Regulatory Capital and Tax Losses

(1)

Regulatory Capital and Tax Losses

Do European banks manage capital ratios using carryforward tax losses

accounted for under IAS 12?

by

Hasan Gürkan

University of Groningen

Faculty of Economics and Business

MSc accountancy

July, 2013

Supervisors:

drs. J.L. (Bert-Jan) Bout RA

prof. dr. R.L. (Ralph) ter Hoeven RA

(2)

Abstract

This study examines if European banks manage regulatory capital ratios using DTAs recognized for carryforward tax losses as accounted for under IAS 12. Based on prior literature different determinants to recognize a DTA for carryforward tax losses are identified. Using a sample of the 50 largest banks in Europe during the period 2008-2011, no significant evidence is found to support capital management through DTAs recognized for carryforward tax losses. However, there are significant results for the DTL/DTA ratio, system banks, size and riskiness. Similar to prior research, I find a negative relation between these four determinants and the amount of unrecognized tax losses. I would also like to propose to count DTAs fully to regulatory capital to the extent of taxable differences (DTLs) against which DTAs could be offset. This study has also some important limitations which could have an impact on the presented results. For example, this study only examined four years. Most other studies examine a much longer period, commonly 10 or more years.

Keywords: IAS 12, IFRS, deferred tax assets, valuation allowance, taxable income, Basel III, CRD IV, tax accounting, banks, capital management.

Data availability: Data are available from public sources identified in this study and database Bankscope

JEL Classification: G01, G21, G38, M41

The author is responsible for the content of this thesis. Copyrights of this thesis are of the author.

Author Hasan Gürkan

Student number S2079739

Email h.gurkan88@gmail.com

University University of Groningen

Faculty Faculty of Economics and Business

Degree Master of Science Accountancy

Supervisors University of Groningen drs. J.L. Bout RA

prof. dr. R.L. ter Hoeven RA

Thesis internship KPMG, Audit Financial Services at Amstelveen

(3)

Preface

‘’Our new Constitution is now established, and has an appearance that promises permanency; but in this world nothing can be said to be certain, except death and taxes’’

Benjamin Franklin, in a letter to Jean-Baptiste Leroy, 1789 The last five months I wrote my thesis to finish my study MSc Accountancy and Controlling at the University of Groningen. The banking sector always interested me. After the 2008 financial crisis and 2010 Euro crisis in Europe my interest for the banking sector has grown immensely and I wanted to research something about banks in my thesis. Besides the banking sector financial reporting is also a very interesting subject, especially complex and technical issues like tax accounting. Therefore my MSc thesis is about regulatory capital and tax losses. Together with four other classmates I wrote my thesis at KPMG at the Financial Services department. This was a great experience. Besides conducting scientific research I have learned a lot and also had a lot fun with my classmates. The major challenge was to define a concrete and specific research question. Unfortunately I have the negative quality that I would like to research everything about my subject. Luckily I got help from some people in this technical and complex subject of my thesis and therefore I would like to thank them.

First, I would like to thank drs. J.L. (Bert-Jan) Bout RA for his role as first supervisor. During my thesis Bert-Jan helped me a lot to define a concrete research area and question. Bert-Jan is also very intrinsically motivated to help students to get the best out of their thesis and to share his (enormous) knowledge. Especially his critical, constructive and straightforward feedback helped me a lot to write a good thesis. Bert-Jan also supported me a lot to understand capital requirements and constantly pushed me to stay critical on my own work. Secondly, I would like to thank prof. dr. R.L. (Ralph) ter Hoeven RA for his role as second supervisor. Ralph is a well known authority on the field of financial reporting. His knowledge positively contributes to my thesis. Finally, I would like to thank drs. W. (Walter) Berger for his support during my internship at KPMG. I knew Walter from last year when he was also my attendant. I enjoyed his way of working.

To understand this technical and complex subject I got some help from two other people. First, dr. P. (Peter) Epe RA helped me to understand accounting for income taxes. Peter also supported me in understanding the main message of his dissertation, which I believe was a very good one on the field of tax accounting. Secondly, I would like to thank drs. E.D.M. (Eveline) Gerrits RA (director of KPMG Meijburg) for her support on the field of tax accounting and especially tax accounting under Basel III. Eveline provided access to all available sources within KPMG for better understanding about tax treatment under capital requirements.

I hope that the readers of my thesis acquire knowledge about tax accounting, capital requirements and especially the interaction – tax treatment in regulatory capital. I also hope the readers enjoy it as much as I enjoyed writing it.

Amstelveen, 31th of July 2013 Hasan Gürkan

(4)

Dutch Summary

Deze scriptie onderzoekt of Europese banken hun kapitaalratio (hierna: toetsingsvermogen) sturen door gebruik te maken van de mogelijkheid tot voorwaartse verliescompensatie. Door een populatie van de vijftig grootste banken voor de jaren 2008 tot en met 2011 te bestuderen is onderzocht of er sprake is van kapitaalsturing.

De financiële crisis die medio 2007-2008 zijn weerslag had op Europese banken heeft aangetoond dat banken onvoldoende waren gekapitaliseerd om hun verliezen op te vangen. Mede onder druk vanuit de politiek zijn kapitaaleisen van banken zowel kwantitatief als kwalitatief verscherpt. Het aanhouden van hogere kapitaalbuffers is een voorbeeld van een kwantitatieve verscherping. Het elimineren van actieve belastinglatenties uit hoofde van voorwaartse verliescompensatie (hierna: vvc) bij berekening van het toetsingsvermogen is een voorbeeld van een kwalitatieve verscherping van kapitaaleisen.

Bijna alle Europese banken stellen hun jaarrekening op volgens International Financial Reporting Standards (IFRS) die verplicht zijn voor alle beursgenoteerde ondernemingen binnen de Europese Unie (EU). Via standaard 12 (IAS 12) bepalen banken in de EU óf zij hun verliezen activeren én ook voor welk deel. Hiertoe dienen banken in de EU een beoordeling te maken van hun toekomstige winst en de omvang van de passieve belastinglatenties. Actieve belastinglatenties ontstaan door het activeren van vvc of door vervroegde belastingen. Passieve belastinglatenties ontstaan door het opnemen van een verplichting voor uitgestelde belastingen. Passieve belastinglatenties kunnen – indien deze verrekend kunnen worden – als alternatieve onderbouwing dienen voor de realisatie van actieve belastinglatenties.

Toezichthouders erkennen de onzekerheid met betrekking tot realisatie van de vvc. Toezichthouders zijn sceptisch over de inschatting van banken over hun toekomstige winstgevendheid en hebben er daarom voor gekozen om vvc te elimineren bij het berekenen van het toetsingsvermogen. Blijkbaar is er een vermoeden dat banken hun toetsingsvermogen kunstmatig sturen door gebruik te maken van vvc. Ook voor belastinglatenties uit hoofde van tijdelijke verschillen geldt een beperking in berekening van het toetsingsvermogen. De verslaggevingsstandaard, voor banken in de EU veelal IFRS, is namelijk ook het uitgangspunt voor het berekenen van het toetsingsvermogen. De nieuwe kapitaaleisen zijn vastgelegd in Bazel III en worden geïmplementeerd in de EU via de kapitaalrichtlijn en -verordening CRD IV.

