The impact of the European Central Bank’s
negative policy rates on the bank interest rate
pass-through in the Netherlands.
Thom de Boer 10280669
BSc Thesis Economics and Business, specialization Economics Supervisor: Alex J. Clymo
Statement of Originality
This document is written by Thom de Boer who declares to take full responsibility for the contents of this document.
I declare that the text and work presented in this document is 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 completion of the work, not for the contents.
Abstract
The strength of the monetary transmission mechanism in the eurozone crucially depends on the response of retail bank rates to policy rates, through their effect to market rates. This is particular the case in countries with predominantly bank-based financial systems. Since the Netherlands is such a country with an influential bank-based financial system, it is convenient to monitor the bank interest rate pass-through. This thesis will have a look at the structural and cyclical factors influencing the speed and size of the pass-through in the Netherlands. An error-correction mechanism model will be used to examine whether the negative policy rates of the European Central Bank have an impact on the pass-through.
The results suggest that retail banking rates are less sticky since the first negative policy rates at the European Central Bank’s deposit facility. Likewise, the speed of adjustment to the long-run equilibrium has also increased since the . Other than expected beforehand, the results also indicate that the bank interest rate pass-through has increased for most retail bank rates.
Table of Content
➢ Introduction 4
➢ Literary review:
The monetary transmission mechanism 6
The natural rate of interest and the Taylor principle 12 Factors explaining the bank interest rate pass-through 15 ➢ Empirical Study: Methodology 21 Data summary 23 Results 25 ➢ Concluding remarks 29 ➢ Bibliography 30 Data resources 34
Figures and tables
❖ Figure 1: Equilibrium in the market for reserves 7 ❖ Figure 2: Plots of bank interest rates with their corresponding market rates 24
❖ Table 1: Bank loans and debt securities in the euro area, the UK and the US 17 ❖ Table 2: Monetary financial institutions interest rate pass-through estimates
Introduction
Since the global financial crisis, inflation has been low worldwide, and growth subdued (ECB, 2016). Central bankers are responding to this low inflation and output below potential by conducting accommodative policies. An important question for monetary policymakers, as well as participants in financial markets, is: “Where are interest rates headed?” In the long run, economists assume that nominal interest rates will tend towards some equilibrium, or “natural,” real rate of interest (Williams, 2003). This equilibrium rate is the interest rate that is consistent with stable inflation and output at its potential level. Setting short-term interest rates above this rate puts downward pressure on economic activity and inflation. Setting them below this rate has the opposite effect (ECB, 2016).
Both the cyclical factors and the decline in the long-term real rate of interest, due to the declining growth trend and ageing societies, which exist in many mature economies (ECB, 2017), have required policy rates to be set at all-time lows (ECB, 2016).
The European Central Bank [ECB] sets policy rates and influences short-term market rates with, for example, open market operations, standing facilities and reserve requirements. The purchases of the ECB, open market operations, increases the price of these securities and creates money in the banking system. As a consequence, a wide range of interest rates fall (ECB, 2016), loans become cheaper and businesses and people are able to borrow more and spend less to repay their debts. As a result, consumption and investment receive a boost, which in their turn support economic growth and job creation. These assets purchases can thus support economic growth across the euro area but are mainly used to help inflation levels to return below, but close to, two percent (ECB, 2000).
Conducting this monetary policy typically has a significant and fairly immediate effect on market rates at various maturities (de Bondt, 2005). In particular, changes in policy rates in normal circumstances have been shown to spill over more or less one-to-one to unsecured short-term money market rates, such as the Euro OverNight Index Average [EONIA] and, to a somewhat lesser extent, the term Euro Interbank Offered Rates [EURIBOR] (ECB, 2009). Through its influence on expectations about future policy actions, changes in the monetary policy stance will often also have a strong impact on longer-term markets rates, such as long-term government bond yields and swap rates, by moving the yield curve (de Bondt,
Mojon, & Valla, 2005). Owing to the importance of banks in the euro area financial system and their role in the transmission of monetary policy rates, the bank interest rate pass-through to retail bank rates is a key issue for central banks.
In June 2014, following in the footsteps of the Danish National Bank, the European Central Bank became the first major central bank to lower one of its key policy rates to negative territory (ECB, 2016). Nowadays, the interest rate on the deposit facility of the ECB is -0.4 percent.
Since retail banks are expected and assumed to offer non-negative deposit rates, this raises questions about the impact of the ECB’s negative policy rates on the speed and size of the interest rates pass-through. Because this kind of research has not been done in the Netherlands, my research question will be: “What is the impact of the European Central bank’s negative policy rates on the bank interest rate pass-through in the Netherlands?”
The first part of this paper will be a literary review. This review will incorporate a description of different channels of the monetary transmission mechanism [MTM], the channels through which the money supply affects economic activity. The review will also describe the natural rate of interest and the Taylor rule as well as a detailed explanation of factors influencing the bank interest rate pass-through.
The second part will be an empirical study on the bank interest pass-through in the Netherlands and will monitor if negative policy rates of the ECB had an impact on this pass-through. I will use an error correction mechanism [ECM] model, which is used by the European Central Bank to observe the pass-through for the period January 1997 to June 2007 of 11 broad categories.
In this modelling framework, changes in a specific bank interest rate, ΔBRt, are regressed on simultaneous and lagged changes in a relevant market rate, ΔMRt, and lagged changes in the bank interest rate itself. The results suggest that examined retail banking rates are less sticky since the negative policy rate at the ECB deposit facility. Nevertheless, the results also indicate that the bank interest rate pass-through has increased for most retail bank rates. This finding was not expected but may be due to the beforehand uncertainty about negative official interest rates, as well as the uncertainty of future monetary policy.
Literary Review
The Monetary Transmission Mechanism
The main goal of the ECB is price stability in the medium term (ECB, 2000). Because price levels and the economy affect one another, it is very useful for the ECB to evaluate monetary policy by looking at the most important transmission channels.
