MSc in Economics and Business Administration (Applied Economics and Finance)
Department of Economics Master Thesis, October 14th 2010
Supervisor: Finn Østrup, Center for kreditret og kapitalmarkedsret Number of pages 97
Characters 233,291 Copenhagen Business School
The impact of financial
distress on external financing
A comparison of non-financial corporations in market- and bank- based financial systems
By Christiane Freund
Executive summary
This paper addresses a so far unexplored area of the difference between market- and bank-based financial systems. It addresses how the ability of non-financial corporations to obtain finance under the recent financial crisis is expected to change and how it actually changed. Non-financial corporations are the main value drivers of any economy. Their ability to obtain external financing affects the profitability of their investments and thus the economy’s ability to cease the crisis.
Large quantities of literature exist in the field of valuation of external finance, market- and bank-based financial systems and the fragility of financial systems towards financial distress.
Very little literature exists in the cross-field of these three areas of external finance. The literature has so far left unexplored the subject of how developed financial markets, with different types of capital structure differ in their reaction to shocks. I use existing literature to formulate suggestions for the unexplored area. 7 hypotheses summarize the suggestions.
The 7 hypothesis claim that credit tightening would force banks to sell securities and decrease their holding in bank loans. Depositors were expected to move their savings to investments in T-bills. Defaults in the Asset Backed Security market reduced the liquidity for banks with the consequence of banks selling assets and transforming the remaining assets to high risk liquid assets. All of these consequences were expected to be larger in market-based countries because systematic risk was expected to be higher. The increase in perceived risk both rationally and irrationally founded, were expected to be higher in market-based countries. Therefore value reductions in all types of external finance should be larger. Irrational behaviour should make lending and borrowing through bank loans more attractive. NFCs’ need for new finance should be reduced because prospects of future income decreased. Expansionary monetary policy should increase the value of especially bonds, but also bank-loans and equity.
It is believed that changes in the value of already existing external financing affect the ability non-financial corporations to obtain new finance. Thus, I measure changes in the value of external finance and the mix of new finance. Regression analysis is used to measure both types of changes. The former showed highly significant results for bond, equity and bank loan financing.
The latter showed insignificant results with respect to the mix of new financing.
The empirical findings were as expected for bonds, equity and bank loans: The value of bonds never decreased, but it almost stagnated through 2008. Equity decreased the most through 2008 and rebounded in 2009. Bank loans already showed signs of a slowdown in 2006 and its value decreased in 2008 and 2009. The empirical findings were unexpected when comparing the change in external finance of bank- and market-based countries. Contrary to the hypotheses, bank- based countries were more volatile. They experienced a larger value reduction of securities in 2007-2008 and a larger increase in 2009.
Contents
_Toc274657970
1 Introduction
... 11.1 Scope of interest ... 1
1.2 Problem statement ... 2
1.3 Methodology ... 3
1.4 Sources ... 4
1.5 Delimitation ... 5
2. Literature
... 62.1 External financing and its nature ... 6
2.1.1 Bonds ... 7
2.1.2 Equity ... 8
2.1.3 Bank loan ... 10
2.1.4 Comparison of securities and bank loans ... 12
2.2 The Market-based and Bank-based Financial Systems... 14
2.2.1 Income and Capital Structure ... 14
2.2.2 Information access, relationship and transaction costs ... 14
2.2.3 Information asymmetry, moral hazard and agency problems ... 15
2.2.4 Market efficiency and risk sharing ... 17
2.2.5 Asset Backed Securities and Credit Default Swaps ... 18
2.2.6 Market instability, fragility and systematic risk ... 20
2.2.7 The financial crisis and capital access ... 21
3 The Hypotheses
... 243.1 The contribution of this paper ... 24
3.2 Proposed effects of the shock on the financial systems ... 25
3.2.1 Banks affect all types of external financing ... 25
3.2.2 Different levels of systematic risk ... 30
3.2.3 External financing and panics under financial crisis ... 33
3.2.4 Non-risk related change in required return and interest rate ... 35
3.2.5 The NFCs’ new finance unrelated to risk and required return ... 36
3.3 The hypotheses in short ... 36
4. Empirical foundation
... 394.1 Structure of reporting the results ... 39
4.2 Data collection ... 39
4.2.1 Data sample ... 40
4.2.2 Country determination and categorisation ... 41
4.2.3 Critical assessment of the country selection ... 43
4.2.4 Measurement techniques and variables ... 43
5 Empirical results
... 515.1 Research question A ... 51
5.1.1 Bonds ... 54
5.1.2 Equity ... 57
5.1.3 Bank loans ... 60
5.2 Research question B ... 62
5.2.1 Results ... 62
5.2.2 Why net-transactions provide insignificant coefficients ... 63
5.3 Research question C ... 64
5.3.1 Mathematical understanding of financial system-differences ... 65
5.3.2 Introducing the results ad interpretation of these ... 67
5.3.4 Change on total external finance ... 75
6 Discussion
... 80Hypothesis I... 80
Hypothesis II ... 83
Hypothesis III ... 86
Hypothesis IV ... 88
Hypothesis V ... 91
Hypothesis VI ... 93
Hypothesis VII ... 95
7 Conclusion
... 97Bibliography ... 99
Equations
Equation 4.1 – Degree of market-orientation, per year and per country ... 41
Equation 4.2 – Average degree of market-orientation, per country ... 41
Equation 4 3 – Weighted degree of market-orientation... 42
Equation 5.1 – The contribution factor ... 55
Equation 5.2 – The marginal contribution factor ... 56
Equation 5.3 – The CCT marginal contribution factor bond level ... 65
Equation 5.4 – The CCT marginal contribution factor for equity ... 65
Equation 5.5 – The CCT marginal contribution factor for bank loan level, total ... 66
Equation 5.6 – Yearly percentage change in external finance ... 75
Graphs Graph 5.1 – Changes in bond level of NFCs ... 52
Graph 5.2 – Changes in NFCs’ equity level over time ... 52
Graph 5.3 – Changes in NFCs’ bank loan level over time ... 53
Graph 5.4 – Net transactions in short-term bonds ... 63
Graph 5.5 – Marginal contribution factor for bond level ... 68
Graph 6.6 – Marginal contribution factor for equity level ... 71
Graph 6.1 – Yearly change in consumer deposits ... 84
Graph 6.2 – Short-term interest rate on government bonds... 84
Tables Table 4.1 – Categorization of countries ... 42
Table 4.2 – Short name and components of variables in Regression I ... 44
Table 4.3 – Properties and brief interpretations of variables in Regression I ... 45
Table 4.4 – Dependent variable ratios of Regression II ... 50
Table 5.1 – Regression I output for bond level ... 54
Table 5.2 – Time estimators for bond level ... 55
Table 5.3 – Cross-time term estimators for bond level... 56
Table 5.4 – Regression I output for equity level... 57
Table 5.5 – The time estimators for equity level... 58
Table 5.6 – The cross time-term estimators for equity level ... 58
Table 5.7 – Regression I output for bank loan level ... 60
Table 5.8 – The time estimators for bank loan level ... 61
Table 5.9 – The cross time-term estimators for bank-loan level ... 61
Table 5.10 – Volatility in the marginal contribution factor for bond level... 69
Table 5.11 – Volatility in the marginal contribution factor for equity level ... 72
Table 5.12 – Marginal contribution factors on bank loan level for each country ... 72
Table 5.13 – Volatility in the marginal contribution factor for bank loan level ... 74
Table 5.14 – Percentage change in total external finance ... 75
Table 5.15 – Example of unrealized return ... 77
Table 5.16 Percentage standard deviation from the average level of external finance ... 78
Appendix
Appendix 1: Country categorization
Appendix 2: Variable coding for Regression I
Appendix 3: Normal distribution of the level data in logged values and a natural scale Appendix 4: Regression 1 output for all types of external finance
Appendix 5: Variable coding for Regression II
Appendix 6: Regression II output for all ratios between different types of external finance Appendix 7: Graphs for net-transactions in long and short-term bonds, equity and bank loans.