Op basis van voorgaand onderzoek zijn determinanten vastgesteld die invloed hebben op de bepaling van de hoogte van de vvc. Dit zijn onder andere de toekomstige winstgevendheid, de aanwezigheid van passieve belastinglatenties en de omvang van de bank. De resultaten van deze scriptie tonen niet aan dat er kapitaalsturing plaatsvindt via vvc. Hiermee kan echter nog niet gezegd worden dat het standpunt van de toezichthouders van tafel geveegd kan worden. Uit onderzoeken naar de voortgang van implementatie van Bazel III blijkt dat belastinglatenties niet van wezenlijk belang zijn in vergelijking tot bijvoorbeeld aftrekposten voor goodwill. Hierdoor zal het effect logischerwijs kleiner zijn. De vvc is namelijk afhankelijk van toekomstige winstgevendheid van de bank en de omvang van de passieve belastinglatenties. In geval van verliezen is het ook nog maar de vraag in hoeverre passieve belastinglatenties in staat zijn om verliezen die zijn geactiveerd uit hoofde van vvc op te vangen. Daarnaast is de periode in deze scriptie vermoedelijk te kort om effecten van invoering van nieuwe kapitaaleisen te meten en is het nog onmogelijk om onderscheid te

(5)

maken in een periode voor en na invoering van de nieuwe kapitaaleisen. De CRD IV is namelijk pas eind juni 2013 gepubliceerd en zal gefaseerd in werking treden vanaf 2014. Dit onderzoek levert ook een aantal interessante uitkomsten op. Allereerst valt op dat 60 van de 200 onderzochte jaarrekening niet het toelichten welk deel van de fiscale verliezen zij niet hebben gewaardeerd. Deze zijn verwijderd uit dit onderzoek. Wellicht erger is dat zij verplicht zijn dit toe te lichten maar hier geen gehoor aan geven. Dit wijst op het niet voldoen aan gestelde regelgeving. Hiernaast valt uit de resultaten af te lezen dat banken met een hoger toetsingsvermogen de vvc niet volledig tot waardering brengen en deze als een ‘stille reserve’ gebruiken. In tegenstelling tot deze groep vertonen banken met lagere toetsingsvermogens resultaten die wijzen op kapitaalsturing via vvc. Zoals reeds vermeld is de periode in dit onderzoek vrij kort, slechts vier jaren. De meeste onderzoekers beginnen vanaf 10 jaren en hierdoor kan uitbreiding van het aantal waarnemingen wellicht leiden tot dezelfde resultaten als voorgaande studies. Daarnaast, is er statisch significant bewijs dat systeembanken een hogere vvc hebben in vergelijking tot niet systeembanken. Deze bevinding is volledig in lijn met voorgaand onderzoek. Tot slot hebben grotere banken ook hogere vvc op hun balansen in vergelijking tot kleinere banken. Dit lijkt logisch. Systeembanken zijn vaak de allergrootste banken en de grootste banken hebben meer mogelijkheden om fiscale verliezen te compenseren. Dit kan zijn via toekomstige winsten, betaalde belastingen in het verleden of door verrekening met uitgestelde belastingen. Deze resultaten zijn daarom van belang voor accountants en toezichthouders. Zij dienen te beseffen dat grotere en systeembanken deze mogelijkheden hebben en de impuls hiertoe – verhoogde kapitaaleisen voor systeembanken – lijkt ook aanwezig.

Een ander interessant object voor onderzoek is de beperking van de omvang van de actieve belastinglatentie uit hoofde van tijdelijke verschillen die meetellen in de berekening van het toetsingsvermogen. Voor banken die nog niet het maximum hebben bereikt van tijdelijke verschillen die worden meegenomen in de berekening van het toetsingsvermogen kunnen prikkels ontstaan voor kapitaalsturing. Hiertoe kan bijvoorbeeld een vvc worden omgezet in een actieve belastinglatentie uit hoofde van een tijdelijk verschil. Deze constructie zal in Nederland werken voor banken die een fiscale eenheid vormen en entiteiten hebben buiten de fiscale eenheid, zie voetnoot 10 op pagina 17 voor het complete voorbeeld. Evenals het meten van kapitaalsturing via vvc is dit vermoedelijk ook niet eenvoudig te operationaliseren in empirisch onderzoek; desondanks verdient dit onderwerp samen met uitgebreider onderzoek naar kapitaalsturing via belastinglatenties mijns inziens de aandacht voor vervolgonderzoek. Tot slot heeft dit onderzoek nog een belangrijke aanbeveling voor toezichthouders. Evenals andere normatieve – en empirische onderzoeken toont ook dit onderzoek aan dat uitgestelde belastingen een belangrijke factor zijn in het opnemen van de vvc. In aansluiting op de vorige alinea, lijkt het mij beter om actieve belastinglatenties volledig toe te staan – tot de omvang van de passieve belastinglatenties waarmee ze verrekend kunnen worden – bij berekening van het toetsingsvermogen. Met dit voorstel is de onzekerheid met betrekking tot de actieve belastinglatenties sterk gereduceerd, vermoedelijk nihil. In geval van excessieve verliezen – of een faillissement – is de bank in staat om alle actieve belastinglatenties te verrekenen met passieve belastinglatenties. Hiermee worden ook allerlei prikkels om de vvc om te zetten in tijdelijk verschillen weggenomen. Actieve belastinglatenties worden bij berekening van het toetsingsvermogen immers slechts meegenomen voor zover er verrekenbare passieve belastinglatenties aanwezig zijn.

(6)

Acronyms

BCBS Basel Committee on Banking Supervision

BIS Bank for International Settlements

CET1 Common Equity Tier 1

CRD IV Capital Requirement Directive IV

CRR Capital Requirements Regulation

CT1 Core Tier 1 Equity

DTA Deferred tax asset

DTL Deferred tax liability

EBA European Banking Authority

EU European Union

GAAP Generally Accepted Accounting Principles

HROA Historical Return on Assets

IAS International Accounting Standards

IFRS International Financial Reporting Standards

LLP Loans loss provision

ROA Return on Assets

ROE Return on Equity

RWA Risk-weighted assets

U.S. United States of America

(7)

1 Introduction

There is no doubt about the footprint that the 2008 financial crisis has left on the world economy. Lehman Brothers, an U.S. investment bank, goes bankrupt after the U.S. government refuses to bail out. More recently, on 1 February 2013, the Dutch government had to nationalize SNS Reaal after real estate losses. Also several other banks were not able to carry their losses and had to be saved by governments or went bankrupt. However a survey of KPMG (2011; 2012) of 15 of Europe’s largest banks shows that banks have huge deferred tax assets (DTAs) on their balances. According to the KPMG survey these banks need a minimum taxable profit of €334 billion in 2010 and €450 billion in 2011, which seems paradoxical in times of economic downturn.

Since 2005 all European listed firms needed to prepare their consolidated financial statements in accordance with one single set of accounting standards, namely International Accounting Standards (IAS) (Commission Regulation (EP-Council) no. 1606/2002). Some of the IASs are revised in International Financial Reporting Standards (IFRS). Reporting about taxes is primarily regulated in IAS 12, Income Taxes. Although other standards, e.g. IAS 1,

Presentation of Financial Statements and IAS 8, Accounting Policies, Changes in Accounting Estimates and Errors, are also applicable to income taxes. DTAs exist of two parts, namely

deductible temporary differences and carryforward due to tax losses. Especially a carryforward asset is based on estimations. Banks need to estimate their future profit to recognize a carryforward tax asset. It is clear that estimations are inherently subject to subjectivity and therefore could be used, for example to manage earnings (e.g. Wahlen, 1994, Dechow and Skinner, 2000; Liu and Ryan, 2006; Kilic et al., 2013).

