Does ING Bank benefit from perceived government support? : an empirical analysis of government guarantees in 2017

(1)

Master Thesis

MSc Economics

Monetary Policy and Banking

Does ING Bank benefit from perceived government

support?

An empirical analysis of government guarantees in 2017

Kristín Arna Björgvinsdóttir

Student number 11665873

Instructor: Prof. Sweder van Wijnbergen

Faculty of Economics and Business

(2)

Statement of Originality

This document is written by Kristín Arna Björgvinsdóttir 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 completion of the work, not for the contents.

Preface

The thesis Does ING Bank benefit from perceived government support? is a 15 ETCS credit project as a final assignment for the authors MSc degree in Economics at the Faculty of Economics at the University of Amsterdam.

The author wants to use this opportunity to express gratitude to the instructor of this thesis Prof. Sweder van Wijnbergen and to her parents for proof-reading and advice.

Abstract

This thesis investigates whether ING Bank, the largest bank in the Dutch financial system, benefits from government guarantees arising from market participants’ expectation of government support to the institution in times of insolvency. Furthermore, in this thesis, the value of this perceived government support to ING Bank in 2017 is quantified. The government guarantees examined in this thesis are the effective tax subsidy of systemically important institutions, the Too Big to Fail status and the fair value of deposit insurance to banks. To examine these guarantees for ING Bank in 2017 the following papers were utilized: Groenewegen, Mosch and Wierts (2016) to estimate the effective tax subsidy. Acharya, Anginer and Warburton (2013) to determine the funding cost reduction of Too Big to Fail institution with ordinary least squares regression analysis and Nobuyuki Oda (1999) explanation of the Merton Method but the method builds upon Merton’s 1977 paper where the Black-Scholes framework is adopted to estimate fair insurance rates for deposit insurance using option pricing theory. The main findings of this thesis suggest that ING Bank benefited from perceived government support in 2017 by EUR 38,360 million or approximately eight times the institution's profit in the same year.

(3)

Figure

Figure 1: The development of the effective tax subsidy of ING Bank in EUR million per year during 2013 to 2017 ... 21 Figure 2: The development of the fair deposit insurance rate for ING Bank in EUR million per year during 2013 to 2017 ... 30 Figure 3: Merton's distance to default and probability of default of ING Bank over the period 2013 to 2017 ... 43 Figure 4: Goodness of fit of individual variable in the Too Big to Fail analysis ... 43

Tables

Table 1: Results of the Too Big to Fail regression analysis with size dummy variables... 26 Table 2: Comparison between the estimated value of government guarantees enjoyed by ING Bank and the banks profit in 2017 ... 31 Table 3: Financial institutions in Too Big to Fail sample, ranked by total assets. (Annual reports of the sample financial institutions) ... 44 Table 4: Correlation matrix of Too Big to Fail variables ... 44

(5)

1 Introduction

The global financial crisis of 2008 brought to light that major financial institutions do benefit from their systemic importance both to the global and local financial systems. This benefit arises from the expectation of market participants that in case of severe distress or insolvency concerns the systemically important institutions are expected to receive assistance. This expectation of support thus allows major financial institutions to benefit from government guarantees and for example, enjoy lower funding cost than the institutions fundamental would suggest.

Financial institutions deemed systemically important benefit in various ways from the perceived government support. This thesis focuses on three implicit and explicit government guarantees for ING Bank in 2017: the Too Big to Fail status, the deposit insurance system and the beneficial tax treatment of excessive leverage. The Too Big to Fail status of a major financial institution is an implicit guarantee because authorities don’t have an explicit commitment to intervene in times of distress although assistance from authorities would be expected. Furthermore, a vast number of countries including the Netherlands have in place explicit deposit insurance schemes, but such a system secures the deposits of customers of a financial institution and thereby secures the institutions funding. The beneficial tax treatment of leverage encourages a significant part of debt financing instead of issuing equity for major financial institutions. However, excessive debt financing makes financial institutions weaker in times of trouble and has a negative externality of the financial system as a whole by making the system more vulnerable in the case of financial distress. These government guarantees affect risk-taking decisions of large financial institutions because of the moral hazard problem that entails such guarantees since important institutions can take on increased risk at limited cost.

The primary purpose of this thesis is to examine these implicit and explicit guarantees for ING Bank and then to quantify the value of these guarantees for the operating year 2017. Thus, shedding light on to what extent the bank benefits from the negative externality introduces by its potential failure to the Dutch financial system. ING Bank is the largest bank in the Netherlands if measured by total assets (TheBanks.eu, n.d) and thus the bank is an important financial institution in the Dutch financial system. Furthermore, ING Bank is the only major bank in the Netherlands that is listed and has the necessary information available

(6)

for the analysis performed in this thesis. ING Bank N.V provides various banking services and operates both in the Netherlands and internationally, but ING Bank N.V. is a subsidiary of ING Group N.V. The principal activities of the bank are providing payments and cash management, corporate finance, treasury services etc. Furthermore, ING Bank serves individuals, governments, institutions and corporates.

The following three academic papers were applied as a basis for estimating the value of the government guarantees. To estimate the value of the Too Big to Fail status of ING Bank in the Dutch financial system I utilise a regression model put forward by Acharya, Anginer and Warburton (2013). With ordinary least squares regressions, the reduction in funding cost for different sizes of financial institutions was estimated. One of the regression analysis yields the reduction in funding cost -benefiting ING Bank in 2017. To estimate the fair insurance rate for the deposit insurance scheme that ING enjoys I followed Nobuyuki Oda (1999) explanation of the Merton Method, but the method builds upon Merton’s 1977 paper and adopts Black-Scholes option pricing theory for the calculation of fair insurance rates. I analysed the fair insurance rate for ING Bank for the year 2013 to 2017 and compared the estimated fair insurance rate to the actual contributions of the bank to the pre-funded deposit guarantee scheme in the Netherlands. To estimate the value of the effective tax subsidy ING Bank benefitted in 2017 I follow the reasoning and estimation method put forward by Groenewegen, Mosch and Wierts (2016). In their paper, the authors compare two different cases of tax deductibility of interest payments to compute the effective tax subsidy that arises for major financial institutions. Systemically essential institutions have incentives to finance with debt because interest payments are tax deductible while dividend payments are not deductible from earnings before tax.

The literature on implicit and explicit government guarantees like the ones covered in this thesis is extensive. Nevertheless, this thesis adds to the existing literature by estimating the monetary value of those guarantees for a particular bank on an annual basis. Developing a firm-specific analysis of the benefit of perceived government support for a single bank shows to what extent a large institution gains from its importance to the financial system without appropriately compensating taxpayers. Furthermore, by estimating the value of these guarantees on an annual basis allows for comparison to be made between the calculated amount of the perceived government support and the bank's profit in the same

(7)

year. Thus, hopefully, this thesis introduces a new perspective on how to estimate government guarantees for a single bank.

