企業價值與信用風險之評估 —現金流量折現基礎信用風險模型∼專案/公司債券及擔保貸款憑證評價之應用 —購併活動市場認同度影響因子之實證研究∼以美國股權交換購併案件為例

Enterprise Valuation and Credit Risk Evaluation —DCF-Based Credit RiskModel~Application in Project/Corporate Bonds & CLO —The Influencing Factors’Research of Market’s Acceptance on M&A Activities~Empirical Tests on Stock Exchange Cases

陳宗岡(Tsung-Kang Chen)

指導教授 ： 廖咸興

共同指導教授 ： 邱顯比(Shean-Bii Chiu)

國立臺灣大學/管理學院/財務金融學研究所/碩士(2004年)

https://doi.org/10.6342/NTU.2004.00730

摘要

企業價值與信用風險之評估

研究一

現金流量折現基礎信用風險模型~專案/公司債券
及擔保貸款憑證評價之應用

論文摘要

在信用風險模型的研究上，迄今已有相當多的研究者提出。依其模型假設及著重點的不同，大致可分為結構型(structural-form)與縮減型(reduced-form)兩大類信用風險模型。而在此兩大類傳統信用風險模型架構下，鮮少有直接以現金流量的隨機模型來作為衡量企業信用風險的研究基礎。這是因為一般人認為現金流量不易估計，且亦無人發展出適合的模型來描述現金流量。本研究透過對現金流量的觀察，發現其具有「一般平均反轉」、「可正可負」的隨機特性，因此本研究結合計量經濟學中系統變異性參數模型的概念，建立「時間相依之現金流量隨機模型」(Time-dependent stochastic cash flow model)，並以此模型為基礎來進行企業價值評估，進而發展出多期的信用風險模型。
利用本研究所設計的多期現金流量折現基礎信用風險模型(DCF-Based Credit Risk Model)，並結合Jarrow-Turnbull Model(1995)對有違約風險債券評價的概念，即能對專案債券與企業債券進行評價。而後，將單一企業之現金流量隨機模型推廣至多個企業，進而建立多資產之現金流量隨機模型。如此一來，對「多個企業抵押貸款債權」重新組合包裝(pool)後所衍生的資產證券化商品—CLO(Collateralized Loan Obligation)、CDO(Collateralized Debt Obligation) ，亦能依此模型進行評價。
本研究自台灣市場中選取24家樣本企業，並依現金流量隨機模型來進行評價與信用風險的驗證。結果顯示﹕評價驗證方面有2/3以上符合，信用風險驗證上有3/4以上符合。所以本模型效力初步是受到實證肯定支持的。
而在資產證券化商品CLO方面，初步結果顯示：多資產現金流量折現基礎信用風險模型會受到是否存在交叉違約(cross-default)的條件限制。未來將嘗試引入傅立葉轉換的概念與本模型相結合來處理此一問題，這也是本研究後續發展的重心。

研究二

購併活動市場認同度影響因子之實證研究∼
以美國股權交換購併案件為例

論文摘要

近年來，由於全球購併活動的盛行，因此市場對於購併活動的反應逐漸受到重視。所以本研究對於市場反應所扮演的角色及其所隱含的意義感到相當的興趣。而市場反應主要受到「預期綜效」(expected synergy, 亦即預期未來所能產生的效益)與「購併溢酬」(merger premium, 為完成購併所需付出的成本)的影響，故兩者之關係成為本研究的主要討論範疇。然而，傳統上對於兩者之關係並無提出一個整合型的分析性架構，對市場的反應亦無法作有效的衡量測度。本研究透過股價反應，建立市場預期綜效的衡量指標；並以L-G model為基礎，納入「預期綜效」與「未來的盈餘績效變動」的考量，重新建置出「企業價值」與「換股比率」之新平面，進而發展出能將預期綜效與購併溢酬整合於同一分析性架構的新模型—市場接受度模型(LMA Model)，克服了以往綜效與溢酬難以比較的限制。此外，本研究透過此分析性架構下的預期綜效與購併溢酬之關係，定義了一個對市場反應的新衡量測度：市場接受度(the level of market’s acceptance, LMA)。
因此，本研究之目的主要有二：一為探討目前市場上之購併案件是否合理。二為尋找出能解釋市場接受度的因子並討論其經濟意涵。
本研究自美國市場中選取251各股權交換購併案件(其中214件為正的購併溢酬，37件為負的購併溢酬)作為研究樣本，並依市場接受度模型來進行購併溢酬合理性的評估。結果顯示：第一，宣告期間中，有148件的購併參與者至少能維持其原有股東財富不變(落在市場接受度模型平面的第一象限)；反之，有103件的購併參與者至少有一方受到股東財富的減損。然而，在本研究所觀察的購併期間(宣告日、完成日、完成日後一月)中，至少有40%的購併案件不落於市場接受度模型的合理區間。第二，落於合理區間的148件購併案件，其主併公司自宣告日前一月即具有超越市場的績效表現。這顯示了可能存在資訊滲漏( information leakage)的現象。第三，在本研究所觀察的三個期間中，購併參與者的相對財富地位比率(相對價格比率、相對資本比率等)皆對於市場接受度有相同方向的影響。
綜合以上所述，本研究的主要貢獻為：提出市場接受度模型，將有效地將預期綜效與購併溢酬整合於同一分析性架構。這不但解決了L-G model 原有的不合理限制，對購併活動中的市場反應，也創造出一個新而有效的市場衡量測度。此外，市場接受度與相對財富地位比率的關係，亦對購併案件的合理性提供了另一方面的解釋。

