
2.1 Unconventional monetary policy (UMP) 5
2.1.1 UMP worldwide 5
2.1.2 UMP in Eurozone 5
2.2. Monetary Transmission Mechanism(MTM) 6
2.2.1. Different types of channels in MTM 6
2.2.2. The empirical linkage between UMP, MTM and bank competition level 8 2.3. Overview of competition and concentration measurements of the banking sector 9
3. Data and Methodology 10
3.1. Theoretical model 10
3.2. Choice of explanatory variables 14
3.2.1 Proxy of bank competition 14
3.2.2. Proxy of UMP 14
3.2.3. Proxy of Other factors 15
3.3 Econometric specification 16
3.4 Data Source 18
3.5 Summary statistics 18
4. Empirical results 20
5. Conclusion & Discussion 23
Appendixes 25
Page 4 of 34 1. Introduction
After the global financial crisis in 2008, policymakers reconsider the credibility of the neoclassical economic theory. At the same time, the financial instability hypothesis (Minsky, 1992) provides alternative insights into the causal chains guiding the economy. Minsky advocates that for an economy to operate within reasonable bounds, the government should conduct interventions in the market to lighten the impact of exogenous shocks. The theory later flourished in the Modern Monetary Theory and popularized the application of unconventional monetary easing policy. However, the monetary easing policy presents a potential trade-off that can encourage bank risk-taking (Woodford 2012). A better understanding of the linkage of monetary policy and financial stability can be in favor of stable economic growth. (Cetorelli, 2004; Bonaccorsi Di Patta and Dell’Aricca, 2004; Beck et al., 2004).
In this paper, we intend to discover the relationship between the competition level in the banking industry and the extent of monetary transmission, by analyzing the change in lending margins of Monetary Financial Institutions to the private sector. The sample is obtained within 12 selected European countries for the period 2007 to 2017. The result can provide a more in-depth understanding of the interest rate pass-through and ideally to improve the effectiveness of the monetary policy. Three alternative hypotheses are tested as follows:
𝐻! ∶ 𝐿𝑜𝑤𝑒𝑟 𝑐𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛 𝑙𝑒𝑣𝑒𝑙 𝑤𝑖𝑙𝑙 𝑤𝑖𝑑𝑒𝑛 𝑙𝑒𝑛𝑑𝑖𝑛𝑔 𝑚𝑎𝑟𝑔𝑖𝑛𝑠 𝐻" ∶ 𝑈𝑛𝑐𝑜𝑛𝑣𝑒𝑛𝑡𝑖𝑜𝑛𝑎𝑙 𝑚𝑜𝑛𝑒𝑡𝑎𝑟𝑦 𝑝𝑜𝑙𝑖𝑐𝑦 𝑤𝑖𝑙𝑙 𝑛𝑎𝑟𝑟𝑜𝑤 𝑡ℎ𝑒 𝑙𝑜𝑎𝑛 𝑚𝑎𝑟𝑔𝑖𝑛𝑠
𝐻# ∶ 𝐿𝑜𝑤𝑒𝑟 𝐶𝑜𝑚𝑝𝑒𝑡𝑖𝑡𝑖𝑜𝑛 𝑙𝑒𝑣𝑒𝑙 𝑡𝑜𝑔𝑒𝑡ℎ𝑒𝑟 𝑤𝑖𝑡ℎ 𝑡ℎ𝑒 𝑢𝑛𝑐𝑜𝑛𝑣𝑒𝑛𝑡𝑖𝑜𝑛𝑎𝑙 𝑚𝑜𝑛𝑒𝑡𝑎𝑟𝑦 𝑝𝑜𝑙𝑖𝑐𝑦 𝑤𝑖𝑙𝑙 𝑎𝑓𝑓𝑒𝑐𝑡 𝑡ℎ𝑒 𝑒𝑥𝑡𝑒𝑛𝑡 𝑜𝑓 𝑙𝑜𝑎𝑛 𝑚𝑎𝑟𝑔𝑖𝑛𝑠
There is a lack of consensus on the optimal way of measuring the market competition level. In this paper, The Herfindahl-Hirschman Index will be used as the proxy of competition level, and the validity of the proximation determined by the Structure-Conduct-Performance (SCP) hypothesis, if the assumption holds, the higher concentration indicates lower competition level (Mason, 1939). The evidence gathered in the current study indicates a lower level of competition induces non-competitive pricing, and the index becomes a natural proxy for competition level.
Page 5 of 34
Moreover, unconventional monetary policy will be proxied by the composition of the National Central Bank balance sheet. The implementation of monetary easing is measured by the growth in central banks’ asset holdings.
The paper will start with the literature review in Section 2. Thenceforth, we will discuss the underlying theoretical framework, the reasoning on variables selecting and method of
approximation in the first part of Section 3, following that will be the data source and summary statistics in the latter part of Section 3. The empirical result will be present in Section 4. At last, the conclusion and discussion will be in Section 4. The detailed statistics will be found in appendices.
2. Literature review
2.1 Unconventional monetary policy (UMP)
2.1.1 UMP worldwide
The Zero Interest Rate Policy (ZIRP) was first introduced by the Bank of Japan (BoJ) in 1999, in the form of the government bond purchasing program. In the program, BoJ purchased sovereign bonds from the domestic banking sector and provided liquidity to the banking industry. The purpose of the program was to convert the excessive reserves into commercial lending and ultimately drive up the asset price and encounter the deflation (Joyce & Miles, 2012). The primary program did not have a sufficient impact on inflation. BoJ, therefore, expanded the program budget, and the inflation peaked at nearly a double to the original target for 2 % per year. As the aftermath of the 2008 global financial crisis, the Bank of England (BoE) and Federal Reserve (Fed) both attempted on Unconventional Monetary Policy (UMP) in the manner of asset purchasing programs with a focus on domestic sovereign bonds.