This thesis is organised as follows. Section 2 presents a literature review of the current state in the literature of the government guarantees covered in this thesis. Section 3 describes what data and information were employed in the estimations in this thesis and from where the data attained. Section 4 introduces the methodology used in estimating the three government guarantees individually. Section 5 provides the results from the quantification of the value of each guarantee. Then section 6 demonstrates how the results attained in section five compares to the operating profit of ING Bank in 2017. Finally, section 7 offers conclusions drawn from the estimation results.

2 Literature review

The literature on implicit and explicit subsidies is vast and has grown substantially since the Global Financial Crisis of 2008. The following are a few examples of the numerous research done on the subject.

Acharya, Anginer and Warburton (2013) used credit spreads of financial institutions to measure funding cost advantages enjoyed by institutions deemed Too Big to Fail and estimated the total value of the implicit subsidy. They determined that the expectation of state support to be embedded in the credit spread on bonds issued by large U.S. financial institutions. Furthermore, the authors conclude that this expectation of public support creates an implicit subsidy to systemically important institutions, permitting them to borrow at a government-subsidised rate.

The International Monetary Fund (2014) explored a wide range of techniques for estimating the implicit funding subsidy to systemically important banks in the Funds’ 2014 Global Financial Stability Report. The methods included comparing bond spread differential, a contingent claim analysis (CCA approach) and a ratings-based approach. The primary results from the study were that the Too Big to Fail subsidies remain substantial but have declined in most countries from their crisis peaks in part due to financial reforms put forward since the financial crisis. Furthermore, based on the funding cost advantages from the CCA approach and the ratings-based method the dollar values of the implicit subsidies could be calculated. The subsidy values obtained from the CCA approach from the period 2011 to 2012 showed

(8)

the support to be around $50 billion for the United States, $110 billion for Japan and the United Kingdom and above $300 billion for the Euro area.

Merton, Tsesmelidakis and Schweikhard (2014) estimated the value of the implicit Too Big to fail guarantees extended to the financial sector during the global financial crisis. They determined that debt security issued by an institution considered unlikely to default on its liabilities than under normal conditions would have a lower spread over the risk-free rate, but this funding cost advantage could be interpreted as a benefit. By comparing bond characteristics and prices in the primary and secondary market, the authors investigated how the reduced debt capital costs affect the position of creditors and shareholders of a financial institution.

Groenwegen, Mosch and Wierts (2016) developed an analytical framework that takes into account the incentive effects of corporate tax deductibility of interest expenses and not dividend payments on a bank’s decision to finance with debt or issuing equity. Since the tax base decreases when debt financing is chosen over equity financing this constitutes in effect a tax subsidy on debt. Furthermore, they estimated the nominal size of these tax subsidies and their findings suggest that the net support falls in the range of 0,5 to 0,8 percent of GDP in recent years. However, the volume of the implicit tax subsidy has been on the decline since 2009 due to lower interest rates and the introduction of bail-in regimes.

In an IMF working paper De Mooij, Keen and Orihara (2013) explored how corporate tax systems favour debt over equity finance and how this potentially amplifies risks to financial stability. They examined the relationship between the tax bias and the likelihood of financial crisis erupting. Their findings suggest that debt bias leads to a significantly higher aggregate bank leverage and the higher aggregate bank leverage increases the likelihood of financial crisis occurring.

Robert Merton pioneered a model to estimate the value of deposit insurance using option pricing theory with the Black-Scholes framework. In Merton (1977) the author explained that the model is based upon the similar relationship between a third-party guarantee on deposits and common stock put options. In the framework, the promised payment on debt corresponds to the exercise price and the value of the firm’s assets corresponds to the underlying asset. An example of such a third-party guarantee is deposit

(9)

insurance where a government guarantees customers deposits stored at a financial institution to keep its face value in case of insolvency of the institution.

Numerous scholars have since the publication of Robert Merton’s 1977 paper put forward models to estimate the fair deposit insurance rate with several modifications. For example, both Marcus and Shaked (1984) and Ronn and Verma (1986) put forward models to evaluate deposit insurance for financial institutions using the option pricing theory.

3 Data

In estimating the volume of each guarantee enjoyed by ING Bank, various data and information had to be utilised. For the analysis on the tax treatment of excessive leverage in section 4.1 both information from ING´s annual reports for the operating years, 2013 to 2017 and the national corporate income tax rate were used (Government of the Netherlands, n.d).

For the analysis on the Too Big to Fail implicit government subsidy in section 4.2 numerous variables had to be estimated for each of the financial institutions in the sample. Table three in the Appendix shows the financial institutions in the sample as well as the size of their total assets. Information on bonds features was retrieved from the institution's websites or stock exchanges. The stock exchanges that were used to attain information are the Boerse Stuttgart Stock Exchange, the Luxembourg Stock Exchange and the Irish Stock Exchange. Firm-specific information was retrieved from annual and quarterly reports located on the pertaining institution's websites. Share price information for the institutions in the sample was attained from the firm’s webpages and Yahoo Finance. Information on the macro control variables, market risk premium and the yield difference between a short-term and long-term government bond, were retrieved from Fernández, P., Pershin, V., & Acín, I. F. (2017) paper and the OECD database respectively. The default risk premium as measured by the yield spread between maturity matched BAA rated and AA rated corporate bonds by Moody’s. Information required to estimate the default risk premium was retrieved from the Markets Insider webpage. Furthermore, for the Too Big to Fail analysis Merton's distance-to-default had to be calculated for each financial institution in the sample in 2017. A detailed explanation of the procedure and required variables to estimate the Merton's distance-to-default is presented in the Appendix. Furthermore, the yield on government bonds in the analysis was retrieved from the European Central bank.

(10)

The analysis of the value of deposit insurance scheme in section 4.3 required various information from ING´s annual and quarterly reports for the years 2013 to 2017 to estimate the market value of the bank's assets during the period as well as the volatility of those assets. Also, in the following analysis, the risk-free interest rate is the deposit rate of the European Central Bank, the interest rate at which institutions may at all times place overnight deposits with the national central bank. Furthermore, to calculate the variance of the value of ING´s assets information about ING´s share performance was utilised. Information about the share performance of ING´s shares was retrieved from the bank's webpage.

4 Methodology

This section explains the methodology behind the three implicit and explicit government guarantees examined in this thesis. Firstly, section 4.1 defines the effective tax subsidy that systemically important financial institutions enjoy. Then, section 4.2 illustrates how financial institutions that are systemically important to either local or global financial systems benefit from their Too Big to Fail status. Lastly, section 4.3 describes how to estimate the fair deposit insurance rate for depository institutions to examine to what extent institutions benefit from explicit deposit insurance schemes.

4.1 The tax treatment of excessive leverage of financial institutions

One of the various ways large financial institutions like ING Bank benefit from the presence of government guarantees is the favourable tax treatment of excessive leverage. During the global financial crisis, ING like many systemically important banks received support from the national government revealing that their government effectively guarantees the bank. Due to this implicit guarantee debt financing is cheaper for these banks than their fundamentals would suggest. Furthermore, the tax bias created since interest payments on outstanding liabilities are deductible from a banks tax base while dividend payments do not affect the tax base benefit the banks that enjoy the implicit guarantee significantly. Due to this tax bias debt financing has proven to be more desirable to systemically important institutions leading them to be excessively leveraged. In their paper Groenwegen, Mosch and Wierts (2016) found that individual banks have substantial scope for influencing their tax return by shifting between debt versus equity financing. Thus, excessive leverage has the benefit of lowering the tax base

(11)

of banks. Therefore, the savings in tax payments can be viewed as an implicit tax subsidy that large banks enjoy because of perceived government support.