The channels through which the money supply affects the economic activity are called the transmission mechanisms of monetary policy. This approach examines the effect of changes in the money supply on economic activity by building a structural model (Mishkin, Matthews & Giuliodori, 2013, p. 526). Mishkin, Matthews and Giuliodori mention that this is a description of how the economy operates using a collection of equations that describe the behaviour of firms and consumers in many sectors of the economy. These equations then show the channels through which monetary and fiscal policy affect aggregate output and spending. A structural model might describe the workings of monetary policy with the following schematic diagram:
M ⇒ i* ⇒ i ⇒ I ⇒ Y
The change in the money supply M affects the key official interest rates i* of the ECB, these interest rates i* influence market interest rates i, which in turn affect investment spending I, which in turn affects aggregate output or aggregate spending Y. Structural model evidence on the relationship between M and Y looks at empirical evidence on the specific channels of monetary influence (Mishkin, Matthews & Giuliodori, 2013, p. 526), such as the link between interest rates and investment spending.
Monetary Policy Tools
Central banks have been signalling the stance of their monetary policy by controlling or targeting the overnight interbank interest rate (de Bondt, 2005). This is the interest rate on loans of reserves from one bank to another. Changes in interbank rates have an impact on the interest rates set by banks on loans and deposits, and therefore play a fundamental role in the way central banks can affect the economy and achieve their goals.
The European Central Bank has several practices to conduct monetary policy. Open market operations, standing (lending and deposit) facilities and reserve requirement will all affect the market for reserves and the equilibrium overnight rate (ECB, n.d., “The Eurosystem’s instruments”).
Open market operations [OMOs] are the most important monetary policy tool. Central banks
buy and sell securities to regulate the money supply in the economy. They are the primary determinants of changes in the interest rates and the monetary base, the main source of fluctuations in the money supply (ECB, n.d., “The Eurosystem’s instruments”). Open market purchases expand reserves and the monetary base, thereby increasing the money supply. When the money supply rises, it will in the short term exceed the money demand. Money becomes relatively cheap and results in lower short-term interest rates. The lower short-term interest rates restore the equilibrium in the money market. Oppositely, open market sales shrink the money supply and increase short-term interest rates.
There are two main types of open market operations. The first are outright sales or purchases of securities in the secondary market (Mishkin, Matthews & Giuliodori, 2013). These outright open market operations permanently drain or add reserve balances. The second type refers to temporary open market operations. A repurchase agreement is such a temporary open market operation (Mishkin, Matthews & Giuliodori, 2013). The central bank purchases securities with an agreement that the seller will repurchase them in a pre-specified period of time. The effects on reserves balances of a repurchase agreement are reversed on the day the agreement matures. When the central bank wants to conduct a temporary open market sale, it engages in a reverse repurchase agreement. The central bank will sell securities and the buyer agrees to sell the securities back at maturity of the agreement.
Standing facilities allow credit institutions, on their own initiative, to either borrow liquidities (until the next morning) from the central banks using the marginal lending facility, or to deposit excess liquidities with their national central banks using the deposit facility (DNB, n.d.). The
lending facilities are operated to provide, against collateral, overnight reserves to banks at a lending rate, typically set above the official overnight interbank rate (DNB, n.d.). The deposit facilities are offered for absorbing overnight reserves of banks wishing to deposit at the central bank at a deposit rate, typically set below the official rate (DNB, n.d.). If the central bank lowers the lending rate, a loan from the central bank becomes more attractive for banks. If the central bank lowers the deposit rate, a commercial banks’ deposit at the central bank becomes less attractive. In general, the central bank’s lending and deposit rate are, respectively, above and below the official overnight interbank rate.
In figure 1, the market for reserves is scheduled. These reserves can be split up in
non-borrowed reserves [NBR] and borrowed reserves. When the overnight rate is above the deposit rate and the overnight rate is decreasing, the opportunity costs of holding excess reserves goes down. Therefore, the demand curve for reserves is downward sloping. Regarding the supply curve, the vertical part of the supply curve starts at a given amount of non-borrowed reserves, this is the amount of reserves that are supplied by the central bank’s open market operations. The horizontal part shows the amount of reserves borrowed from the central bank, priced at the lending rate. If the overnight rate is well in between the lending and deposit rate, most changes in the lending or deposit rate, holding everything else constant, would not affect the overnight rate and no excess reserves would be held.
However, if the overnight rate is above the lending rate, retail banks would borrow from the central bank. And, if the overnight rate is below the deposit rate, retail bank would deposit at the central bank. Therefore, in both cases, as shown in figure 1, when the demand curve intersects the supply curve on its flat section, a change in the lending rate will affect the overnight interbank rate. Likewise, when the supply curve intersects the demand curve on its flat section, a change in the deposit rate will alter the overnight rate.
When central bank requires banks to hold less reserves, the demand for reserves will decrease for any given interest. The fall in the required reserve ratio will thus shift the demand curve to the left. An increase in reserve requirement will increase the demand for reserves and thus shift the demand curve to the right. For this reason, a change in the reserve
Monetary Policy Tools used by the European Central Bank
Of these main broad categories of conducting conventional monetary policy, the ECB uses open market operations as its primary tool (ECB, n.d., “Open market operations”). The Eurosystem’s regular open market operations consist of one-week liquidity-providing operations, the main refinancing operations [MROs], as well as three-month liquidity-providing operations, called the longer-term refinancing operations [LTROs] (ECB, n.d., “Open market operations”). The interest on these refinancing operations is called the
main refinancing rate.The main refinancing operations serve to steer the overnight rate and affect the supply of reserves, to manage the liquidity situation and to signal the monetary stance. The longer-term refinancing operations are not uses for these purposes, but instead are aimed at providing euro area banks access to longer-term funds and thus provide additional, longer-term refinancing to the financial sector (ECB, n.d., “Open market operations”).
Although the ECB has decentralized its open market operations, they are conducted by national central banks, the main refinancing rate is set by the Governing Council of the ECB (ECB, n.d., “Governing council decisions”). In recent years, due to the economic crisis, the
regular operations have been complemented by long-term refinancing operations with a
three-year maturity and targeted longer-term refinancing operations [TLTROs] (ECB, n.d.,
“Open market operations”). The targeted operations provide financing to credit institutions for periods up to four years.
Channels of the Monetary Transmission Mechanism
The channels through which the money supply affects the economic activity are called the transmission mechanisms of monetary policy. As stated above, because price levels and the economy affect one another, it is very useful for the ECB to evaluate monetary policy by looking at the most important transmission channels.
The traditional interest-rate channel can be defined as the channel where an expansionary monetary policy, as a consequence of the ECB doing open market operations, will lead to declines in short-term nominal interest rates. Through the sticky prices, the real long-term interest rates will decline as well (Mishkin, 1995). The decrease in real interest rates will in turn lower the cost of borrowing and cause investment spending to go up. This will result in an increase in aggregate demand and rising output.