2.1.2 UMP in Eurozone
After the sovereign debt crisis in 2010, the Eurozone had experienced target balances deficit in the national central bank (NCB), the Target balances system represent "Trans-European
Page 6 of 34
similar level current account balances with other members (Sinn & Wollmershauser, 2012). Since the NCBs are obligated to issue public debt in euro- denominated, the market has a consensus that the NCB cannot guarantee the creditworthiness of the bonds (Grauwe, 2012). Especially in Greece, Ireland, Italy, Portugal, and Spain (GIIPS), these countries recorded significant target deficits, and the liquidity flowed to particularly in the Netherlands and Germany (Sinn & Wollmershauser, 2012).
The sovereign bank risk also induced a negative feedback loop between the EU bank
systematic risk and sovereign risk (Dieckmann & Plank, 2012; Giannone et al., 2011; Lucas et al., 2014, Pelizzon et al., 2016, Eser & Schwaab, 2016, Krishnamurthy et al., 2018) and the correlation of credit default swap between two parties have increased from 0.1 to 0.7 (Fratzscher & Rieth, 2019). As a result, ECB carried out bank rescue policies to restore the confidence of the market (Allen et al., 2013; Cooper & Nikolov 2013; Acharya et al., 2014; Leonello 2017; Farhi & Tirole, 2018) by authorizing the NCBs with target surpluses to borrow and withdraw the euro from liquidity circulation and to provide liquidity to countries with a target deficit by UMP (Manganelli, 2012; Ghysels et al., 2016). The UMP constructed by the Securities Market
Programme (SMP) started in 2010, and Outright Monetary Transactions (OMT) started in 2012. Since 2015, the European Central Bank has launched the extensive asset purchasing program, which is one of the examples of UMP. The idea of UMP is to lower the funding costs for the commercial bank convert. By doing this, the bank charge on loan rate will be lowered, thus, transmitted to higher consumption and investment level to stimulate the economy and stabilize the financial system.
2.2. Monetary Transmission Mechanism(MTM)
2.2.1. Different types of channels in MTM
The change of short term real interest rate set by monetary authority with UMP contributed to different channels of monetary transmission. When the monetary authority changes the short term real interest rate in order to stimulate consumption and investment (Kuttner & Mosser, 2002), it is hard to explain the striking macroeconomic responses caused by the initial monetary shocks itself. The existence of MTM is widely verified around the world under the different
Page 7 of 34
market structure, particularly in competition level, this will be one of the factors influencing the bank's risk-taking response through the change in price that determine the effectiveness of the transmission (Chen et al., 2017; Delis & Kouretas, 2011; Buch et al., 2014; Ioannidou et al., 2014; Jiménez et al., 2014). Several research (Taylor 2007, 2014; Mishkin, 2011) pointed out that the low-interest rate is one of the significant factors for the 2008 global financial crisis due to excessive risk-taking by the financial institution.
As an enhancement of traditional interest rate pass-through, scholars have tested the existence of the credit channel (Bernanke & Gertler, 1995) that consist of a bank lending channel
(Bernanke & Blinder, 1988). After the monetary easing of UMP, the following monetary tightening reduced banks' capital reserves. Although the legal requirement of the ratio has been largely erased by globalization within the banking sector and deregulated, in order to meet the shareholder target reserve to deposit ratio, the level of the deposit of the bank that has to scale back (Li, 2019). Thus, the loan market supply will be lowered, and the raised funding cost will pass to the borrowers.
The change in interest rate will change the valuation of bank assets and liabilities through the balance sheet channel (Bernanke & Gertler, 1989). The interest rate channel focuses on the endogenous change in external funding premium for a given composition of funding sources (Li, 2019). The monetary tightening policy will push up the interest rate and reduce the value of assets and collateral on bank balance sheet. Together with the information asymmetries, banks have to pay a premium to compensate for the risk associated with the increase in funding cost (Kuttner & Mosser, 2002).
Combined with all of the channels, it outlined the bank risk-taking channel. It has three components. Firstly, the interest rate affects the valuation in the balance sheet since banks are naturally exposed to mismatched maturity. Their primary liability is the deposit from clients, and it often has a shorter maturity than their primary asset, loans. The lower interest rate boosts up the value of the asset, and subsequently, raises the upper ceiling of the banks’ risk-taking,
loosened lending standards and therefore, more credit is granted to the riskier loaner (Altunbas et al., 2012, 2014; Jimenez et al., 2014, Borio & Zhu, 2012; Maddaloni & Peydró, 2011;
Dell'Ariccia et al., 2012; Paligorova & Santos, 2012). Secondly, the lower interest rate
Page 8 of 34
(Rajan, 2006). Thus, managers are forced to hold riskier assets or increase the level of leverage. Also, managers were motivated by compensation. The excessive risk-taking will shift to the public due to government guarantee and limited liability (De Young et al., 2013; Challe et al., 2013). Lastly, the adverse selection problem, the bank incentive in due diligence to loan
applicants will be lower (Dell'Ariccia & Marquez 2009; Dell'Ariccia et al., 2014). Lower interest rates also disproportionately change the expected cash flow of riskier clients, less risky clients, and respectively, the riskier borrower will take a more significant part within the borrowing pool, which can jeopardizing the financial systematic stability (Blommestein et al., 2011).
Besides, the extent of monetary transmission to the risk-taking level will be more significant to smaller and undercapitalized banks via bank capital channel (Kashyap & Stein, 1995,2000; Kishan & Opiela, 2000; Buch et al., 2014). Furthermore, the vulnerable bank will choose to reduce its lending (Van den Heuvel, 2002).
2.2.2. The empirical linkage between UMP, MTM and bank competition level
UMP usually refers to the monetary easing policy in order to reduce the term premium on assets returns in the long run. Therefore, the interest rate of assets will be lower. In the studies of Christensen & Rudebush (2012) and Wright (2012), they found evidence that the interest rate with above ten years maturity declined after the UMP was taken into account by the market. However, the impact is short-lived and will soak up within a few months. The idea of MTM is to measure the extent of the impact of policy rate set by the central bank to the bank response. Banks can generate abnormal marginal profit in higher market concentration with more access to additional deposits funding, while the incentive to search for yield will ease and the dependent to MTM will be lower (Adams and Amel 2011; Brissimis et al., 2014; Dell'Ariccia et al., 2014; Jimenez et al., 2014; Olivero et al., 2011).