The financial crisis created input for the Basel Committee on Banking Supervision (BCBS) to publish a revised capital standard, Basel III (BIS, 2011a). Basel III also pays attention to the loss absorption capacity of DTAs. The BCBS acknowledges the uncertainty with respect to reported DTAs in financial statements of banks. Paragraph 69 of Basel III determines that DTAs based on future profitability of the bank are deducted in calculation of Common Equity Tier 1 (CET1). Paragraph 87 contains threshold for deductions for DTAs that arise from temporary differences. With these deductions in calculating regulatory capital it seems clear that supervisory bodies have suspicion about the loss absorption capacity of DTAs.

1.1 Relevance

DTAs are part of the accruals used in financial statements and prior research showed that investors react positive on accruals quality (Francis et al., 2004; 2005; Ecker et al., 2006; Core et al., 2008). Recently, Epe (2010) argued that listed companies in the Netherlands should recognize all deferred taxes for a fair view in the statements of financial position, based on his assumption that recognizing all deferred taxes should present a fair view. Prior research also investigated several incentives to manage regulatory capital by banks (Chen and Daley, 1996; Leventis et al., 2011; Shen and Huang, 2013). Ahmed et al. (1999) found strong evidence that banks use loan loss provisions (LLPs) for capital management, but found no evidence for earnings management via loan loss provisions. On the contrary, Alali and Jaggi (2011) found evidence of stronger earnings management by U.S. commercial banks during the financial crisis compared to the pre- financial crisis period, but not find evidence to conclude that managers use LLPs to manage capital ratios. Pérez et al. (2008) conclude almost the same for Spanish banks. In addition, Cheng et al. (2011) found evidence that bank

(10)

managers are more likely to manage earnings if capital ratios are closer to the minimum required capital and Herbohn et al. (2010) found evidence for income increasing earnings management via unrecognized tax assets due to carryforward tax losses. Although other researchers found similar evidence with respect to DTAs (Schrand and Wong, 2003; Christensen et al., 2008), explanations for European banks still remain unexamined.

To date, to the best of my knowledge, empirical evidence to detect capital management through (un)recognized deferred tax assets due to carryforward tax losses accounted for under IAS 12 is inconclusive. Therefore this study extends prior research by examining the

impact of unrecognized deferred tax assets due to tax losses (VA) accounted for under IAS 12 on regulatory capital by European banks. With this extension this study creates the bridge to banking regulation where other studies, for example Alford et al. (1993) did not. This study also extends a survey of KPMG (2011; 2012) by providing empirical evidence if European banks manage capital ratios through carryforward tax assets due to tax losses accounted for under IAS 12. Finally, this study responds to Epe (2010) and Epe and Langendijk (2012) who argue that firms should recognize all deferred taxes in financial statements. Results of this study could be used to determine if Basel III will probably achieve its goals to reduce information asymmetry and increase the resilience of banks through stronger and larger capital buffers. These results will also contribute to existing literature about the theory of agency, earnings management and financial reporting by providing evidence whether banks reduce or increase information asymmetry via unrecognized deferred tax assets due to tax losses (e.g. Ball and Brown, 1968; Warfield and Wild, 1992; Aboody and Kasznik, 2000; Bergstresser and Philippon, 2006).

1.2 Research question

EU banks need to comply with Basel III capital requirements as of January 2014. Deferred taxes are specifically highlighted in Basel III and specific parts of DTAs need to be deducted in calculation of CET1. The BCBS recognized the possibility by managers to use DTAs for capital adequacy. DTAs have importance for financial reporting as well as prudential reporting. DTAs are part of the accruals used in financial reporting to present fair financial position, financial performance and cash flow of an entity. Loss absorption capacity of DTAs is assessed for prudential reporting. This study will conduct research on the question if European banks use their discretion on valuation allowance (VA) for capital management purposes. This study will research this by answering the following main research question:

Do European banks use their discretion to determine valuation allowance accounted for under IAS 12 to manage regulatory capital?

By answering the main research question this study investigates whether there is a relation between unrecognized deferred tax assets due to carryforward tax losses and capital management by European banks. This answer enables to identify recommendations for standard setters and supervisory boards on the (possible) effectiveness of the new capital requirement, known as Basel III. After all Basel III aims to contribute to a more stable financial system. Before answering the main research question, the following sub questions need to be answered:

1) How do deferred taxes arise?

2) When are deferred taxes recognized and disclosed in financial statements under IFRS? 3) How are deferred taxes measured in financial statements under IFRS?

(11)

5) Do risky and/or well- and poorly capitalized banks recognize DTAs for carryforward tax losses differently?

6) What are the relevant capital rules for deferred taxes under Basel III?

7) Do system banks have significantly higher or lower VA compared to non-system banks?

1.3 Conceptual framework

This paragraph explains the main concepts of this thesis and gives insight in (mutual) relationships. The main research question in this study is whether European banks use unrecognized DTAs due to carryforward tax losses (hereafter: VA) for capital management. The stippled arrows display the main elements studied in this main research question. In order to obtain adequate data it is necessary to use proxies in this study, like accounting income as a proxy to determine (un)recognized DTAs due to carryforward tax losses. Figure 1 illustrates these elements. These elements and their relations will be explained in subparagraph 1.3.1.

Figure 1: Conceptual Model

1.3.1 Explanation and relations of concepts

EU banks use three different standards to report. First, financial reporting which provides information about the economic performance of a firm (Scott, 2012). Financial reporting stands for the accounting standard applied by EU banks (i.e. IFRS). Second, tax reporting (i.e. tax laws) to determine annual profit so accurate that it justifies paying tax (Bruins Slot and Gerrits, 2009). Tax reporting stands for tax laws which are applied to determine taxable income. Finally, prudential reporting which determines capital adequacy through loss absorption capacity of equity (BIS, 2006; 2011a). Prudential reporting stands for capital requirements, in this study Basel III which is effectuated via the EU Capital Requirement Directive IV (CRD IV). The main differences between these standards and definitions, as stated in figure 1, are explained below.

(Taxable) Income or Loss Current Tax Expenses Deferred Tax Expenses Deferred Tax Assets Temporary Differences (un)recognized Carryforward Tax Losses Adjustments Accounting Equity Regulatory Capital

(12)

Financial reporting vs. tax reporting

Reported income is determined by IFRS. Taxable income is determined by tax laws.

Differences between IFRS and tax laws lead to differences in tax payments and tax expenses.

Current tax expenses represent the payable taxes and these are based on tax laws and so on

taxable income. Deferred tax expenses represent the amounts of taxes which are taxable (i.e. payable) or deductible (i.e. receivable). On the one hand, deferred taxes arise because of

temporary differences and will offset in the future, for example differences in depreciation

periods (Epe, 2010). On the other hand, deferred taxes arise because of compensation of tax

losses. According to tax laws these losses could be offset with profits, but there is no certainty

because offsetting depends on future profitability (IAS 12.36). Total tax expenses in financial statements are the sum of current and deferred tax expenses.