High leverage in banks funding produces negative externalities in the economy and is often regarded as one of the leading causes of fragility in the financial system. Vast debt financing by banks can increase their vulnerability to bank runs, especially if their debts have short maturities. A single banks vulnerability to bank runs can have substantial spillover effects on other banks in the financial system through financial contagion where liquidity or insolvency risk is transmitted from one financial institution to another. Due to these negative externalities policy measures have been adopted in recent years to limit debt financing. For example, the Basel Committee on Banking Supervision has introduced measures that directly affect funding structures of financial institutions, such as more conservative capital and leverage requirements. Furthermore, the Dutch government decided recently to abolish the tax deductibility of Contingent Convertibles (CoCos) so that the coupon on the CoCos bonds will no longer be tax-deductible from January 1st 2019. CoCos are bonds that absorb losses when the capital of a financial institution falls below a critical level. The goal of this decision was to encourage financial stability by encouraging institutions to hold more equity. (Government of the Netherlands, 2018)

Thin capitalisation relates to firms that are financed through a relatively high level of debt compared to equity. Therefore, firms that can be characterised as being thinly capitalised are firms that are highly leveraged. Because the way in which firms are capitalised has a significant effect on the profit the firm and thus the amount of tax, the firm pays countries have often proposed rules that set a limit on the amount of interest that can be deducted. The thin capitalisation rule that was in place in the Netherlands was abolished on 1st of January 2013, and currently, there is no thin capitalisation rule in the corporate income tax code in the Netherlands (Deloitte, 2018). Since banks tend to be thinly capitalised because of the nature of their business, it is interesting to examine if the tax shield of companies is especially large for banks, but in section 5.1.1 the additional tax shield that ING Bank enjoyed over traditional Dutch firms in 2017 was estimated.

(12)

4.1.1 The model for estimating the value of the effective tax subsidy

This thesis follows the model put forward by Groenewegen, Mosch and Wierts (2016) where they construct a model for two scenarios. First, a scenario called a neutral benchmark where interest rates are non-deductible from before-tax earnings and then a second scenario where interest payments are deductible from earnings before the tax is calculated.

In the model the following are the definitions of the variables: t is the corporate tax rate on earnings, x is the bank's gross earnings on its assets before interest and taxes, 𝑟_"#$_is the interest rate on short-term debt, 𝐷"#_{is the short-term debt level of a bank, 𝑟}

&#$ is the interest rate on long-term debt and 𝐷&#_{is the long-term debt level of a bank.}

4.1.1.1 A neutral benchmark

Under the neutral tax regime, banks earnings, 𝑥, would first be taxed in its entirety and the after-tax earnings then redistributed between the institution's stakeholders, holders of equity or debt. The cash flow to bank’s financiers in the case of non-deductibility of interest payments is the following:

To shareholders (1 − 𝑡)𝑥 − 𝑟_"#$_𝐷"#_{− 𝑟} &#$𝐷&# To short-term debt holders 𝑟_"#$_𝐷"#

To long-term debt holders 𝑟_&#$_𝐷&#

By adding up different elements of the cash flow we get the total after tax earnings to be: (1 − 𝑡)𝑥 (1) As equation one demonstrates, the bank cannot increase the total cash flow to its shareholders and creditors by modifying the debt and equity combination in the funding structure of the financial institution.

(13)

4.1.1.2 Deductibility of interest payments

When interest payments are deductible from earnings before the tax is calculated the cash flow to the bank's stakeholders will be:

To shareholders (1 − 𝑡)(𝑥 − 𝑟_"#$_𝐷"#_{− 𝑟}

&#$𝐷&#)

To short-term debt holders 𝑟_"#$_𝐷"#

To long-term debt holders 𝑟_&#$_𝐷&#

By adding up the elements of the cash flow we now get the total after-tax earnings to be: (1 − 𝑡)𝑥 + 𝑡𝑟_"#$_𝐷"#_{+ 𝑡𝑟}

&#$𝐷&# (2) Note, that the difference in the total after-tax earnings between the two cases, the difference between equation one and two, is the corporate income tax multiplied with the interest payments on short-term and long-term debt, 𝑡𝑟"#$𝐷"#+ 𝑡𝑟&#$𝐷&#. These two terms are the tax advantage of debt, or the effective tax subsidy. Since the debt level, D, is to a certain extent a choice variable for financial institutions a bank can increase its after-tax earnings by expanding the debt level. Therefore, the introduction of tax deductibility eliminates the irrelevance of the capital structure of a bank that was noticeable in the benchmark case. Furthermore, since increased debt level can increase after-tax earnings then increasing leverage is beneficial to a bank.

Section 5.1 shows the quantification of the effective tax subsidy enjoyed by ING Bank in 2017. In the section, the tax advantage of debt is estimated by multiplying the national corporate income tax rate to the interest payments made by ING Bank in 2017. Also, in section 5.1, an analysis of the additional tax shield that ING Bank enjoyed over traditional Dutch firms in 2017 is presented.

4.2 The Too Big to Fail status of vast financial institutions

Financial institutions that are considered Too Big to Fail enjoy a funding subsidy resulting from implicit government support. This implicit support arises from the expectation of investors in large financial institutions of government support in times of severe distress for the institution. However, in the absence of such an implicit guarantee for large financial institutions investors would be forced to evaluate an institutions financial health and take

(14)

that into consideration when investing in the firm. Thus, investors would most likely demand higher yields on uninsured debt in cases of increased risk taking of the institutions. For market participants to excerpt such discipline on institutions debtholders must truly believe that they will bear the cost in case of insolvency or financial distress. When institutions are assumed to be Too Big to Fail investors believe they will not have to carry these costs themselves and thus the presence of such a guarantee dulls market discipline.

When estimating the Too Big to Fail implicit subsidy for systemically important financial institutions it is important to examine what defines a systemically important institution. In the analysis presented in this thesis the size of the firm’s assets is assumed to define a systemically important institution. However, it is important to note that size is not the only attribute of a systemically important institution. Nevertheless, according to recent literature the size of a firm’s assets is a significant attribute. For example, Adrian, T. and Brunnermeier, M. (2011) conclude that larger size financial institutions tend to be associated with more substantial contributions to systemic risk.

In section 5.2 the results to the estimation of the value of the implicit Too Big to Fail subsidy for ING Bank in 2017, is presented.

4.2.1 Regression analysis of the Too Big to Fail implicit subsidy of ING Bank in 2017

To estimate the reduction in funding costs for a Too Big to Fail financial institution in 2017 as a result of implied government support the following regression from Acharya, Anginer and Warburton (2013) was run for the year 2017.