The degree of interest rate pass-through was tested by many researchers (Hofmann & Mizen 2004; Sander & Kleimeier 2004; Bondt 2005; Kleimeier & Sander 2006; Kopecky & Hoose 2012) using Vector Autoregressive (VAR) models. Typically, more concentration markets will lead to higher asymmetric interest rates pass-through, that can widen the profit margin. The result of Baarsma & Vooren (2018) further confirmed that the degree of competition has a
Page 9 of 34
positive relationship with the interest rate pass-through with standardizing scales of the Boone indicator, the H-statistics, and the Herfindahl–Hirschman Index (HHI).
Nonetheless, they also mentioned that the concentration level and market competition are not the same in the banking sector. The interest rate pass-through can be interpreted by financial stability instead of banking concentration or competition (Baarsma & Vooren, 2018).
The impact of interest rate pass-through is often estimated by using the error correction model (ECM) and performing a regression analysis to test the coefficient of incorporation with bank competition level. Various Scholars developed different proxies on the competition level. Return on equity (ROE) can enlarge the effect of interest rate pass-through (Gigineishvili, 2011).
Likewise, Sorensen & Werner (2006) combined the concentration ratios with ROE to estimate the competition level. The results showed that the more competitive environment makes a quicker adaptation in the new policy rate. Instead of proximation the effect of competition level on interest rate pass-through, scholars could directly apply the Lerner index (Leroy & Lucotte, 2015) or Boone indicator (Van Leuvensteijn et al., 2011) for the competition measure in the ECM. Both of the studies found the positive relationship of interest rate pass-through and competition level.
2.3. Overview of competition and concentration measurements of the banking sector
The two popular approaches to the measurement of the banking competition level are structural or non-structural (Leon, 2014). The SCP hypothesis is mainly originated by Mason (1939, 1949) and Bain (1951, 1956) to explain the performance by analyzing the structural characteristics of the market. The elements include product differentiation, entry barriers, number of business, and the absolute and relative sizes. Thus, the companies can use their market power and use strategic behavior, as shown in figure 2.1, to earn an abnormal return. One crucial component of the SCP hypothesis is the efficiency hypothesis (Demsetz, 1973), which in contrast to Mason and Bain, indicates the excess performance need not caused by the collision but the economy of scale gained by its size. Both hypotheses suggested a positive correlation between marginal revenue and competition level (Berger & Hannan, 1989).
Page 10 of 34
Figure 2.1 The structure conduct performance paradigm (Hong et al., 2013)
Herfindahl-Hirschman Index (HHI) is one of the most often used concentration measures in bank concentration research, and the measure based on the SCP hypothesis (Bikker et al., 2012) and according to SCP hypothesis, the higher concentration will encourage banks to engage in unethical strategies like collusion. Hence, the HHI can be used as a proxy of imperfect competition when there is evidence to hold the validity of the SCP hypothesis.
The efficiency hypothesis has challenged the assumption of Mason on the abuse of market power. It recommended the marginal performance is simply due to the higher market power (Berger et al., 2015). Some scholars criticized SCP's ability to capture the market competition level (Bolt & Humphrey, 2015). The criticism later flourishes the New Empirical Industrial Organisation (NEIO) approaches (Shaffer, 1994), the non-structural measure of competition including the Panzar-Rosse H-statistic, Boone indicator, and Lerner index.
For the NEIO approaches, due to the limitation on sample size, it would be a probable to generate a negative Lerner index or biased Panzar-Rosse H-statistic (Bikker et al., 2012). The akin limitation applies to the Boone indicator, even though the indicator has proved the robustness in theory (Boone, 2008). Meanwhile, the feasibility of performing an empirical quality test is ambiguous (Schiersch & Schmidt-Ehmcke, 2010).
The article by Hannah and Kay (1977) argues that a measurement must fulfill four key criteria in order to obtain the market structure. Firstly, the concentration measurement should be a relative measurement, and it should be a ranking order of the cumulative share of output. Secondly, if the smaller firm sells its shares to a larger firm, the concentration ratio should always increase. Thirdly, if a newcomer below a specific standard size joins the competition, the event should reduce market concentration, and vice versa. Lastly, the merger within the market should always increase concentration. Nevertheless, no matter on the approaches of
Page 11 of 34
measurement, there will never be a perfect measure, and all measures have their specific limitations (Lipczynski et al., 2009).
In this paper, we intend to test the evidence of the SCP hypothesis by examining the
relationship between lending margin and concentration. If the positive link is definite, it indicates the evidence holds, and HHI can be a good proxy for the competition level.
3. Data and Methodology
3.1. Theoretical model
This paper will explore the extent of bank competition and the impact of UMP in determining the risk-taking level. The underlying theoretical model derived from the oligopolistic version of the Monti-Klein model (Klein, 1971) with an assumption that the bank profit is determined by the market structure. The theoretical explanation summarised by Freixas and Rochet (1997) and the extension interpretation on aggregate demand of loan and market elasticity were clarified by Li (2019). The original model devised for perfect competition (monopoly). The imperfect
competition model (oligopoly) is a more appropriate assumption for the banking market. Under the oligopolistic market, bank profits are decided by its share of total loan and deposit quantities, taking the others’ party quantities as given. Thus, strategic behavior will conjointly determine the bank sector risk-taking level.
The model simplified the balance sheet of the bank as follows:
Assets Liabilities
Bank Reserves (R) Deposits (D) Bank Loans (L)
On the assets side, bank reserves (R) are constructed by two components, cash reserves (CR) and its net position (N) in the interbank market. As mentioned in section 2.2.1, the reserve/ deposit ratio is not obligated be followed but set by the bank's will. Hence, the reserves/ deposits ratio measured as α.