The size of the interest-rate channel is, in general, not likely to be equal for every country (Mojon, 2000). The heterogeneity of retail banking markets may contribute to national asymmetries in the interest rate channel. Also the balance sheet structures of households and firms are important. These structures are the primary factors in determining the income and wealth effects of changes in interest rates. These effects will be discussed later.
Aside from the traditional interest-rate channel, conducting monetary policy could also affect economic activity through other assets-price channels.
An example of such an assets-price channel is the exchange rate effect (Mishkin, 1995). In open economies with flexible exchange rates, a monetary expansion would lead to lower real interest rates. These lower interest rates will negatively affect the value of domestic assets compared to other currency assets. Therefore the domestic currency, in our case the euro, will depreciate and the exchange rate, denoted in euros to, for example, US dollars, will go up. Because of the rises in the exchange rate, domestic goods will become relatively cheap. Consequently, there will be an increase in net exports and hence in aggregate output.
In addition, JamesTobin’s q theory explains how monetary policy could have effects on
economic activity through the effects of monetary policy on the value of equities. The suggested theory by Tobin, is that the rate of investment is a function of q, the ratio of the market value of new additional investment goods over their replacement costs (Hayashi, 1981). With a monetary expansion, people will have more money and therefore buy more equities. This will result in higher equity prices for the issuing firms and will consequently increase their market values. The increase in market values will lead to a higher qand thus to higher investment spending, which in turn will lead to an increase in aggregate demand (Mishkin, 1995).
Furthermore, a powerful channel is the income and wealth effect. Modigliani's life cycle hypothesis of consumption (1971) looked at how consumers' balance sheets might affect the spending decisions (Mishkin, 1995). Modigliani's theory most important assumption is that consumers smooth out their consumption over time. Therefore today's consumption depends not only on today's income, but also depends on future earnings. With stock prices being an important factor of consumers' financial wealth and future earnings, a monetary expansion would lead to higher stock prices and thereby increase wealth (Mishkin, 1995). The increase in wealth would lead to a higher consumption and in turn to a higher aggregate demand. This wealth effect could also applied for housing and land prices. If, because of monetary
expansions, prices of these would increase, consumer's wealth would increase and thus leads to a rise in aggregate demand.
Two type of channels of the monetary transmission mechanism arise in credit markets as a result of asymmetric information problems.
Since larger firms are more likely to access the credit market through stocks and bonds than small firms, the bank-lending channel is especially important for small firms (Mishkin, 1995). These will not have access to the credit market unless they borrow from banks. A monetary expansion would increase banks' reserves and lead to more bank deposits. With an increase of bank deposits, more money is available for new loans and investment spending will rise. Bernanke and Gertler (1995) suggest that more bank lending, induced by expansionary monetary policy, should also cause an increase in durable goods consumption and housing purchases for households. Nevertheless because of financial innovation, the importance of the bank-lending channel has probably decreased compared to the 50s, 60s, 70s. This is mostly due to banks playing a less important role in the credit market (Mishkin, 1995).
The second credit channel due to asymmetric information problems is the
balance-sheet channel. Adverse selection and moral hazard are likely to decrease when
lending to firms who's net worth has increased. When expansionary monetary policy would lead to stock prices to rise, the net worth of the borrowing firms would also increase and thus their collateral (Mishkin, 1995). The increase in collateral decreases adverse selection and moral hazard, which in turn would lead to more investment spending and an increase in aggregate demand.
Another balance-sheet channel is called the cash-flow channel. This channel argues that a monetary expansion would decrease the nominal interest rate and, through the nominal interest rate, affect firms' cash flow (Mishkin, 1995). The cash-flow channel thus differs from the traditional interest-rate channel, in which investment is affected by real interest rates. The decrease of nominal interest rates improves the firms' balance sheets because it raises their cash flow. With increased cash flow, liquidity of the firms rises. It will be easier for firms to make payments and thereby lowers adverse selection and moral hazard when lending to these firms (Mishkin, 1995). Investment spending and economic activity would therefore go up.
The natural rate of interest and the Taylor principle
In macroeconomic and monetary theory, the natural rate of interest has played a significant role. The natural or equilibrium real rate of interest is the real interest rate where output equals its natural rate and inflation is constant. Wicksell defined the natural rate of interest in 1898 as:
“a certain rate of interest on loans which is neutral in respect to commodity prices, and tends neither to raise nor to lower them.”
Although the natural rate of interest is not observable and difficult to estimate, monetary policy makers are interested in the natural rate of interest because it provides a benchmark for monetary policy measurement (Williams, 2003). If the short-term real interest rate lies below the natural rate, monetary policy should be contractionary. If it lies above the natural rate, monetary policy should be expansionary (Laubach & Williams, 2003). As well as policy makers, investors are interested because it would be helping to forecast interest rates in the future. These future interest rates forecasts could also be helpful to determine long-term yields of securities.
Most central banks target short-term interest rates, in the eurozone for example the overnight money market rate EONIA, by setting certain policy rates. In view of the natural rate providing a benchmark for most central banks formulating monetary policy, the policy rates can be measured against the natural rate of interest rate directly (Amato, 2005)
A monetary policy rule which easily clarifies what would be desirable policy is the Taylor rule. The Taylor rule links the level of the policy rate to deviations of inflation from its target and the output from its potential (Hofmann & Bogdanova, 2012). According to this rule, the real interest rate exceeds the natural rate when inflation exceeds its target rate and vice versa, all else equal (Laubach & Williams, 2003). It takes the following form:
𝑖 = 𝑟* + 𝜋* + 1.5(𝜋 − 𝜋*) + 0.5 y
where i is the nominal policy rate, r* is the natural rate of interest, π* is the central bank’s inflation objective, π is the current period inflation rate, and y is the current period output gap defined as the percentage deviation of real GDP from its potential level.
Contractionary or expansionary monetary policy changes should occur when variables deviate from their target or potential. In order to follow the taylor rule, the nominal policy rate should go up if inflation is above its target level. Economic activity would, because of the contractionary monetary policy, go down down. The same reasoning holds for the output gap, when there is a negative output gap, the nominal policy rate should be adjusted downwards. The expansionary monetary policy leads to an increase in economic activity.
