The effect of regulatory intervention in the
American banking sector corresponding
profitability.
University of Amsterdam,
Faculty of Economics and Business Bachelor Thesis
Student: Vincent de Wit Student number: 10750061
Abstract
This thesis analyses whether regulatory intervention in the banking sector affects the profitability in practice. For this research an econometrical model is designed with ROAA as dependent variable and the tier one capital ratio as tested independent variable. In the model six more independent variables are added as control variables. For the testing of this model a fixed effect panel analysis has been executed over 109 American banks. The result is a negative relation between the ratio and the ROAA.
Statement of Originality
This document is written by Vincent de Wit, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of content
1 INTRODUCTION... 4
2 LITERATURE REVIEW ... 6
2.1THE FINANCIAL CRISIS OF 2008 ... 6
2.2BASEL COMMITTEE OF BANKING SUPERVISION ... 7
2.3COMPARISON BETWEEN THE CANADIAN AND THE US BANKING MARKET, WITH RESPECT TO REGULATIONS ... 7
2.4PREVIOUS RESEARCH WITH RESPECT TO BANKING PROFITABILITY ... 8
3 DATA & METHODOLOGY ... 10
3.1INTRODUCTION... 10 3.2REGRESSION MODEL ... 10 3.3VARIABLE SELECTION ... 11 3.4DATA ... 12 3.5HYPOTHESIS ... 13 3.6METHODOLOGY ... 13 4 RESULTS ... 14
4.1ELABORATION OF THE RESULTS ... 16
4.2TESTING THE HYPOTHESIS ... 17
5.0 CONCLUSION ... 18
1 Introduction
After the financial crisis in 2008, the banking sector became a controversy of itself due to all the flaws that were exposed after the first banks filed for bankruptcy. As a result of the crisis, a new regulatory intervention was a reaction from the Basel Committee to restore the
financial stability and trust in the sector (Bank for International Settlements, 2017). These new regulations make banks more solvent and liquid, but at the cost of banks’ efficiency and thus their profitability (Tung-Hao & Shu-Hwa, 2013). Hence, banks request the Committee to mitigate the proposed regulations. The reformed regulations have three main pillars
consisting of increasement of the capital ratio, increasement of liquidity ratio and increasement of leverage ratio. All of these affecting a banks’ efficiency when the new regulations are faced. However, is banking profitability altered in practice? According to De Grauwe (2011), banks can develop innovative products which can undermine new rules imposed. Which was feasible with the former Basel agreements. He also explains that the capital requirements cannot be calculated because the market is not efficient (assumed by the Basel Committee), due to the fact that tail risks - like crashes and bubbles – cannot be
quantified. In previous research many variables have been considered in what extent they would affect the average return in banking. In Dietrich and Wanzenried (2009) their study, the outcome shows that having more equity rises the profit, simultaneously the cost-income ratio presents (negative) significance. In addition, the interest income and the size of the bank (in assets) are significant variables. However, during this research the angle of approach is rather different. With the same dependent variable, the goal of this study is not to find which variables have influence on the profit. Yet, it is about tracking down in what direction the specific capital ratio will influence the ROAA. To examine this question an econometric model is created where return on average assets will be used to reflect a bank’s profitability. The tier one capital ratio is the variable tested, to see whether it influences the profit
negatively. This variable will be subjected to a time distortion, implicating it will be
regressed at time: t+1, because the ratio of last year will affect the contemporary profitability. Six control variables will be part of the model, consisting of four bank-specific and two macroeconomic factors. The bank-specific variables include the leverage ratio, logarithm of assets, the net interest income share and provisions for loan losses. The macroeconomic part consists of tax rate and the US GDP growth. The method used will be a fixed effect panel data analysis. According to theory the ROAA will be negatively influenced by the tier one
capital ratio. Nevertheless, in this study it is assumed that in practice this outcome is not a
certainty.