𝑆𝑝𝑟𝑒𝑎𝑑_6,7,# = 𝛽_:+ 𝛽_;𝑖𝑠𝑠𝑢𝑒𝑠𝑖𝑧𝑒_6,7,#+ 𝛽_@𝑡𝑡𝑚_6,7,#+ 𝛽_B𝑙𝑒𝑣𝑒𝑟𝑎𝑔𝑒_6,#+ 𝛽_F𝑟𝑜𝑎_6,#+ 𝛽_H𝑚𝑏_6,# + 𝛽_J𝑚𝑖𝑠𝑚𝑎𝑡𝑐ℎ_6,#+ 𝛽_M𝑚𝑒𝑟𝑡𝑜𝑛𝑑𝑑_6,#+ 𝛽_O𝑑𝑒𝑓_#+ 𝛽_Q𝑡𝑒𝑟𝑚_#+ 𝛽_;:𝑚𝑘𝑡_#

+ 𝛽_;;𝑠𝑖𝑧𝑒90_6,#+ 𝜀_6,7,# (3) In equation three, various bond characteristic, macro and firm-specific factors are controlled for. The bond characteristics control variables are the following: Spread is the difference between the yield of a firm’s bond and the yield of a maturity matched government bond. Issuesize is the log value of the size of the bonds issue and ttm is the effective time to maturity of a bond. The firm-specific control variables are the following: Leverage is the firm’s total liabilities divided by total assets. ROA is the return on a firm’s assets, measured as the annual net income divided by total year-end assets. Mb is the market to book value of total

(15)

equity and mismatch is the short-term liabilities minus cash divided by total liabilities. MertonDD is the distance to default of each institution in the sample in 2017. The macroeconomic control variables are the following: Mkt is the market risk premium, and term is the term structure premium measured by the yield spread between long-term, ten-year maturity, and short-term, three-month maturity, government bonds. Def is the default risk premium as measured by the yield spread between maturity matched BAA rated and AA rated corporate bonds. The variable of interest in the Too Big to Fail analysis is size90 which is a dummy variable equal to one if the bank's total assets are in the top 90th percentile rank of firms in the sample and otherwise equal to zero. In the sample for this analysis ING Bank is the only firm in the 90th percentile and thus the coefficient indicates the reduction of the funding cost for ING Bank in percentage points in 2017, the implicit TBFT subsidy. The results from the regression analysis appear in table one in section 5.2.

The sample of financial institutions utilised in the regression analysis consists of ten European banks of variable sizes measured by total assets in 2017. These institutions and the size of their total assets in euros in 2017 are shown in table three in the Appendix. In equation three the subscript i, b and t indicate the financial firm, the bond and the year respectively. The analysis in this thesis was performed for the operating year 2017 and thus the subscript t indicates 2017. As aforementioned the variable of interest in equation three to this analysis is size90 but to quantify the amount of the implicit subsidy for ING Bank the annual reduction in funding cost, the coefficient on size90, was multiplied by the total uninsured liabilities of ING Bank in 2017. The quantification of the implicit Too Big to Fail subsidy for ING Bank in 2017 is presented in section 5.2.

4.2.2 Merton’s distance to default

To measure the Too Big to Fail implicit subsidy for ING Bank in 2017 the Merton’s distance to default had to be calculated for each of the firms in the sample, presented in table three in the Appendix, for the operating year 2017. In the analysis, the firm's distance to default is the primary risk measure but a firm’s distance to default can easily be linked to the firm’s probability of default. Distance to default is based upon Merton (1974) structural credit risk model, and measures the number of standard deviations the market value of firms assets is away from the default point, where the value of liabilities exceeds the value of the assets. A high Merton distance to default signals a lower probability of insolvency and default and vice

(16)

versa. Details on the Merton’s distance to default calculations are put forward in the Appendix. Furthermore, figure three in the Appendix shows the development of Merton’s distance to default and probability of default for ING Bank during 2013 to 2017.

4.3 Deposit insurance

The essential functions of banks are to lend money to individual or firms as well as to store safely deposits of customers. Banks thus charge interest on their loans to individuals and firms and pay interest to depositors for the use of their funds. Deposit insurance has benefits both to depositors of the bank as well as the bank itself. In the absence of insurance on deposits depositors would have to exercise market discipline and determine in which bank to store their deposits. Thus, depositors would have to analyse the balance sheet of each bank as well as its management and overall market risk before deciding on which bank to use. Therefore, depositor’s insurance benefits small depositors particularly since there are large information and surveillance costs to be saved if they believe that their deposit will be insured. Also, deposit insurance schemes benefits banks since they are both subject to less scrutiny and the scheme provides liquidity to banks in case of banks runs, thus, making the financial system more stable by preventing bank runs.

Due to the aforementioned advantages of deposit insurance schemes numerous countries have put into place explicit deposit insurance system, although certain elements like coverage and funding of these systems may differ across countries. Furthermore, explicit deposit insurance schemes may be desirable for policymakers since they set ground rules regarding coverage, participants and funding.

However, various costs are also associated with deposit insurance schemes. An example of such costs is reduced incentive for depositors to excerpt market discipline on banks because depositors expect their savings to be secure they store their savings at a bank without examining the bank's financial position. Deposit insurance intensifies the ability of banks and incentives of shareholders to increase risk since their potential cost is limited. Deposit insurance limit potential losses for banks since their debts are in large part guaranteed by a third party in case of insolvency. Therefore, the presence of deposit insurance schemes can threaten financial stability due to possible risk-shifting of financial institutions where the institution can increase risk without adequately compensating

(17)

taxpayers by increasing their capital ratios or by paying higher premiums for government guarantees.

Since deposit insurance schemes are in place in a vast number of countries across the globe and have a direct effect on depositors, banks and financial stability it is vital to realise the real cost of such a system on the guarantor which in the case of the banking sector due to its size is often the government or a government agency. Therefore, as the insurer is the government the actual cost of such a system is de facto a cost to taxpayers.

Various methods have been developed to measure the fair value of deposit insurance guarantees by applying option pricing theory. Robert Merton pioneered a model to value deposit insurances using option pricing theory with the Black- Scholes framework in his 1977 paper "An analytic derivation of the cost of deposit insurance and loan guarantees". The primary idea behind this valuation of deposit insurances is that the issuing of a guarantee imposes a cost on the guarantor and this cost can be estimated with the Black- Scholes framework. Under normal circumstances, the guaranteed would be expected to pay an amount equalling at least the cost for the guarantee to the guarantor. Since the Merton method can be applied to estimate the cost for the insurer we can use the Merton method to determine the fair insurance rate or the value of the deposit guarantee from the Dutch government to ING Bank.

The development of deposit insurance pricing models like the Merton method is founded on the isomorphic relationship between deposit insurance and common stock put options. In the Black- Scholes framework estimating the fair insurance rate can be viewed as a European option where the underlying asset is the value of the bank’s assets and the promised payment on the debt on the exercise date is equal to the strike price. The term European put option is used to describe options that can only be exercised on the expiration date while an American put option can be exerted on or before the expiration date. The owner of the put option has the right to exercise his option to sell at a specified time. If the put owner chooses not to exercise his option on the specified date the contract expires, and it is worthless.