關鍵字

購併 ； 信用風險 ； 市場接受度 ； 平均反轉 ； 時間相依之現金流量隨機模型 ； 企業評價

並列摘要

Enterprise Valuation and Credit Risk Evaluation

Research I

DCF-Based Credit Risk Model—Application in Project/Corporate Bonds & CLO

ABSTRACT

There are many research models about enterprise credit risk and can be roughly categorized into structural-form and reduced-form according to their assumptions and focuses. Under these two types of credit risk models, few people use stochastic cash flow model to evaluate enterprise credit risk. It’s because people generally think it difficult to estimate cash flow, and no one had ever developed an applicable model to describe the stochastic characteristics of cash flow. Through our observations of cash flow, however, we discover that the behavior of cash flow exist some stochastic characteristics, including mean-reversion and fluctuating above or under zero. We therefore use the concept of “varying coefficient model” (also called time-varying parameter models) to construct a “Time-dependent stochastic cash flow model”. And then based on this model, we further establish a “Multi-periods DCF-Based Credit Risk Model” to estimate a firm’s value and to evaluate a firm’s credit risk.
By using these two models and the concept of expected payoff ratio used to evaluate risky bonds developed by Jarrow-Turnbull Model (1995), we can now value the project bonds and corporate bonds. Furthermore, we expand the single firm’s “Multi-periods DCF-Based Credit Risk Model” to multi-firms and then construct a “Multi-assets & periods DCF-Based Credit Risk Model”. In this way, we then can valuing the securitized products including CLO(Collateralized Loan Obligation), CDO(Collateralized Debt Obligation) re-pooled from all kinds of collateral debts .
In our research, we select 24 companies as our sample and then make empirical tests of valuation and credit risk by our models. The empirical results show that our models precisely value more than two-third of sample companies and successfully assess credit risks of more than three-fourth of sample companies. These results indicate that our models are preliminarily supported by the empirical evidences.
However, in valuing the securitized products such as the CLO by using the “Multi-assets & periods DCF-Based Credit Risk Model” is limited because the model correctly can only deal with the portfolio that assets are cross-default. To handle with this limitation, the Fourier Transform Method (FTM) can be introduced into our models. Fortunately, it seems very promising that the limitation can be relieved by FTM. It is also the focus of our further study.

Research II

The Influencing Factors’ Research of Market’s Acceptance on M&A Activities—Empirical Tests on Stock Exchange Cases

ABSTRACT

Due to hot merges and acquisitions in the recent years in the world, we are interested in the roles and hidden meanings of market’s reaction toward M&A activities. However, market reaction is primarily influenced by expected synergy and merger premium, which individually represents expected benefit in the future and cost for mergers. For this reason, the relationship between expected synergy and merger premium becomes the focus of our research. But in historical researches, there doesn’t exist an integrated-analytical framework to describe the relationship. In this paper, we set a measure for expected synergy through stock price’s fluctuating. Additionally, we construct a new plane for firm’s value and exchange ratio based on L-G model by consideration of the expected synergy and the time-varying earning performance. Furthermore, we develop a LMA model which can integrate expected synergy and merger premium in the analytical framework. This model overcomes the difficulties that expected synergy is hard to compare with merger premium in the past. Moreover, we also create a new measure “the level of market’s acceptance” (LMA) through the analytical framework.
Therefore, there are two main purposes in our research. First, to examine whether the merger case is reasonable or not. Second, to find out what are the factors which can explain market’s reaction (LMA) and discuss the meanings of these significant factors.
In our research, we select 251 stock-exchange mergers (214 for positive premium, 37 for negative premium) as our sample and make judgments for the merger reasonableness by LMA model and statistical tests. As a result, we make some important conclusions: First, 148 of the 251 proposed mergers produced sufficient combined value to maintain each participant’s quadrant I status at the time of merger announcement. Alternatively, the announcement effects produced wealth losses for one or both firms in 103 of the mergers. In the merger period examined (i.e., announcement to the month following completion), at least 40% do not conform to the rationality assumption of LMA model. Second, acquiring firms of the 148 cases outperform the market since one month prior to announcement. It reveals that there may be information leakage. Third, some ratios of relative wealth status significantly influence the level of market’s acceptance in the same direction during the three periods.
To sum up, the main contributions of this study is that, by constructing LMA model and making expected synergy can be efficiently compared with merger premium. We not only solve the constraint of the original L-G model but also create a new and effective measurement of market‘s reaction toward M&A activities. In addition, we explore the relationship between LMA and relative wealth’s ratio and try to discuss its economic meanings.

並列關鍵字

Mean-reversion ； Time-dependent Stochastic Cash Fl

參考文獻

2. Darrell Duffie and Nicolae Garleanu, “Risk and Valuation of Collateralized Debt Obligations”, Working Paper, Sep 2001

3. Edward I. Altman, Andrea Resti and Andrea Sironi, “Analyzing and Explaining default Recovery Rates”, A Report Submitted to The International Swaps & Derivatives Association , 2001

5. John Hull and Alan White, “Valuation of a CDO and an nth to Default CDS Without Monte Carlo Simulation”, Working Paper, Jan 2004

6. John Hull ;AlanWhite, “Pricing Interest-Rate-Derivative Securities”,The Review of Financial Studies, 1990

8. Masaaki Kijima and Katsuya Komoribayashi, ”A Markov Chain Model for Valuing Credit Risk Derivatives”, Journal of Derivatives, Fall 1998