In uncertainty financial markets, capital buffer is essential part of risk management to absorb unexpected losses and reduce the risk of failure. In past ten years, the Basel Committee has put attention on the capital adequacy of international banks to strengthen the stability of international financial system. Since the recent global financial crisis, there was a challenge for banks in allocating regulatory capital in difficulties period. Some authors judge that inadequate risk models and capital standards in the Basel II Accord is a major reason of bank failures. The purpose of this research is to examine how British banks react on their capital ratios particularly the impact of capital adequacy requirements of the Basel II during the financial crisis period 2008-2011. In our sample, 76 UK banks were selected with a total of 380 observations for testing the relationships between changes in capital ratios and bank size, profitability, changes in risk and regulatory pressure. Our results show that all UK banks' total capital ratio satisfied the 8% minimum capital requirement in the observation period, only one bank with tier 1 capital ratio failed to meet the 4% criteria in 2008 and 2009. In addition, bank size, risk and regulatory pressure were significantly related to capital ratios, while bank profitability was not statistically significant. In the end, the results suggest that the Basel II Accord was effective in increasing British banks capital levels and reducing their risk levels between 2008 and 2011.
Introduction to bank capital requirements and allocation
In high-volatile financial markets, banks are managing different financial risks in their daily business such as credit risk, market risk and operational risk. To minimise the risk of loss, they must effectively manage their assets and liabilities in safeguarding their solvency and overall economic stability (Bessis, 2010). Successful capital allocation is a long term goal of risk management practice. Banks combine risk management framework and capital modelling methodology to determine the capital needed in absorbing losses as well as protecting their depositors' funds. Since 1988, the Basel Committee's capital measurement framework and standards has been introduced in the Basel I. It provides supervisory rules for example the minimum capital requirement of total capital ratio and tier 1 capital ratio for banks to allocate sufficient amount of capital as a cushion against unexpected losses and failure risk. This is known as 'regulatory capital'. In 2004, the original Basel II has been published and the comprehensive version released in 2006. It revised the capital adequacy framework in facilitating banks to hold minimum level of capital to sustain wider range of risks including credit risk, operational risk and market risk. Now, 27 member countries of the BCBS are expected to meet financial regulatory and supervisory expectations.
One of the principles of the regulatory minimum capital requirement is to enhance banks' risk sensitivity and capital level. This is expected that regulatory pressure would be able to boost their total capital ratio and tier 1 capital ratio. Thus, regulatory pressure and bank capital allocation behaviour should be closely related and the efficiency of the Basel II Accord could be reflected from bank capital levels. However, the regulatory capital requirement may not always contribute to increase bank's capital levels due to bank capital decision that may affect by internal factors such as bank size, profitability and risk level (Jacques and Nigro (1997), Aggarwal and Jacques (1998), and Rime (2001)).
In the third quarter of 2007 the current global financial crisis has begun. This is the U.S. subprime crisis overspread the global financial markets that cause the whole world economy into downturn, especially in the banking industry. Banking sectors of many countries, include the UK, were severely affected by credit crunch. By 2009, the UK government spent a large amount of money in bailouts of commercial banks such as Lloyds TSB and Royal Bank of Scotland (RBS). Poor risk management, i.e. inefficient corporate governance and internal control, is considered as the major reason of the failures. But Danielsson (2008) argue that the main factor is inadequate risk models and financial regulations in the Basel II Accord. Financial crisis increases difficult challenges and dilemmas to banks in allocating capital adequacy level. This therefore raises a question: 'does the Basel II Accord is efficient in increasing banks' capital levels during financial crisis period?'. This research will investigate the relationship between regulatory pressure and bank capital ratios to identify the impact of regulatory capital standard on British banks capital behaviour in 2008-2011.
Motivation
Regulatory capital is the need for banks to set aside the required amount of capital in a prudent manner. The goal of the regulatory capital requirement is to encourage banks in increasing their capital ratios in order to minimise their unexpected losses and the probability of failure, and to protect their depositors' funds (Allen et al. 2004). Due to the collapse of 2008, the Basel Committee on Banking Supervision (BCBS) has been reformed. The Basel III initially developed in December 2010 and the implementation by member countries will begin on 2013. It strengthens the global capital framework such as increases the minimum capital requirement levels in order to further improve bank's ability in risk management and solvency. An effective capital allocation is a critical component to risk management and necessary to a long-term success of any banking organisation as well as economic growth (Goldsmith (1969), McKinnon (1973), Shaw (1973) and Greenwood and Jovanovic (1990)). Review banks' capital ratios can gain insight on their capital allocation behaviour and how they react to the current Basel Accord, the Basel II. And therefore we may able to predict the efficiency of a new Accord.
Until now, some empirical studies have already been made to the literature on bank capital allocation behaviour, for example, Jacques and Nigro (1997), Ediz et al. (1998) and Roy (2005). However, all such works are reviewing the impact of the Basel I on bank's capital ratios. And most of the works are investigated the capital allocation practice of the US banks, while lack of answer for the performance of UK banks. Consequently, this research will focus on the impact of the Basel II's capital requirements on British bank behaviour during the global financial crisis period 2008-2011. Through this research, theoretical and empirical frameworks will use to enhance understanding of current capital allocation behaviour of British banking organisations. This could not only help managers in effective risk management and communication with related parties, but also could enable decision makers in determination of optimal capital adequacy levels and regulators in the design of capital standard.
Aim and objectives
This study aims to critically examine the factors affecting UK bank's capital allocation behaviour particularly the impact of regulatory capital requirements in the periods of 2008 and 2011. 76 UK banks have been selected with a total of 380 observations for testing two hypotheses:
Hypothesis 1: Bank internal factors (i.e. size, probability and risk) and regulatory factors will significant influence UK bank's total capital ratio and tier 1 capital ratio.
Hypothesis 2: Regulatory pressure will increase bank's capital levels effectively.
This will therefore seek to answer the research question: 'Does the Basel II Accord is efficient in increasing banks' capital levels?'.
In order to observe capital levels of British banks, the study will answer the research question based on the following objectives:
To explore the literatures on the concept of capital allocation and capital regulation to gain insights into the potential factors affecting capital decision.