Implications of the Taylor rule are that the central banks aim at stabilising output around its potential and stabilising inflation around its target level (Hofmann & Bogdanova, 2012). In our case the ECB which aims to maintain interest rates below, but close to, two percent over the medium term (ECB, “Definition of price stability”).
Another implication of the Taylor rule, critical to the success of monetary policy, is the Taylor principle (Hofmann & Bogdanova, 2012). Monetary authorities should raise nominal interest rates by more than the increase in the inflation. For example, to follow the Taylor principle, the interest rate target should be raised by 1.5 percentage points if the inflation rate increases by 1 percentage point. If the Taylor principle would not be followed and the nominal policy rate would be raised by less than the increase in inflation, real interest rates fall. Effectively this means an easing of monetary policy, which will lead to further rises of inflation in the future.
When comparing policy rates with Taylor rates, Hofmann and Bogdanova findings (2012) suggest that monetary policy has been systematically too accommodative for the past few decades. This is especially the case in major advanced economies. They argue that asymmetric reactions of monetary policy to different stages of the financial cycle and global spillovers are possible explanations their findings. Since 2000, advanced economy policy rates did fall strongly and rapidly in the occurrence of financial market crashes, but rose slowly or not at all during the recovery. Concerns about exchange rate movements and unwelcome capital flows could have channeled low interest rates from advanced economies to emerging market economies and other advanced economies (Hofmann & Bogdanova, 2012). Out of these concerns, Gray (2012) reasons that central banks may aim to avoid large and volatile interest rate differentials so that their policy rates become implicitly tied to those prevailing in core advanced economies.
The main advantage of the Taylor rate is that it is relatively simple and easy to compute. On the other hand, monetary policy has long lags, it takes a long time for policy changes to affect the economy. Monetary policy therefore needs to be forward-looking.
Therefore, for central banks to correctly forecast the future inflation and economic activity, other variables, besides the current inflation and output gap, should be looked at as well (Mishkin, Matthews & Giuliodori, 2013). Furthermore, the natural rate of interest and the output gap are difficult to measure (Laubach & Williams, 2003). Since
Because of these difficulties and problems of the Taylor rule, it does not explain all changes in policy rates. It can, however, still be used as a good benchmark (Mishkin, Matthews & Giuliodori, 2013). Whenever policy rates deviate from the Taylor rule, monetary policy makers should ask themselves whether they have good reasons for the deviations of the policy rates. They might be making a mistake if they don’t.
For the ECB, the primary objective is to maintain price stability (ECB, Definition of price stability). By reason of the dual mandate of the Federal Reserve Bank [Fed] in the United States, where they not only aim for price stability but also for maximum employment (Bernanke & Mishkin, 1997), the Taylor rate is a more important benchmark for the Fed than for the ECB.
Factors explaining the bank interest rate pass-through
The adjustment of retail bank rates in response to changes in policy rates plays a key role in the monetary transmission mechanism. Timing and size of the pass-through may considerably affect the strength of the monetary transmission mechanism because a quick and complete pass-through will have a larger impact on borrowing and saving behaviour of firms and households, whereas a slow and incomplete pass-through will make monetary policy less effective (Kwapil & Scharler, 2010). For policymakers this raises the crucial question about likely factors that affect banks’ interest-rate-setting behaviour when monetary policy conditions change. Some theories about factors explaining the pass-through are summed up in the textbook of Mishkin, Matthews & Giuliodori (2013):
According to these theories, the degree of bank competition is a crucial factor for the interest rate pass-through. In a fully competitive banking sector a decrease in the policy interest rate is likely to lower retail bank loan rates relatively quickly. If these banks would not adjust their rates they would lose their market share. When for example a certain bank would be a monopolist, a decrease in the policy interest rate, does not have to lead into a relatively quick decrease of retail loan rates of that certain bank. This certain bank might postpone a reduction in loan rates to increase its profit margin (Mishkin, Matthews & Giuliodori, 2013). Hence less competition among banks may be accompanied with a relatively slow and incomplete interest rate pass-through. Since a decrease in policy rates also imply lower market rates (de Bondt, 2005), less competition among banks and non-bank financing sources may also be accompanied with relatively slow and incomplete interest rate pass-through.
A second factor explaining the speed and size of the interest rate pass-through is the uncertainty about the future course of monetary policy. An uncertain monetary policy may prevent a smooth transmission process (Mishkin, Matthews & Giuliodori, 2013). If a change in monetary policy is believed to be temporary and is likely to be reversed in near future, banks may decide not to react or to adjust only contracts with short maturities.
Also the existence of adverse selection and moral hazard may prevent a bank from an upward adjustment in yields offered on loans (Mishkin, Matthews & Giuliodori, 2013). If banks
increase their lending rates they may attract less solvent borrowers or this increase in lending rates could induce conflicting incentives for borrowers to choose riskier projects (de Bondt, 2005). Since riskier borrowers are more likely to fail on their debt, the expected returns on banks’ loan portfolios may actually fall when banks increase their lending rates. In this case of credit rationing, bank lending rates exhibit upward stickiness (de Bondt, 2005). In order for banks to maximize the expected return on bank loans, banks might choose not to fully adopt official interest rates. Adverse selection and moral hazard are thus possible factors which restrict the completeness of interest rate pass-through.
An other theory is stressed by Sørensen and Werner (2006). They argue that banks’ excess liquidity and excess capital may act as buffer against market fluctuations and would hence be expected to show a negative relation with the speed of adjustment. Contrarily, Sørensen and Werner (2006) also reason that in a competitive banking market, banks’ excess liquidity and excess capital might also be used to offer extra attractive interest rates in order to gain market share. This kind of reasoning would imply a relatively quick interest rate pass-through.
The bank-customer relationship (de Bondt, 2005) might affect the pass-through as well. De Bondt (2005) refers to the paper of Kim, Klinger & Vale (2003) where they present an empirical model of firm behaviour in the presence of switching costs, as well as on customers’ transition possibilities. Switching costs may arise when bank customers consider switching from one bank to another. These switching costs for customers, costs such as administrative costs, acquiring information and studying that information, are potentially important in markets where significant information or transaction costs exist (de Bondt, 2003). Kim, Klinger & Vale expected these costs to be higher in markets with long-term customer relationships. However, even in the presence of small switching costs, Kim, Klinger & Vale predict that the smaller the proportion of customers that are ‘new’ to the market, the less competitive prices will be.