Page 12 of 34 The net position (N) is expressed as:
𝑁 = 𝐷 − 𝐿 − 𝛼𝐷 = (1 − 𝛼) 𝐷 − 𝐿 (1) The maximization of the profit function occurs as follows:
𝑚𝑎𝑥$!,&!𝜋' = 𝑟(! + (𝑟$(𝐿' + ∑')* 𝐿*∗)𝐿
'− 𝑟&(𝐷'+ ∑')* 𝐷*∗)𝐷' − 𝐶'(𝐿', 𝐷')
(2)
i denotes the number of banks = 1… N. is a concave function, and C is the convex function increased by L or D, the cost function represents the bank operating costs. The interbank rate r took as given the assumption of the Cournot model. Substituting (1) into (2), the profit function can be expressed as:
𝑚𝑎𝑥$!,&!𝜋' = [𝑟$(𝐿'+ ∑')* 𝐿*
∗)− 𝑟]𝐿
'+ [𝑟(1 − 𝛼) − 𝑟&(𝐷' + ∑')* 𝐷*∗)]𝐷' − 𝐶'(𝐿', 𝐷')
(3)
In the Cournot equilibrium, the optimal quantities are𝐿∗ = 𝐿 ' ∗+ ∑
')* 𝐿*∗ and𝐷∗ = 𝐷'∗+
∑')* 𝐷*∗ . Each player contributes part of the total quantities𝐿∗ = 𝑚𝑠
!'𝐿∗ and𝐷∗= 𝑚𝑠"'𝐷∗,
where 𝑚𝑠!' = $∗!
$∗ 𝑤𝑖𝑡ℎ 1 > 𝑚𝑠!' > 0 and 𝑚𝑠"' =
&∗!
&∗ 𝑤𝑖𝑡ℎ 1 > 𝑚𝑠"' > 0.
To obtain the maximum, first-order conditions can express as: 𝜕𝜋' 𝜕𝐿'|$∗!,$∗ = 𝑟$𝐿 ∗− 𝑟 +𝑑𝑟$(𝐿∗) 𝑑𝐿∗ 𝜕𝐿∗ 𝜕𝐿∗'𝑚𝑠!'𝐿∗− 𝜕𝐶' 𝜕𝐿∗' = 0 (4) 𝜕𝜋'
𝜕𝐷'|&!∗,&∗ = 𝑟&𝐷
∗− 𝑟 +𝑑𝑟&(𝐷∗) 𝑑𝐷∗ 𝜕𝐷∗ 𝜕𝐷'∗𝑚𝑠"'𝐷∗− 𝜕𝐶' 𝜕𝐷'∗= 0 (5) -$∗ -$!∗ and -&∗
-&!∗ are equal to 1, by rearranging the equations and dividing both sides by equilibrium
interest rate, the optimal interest margin depends on aggregate market elasticities: 𝑟$(𝐿∗) − 𝑟 − 𝜕𝐶𝜕𝐿$ ' ∗ 𝑟$(𝐿∗) = − 𝑑𝑟$(𝐿∗) 𝑑𝐿∗ 𝐿∗ 𝑟$(𝐿∗)𝑚𝑠!' = 𝑚𝑠!' 𝜀$(𝑟$∗) (6) 𝜀$ = − 𝑑𝐿 𝑑𝑟$ 𝑟$ 𝐿 > 0 (7) 𝑟(1 − 𝛼) − 𝑟&(𝐷∗) − 𝜕𝐶$ 𝜕𝐷'∗ 𝑟&(𝐷∗) = 𝑑𝑟&(𝐷∗) 𝑑𝐷∗ 𝐷∗ 𝑟&(𝐷∗)𝑚𝑠"' = 𝑚𝑠"' 𝜀&(𝑟&∗) (8)
Page 13 of 34 𝜀& = 𝑑𝐿 𝑑𝑟& 𝑟& 𝐿 > 0 (9) The equations (6) and (8) are the margins that banks can obtain from the market (Belleflamme & Peitz, 2010). Higher market share of banks or greater inelasticity implies banks can charge more mark-ups. The evidence of the positive relationship between interest margin and market power has been proven by Li (2019). In this paper, the interest margin will remain as the dependent variables, and we will extend the study to find the relationship of interest margin, market power, UMP, bank characteristics, and macroeconomic factors. Under the oligopolistic model, the relations between interest margin and market competition will be recognized as the relationship of mark-up pricing and market competition as equation (6) and (8) (Rousseas, 1985). The interest margin only implies optimal values at the equilibrium and the bank could only modify its quantity. Thus, the model cannot reflect banks’ pricing strategy, but it is still providing insight into the determinants of interest margins (Li, 2019). This study accompanies the parent empirical model introduced by Corvoisier and Gropp (2002), which uses six
relationships between interest margins and concentration level derived from a similar model. For the validity of HHI to be able to interpret as a measure of competition, evidence of SCP must be held. If so, the clashing hypothesis mentioned in section 2.3 will become trivial (Berger & Hannan, 1989). Higher HHI, therefore, will imply a lack of competition in the market.
Moreover, the endogeneity problem on the aggregate level in the bank lending channel literature will be averted (Gambacorta, 2008).
The paper will estimate a panel-data regression analyzes banks’ loan interest margins on a country level. The selected sample covered 13 countries within the Eurozone from the year 2007 to 2017 cumulative for 11 crests. The fixed-effects model can control the unobserved country-specific effect that is time-invariant to secure the consistent and unbiased estimation for the bank competition.
According to the interest rate pass-through literature in section 2.2. The bank loan or deposit rate will influence the funding cost to enterprise and interest income to the household, which determines the wiliness of consumption and investment, the effect will later radiate to the real economy if the borrowers are having difficulty to have alternative funding sources (Bernanke, 2007). Nonetheless, this paper will not cover the effect of the real economy.
Page 14 of 34 3.2. Choice of explanatory variables
3.2.1 Proxy of bank competition
This paper will carefully follow Corvoisier and Gropp (2002) and Li (2019) to identify the relevant control variables based on the theoretical model.
Assuming the evidence that the SCP hypothesis holds, the HHI will be a natural proxy of bank competition under the Monti-Klein model. The higher HHI represents a lower competition. Moreover, since the 𝐻𝐻𝐼 = ∑ /
' 𝑚𝑠", related to equation (6) and (8). The index is calculated
by squaring the sum of the market share of each bank within the market, i denoted the number of banks.