To state the levels of banks' capital ratios during the four years, whether they satisfied the current capital requirements of the Basel II.
To identify the relationships between bank capital ratios and regulatory pressure, and bank internal factors including firm's asset size, profitability and risk.
Research methodology and findings
The study has conducted a review of the literatures on the rationale of capital allocation, functions of regulatory capital, the developments of capital requirements of Basel Accords, the historical trends of UK bank capital ratios and potential factors affecting bank capital allocation behaviour in order to develop research hypothesis.
The study has carried out positivism as research philosophy with deductive approach to investigate UK banks' capital allocation practices between 2008 and 2011. We applied survey as research strategy and mono method in collecting, testing and analysing data for a cross-section study.
Large qualitative data were gathered from the internet information gateway, Bankscope database. Before data analysis, we prepared data for analysis following the three-stage process: editing, coding and entering data. And then we have adopted SPSS software for statistical analysis including descriptive statistics, significance test, correlation and multivariate linear regression.
Bank size, profitability, risk and regulatory pressure have put into quantitative analysis to discuss bank total capital ratio and tier 1 capital ratio. Our results show that bank size, risk and regulatory pressure were significantly related to capital ratios, while bank profitability was not statistically significant. Consequently, the results suggest that the Basel II Accord was efficient in raising British banks capital levels and reducing their risk levels in the surveyed period.
An overview of the contents
The purpose of this paper is to examine empirically the factors affecting UK banks' capital ratios in particular the impact of regulatory pressure. Chapter 1 of the paper provides a general research background on capital allocation and requirements in banking industry, a discussion of the need of the research, the research aim and objectives, and briefly highlights the methodology to be applied for the research. Chapter 2 presents a critical evaluation of views and experiences on capital allocation practice from related literatures, such as identifying the importance of capital allocation, functions of regulatory capital, Basel standards, historical trends of UK bank capital ratios and the impact of potential factors. And then detailed research methodology explains in chapter 3 to understand the research paradigm, sampling method, data collection method, factors selection, research hypothesis, the process of data analysis as well as the benefits and limitation of methodology. Chapter 4 is about the empirical results and analysis that presents the findings and answers the research question. Chapter 5 discusses the implications of research findings including compares between our findings and pervious literatures to address similarities and differences, and explains the implications of findings to the practice of risk management. In the end, chapter 6 concludes the dissertation by summarising key points and providing limitations and suggestions for further study.
Chapter 2: Literature Review
This chapter reviews bank capital allocation and regulatory requirements from pervious literatures. First, section 2.1 and 2.2 define the importance of capital allocation and functions of regulatory capital in banking organisations respectively. The development of the Basel Committee's capital requirement introduces in section 2.3 including the 1988 Basel Accord and the current Basel II. And then section 2.4 highlights the historical trends in UK bank capital ratios. Finally in section 2.5, bank capital allocation behaviour discusses by investigating a number of potential factors.
2.1 The importance of capital allocation
Figure : UK household bank deposits and loans in 1964-2009 (FSA, 2011)
According to Financial Services Authority (FSA) (2011), UK household bank deposits and loans rose significantly over the years 1964-2009. The Figure 1 represents that UK household bank deposits are greater than their loans until 1988, their borrowing more than savings. This shows that the increased level of loans is higher than the increased level of deposits. The level of loans went up around 60% during the forty-five years where it was only at approximately 15% in 1964 jumped to 75% of GDP in 2009, while the level of deposits increased just about 30%. Banks play an important role in societal and economic development. They act as financial intermediary in allocating money from depositors' funds to borrowers for various investments, such as housing, insurance, education and business growth (Gestel and Baesens, 2009). This intermediation process does not only provide deposit and loan supply but also for the liquidity of economy. Berger et al. (2010) believe that transaction costs can be reduced if intermediation is undertaken in an efficient manner. Low cost for both deposit and credit demands increases benefit to the related parties as well as the economy overall.
Figure : An overall of the banking system (Allen at el., 2004)
Bank for International Settlements (BIS) (2012) reports that the G10 central banks had 23,175 billion of US dollars in liabilities in 2009 increased to 24,703 billion by the end of 2011. Large growth in liabilities increases the risk of bankruptcy. In theory, bank's asset side and its liabilities and equity capital side of the balance sheet must be equal. This means increases in liabilities or equity, the levels of asset must go up to match it. When banks issue loans to individuals or businesses, there is a risk and uncertainty issue that loans may delay or may not get back from borrowers. If those borrowers default on their loans, bank loses money. Thus bank's equity capital acts a buffer to cover losses on assets. However, the decreasing equity capital decreases deposits or other funding sources and thereby bank faces to liquidity crisis. The collapse of Northern Rock in 2007 is the typical example. British bank Northern Rock was severely suffered in funding problem to repay loans. It also had liquidity problem due to its depositors withdraw their savings during its difficulties period (Shin, 2009). This case highlights the importance of capital allocation where banks must set aside sufficient amount of capital to absorb losses and protect their depositors to allow them to give back deposits at all times. Thereby the liquidity and solvency of banks are guaranteed.
Since a bank holds insufficient buffers, small shocks may trigger large losses or even insolvency. The determination of capital needs is a part of risk management activities. The practice of risk management is integrated with the capital modelling process and methodology, allowing management to effectively measure the capital needed and enhance capital levels against losses and failure. Since bank activity directly affects the economy, better capital allocation can be explained the relationship between financial development and economic growth. This supports by several empirical literatures including Goldsmith (1969), McKinnon (1973), Shaw (1973) and Greenwood and Jovanovic (1990). Furthermore, Bagehot (1873) documents that better capital allocation is a major reason for England's fast growth in the mid 19th century. Therefore effective capital allocation is necessary in order to minimise liquidity crisis and the risk of insolvency in stressed conditions.
2.2 Definitions and functions of regulatory capital
According to Mitchell (1984), capital is the first considering issue on bank daily operation and managerial decision because it plays the central role in the stability and soundness of bank. There are two types of capital which banks must take account to absorb losses and they are economic and regulatory capital. Gestel and Baesens (2009) draw a clear distinction between two types of capital. Economic capital is the amount of capital that a bank is required to cover losses in safeguarding economic solvency. It used to assess and evaluate performance of its venture, ensuring adequacy balance sheet capital for its operation (Bessis, 2010). In fact, there is no regulation for banks to hold economic capital. This means that economic capital is chosen by firm. Hence each bank has different models to calculate this capital in order to handle its volatility of business performance. Conversely, regulatory capital is the minimum capital required by regulation, the Basel Committee on Banking Supervision's (BCBS) Basel standards. Banks are required to comply with a set of supervisory rules and guidelines to determine the amount of capital and meet the minimum capital requirements.