The results are like the theory predicts: tier one capital ratio shows negative significance with respect to the ROAA, rejecting the null hypothesis (H0: β1=0). Although having a low
correlation with the dependent variable, it can be concluded that when the ratio increases, a bank becomes less beneficial. Beside the β1, the efficient tax rate and the real GDP growth showed likewise significance. Main implication in this research was the low R-squared (0.1311) which can be solved by adding more data. Many factors, influencing the net income of banks are not straightforward and are therefore hard to point out, in a sense that complex factors are hard to translate into variables (if data is available already). In future research, more variables should be added to come to improved estimations.
In the next chapter, the literature review will explain some subjects regarding this study. Starting with the financial crisis, which will be elaborated on the basis of the main issues triggering this economic turmoil. The Basel Committee of Banking Supervision will be addressed to explain what the new regulations are. Followed by a comparison between the US and Canada with respect to their regulatory policies. Concluding with literature
concerning profitability in banking. Some background information is given, the methodology & data section will go deeper into the subject by elaborating the model, the hypothesis, variable selection, data processing and previous studies regarding this topic. Afterwards, the results will be discussed and compared to previous empirical outcomes. In the final chapter, the conclusion and discussion will be addressed.
2 Literature Review
2.1 The financial crisis of 2008
Goldberg (2009) states that international risk-sharing and diversification are the positive effects globalisation generates. Although it has a negative part as well according to De Grauwe, which is the exposure to bubbles and crashes banks have to deal with (2011). He states that risks that are encountered by universal banks cannot be quantified in such manner like imposed by the Basel Committee due to tail risk like bubbles and crashes. The financial crisis of 2008 began with the housing bubble in the US generated by abundancy of liquidity and low interest rates which caused that new products were designed with higher profit and more risk (like: CDO’s, ABS’s, SPV’s etc.). As a result of the rising of new complex products, risk became mispriced and leverage increased (de Larosière & al., 2009).
The second segment which triggered the crisis is risk management. The main failures were the disability to fully verify the leverage and the misinterpretation between credit and liquidity (de Larosière & al., 2009). These complex products did not reflect the contained risk and were often evaluated through mild tests by credit rating agencies. This resulted in
underestimated risk exposure and due to the lack of transparency in the financial sector, the location of the risk was unknown (Barrell & Davis, 2008). De Grauwe argues that banks were pushed to take risk off-balance by the new rules implied by Basel I, although Basel II took away this incentive partially (2011). When Lehman Brothers signed for liquidation, these ‘toxic’ assets came to light which triggered a worldwide shock, questioning the liquidity of severe financial markets (De Grauwe, 2011). According to Barrel & Davis this shadow banking is another reason why this market is not efficient and cannot be controlled by capital requirements only, since banks can undermine these rules imposed by untraceable routes (Barrell & Davis, 2008). Due to these innovative products, boards of the financial companies neither understood what kind of commodities were sold by their company and did not know to what extent the risk exposure was risen. Likewise, was the incentive mechanism changed from long-term profits, to short-term volume expansion involving higher risk (de Larosière & al., 2009).
2.2 Basel Committee of Banking Supervision
The Basel Committee of Banking Supervision (BCBS) sets standards for banks that are internationally active with respect to minimal capital requirements. The Committee made a new proposition (Basel III) which should result in a more stable banking environment, more confidence in prudential ratios and transparent regulatory environment (Bank for
International Settlements, 2017). In this imposition it is stated that the requirement of Tier 1 capital ratio needs to increase to 6% (which is an addition of 2 percent point to Basel II). Another imposition the banks have to follow is about the leverage. In this ratio, the tier 1 capital will be derived by on- and off-balance sheet exposures and must be minimal 3%. The last pillar is the Liquidity Coverage Ratio, which measures the bank’s ability to sustain a 30-day stressed scenario with high-quality liquid assets and needs to be >100%. G-SIB’s (global systematically important banks) have higher norms with respect to aforementioned ratios to fulfil (Bank for International Settlements, 2017).