Appendix 8: Marginal contribution factors and associated graphs
Appendix 9: Capital structure for each country and each type of finance over time Appendix 10: Correlation coefficients between level and capital structure
Appendix 11: Discussion of the hypotheses Appendix 12: Bank assets
1
1 Introduction
1.1 Scope of interest
For many years it has been discussed whether bank-based financial systems contribute better to growth than market-based financial systems. A bank-based financial system is a region (typically a country) where financing through loans dominates bonds and stocks and vice versa. Several authors have described relative efficiency, stability, strength and weaknesses of the two different systems. However, no authors have addressed how and to what extent these two systems affect the non-financial sector’s ability to obtain finance under periods of financial distress. This is what I want to do in this paper.
The non-financial corporations (NFCs) represent the part of the economy that produces services and products; both private and public. Hence the non-financial sector holds a significant share of jobs and value creation in all countries. In the end, financial intermediaries (incl. pension funds, insurance companies and commercial banks) invest their money in the non-financial sector; directly or indirectly.
Financial distress can evolve into financial crises and this can begin and develop in many ways. Depending on how financial instability starts and how it evolves, the impact on both the financial and non-financial sector will differ. Therefore one must be careful not to generalize the findings on how and to what extent macroeconomic determinants affect the NFCs. Yet, understanding why and how NFCs’ ability to obtain finance was influenced by the recent crisis can lead us to better future decisions about the optimal financial structure.
Therefore the paper only considers the recent financial distress beginning in 2007 and the scope is purely the NFCs. In addition the focus is on European countries and well as industrialized Anglo-Saxon countries in order to control as much as possible for other major discrepancies than financial structures.
Since NFCs are the main value driver in all countries, their access to finance is absolutely crucial for how fast a real economy of a country tackles financial distress. Finance is not the only remedy to cure low growth rates. Nevertheless access to finance at favourable prices increases the number of profitable businesses benefitting the creditors and investors as well as the labour force and hence countries in general. Finance is in this paper defined as external finance, not internally generated cash. External finance is roughly divided into bond financing, equity financing and bank loan financing.
2
1.2 Problem statement
1) How would one expect new financing and the level of finance for non-financial corporations to change in the event of financial shocks similar to the one occurring in 2007-2009? Would such expectation differ between market-based and bank-based financial systems?
2) In what way have new financing and level of finance for the non-financial corporations actually changed during the recent financial crisis? Do such changes differ between market-based and bank-based financial systems?
The theoretical answers to problem statement 1 are formulated as hypotheses. These are based on existing literature in the field and they distinguish between the effects that strike bonds, equity and bank-loans in market- and bank-based financial systems.
The empirical answers to problem statement 2 are provided through three research questions that differentiate between the types of external finance:
A. How does the overall level of finance change under the crisis?
B. How does the mix of the three kinds of new finance change during the crisis?
C. How do the results of A and B differ between bank and market-based economies?
Research questions A and B use the level of finance and kind of new finance as variables.
The level is a measure for the total value of finance as stated on the balance sheet of NFCs.
Changes in level equal net-transactions, value changes and accounting related changes in the period of interest. New finance is in practice measured as net-transactions. Thus the difference in the changes in the level and new finance lays in valuation and other accounting related changes. New finance is the most interesting measure as it directly addresses the NFCs’ ability and willingness to obtain new finance. However, changes in the level through revaluation are expected to affect net-transactions and highlight why new external finance would change from one period to another. Therefore both variables are included in the analysis.
Research question B concentrates on whether one type of new external finance changes more than another type of new financing. It highlights relative change and consequently it addresses if one type of external finance was more exposed than another to the crisis, regardless to the total change in external finance.
Answers to research question C uses the results provided in research question A and B.
3
1.3 Methodology
The methodology serves to describe the general structure, data treatment and limitations of the paper. In-depth description of the purpose of each chapter as well as the structure and limitations of these are presented in the respective chapters.
Figure 1 below presents an overview of the paper, the order in which it is presented and how the different parts are related.
Figure 1- Model of the structure
4 Chapter 2 provides the informational foundation which serves three purposes: (1) To obtain the tools needed to create solid hypotheses, (2) to be able to treat relevant empirical data correctly and (3) to create a framework for interpreting the results. The informational foundation is divided in two subchapters; one presenting tools for valuation of financial assets and one introducing existing literature in the field of market and bank-based financial systems and financial volatility.
Chapter 3, The Hypotheses, first discusses how the theory and existing research coincide (and collide) and the literature gap this leaves. The literature gap is consistent with the problem statements. Secondly, expectations on how the level of, and access to, external financing changed during the recent crisis are formulated and thus problem statement 1 is answered.