Regulatory capital is regarded as a prudent risk management. It sets by regulators to encourage banks to provide higher liquidity in order to reduce unexpected losses and the probability of failure (Allen et al. 2004). Thus banks would be able to effectively protect their depositors as well as themselves against the costs of bankruptcy and agency problems. Myers (1977) argues that a bank with high default risk tends to reduce shareholders' incentives to contribute new capital even to fund profitable investment project. Regulatory capital enables banks to control the risk of default and the volatility of the trading book (Gestel and Baesens, 2009). This fosters enhance risk management in safeguarding the safety and soundness of capital structure and increasing investment opportunity. However, Allen et al. (2004) highlight a conflict of interest between regulators and banks. They state that regulators seek to place highest possible capital requirements on banks, while banks consider that capital held are a waste of resources as that could be invested elsewhere.
2.3 The Basel Committee's capital requirements
2.3.1 The 1988 Basel Accord
In 1988, BCBS established the Basel I 'International Convergence of Capital Measurement and Capital Standards'. It introduced a set of capital measurement framework and minimum capital requirements. The central purpose of the Committee's work is, first, to enhance the safety and soundness of the international banking system; and second, to provide homogenous framework for banks in different countries against risk. The minimum capital requirements such as the total capital ratio and tier 1 capital ratio are put into place for banks to hold a certain level of capital to sustain operating losses and decrease the risk of default. This is known as 'regulatory capital'. BCBS sets a framework for measuring capital adequacy mainly in relation to credit risk and the minimum standard to be achieved. (BCBS, 1988) This framework and the standards approved by the central bank Governors of Group of Ten (Belgium, Canada, France, Germany, Italy, Japan, the Netherlands, Sweden, Switzerland, the United Kingdom and the United States) and Luxembourg for their internationally active banks (Eyssell and Arshadi, 1990).
Table : Tier 1 and 2 capital elements (BCBS, 1988)
At the end of 1990, there was an interim minimum capital standard of 7.25% of which at least 3.6% is core capital element. By the end of 1992 the transitional period ends, a target standard ratio of total capital to weighted risk assets is set at 8%, of which core capital element is at least 4%. As the following equations:
The total capital is defined in two tiers which are required to hold at least 50% of tier 1 and the rest is tier 2. Table 1 shows that tier 1 is a bank's core capital consisted of permanent shareholders' equity and disclosed reserves and may be minority interests in the equity of subsidiaries in the case of consolidated accounts. The committee designs equity capital and disclosed reserves as the key elements of capital because they are the only common elements in all countries' banking systems and the most market concerns in the published accounts reflecting bank's capital adequacy, probability and ability in competition. Tier 2 is supplementary capital to hold against unforeseeable losses, comprising undisclosed reserves, asset revaluation reserves, general provisions or general loan-loss reserves, hybrid capital instruments, subordinated term debt. It is limited to a maximum of the total of tier 1. In calculating risk-based capital ratios, some deductions are made from the capital base, including goodwill deducts from tier 1 capital elements and investment in financial subsidiaries and other financial institutions deducts from tier 2 capital elements. (BCBS, 1988)
Risk weighted assets is about the riskiness of assets. The Basel I Accord sets the five risk weights for different categories of asset: 0, 10, 20, 50 and 100% (see Appendix I). Therefore the total risk weighted assets is the sum of multiplying these risk weights by individual asset values. This risk-weight approach provides an easy way to measure off-balance-sheet exposures and to compare between countries' banking systems in fairly basis. (BCBS, 1988) When risk weighted assets becomes higher, more cost of regulatory capital will be required by bank to meet minimum capital requirements (Gestel and Baesens, 2009). In other words, the decreasing riskiness of assets decreases the amount of capital to set aside.
Clearly, the aim of the Basel I is to govern the capital adequacy of member countries' international banks. The capital measurement framework with a minimum standard of 8% is used to encourage international banks to boost their capital levels in order to ensure that they are able to give back depositors' funds at all times. Hence, the Basel regulations is not only to protect the banks themselves but also to their customers as well as the economy overall.
2.3.2 The current Basel II
In 2004, The BCBS published a new Accord called 'Basel II Accord' and the comprehensive version released in June 2006. The committee revised the 1988 Basel Accord capital adequacy framework in further strengthening the soundness and stability of the banking systems and facilitating the implementation of stronger risk management practices (BCBS, 1988). It extends the risk considerations on capital requirements. In the past, the 1988 Basel Accord has only focused on credit risk. But in the current Basel II, operational risk and market risk also take into accounts in setting regulatory capital, benefiting banks to hold capital to sustain different types of risk.
The revised framework is divided into three pillars approach and they are minimum capital requirement, supervisory review process and market discipline and public disclosure. The Pillar 1 of the Basel II, 'minimum capital requirements', provides a detailed calculation of total capital requirements with full risk assessments for credit risk, operational risk and market risk. It offers a range of options such as the simple standardised approach and IRB approach for determining the amount of regulatory capital. This allows banks to select the most appropriate approach for their business operations and financial market infrastructure. It also highlights that key elements of the 1988 Basel Accord capital adequacy framework did not change, including the minimum requirement for banks to hold at least 8% total capital to their risk weighted assets, of which minimum 4% is tier 1 capital, and the definition of eligible capital that are retained in the revised framework. (BCBS, 1988) The Pillar 2 'supervisory review process' discusses the principles and processes of supervisory review; and the Pillar 3 'market discipline and public disclosure' explains the general market considerations and disclosure requirements.
Basel II has been implemented by national regulators include the Financial Services Authority (FSA) in the UK. Nowadays, almost all banks in the world are subject to the uniform capital requirements under the Basel II Accord.