Financial reporting vs. prudential reporting

According to article 4 (53) of the CRD IV prudential filters are applied to accounting equity which is based on IFRS. These prudential filters are called adjustments and remove equity components which have little or no value to a bank in ‘’stress’’ situation (Gunst, 2013). According to article 33 (c) of the CRD IV, DTAs that rely on future profitability subtracted from CET1.

Since the purpose of financial reporting, tax accounting and prudential reporting differs, it is not surprising that these standards will lead to different tax treatment. For a better understanding of the purposes of these standards, figure 2 is composed.

Figure 2: Purpose of reporting standards

1.4 Structure

This thesis has the following structure. Chapter 2 contains the institutional framework of this thesis. Important definitions of IAS 12, other IFRS standards and Basel III will be explained and illustrated. For a better understanding of this technical complex subject also journal entries are given when necessary. Chapter 3 will contain literature review and hypothesis. Chapter 4 will give insight in the research population and design. The chosen research method is also explained in chapter 4. Results will be discussed in chapter 5 and chapter 6 will contain conclusions and limitations of this study.

Reporting Standard Purpose Source

Financial IFRS Provide useful information to stakeholders about economic performance

Scott, 2012 Tax Tax laws Determine the annual profit so accurate that it justifies

immediately paying tax

Bruins Slot and Gerrits, 2009

Prudential Basel III Provide useful information about the loss absorption capacity of equity

(13)

2 Accounting for taxes under IFRS and Basel III

All banks have to deal with taxes and capital requirements. Differences between IFRS and tax laws lead to differences between reported and paid taxes. Tax treatment in financial reporting is called tax accounting. The main difference is because of accrual and cash based accounting. Knowledge about tax regulation and determination of taxable income is necessary to understand tax treatment in financial reporting. Therefore, paragraph 2.1 will provide insight with regards to determination of taxable income and carryback and carryforward facilities. Tax accounting under IAS 12 is explained in paragraph 2.2.

Due to Basel III capital requirements banks need to increase their capital ratios and deduct (some parts of) DTAs in calculation of CET1. For DTAs arising from temporary differences, thresholds are embedded in calculation of regulatory capital. Paragraph 2.3 explains the relation between DTAs and capitalization.

2.1 Tax treatment according to tax laws

Since it is impossible for this thesis to discuss national tax regulation for all EU countries, tax treatment will be explained from the perspective of the Netherlands. The main difference will probably arise in determining taxable income. EU countries will use their discretion to determine which income is subject to taxes. For example, the Netherlands have an associates exemption. The carryback and carryforward facilities regarding tax losses are almost similar for all EU countries. Difference between countries arise mostly in the length of the period tax losses could be utilized. Current tax regulation in the Netherlands decides that tax losses could be utilized with profits from a year-earlier and future profits in the following nine years. Under Sec. 3.25 of the Netherlands Income Tax Act of 2001, the yearly profit in the Netherlands is determined ‘in accordance with sound business practice’ (see also Bruins Slot and Gerrits, 2009). The Netherlands Supreme Court decided that sound business practice could be read as ‘accepted business economic principles regarding profit determination’ (see Bruins Slot and Gerrits [2009] for further discussion).1 Sound business practice seems similar to IFRS regarding determination of income. This means that accounting income need to be adjusted; (1) in case it conflicts with tax law and (2) in case it conflicts with the purpose of tax laws (Bruins Slot and Gerrits, 2009). Assume that a bank has a profit of €10 million based on IFRS, including non-deductible cost for tax purposes of €1 million. The profit for tax purposes will be €11 million because those cost or not deductible according to tax laws. If the bank has again a profit of € 10 million based on IFRS, including € 15 million profit from associates. The loss for tax purposes will be €5 million because of the associates exemption. In case of a loss taxes paid a year-earlier could be claimed through an carryback. A carryback could be claimed to the amount for which a year-earlier tax was paid (Epe, 2010). For the remaining or for all tax losses (in case a carryback is not possible) a carryforward ‘receivable’ on the tax authority could be recognized. Losses from the past could be offset against future profits. Blokdijk (1985) refers to this as the ‘final balance’ for tax charging, taxable profit minus tax losses (see paragraph 2.2 and especially Epe [2010] for a detailed discussion).

1

(14)

2.2 Tax accounting under IAS 12

Financial reporting is based on accrual accounting. This means that costs and revenues need to be allocated to years in which they occur. Income taxes are a part of the total costs in financial statements. Therefore, tax expenses need to be allocated to periods in which they occur and not to periods in which they will be paid (Epe, 2010; IAS 12.5). As mentioned in IAS 1.15, ‘Financial statements shall present fairly the financial position, financial

performance and cash flows of an entity’. Tax accounting deals with the question how taxes

should be processed in financial reporting for a fair presentation (Gerrits et al., 2011). This means that a deferred tax asset (liability) needs to be recognized for all differences which could lead to smaller (larger) future tax payments (IAS 12.10). IAS 12 is the standard which shall be applied in accounting for income taxes (IAS 12.1). In this case income taxes also include other taxes, such as withholding taxes (IAS 12.2). Simplified, it can be said that IAS 12 is applicable for all tax issues in financial statements.

Differences between financial reporting and tax reporting which will offset in the future are because of allocation, so-called temporary differences (Epe, 2010). These differences exist of two parts. First, because of differences in timing of tax expenses or payments and differences in carrying amounts of assets and their tax base (IAS 12.IN2; IAS 12.5). These differences will disappear in the future and have no effect on reported tax expenses (IAS 12.6; Epe, 2010). Second, because of carryforward of unused tax losses (IAS 12.34). These differences are able to affect reported tax expenses and so the effective tax rate (Gerrits et al., 2009). Further explanation is given in paragraph 2.2.1 for temporary differences and in paragraph 2.2.2 for unused tax losses and credits.

Differences between financial reporting and tax accounting which will not offset in the future are because of accounting policies, so-called permanent differences. These differences have impact on the effective tax rate (Gerrits et al., 2011). For example, profit (loss) of associates increase (decrease) reported accounting income but have no effect on taxable income. The Dutch associates exemption and other types of permanent differences are further explained in paragraph 2.2.3. Recognition, measurement and disclosure are explained in paragraph 2.2.4. 2.2.1 Temporary differences

As mentioned before, temporary differences will normally offset in the future and have no effect on reported tax expenses. IAS 12.5 defines temporary differences as a difference between the carrying amount of an asset or liability in the statement of financial position and its tax base. The amount attributed for tax purposes is the tax base of an asset or liability. IAS 12.5 also distinguishes between taxable – and deductible temporary differences. Differences which will result in taxable (deductible) amounts in determining taxable profit or loss of future periods when the carrying amounts of an asset or liability is recovered or settled, are called taxable (deductible) temporary differences. These differences arise because of different accounting policies. An asset with an economic useful life of 5 years could be treated differently for accounting and tax purposes in case of depreciation. Early depreciation could lead to delay of current tax expenses and an interest benefit. There will be no impact on the effective tax rate and differences will disappear at the end of year five (Epe, 2010). IAS 12.86 defines the effective tax rate as the tax expense/income divided by accounting profit. This phenomenon will be illustrated in an example.