2.3 Comparison between the Canadian and the US banking market, with respect to regulations
Fonseca et al found that banks with a higher capital buffer, charge lower interest rates on their lending and pay less funding costs on their borrowing (Fonseca, Gonzalez, & Pereira da Silva, 2010). On the other side, in working paper 494 of the Bank of England it is discussed that during a macroeconomic boom, a 1 percent point higher capital buffer, leads to 4.5% less lending due to pro-cyclicality capital requirements (Noss & Toffano, 2014). Nevertheless, this lost revenue due to the higher capital requirements decreases the chance of default, which add value to the firm and gives consumers more confidence in the financial sector. For
example: in the Canadian banking sector, no governmental support was necessary for any of the six biggest banks (the only country in the G7) (Bordo, Redish, & Rockoff, 2015). One of the main reasons for this event is the stricter rules imposed by the Canadian regulator: Office of the Superintendent of Financial Institutions (OSFI). Higher capital requirements and a more conservative financial market with less innovative (complex) products made their banking sector more resilient during the financial turmoil. A different funding structure than most other OECD banks and an oligopolistic market were key determinants (Huang &
Ratnovski, 2009). Besides, the Canadian mortgage market is insured in greater extent than the American, by the National Housing Act (Kiff, 2009).
The reason of the Canadian market being more resilient to systemic risk is due to the acknowledgement that the mortgage and investment market are the core of this risk (Bordo, Redish, & Rockoff, 2015). This conservative mortgage market is a result of the consolidated financial market that Canada created through the last century. Bordo et al. (2015) discuss that the robustness of the financial market is due to the concentration and diversity of these banks.
Three main pillars creating the withstanding financial market. First off, The Canadian regulator Office of the Superintendent of Financial Institutions (OSFI) set capital
requirements at 7% instead of the mandatory 4% imposed by BCBS. This capital can only exist for 15% out of innovative instruments, compared to no obligation to a certain ratio from the Committee. The last part is a leverage ratio of 5%, where Basel II did not appoint any percentage (Huang & Ratnovski, 2009). According to Seccareccia (2014), due to the
oligopolistic character of the market with high equity and low leverage, the barrier to entry is high, therefore less competitive pressure prevails. Consequently, the Canadian banks yielding revenues from highly lucrative domestic activities.
2.4 Previous research with respect to banking profitability
The return on average assets is a common dependent variable to measure the profitability and the independent variables in general are macroeconomic, bank-specific and industry-specific. Demirguc-Kunt & Huizinga found positive signifiance in GDP growth and inflation tested againt ROAA, while the effective tax rate reacts the opposite (1999). However, this tax rate has a small burden on the profitability because banks shift most of these expenses through to their clients (Albertazzi & Gambacorta, 2009). While some factors are straight forward and have a predictable outcome corresponding bank profit, some studies do not conclude inherently. For instance, according to the market-power paradigm, enlarged market power accomplishes monopolistic benefits, which is proved by the study of Bourke (1989) and Molyneux and Thornton (1992), yet the theory’s being refute in two different studies by Demirguc-Kunt & Huizinga (1999) and Staikouras & Wood (2004). The validity of this theory is questioned by last mentioned since the latter two studies find nagative though insignificant effects concerning net income. In prior research profitability determinants of commercial banks across countries with different income levels diverge extensively with respect to significance, sign and size of the effect (Dietrich & Wanzenried, 2014). From a theoretical point of view the relation between capital and return is negative. When capital rises, the market demands less return due to descending risk assumed investors are risk averse
and cannot diversify all the banks’ risk away. Regardless, some other theories come to the opposite relation, with assumptions added. Bourke (1989) assumes that better capitalized banks access to cheaper sources of funds, accompanied with higher quality asset markets and proves empirically that a positive correlation exists, with the signalling hypothesis. Another hypothesis regarding the positive relation is the expected bankruptcy cost hypothesis which links the capital ratio to the chance of bankruptcy and the costs accompanied. Scoped into the US corresponding this topic, research confirms a negative relation between capital and profit in the period between 1995-2007 (Hoffman, 2011). In this study the efficiency-risk and the franchise-value hypothesis are the most important factors explaining the findings. The first hypothesis assumes more efficient banks will have lower capital ratios, in the results this hypothesis prevails when the ratio is below the 41percent. If the capital ratio has a higher value than prementioned percentage the second hypothesis will dominate including that banks hold a high ratio when being most efficient. It can be concluded that many different theories and hypothesis have empirically been proved, with all different assumptions and substantiations.