2.4 Historical trends in UK bank capital ratios
Figure : Capital ratio of U.S and U.K banks (FSA, 2011)
In 1880-1980 before the Basel's capital standard is made, FSA (2011) reports that the evolution of U.S. and U.K banks' capital adequacy towards lower capital ratio. The Figure 3 illustrates that their capital ratio continuously declined until the 1920, rebounded during the 1920s and dropped sharply to 6% and 3% in the 1950 respectively. In 1980, they both hit the same level at 6%. It can be seen that the movement of UK banks' capital ratio is unstable over the hundred years.
Since the Basel's minimum capital requirement introduced in 1988, a number of papers state that UK banks' capital ratio tend to steady increase. The BIS (1999) reviews the trend of capital ratios of the G-10 countries between 1988 and 1996. It reposts that the average ratio of the G-10 countries has dramatically increased around 5% in nine years as represented in the Figure 4. It also shows that both UK and US banks have an upward direction but UK banks' capital ratio is slightly higher than US banks. In addition, Francis and Osborne (2012) investigate the movement of total capital ratio and tier 1 ratio in the UK over the years 1989-2007. They find both capital ratios move in the same direction. The Figure 5 shows that a significant upward adjustment during the first nine years. Total capital ratio rose from 10% to 13.5%, while tier 1 capital ratio increased from 4.5% to 7% approximately. In the eight years between 1999 and 2007, they both steady fell 2% and 1% respectively.
Figure : Capital ratios in UK and US (BIS, 1999)
Figure : Trends in total capital and tier 1 ratios for the UK Banks (Francis and Osborne, 2012)
Through the historical trends of the UK banks' capital ratio reviewed above, the UK banking industry appears better capitalised after the Basel standards is made than in the 1980s.
2.5 Potential factors affecting bank capital ratios
Changes in capital ratios reflect bank behaviour in capital allocation. A number of theories examine the factors affecting bank's total capital ratio and tier 1 capital ratio. They suggest that bank's capital decision can be explained by regulatory factors and bank internal factors including bank size, probability and risk exposure. The following sub-sections provide a detailed discussion of each of these variables.
2.5.1 Bank size
Bank size may influence the adjustment of its capital ratios due to the scale of operations is related to ownership equity investments, business diversification and investment opportunity set. The early study of Shriever and Dahl (1992), Aggarwal and Jacques (1998) and Gropp and Heider (2007) discuss the determinants of bank capital allocation. They all find that bank's asset size has a great and negative impact on capital ratios. This reflects that larger banks allocate lower levels of capital than smaller banks for non-performing loans. This may result because they have greater diversified asset-portfolio, scale economies in risk management, or lower levels of non-performing loans. Hence, reduce in risk and lower capital needs. In addition, correlated in an inverse direction could be interpreted as the 'too big to fail' effect (Roy, 2009). The failure of large banks is likely to trigger larger losses and failures to the whole banking industry, financial markets as well as the economy (Bessis, 2010). Because of afraid of such effect, government must support larger banks when they face difficulty, and thereby larger banks have less incentive than smaller banks to allocate higher capital in preventing risk.
2.5.2 Probability
Probability is about the return on capital provided by investors and creditors which reflects the efficiently of a bank in resources management. It is also considered as an important element to bank capital ratios. From the empirical findings of Jacques and Nigro (1997), Aggarwal and Jacques (1998), Gropp and Heider (2007) and Brewer et al. (2008), profitability is positively and significantly related to capital ratios. More profitable banks tend to have higher return and thus can retain more capital relative to assets. In other words, bank's income level increases as its capital adequacy ratio increases simultaneously.
2.5.3 Risk exposure
Dhaene et al. (2012) suggest that capital is close to the risk, thus allocated capital should be able to reflect the associated risk. A number of studies of bank behaviour on capital allocation observe the relationship between changes in capital and risk and conclude that there is a significant effect. Shriever and Dahl (1992) and Gropp and Heider (2007) show that bank's asset risk and capital ratios are positively correlated to one another. Greater risk is associated with higher capital ratios because it increases the probability of bank failure, the costs of bankruptcy and the rate of return on equity required by investors. The increase of likelihood of bankruptcy increases the value of bankruptcy costs and investors require higher return from company to compensate their risk of loss. Therefore banks would tend to set aside more capital when asset risk increases. Besides, bank's risk taking behaviour may affect by its capital levels. Banks with higher capital ratios position tend to more prone to risk-taking.
More recently, Junarsin (2011) provides evidence that banks with higher risk are inclined to have lower capital ratios. Shriever and Dahl (1992) note that a negative relationship between capital and risk may occur because bank maximises deposit insurance subsidy by reducing capital and increasing risk, while Jacques and Nigro (1997) argue that methodological flaws in the capital standards may explain for a negative association. Berger (1995) suggests that banks with higher capital and lower risk may be more easily to borrow uninsured funds to pursue greater profit on investment opportunities.
Moreover, Shriever and Dahl (1992) find that bank size and risk are positively related. Larger banks are more active to risk-taking than smaller banks, reflecting they have higher willingness to gamble in order to raise the potential return on investments. Conversely, smaller banks have lower investment opportunity sets and get harder access to capital markets so that they tend to risk averse.
2.5.4 Regulatory standard
As noted before, the minimum capital requirement in the Basel Accord is to ensure that banks allocate adequacy capital to sustain operating losses. The impact of the Basel capital standard on bank's capital ratios has reviewed by several literatures, for example, Jacques and Nigro (1997), Ediz et al. (1998) and Roy (2005). Jacques and Nigro (1997) assess the impact of the 1988 Basel Accord for FDIC (Federal Deposit Insurance Corporation) insured commercial banks during the first year the standards entered into force, from year end 1990 and year end 1991. They state that regulatory pressure brought significant increases in bank's capital ratios. Ediz et al. (1998) also find a positive relationship between capital requirements and capital ratio decisions in the study of British banks behaviour during the period 1989 and 1995. Roy (2005) analyse G-10 banks' reactions to regulatory constraints over the period 1988 and 1995. The finding shows that regulatory standard successfully increases the capital to banks in Canada, Japan, the UK and the US. In short, all these findings suggest that regulatory standard, the Basel I, is effective in raising bank's capital levels.