Chapter 4 describes the most appropriate way for measuring research question A and B.
Research question C uses the output of research question A and B. The chapter exposes the statistical implications that must be dealt with to ensure the validity of the results. The chapter implicitly uses knowledge obtained in chapter 2 for formulating mathematical relational arguments, i.e. causal relations. The mathematical arguments are formulated in terms of regressions and treated in SAS Enterprise 4.2.
The results are presented and analyzed in chapter 5. They are divided into three parts, each corresponding to one of the three research questions. Each of these parts is again divided into subchapters concerning bond, equity and bank loan financing respectively. Chapter 5 merely points out the findings and a deeper argumentation for the underlying reasoning for the results is not carried out until chapter 6.
The Discussion, chapter 6, compares the three results. It is the part of the paper where information from all chapters are woven together in order to attain a higher level of analysis, reflection and discussion. All arguments are structured according to the hypotheses and they therefore highlight discrepancies and unanimities between the hypotheses and the results.
In the end, the conclusion sums up all findings in light of the problem statements and the hypotheses.
1.4 Sources
My second-hand sources are primarily used to answer problem statement 1 and consist of two types of input:
5 (i) Models and theory describing the nature of external finance. These are used in chapter 2.1 and build on literature published with an educational purpose and the chosen books are used at Copenhagen Business School in courses. One, a statistic book, is used in chapter 4.
(ii) Research papers. Chapter 2.2 uses research papers where theoretical models associated to market- and bank-based finance as well as financial volatility are tested and accepted or rejected. A significant amount of researchers contribute to the empirical findings. Often they disagree and as the macroeconomic structures have changed, what seemed true once is perhaps wrong today. I quote as many authors as possible and hold their arguments against each other. I do not judge who is right and who is wrong. I merely highlight all views in order to expose the uncertainty in the literature and use the perspectives to create my own picture of the current literature.
My first-hand sources are used to answer problem statement 2:
(i) I have collected empirical data from national statistical databases and the central banks in each country. These databases are also used by the OECD and Euro Stat. Since I consider these as valid data sources, I also consider the national data bases to be valid.
(ii) I have backed statistical theoretical arguments and methods with dialogues with a statistic teacher at CBS, Cedric Schneider.
1.5 Delimitation
Specified delimitations will be described in details throughout the paper where appropriate. In general I have delimited my thesis in three areas:
(i) Scope delimitation. The paper does not distinguish between different types of NFCs and their associated risk level. No relations between NFCs access to external finance and the real economy’s situation is examined, even though the paper claims such relation to exist.
(ii) Data and literature delimitation. Hundreds of articles and theoretical books have been written on the broader subject of financial systems (though none on my specific subject). I strive to use as wide a pallet of perspectives as possible, but I trust that the amount of information used to cover the general view is sufficient.
(iii) Expectations to the reader. In general it is expected that the reader possesses thorough knowledge in the area of finance, valuation and risk. Furthermore a basic understanding of statistic regression analysis is required.
6
2. Literature
Chapter 2 is divided into theories and models on external financing (2.1) and research papers describing the characteristics of market and bank-based systems (2.2). Then this chapter is used throughout the rest of the paper as a framework for intuitive treatment of problem statement 2 and directly to answer problem statement 1.
2.1 External financing and its nature
At its broadest level, one can split financing into three categories: debt, equity and internal financing. Debt and equity are together referred to as external financing. Internal financing of a project originates from either liquid assets sold to obtain the necessary finance or from net income earned in the previous periods. As this paper is interested in external financing, internal financing will not be further addressed.
External financing as debt includes primarily two types of debt: Bank-loans and bond securities. Both of these exist in many forms and with different maturities. Bank-loans are credit by banks and differ mostly from bonds in the sense that they seldom are traded on the secondary market (Brealey et al, 2008). The bank tends to keep its investment to maturity.
Bonds on the other hand are sold all the time at the secondary market.
External financing in form of equity is distinct from bonds and loans, as equity investors stand as owners of the company with the right to all residual income.
External finance can be addressed based on whether it is publicly traded (secondary market trading) or held by the original creditor (primary market). Bonds and equity are traded on the secondary market and hence referred to as securities. Financial markets where most of the external finance consists of securities are referred to as market-based financial systems. When NFCs rely heavily on bank-loan the system in question is named the bank-based financial system (Demirgüς-Kunt & Levine, 1999). This distinction is used in the rest of my paper as we check if and how the two systems react to financial distress.
In order to conduct a thorough analysis of shock effects, one has to understand the similarities and differences between the three types of external financing. The next four subchapters serve the purpose of explaining the role and terms of bond-financing, equity-financing and bank- loans. In the end, they will be compared and their relative attractiveness is addressed; first from the NFC’s perspective, then from the investor’s and creditor’s perspective.
7
2.1.1 Bonds
The explanation of bond characteristics is divided into contractual composition and valuation of bonds. The same division is used for equity.
2.1.2.1 The contractual agreement
Bonds are debt securities which a debtor can issue in order to receive a certain amount of cash today. It is a certificate stating how much a debtor owes to a creditor, the payment terms, price, trustee, sinking fund terms and potential call provision (option to repay the debt before it is due). It does not state who the creditor is as the security in principal could shift hands 1000 times a day (Brealey et al, 2008, p.670).
Debtors have the option to pay the principal (the money borrowed) back to the creditor. If he decides not to, the creditor can claim the debtor’s assets corresponding to the value of the debt. This option is referred to as the option to default. If the value of the assets is larger than the value of the debt, the residual value belongs purely to the debtor, i.e. equity holders. If the assets are worth less than the debt, the creditor can only claim the value of the assets and not the value of the debt. This is a valuable option for the debtors, as the loss is limited to the value of equity, i.e. assets minus debt (Brealey et al, 2008). The option increases the creditor’s risk and thus bondholders require a premium to take such additional risk.
The bondholders will never earn more than they were promised if they keep the bond until maturity. If the promised cash flow differs from what they expect (e.g. due to the risk of default) creditors adjust the price to their expectations (Brealey et al, 2008, chapter 4). Bond creditors are more risk adverse than stockholders. They can never receive more than promised, only less (in event of bankruptcy), but on the other hand they are entitled to the NFC’s cash flow before the stockholders.