3 Data & Methodology
3.1 Introduction
In this part the model will be explained, followed by the variable selection and data-gathering process ending with the hypothesis. Concluding with the statistical method detailed, used in this research.
3.2 Regression model
ROAA i,t = αi + β1T1CR_1i,t + β2LR_1i,t + β3LNa i,t + β4POLLi,t + β5 NII i,t + β6TR i,t +
β7GDPi,t + ui + ε i,t
- ROAA= return on average assets - i= the non-random bank that is used
- t= time indicator, where t is 1 for the year 2006 - α= constant
- u= individual error-term - ε= idiosyncratic error-term
-T1CR_1= tier 1 capital ratio = Common Equity Tier 1Risk Weighted Assets. This ratio is calculated one year prior the ROAA
-LR_1= leverage ratio = Total CapitalDebt . This ratio is calculated one year prior the ROAA -LNa=LN(assets)
-PLL= provision for loan losses -NII= net interest income -TR= effective tax rate -RGDP= real US GDP growth
3.3 Variable selection
In the regression model of the 109 biggest banks of the US by asset size will be tested with
ROAA (return on average assets) as the dependent variable. In former research this variable
is also used as dependent variable because it shows how efficient a bank uses its assets to generate profit. Unfortunately, the Liquidity Coverage Ratio cannot be part of the model since databanks do not supply this data. Although some banks do show this ratio in the newest balance sheet (2017), banks are not obligated to disclose this ratio, hence this content cannot be derived from loose balance sheets from all banks needed either. The tier 1 capital
ratio is supplied by databanks and can be tested as desired. However, also with the leverage ratio a problem has risen: the leverage ratio the Basel Committee proposed is not available in
databanks. This ratio is available in balance sheets of G-SIBs (global systematically important banks) because these banks are obligated to disclose this information annually, hence most of the banks in this dataset do not have to attach this intel to their annual reports. Accordingly, a different approach is needed with another leverage ratio compared to the one which was intended to test.
The tier 1 capital ratio should give a negative outcome due to less available capital which could be used to generate revenue. This ratio is calculated one year prior to the profit measure because the capital requirement of last year influences next year’s profit. The standard leverage ratio, as shown in the model specifications, calculates a rather different aspect, it shows the relation between the debt to total capital, compared to a non-risk adjusted tool that divides the capital measure by an exposure measure (off-balance items included). The ratio used in the model is less sophisticated and only measures a banks debt with respect to its total capital. Hence, the estimated value of this outcome cannot be compared to the actual ratio we want to estimate and acts more as another control variable. The leverage ratio estimate used, should have a positive sign as a result of a higher amount of debt a bank owns, the more loans and other financial products a bank can sell.
The control variables used, are selected because of theoretic support and formerly executed research pointed out that there is significance with most of the variables with respect to profit, to begin with the firm specific variables (Dietrich & Wanzenried, 2009). First off, the size of a bank (LNa), where the optimal size has been argued in literature, with one outcome that a bigger bank makes higher relative profits (Pasiouras, 2007). Larger firms have more diverse portfolios and loans which reduce risk, beside increasing returns to scale. However, bigger banks are exposed to higher agency and bureaucratic costs (for example)
due to more complexity (Dietrich & Wanzenried, 2009). Reasoning, it is not sure what influence the size will have on the profit. The logarithm for the asset size is chosen to reduce variance. Provision over loan losses: this ratio is a bank’s measure for the credit quality and divides loan losses over total loans. Expected is a negative correlation with the profitability.