Furthermore, the study results of Jacques and Nigro (1997) show a negative relationship between regulatory pressure and risk, indicating the regulatory standard brought decreases in risk of banks.
In this chapter, the importance of capital allocation and functions of regulatory capital in banking organisations are defined. This calls for the study of bank capital ratios to understand current capital allocation behaviour of banking organisations. This chapter then is highlighted the development of the Basel Committee's capital requirements and the historical trends in UK bank capital ratios. The possible determinants of bank capital ratios are also discussed in providing an overview theoretical background for the research methodology and data analysis in the following chapters.
Chapter 3: Methodology
Based on the literatures in the previous chapter, this chapter provides a description of an appropriate research methodology for a capital allocation survey. In the first part of this chapter section 3.1 discusses the adopted research paradigm which refers to a way of examining social phenomena including research philosophy, approach, strategy, choice, and time horizon. Secondly, section 3.2 and section 3.3 address sampling method and data collection method for the chosen topic respectively. And then factors selection explains in section 3.4 and testable research hypothesis states in section 3.5. In addition, section 3.6 discusses the required procedures for data analysis comprising data preparation and appropriate statistics. Lastly, benefits and limitations of research methodology represent in section 3.6.
3.1 Research paradigm
In the section of the literature reviews it allows us to understand more clearly what information is needed to solve our problem and possible to discover the answer to the research question. However, it does not mean that it is able to provide an accurate solution for it (Baker and Foy, 2008). To effective identify a research issue, it is necessary to select an appropriate research methodology for guiding the investigation before beginning the research.
Research paradigm, i.e. positivism, interpretivism and pragmatism, is a framework based on a set of philosophies and assumptions to interpret social reality. It is important for the research as Guba and Lincoln (1994) state that it will strongly impact not only ways of conducting research but also the nature of ontology and epistemology. Saunders et al.'s (2009) 'research onion' divided research into six layers which provides simple and systematic model to process research from the outer to inner layer (see Figure 6).
Figure : The research 'onion' (Saunders et al., 2009)
3.1.1 Research philosophy
To begin with the first layer of the research onion is 'philosophies'. It helps to understand the research perspectives to view the world in terms of ontology, epistemology and axiology in order to determine the most suitable philosophy for the research.
According to Baker and Foy (2008), ontology is defined as the nature of existence of reality. It classifies into two aspects which are objectivism and subjectivism. The primary purpose of this research is to study capital management of banks in a particular country, United Kingdom. Banks recognised the importance of providing stability at a time of uncertainty which set financial strength as the central importance to organisations. They have a set of formal model to determine the amount of capital needed for maintaining stability. Aspects of capital allocation may differ but the essence of the function is similar in all banking organisations. All British banks comply with the minimum capital requirement of the Basel rules on capital ratios, aiming to sustain capital levels against risks. The effects of regulatory requirements and bank internal factors are expected to influence banks' capital ratios. This emphasises that the subjectivist aspect of ontology cannot help to understand objective aspects of changes in their capital ratios. Therefore in this research, the capital management is an objective entity that exists in reality external to social actors.
Epistemology is concerned with the valid knowledge in a particular field of study (Saunders et al., 2009). This research considers data on resources needed is the most important in the study of the capital allocation. It aims to collect and analyse the facts. Large samples and statistical measurement are required to support the research. Following this, the findings of the research produce objective reality and far less bias.
In the philosophical position, axiology is great importance to the results and judgements. Saunders et al. (2009) describe axiology is about the role that our values play in research process. The choice of research approach reflects our values and the credibility of the research. This research is undertaken in a value-free way. We are independent of the data and maintain an objective stance.
To sum up briefly, this research reflects an objective and observable social reality. Large quantitative data tend to collect and analysis statistically. Moreover, researcher is independent from that being researched. It concludes that positivism is the best research philosophy to this research.
3.1.2 Research approach
From the choice of research philosophy, it becomes clear that the research is based on quantitative method using deductive approach rather than inductive approach. According to the research questions and objectives, this research aims to explain potential factors that could affect capital ratios set by banks. It assumes that the variation of capital ratios depends upon regulatory pressure and banks' internal factors related to their asset size, profitability and risk. This hypothesis is deduced from the literatures.
Research approach is matter to the findings and evaluation of the research. Because it influences type of data we collect, the choice of data collection method and data analysis techniques. Deductive approach refers to scientific research that involves theory testing (Saunders et al., 2009). It contains the basis of explanation and anticipation of phenomena to predict occurrence and thereby a test to be controlled (Collis and Hussey, 2003). This research is to collect quantitative data using large samples for testing hypothesis. After that, it needs to explain the relationships between variables. It will either tend to confirm or modify the theory (Robson, 2002).
3.1.3 Research strategy, choice and time horizon
Research strategy provides logic for answering particular research questions and meeting objectives. Blaikie (2000) defines this term as the logic of enquiry. Under the 'strategies' layer of the research onion, we have chosen survey as our research strategy. This strategy allows us to collect large enough quantitative data for the investigation of relationships between variables. Therefore we can produce numerical analysis using descriptive and inferential statistics, and a model for testing correlation (Saunders et al., 2009). From the research strategy, it can be seen that this research adopts a single data collection technique and corresponding analysis procedures based on quantitative. This represents that our 'research choice' is mono method. Moreover, regarding research time horizon, this research is a cross-sectional study which provides a series of snapshot of UK banks' capital management over a given period, from 2008 to 2011.
On the whole, the research paradigm of this research follows a quantitative method for the core of the research onion, 'data collection and data analysis' (see Figure 7). This research adopts positivism as the research philosophy with deductive approach. Survey has been chosen as the research strategy along with mono method in collecting, testing and analysing data for a cross-sectional study.
Figure : The research onion of the project
3.2 Sampling method
For this research it is impossible and impracticable to collect and analyse all data from the entire population (Saunders et al, 2009) because of the constraints of time and access. Therefore it needs to select a sample. In the case of capital allocation survey, probability sampling is the most suitable sampling method for the quantitative research. It would be able to minimise bias and statistically measure the characteristics of UK banks from the sample.