Retail banks experience administrative costs too. These costs, when adjusting their interest rates on new loans and deposits, may exceed the losses of profit keeping when their rates at non-equilibrium levels. Banks’ administrative costs may therefore prevent a quick and complete pass-through (de Bondt, 2005).
Empirical findings on the determinants explaining the interest rate pass-through
Regarding the competition in the banking market, Leuvensteijn, Sørensen, Bikker and Rixtel (2013) their results show that bank interest rate spreads on mortgage loans, consumer loans and short-term loans to enterprises are significantly lower in more competitive markets. Under heavier competition, bank rates tend to be lower which increases social welfare. To compensate for stronger loan market competition, banks lower their deposit rates. Leuvensteijn et al. (2013) also found evidence that, in the long run, bank loan rates are closer in line with market rates where competition is higher. Leuvensteijn et al. (2013) conclude that the results show that bank loan rates will be lower in loan market with stronger competition. Hence bank competition is an important determinant of the interest rate pass-through and thus may also have substantial impact on the monetary transmission mechanism
Concerning the certainty of monetary policy actions, an important conclusion in Kleimeier and Sander their paper (2006) is the establishment of a significant positive impact of the interest rate anticipation on the pass-through in loan markets. More specifically, banks will increase rates faster when they expect increases in market interest rates as opposed to unexpected increases. Given this faster response, Kleimeier and Sander (2006) conclude that a good communication policy is an important feature for the effectiveness of monetary policy.
In addition, as briefly mentioned in the introduction, the transmission of policy rate changes to retail bank loan and deposit rates is a crucial part of the monetary policy transmission mechanism. This is especially true in the euro area which has a predominantly bank-based financial system (ECB, 2009).
In table 1, where data refers to the end of 2009, it is shown that bank loans in the euro area were around 141.5% of GDP, while for the US this ratio was around 55.6%. Additionally, bank loans to non-financial corporations in the euro area were around 52.4% while this ratio was
around 22.0% for the US. Because also debt securities issued by non-financial corporations were around 7.8% in the euro area and around 19.5% in the US, European firms rely much more on bank sources compared to firms in the US. Concluded could be that the euro area has a predominantly bank-based financial system. Same sort of reasoning also applies in the case for the Netherlands in specific (OECD, 2014).
In a working paper of the ECB, Ehrmann, Gambacorta, Martínez-Pagés, Sevestre and Worms (2001) compared the banking systems between Europe and the US and matched the empirical findings with differences in these bank structure. Since various features of European banking markets are significantly different from banking market features of the US, it is not likely that the distributional effects across banks will be alike (Ehrmann et al., 2001). They found that the way banks respond to monetary policy can be explained by the structure of banking market.
One important issue is the relevance of informational frictions in the banking market. Assuming that there are strong informational asymmetries in the interbank markets, it has been shown that a change in interest rates can lead to distributional effects across banks that are informationally vague to a different degree (Ehrmann et al., 2001).
These informational frictions in banking markets may substantially be reduced by an active role of the government in the banking market. Publicly owned or guaranteed banks are unlikely to be at the disadvantage of significant decrease in funds after a monetary tightening. Distributional effects in their loan reactions are hence unlikely to occur (Ehrmann et al., 2001). On the other hand, Ehrmann et al. conclude (2001), weaker publicly owned or guaranteed banks might also extended their loan portfolios despite potential increases in attracting non-creditworthy borrowers. These banks more or less ignore the fact that these non-creditworthy borrowers are more likely to fail on their debt, because the publicly owned and guaranteed banks will be prevented from bankruptcy by the government. Overall, Ehrmann et al. (2001) expect the consequences of informational frictions to be much more important in the US than they are in the euro area.
Empirical results of the interest rate pass-through in the euro area and the Netherlands
Given the above discussion, the list of factors explaining the interest rate pass-through is quite extensive. Since there is uncommonly little empirical evidence of the interest rate pass-through in the Netherlands, it is interesting what the empirical evidence on the interest rate pass-through is for the euro area.
Money market rates, such as EONIA, sharply follow changes in the official interest rates set by the ECB (ECB, 2009). As mentioned in the introduction, also De Bondt (2005) claims that monetary policy typically has a significant and fairly immediate effect on market rates at various maturities. In particular, changes in policy rates in normal circumstances will approximately spill over one-to-one to unsecured short-term money market rates. For example, the EONIA and, to a somewhat lesser extent, the term EURIBORrates (ECB, 2009). In the paper of de Bondt (2005), he examined the interest rate pass-through from policy rates to market rates. He used, like I will, an error correction model. He used his model to monitor the pass-through for time periods between 1996 to 2001. Concerning the immediate pass-through of market interest rates to retail bank interest rates, De Bondt (2005) concludes it is incomplete. The proportion of a given market interest rate change that is passed through within one month is found, at its highest, to be around 50%. In the longer term, the pass-through is higher. This is especially the case for longer-term lending rates with interest rate pass-throughs close to 100%. De Bondt (2005) also concludes that the most sticky retail bank interest rates are the interest rates on overnight deposits and deposits at notice up to three months.
This sluggishness of short-term deposits is also found by Sander and Kleimeier (2004a,b). Since pass-through analyses are also used to investigate the degree of competition or imperfection in the banking market, Sander and Kleimeier (2004a) argue that their findings for short-run rigidity of banking product prices indicates imperfectly competitive banking markets. With a marginal cost pricing model, the bank interest rate is set by banks depending on a marginal cost price approximated by a market interest rate. In their model, they argue that the size of the pass-through depends on the demand elasticity for banking products, such as deposits and loans. A not fully demand elasticity would imply an imperfect interest rate pass-through, which could therefore be seen as a results of imperfect competition (Sander & Kleimeier, 2004a).
Like de Bondt (2005), the paper of Kwapil and Scharler (2010) suggest that the interest rate pass-through for the euro area is not only slow but incomplete too. It usually takes several months for retail rates to adjust, in particular deposit rates.
The little evidence for the Netherlands in specific can be found in the paper of Sander and Kleimeier (2004b) and Marotta (2008). They find, for short-term lending rates, a short-run pass-through of 40%. This means that a proportion of 40% of a given market interest rate change is passed through within one month. The final proportion of a given market interest rate change that is passed through, was found to be 99% for the time period September 1997 to October 2002. Compared to the findings regarding the euro area (de Bondt, 2005), these findings for the Netherlands are to a certain extent in line with the findings of de Bondt.