3.2.2. Proxy of UMP
The Cournot pricing formula will be extended to account the UMP, bank characteristics and macroeconomy factors to proxy the aggregate loan demand conditions and market elasticity. Following the identification strategy of (Kenourgios & Ntaikou 2019), instead of measuring UMP countries by country on a national level, we could simply measure the amount of ECB Central Assets Purchases. In this paper, we adjust the measurement to the growth of the amount of ECB Monetary policy operations (MPOs), the MPOs include the open market operations, standing facilities, minimum reserve requirements for credit institutions, and the asset purchase programs as the UMP will associate an increase of NCBs portfolios. The alternative
measurement of using the country-specific yield on long-term government bonds correlated to non-economic reasons such as exposure to political risk or market risk (Baarsma & Vooren, 2018). Therefore, the alternative measure will not be used in this paper.
3.2.3. Proxy of Other factors
In the bank characteristics, first, we control the capitalization to control the leverage level of the banking industry, which will proxy by the bank capital over total assets. In a general view, higher capitalization has a negative relationship with the incentives of banks to issue risky loans,
Page 15 of 34
thus, lowering the loan margin (Delis & Kouretas, 2011). Despite that, if the capital ratio exceeds a certain threshold, it becomes an incentive for banks to take risks (Calem & Rob, 1999).
Second, we control the liquidity of funding sources, which will proxy by liquidating assets to deposits and short term funding. The liquidity position will insulate the loans from economic or monetary shocks (Cornett et al., 2011).
Lastly, we control the diversification of banks’ income, which is proxied by bank non-interest income to total income. The idea of diversification is to lower the risk-taking and stabilize the bank return. However, the empirical evidence is conflicting to these views (Demsetz and Strahan, 1997; Stiroh, 2004).
In macroeconomics, the cyclicality of bank system stability was studied by many scholars. The two main variables that affect the stability will be the inflation rate, which will proxy by Harmonised Index of Consumer Prices (HICP) and growth rate of real GDP will proxy by real GDP per capita growth rate (Demirgüç-Kunt & Detragiache, 1998; Marcucci & Quagliariello, 2009). Besides, to control the impact on the banking crisis to loan margin, we included a dummy variable of the event of the banking crisis (Laeven & Valencia, 2013).
3.3 Econometric specification
Following the identification strategy mentioned above, the panel data models are constructed as: 𝐿𝑜𝑎𝑛 𝑀𝑎𝑟𝑔𝑖𝑛',0 = 𝛼 + 𝛽! 𝐻𝐻𝐼 + 𝛽" 𝑈𝑀𝑃 + 𝛽# (𝐻𝐻𝐼 ∗ 𝑈𝑀𝑃) + 𝛽1 𝑌',0+ 𝛽2 𝑀',0
i denoted the number of banking sectors = 1… 12, t denoted to time waves (2007-2017). The loan margin is the difference between the interbank market rate and the bank loan rate. The HHI is the primary identifier in this paper to measure the competition level within the banking sector under the assumption that the SCP hypothesis holds, higher HHI implies a lower competition level. The UMP is to measure the extent of interest-rate pass-through. The interaction variables (HHI * UMP) designed to capture the joint effect of two variables. Y is the summary of bank characteristics, which include bank capital & reserves over total assets, liquid assets to deposits, and short-term funding and bank non-interest income to total income. M is the summary of macroeconomic factors, which include inflation rate, the growth rate of real GDP, and a dummy of the banking crisis.
Page 16 of 34
Higher HHI is expected to have a positive relationship with the loan margin since the bank has a higher market power to increase its profit. Higher UMP implies a more significant effect of the interest-rate pass through, which is likely to have a negative relationship to the loan margin since the UMP will drag down the Intermarket rate, and the bank will lower its loan rate accordingly due to competition. Higher capitalization, liquidity, and diversification expected to have a negative relationship with the loan margin since it lowers the incentive of risk-taking. Higher growth in real interest rate, inflation rate is expected to have a positive relationship with the loan margin since it increases the demand for loans. The occurrence of a banking crisis is expected to have a positive relationship with the loan margin since it increases the volatility of the return.
3.4 Data Source
The ECB statistics warehouse and World Bank Open Data obtain the bank loan margins, HHI, and UMP. The Harmonised monetary financial institution (MFI) statistics was introduced in 2003 to ensure the efficiency and quality of the cross-country research (Paries et al.,2014). MFI interest rates are based on different initial rate fixation (IRF) periods within each category. The loan margins in this paper are based on the monthly MFIs lending margins on new loans to households and non-financial corporations. HHI are based on the Herfindahl index for Credit institutions (CIs) total assets, Domestic, and the UMP are captured from the ECB consolidated balance sheet.
The bank capital over total assets included all funds contributed from the owner and the bank reserves; the assets included financial and non-financial assets. The data gathered by Financial Soundness Indicators Database under the International Monetary Fund (IMF). The liquid assets to deposits and short term funding and Bank non-interest income to total income gathered by Bankscope and Orbis Bank Focus, Bureau van Dijk (BvD). The inflation rate and real economic growth gathered by the ECB statistics warehouse. Finally, the dummy of the banking system crisis is based on the result of Laeven & Valencia (2018).
Its average will annualize the variables collected in 12 EU countries, which included Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, and Spain from the year 2007 – 2017, in order to perform the regression analysis.
Page 17 of 34 3.5 Summary statistics
The results of the summary statistics are shown in Table 1 below. Three dependent variables were used; thus, three regressions will be performed for the empirical analysis. The summary statistics by country can be found in Table 1 to Table 3 under Appendix A. Three missing data occur in the year 2017 under the bank characteristics. We will ignore it since it is unlikely to affect the result. In this paper, we will ignore the missing value and set it equal to 0. In the following part, we will describe the summary statistics of the panel data.
The lending margin among three subsets is gradually close. The lending margin to non-financial corporations has the lowest with 1.62%, the volatility is 0.71%, and the loan margin to households is highest with 1.65%, and its volatility is also highest with 0.84%. Overall, the new loan margins to both parties are 1.64%, with the lowest volatility of 0.66%. The overall lending margin was once widened to 3.7% and narrowed to 0.47%. The statistical number is stable and close among the selected sample.