(18)

In this thesis, I follow the Merton method as described in the paper “Estimating Fair Premium Rated for Deposit Insurance Using Option Pricing Theory: An Empirical Study of Japanese Banks” by Nobuyuki Oda (Oda, 1999).

4.3.1 The Merton Method of estimating the fair value of deposit insurance guarantees Merton (1977) utilises the Black-Scholes option pricing framework for the valuation of fair insurance rates as follows:

The Merton method assumes that a financial institution issues debt of B euros to be paid on a specified maturity date. On the maturity date if the value of the firm's assets, V, is higher than the promised payment on the debt, B, then it is in the interest of the shareholders for the management of the institution to make the promised payment. If the management is unable to meet the payment the institution defaults, and the bondholders will acquire the institution's assets. If on the maturity date the value of the institution's assets is less than the promised payment than the management of the institution will be unable to make the promised payment. Since even if the administration would try to sell the intuitions assets, it would still be unable to meet the payment. Thus, the institution will default its assets to institutions bondholders, then the value of the debt issued will be equal to V, and the value of equity will be zero.

Introducing a third-party guarantee of the promised payment on the maturity date could eliminate all uncertainty about the payment of the debt being met. Such a guarantee would assure the bondholder that in the case of the financial institution not being able to make the promised payment a guarantor would step in and make these payments. However, in the case of such an event, the institution would default all its assets over to the guarantor. Thus, in effect, the guarantor ensures that the value of the firm's assets at the maturity date will be at least equal to B euros.

This rationale can easily be applied to banks where generally the vast majority of their total liabilities are deposits kept at the bank that can either be withdrawn on demand or a specified date. A bank’s assets need to be sufficient to repay depositors that withdraw their deposits from the bank but in case of insolvency of a bank deposit insurance covers the customer's deposits. Furthermore, the third-party guarantor described above is generally the

(19)

government or a government agency in the case of banks due to the vast size of banks in the financial system.

The following are the formulas as put forward by Nobuyuki Oda (1999) to estimate the value of a deposit guarantee to a financial institution. From the viewpoint of the insurer providing the deposit insurance to a bank the required cash flow G at time of maturity of the banks liabilities may be defined as follows:

𝐺 = 𝑀𝑎𝑥[0, 𝐵 − 𝑉] (4)

Equation four may be viewed as a European put option in which at the maturity date the value of the guarantee is determined by which of the following elements are greater, the difference between the promised payment on the institution's debt and the value of the firm’s assets or zero. If at the maturity date the face value of the promised payment of liabilities excluding the value of the firm’s assets is positive than the option benefits to the shareholders. However, the insurance is worthless if the institution has sufficient amount of assets so that the protection will not be employed at the maturity time.

With this as background the fundamentals of the Black-Scholes model can be utilized and the formula to estimate the value of a put option can be applied to estimate the value of deposit guarantee: 𝐺 = 𝐵𝑒^_`_{𝑁b𝑥 + 𝜎} d√𝑇g − 𝑉𝑁(𝑥) (5) Where: 𝑥 = ln k𝐵𝑉l − m𝑟 + 𝜎_d@ 2 n 𝑇 𝜎d√𝑇 (6)

In the equations above N(.) is the cumulative normal density function, V is the market value of the bank’s assets, and B is the face value of the bank’s liabilities. In the analysis of the annual fair insurance rate both the market value of the bank’s assets and the face value of the banks liabilities are measured at year-end. T is the time to maturity, r is the risk-free interest rate, e is a mathematical constant approximately equal to 2,71828 and 𝜎d@ is the variance of the value of the firm’s assets. The market value of ING Bank at year end 2017 was

(20)

measured as the sum of the banks market capitalisation and the face value of the institution's debt.

The primary purpose of this analysis of the fair insurance rate is to evaluate the fair insurance rate for ING Bank in the period 2013 to 2017. As in the case for most traditional banks, ING Banks liabilities are too large extent customers and wholesale deposits that can be withdrawn on demand or a specified date. Thus, following Nobuyuki Oda (1999) description of the Merton method, we assume the face value of the bank's liabilities, B to be the exercise price in the put option framework. Since a vast part of the bank's liabilities consist off deposits available on demand defining time to maturity can be complicated. Following Merton (1977) the time until maturity is assumed to be the length between audits on the bank's assets. Since ING Bank has an external audit performed annually, T is assumed to be equal to one. This assumption provides us with the estimated fair insurance rate for ING Bank, the fair value of the deposit insurance guarantee, on an annual basis.

In section 5.3 the results from this analysis of the fair insurance rate for ING Bank during the period 2013 to 2017 is presented, and the value of the deposit insurance system is displayed on an annual basis. Therefore, the development of the fair insurance rate over the five-year period can be examined and interpreted.

5 Results

This section presents the findings of the analysis of the scope of government guarantees enjoyed by ING Bank from the Dutch government in 2017. Firstly, the effective tax subsidy that ING Bank experiences due to the tax treatment of excessive leverage is examined. Secondly, the value of the Too Big to Fail subsidy that ING Bank enjoys due to its vast size in the Dutch banking system is reviewed. Lastly, the value of the deposit insurance that ING Bank enjoys from the Dutch government is estimated.

5.1 The effective tax subsidy to ING Bank during 2013 to 2017

The difference between the neutral benchmark and the case of tax deductibility of interest payments scenarios put forward in section 4.1 is that the earnings after tax of a financial institution are higher in the tax deductibility scenario by 𝑡𝑟"#$𝐷"#+ 𝑡𝑟&#$𝐷&# than in the neutral benchmark. These two terms are the tax advantage of debt, or the effective tax subsidy. Thus, to estimate the aggregate the size of the effective tax subsidy ING Bank enjoys I follow

(21)

Groenwegen, Mosch and Wierts (2016) and multiply the corporate income tax, t, with interest paid both on short-term and long-term debt. Notice that distinguishing between interest paid on short-term and long-term debt is unnecessary because they are both included in the total interest paid on debt presented in the annual reports of ING Bank.

𝑆𝑖𝑧𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 𝑡𝑎𝑥 𝑠𝑢𝑏𝑠𝑖𝑑𝑦 = 𝑡 ∗ (𝑟_"#$_𝐷"#_{+ 𝑟}

&#$𝐷&#) (7) The Annual Report of ING Bank from 2017 shows that the bank’s interest paid in 2017 amounted to 30,978 in EUR million. The corporate income tax in the Netherlands in 20% on taxable profits up to 200,000 in EUR and 25% on taxable profit exceeding that amount (Government of the Netherlands, n.d). Since the profit of ING Bank after taxation in 2017 amounted to 5,019 in EUR million, I will use the 25% corporate tax rate in the calculations. Utilizing this information yields an effective tax subsidy of 7,745 in EUR million or about 1,05% of the GDP of the Netherlands in 20171_.