Empirical Study
Methodology
This second part of my thesis will be an empirical study on the bank interest pass-through in the Netherlands and will monitor if the negative policy rates of the ECB had an impact on this pass-through.
According to the bank interest rate pass-through literature, the transmission of monetary policy rates, via changes in market rates, to bank interest rates can be modelled using an error correction modelling framework (ECB, 2009). In this modelling framework, changes in a specific bank interest rate, ΔBRt, are regressed on simultaneous and lagged changes in a relevant market rate, ΔMRt, and lagged changes in the bank interest rate itself. An error correction term is included reflecting the extent to which the bank rate had diverged from it long-term equilibrium relationship with the market rate in the previous period. This approach, also used by the ECB (2009) takes the following form:
ΔBRt = ϑ + γ(BRt-1 - βMR t-1 - 𝜅) + α1ΔMR t + α2ΔMRt-1 + ηΔBRt-1 + εt
α1 = immediate pass-through; the proportion of a given market interest rate change that is passed through within one month
β= final pass-through
γ = speed of adjustment to the long-run equilibrium
𝜅 = can be interpreted as representing all other factors, apart from the market rate, that determine the level of the bank rate
An OLS-regression will be run in STATA where I regress ΔBRt on BRt-1, MRt-1, ΔMRt, ΔMRt-1 and ΔBRt-1, corrected for White heteroscedasticity-consistent standard errors by using the
“robust” option in STATA. When performing the regression this way, γ will be found. Likewise an estimate for the coefficient of MRt-1 will be found. When this estimate is divided by -γ, β is found. In this paper I will mainly be looking at the core estimates γ, β and α1.
To determine whether the bank interest rate pass-through has been influenced by the negative policy rates of the ECB deposit facility, I will use the error correction modelling
framework for a period of 35 months before the first implementation of a negative ECB deposit rate, and 35 months after. To be more precise, the first implementation of a negative policy rate was in June 2014, therefore the specific time periods are July 2011 to May 2014 and June 2014 to April 2017.
Data summary
I will be focussing on the overnight deposits and deposits with an agreed maturity up to one year and will differentiate between the group households and non-profit institutions and the group non-financial corporations.
The first reason why I will be focussing on these relatively short-term deposits, is because only the deposit rate at the ECB is negative, -0,40%. At this moment, the main refinancing interest rate and the lending rate at the ECB are 0,00% and 0,25% respectively. The second reason is that it might be more interesting to monitor just the short-term rates, because, in the paper of Demertzis and Wolff (2016), a survey among professional forecasters expect inflation and market interest rates to be at normal levels again in five years time. If that would be the case, as mentioned above, banks may decide to adjust only contracts with short maturities.
Since banking rates will be influenced by corresponding market rates, for this modelling framework to make sense, it is essential to regress the banking rates on relevant market rates.
For the overnight deposit banking rates, the EONIA rate would be the most relevant market rate. The most relevant market rate for banking rates on deposits with an agreed maturity up to one year, is the EURIBOR 12-months rate. Plots of these specific bank rates [BRt] and their correspondent market rates [MRt] can be found below. The vertical axes of the plots represent the specific bank- and market interest rates. The horizontal axes of the plots represents the 70 month timeframe from July 2011 (left) to April 2017 (right). All these banking and market rates can be found online at the statistical data warehouse of the ECB . 1
❖ Figure 2: Plots of bank interest rates with their corresponding market rates
correlation: 89,0% correlation: 95,2%
correlation: 96,3% correlation: 84,6%
With the above stated method and data, I expect the interest pass-through to be lower in the time period June 2014 to April 2017. Mostly because retail banks are expected and assumed to offer non-negative deposit rates. Because of this assumption, policy rates, through market rates, can’t influence retail banking rates to the usual extent when the policy rates are negative. Therefore, the bank interest rate pass-through is expected to be lower.
Results
❖ Table 2: Monetary financial institutions interest rate pass-through estimates based on the error correction model
Notes: : “*”, “**” and “***” indicate significance at the 90%, 95% and 99% level respectively. Corrected for White heteroscedasticity-consistent standard errors
Observing the interest rate pass-through in general
The results of the immediate pass-through estimates are, to some extent, in line with the earlier conducted research by de Bondt (2005) and Sander and Kleimeier (2004a). Both papers conclude relatively sticky retail bank rates. De Bondt found that the proportion of a given market interest rate change passed through within one month is, at its highest, to be around 50%. For the overnight deposits for household and non-profit institutions and non-financial corporations, I find immediate pass-throughs of 4,8% and 39,8% respectively. The immediate pass-through for deposits with an agreed maturity up to one year for non-financial corporations is 56,3%. A rather unlikely estimate is the immediate pass-through for deposits with an agreed maturity up to one year for households and non-profit institutions. Because of the estimate of -42,9%, when a given market rate decrease with 1%, banks will in general not respond by increasing their retail rates by 0,429%. A negative pass-through has not been found in any of the above stated papers, but could be explained by the lack of significant, when looking at the estimates of the immediate pass-through.
Regarding the interest rate final pass-through, de Bondt (2005) and Kwapil and Scharler (2010) found that this final pass-through is higher than the immediate pass-through, but is still incomplete. Only for longer-term lending rates the final interest rate pass-through is found to be complete (de Bondt, 2005). The results I found are more or less in line with the findings of de Bondt (2005) and Kwapil and Scharler (2010). For all but deposits with an agreed maturity up to one year for non-financial corporations, the final pass-through is higher than the immediate pass-through. This means banks are likely to adjust their retail rates to follow market rates also after one month, which is incorporated in the immediate pass-through. Though, the final pass-through is incomplete for all examined retail bank rates. For the overnight deposits for household and non-profit institutions and non-financial corporations, the results indicate final pass-throughs of 53,8% and 69.7% respectively. The results for overnight deposits are more apart. The final pass-through for overnight deposits for households and non-profit institutions is relatively low, 35,1%, compared to all others. Yet, the final pass-through for overnight deposits for non-financial corporations is relatively high, 87,8%. This results suggest that 87,8% of the change in a specific market rates is passed through to retail bank rates. This estimate is the most complete final pass-through for the examined retail bank rates.