In deciding a sample, sample size is an important factor to the research as it directly impacts the accuracy of findings as well as a matter of judgement. According to Saunders et al (2009), the larger size of sample the lower the possibility of error of the research. However in consideration of the amount of time in collecting, processing and analysing data, sufficient sample size would be suggested. Stutely (2003) provides a rule of thumb that the minimum number of sample size is 30. Any sample is less than 30 that results will not able to represent the whole population. This means that precise estimations and judgement will not be made. This research focuses on UK banking industry and randomly selected 76 British banks for the sample using simple random sampling technique. Each UK bank in the population, whether is large or small bank and any type of bank, had equal chance of being included in the sample. This implies that the selection is without bias and thus the sample can be representative of the population.
3.3 Data collection method: Bankscope
A large amount of quantitative data is required for testing the factors affecting UK bank capital allocation behaviour. This research focuses on four potential factors and they are bank size, probability, risk and regulatory pressure. Therefore the targeted data for the study are limited to banks' statistical records of total capital ratio, tier 1 capital ratio, total assets, return on assets (ROA) ratio and risk-weighted assets during the last four years.
Secondary data provide useful source such as raw data and published reports that have been collected for other purpose (Saunders et al, 2009). Stewart and Kamins (1993) believe that researcher is at an advantage when using secondary data. Because the data already exist which save time and money in collecting suitable data. There are a number of information gateways locating secondary data. For this research, data are obtained from Bankscope database of the Bureau van Dijk. Bankscope is one of online computer databases via the internet. It combines wide range of data sets on worldwide banks which contains up to 16 years of comprehensive information on 30,000 banks (Bureau van Dijk, 2012). This is called multiple-source secondary data. Bankscope also provides a convenience and flexible software for searching data. It allows us to select numerical data based on target variables about particular country in particular time period.
Data are gathered from 76 UK banking organisations for the years 2008 through 2011 with a total of 304 observations. Such data sources have been evaluated the suitability to answer the research question. The reasons of assessments are that secondary data may not specifically achieve the research objective or out-of-date. Thereby quality check of data is necessary in order to ensure that data are reliable and valid to this research. Saunders et al. (2009) suggest three steps to determine the suitability of secondary data set in the following order:
Assess overall suitability of data to the research question;
Evaluate precise suitability of data for analysis and meet objective; and
Judge data sources based on costs and benefits.
The following table, Table 2, shows that secondary data have been tested by a set of questions following Saunders et al.'s (2009) three-stage process. In summary, it represents that data are appropriate for this research to answer question and meet objective.
Table : Evaluation of secondary data sources (Sources from Blumberg et al. (2008); Dale et al. (1998); Dochartaigh (2002); Jacob (1994); Kervin (1999); Stewart and Kamins (1993).)
3.4 Factors selection
3.4.1 Dependent variables
The aim of this study is to investigate the factors affecting UK banks' capita allocation. In order to capture their capital decision, we define capital levels as a dependent variable through two different measures of capital ratios: total capital ratio and tier 1 capital ratio. Total capital ratio is the ratio of bank's total capital to weighted risk assets; and tier 1 capital ratio is the ratio of bank's core capital to weighted risk assets. These risk-based capital ratios have been used by previous literatures such as Jacques and Nigro (1997), Aggarwal and Jacques (1998), Rime (2001) and Brewer et al. (2008). The leverage ratio (total capital or tier 1 capital to total assets) is another way to measure capital which is also employed by the literatures. However, it does not apply in this research. This research uses the observed changes in capital ratios (ΔCAP) which are modelled using partial adjustment framework. The framework is developed by Shrieves and Dahl (1992). It consists of two components, a discretionary adjustment and an exogenously determined random shock. As a result:
where is the total change in capital ratios for bank j in period t, is the endogenously determined adjustment, and is the exogenous shock.
where is proportional to the difference between a bank's target capital levels in period t and the exiting capital levels in period t - 1.
In this study, bank changes in capital ratios (ΔCAP) may influence by a number of independent variables including the size of bank (SIZE), profitability (ROA), changes in risk (ΔRISK) and regulatory pressure (REG). All these variables were used by Jacques and Nigro (1997), Aggarwal and Jacques (1998) and Rime (2001).
3.4.2 Independent variables
Following previous research, SIZE is bank's asset size. It is measured as the natural log of total assets to capture size effects.
A bank's profitability, ROA, represents by the ratio of return on asset which is calculated as bank's net income divided by total assets. It uses to identify the ability of profitable banks to increase their capital ratios (Aggarwal and Jacques, 1998).
Risk is the ratio of a bank's risk weighted assets to its total assets, presenting assets allocation across the quality of loan and risk types (Shriever and Dahl, 1992). This research uses the observed changes in risk (ΔRISK). The concept is the same as ΔCAP which also adopts partial adjustment framework for modelling.
where is the total change in risk ratios for bank j in period t, is the endogenously determined adjustment, and is the exogenous shock.
where is proportional to the difference between a bank's target risk levels in period t and the exiting risk levels in period t - 1.
The main interest of this study is on regulatory pressure. Regulatory pressure, REG, can be evaluated in many different ways. In this research, we explain the degree of regulatory pressure in two ways, denoted by REG1 and REG2. REG1 is a dummy variable that equals to one if a bank is adequately capitalised which with total capital ratio (tier 1 capital ratio) at or above the 8% (4%) regulatory minimum requirement, zero otherwise. This measure is similar to the regulatory pressure variable used by Shrieves and Dahl (1992). REG2 is introduced by Ediz et al. (1998). It is measured by a dummy variable that takes the value of one if a bank's capital ratio has experience an upward adjustment, zero otherwise.
Table : Overview of the calculation of variables
3.5 Research hypothesis
Hypothesis is the statement about the prediction of the outcome of experiment. Since the four potential factors were used and suggest by a number of literatures, this leads to two hypotheses that are:
Hypothesis 1: Bank internal factors (i.e. size, probability and risk) and regulatory factors will significant influence UK bank's total capital ratio and tier 1 capital ratio.
Hypothesis 2: Regulatory pressure will increase bank's capital levels effectively.
The first hypothesis assumes that all independent variables will have relationships with both bank's capital ratios in the periods of 2008 and 2011. And the second hypothesis refers to the research question: 'Does the Basel II Accord is efficient in increasing banks' capital levels?'.