HHI in 12 countries has a mean of 10.67%. It shows that the market is under imperfect
competitive conditions, the volatility of it is 8.9%. The maximum HHI of the banking industry is 38%, which implies high concentration and will translate to high market power and low
competition, its minimum is 1.83%. In the selected countries, the Netherlands (20.72%) and Finland (31.91%), are significantly more concentration markets. In contrast, Germany (2.61%) is the closest to the perfect competition market.
The UMP is uniformly based on the data of ECB with a mean of 24.26%, the volatility of it is 33.63%, the minimum is 29.57%, and the maximum is 66.87%. The higher UMP will imply a higher level of monetary easing policy. At last, the interaction variables will depend on the level of HHI and UMP. The highest mean is 1.22% in France, with the highest volatility of 3,68%. The lowest mean is 0% and near 0% volatility in Luxembourg.
The definition of the indicator and the reason for selection has been explained in section 3.2. The remaining variables are designed to control the impact on banking-sector specific
Page 18 of 34
We also report the Pearson correlation coefficient, and the table can found in Table 1 under Appendix B. Surprisingly, the correlation on lending margins to households and lending margins to the non-financial corporation is only 0.0984, which may suggest the market segmentation exists, and banks will treat sub-sectors differently. The bank characteristics have a relatively higher correlation to the lending margins. Meanwhile, the correlation among the variables is lower than the cut-off of 0.7, which can conclude that the variables are not excessively highly correlated and, therefore, not caused by any serious multicollinearity issues.
Table 1
4. Empirical results
Appendix C presents the results of the panel data regression and the conditions of regressions specified in section 3.3. The Hausman test rejected the null hypothesis of the validity of a fixed-effect. Therefore, the random effect model will be used for the study. Three dependent variables used to detect the extent of interest rate pass-through in different market segmentation. The first model depicts overall loan margins to the new loan, the HH stands for households, and the NFC stands for the non-financial corporation. The R-squared (Within) statistics is calculated by using the xtreg command under Stata, which provided the quality of prediction within the underlying dataset.
The three different loan margins analyzed by the same set of variables. Based on the result, HHI is significant at 0.1% on the overall loan margins and for the non-financial corporation.
Page 19 of 34
Furthermore, HHI is insignificant to explain the lending margin in the household market, which is contrary to the result from Li (2019) and Corvoisier and Gropp (2002). The HHI is a
percentage of market concentration, and the range of it is from 0 to 1. Thus, every percent increase in HHI will increase the overall lending margin by 0.04%. As the HHI is a nature proxy for imperfect competition under the SCP hypothesis, correspondingly, the results inferred that weaker competition could lead to the non-competitive pricing strategy in the NFC market and overall strategy. However, the impact on the household market is limited.
UMP is a significant factor between all of the three models, and the p-value is smaller than 1% on the household market with a coefficient -0.007. The UMP measured the percentage of
national central bank assets to its national GDP. Therefore, the range of it is from 0 to 1. As a result, every percent increase in UMP will lead to 0.007% decrease in the loan margin to
household, in the other two segmentation, UMP is significant at 10%, and the negative is in line with the expectation since the UMP will drag down the policy rate.
The HHI * UMP term is statistically insignificantly different from 0, and it is in unison with the result of Baarsma & Vooren (2008). It suggested within the observation window, that the interaction of concentration ratio and UMP did not affect the level of loan margins statistically, and the impact of the factor operates differently in the overall & NFC market compared to the household market.
The remaining variables are the control variables, which are designed to control the quality as a supplement of the main explanatory variables. The direction on its impact is almost identical, although the magnitude is different.
In the class of the bank characteristics, the capitalization ratio is significant at 0.1% on NFCs and the overall model with a positive coefficient suggested that the bank may engage in riskier behavior in price setting when the banking industry is well-capitalized. In these models, the impact of the funding liquidity and income diversification is minor to the loan margins within the samples.
Out of the macroeconomy factors, a positive growing GDP per capita reflects a right-hand side shifted aggregate demand for loans, and therefore, a positive impact on loan margins. However, the coefficient of GDP growth is negative. Higher inflation should indicate a lower elasticity of loan demand. Thus, the loan margin should widen, which is also contrary to the result. Though
Page 20 of 34
most of the macroeconomic factors are insignificant, the crisis variables under lending margin to NFC are significant at 5%. The event of the banking crisis is presumed to have a positive
relationship with loan margins since the increased probability of default can drag up the spreads. The contradictory result can be explained by the bank risk-taking behavior that banks lower their risk-taking by taking the less risky loan, and thus, reduce the lending margins. The evidence supports the impact of HHI, while the evidence on the effect of UMP is merely confirmed, and the interaction effect of HHI and UMP has no supporting evidence. More detail will be discussed in Section 5.
5. Conclusion & Discussion
This paper intends to replicate the research based on Li (2019), to test the significance of bank competition proxied by HHI, besides, an extension to the model to test the effectiveness of UMP and the extent of interaction between HHI and UMP on the loan interest margins. HHI confirmed its significant impact on the NFC market and overall lending margin. Based on the evidence, the results are conflicted to Li’s (2019) conclusion, the positive relationship on HHI suggests that SCP does hold and the HHI is a good proxy for competition. An increase in HHI can indicate a more concentrated market, and banks have higher market power, which can cause
non-competitive pricing to explain the positiveness of the relationship. Thus, the first null hypothesis is partially rejected and shows that higher HHI (lower competition) will widen the loan margins. The effectiveness of UMP is providing weaker evidence within the sample. It is statistically significant on the NFC markets and overall market at 10%, and the negative of the relationship can be explained by UMP. The banking industry is relatively stable, which drags down the lending margin lower. The result suggested that the second null hypothesis is again partially rejected, and the UMP can narrow the loan margin in the NFC market and overall market. The interaction relationship is providing none of the evidence in this paper, and the variables are insignificant among all models, worth mentioning the interaction variables operate in a different direction, which can imply a different mechanism between various market
segmentation. However, the study fails to reject the third null hypothesis, that is, the interaction has no statistical power to explain the loan margins.