Bonds differ with respect to maturity, type of interest rate (floating or fixed), currency and contractual promises such as collaterals and payment structure. The price of such a bond will depend on all the variables above. In addition a fee is charged by the underwriter to manage issuing and reselling of bonds. The price in the US is about 7 % of the sum raised from creditors. No significant scale economies seem to exist (Brealey et al, 2008, p.418).
2.1.2.2 Valuation of bonds
Bond holders required return is called yield to maturity (YTM). It is used to determine the market price of a bond given the time to maturity, coupon rate, face value and payment structure. YTM can be seen as the creditor’s required risk adjusted compensation over time for lending out money. YTM is an average rate of what the creditor is promised to receive in
8 nominal return for investing in 1-year 0-coupon bonds in sum replicating the terms and cash flow values of the original investment.
∆ in bond value is a function of the ∆ in expected yield to maturity (YTM) (Brealey et al, 2008, chapter 4). YTM consists of different types of compensation which can be divided into (I) compensation related to default free treasury bonds and (II) the additional premium related to all other types of bonds (Ross et al, 2006, chapter 7). In detail:
I.a. Real interest rate or the time value of money (Brealey et al, 2008).
I.b. Inflation premium: Compensation for the value reduction of money over time.
I.c. Interest rate risk premium: It points back at the two former premiums and express compensation related volatility in the nominal interest rate.
II.a. Risk of default: It is already explained above.
II.b. Taxability premium: The additional tax the government charges on corporate bonds.
II.c. The liquidity premium: Compensation for illiquidity of assets (more about this in the subchapter 2.1.3 Bank loans)
2.1.2 Equity
The explanation of equity characteristics is divided into contractual composition and valuation of bonds.
2.1.2.1 The contractual agreement
Equity is the capital belonging to a firm’s owners. Equity and bonds are first issued at the primary market by a given company, and then it is sold and bought on the secondary market.
It is the price available on the primary market that is relevant for the NFC, but also the signalling price on the secondary market, since it affects the primary market price. Signals indicate the risk/stability of a company’s strategies and operations as well as the profitability of the company.
The equity value = because the owners can refuse to pay debt if it exceeds the value of assets (Brealey et al, 2008). Ones they have repaid their obligation, there is in theory no limit for how much they can earn. The result is (i) higher volatility in income and (ii) a non-normal distribution of income, thus enhancing equity owners’ motivation to take risk.
If only one owner exists and he operates his company, the interest of owners and the management is aligned. The risk of conflicts arises when the management has different incentives than the owners. A relating monitoring cost is encountered and contracts and
9 incentive schemes are used to reduce the problem. Such problem stems from asymmetric information and moral hazard (Bennedsen & Nielsen, 2009) and it cannot be eliminated, only reduced. Hence a risk premium is added by owners.
2.1.2.2 Valuation of equity
One crucial factor separates the valuation method related to equity from that of bonds and bank-loans: A corporation is assumed to live forever and therefore it does not mature at a certain point in time. Not all products are relevant in infinity, but ones a product is pulled back from the market it is assumed the money is used for a new product, just as profitable. If not, the owner will receive the money, and given the market is efficient the owners can freely reinvest their money in other firms. The market value is the sum of all discounted cash flows, where the discount rate is the cost of equity, also called the required return, R (Brealey et al, 2008).
R is determined by the general market expectations given a perceived risk level. It includes three types of compensation:
1. Real interest rate or the time value of money 2. Inflation premium
3. Risk premium: Just as with bonds, equity holds risk. Risk is associated with volatility in (i) the future cash flow and (ii) interest rate risk premium.
(i) The variance of the cash flow is a function of the volatility in the difference underlying cash flows, e.g. the cash flow of purchasing costs and the cash flow of income, and the weights of these cash flows. In addition the correlation between the underlying cash flows also affects the total volatility. For accepting cash flow risk, a risk premium is required. Risk can be separated into two categories, unique risk and market risk. Unique risk refers to the volatility in cash flow which affects one company or a few companies. It can be diversified away by investing in negative correlated assets. Market risk refers to the type of volatility in cash flow that to some extent affects all companies in the same way. For the same reason one cannot hedge away this type of risk (Brealey et al, 2008) and thus a premium is required for absorbing market risk, also named systematic risk. (ii) The change in real interest rate and inflation follows the same logic as for bonds.
The price of equity changes as the expected cash flow and components in R changes. Reasons for this are discussed later in the paper. Given a certain required return of equity or interest of
10 bonds, the volatility in equity prices (stemming from either changes in expected cash flow or changes in R) is larger than for bonds, because the duration is infinite.
2.1.3 Bank loan
The 1st subchapter concerns the activities characterizing banks, as banks are the only who can grant bank-loans. Understanding them is important to understand bank-loans. In the same manner the 2nd subchapter addresses the competitive advantage of banks and bank-loans.
2.1.3.1 Structural form
Financial intermediaries (from now on FIs) are, in contrast to NFCs, firms who hold large quantities of financial claims as assets (Greenbaum & Thakor, 2007, p. 50). Parts of these assets correspond to the liabilities of NFCs and households. FIs take many forms such as commercial banks, pension funds or insurance companies. The financial claims they hold exist in numerous versions. Banks hold both securities and loans. Loans are issued in shape of consumer credit, business loans and mortgage (Greenbaum & Thakor, 2007, p. 174).
Therefore, when explaining the nature of bank loans one has to be careful venturing into an analysis of banks as a whole.
The fact that bank loans exist implies that they can offer something that the capital market cannot. When banks, many years ago, started to lend, it was in shape of loans. Today, credit exists in many hybrid versions, which is partly a result of a more efficient capital market. It has reduced the benefit of bank loans and forced banks to compete in and with the secondary market where securities are traded (Greenbaum & Thakor, 2007, p. 179). Commercial banks held around 40 % of their assets as loans in the US in 20071. Of this, about 50 % went into mortgage, 40 % into business loans and 10 % into consumer credits. Thus, loans are still important to banks and understanding the role of bank loans is to understand banks and their competitive advantage.
2.1.3.2The role of banks
Banks occupies two major activities: Brokerage and quantitative asset transformation (QAT) ((Greenbaum & Thakor, 2007, chapter 2). Brokerage concerns matching creditors and debtors by screening them and categorizing them into different types of debtors and creditors (depositors). QAT involves transforming deposits with one type of characteristics into loans with other types of characteristics.