Interest income share is the net income a bank earns from its interest minus the interest cost.
This value varies considerable from traditional banks, which earn profit through interest, to banks that focus on asset management. The latter creates higher profit margins on average and thus will have a relatively higher profit. Expected is that banks with more weight on interest income will earn lower profits.
Secondly, some macroeconomic values will be tested starting with the effective tax
rate. This reflects the effective tax rate the firm is subjected to, expressed in a percentage,
through complex structures firms can reduce the effective tax rate to such degree differences arise between banks. This estimation will have a negative impact on the dependent variable. In addition, the tax rate is not as pure as it should be for the regression, the tax rate is namely the total amount of tax a firm pays divided by its globally gathered income. The isolation of tax paid - only in the US - would be ideal but cannot be executed with the available data. But, since lion’s share of the banks in the regression gain their revenue mostly in the US, it causes particularly bias within the variables of the bigger banks. The second macroeconomic factor is real growth GDP. This will have a positive correlation with the profitability of the banking sector due to the increased (decreased) demand of lending during upward (downward)
economic movements.
3.4 Data
In this research, data of the 109 biggest banks in the United States will be gathered through
Datastream, at least for the bank specific information. The real GDP of the US will be
delivered by OECD.stat. Only American banks will be tested so the banks are committed to the same rules and have the same pro-cyclicality when looked at GDP. The dataset range of 109 banks has been chosen because the benchmark was a minimum of $1 billion in assets in the year 2017, this will create enough statistical power. A bigger set could also have been chosen, however, many banks above the largest 100 are comparable in asset size so it would be questionable in what manner it would add value to the outcome of the analysis. The starting year 2006 is just before the financial sector began showing its flaws, which should give a good insight in how profitability changes during economic downturns. In addition,
some banks are added to the regression that went bankrupt, merged or had fiscal aid and were not publicly tradable for some years. Reasoning behind adding these banks is that otherwise only the healthier banks would enter the regression which would lead to biasedness.
3.5 Hypothesis
H0: β1=0
H1: β1<0
H0: If β1 is zero, it can be assumed that tier one capital ratio does not influence the profit
significantly in practice.
H1: If β1 shows negative significance, it can be assumed that the tier one capital ratio does
have a negative effect on the ROAA in practice.
The model will be tested with the p-value <0.01 (t<2.327), this value is justified by the number of observations used (n=925). Beta 2 is left out because it is not inherent to the
leverage ratio as it is proposed in the Basel agreements.
3.6 Methodology
In this research, the hypothesis will be tested through a fixed effect regression. For this regression, a panel analysis will be done. The choice for fixed effect instead of random effect is appropriate since the Hausmann test gave a value less than the 0.05 significance. The main assumption of a fixed effect regression is that the variables which are going to be analysed vary over time. Within this regression method, the relation between the dependent variable and the independent variable is explored within the firm (Stock & Watson, 2003). ROAA i,t
is the dependent variable of bank i at time t . The timespan will be from 2006-2017. Furthermore, heteroscedasticity is assumed so robust errors are calculated. Alpha will be a constant and the two error terms are split up as ui (individual component) which is time
4 Results
In this paragraph the summary statistics and results are shown in two tables, followed by the elaboration of the outcomes. These outcomes will be linked to previous empirical studies and will be tested by the aforementioned hypothesis.
