Based on Leland (1994), the paper tries to examine the changes in probability of bankruptcy when direct and indirect costs of bankruptcy are considered. Besides, the paper tries to study the changes in probability of bankruptcy when agents have different goals, and explore the changes in bankruptcy costs and agency costs. The paper finds that, by taking into account of indirect cost of bankruptcy, the probability of bankruptcy tends to increase when agents try to maximize the value of shareholders, but tends to decrease when agents try to maximize the value of the firm. The probability of bankruptcy also varies by the changes of the parameters in different goals of agency models. Three major findings in the paper are: (1) when agents try to maximize the value of firm, the optimal probability of bankruptcy is close to zero; (2) marginal risk premium incurred by additional debts have to be considered in maximizing the value of firm; (3) the financing decision made by agents are affected by indirect costs of bankruptcy as they will increase the optimal probability of bankruptcy and marginal risk premium of debt in comparison with the case of direct costs being considered only.

Inledning: Modigliani och Miller-teoremet betraktas som en milstolpe och ett paradigmskifte inom Corporate Finance. Forskningen bakom teoremet har gett upphov till den kontroversiella slutsatsen att ett företags kapitalstruktur är irrelevant på en jämviktsmarknad, utan beskattning. När beskattning däremot förekommer, ökar företagsvärdet motsvarande den avdragsgilla skatten på räntekostnader vid belåning. Teoremet har dock inte varit utan kritik. Flera forskare har undersökt teoremet och kritiserat teoremets antaganden. Syfte: Syftet med studien är att undersöka Modigliani och Miller-teoremet i avseende av kapitalstrukturens inverkan på företags aktiekurser. Tidigare forskning: Modigliani och Miller-teoremet med dess antaganden presenteras inledningsvis som studiens grundläggande teori. Därefter presenteras agentteorin, signaleringsteorin, trade-off-teori och effektiva marknadshypotesen tillsammans med tidigare forskningsstudier om teoremet och antagandena. Metod: Studien tillämpar en kvantitativ metod, med tre regressionsmodeller. Det slumpmässiga urvalet består av tio företag inom industribranschen som är registrerade på Stockholmsbörsen. Urvalet består av paneldata från företagen under perioden 2005 till 2012. Studiens empiri består av historiska aktiekurser och årsredovisningar. Resultat: Skuldsättningsgraden visar en låg och icke signifikant korrelation med aktiekurserna för de undersökta företagen, i samtliga tre regressionsmodeller. Skatteskölden visar en högre och signifikant korrelation med aktiekursen, medan kontrollvariabeln vinst per aktie EPS visar den högsta korrelationen med aktiekursen, som responsvariabel. Slutsats: Studiens resultat visar inget empiriskt stöd för Modigliani och Miller-teoremet i dess helhet. Förklaringen till resultaten kan vara att antagandena inte tillfredsställs inom empirin. Agentteori, signaleringsteori, trade-off-teori och effektiva marknadshypotesen kan stå som förklaringar till studiens resultat. / Introduction: Modigliani and Miller theorem is regarded as a milestone and a paradigm shift in Corporate Finance. The research behind the theorem has given rise to the controversial conclusion that a company's capital structure is irrelevant in an equilibrium market, without taxation. In contrast, when taxation occurs firm value will increase equivalent to the deductible tax on interest expense when debt is issued. The theorem has not been without criticism. Several researchers have examined the theorem and criticised its assumptions. Purpose: The purpose of the study is to investigate the Modigliani and Miller theorem in respect of capital structure’s impact on companies' share prices. Literature Review: The Modigliani and Miller theorem with its assumptions is initially presented as the study’s main theorem. The presentation of the theorem is followed by a description of agency theory, signalling theory, trade-off theory and the efficient market hypothesis, along with previous studies on the theorem and its assumptions. Research Methodology: The study applies a quantitative approach, with three regression models. The random sample consists of ten companies in the industrial sector, that are listed on the Stockholm Stock Exchange. The sample is based on panel data of the companies during the period 2005 to 2012. The study's empirical data consists of historical stock prices and annual reports. Empirical Results: The leverage ratio, i.e. debt to equity ratio, indicates a low and non-signif-icant correlation with the stock prices of the examined companies, in all three regression models. The tax shield indicates a higher and significant correlation with the stock price, while the control variable earnings per share EPS indicates the highest correlation with the stock prices, as the response variable. Conclusions: The study’s results show no empirical support for the Modigliani-Miller theorem in its entirety. The explanation for the findings may be that the assumptions are not satisfied in the empirical data. Agency theory, signalling theory, trade-off theory and the efficient market hypothesis may serve as explanations of the study’s results.

本文建立一公司稅、累進個人稅、非債稅盾和破產可能之不確定經濟模型，探討公司的投資理財決策行為。和以前文獻最大的不同點，在於公司風險債利率水準明顯且完全的內生化處理，此內生化利率不僅受公司舉債額度的財務決策影響，也會反應公司現行投資或資本存量的實質決策，模型中的財務和實質決策具有完全互動之特性。 累進個人稅下，只要存在一組投資人間的邊際稅率結構，能排除套稅資產配置，則市場無套稅價格和競爭均衡便存在。此一結果，使本文模型中的資產評價得以建立於競爭財務市場均衡的基礎。市場均衡下，公司債利率會隨公司舉債額度而同向、投資額度而反向調整，此時股東與債權人間的代理成本問題完全消失、市場具有公平保護債權人的機能。如果公司債利率外生而無法依市場均衡評價，則由於代理成本問題存在，使公司最適投資相對低於、舉債額度則高於利率內生之最適水準。另外，傳統一些稅制分析法的資本結構決定理論，只是本文之特例。 稅盾變化，如投資減免率或定額減免等，對公司最適投資和舉債的影響，除以前文獻之結果外，還包含一利率回饋效果，後者可能改變以前文獻之結論。例如，本文發現獎勵投資之折舊率或投資扣抵的政策美意，可能反而造成抑制投資的結果，本文這部分之結論，或可提供有關公司資本結構決定之實證研究新的詮釋，另外在稅制設計上亦具政策上的含意。