1 Among the country sample used in this paper, US NFCs are the ones holding least liabilities in form of loans and it is therefore reasonable to believe banks in other countries to hold at least 40 % of their assets as loans.
11 Banks earn money on difference between the interest paid to depositors and the interest required by the debtors. Loans are usually more expensive than bond-loans collected on the free market, so to stay attractive banks must provide something the capital market cannot.
The role of the broker is to reduce the problems that can occur due to asymmetry between information held by the bank and the information held by the debtor. Informational symmetry is only a problem if the debtor uses his private information to maximize own utility to exploit the creditor. Such action is referred to as moral hazard. Brokers seek to align both pre and post informational asymmetry and they succeed better than individual creditors, because they are specialized and because they serve so many debtors that scale advantages exist and cost of screening per debtor decreases. However, the broker role can exist without the bank holding any significant assets, i.e. by servicing other investors on their respective investments.
Yet, banks choose to conduct the broker role by using their own assets and liabilities.
Matching assets in the form of loans with liabilities in the form of deposits (incl. interbank lending), involves four overall tasks and hence four types of risk:
1. Matching assets and liabilities with different durations. Creditors tend to prefer short term investments and debtors long term investments. Higher price volatility and hence risk exists for long term assets than for short term liabilities. The related interest rate risk premium results in YTMLong > YTMShort and this creates a base for profit.
2. Matching liquid deposits with illiquid bank loans. It benefits the depositor as he can withdraw instantly keeping risk low. It benefits the NFC as it expands the capital market.
Withdrawal risk generates liquidity risk, as banks ability to meet depositor claims is in danger if a large number of depositors withdraw at the same time. Governments partly insure deposits which reduce overall liquidity risk and hence premium of the bank loans.
3. Matching deposit and loans with different default risk. Banks absorb debtors’ risk of defaulting, partly by diversifying risk (only possible because they serves many debtors), partly by monitoring debtors and partly with help from the government’s deposits insurances. This attracts money from very risk adverse depositors thus expanding the credit market available for the NFC. For this service and risk absorption banks charge a credit risk premium.
4. Matching depositor savings with debtor loans. Savings are usually smaller than demanded loans, so the task enhances access to loans and in this way banking increase market efficiency.
12
2.1.4 Comparison of securities and bank loans
The properties and valuation of bonds, equity and bank loans were described above, because potential changes in NFCs’ financial structure and level during the crisis somehow is caused by changes in the components used to value external finance. These components, i.e. discount rate and expected cash flow, differs depending on the specific type of external finance which is characterised by its contractual agreement. This chapter compares the contractual
agreement and price of external finance with the relationship between the creditor and debtor.
It discusses how these variables determine the creditor’s and debtor’s preference for one type of finance over another.
2.1.4.1 Comparison from the NFC’s point of view
It has been shown that NFCs tend to borrow from venture capitalist when they possess few tangible assets and as they grow larger they turn to banks for capital. The mature firms prefer the capital market, but they also use bank loans and of course equity (Greenbaum & Thakor, 2007, p. 61). Lack of tangible assets means lack of insurance against the loan and thus the borrower has a larger incentive for hazardous behaviour. Therefore venture capital works as equity which gives the creditor legal right to shape and affect decision making. In turn, this reduces the risk of moral hazard. Since this is the only way the borrower can obtain any capital, they will pay the price the venture capitalist requires.
Banks do not work as venture capitalist. Banks (with blurry limits) are creditors, not owners.
They use contracts including covenants in form of collateral to reduce monitoring cost. In addition, the governments partly insure deposits. Therefore, the deposits carry a lower interest rate than venture capital does: Seen from the depositor’s perspective, the default risk is lower and so is monitoring costs. NFCs prefer bank loans to venture capital when they have sufficient assets to obtain the lower interest on bank loans.
Bank loans usually mature faster than bonds and therefore loans need to be refinanced from time to time (James, 1987). The security market uses the information of refinancing as a sign of creditworthiness, so the longer a NFC has obtained loan from one specific bank, the more the security market interpret these signals as signs of high creditworthiness. In the end, it enables a trusted NFC to avoid the banks and their broker fees and instead enter the security market. The lower interest rate in the security market is an incentive for NFCs to protect their reputation. The value of reputation works as self-control not to conduct moral hazard.
13 Therefore the creditors’ monitoring costs are reduced, but not eliminated, and this is part of the reason why interest rates on bonds are lower than those on bank loans.
Concerning equity, NFCs do not issue new equity each day, as it can send mixed signals:
Some could interpret new issuing as a last resource to external capital and as a rejection from banks and other bondholders. Others could interpret new issuing as a sign of yet another profitable investment where equity financing would keep the weighted average cost of capital (WACC) at a minimum. The fact is still that loans and bonds are used much more for refinancing and new investment financing than stocks. This is referred to as the pecking order of finance (Myers and Majluf, 1984).
2.1.4.2 From the Investor’s Point of View
When an investor requires a return or price, it covers a basic compensation premium plus a risk premium and perhaps intermediary costs (broker fees). This is valid for all types of new external financing.
According to the capital asset pricing model (CAPM), an efficient market holds a constant relation between the price required by investors (return) and the market risk related to a certain portfolio. If asset prices divert form the constant relation the market will correct the price. In reality, many critics have attacked the model, but intuitively it makes sense if we recognize that no one will take risk without perfect compensation and that the market is efficient enough to prevent superior profit in the long term.
If the CAPM is accepted, then FIs’ investments in bank loans or securities illustrate how the different financial sources with different risks meet different preferences. The interest rate merely balances the risk and the expected return. When banks invest in all three types of finance they create a portfolio corresponding to their risk tolerance.
The same is true for the borrower, i.e. the NFC, who also chooses a portfolio of liabilities which ideally creates the lowest WACC and all things equal the highest firm value (Ross et al, 2006, chapter 15.6). The relevance for this paper is not to determine appropriate WACC and thus capital structure, but rather the change in the balance in the event of financial distress.
14
2.2 The Market-based and Bank-based Financial Systems
This chapter is divided into different subjects relevant to problem statement 1. The subjects present research findings and proposed theoretical relationships by other authors. In general they are structured according to the two main division of the literature: (i) the characteristics of the two systems disregarding point in time and (ii) how these systems reacts to changes over time, such as financial shocks.