Furthermore, chart 1 shows the development of the tax subsidy of ING Bank over the last five years in EUR millions. As the chart reveals, the support has declined steeply during

1_{The value for the Netherlands gross domestic product (GDP) was retrived from the Eurostat database.}

According to the database the GDP of the Netherlands was 737,048 million euros in 2017.

(22)

the period or by 26,54%. Interest paid by ING Bank was the highest in 2013 at 42,171 EUR million and had since then continuously fallen to 30,978 EUR million in 2017.

5.1.1 The tax shield that ING Bank enjoyed over traditional Dutch firms in 2017

In the context of tax shields that arises due to the deduction ability of interest payments to shield portions of the firms' earnings from taxation, it is essential to recognise that it is in the nature of banks to be leveraged since the primary purpose of banks is to store deposits of the public securely. Thus, it is interesting to examine if the tax shield is especially large for banks and how the tax shield that ING Bank experienced in 2017 compares to the tax shield of other traditional Dutch firms in the same year.

To examine the additional tax shield that ING Bank experienced over traditional companies in the Netherlands in 2017 the average leverage of ten Dutch firms was calculated and compared to the leverage of ING Bank and the effective tax subsidy in 2017, presented in section 5.1. The ten firms that were used to estimate average leverage were the following: Heineken N.V., Philips, Royal FrieslandCampina N.V., Koninklijke Ahold Delhaize N.V., Akzo Nobel N.V., TomTom N.V., Aalberts Industries N.V., ASML Holding N.V., Beter Bed Holding N.V. and Randstad N.V. The average leverage for the ten firms in 2017 was 54,64% while the leverage of ING Bank in the same year was 94,76%.

A company with high levels of debt, a highly leveraged firm, ordinarily bear greater interest expenses than less leveraged firms. Because interest paid by a firm on its debt must be proportional to the firms leverage the additional tax shield that ING Bank experienced in 2017 over traditional Dutch firms can be estimated. Utilising that ING Bank had 40,11% greater leverage than the estimated average of the sample of Dutch companies the additional tax shield that ING Bank enjoyed over traditional Dutch firms in 2017 amounted to EUR 3,278 million.

5.2 The implicit Too Big to Fail subsidy to ING Bank in 2017

Table one presents the results from three different regression analyses of the reduction in funding costs enjoyed by major financial institutions. The sample of financial institutions utilised in the regression analysis consists of ten financial institutions in variable sizes measured by total assets and 134 observations. A list of the financial institutions in the sample along with their size of total assets in euros are presented in table three in Appendix.

(23)

The primary difference between the three regression models in table one is that they capture the reduction in funding cost enjoyed by financial institutions depending upon in which percentile the institution ranks by their total assets in 2017. Thus, the main difference between the models is the definition of the size variable, but in these regressions, the size variable is a dummy variable which equals one depending upon to which percentile the bank in the sample belongs. Therefore, in model A the size90 dummy variable captures the reduction in funding cost enjoyed by the bank in the top 90th percentile ranked by total assets in 2017. In our sample, size90 captures the funding advantages experienced by the institution of interest in this thesis, ING Bank, in 2017. In model B the size60 variable captures the funding cost advantages of Intesa Sanpaolo S.p.A, Credit Suisse Group AG, Banco Bilbao Vizcaya Argentaria S.A and ING Bank in 2017, the banks in the top 60th percentile of the sample ranked by total assets. In model C the size30 variable captures the funding cost reduction for the banks in the top 30th percentile of the sample, Intesa Sanpaolo S.p.A, Credit Suisse Group AG, Banco Bilbao Vizcaya Argentaria S.A, ING Bank, Commerzbank AG, Erste Group AG and Swedbank AB.

In the analysis bond information for all of the financial institutions in the sample listed in the Appendix were utilised to estimate the Too Big to Fail subsidy. The bonds used are all corporate bonds without equity or derivative features (i.e., callable, puttable and convertible bonds) and all bonds with floating interest rates were also excluded. Also, following Acharya, Anginer and Warburton (2013) all bonds with less than one year to maturity were eliminated before the estimation. Furthermore, to ensure that outliers do not profoundly influence the statistical results I eliminate observations higher than the 99th percentile value and lower than the 1st percentile value. To estimate the reduction in funding cost for ING Bank in 2017, I use the statistical software package STATA and the regressions performed are ordinary least squares regressions.

Table one shows the findings of three regressions of bond characteristics, firm-specific information, size dummy variables and macroeconomic control variables on credit spread on the financial institution's bonds. Each variable in the table has two values. The upper value shows the estimated coefficient value for the variable while the standard errors are reported in parenthesis below their coefficient estimates and are adjusted for heteroskedasticity using robust standard errors.

(24)

The regressions results allow for some conclusions to be drawn about how the size of major financial institutions influences the institutions borrowing costs. As table one shows the variables that capture the reduction in funding costs of large financial institutions, the size dummy variables indicate a negative and significant relationship in all of the regressions. Concluding that large financial institutions have lower credit spreads on their bonds and therefore face lower funding cost than smaller banks. Furthermore, the estimated coefficient values for the size variables indicate that the largest institution, ING Bank indeed enjoy the most substantial funding cost advantages, thus affirming the Too Big to Fail status of the Dutch bank.

In the regressions three size variables interpret the effect of the size of the financial institutions on the spread difference on institutions bonds and maturity-matched government bonds after controlling for firm-specific information, bond characterises and macroeconomic factors. The coefficients that capture the effects of size of financial institutions range from a reduction in funding cost by 6,63 percentage points for the 90th percentile to 0,952 percentage points for the 60th percentile and 1,18 percentage points for the 30th percentile.

In regression model A, the issue size of bonds, mismatch, Merton’s distance to default and the size dummy variable for ING Bank were statistically significant. In regression model B, the issuesize, leverage, return on assets, market to book ratio of equity, the market risk premium and the dummy size variable for the 60th percentile of financial institutions were all statistically significant. In regression model C the issue size of bonds, leverage, return on assets, market to book ratio of equity, the term premium and the dummy variable that captures the 30th percentile of the financial institutions on the sample were all statistically significant.

The coefficients on the size dummy variables represent the funding cost advantages enjoyed by the large financial institutions. For this thesis, the primary coefficient of interest is the coefficient of size90 because it captures the implicit government insurance accruing to ING Bank in 2017. Model A shows the reduction in funding cost for ING Bank in 2017 was 6,63 percentage points but to quantify the euro value of the implicit Too Big to Fail subsidy for ING Bank in 2017 the bank’s total uninsured liabilities was multiplied by the reduction in funding cost. To determine the total uninsured liabilities of ING Bank in 2017 deposits backed by

(25)

government insurance had to be excluded from the bank's total liabilities in 2017. Following the "The state of Dutch banks in 2015" issued by KPMG in 2015 I assume the proportion of covered deposits versus total deposits to be 51,8%, meaning that the Deposit Guarantee Scheme will cover 51,8% of all deposits. The percentage was initially derived from an earlier study performed by the European Commission in 2014. As the deposit insurance coverage has remained unchanged since the protection provided by the deposit guarantee scheme was raised to EUR 100,000 in all EU member states on 31 December 2010 (Government of the Netherlands, n.d). I assume that the ratio of uncovered to covered deposits have remained relatively unchanged since the issuance of the KPMG report. Information on ING Banks liabilities was retrieved from the bank's annual report for 2017. The analysis yields that the euro value of the implicit Too Big to Fail subsidy of ING Bank in 2017 amounted to 30,461 million EUR.