Table 1: summary statistics
Variables Mean Standard deviation Median Interquartile range
ROAA 1.065 1.019 1.080 0.510
Tier one capital ratio 0.119 0.056 0.122 0.037
Leverage ratio 0.452 0.223 0.470 0.299 Natural Logarithm of assets 16.845 1.685 16.467 1.853
Provisions for loan losses 665270 3164090 26819 111223
Net interest income 0.575 0.155 0.582 0.207
Tax rate -0.25 0.155 -0.297 0.144 Real US GDP growth 1.168 0.511 1.020 0.010
Table 2: fixed effects regression
Variables ROAA
Tier one capital ratio -3.274**
(-2.78)
Leverage ratio 0.145
(0.38)
Natural Logarithm of assets -0,07
(-0.71)
Provisions for loan losses -0,001
(-1.41)
Net interest income -0.001
(-0.98) Tax rate -2.088* (-2.44) Real US GDP growth 0.0591* (-2.07) Constant 1,908 (-1.22) N 925 R-sq. 0.1311 t-statisticus in parentheses * p<0.05, ** p<0.01, *** p<0.001
4.1 Elaboration of the results
Looking at the summary statistics in table 1 it can be seen that the ROAA is very volatile with a standard deviation the same size as the mean, this might be due outliers because the
interquartile range shows 0.51 compared to the 1.065 mean. If compared to the ROAA of 2 percent in Sub-Saharan African countries, a significant difference can be determined (Flamini, Schumacher, & McDonald, 2009). The tier one capital ratio does not show
deviating results. The mean of 0.452 considering leverage is lower than expected because this reflects that banks on average facilitate their funding with more capital than debt. The
provisions for loan losses shows a great standard deviation, this makes sense because the
variable is not size adjusted, leading to not very useful information. Banks revenue 57.5 percent out of interest income with a standard deviation of 15.5 percent. The tax rate mean is 0.25, implicating that banks pay 25 percent taxes on their net income. The last variable, does not unveil much because the real GDP growth only contains 12 different data points.
Considering table 2, the fixed effects regression output shows significance although it has a (within) R-squared of 0.1311. This is not a high measure but can be explained by the number of variables used, in comparison to all variables that influence the profit outside this model. Because of the scarcity of available data and the complexity of some variables, it is not feasible to add these variables to this model for this study.
In the regression, 3 variables show significance (at least p<0.10). All the variables will be discussed, and theories will be linked to see what value this analysis has. The tier one capital
ratio shows negative significance, this is perfectly in line with the theory and empirical
outcomes of Hoffman (2011). Furthermore, both the logarithm of the assets and the leverage ratio are not significant. The ultimate size of a bank is a widely discussed issue in economics, as addressed in the variable selection section. The outcome of this variable varies through different studies. Dietrich and Wanzenried (2009) conclude there is no difference in relative profitability by assets size, however in research executed in Kenya, significant outcomes were hatched (Mathuva, 2009). Along these lines, the result of this study does not reveal much. Although most of the studies regarding leverage analyse a positive outcome, it can be reasoned that leverage costs profit because of increased chance of default (Athanasoglou, Brissimisa, & Delis, 2008). Since this time span of this research is during the crisis, the positive and negative part of leverage concerning profit might level each other out, causing a
non-significant result. The provision for loan losses does not add significant value to the profit. In the research of Dietrich and Wanzenried (2009) this variable shows a comparable outcome. The net interest income is additionally insignificant. This is a deviation from the finding of Dietrich and Wanzenried (2009). In their research, only banks in Switzerland are observed and within a different time period (1999-2006). These different circumstances could be the difference, however, omitted variable bias is considered as well. The tax rate is
negatively significant, this outcome is inherent to the empirical analysis of Demirguc-Kunt & Huizinga (1999). The real GDP has a positive significant effect on the profit. This makes perfect sense, because overall demand in all sectors increases when GDP rises, and thus the demand for loans and securities rise as well, which affects bank profitability positively (Demirguc-Kunt & Huizinga, 1999).