How the negative interest rate affected the interest rate pass-through
Previously to the first implementation of negative deposit policy rates, the proportion of a given market interest rate change that is passed through within one month is found to be higher than 50% for deposits with an agreed maturity up to one year. This does not suggest sticky retail bank rates. The results do indicate relatively sticky retail bank rates for all other but deposits with an agreed maturity up to one year for households and non-profit institutes. Since the implementation of negative deposit policy rates, the immediate pass-through estimates of June 2014 to April 2017 indicate less sticky retail bank rates for overnight deposits. Though, the immediate pass-through for deposits with an agreed maturity up to one year for households and non-profit institutes fell. Again, looking at these estimates of the immediate pass-through, there is a lack of significance. This lack could be a reason for the relatively extreme -0,918 estimate at the deposits with an agreed maturity up to one year for households and non-profit institutes.
The speed of adjustment to the long-run equilibrium has been substantially influenced by the negative policy rates. Whereas for overnight deposits for non-financial corporations, previously to the negative policy rates, the speed of adjustment was -0,125. Since the negative policy rates, this estimate is -0,349. This estimate means the model corrects its previous period disequilibrium at a speed of 34,9% monthly to return to the long-run equilibrium. The same reasoning also holds for deposits with an agreed maturity up to one year. Nevertheless, the speed of adjustment toward the long-run equilibrium for overnight deposits for households and non-profit institutes has gone down.
Looking at the interest rate final pass-through for the examined relative short-term deposits in the time period June 2011 to May 2014, the result suggest the final pass-through is far from complete. The negative policy rates did, however, influence the final interest rate pass-through, but not in the direction I expected the pass-through to be influenced in. Most pass-throughs have turned out to be higher since the implementation of negative interest rates at the ECB deposit facility.
An explanation could be that before the implementation of negative policy rates, retail banks might have assumed and expected that these ECB deposit rate would not turn negative. Therefore, they might not have adjusted their retail bank rates to a relatively full extent expecting policy rates to be still positive. Nowadays, uncertainty about future monetary policy in the euro area may play its part. Uncertainty about for example the exit strategy of an asset
purchase program like quantitative easing (Mishkin, Matthews & Giuliodori, 2013, p. 547), and with the knowledge that key official interest rate are able to turn negative, retail banks might follow up policy rates to a higher degree. Market participants expect policy rates to remain low in the future.
An other explanation for higher interest rate pass-through could be that in most cases the variances of the market rates were significantly higher than the variances of the retail banking rates. Further research should be needed.
Noticed should be that, since regression analysis is a statistical technique, correct application of the statistics is required. Many observations are need for statistical power or efficiency. The lack of significance of the immediate pass-through estimates is likely to be due to the number of observation run in the regressions. On the other hand, when for example monetary policy is actively managed, coefficients could be quite volatile and change over time. A relatively small sample size is required (Bodie, Kane & Marcus, 2014, Chapter 24). Essentially this is a trade-off between bias and efficiency. Nonetheless, Ang and Kristensen (2012) derived an optimal sample size using advanced statistical techniques. They find, for using monthly data, sample sizes between 1,5 and 8,5 years to be optimal. Using a sample size of 35 months should therefore be accurate.
Concluding remarks
Nowadays, interest rates in the euro area are at historic lows. Central bankers are responding to this low inflation and output below potential by conducting accommodative policies. Because of the assumed long-run equilibrium interest rate, the central bankers vital question is where interest rates are headed. By adjusting lending and deposit rates, open market transactions and reserve requirements, the ECB tries to affect market rates. The main objective with altering these market rates, is to affect inflation to be below, but close to, two percent. Moreover, through monetary policy mechanism channels, economic activity is also influenced when conducting monetary policy.
Since the Netherlands also have a bank-based financial system, the dependence of the monetary transmission mechanism on the response of bank rates to changes in policy rates is crucial. Hence, monitoring the interest rate pass-through is convenient.
With the use of an error correction model the results suggest that examined retail banking rates are less sticky since the negative policy rate at the ECB deposit facility. The speed of adjustment to the long-run equilibrium has increased too. Nevertheless, the results also indicate that the bank interest rate pass-through has increased for most retail bank rates. This finding was not expected but may be due to the beforehand uncertainty about negative official interest rates, as well as the uncertainty of future monetary policy.
Bibliography
Amato, J. D. (2005). The Role of the Natural Rate of Interest in Monetary Policy. SSRN
Electronic Journal. doi:10.2139/ssrn.699221
Ang, A., & Kristensen, D. (2012, 10). Testing conditional factor models. Journal of Financial
Economics, 106(1), 132-156. doi:10.1016/j.jfineco.2012.04.008
Bernanke, B., & Gertler, M. (1995). Inside the black box: the credit channel of the monetary policy transmission. Journal of Economic Perspectives, 9(4), 27-48
Bernanke, B., & Mishkin, F. (1997, 01). Inflation Targeting: A New Framework for Monetary Policy? doi:10.3386/w5893
Bodie, Z., Kane, A., & Marcus, A. J. (2014). Investments - Global Edition. London: McGraw Hill Higher Education.
de Bondt, G. J. (2005). Interest Rate Pass-Through: Empirical Results for the Euro Area.
German Economic Review, 6(1), 37-78.
Demertzis, M., & Wolff, G. B. (2016). The effectiveness of the european central bank’s asset purchase programme. Bruegel policy contribution, 10.
DNB. (n.d.). Monetary policy instruments. Retrieved June 24, 2017, from
https://www.dnb.nl/en/interest-rates-and-inflation/monetary-policy/monetary-policy -instruments/
Ehrmann, M., Gambacorta, L., Martínez-Pagés, J., Sevestre, P., & Worms, A. (2001). Financial systems and the role of banks in monetary policy transmission in the euro area.
Monetary Policy Transmission in the Euro Area, 235-269.
doi:10.1017/cbo9780511492372.018
European Central Bank. (n.d.). Governing council decisions. Retrieved July 11 , 2017, from https://www.ecb.europa.eu/press/govcdec/html/index.en.html
European Central Bank. (n.d.). The Eurosystem’s instruments. Retrieved July 11 , 2017, from https://www.ecb.europa.eu/mopo/implement/html/index.en.html
European Central Bank. (n.d.). Definition of price stability. Retrieved June 26, 2017, from https://www.ecb.europa.eu/mopo/strategy/pricestab/html/index.en.html
European Central Bank. (n.d.). How quantitative easing works. Retrieved June 24, 2017, from https://www.ecb.europa.eu/explainers/show-me/html/app_infographic.en.html European Central Bank. (n.d.). Objective of monetary policy. Retrieved June 24, 2017, from
https://www.ecb.europa.eu/mopo/intro/objective/html/index.en.html
European Central Bank. (n.d.). Open market operations. Retrieved June 24, 2017, from https://www.ecb.europa.eu/mopo/implement/omo/html/index.en.html
European Central Bank. (2000, July). Monetary Policy Transmission in the Euro Area. Monthly
bulletin, July 2000, 43-58.