A bank operates in an area with a tax rate of 25% and calculates depreciations of property in own use to determine accounting and taxable income, in accordance with e.g. IAS 12 and IAS

(15)

16. For accounting purposes depreciations will be €1.000 for all 5 years. For tax purposes depreciations will be €2.500 for year 1 and 2 and zero for years 3, 4 and 5. Annual profit is €10.000 and there are no other differences. Table 1 will illustrate the movements of depreciations and their tax effects. Table 2 will provide the journal entries and show how these differences disappear and have no impact on the effective tax rate.

Table 1: Example with depreciation and deferred taxes

Table 2: Example with the effective tax rate and temporary differences through different depreciations policies

This example was a first step to understand the relation between the effective tax rate and deferred taxes. By accounting for deferred taxes, the effective tax rate remains 25% and contains a current tax expense of 21% (1.875/[10.000-1.000]*100%) and a deferred tax expense of 4% (375/[10.000-1.000]*100%) for the first two years. The accounting profit is calculated by subtracting depreciations from the annual profit. For years 3, 4 and 5 the effective tax rate is also 25% and contains a current tax expense of 28% (2.500/[10.000-1.000]*100%) and a deferred tax expense of -3% (-250/[10.000-(2.500/[10.000-1.000]*100%). A reversed journal entry for deferred tax liability (DTL) is made during year 3, 4 and 5. The logic behind this reversed journal entry is given by IAS 12. Imagine a revaluation reserve which is accounted in a company’s equity. In that case, IAS 12.61A determines that tax consequences of a revaluation reserve should also be processed directly in equity.

Directly accounting tax consequences points out another issue which shall be illustrated by an example.2 A bank decides to use properties to earn rentals and for capital appreciation. Therefore, properties are qualified as investments in accordance with IAS 40.5. This change of accounting policy leads to movements in financial position and profit and loss statements.3 In line with IAS 40.35, a gain or loss arising from a change in fair value is recognized in the

2

This example is based on Epe (2010), casus 3.4 on page 87. Contrary to Epe’s case the change of accounting policy is only applied for financial reporting and not for tax reporting. This example intends to provide insight in temporary differences were Epe’s example provides also insight to permanent differences.

3_{The official definition as used in IFRS is ‘Statement of comprehensive income’. In this study profit and loss} statement will be used.

Year Accounting depreciation Tax depreciation DTA DTL 1 1.000 2.500 375 2 1.000 2.500 375 3 1.000 0 250 4 1.000 0 250 5 1.000 0 250 5.000 5.000 750 750

Debet Current Tax Expense 1.875 Debet Current Tax Expense 2.500

Credit Taxes Payable 1.875 Credit Taxes Payable 2.500

Debet Deferred Tax Expense 375 Debet Deferred Tax Liability 250 Credit Deferred Tax Liability 375 Credit Deferred Tax Expense 250

Year 1 and 2 Year 3, 4 and 5

(16)

profit and loss statement for that period. According to IAS 8.26, the cumulative effect of the change of policy needs to be accounted in the bank’s equity for the opening and closing period. For tax purposes properties are still measured at historical costs. Property has a carrying amount of €250.000 based on historical costs and €330.000 based on fair value. Tax rate is again 25%. Change of policy leads to the following journal entry:

(1) Debet Property €80.000 (330.000 – 250.000)

(2) Credit Retained Earnings €60.000 (330.000 – 250.000)*75% (3) Credit Deferred Tax Liability €20.000 (330.000 – 250.000)*25% Journal entry (1) shows the capital appreciation on property through fair value and journal entry (2) shows the net amount transferred to the bank’s equity (100% - 25% = 75%). IAS 8.26 does not permit use of other components of equity to account the net amount of capital appreciation. In accordance to IAS 12.61A, journal entry (3) shows the tax consequences of this change in policy. The following movements have also occurred during the year. Rental earnings €10.000 and change in fair value of €40.000. This leads to the following journal entries:

(1) Debet Cash €10.000

(2) Debet Property €40.000

(3) Credit Interest Income €50.000

(4) Debet Current Tax Expense € 2.500 (10.000*25%) (5) Debet Deferred Tax Expense €10.000 (40.000*25%)

(6) Credit Taxes Payable € 2.500 (10.000*25%)

(7) Credit Deferred Tax Liability €10.000 (40.000*25%) According to IAS 40.35 a gain or loss arising from a change in fair value is recognized in the profit and loss statement for that period, see entries (1), (2) and (3). For tax reporting the upward raise of fair value is recognized when the property is sold. Therefore a deferred tax liability arises over the change of €40.000 in the fair value and tax payment is only required over the earned rentals, see entries (4), (5), (6) and (7). It is important to notice that in case of a decrease in the property’s fair value a loss needs to be deducted immediately in determining

taxable income (Bruins Slot and Gerrits, 2009).4

With the previous two examples tax consequences for differences in timing of expenses (i.e. depreciations) and accounting directly in equity (i.e. change in policy) are clarified. However, a final issue of these two combined is not discussed, for example a compound financial instrument. IAS 12 (example 4) and Epe (2010) provide an excellent example to illustrate this phenomenon and the following example is based on both. A bank has a convertible loan of €100.000 with a repayment date over 5 years and interest payments at the end of each year. According to IAS 32.28 and IAS 32.29 a compound financial instrument exists of an equity and liability component. The coupon interest rate is 8% and the market interest rate for a similar instrument is 10%. For tax reporting this separation is not made and the convertible loan is fully recognized as a liability. The tax rate will be 25%. This leads to the following journal entry in financial statements:

(1) Debet Cash €100.000

(2) Credit Equity Component € 5.686,18 [(1) – (3)]*75%

4

(17)

(3) Credit Liability Component5 €92.418,43 see footnote 5 (4) Credit Deferred Tax Liability € 1.895,39 [(1) – (3)]*25%

In line with IAS 12.23, journal entry (4) represents the tax consequence of the equity component. The difference between (1) and (3) represents the premium for this loan. Since this component is not accounted for under tax reporting, there is a tax claim on this premium in the future. This DTL needs to be subtracted from the equity component. This subtraction is in line with the same principle as mentioned in the previous example (IAS 12.61A).

As mentioned, interest payments are at the end of each year. For year 1 this leads to the following journal entry:

(1) Debet Interest Expense €9.241,84 (92.418,43*10%)

(2) Credit Cash €8.000 (100.000*8%)

(3) Credit Liability Component €1.241,84 [(1) – (2)]

Journal entry (1) presents the interest expense in the profit and loss statement, which is calculated by the market interest rate (10%) and the carrying amount of the convertible loan (the net present value). Journal entry (2) represents the interest payment. Journal entry (3) represents the liability component as the difference between interest expense and interest payment. The sum of the liability components over five years and the net present value at year 1 makes sure that the convertible loan reaches the nominal value of €100.000, see table 3.

Table 3: An overview of the carrying amount and nominal value of the compound financial instrument

At year-end the tax consequences need to be processed and that will lead to the following journal entry:

(1) Debet Taxes payable €2.000 (8.000*25%)

(2) Debet Deferred Tax Liability € 310,46 [(3) – (1)] (3) Credit (Deferred) Tax Expense €2.310,46 (9.241,84*25%) Journal entry (1) represents the amount which can be deducted from taxes payable due to interest payments for liabilities. Journal entry (2) represents the amount of interest expense which are not deductible in calculating taxable income. Journal entry (3) represents the amount which shall be deducted in calculations of tax expenses for financial reporting. If this were the only activities of the bank, the journal entry would be exactly the opposite. It is very important to understand journal entry (2) because this entry proves that the DTL will

disappear after year 5. To increase the understandability, movements in DTL are displayed in

table 4.