2.2.1 Income and Capital Structure
Several authors have proven that as GDP increases countries tend to move towards a more market-oriented financial system (Murinde et al, 2004). Murinde, Agung & Mullineux demonstrate how especially equity plays a still larger role for European countries as these countries get richer, while bonds hold a more constant position and bank-loans decreases.
Internal finance plays a still larger role. Demirgüς-Kunt and Levine (1999) found the same relationship between the degree of market-orientation and income to be true, but they used a worldwide selection of countries. They defined the degree of market orientation as a function of the stock market relative to bank loans; more specifically measured according to size, activity level and efficiency and limited to the private sector. Bonds were excluded in the survey. My paper has limited its sample to high income countries, so in light of Murinde et al and Demirgüς-Kunt & Levine, these countries should all be relatively market-oriented. In so, even though some of them are defined as bank-based and some of them as market-based their financial structure should not divert as much as if we included low income countries.
Therefore we would not expect the sample countries’ potential reaction to the crisis, as a function of capital structure, to divert as much.
2.2.2 Information access, relationship and transaction costs
Banks fulfil an intermediary role that on one hand increases the price of bank-loans relative to bonds (Holmstrom & Tirole, 1997) and on the other hand provide the more efficient and thus cheaper access to information for certain types of borrowers. Hence, on average bank loans are more expensive than bonds, and NFCs would all things equal prefer bonds. But for some NFCs, such as small companies with little collateral, banks can be the only way to obtain external capital. In the end, the cost of external financing depends on risk and information cost, where the latter partly is a function of the former.
Ross Levine includes information cost implicitly in his estimation of relative structural efficiency, as bank efficiency is measured partly by the interest rate level and partly by overall
15 cost to asset value (Levine, 2001). Hence he believes that informational costs affect relative efficiency and therefore it does not only affect NFC’s financing choice, but also the overall access to external financing and hence growth.
The harder it is to gain information, the more expensive it is. Thakor refers to informational economies of scale as information reusability and it originates from the fact that specialists of screening, such as banks, have many clients with either cross-sectional or inter- temporal similarities (Chan et al, 1986). The price of such information is less than the price paid by the free market investor, who has to gain company specific information from scratch every time a new potential investment appears.
The cost of gaining proper information to avoid adverse selection and moral hazard is reduced as investor protection is improved by La Porta (2005). Levine and La Porta agree that both the market and the bank-based system benefit from good accounting standards, investor protection, law enforcement and common law traditions. All these criteria seem to create an overall more developed financial system which in turn decrease the borrowing costs (Boyd &
Smith, 1998) and increase growth. As the banks’ strength mostly lie in NFCs where information reusability is significant, and as these NFCs for the same reason only can obtain external finance through bank loans, a low level of bank-financing enhances income inequality.
According to Chakraborty & Ray (2006) none of the systems are better for growth as long as the overall financial market is efficient. Levine supports this view. Given a certain level of financial development, the security market obtains information through signalling and through the reliance of how important the market is for the debtor. The lower the debtor’s wealth, the higher the gambling incentive. The higher the gambling incentive, the more the debtor will need to get his money from the banks. As the solvency decreases under financial crisis, more debtors might be forced to use bank loan financing.
2.2.3 Information asymmetry, moral hazard and agency problems
Carey did in 1995 find evidence that information asymmetries are not an important factor in bank loan contracting with large borrowers, but moral hazard is (Boot & Thakor, 1997, p.
727). This is interesting because information asymmetry itself is a precondition for moral hazard to occur: The debtor uses the asymmetry to increase own utility at the expense of the creditor and owners. The problem of moral hazard is twofold: First, depending on the degree of asymmetric information and the perceived likelihood of agency problems, the cost of
16 reducing asymmetric information (through screening and monitoring) and moral hazard (through contracts and monitoring) can be significant. Secondly, even though the highest level of monitoring and contractual agreements is conducted, contracts are incomplete and agency problems (moral hazard is a type of agency problem) cannot be prevented completely (Allan
& Gale, 2001). Thus, the associated risk premium and cost of screening and monitoring increases with moral hazard. According to Allen & Gale renegotiation can reduce the agency problems associated with incomplete contracts, because it gives the borrower an incentive to deliver the agreed result if refinancing must be an option. Does it mean that agency problems are controlled better in a bank-based system, assuming it is managed in a cost-efficient way?
Dewatripont & Maskin (1995) argues that debtors using the secondary market for external financing face several creditors and hence the debtor is not as important for the single creditor as he will be for the financial intermediary from whom he receives a larger fraction of his total loan. Renegotiation in the market is costly, because it involves more creditors. The debtor is more depended on the creditors as a whole than the creditors are on the single debtor. As a result, the capital market’s threat of no refinancing in event of moral hazard is more real and this can reduce the agency problem.
The two arguments reflect upon the relative power and interdependences between the borrower and creditor. Note that the discussion takes for granted the level of asymmetric information and focus on the relative agency problem of the two systems and how this can be reduced through renegotiation. Both markets can be just as efficient in reducing the agency problem; banks by creating tight relationships and the secondary market by reducing its dependence of the debtor. Relative efficiency depends on the cost of reducing agency problems of one specific debtor. Thus relative efficiency is not a constant unrelated to the characteristics of the debtor, but differs depending on how the strength of the two systems fits the profile of the debtor.
Moral hazard arising from agency problems, i.e. the fact that the interest of the creditor and debtor is not aligned, is linked to a non - normal distribution of the NFC’s profit. A debtor can maximum lose his equity, but profit can be infinitely large. As debt relative to equity increases, the incentive to take more risk increases at the expense of the creditor. Thus, the probability of moral hazard increases (Allan & Gale, 1998).
Owners should ideally be represented perfectly by the managers of the company and in this case investors face no agency problems. Still, La Porta et al emphasise the difference between insiders and outsiders, i.e. managers and controlling shareholders versus residual
17 shareholders and creditors (La Porta et al, 2000). Under financial crisis the solvency degree decreases and therefore controlling shareholders and management use a higher degree of self- dealing and risk taking (La Porta et al, 2005). Less profit is left for the remaining shareholders and thus securities to sell for a discount.
2.2.4 Market efficiency and risk sharing
Closely related to monitoring debtors is the concept of risk-sharing, which cover the situation in which a principal and an agent somehow align their interests in accordance with their risk- preferences (Shavell, 1979). It involves transferring of risk and associated required return from one actor to another actor.