European Central Bank. (2009, August). Recent development in the retail bank interest rate pass-through in the euro area. Monthly bulletin, August 2009, 93-105.
European Central Bank. (2016, July 28.). Assessing the implications of negative interest rates. Retrieved June 24, 2017, from
https://www.ecb.europa.eu/press/key/date/2016/html/sp160728.en.html European Central Bank. (2017, February 10). Behaving responsibly in a low interest rate
environment: A central banker's perspective. Retrieved June 24, 2017, from https://www.ecb.europa.eu/press/key/date/2017/html/sp170210.en.html Gray, C. (2012, 12). Responding to a Monetary Superpower: Investigating the Behavioral
Spillovers of U.S. Monetary Policy. Atlantic Economic Journal, 41(2), 173-184. doi:10.1007/s11293-012-9352-0
Hayashi, F. (1981). Tobin’s marginal q and average a: A neoclassical interpretation. Discussion
Hofmann, B., & Bogdanova, B. (2012). Taylor rules and monetary policy: a global “Great Deviation”? BIS Quarterly Review, September 2012, 37-49.
Kim, M., Kliger, D., & Vale, B. (2003, 01). Estimating switching costs: The case of banking.
Journal of Financial Intermediation, 12(1), 25-56. doi:10.1016/s1042-9573(02)00005-0
Kwapil, C., & Scharler, J. (2010) Interest rate pass-through, monetary policy rules and macroeconomic stability. Journal of International money and Finance, 29, 235-251 Kleimeier, S., & Sander, H. (2006, 07). Expected versus unexpected monetary policy impulses
and interest rate pass-through in euro-zone retail banking markets. Journal of Banking &
Finance, 30(7), 1839-1870. doi:10.1016/j.jbankfin.2005.07.006
Laubach, T., & Williams, J. C. (2003). Measuring the natural rate of interest. The Review of
Economics and Statistics, 85(4), 1063-1070.
Leuvensteijn, M. V., Sørensen, C. K., Bikker, J. A., & Rixtel, A. A. (2013, 04). Impact of bank competition on the interest rate pass-through in the euro area. Applied Economics,
45(11), 1359-1380. doi:10.1080/00036846.2011.617697
Mishkin, F. S. (1995). Symposium on the Monetary Transmission Mechanism. Journal of
Economic Perspective, 9, 3-10.
Mishkin, F. S., Giuliodori, M., & Matthews, K. (2013). Economics of money, banking and financial markets. Harlow: Pearson.
Modigliani, F. (1971). Monetary policy and consumption. Consumer Spending and Money Policy: The Linkages. Boston: Federal Reserve Bank, 9–84
Marotta, G. (2008).Structural breaks in the lending interest rate pass-through and the euro.
Economic Modelling, 26, 191–205.
Mojon, B. (2000). Financial structure and the interest rate channel of ECB monetary policy.
ECB Working Paper, 40.
doi:10.1787/eco_surveys-nld-2014-en
Sander, H., & Kleimeier, S. (2004a). Convergence in euro-zone retail banking? What interest rate pass-through tells us about monetary policy transmission, competition and integration. Journal of international money and Finance, 23, 461-492.
Sander, H., & Kleimeier, S., (2004b). Convergence in Euro-Zone Retail Banking? What Interest Rate Pass-Through Tells Us About Monetary Policy Transmission, Competition and Integration. LIFE Working Paper No 04-005. University of Maastricht, Maastricht Sørensen, C. K. & Werner, T. (2006). Bank interest rate pass-through in the euro area: a cross
country comparison, ECB Working Paper, 580.
Werden, G. J. (1998). Using the Herfindahl–Hirschman index. Applied Industrial Economics, 368-374. doi:10.1017/cbo9780511522048.021
Wicksell, K. (1907, 06). The Influence of the Rate of Interest on Prices. The Economic Journal,
17(66), 213. doi:10.2307/2220665
Williams, J. C. (2003, October 31). The Natural Rate of Interest. Retrieved June 24, 2017, from http://www.frbsf.org/economic-research/publications/economic-letter/2003/october /the-natural-rate-of-interest/
Data resources
❖ Figure 1: Equilibrium in the market for reserves
source: Mishkin, Matthews & Giuliodori, 2013, p. 328, Figure 15.1
❖ Table 1: Bank loans and debt securities in the euro area, the UK and the US source: Mishkin, Matthews & Giuliodori, 2013, p. 543, Table 23.1
➢ Money market rates:
○ Euro OverNight Index Average
source: Statistical Data Warehouse European Central Bank
http://sdw.ecb.europa.eu/quickview.do?SERIES_KEY=143.FM.M.U2.EUR.4F. MM.EONIA.HSTA
○ Euro Interbank Offered Rate 12-Month rate
source: Statistical Data Warehouse European Central Bank
http://sdw.ecb.europa.eu/quickview.do?SERIES_KEY=143.FM.M.U2.EUR.RT. MM.EURIBOR1YD_.HSTA
➢ Retail bank rates for the Netherlands:
○ Overnight deposits for households and non-profit institutes source: Statistical Data Warehouse European Central Bank
http://sdw.ecb.europa.eu/quickview.do?SERIES_KEY=124.MIR.M.NL.B.L21.A .R.A.2250.EUR.N
○ Overnight deposits for non-financial corporations
source: Statistical Data Warehouse European Central Bank
http://sdw.ecb.europa.eu/quickview.do?SERIES_KEY=124.MIR.M.NL.B.L21.A .R.A.2240.EUR.N
○ Deposit with an agreed maturity up to one year for households and non-profit institutes
source: Statistical Data Warehouse European Central Bank
http://sdw.ecb.europa.eu/quickview.do?SERIES_KEY=124.MIR.M.NL.B.L22.F .R.A.2250.EUR.N
○ Deposit with an agreed maturity up to one year for non-financial corporations
source: Statistical Data Warehouse European Central Bank