3.6 Data analysis
Quantitative data in a raw form are no meaningful unless these data have been processed and analysed to become information. Before statistical analysis, this section identifies the main issues of data preparation and the most appropriate statistics for answering the research question.
3.6.1 Data Preparation: Data editing, coding and entering
The data are already in an Excel file with spreadsheet format. There are three important steps to prepare data for analysis in the following order: data editing, data coding and data entering.
Step 1: Data editing
Data editing includes process of checking and adjusting data for legibility and consistency. Data also have been checked for omissions because there may be some errors or missing data in the collection. And then we calculate variables including ΔCAP, SIZE and ΔRISK from numerical data.
Step 2: Data coding
Some data need to be identified and assigned a numerical score, for instance, for regulatory pressure variables and missing data. As mentioned in the section 3.4.5, regulatory pressure variables are dummy variables with the value either 0 or 1. In our data set, some data are missing and we use '999' as a specific code instead of blank grid for the whole sample.
Step 3: Data entering
The final step of data preparation is to transfer quantitative data from Excel file to SPSS software. In this stage, error checking is essential to ensure that data are entered and labelled correctly. Once the data have been checked, SPSS is then used for statistical analysis including descriptive statistics, significance test, correlation and linear regression model which discuss in subsequent sections.
3.6.2 Descriptive statistics
Descriptive statistics is statistical summary of the data in the sample (Saunders et al., 2009). We use mean to calculate the average levels of capital ratios in each year to describe the central tendency of UK banks' capital allocation between 2008 and 2011. We also measure the mean of changes in capital ratios per year to show the variation of capital level in the four years. Furthermore, the mean of banks' profitability and of risk can represent the UK banking financial health during the financial crisis period.
3.6.3 Significance test and correlation
The purpose of this study is to investigate the factors affecting bank's capital allocation behaviour particularly the regulatory pressure variables. To do this, a two tailed significance test will carry to analyse whether there is a significant difference between two variables at 95% or 99% significance level. The p-values need to be less than either 0.05 or 0.01 to be statistically significant. The research hypothesis has been discussed in earlier section 3.5, which we assume that each independent variable significantly impact on the dependent variable, ΔCAP. The decision will make between the null hypothesis and alternative hypothesis: means that there is no relationship between two variables; and means that they are correlated.
All pairs of variables are put into SPSS to calculate Pearson's product moment correlation coefficients (r). Correlation indicates the direction of a linear relationship between two variables (Saunders et al., 2009). A correlation coefficient takes on any value between -1 and +1 (see Figure 8). A value with positive sign represents a positive correlation, meaning that two variables move in the same direction, while a value with negative sign is negative correlation that they move in an inverse way.
Figure : Value of the correlation coefficient (Saunders et al., 2009)
3.6.4 Multivariate linear regression
Regression analysis is a technique to examine the relation of dependent variable to specified independent variables (Saunders et al., 2009). It can be used to closely observe a cause-and-effect relationship where a change in independent variables causes a change in dependent variable. We developed a multivariate linear regression model to calculate a coefficient of determination () and predict the values of a dependent variable given the values of independent variables. As a result:
3.7 Benefits and limitations of methodology
This research adopted positivism as the research philosophy with deductive research approach to examine social phenomena and the facts. This leads to collect large quantitative data and analyse statistically, and that researcher is independent from that being researched. Researcher is external to the process of data collection and cannot change the facts that increase the level of objectivity and accuracy of our data. Therefore the reliability of our research would be high. However, low in validity. The research paradigm leads to the hypothesis that the facts are true and the same for the whole population. Because of the complexity of human behaviour, banks' capital decision do not all act in the same way and bring the same effect. Hence it could be misleading the results.
Probability sampling has been chosen along with simple random sampling technique for this quantitative research. Randomly select the sample from the sampling frame, meaning that each UK bank has an equal chance of being selected. Thus, this minimise bias and maximise validity to the results.
As noted before, the larger sample size the greater the reliability and accuracy of the results. Due to the constraints of time and access, only 76 British banks have been selected with a total of 380 observations for the four years study. Despite the number of sample size exceeds Stutely's (2003) advice of a minimum number of 30, but this sample size might not sufficient enough to lower the margin of error. Thereby there is a potential bias on the representativeness of the sample. Bankscope is an internet information gateway locating secondary data, benefiting the researcher to save time and money in data collection. Nevertheless, there is the potential bias since secondary data are not primary function for our research question. And regarding the coverage of Bankscope, some numerical data are missing. Therefore such limitations could be reducing the level of accuracy and reliability of the research results.
There is the potential bias in factors selection. The representativeness of the factors in bank capital behaviour is very important to the research question. Although the five independent factors were selected according to several literatures, but there still could be other factors influence bank's capital ratios. This implies that the five selected factors may not sufficient enough to explain the variation of capital ratios.
Data preparation is a necessary step before statistical analysis, benefiting us to minimise the risk of error, inconsistency and illegibility in order to provide accurate analysis and results. Once the data have been checked, quantitative analysis will process using SPSS software. SPSS is statistical analysis software package that provides an easy and fast way for researcher to analyse large numerical data for a complex research project. It offers a broad range of statistical tests and models to completely answer research question and meet objective such as descriptive statistics, significance test, correlation and linear regression model. Correlation test is a major statistical technique to our research. It allows us to investigate the factors affecting UK banks' capital ratios and answer our research question 'Does the regulatory standards is efficient in increasing banks' capital levels?'. Consequently, this data analysis method would be able to produce the desired and precise research outcomes.
Chapter 4: Empirical results and analysis
This chapter presents the findings of UK banks capital allocation survey through SPSS statistical analysis: descriptive statistics, significance test, correlation and linear regression model. Firstly, in section 4.1, we summarise the collected data. Secondly, section 4.2 addresses the trends of banks capital ratios, while the changes in capital ratios highlights in section 4.3. The financial health of banks and key factors to capital decision discuss in section 4.4 and 4.5 respectively. Finally, we conclude the overall results and consider the reliability of findings in section 4.6.