5_{8.000/1,1^1 + 8.000/1,1^2 + 8.000/1,1^3 + 8.000/1,1^4 + 108.000/1,1^5 = 92.418,43. This represents the net} present value of the convertible loan.

Year 1 Year 2 Year 3 Year 4 Year 5

NPV Liabilitty Component € 92.418,43 € 93.660,27 € 95.026,30 € 96.528,93 € 98.181,82 Interest Expense € 9.241,84 € 9.366,03 € 9.502,63 € 9.652,89 € 9.818,18 Cash (Current Expense) € 8.000,00 € 8.000,00 € 8.000,00 € 8.000,00 € 8.000,00 Liability Component € 1.241,84 € 1.366,03 € 1.502,63 € 1.652,89 € 1.818,18

(18)

Table 4: Movements in DTLs and disappearance of DTL after year 5

Finally, Epe (2010) gives several examples where the settlement date of temporary differences is foreseeable. There are also situations where the settlement is not foreseeable. In accordance with IAS 16.31 properties could be measured at fair value for financial reporting and at historical cost for tax reporting. In contrary to financial reporting commonly there is no depreciation for tax purposes. Due to unforeseeable settlement date this situation leads to the issue if these differences should be recognized or not. It could take a long time before these differences will offset. IAS 12.15 requires a DTL for all taxable temporary differences (comprehensive allocation), so also these kind of differences need to be recognized (see also Epe, 2010). According to IAS 12.24 ‘a DTA shall be recognized for all deductible temporary

differences to the extent that it is probable that taxable profit will be available against which the deductible temporary difference can be utilized’. Epe and Langendijk (2012) give

arguments that not only the extent of taxable profit but also the extent of DTLs is sufficient evidence to recognize a DTA (see also Blokdijk, 1985; Schrand and Wong, 2003; Epe, 2010). At this point important sources of temporary differences are discussed and explained. Measurement, recognition and disclosure will be discussed in paragraph 2.2.4. Therefore little attention is paid on these items. The following paragraph will discuss the main element of this study, unused tax losses.

2.2.2 Unused tax losses and credits

The previous paragraph explained different types of temporary differences, also called certain deferred taxes. Contrary to certain deferred taxes, this paragraph discusses uncertain deferred taxes. Uncertain deferred taxes rely on future profitability (Epe and Langendijk, 2012). Unused tax credits are called carryback tax losses and represent certain DTAs. This will be shown by the following example based on Epe (2010).

A bank has an annual profit of €100.000 in year 1 and an annual loss of €200.000 in year 2. There are no other differences and the tax rate is 25%. In year 1 regular tax expenses of €25.000 (25%*100.000) are reported and paid. The reported negative tax expenses in year 2 exist of two components, a carryback tax loss of €25.000 (25%*100.000) and a carryforward tax loss of €25.000 [(200.000 – 100.000)*25%]. Table 5 shows that the effective tax rate remains 25% when a bank uses the carryback and carryforward facilities. If the carryforward facility was not used the effective tax rate would be 12.5% [(25.000/200.000)*25%], which means a higher effective tax rate exist in case of a loss!

Year 1 Year 2 Year 3 Year 4 Year 5

Current Tax Expense € 2.000,00 € 2.000,00 € 2.000,00 € 2.000,00 € 2.000,00 Deferred Tax Liability (DTL) € 310,46 € 341,51 € 375,66 € 413,22 € 454,55 (Deferred) Tax Expense € 2.310,46 € 2.341,51 € 2.375,66 € 2.413,22 € 2.454,55 Opening DTL at 25% € 1.895,39 € 1.584,93 € 1.243,43 € 867,77 € 454,55 Deferred Tax ''Income'' € -310,46 € -341,51 € -375,66 € -413,22 € -454,55 Closing DTL at 25% € 1.584,93 € 1.243,43 € 867,77 € 454,55 € 0,00

(19)

Table 5: Accounting for carryback and carryforward tax losses

The amount of the carryback facility is a certain deferred tax and therefore needs to be recognized (IAS 12.13). To recognize the carryforward tax loss there are some criteria. According to IAS 12.36 a company should asses the following criteria before recognizing a carryback or carryforward tax loss:

(a) whether the entity has sufficient taxable temporary differences relating to the same taxation authority and the same taxable entity, which will result in taxable amounts against which the unused tax losses or unused tax credits can be utilized before they expire;

(b) whether it is probable that the entity will have taxable profits before the unused tax losses or unused tax credits expire;

(c) whether the unused tax losses result from identifiable causes which are unlikely to recur;

(d) whether tax planning opportunities are available to the entity that will create taxable profit in the period in which the unused tax losses or unused tax credits can be utilized.

Like Gerrits et al. (2011) and Brouwer et al. (2012) mention main issues around recognizing DTAs due to carryforward tax losses arise at point (a). IAS 12.35 points out that in case of a history with losses an entity shall only recognize a DTA due to carryforward tax losses to the extent that the entity has sufficient taxable temporary differences. An illustrative example of Epe (2010) will be used to explain the difficulty with carryforward losses and the extent of taxable temporary differences depends on future profitability, so-called uncertain DTAs. In year 1 bank X has an annual profit of €500.000. For financial reporting purposes it charges costs of €100.000 fully to year 1 and for tax reporting the amount will be €50.000 for each year. There are no other differences and the tax rate will be 25%. This leads to the following journal entry in year 1:

(1) Debet Tax Expense €100.000 (400.000*25%)

(2) Debet DTA € 12.500 (50.000*25%)

(3) Credit Current Tax Expense €112.500 (450.000*25%)

In year 2 the annual profit is €20.000 for financial reporting, the annual loss for tax reporting will be €30.000 (see table 6). In that case the journal entry for year 2 will be:

(2) Debet Carryback €7.500 (30.000*25%)

(3) Credit DTA €12.500 (50.000*25%)

Journal entry (3) proves that the DTA is offset against a carryback claim of €7.500 and an allocation of tax payment of €5.000 in year 1 to a tax expense in year 2.

Accounting Tax Accounting Tax

Profit or Loss € 100.000 € 100.000 € -200.000 € -200.000 Tax 25% € -25.000 € -25.000 € 50.000 € 50.000 Profit or Loss after Taxation € 75.000 € 75.000 € -150.000 € -150.000

Effective Tax Rate 25% 25% 25% 25%

(20)

In addition to earlier information this bank has also research and development (R&D) costs. R&D costs are recognized as intangible assets for financial reporting purposes following IAS 38 and are depreciated in five years, starting at the beginning of year 2. For tax reporting these costs are directly recognized in the profit and loss statement. Finally, there will be no profit until year 6. Consequences for accounting and taxable income are displayed in the table below. The only difference in years 3 until 6 are depreciations on R&D costs, leading to a accounting loss of €20.000.

