The thesis examines subordinated debt holder market discipline in UK credit institutions during the period 1995 to 2002. The topic is relevant as current research is questioning the role and effectiveness of rules-based bank regulatory oversight, and favouring, instead, incentive-compatible regulatory design and market discipline. In particular, the literature proposes using signals from subordinated debt holders to constrain bank risk-taking. In addition, this market oversight may provide information signals to regulatory agencies that are useful in improving bank regulatory design. The thesis researches two prominent issues related to subordinated debt holder market discipline and, therefore, contributes to the debate in introducing incentive-compatible polices in bank regulatory design. First, testing the risk sensitivity of UK credit institution subordinated debt spreads assesses whether investors are signalling bank risk in market prices. The UK evidence supports the theoretical literature in claiming that eliminating too-big-to-fail policies can encourage effective incentive-based mechanisms. Secondly, the research examines the appropriateness of introducing a mandatory subordinated debt policy in the UK. The empirical analysis raises a number of themes, many of which are in stark contrast to US and other European banks' subordinated debt characteristics. The conclusion is that the regular issuance of subordinated debt should be the overriding policy tool to signal and constrain bank risk-taking (i.e. direct discipline). Extending the policy to include indirect market discipline through a standardised mandatory subordinated debt requirement would impose substantial costs and should not be implemented.

Three independent models are built based on information asymmetric to analysis discipline effects brought by banks` issuance of subordinated debts. The research results offer the theory of subordinated debts discipline effects. First, an adverse selection model is built to examine banks¡¦ policies when issuing subordinated debts based on the banks¡¦ credit risks. The result shows that banks with lower credit risks are more likely to issue subordinated debts, since after the issuance, the yields of these debts are lower. When depositors observe the lower yields of the subordinated debts, they would presume that these banks have lower credit risks. As a result, banks with lower credit risks can decrease their operational costs by issuing subordinated debts. This model has demonstrated that the market can discipline banks indirectly through the issuance of subordinated debts. Second, a moral hazard model to show that issuing of subordinated debts by banks can bring direct market discipline and indirect market discipline to make their loans less risky. The direct market discipline means that the risk level of bank will be evaluated by professional investors. The investors will require that the yields accord to the banks risk. For lowering the cost of issuing subordinated debts, banks will make their loans less risky. The indirect market discipline means that the depositors would take the yields of these debts as a significant signal indicating banks risk levels. The depositors will decide to withdraw their savings when the bank signals a higher risk, and keep their saving when the bank signals a lower risk. I prove that issuing of subordinated debts by banks can bring about these two kinds of market discipline. The model has also demonstrated that if the bank supervisor can utilize the information of issuing subordinated debts effectively, they will achieve higher supervisory goal. Third, a reputation model is built to show that for reputation concerns, a bank would change its monitoring decisions if it issues subordinated debts. Reputation effect in banks is different in different scenarios. When the good banks probability of success is very high, reputation effect would induce the bad bank to start monitor it`s borrowers, and the efforts of bad banks monitoring would be increased by time. When the bad banks probability of success is very low, reputation effect would induce the good bank start monitoring its borrowers. and the efforts of bad banks monitoring would be decreases by time.

x, 99 p. / Using data for publicly listed commercial banks and bank holding companies around the world, I investigate the market discipline effect of subordinated debt on banking firm risk taking in the period 2002-2008. In addition, I examine whether this effect depends on national bank regulations and legal and institutional conditions. I provide evidence that subordinated debt has a mitigating effect on banking firm risk taking. Further, the results suggest a threshold level of national bank regulations and economic development above which subordinated debt mitigates risk taking. Overall, the evidence supports the efficacy of proposals calling for increased use of subordinated debt in banking firms. / Committee in charge: Wayne Mikkelson, Chairperson; Ekkehart Boehmer, Member; Diane Del Guercio, Member; Wesley Wilson, Outside Member

本文立基於Merton(1974)或有求償權評價法及Leland(1994)內生破產資本結構模型上，加入次順位債券的考量，建立一個連續時間資本結構分析模型，量化的資本結構可作為企業融資之決策依據。內生破產行為與風險移轉行為都是股東與債權人之間可能產生的代理問題，本文的模型不但分析這兩種代理問題，更計算其所產生的代理成本。由模擬的結果發現內生破產成本雖然不大，但是其對無風險利率及公司非槓桿價值的波動度卻是非常的敏感。在本文的模型下，只要債務契約不能重新訂立，次順位債券的發行永遠都會降低優先債務的價值。在本模型中次順位債券的風險貼水比優先債券的風險貼水高；但是本文發現不管是優先債券或是次順位債券的風險貼水，當公司接近宣告破產時，公司非槓桿價值的波動度對兩種債券風險貼水的影響出現高風險低報酬的現象，可能的解釋理由如下：當公司接近宣告破產時，公司非槓桿價值波動度的增加使破產可能性提高，進而使風險貼水減少的間接力量大於波動度增加直接使風險貼水上升的力量。 / Based on Merton(1974) and Leland(1994), we construct a continuous-time capital structure model with subordinated debt. Quantitative results may serve the guidance of financial policy of the firm. Both endogenous bankruptcy and risk shifting behaviors are the agency problems between the equityholders and debtholders. Based on our model, the agency problems is considered and the agency costs are calculated. From the result of simulation, endogenous bankruptcy agency cost is small but sensitive to the volatility of unlevered asset value of the firm. Under renegotiation-proof, the senior debt is harmed by the issuance of the subordinated debt in our model. The risk premium of the subordinate debt is higher than that of the senior debt is confirmed by the model, however, when the firm is near bankruptcy, the behavior of "high risk high return" of both debts is reversed. The reason for the junk bond behavior may be explained as the negative effect of risk premiums due to the increase of the probability of bankruptcy is higher than the positive effect of the greater risk caused by higher volatility.