4.2 Testing the hypothesis
To test the hypothesis, some underlying shortcomings must be considered. Due to the small number of variables contained in this model, omitted variable bias is inevitable. Although some banks were included in the model suffering default or other deficiencies, the vast majority of the tested banks survived the crisis properly, resulting in survivorship bias. Concluding that within this research project, there is enough statistical evidence to reject the H0 since P<|0.01| given P = -0.007. Since the hypothesis can be rejected it can be concluded
5.0 Conclusion
In this study, it is tested if the new capital requirement regarding the: tier one capital ratio imposed by the BCBS has a negative influence on the ROAA of banks, in practice. The requirements have a negative influence on the profitability according to the banking lobby, but is this the case in practice? After elaborating theories and empirical studies regarding the subject, not a straight forward answer can be formulated because different studies conclude with alter outcomes. However, after statistical analysis, it can be concluded that the
aforementioned ratio has a negative effect on the banks profitability, in the US. Since the different conclusions drew by various studies, conclusions should be formulated carefully. And ever since the banking industry is opaque and a complex sector, results from one country cannot be generalized (with such a study as this for the least). For further research more data should be gathered, with more independent variables (to reach higher R-squared), especially the two variables intended to test which were not available (yet) in databanks offered by the UvA. In addition, a cross-country analysis would give more insight in the banking sector as a whole, which could lead to a more generalizable conclusion.
References
Aboura, S., & Lepinette, E. (2013). Do Banks Satisfy the Modigliani-Miller Theorem ? 13. Admati, A., DeMarzo, P., Hellwig, M., & Pfleiderer, P. (2011). Fellacies, irrelevant facts,
and myths in the discussion of capital regulation: Why bank equity is not expensive.
Stanford University: working paper.
Albertazzi, U., & Gambacorta, L. (2009). Bank profitability and the business cycle. Journal
of Financial Stability, 393-409.
Athanasoglou, P. P., Brissimisa, S. N., & Delis, M. D. (2008). Bank-specific, industry-specific and macroeconomic determinants of bank profitability. Journal of
International Financial Markets, Institutions and Money, 121-136.
Bank for International Settlements. (2017, december 7). Basel III: international regulatory
framework for banks. Retrieved from https://www.bis.org:
https://www.bis.org/bcbs/basel3.htm?m=3%7C14%7C572
Barrell, R., & Davis, P. E. (2008, October). The Evolution of the Financial Crisis of 2007—8.
National Institute Economic Review, 206, 6-14.
Benink, H., Daníelsson, J., & Jónsson, Á. (2008). On the Role of Regulatory Banking Capital. Financial Markets, Institutions & Instruments, 17(1), 85–96.
Bikker, J. A., & Hu, H. (2002). Cyclical patterns in profits, provisioning and lending of banks and procyclicality of the new Basel capital requirements. PLS, 55(221), 144-175. Bordo, M. D., Redish, A., & Rockoff, H. (2015). Why didn't Canada have a banking crisis in
2008 (or in 1930, or 1907, or …)? Economic History Review, 68(1).
Bourke, P. (1989). Concentration and other determinants of bank profitability in Europe, North America and Australia. Journal of Banking and Finance, 65-79.
Bridges, J., Gregory, D., Nielsen, M., Pezzini, S., Radia, A., & Spaltro, M. (2015). The
impact of capital requirements on bank lending. London: Bank of England Working
Paper.
De Grauwe, P. (2011). The Banking Crisis: Causes, Consequences and Remedies. World
Scientific Studies in International Economics, 23-46.
de Larosière, J., & al., e. (2009). Report of the de Larosière group - European Commision. Brussels: European Commision.
DeAngelo, H., & Stultz, R. (2013). WHY HIGH LEVERAGE IS OPTIMAL FOR BANKS. Camebridge, Massachusetts: NATIONAL BUREAU OF ECONOMIC RESEARCH. Demirguc-Kunt, A., & Huizinga, H. (1999). Determinants of Commercial Banks, Interest
Margins & Profitability: Some International Evidence. World Bank Economic Review, 379-408.
