Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/53860
標題: General Equilibrium and Risk Neutral Framework to Estimate Systematic Risk, Nonsystematic Risk, and Cost of Equity
一般均衡風險中立評價系統風險、非系統風險、與權益資金成本之研究
作者: 林丙輝
關鍵字: 管理科學;基礎研究
摘要: 
本計畫預計應用風險中立評價關係,考慮以 g distribution 或是其它的常態轉換(transformed-normal)分配來描述系統性與非系統性風險變數之分配,進而發展一個更一般化的選擇權評價模型來估計前瞻性(forward-looking)的資產系統性與非系統性風險與資金成本等市場資訊。Lin, Paxson, Wang, and Kuo (2010)以風險中立評價關係(Risk Neutral ValuationRelationship, RNVR)架構為基礎,整合資本市場模型(Capital Asset Pricing Model, CAPM)進來,發展出一個以股票市場指數波動度、個股beta 值與個股非系統風險之波動度為基礎的選擇權評價模型。此模型不但可以解釋個股選擇權的volatility smile,還可以讓投資人以此為基礎來估計個股的系統性與非系統性風險。該研究運用兩階段的校正方法,首先用股票指數選擇權的市價校正出股票市場指數波動度,再用個股選擇權的市價來校正出個股之個股beta 值與個股非系統風險之波動度。Lin, Paxson, Wang, and Kuo (2010)雖然能成功地解決的個股選擇權的volatility smile與beta 估計值的問題,但卻無法符合股票指數選擇權的volatility smile 之現象。因此本此計畫的主要目的,即是要假設股票指數服從g distribution 或是其他常態轉換分配,此因g distribution 為一種常態轉換分配,比一般的對數常態分配擁有更負的skewness 來解釋股票指數選擇權的volatility smile 之現象。而根據Câmara (2003),只要標的變數服從常態轉換分配,都可從RNVR 架構中,導出無風險偏好(preference-free)的選擇權評價模型。本計畫預計分兩年執行:第一年部分,預計在之前結合了 RNVR 與CAPM 的架構中,考慮g distribution 或是其他的轉換常態分配,例如displaced lognormal distribution、negatively skew lognormal distribution、SU distribution、甚至inverse Gaussian distribution與gamma distribution,以導出一個新的選擇權評價模型,能更適切的描述股票指數選擇權與個股選擇權的volatility smile 現象。在此transformed-normal 分配假設下,CAPM 模型除了考慮beta (即covariance 的效果)外,還需將coskewness 與cokurtosis 的效果考慮進去,以確保整個模型的一致性。本計畫的第二年部分為廣泛且深入的實證研究,將運用其他forward-looking beta 的模型,並與本計畫所發展的模型比較。除了希望對股票指數選擇權與個股選擇權的volatility smile 現象做最好的解釋之外, 也能比較並驗證這些模型所得出的forward-looking beta,是否對於未來期間的beta 與權益的資金成本(cost of equity, COE),提供更好的預測與解釋能力。

In this research project, we will adopt the risk-neutral valuation relationship (RNVR) andconsider the g distribution or other transformed-normal distributions for the systematic andnonsystematic components of equity pricing, to describe an option pricing model which canbe used to obtain the forward-looking beta as well as the nonsystematic risk of equity and itscost of capital in the market.Lin, Paxson, Wang, and Kuo (2010) utilize the framework of RNVR and integrate thestandard capital asset pricing model (CAPM) to derive an option valuation model, whichincludes the market volatility, equity beta and its nonsystematic risk in the model. With thismodel, the equity option implied volatility smile can be properly accounted, and one can usethis model to estimate forward-looking beta and nonsystematic risk of equity by calibratingthe market information.Although Lin, Paxson, Wang, and Kuo (2010) can consider the effect of equity optionimplied volatility smile and estimate the forward-looking equity beta, however it could notadapt itself to the effect of index option implied volatility smile, since their model convergesto the Black-Scholes formula in the case of index options. The aim of this project is toassume the stock index following the g distribution or other transformed-normal distributions,in order to overcome the disadvantages. The g distribution is one of the transformed-normaldistributions with more negative skewness than the (log) normal distributions, to describe thevolatility smile effect. According to Cmara (2003), as long as the stock price returns followthe family of transformed-normal distributions, one can utilize the RNVR to derive thepreference-free option pricing model.The project spans into two years: In the first year we will integrate the RNVR andCAPM, considering the g distribution or other transformed-normal distributions, such as thedisplaced lognormal distribution, the negatively skew lognormal distribution, the SUdistribution, the inverse Gaussian distribution, or the gamma distribution, to derive a newoption valuation model in order to adapt both the equity and the index options impliedvolatility smile effects. With these assumptions, the CAPM should include the coskewnessand cokurtosis effects, in addition to the covariance effect, the beta, in order to assure theconsistency of the theoretical model.In the second year of this project, we will conduct extensive empirical studies using themodel developed in the first year, and other models of forward-looking beta to compare theresults of the forward-looking beta and the historical beta, as well as the cost of capital forasset pricing.
URI: http://hdl.handle.net/11455/53860
其他識別: NSC100-2410-H005-013-MY2
Appears in Collections:財務金融學系所

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.