Risk-sharing is closely connected to market liberalisation as less regulations encourage innovative products such as securitization (more about this later) and as these products enables intermediaries to repackage assets into new type of assets with different risk levels.
As a result the market of financial assets expands in form of higher overall supply and demand driven by more diversified products. The effect is increased liquidity, motivated by increased efficiency (less deadweight loss and lower transaction cost as the markets become more competitive). Gallegati, Greenwald, Richiarti & Stiglitz did in 2008 write that
“securitization proved an effective means to obtain diversification and risk reduction in good times. However, when the economic conditions started to deteriorate these linkages became detrimental and dangerous, building up a significant risk of systemic failure to the U.S.
financial system and imposing a significant a threat to global financial stability” (p. 3).
Risk-sharing can reduce default risk because more actors share a potential loss which decreases the risk of a chain-reaction in defaults. But risk-sharing in a combination with new innovative products has a downside in that the claims on the same underlying assets are sold and repackaged many time. First, this diminishes the final creditor’s ability to monitor the debtor which enhances the debtor’s incentive to take more risky positions. Secondly, the risk of miscalculating the risk associated to the underlying cash flow increases as information access decrease. Still, risk-sharing increases competition and thus efficiency in good times.
Financial market efficiency is not a matter of the bank- or market-based systems’ relative efficiency, but of their accumulated efficiency. In liberalized markets the two systems supplement each other, but bank-activities tend to move toward market-activities. Parallel to this, Boot & Thakor mention that an observable quality cut-off, where borrowers below this level is served by banks, tends to increase as banks become more competitive and the overall
18 market more efficient (Boot & Thakor, 1997). Hence, more debtors obtain finance from banks, but now in form of securities and loans later transformed into Asset Backed Securities.
Allen & Gale describe how banks have a long-term perspective and therefore diversify risk inter-temporally much more than other investors and creditors, (Allan & Gale, 2001 and 1995)2. In contrast, these other investors and creditors do not allow such a long perspective and therefore risk-sharing is more cross-sectional here. During the past years, banks have reduced inter-temporal risk sharing in favour of cross-sectional risk sharing in order to be more competitive and market-oriented,. In context of this paper it might be relevant to ask which of the two types of risk diversification is more fragile under financial distress.
2.2.5 Asset Backed Securities and Credit Default Swaps
Securitization can be seen as a fourth way of financing. It requires that the originator (a bank) holds cash-flow generating assets (securities, rental, leasing, loan etc). If the originator sells the cash flow (not the loan itself) to someone else, it can receive money today corresponding to the future cash flow. A claim is changed into something tradable; bank loans are transformed into a financial asset that is tradable on the secondary market. Securitization results in an increase in the bank’s assets corresponding to the cash received. The liabilities increase with the same value and the debt account is referred to as Asset Backed Securities.
The process of “transforming” loans into Asset Backed Securities is referred to securitization.
Securitization is a benefit for the originator as it isolates risk of a specific loan and free capital that can then be used for new investments (Greenbaum & Thakor, 2007, p. 180).
Securitization of assets that are better rated than the total assets, result in lower interest requirement by the Asset Backed Securities - buyers and correspondingly cheaper access to capital for both the bank and the original debtor.
From the buyer’s point of view, the securities can take the form and risk preferred, because many different loans are repackaged into new format that suits different risk preference of the investor. Consequently, more bondholders become interested in investing. In turn, market liquidity and money aggregates increases (Weithers, 2007).
If the risk is perceived as too big, Credit Default Swaps can be offered by a bank, an agency or others to design the perfect product for the bondholder. The seller of the Swap guarantees a pre-specified compensation level to the buyer of the Credit Default Swap in the event of
2 It means that banks hold larger reserves than other actors in the financial market, because it is in their interest to ensure future profitability. Such interest serves future generations, and results in less volatile asset prices today.
19 default. To provide this service, the credit enhancer, e.g. Freddie Mac, Fannie Mae or Ginnie Mae, receive an insurance fee. Credit Default Swaps do not have to be combined with Asset Backed Security and they can be sold by everyone. But they often are combined, and they are often bought and sold by banks. The reasons for combining them are many and complicated and thus out of the scope of this paper. The consequences, however, are important:
1. The risk of default is reduced to the extent that the seller of the swap does not default, i.e. to the extent that a double-credit event does not occur.
2. The liquidity of the Asset Backed Securities might increase, because the information gap is less important: As the risk is transferred to the Credit Default Swap-seller, risk assessment is reduced and so is the perception of risk while the interest rate might not fall correspondingly. Thus the demand increases. If the perception of risk is reduced, NFCs can possibly obtain capital cheaper thus making the market more efficient and growth-enhancing (Kaufmann & Valderrama, 2007).
3. Since the bank does not bear the risk of default, it might lower the quality of its monitoring and screening to increase profitability. The risk of this happening should increase the interest of Asset Backed Securities and the insurance fee of Credit Default Swaps. But if good lasts long, the credit enhancement institutions might underestimate the true risk and buy too many and too expensive Asset Backed Securities. Thus, the banks will earn profit on taking on additional risk. This profit relates to (i) fees and (ii) a potential spread between the interest rate on Asset Backed Securities and the interest received from the original loan.
Asset Backed Securities are owned by the same investors as the ones holding other securities (Mengle, 2007). Banks are the major sellers and buyers of Credit Default Swaps (PIMC, 2006). NFCs barely buy or sell Credit Default Swaps. As a result NFCs are affected by Asset Backed Securities and credit enhancers to the extent it affects their access to capital. Many market players are involved in the Asset Backed Securities and Credit Default Swaps markets and the latter is often issued many times the principal value of the original cash flow.
Therefore, ones a borrower defaults it affects investors all over the market. Stiglitz &
Greenwald (2003) describes cross-sectional risk sharing as an efficient shock absorber, because many parties lose some which decreases the risk of default. If defaults are comprehensive enough, cross sectional risk-sharing appears to be contagious and thus create systematic liquidity problems. It could reduce NFCs’ access to new finance because creditors’
20 liquidity is reduced, but also reduce the value of NFCs’ debt and equity as its creditors and investors default.