Page 21 of 34
Among the regression, the within-sample R-squared is 0.4697, 0.1707, 0.414 respectively of overall lending margin, lending margin to households, and lending margin to NFC. The data shows that the regression is providing good prediction power within the underlying sample set except the lending margin to households, which opens up a discussion that the interest-rate pass through might operate differently among different market segmentation. The same independent variables set are not proving the same statistical power.
At last, although this paper does not discover the significant relationship in all intended variables, it could be caused by three reasons. First, the sample size is relatively small since we have only 12 countries from 2007 to 2017, and the trend of NCBs massive adoption of UMP began in the year of 2015. Secondly, in this paper, we assumed the SCP held, and HHI is a good proxy for market competition. Meanwhile, we have different approaches that can be used to proxy market competition. The selection of the proxy will determine the result of the paper. Lastly, this paper only covers a specific relationship between loan margins and MTM. The link between the loan margins and the private sector with different market segmentation needs to be examined. Future studies with more all-rounded competition measures and more precise sample sets may discover more shreds of evidence on how the interest-rate pass-through operates in a different market segmentation under the unconventional monetary policy.
Page 22 of 34 Appendixes
Appendix A Table 1
Page 23 of 34 Table 2
Page 24 of 34 Table 3
Page 25 of 34 Appendix B
Page 26 of 34 (Continue…)
Page 27 of 34 Appendix C
Page 29 of 34 References
Acharya, V., and H. Naqvi. 2012. “The seeds of a crisis: a theory of bank liquidity and risk-taking over the business cycle.” Journal of Finance & Economic 106 349-366.
Acharya, V., I. Drechsler, and P. Schnabl. 2014. “A pyrrhic victory? Bank bailouts and sovereign credit risk.” The Journal of Finance 69 2689–2739.
Adams, R.M., and D.F. Amel. 2011. “Market structure and the pass-through of the federal funds rate.” Journal Banking & Finance 35 1087–1096.
Adrian, T., and H.S. Shin. 2010. “Liquidity and leverage.” Journal of Finance 19 418–437. Agur, I., and M. Demertzis. 2012. “Excessive bank risk taking and monetary policy.” European
Central Bank Working Paper Series 1457.
Allen, F., E. Carletti, I. Goldstein, and A. Leonello. 2013. “Government guarantees and financial stability.” Mimemo.
Altunbas, Y., L. Gambacorta, and D. Marques-Ibanez. 2012. “Do bank characteristics influence the effect of monetary policy on bank risk?” Economics Letters 117 220-222.
Altunbas, Y., L. Gambacorta, and D. Marques-Ibanez. 2014. “Does monetary policy affect bank risk?” International Journal of Central Banking (March) 95 - 135.
Angeloni, I., E. Faia, and M. Lo Duca. 2015. “Monetary policy and risk taking.” Journal of Economic Dynamic Control 52 285–307.
Baarsma, B., and M. Vooren. 2018. “Bank Competition and The Transmission of Unconventional Monetary Policy in the Euro Area.” European Compeition Journal 14:1 174-193.
Beck, T., A. Demirguc-Kunt, and V. Maksimovic. 2004. "Bank competition and access to finance." Journal of Money Credit and Banking, 36 627–648.
Bernanke, B.S., and M. Gertler. 1995. "Inside the black box: the credit channel of monetary policy transmission." Journal of Economic Perspectives 9(4) 27-48.
Page 30 of 34
Blommestein, H., S. Eijffinger, and Z. Qian. 2011. “Monetary policy rules, adverse selection and long-run financial risk.” Tilburg University, Center for Economic Research Discussion Papers 111-121.
Bonaccorsi Di Patta, E., and G. Dell’Aricca. 2004. "Bank competition and firm creation,." Journal of Money Credit and Banking, 36 225–252.
Bondt, GJ De. 2005. “Interest Rate Pass-Through: Empirical Results for the Euro Area.” German Economic Review 37 6(1).
Borio, C., and H. Zhu. 2012. “ Capital regulation, risk-taking and monetary policy: a missing link in the transmission mechanism?” Journal of Finance Stability 8 236–251.
Boyd, John, and David Runkle. 1993. “Size and Performance of Banking Firms: Testing the Predictions of Theory.” Journal of Monetary Economics 31 47-67.
Brissimis, S.N., M.D. Delis, and M. Iosifidi. 2014. “Bank market power and monetary policy transmission. .” International Journal of Central Banking 10 173–214.
Buch, C.M., S. Eickmeier, and E. Prieto. 2014. “In search for yield? Survey-based evidence on bank risk taking.” Journal of Economic Dynamics and Control 43 12-30.
Cetorelli, N. 2004. "Real effects of bank competition." Journal of Money, Credit and Banking 36 543–558.
Challe, E., B. Mojon, and X. Ragot. 2013. “Equilibrium risk shifting and interest rate in an opaque financial system.” European Economic Review 63 117–133.
Chen, M., J. Wu, B.N. Jeo, and R. Wang. 2017. “Monetary policy and bank risk-taking: Evidence from emerging economies.” Emerging Markets Review 31 116-140.
Christensen, J.H., and G.D. Rudebush. 2012. “The Response of Interest Rates to US and UK Quantitative Easing.” The Economic Journal F385 122(564).
Čihák, M., and H. Hesse. 2010. “Islamic Banks and Financial Stability: An Empirical Analysis.” Journal of Financial Services Research, 38 (2-3) 95–113.
Cooper, R., and K. Nikolov. 2013. “ Government Debt and Banking Fragility: The Spreading of Strategic Uncertainty.” National Bureau of Economic Research No. 19278.
Page 31 of 34
Delis, M.D., and G., Kouretas. 2011. “Interest rates and bank risk-taking.” Journal of Banking & Finance 35 840-855.
Dell'Ariccia, G., D. Igan, and L. Laeven. 2012. “Credit booms and lending standards: evidence from the subprime mortgage market.” Journal of Money Credit Bank 44 367–384. Dell'Ariccia, G., and R. Marquez. 2009. “Lending booms and lending standards.” Journal of
Finance 61 2511–2546.