From viewing the regression results in table one, it is clear that the size of financial institutions in the sample has a significant effect on the funding costs of the institutions. Thus, the results in table one support the notion that systemically important institutions do indeed benefit from the perceived government support attributed to them because of their size in the local or global financial system.

The p-value of an analysis determines the significance level of a coefficient. In the following table ***, ** and * indicate significance at the 1%, 5% and 10% level, respectively.

(26)

Model A Model B Model C

Spread Spread Spread

Issue size -0,00493 * -0,00493 * _-0,00493 * [-2,17] [-2,17] [-2,17] Time to maturity -0,000767 -0,000767 -0,000767 [-1,76] [-1,76] [-1,76] Leverage 0,742 3,834 *** _2,932 * [0,45] [3,65] [2,55] Return on assets -0,951 13,17 *** _10,63 * [-0,14] [3,49] [2,60]

Market to book ratio of equity 0,00509 -0,158 *** _-0,121 ** [0,07] [-4,01] [-2,77] Mismatch 0,0797 * 0,000185 -0,0202 [2,39] [0,01] [-1,29] Merton’s distance to default 0,0154 * 0,00506 0,00384 [2,25] [1,15] [0,91]

Default risk premium 0,375 -0,0956 -0,151

[1,90] [-0,87] [-1,36]

Term premium 2,246 -1,442 -1,768 *

[1,28] [-1,76] [-2,25]

Market risk premium -3,903 3,552

** _2,334 [-1.26] [2,86] [1,67] Size90 -0,0663 ** [-2,88] Size60 -0,00952 ** [-2,88] Size30 -0,0118 ** [-2,88] Constant -0,462 -3,622 *** -2,703 * [-0,28] [-3,57] [-2,41] R2 _0,2895 _0,2895 _0,2895 Number of observations 134 134 134

(27)

5.2.1 Further evidence of the implicit Too Big to Fail subsidy in the sample 5.2.1.1 The goodness of fit of variables on the dependent variable Spread

A variety of variables affected the credit spread difference between the yield on a financial institution's bonds and the yield on a maturity matched government bond. A useful way of judging how well a variable explains these credit spread differences is by viewing the goodness of fit, measured as 𝑅@_{. Figure four in the Appendix presents the goodness of fit of} single variables in separate OLS regressions explaining the spread difference on bonds of different financial institutions. In these regressions, the spread was the dependent variable, and the independent variables were the ones presented in figure four.

The most influential factor in determining the spread difference on bonds of the financial institutions in the sample was the size variable indicating if the bank was in the 30th percentile of banks ranked by assets. The second most prominent factor was the market to book ratio of the equity of the financial institutions. The third most influential factor in determining the spread difference was the market risk premium. The following affected credit spreads respectively: the bonds time to maturity, the default risk premium, the size variable signalling if the bank was in the 60th percentile of banks ranked by assets, the banks total assets and the bonds issue size.

5.2.1.2 Correlation matrix

The correlation matrix of different Too Big to Fail variables is showed in table four in the Appendix. Table four shows a negative correlation between the size of a financial institution measured by banks total assets and the credit spread on bonds issued by major financial institutions. Thus, suggesting that larger financial institution enjoy a funding cost advantage reflected in lower credit spreads of their bonds further confirming the Too Big to Fail hypothesis.

5.3 The fair deposit insurance rate for ING Bank during 2013 to 2017

In section 4.3 the Merton method as described in Nobuyuki Oda (1999) is explained in detail, but the Merton method is utilized in this thesis to estimate the fair insurance rate for ING Bank for the period 2013 to 2017. This section presents the mathematical results from that analysis.

(28)

The Black-Scholes formulas presented in section 4.3 were used to estimate the value of the deposit insurance enjoyed by ING Bank from 2013 to 2017. The required accounting data for ING Bank was retrieved from the bank's annual and quarterly reports from 2013 to 2017. The variance of the value of the bank’s assets was calculated with an iterative procedure of the value of ING Banks assets on an annual basis for the years 2013 to 2017. This iterative procedure is explained in detail in the Merton distance to default section in the Appendix. The risk-free interest rate used in the analysis is the deposit rate of the European Central Bank, the interest rate at which financial institutions may at all times place overnight deposits with the national central bank. Therefore, the deposit rate is the rate at which banks can earn a risk-free interest.

Applying the required data to the formulas in section 4.3 yields the result that the fair insurance rate and thus the value of the deposit insurance for ING Banks was EUR 154 million for the operating year 2017, roughly 3% of the banks profit during that year. Figure two shows the development of the value of the deposit insurance enjoyed by ING Bank from 2013 to 2017. The estimated fair insurance rate in 2013 amounted to EUR 2,080 million, in 2014 the amount was calculated to be EUR 1,490 million, in 2015 the estimated amount totalled EUR 623 million and in 2016 the estimated fair insurance rate amounted to EUR 1,354 million.

During the five-year period that this analysis covers there was variability in the estimated fair deposit insurance rate for ING Bank between the years. The variables that affect the estimation results are the face value of ING Banks debt at year-end as well as the estimated market value of the bank’s equity. These two variables increased steadily over the period. The volatility of the market value of ING Banks assets over the period remained relatively stable although the changes that did occur in the volatility seem to have affected the estimated fair value of ING Banks deposit insurance greatly. The deposit rate of the European Central Bank also remained reasonably stable ranging from 0 percent to negative 0,4 percentage over the five-year period.

The volatility of the estimate fair deposit insurance rate raises the question of that to achieve a better sense of the actual value of the deposit insurance scheme to a bank if it would prove better to average a group of operating years instead of focusing on a particular operating year.

(29)

When examining the fair value of the deposit insurance scheme for ING Bank, it is also important to acknowledge that the bank has contributed annually to a pre-funded deposit guarantee scheme for Dutch financial institutions since 2015. The EU Deposit Guarantee Scheme Directive (DGSD) obliges EU member states to have an ex-ante funded deposit insurance scheme (DGS) although there is currently no agreement on a euro-area DGS. In November 2015 the Netherlands started a new pre-funded Deposit Guarantee Scheme making the Dutch DGS more risk-based than it was before, De Nederlandsche Bank (2015). In the current Dutch DGS financial institutions contribute in advance to the DGS fund where the amount demanded from a financial institution depends on the amount of guaranteed deposits at that bank and the bank’s risk profile. The primary purpose of the DGS is to limit the danger to financial stability in case of distress since the De Nederlandsche Bank (DNB) can activate the DGS fund in the event of bank failure.