Table 6: Movements between accounting and taxable income to determine carryforward tax losses

This leads to the following journal entry in year 1:

(2) Debet DTA € 12.500 (50.000*25%)

(3) Credit Current Tax Expense €87.500 (350.000*25%)

(4) Credit DTL €25.000 (100.000*25%)

For year 2 the following journal entry is made:

(1) Debet Tax Expense € 0 (0*25%)

(2) Debet DTL €5.000 (100.000*1/5*25%)

(3) Debet Carryback €7.500 (30.000*25%)

(4) Credit DTA €12.500

Journal entry (4) represents the reversal of the DTA. This is because of two reasons. First, no tax payment is necessary in year 2 so the DTL is also reversed (€25.000 – €5.000). Second, a carryback is claimed on the €87.500 taxes paid in year 1. Until now it is certain that the DTA will be offset with the DTL and therefore is fully recognized in year 1 (IAS 12.28).

To make this example complete, assume that the bank has existed already a couple of years and has no possibility for a carryback because of losses in previous years. Accumulated tax losses are €500.000 and therefore a carryforward of €125.000 (€500.000*25%) exists. This will lead to the following journal entries:

For year 1:

(1) Debet Tax Expense €100.000

(2) Debet DTA € 12.500

(3) Credit Carryforward Tax Loss €87.500

(4) Credit DTL €25.000

For year 2:

(1) Debet Tax Expense € 0

(2) Debet DTL €5.000

(3) Debet Carryforward Tax Loss €7.500

(4) Credit DTA €12.500

Accounting Tax Accounting Tax Accounting Tax

Profit or Loss € 500.000 € 500.000 € 20.000 € 20.000 € - € -Costs € -100.000 € -50.000 € - € -50.000 € - € -R&D costs € - € -100.000 € -20.000 € - € -20.000

Profit Before Tax € 400.000 € 350.000 € - € -30.000 € -20.000 €

(21)

For years 3 until 6:

(1) Debet DTA €5.000

(2) Debet Deferred Tax Expense € 5.000

The difference with the previous two journal entries are the carryforward tax losses. DTAs are reversed in carryforward tax losses for years 1 and 2. For years 3 until 6 there are no such journal entries. Therefore, it can be said that DTAs can only be offset with DTLs if there is tax payment required (Epe, 2010; Gerrits et al., 2011). There is after all no certainty that carryforward will lead to a cash flow. In this example an amount of €5.000 of DTAs is reversed because of the presence of DTLs with also the same settlement date. In line with IAS 12.28 DTAs should only be recognized because these DTAs reflect a cash flow (see also Schrand and Wong, 2003; Epe and Langendijk, 2012). Recognizing carryforward tax losses depends on if there will be enough sufficient future taxable differences and if there will be enough taxable income in the future, so-called uncertain DTAs (IAS 12.36).

Recently, Epe (2010) introduced a ‘norm’ to recognize all deferred taxes. This differs from requirements in IAS 12. Epe (2010) argues that allocation and going-concern principles have huge importance for financial reporting and therefore it is better to estimate a higher amount of carryforward tax losses instead of a lower amount (see also Blokdijk, 1985; Epe and Langendijk, 2012). If the carryforward could not be offset, this is because of insufficient future taxable profit and it still remains a fair presentation of economic performance. However, Epe (2010) provides a deviation of his own ‘norm’. If a company is able to determine the unusable part of the tax losses a VA can be established. This discussion is interesting when it is compared to capital regulation like Basel III. This will be further discussed in paragraph 2.3.2.

Finally, it is imperative to mention that in contrary to temporary differences recognition of carryforward tax losses could have impact on a firm’s effective tax rate (Epe, 2010). Previous examples have shown that recognizing a DTA due to carryforward tax losses does not have any impact on the effective tax rate. If a bank manager decides to not recognize a DTA he could manage reported tax expenses, since DTAs due to carryforward tax losses are part of the accruals used in financial statements. Epe (2010) mentions this phenomenon with evidence of the Dutch telecom provider KPN who reports a net profit of €2.6 billion of which €1.2 billion is due to tax benefits. Incentives to manage recognition of these DTAs are discussed in chapter 3 (see e.g. Kampschöer, 1993; Burgstahler and Dichev, 1997; Bartov and Mohanram, 2004; Herbohn et al., 2010).

2.2.3 Permanent differences

Permanent differences will not offset in the future and therefore have an impact on a firm’s effective tax rate (Epe, 2010). This phenomenon will be explained by an example of ING’s annual report over 2011. ING’s associates exemption and income which is no subject to tax decrease the effective tax rate of ING with 5.4% and 4.3%, respectively.6 It is important to understand that permanent differences are not subject to deferred taxes. Permanent differences arise because of differences in purposes of tax and financial reporting, see figure 2 in paragraph 1.3.1. Permanent differences will not be discussed further in this thesis.

6_{See ING annual report page 197, amounts in millions EUR. Associates exemption €329 and income which is} no subject to tax €263 are both divided by profit before tax €6.109.

(22)

2.2.4 Recognition, measurement and disclosure

All temporary differences shall be recognized to the extent there is evidence for enough future taxable income or the presence of sufficient taxable temporary differences (IAS 12.15; IAS 12.24; IAS12.35). Sufficient taxable temporary differences are also an important criterion to determine a carryforward tax loss (IAS12.35). DTAs are required in cases of measurement at higher fair value, for example revaluation reserves (IAS 12.20). To offset a DTA with a DTL a legal right to offset and the intention to settle on a net basis must exist (IAS 12.71; IAS 12.72).

Brouwer et al. (2012) argue that deferred taxes should be discounted. IAS 12.53 prohibits any form of discounting and deferred taxes should be measured at nominal value. Current tax expenses (IAS 12.54n) and deferred taxes (IAS 12.54o) are disclosed separately in the statement of financial position. According to IAS 1.56 DTAs (DTLs) are disclosed as non-current assets (liabilities).This means deferred taxes are not provisions as accounted for under IAS 37 Provisions, Contingent Liabilities and Contingent Assets. Finally, IAS 12.81e requires a disclosure for unused tax losses, also called ‘valuation allowance’ (further referred as VA).

2.3 Capital requirements

After the First World War the Bank for International Settlements (BIS) was established in 1930. Their main goal quickly moved from dealing with issues of the reparation of payments on Germany after the First World War to realize monetary and financial stability.7 In 1988 the first Basel Capital Accord was published. Basel I was mainly focused on mitigating credit risk. The minimum capital requirements were 8% of the risk-weighted assets (RWAs). In the period 2001-2006 Basel II was developed. With this revised capital standard a link to financial reporting was made very easy. The title of Basel II is as follows: ‘’International

Convergence of Capital Measurement and Capital Standards’’ (BIS, 2006). The RWAs are

based on credit assessments of the assets. For example, governments with an AAA up to AA rating had a risk-weight of 0% according to Basel II credit assessments (BIS, 2006).8 Banks could also decide to use their own credit ratings based on their own developed credit rating systems, so-called internal ratings-based (IRB) approach (BCBS, 2006, p.12). Banks use this possibility to determine a ‘more precise’ Probability of Default (Bout, 2011). To increase comparability banks disclose reconciliation between their IRB approach and standardized ratings. To give an example, ABN AMRO’s IRB approach, the so-called ‘Uniform Counterparty Rating’ is displayed in figure 3.9

Since these different approaches are not in the scope of this thesis further details are not discussed.