Dietrich, A., & Wanzenried, G. (2009). What Determines the Profitability of Commercial Banks? 12th Conference of the Swiss Society for Financial Market Researches. Geneva.
Dietrich, A., & Wanzenried, G. (2014). The determinants of commercial banking profitability in low-, middle-, and high-income countries. Elsevier, 337-354.
European Central Bank. (2015, Juli 10). What is a central bank? Retrieved from
https://www.ecb.europa.eu: https://www.ecb.europa.eu/explainers/tell-me/html/what-is-a-central-bank.en.html
Fiordelisi, F., & Cipollini, A. (2012, November). Economic value, competition and financial distress in the European banking system. Elsevier, 36(11).
Flamini, V., Schumacher, B. L., & McDonald, L. A. (2009). The determinants of commercial
Fonseca, A., Gonzalez, R. F., & Pereira da Silva, L. (2010). Cyclical Effects of Bank Capital
Buffers with Imperfect Credit Markets: International Evidence. Brasilia: Central do
Brasil Working Papers.
Goldberg, L. S. (2009). Understanding Banking Sector Globalization. IMF Staff Papers,
56(1).
Hoffman, P. S. (2011). Determinants of the Profitability of the US Banking Industry.
International Journal of Business and Social Science, 255-269.
Huang, R., & Ratnovski, L. (2009). Why Are Canadian Banks More Resilient? IMF Working
Paper, WP/09/152, 1-19.
Kashyap, K., Stein, J., & Hanson, S. (2010). An analysis of the impact of sustainability
heightened capital requirements on large financial institutions. Cambridge: working
paper.
Kiff, J. (2009, June). Canadian Residential Mortgage Market: Boring But Effective? IMF
Working Paper(WP/09/130), 18.
Mathuva, D. (2009). Capital Adequacy, Cost Income Ratio and the Performance of Commercial BAnks: Kenyan Scenario. The Internetional Journal of Applied
Economics and Finance, 35-47.
Miller, M. (1995). Does the M&M proposition apply to banks? American Economic Review, 261-297.
Mishkin, F. S., & Serletis, A. (2011). The economics of money, banking and financial markets. In F. Mishkin, The economics of money, banking and financial markets (pp. 17-42). Toronto: Pearson Canada Inc.
Molyneux, P., & Thornton, J. (1992). Determinants of European bank profitability: A note.
Journal of Banking and Finance, 1173-1178.
Morgan, D. P. (2002). Rating Banks: Risk and Uncertainty in an Opaque Industry. American
Review, 874-888.
Noss, J., & Toffano, P. (2014). Estimating the impact of changes in aggregate bank capital
requirements during an upswing. London: Bank of England Working Paper.
Pasiouras, F. K. (2007). Factors influencing the profitability of domestic and foreign commercial banks in the European Union. Research in International Business and
Finance, 222-237.
Regling, K., & Watson, M. (2010). Global and European Setting. In K. Regling, & M. Watson, A Preliminary Report on The Sources of Ireland’s Banking Crisis (pp. 11-20). Dublin: GOVERNMENT PUBLICATIONS SALE OFFICE.
Seccareccia, M. (2014). Financialization and the transformation of commercial banking: understanding the recent Canadian experience before and during the recent financial crisis . Journal of Post Keynesian Economics , 277-300 .
Staikouras, C., & Wood, G. (2004). The determinants of European bank profitability.
International Business and Economics Research Journal, 57-68.
Stiglitz, J. E. (1969). A Re-Examination of the Modigliani-Miller Theorem. The American
Economic Review , 784-794.
Stock, J., & Watson, M. (2003). An Introduction to Econometrics. London: Pearson. Tung-Hao, L., & Shu-Hwa, C. (2013). Does financial regulation affect the profit efficiency