Dimitris, K. & Despoina, N. 2019. “ECB’s unconventional monetary policy and bank lending supply and performance in the euro area.” Journal of Economics and Finance De Young, R., E.Y. Peng, and M. Yan. 2013. “Executive compensation and business policy
choices at U.S. commercial banks.” Journal of Financial and Quantitative Analysis 48 165–196.
Diamond, D.W., and R.G. Rajan. 2012. “Illiquid banks, financial stability, and interest rate policy.” Journal of Political Economy 120 552–591.
Dieckmann, S., and T. Plank. 2012. “Default risk of advanced economies: an empirical analysis of credit default swaps during the financial crisis.” Review of Finance 16 903-934.
Eser, F., and B. Schwaab. 2016. “Evaluating the impact of unconventional monetary policy measures: empirical evidence from the ECB’s Securities Markets Programme.” Journal of Financial Economics 119 147-167.
Farhi, E., and J. Tirole. 2018. “Deadly embrace: sovereign and financial balance sheets doom loops.” The Review of Economic Studies 85, 1781–1823.
Fratzscher, M., and M. Rieth. 2019. “Monetary Policy, Bank Bailouts and the Sovereign-Bank Risk Nexus in the Euro Area.” Review of Finance 745-775.
Ghysels, E., J. Idier, S. Manganelli, and O. Vergote. 2016. “A high frequency assessment of the ECB securities markets programme.” Journal of the European Economic Association 15 218–243.
Giannone, D., M. Lenza, H. Pill, and L. Reichlin. 2011. “Non-standard monetary policy measures and monetary developments.” ECB Working Paper No. 1290.
Page 32 of 34
Hofmann, B., and P. Mizen. 2004. “Interest Rate Pass-Through and Monetary Transmission: Evidence from Individual Financial Institutions' Retail Rates.” Economica 99 71. Ioannidou, V., S. Ongena, and J.L. Peydró. 2014. “Monetary policy, risk taking, and pricing:
evidence from a quasi-natural experiment.” Review of Finance 19 95-144.
Jiménez, G., S. Ongena, J.L. Peydró, and J. Saurina. 2014. “Hazardous times for monetary policy: what do twenty-three million bank loans say about the effects of monetary policy on credit risk-taking?” Econometrica 82 463-505.
Joyce, M, and D. Miles. 2012. “Quantitative Easing and Unconventional Monetary Policy, an Introduction.” The Economic Journal F271 122(564).
Kashyap, A.K., and J.C., Stein. 1995. “The impact of monetary policy on bank balance sheets.” Carnegie-Rochester Conference Series on Public Policy 42 151-195.
Kashyap, A. K.., and J. C., Stein. 2000. “What Do a Million Observations on Banks Say about the Transmission of Monetary Policy?” American Economic Review 90 407-428.
Kishan, R.P., and T.P., Opiela. 2000. “Bank size, bank capital, and the bank lending channel.” Journal of Money Credit Banking 32 121–141.
Kleimeier, S., and H. Sander. 2006. “Expected Versus Unexpected Monetary POlicy Impulses and Interest Rate Pass-Through in Euro-Zone Retail Banking Markets.” Journal of Banking & Finance 1839 30.
Kopecky, K. J. , and D. Van Hoose. 2012. “Imperfect Competition in Bank Retail Markets.” Journal of Money Credit and Banking 1185 44(6).
Krishnamurthy, A., S. Nagel, and A. Vissing-Jorgensen. 2018. “ECB policies involving government bond purchases: impact and channels.” Review of Finance 22 1-44.
Leonello, A. 2017. “Government guarantees and the two-way feedback between banking and sovereign debt crises.” ECB Working Paper Series No 2067.
Lucas, A., B. Schwaab, and Z. Xin. 2014. “Conditional and joint credit risk.” Journal of Business & Economic Statistics 32 271-284.
Manganelli, S. 2012. “The impact of the securities markets programme.” ECB Research Bulletin 17 2-5.
Page 33 of 34
Maddaloni, A., and J. L. Peydró. 2011. “Bank risk-taking, securitization, supervision, and low interest rates: evidence from the euro-area and the U.S. lending standards.” The Review of Financial Studie 24 2121–2165.
Mishkin, F. 2011. “Monetary Policy Strategy: Lessons From the Crisis.” NBER Working Papers 16755.
Minsky, H. P. 1992. "The Financial Instability Hypothesis*." Levy Economics Institute Working Paper (74).
Olivero, M.P, L. Yuan, and B.N. Jeon. 2011. “Consolidation in banking and the lending channel of monetary transmission: evidence from Asia and Latin America.” Journal of International Money Finance 30 (6) 1034–1054.
Paligorova, T., and J.C. Santos. 2012. “Monetary Policy and Bank Risk-Taking: Evidence From the Corporate Loan Market.” Federal Reserve Bank of New York Working Paper.
Pelizzon, L., M. Subrahmanyam, D. Tomio, and J. Uno. 2016. “Sovereign credit risk, liquidity and ECB intervention: deus ex Machina?” Journal of Financial Economics 122 86-115.
Rajan, R.G. 2006. “Has finance made the world riskier?” European Financial Management 12 499-533.
Sander, H., and S. Kleimeier. 2004. “Convergence in Euro-Zone Retail banking? What Interest Rate Pass-Through Tell Us About Monetary Policy Transmission, Competition and Integration.” Journal of International Money and Finance 461 23.
Sinn, H.W., and T. Wollmershauser. 2012. “Target Loans, Current Account Balances and Capital Flows: The ECBs Rescue Facility.” International Tax and Public Finance 468 19(4). Taylor, J.B. 2007. “Housing and Monetary Policy.” NBER Working Papers 13682.
Taylor, J.B. 2014. “Causes of the financial crisis and the slow recovery: a ten-year perspective.” Across the Great Divide: New Perspectives on the Financial Crisis 51-65.
Valencia, F. 2014. “Monetary policy, bank leverage, and financial stability.” Journal Economic Dynamic Control 47 20-38.
Page 34 of 34
Wright, J.H. 2012. “What Does Monetary Policy Do to Long Term Interest Rates at the Zero Lower Bound.” The Economiic Journal F447 122(564).