4.1 Data summary
The purpose of the research is to investigate the capital levels of the United Kingdom banking industry in the period of 2008 and 2011. 76 British banks data with a total of 380 observations are taken from the Bankscope database and cover years 2007 to 2011. The data involve a numerical data set of banks' total capital ratio, tier 1 capital ratio, total asset, the ratio of ROA and risk weighted assets during the five years. All these data have been checked for omissions and coded with keywords as discussed in section 3.6.1. And then data were calculated and used for defining variables: changes in capital ratios (ΔCAP), asset size (SIZE), profitability (ROA), changes in risk (ΔRISK) and regulatory pressure (REG) following the detailed description in section 3.4. Finally, period observed is from 2008 to 2011 so that 76 UK banks with 304 observations were input into the SPSS software for statistical data analysis. In this section, the following tables and figures show the breakdown of the collected data according to year, specialisation and total assets.
Table : The collected data according to year
In Table 4, a total of 380 observations from data collection split into five years data and each year consist of 76 banks' annual financial information. Data of 2007 are used to capture the changes effects in year 2008, and so on. Thus these data have been put into the measurement to create the 4 years data with 304 observations for the data analysis.
Table : Specialisation of UK banks
Figure : Banks grouped by specialisation in percentage
Table 5 and Figure 9 present the specialisation of UK banks. Half of banks are commercial banks that mainly provide savings and investment services for personal finance and commercial banking. In 76 banks, only one bank is Islamic bank who offers banking products and services under the principles of Islamic, Sharia law; and one securities firm operates lines of business in securities brokerage, financial advisory services as well as securities trading.
Table : Banks grouped by total assets
Figure : Number of banks grouped by total assets in each year
Banks have been classified into six categories by their total assets. As shown in Table 6 and Figure 10, most of UK banks had total assets between 1 billion and 10 billion pounds. The number of banks in this category increased during 2008 and 2010 but decreased in 2011. However, the number of banks with total assets less than 1 billion pounds, and between 10 billion and 100 billion pounds decreased in the whole observation period. Meanwhile, banks with total assets between 100 billion and 1 trillion pounds, and over 1 trillion pounds did not change much in numbers during the four years.
4.2 Capital allocation of UK banks
According to the Basel II Accord's minimum capital requirements, banks are expected to allocate at least 8% total capital ratio and at least 4% tier 1 capital ratio. During 2008 and 2011, all UK banks had total capital ratio in excess of the minimum capital requirement, meaning that they are adequately capitalised. Only one bank's tier 1 capital ratio (2 out of 304 observations) did not meet criteria of the Basel II in 2008 and 2009. This implies that that bank is undercapitalised in terms of its core capital elements: permanent shareholders' equity and disclosed reserves. The minimum, maximum, mean and standard derivation of capital ratios in each year are shown in Table 7.
Figure 7 presents that the minimum and maximum levels of total capital ratio in 2008 are 9.1% and 44.4% respectively which give the difference of 35.3%. The biggest difference in total capital ratio is 158.2% in 2011, and the biggest difference in tier 1 capital ratio is also at 2011, given 109%.
Table : Descriptive statistics of capital ratios in 2008-2011
Figure : Minimum & maximum banks' capital ratios
From Figure 10, the trends of minimum level are increasing for both capital ratios between 2008 and 2011. The minimum total capital ratio rose from 9.1% to 13.4%, while the minimum tier 1 capital ratio was at -0.4% far below the capital standard and reached to 8.7% by the end of 2011. This indicates that there is a significant upward adjustment during the four years. The trend of maximum capital ratios are also increasing, except the total capital ratio in 2009. The maximum tier 1 capital ratio turned out to be 50.3% to 117.7%, while the maximum total capital ratio was 44.4% in 2008, dropped to 40.2% and then grew up for the following two years.
Figure : Means of capital ratios
Figure 11 illustrates the mean of bank capital ratios in each year. It represents that both capital ratios move in the same direction with increasing adjustment. On average, bank total capital ratio is 16.2% in 2008. It has straight line moving during the first 3 years and jumped to 22% by the end of 2011 as presented in Figure 4. Similarly, the mean of tier 1 capital ratio was at 13.5%, moved in a straight line between 2008 and 2010 and slightly increased to 18% by the end of 2011. Resulting from this, on average, UK banks had significant upward adjustment on both capital ratios in the years of 2008 and 2011, and they are adequately capitalised that hold sufficient capital far above the Basel II's minimum capital requirements.
4.3 Variation of capital ratios
Since actual capital ratios have been described, this section addresses the variation of capital ratios. The variation of capital ratios is the proportional to the difference between a bank's target capital levels in period t and the exiting capital levels in period t - 1 based on the Shrieves and Dahl's (1992) partial adjustment framework. The maximum negative, maximum positive, mean and standard derivation of changes in capital ratios during the four years are shown in Table 8.
Table : Descriptive statistics of changes in capital ratios
Figure : Maximum negatively & maximum positively changes in capital ratios
According to Figure 13, the maximum negatively changes in total capital ratio and tier 1 capital ratio have the same pattern in the movement and nearly same value during the four years. In 2008, the maximum negatively changes in total capital ratio and tier 1 capital ratio were the same, -19%. They both decreased negative value in 2009 and 2010 and then increased to -15.1% and -14.9% respectively by the end of 2011. Similarly, the maximum positively changes in total capital ratio and tier 1 capital ratio also moved in the same trend but unstable. The maximum positively changes in total capital ratio was 12.8% and changes in tier 1 capital ratio was 13% in 2008. They went up to 27% and 25.1% respectively, declined and then rebounded at year end 2011.
In addition, the variation of total capital ratio in 2008 has a small range of fluctuation, ranging only from -19% to 12.8%. Given that, its mean value is -0.8%. In comparison with the four years, on average, Δ total capital ratio in 2011 has the highest variation with 3% as presented in Figure 14. In Figure 14, it also shows that the trend of the mean of Δ tier 1 capital ratio is increasing and thereby Δ tier 1 capital ratio in 2011 has the largest variation, on average, 2.2%. It follows that, on average, UK banks have negative variation on both capital ratios in 2008, increasing to the biggest positive variation is in 2011.
Figure : Means of changes in capital ratios
4.4 Financial health of banks
Average total assets, profitability and risk level of UK banks are shown in the below table, Table 9, to explain their financial health during the global financial crisis period. Total assets are the sum of fixed and current assets owned by a company, ranging from equipment and investment securitie