ING Bank has contributed annually to the deposit insurance scheme (DGS) since 2015, and the bank displays information in their annual reports on the exact amount the institution has provided to the DGS on a yearly basis. ING Bank contributed to the fund for the first time in 2015 when the new ex-ante scheme came into effect in the Netherlands and then again in 2016 and 2017. The regulatory costs associated with payments to the DGS amounted to EUR 341 million in 2017, EUR 316 million in 2016 and EUR 233 million in 2015. Thus, making the total amount that ING Bank has paid into the DGS since its founding EUR 890 million.

During the analysis period, the estimated fair deposit insurance rate amounts to EUR 5,710 million or an average of EUR 1,142 million a year. Thus, the estimated value of the deposit insurance scheme for ING Bank in the period 2015 to 2017 was roughly 2,4 times larger as the bank's contributions to the ex-ante funded deposit insurance scheme during the period.

(30)

Figure 2: The development of the fair deposit insurance rate for ING Bank in EUR million per year during 2013 to 2017

6 Comparison of the value of government guarantees and the profit of ING

Bank in 2017

The estimation results in section five confirm that ING Bank benefited considerably from perceived government support due to its size in the Dutch financial system in 2017. Now that the government guarantees have been quantified for the bank in 2017 it would be interesting to compare the estimated size of these subsidies to ING Banks operating profit in the same year. This section presents a comparison between the value of the government guarantees that ING Bank enjoyed in 2017 and the operating profit of ING Bank in 2017.

Table two presents the estimated euro values of the three government guarantees: the effective tax subsidy, the implicit guarantee of ING Bank due to its Too Big to Fail status in the Dutch financial system and the fair value of the deposit insurance scheme that the bank enjoys. In 2017 the operating profit of ING Bank was EUR 5,019 million. In the same year, the value of the effective tax subsidy of ING Bank estimated in section 5.1 amounted to EUR 7,745 million. The value of the Too Big to Fail implicit subsidy quantified in section 5.2 was EUR 30,461 million in 2017. The estimated fair deposit insurance rate for ING Bank in 2017 was

(31)

EUR 154 million although in the same year the bank contributed EUR 341 million into the Deposit Guarantee Scheme (DGS) in the Netherlands. Since the bank started to contribute to the DGS only in 2017 exceeded the contribution the estimated fair insurance premium.

Table two displays the values of each of the three guarantees individually in 2017 as well as the total amount of the government guarantees in the year. In 2017 the total value of the implicit and explicit government guarantees to ING Bank amounted to EUR 38,360 million, but after taking into account, the bank's contribution to the DGS in 2017 the total value decreases to EUR 38,019 million. In the year 2017 ING Bank had an operating profit of EUR 5,019 million, but the estimated value of the government guarantees exceeded the operating profit by EUR 33,000 million in the year. Thus, these results yield a net result corrected for government guarantees for ING Bank of EUR 33,000 million in 2017 an amount that is approximately seven times as large as the bank's profit in the same year.

Amount in EUR million

The value of the effective tax subsidy of ING Bank in 2017 7,745

The fair deposit insurance premium for ING Bank in 2017 154

The value of the Too Big to Fail subsidy for ING Bank in 2017 30,461

The total value of implicit and explicit government guarantees

to ING Bank in 2017 38,360

ING Bank contribution to the deposit guarantee scheme in 2017 341

The total value of implicit and explicit government guarantees

to ING Bank in 2017 corrected for contributions to DGS 38,019

ING Bank profit in 2017 5,019

Net result of ING Bank in 2017 corrected for the implicit and

explicit government guarantees (33,000)

Table 2: Comparison between the estimated value of government guarantees enjoyed by ING Bank and the banks profit in 2017.

(32)

7 Conclusion

Examining to what extent ING Bank enjoys implicit and explicit support from the Dutch government is vital because implicit subsidies serve as a transfer of resources from one set of agents, the government and ultimately taxpayers to the financial sector. The estimation results presented in this thesis suggest that ING Bank creditors and shareholders benefit to some degree at the expense of taxpayers.

Quantifying the values of the implicit subsidies that the bank benefits from is complicated because unlike other types of explicit agreements implicit support does not have clear terms nor an observable price. The estimation results presented in section five of this thesis shows that the estimated values of these government guarantees are striking in scale and suggest a notable transfer of resources from the government to ING Bank in 2017. Therefore, the presence and volume of these government guarantees should be of interest to taxpayers because at times of severe financial distress the government will offer support to institutions whose failure may result in unacceptable economic cost, but this support is often at the expense of taxpayers.

The estimation results presented in section five and summarised in section six show that ING Bank does indeed benefit substantially from perceived government support from the Dutch government. The estimated support that ING enjoyed in 2017 totals to EUR 38,360 million which is approximately eight times the profit of the institution in the same year.

The effective tax subsidy that ING Bank enjoys was quantified by multiplying the corporate tax rate in the Netherlands to the interest payments of the bank on debt in each operating year. The theory behind this estimation was explained in section 4.1 where two cases of tax deductibility of interest payments were compared, the difference between the cases is assumed to equal the effective tax subsidy. Figure one in section 5.1 presents the estimated subsidy for the period 2013 to 2017. The estimation revealed an EUR 7,745 million subsidy in 2017 alone. During the period, 2013 to 2017 the effective tax subsidy diminished, reaching its highest level in 2013 at EUR 10,543 million and then continuously declining to the 2017 value of EUR 7,745 million. Thus, the estimated tax subsidy fell by approximately 26% over the period. Over the five-year period, the effective tax subsidy totals to EUR 43,739 million.

(33)

Furthermore, the additional tax shield that ING Bank enjoyed over traditional Dutch firms in 2017 was estimated to be EUR 3,278 million in section 5.1.

The quantification of the Too Big to Fail status of ING Bank was performed with regression analysis with a sample of ten financial institutions to estimate the funding cost advantage of the bank in 2017. The aggregate reduction in the cost of bank funding due to the implicit guarantee was calculated by multiplying the reduction in funding cost to the banks total uninsured liabilities. The funding cost advantage of ING Bank in 2017 was 6,63 percentage points and the aggregate Too Big to Fail implicit subsidy was EUR 30,461million.

The fair insurance premium for ING Bank during 2013 to 2017 was estimated using the Merton method, but Robert Merton pioneered a model to determine the fair insurance rate for deposit insurance using option pricing theory with the Black-Scholes framework. The results from this estimation show that the appropriate insurance rate was EUR 154 million in 2017. However, the average estimated value of the deposit insurance scheme for ING Bank in the years 2013 to 2017 was EUR 1,142 million annually. The fair deposit insurance rate was subject to volatility during the five year period, but the highest estimated insurance rate was EUR 2,080 million in 2013, and the lowest estimated value was in 2017.

From the analysis performed in this thesis the answer to the research question of this thesis “Does ING Bank benefit from perceived government support?” is a resounding yes. The bank benefits from the negative externality introduced by their potential failure to the Dutch financial system to the extent that the value of the government guarantees that the bank enjoys amounted to eight times the profit of the bank in 2017.