2.2.6 Market instability, fragility and systematic risk
Kaufmann & Valderrama (2007) show that there is a relationship between credit aggregates3 and asset prices4: (i) Credit and asset prices reinforce each other. (ii) Asymmetric dynamics mean that the relationship between credit aggregates and asset prices is not constant over time; it changes with the level of financial stability. (iii) Credit aggregates and asset prices reinforce each other more in the market-based financial system than in the bank-based system.
It means that a credit tightening in market-based countries is expected to decrease asset prices more than in bank-based countries.
According to Allen & Gale (2004) intermediaries in the developed financial markets are so closely linked and engage in the same activities that financial fragility increases. Fragility enhances systematic risk. But for such risk to result in a systemic slowdown requires that something trig a reaction. This could be withdrawal of bank deposits. Banks operate on competitive terms and invest in many different financial instruments and their access to capital depends on deposits with short-term withdrawal rights. Banks can experience liquidity squeezes, if all depositors withdraw at once. In that case they can be forced to sell rather illiquid assets (e.g. mortgage) below their true value. As a result, the market value of banks’
assets quickly falls. Thus banks’ equity decreases which reduces the portfolio value of bank investors. If the reduction in bank assets and equity is large enough it can trigger insolvency for many market actors. Thus interaction between different market actors, and similarities in their asset investments, combined with a heavy weight of illiquid assets, enhances fragility (Allen & Carletti, 2007). The more risk-sharing the financial system engage in, the more interdependent and fragile it will be, but it will at the same time become more competitive and efficient. Yet, Allen & Carletti argue that fragility is a sign of inefficient liquidity. Berger et al (2008) add that they have found evidence supporting this competition-fragility view; the more competitive the financial system is, the more exposed to risk it is.
Combining the perspective presented by Allen & Gale (2004) with the fact that Asset Backed Securities and Credit Default Swaps enhanced overall leverage before the crisis and thus increased credit risk, I argue that the level of interaction between the market and banks
3 Credit aggregates are the financial assets belonging to banks (Collins and White, 1999), but with special attention to credit, not equity investments.
4 Asset prices include equity and therefore more indirectly the investments conducted by the actors who borrow from the banks.
21 was so high that even small shocks in the financial market could trigger large changes in asset prices.
In contrast to Allen & Gale, Finn Østrup (2010) and Greenwald, Richiarti & Stiglitz (2008) argue that risk is more concentrated in bank-based economies than in market-based economies, and that this means less risk-sharing and thus larger risk of a crisis. Whether one system is more exposed than another probably depends on the magnitude of the shock (more on this later).
We end up with a paradox: Volatility in market prices can cause both stability and financial distress. Efficient markets demand that assets be sold quickly to their true value and that capital be efficiently distributed in the markets where it gives the highest risk-adjusted return.
Intuitively such markets must allow for rational founded volatility which ensures that a decrease in true value is reflected directly and immediately in market prices. Since volatility is exactly what characterizes financial distress, it seems as if well-developed markets all things equals are more likely to experience financial distress. On the other hand volatility also ensures stability in the way that market prices are constantly corrected and thus never stays long below or above true value.
2.2.7 The financial crisis and capital access
It is not relevant for this paper to describe the recent financial crisis in detail, but it is crucial to understand how it hit the financial market and especially secondary market.
“It is often argued that capital regulation is necessary to control the moral hazard problems generated by the existence of deposit insurance. In other words, the bank has an incentive to make excessively risky investments, because it knows that in the event of failure the loss is borne by the deposit insurance fund and in the event of success the bank's shareholders reap
the rewards” (Allen & Carletti, 2007).
Again, the asymmetric outcome distribution triggers an actor’s incentive to maximize his own utility at the expense of other market actors. Plender (2006) support this view. Diamond &
Dybvig agree but emphasise that the deposit insurance also enables banks to provide liquidity insurance to NFCs so they can obtain capital in the bond market and thus stabilizing NFCs’
liquidity (Diamond & Dybvig, 1983).
Many factors reinforcing each other can explain the crisis, but the statement above is a good starting point to understand the recent crisis. In addition deregulation of the financial markets,
22 and thus increased liquidity, motivated investors to bid up asset prices (e.g. housing prices).
Bidding up prices was largely enhanced by creditors’ lack of observance of risk and true value of the debtors’ investments. Such passive behaviour was caused by risk-sharing5 and thus a higher incentive to conduct moral hazard. This view is supported by Kaufmann and Valderrama’s survey of how credit aggregates and asset prices reinforce each other: Creditor investments in securities were for a moment selling at a price above its fundamental value.
Thus the perceived credit risk and associated premium primo the crisis was wrongly estimated too low. Debtors’ investments in relative high-risk assets (houses) were possible for a price (interest rate) that did not reflect the risk. As mentioned before, profit is earned when the underlying asset price is volatile. Return is even more volatile when the supply of assets is relatively fixed such as for houses (Allen and Gale, 1998). Combining (i) the asymmetric outcome distribution for the end-borrower/owner with (ii) the banks’ higher risk incentive (due to secured deposits and Credit Default Swaps selling with too low risk spreads), leads to an attractive environment for high return high risk high leverage investments (Whalen, 2007).
Thus the price is suddenly not based on the fundamental value of housing, but on liquidity in the market.
If investments result in lower returns than expected, some people might withdraw from the market leading to overall reduction in asset prices. Where individual investors do not act in accordance to their private information, but act as everyone else, aggregated information is not reflected in the price. If asset prices fall in this context, herd behaviour intensifies the trend and panics can occur (Cipriani & Guarino, 2008). The remaining investors can suddenly find themselves insolvent and this will be reflected on the banks’ balance sheets as well.
Credit tightening has the same result, as it might force investors to sell their assets, because they cannot repay debt (and they cannot repay debt, because the asset prices have fallen).
Thus, bubbles burst and, due to the interdependence of financial market actors and high level of systematic risk, panics occur. Housing assets are rather illiquid and liquidity shocks also happen because price reduction is larger than it would have been for more liquid assets.
Liquidity shocks in the housing market proved to be (i) very contagious, i.e. the shock multiplied from one market to another and (ii) risk-sharing resulted in a shared loss (Gallegati, Greenwald, Richiarti & Stiglitz, 2008). It went from end-borrowers to banks and/or other Asset Backed Securities -buyers to credit enhancers (Freddie Mae etc) and to the
5 For an example through Asset Backed Securities and Credit Default Swaps.