Figure 3: Reconciliation between internal credit ratings with external credit ratings, Annual Report ABN AMRO 2010, p. 73.

7_{For further information see: http://www.bis.org/about/history.htm}

8_{The previous example was based on the standardized approach where credit ratings are used from credit} agencies like Moody’s and Standard & Poor’s.

9

For further explanation about these ratings see Bout (2011).

Uniform Counterparty Rating: 1 2+ to 2- 3+ to 3- 4+ to 4- 5+ to 5- 6+ 6 to 8

Expected default rates (%) Fitch 0-0.03 0.03-0.13 0.13-0.46 0.46-2.22 2.22-16.97 16.97-100.00 100 Standard & Poor's/Fitch AAA/AA- A+/A- BBB+/BBB- BB+/BB- B+/B- CCC+/C -Moody's AAA/Aa3 A1/A3 Baa1/Baa3 Ba1/Ba3 B1/B3 Caa1/C

(23)

-Basel II required banks to have stronger and higher qualified capital buffers and therefore divided capital in different tiers (‘subgroups’). The highest quality capital is known as Core Tier 1 Equity (hereafter: CT 1), consisting out of equity and disclosed reserves (BIS, 2006, 49i). CT 1 should be at least 2% of the total tier 1 capital ratio. Other Tier 1 capital elements could be for example minority interests. The total Tier 1 ratio should be at least 4%. Tier 2 ratio will be maximized at 100% of the Tier 1 ratio and consists of for example revaluation reserves transferred from Tier 1.

During the financial crisis the capital requirements were criticized. Banks were not able to ‘carry’ their losses and governments had to bail out. Therefore Basel II has been revised. The new standard, Basel III, requires higher and stronger capital buffers. Since Basel III is only a principle, EU countries will need to implement new capital requirements through the Capital Requirements Directive IV (CRD IV). To increase comparability between Basel II and III, paragraph 2.3.1 will discuss changes in Basel III. Hereafter, paragraph 2.3.2 will discuss the adjustments required by Basel III and CRD IV. Finally, paragraph 2.3.3 will discuss and explain how DTAs could affect regulatory capital.

2.3.1 Basel III

A stable financial system still remains the main goal. A main difference between Basel II and Basel III is classification of equity. This classification is made through a quality and quantity dimension (BIS, 2011a, paragraph 3-10). On one hand, Basel III tries to reduce the information asymmetry with respect to stakeholders by strengthening capital buffers (quality dimension) and requiring higher capital ratios (quantity dimension) to ensure money of the stakeholders. On the other hand, Basel III aims to increase the loss absorption capacity of banks through higher and stronger capital buffers. A comparison between Basel III and Basel II capital requirements is displayed in table 7. Basel II does not contain a countercyclical, conservation and systematically important banks buffer. For reasons of comparability no value is given to these buffers under Basel II.

Table 7: Comparison of capital requirements for Basel III and Basel II. Source: BIS, 2006; 2011a, p. 69 and FSB, 2012

CET1 exists for example of common shares, share premium and retained earnings; additional Tier 1 capital exists for example of equity components which are subordinated to depositors and general creditors (BIS, 2011a, paragraph 55). Tier 2 capital exists of capital subordinated to depositors and general creditors of the bank. For example, general provisions for LLPs which are held against future, presently unidentified losses. There is a limit up to 1.25% of the RWAs (BIS, 2011a, paragraph 60). The conservation buffer tends to minimize the impact of

Basel III Basel II

Common Equity Tier 1/Core Tier 1 4,5% 2,0%

Capital Conservation Buffer 2,5% n.a.

Minimum common equity plus capital conservation buffer 7,0% 2,0%

Minimum Tier 1 Capital 6,0% 4,0%

Minimum Total Capital 8,0% 8,0%

Minimum Total Capital plus conservation buffer 10,5% 8,0%

Countercyclical buffer 2,5% n.a.

Buffer for systematically important banks 3,5% n.a.

(24)

losses. Increasing financial stability and depositors’ interests are the main reason for this buffer (BIS, 2011a, paragraph 123-129). The countercyclical buffer tends to mitigate procyclical effects in times of economic downturn (BIS, 2011a, p. 57). Finally, the buffer for systematically important banks is introduced. Size, interconnectedness, financial institution infrastructure, cross-jurisdictional activity and their complexity are indicators to determine systematically importance (BIS, 2011b). FSB (2012) published additional capital ratios between 1%-3.5% for systematically important banks. Citigroup, Deutsche Bank, HSBC and JP Morgan Chase are the four most systematically important banks and need to have 2.5% additional loss absorption buffers (FSB, 2012).

Since this thesis researches the relationship between VA and capital management, further details of Basel III (and also Basel II) are not discussed. In the following paragraph deferred taxes and regulatory capital will be discussed in more detail.

2.3.2 Deductions for DTAs required by Basel III and CRD IV10

As mentioned in paragraph 2.2 realization of deferred taxes is not always certain, especially this is the case with DTAs recognized due to carryforward tax losses. Generally, financial reporting is used as starting point to determine regulatory capital (Gunst, 2013). There are so-called ‘prudential filters’ and deductions which remove or reduce certain elements in the calculation of regulatory capital because of their limited ‘loss absorption capacity’. Mostly these elements have no (added) value in case of a ‘stress’ situation. Under Basel II capital requirements deferred taxes due to carryforward tax losses count fully towards regulatory capital (BIS, 2006; Gunst, 2013).

Under Basel III the BCBS acknowledges the uncertainty with respect to reported DTAs in financial statements of banks. The first deduction is required for DTAs recognized for carry forward tax losses. Paragraph 69 of Basel III removes DTAs that rely on future profitability (CRR article 36.1c). Basel III provides a transitional period. In 2014, at least 20% must be deducted and with steps of 20%, in 2018 this must be a 100% deduction (BIS, 2011a; CRR article 478). Article 478.3 (CRR) mentions that ‘competent authorities’ shall determine applicable percentages and for example the central bank of the Netherlands (DNB) has decided to follow the same transition periods (Lubberink and Bout, 2012). The second deduction is for DTAs arising from temporary differences. These differences are fully counted in CET1 unless they exceed a threshold. These thresholds are the so-called ‘bucket deductions’ (BIS, 2013b). First, paragraph 87 requires that a DTA must remain below 10% of CET1 (CRR article 48). In addition, paragraph 88 determines that DTAs may not exceed 15% of CET1 when they are aggregated with one of the two other elements listed in paragraph 87 (BIS, 2011a; Gunst, 2013).11 Any excess must be subtracted from CET1. CRR article 48.2(b) have raised this percentage to 17.65%.

As mentioned, financial reporting (i.e. IFRS for listed banks) is the starting point to determine regulatory capital. Title I definition 106 of CRR mentions that DTAs have the same meaning

10_{Although CRD IV (Commission Directive (EP-Council) no. 36/2013a) and CRR (Commission Regulation} (EP-Council) no. 575/2013b) are the legal basis for capital requirements, its content with regards to DTAs are almost similar to Basel III. In case of differences CRD IV and CRR are explained, otherwise only corresponding articles of CRD IV and CRR are mentioned.

11_{These are mortgage servicing rights and significant investments in common shares of unconsolidated financial} institutions.

Regulatory Capital and Tax Losses