劉超??
摘 要 在理論上通過推導(dǎo)首次得出了BlackLitterman模型(BL模型)最優(yōu)權(quán)重與信心水平的公式.在各資產(chǎn)收益不相關(guān)及單一絕對觀點的假設(shè)下,得出各資產(chǎn)的BL模型最優(yōu)權(quán)重與信心水平的簡化表達(dá)式.借助于此,還對信心水平與最優(yōu)權(quán)重公式的進(jìn)一步理論分析,并以光大證券的“烏龍指”做實證,詳細(xì)分析投資者在沒有市場觀點、擁有內(nèi)幕信息、以及信心水平在某范圍變化時,其所持各投資品權(quán)重的特點.
關(guān)鍵詞 資產(chǎn)配置;BlackLitterman 模型;信心水平;內(nèi)幕交易;烏龍指
中圖分類號 F830.91 文獻(xiàn)標(biāo)識碼 A
The Further Theoretically Derivation and Empirical
Analysis of the BlackLitterman Portfolio Model
LIU Chao
(School of Mathematical Sciences, Peking University, Beijing 100871, China)
Abstract This paper deduced the function of optimal weight and confidence levels, and then simplified this formula under the assumptions that all the assets to be invested are uncorrelated, and the investor has only one absolute view. Based on this, this paper also analysed the formula of the confidence levels and optimal weight. Finally, this paper figured out how the investor would establish his portfolio and how much his profit would be (including the insidertrading case), according to his view, and the confidence level of this view. The ‘Fat finger trading error, caused by Everbright Securities Company Limited on August 16, 2013, was taken as an example to analyse.
Key words asset allocation; BlackLitterman Model; confidence level; insider trading; Fat finger trading error
1 引 言
BlackLitterman投資組合模型由Fisher Black和Robert Litterman[1]在1991年首先提出,簡稱BL模型,其理論在Black& Litterman[2]得到詳盡闡述.該模型從市場數(shù)據(jù)出發(fā),結(jié)合投資者觀點,通過變換投資者效用函數(shù)、結(jié)合Markowitz[3]的均值方差理論和貝葉斯分析等導(dǎo)出最優(yōu)權(quán)重.該模型在市場的一致預(yù)期和投資者自己的觀點間達(dá)到一種均衡,近20年被國內(nèi)外多家著名量化投資機(jī)構(gòu)使用,同時其理論也得到發(fā)展與完善.此后,He與Litterman[4]給出了該模型的變量在理論上的意義,Idzorek[5]率先將信心水平引入BL模型,并將其理論進(jìn)一步發(fā)展[6].此外Mankert[7],溫琪[8]等也對該模型進(jìn)行了理論完善和實證補(bǔ)充.
BL模型是由市場數(shù)據(jù)與投資者觀點相綜合而得到最優(yōu)權(quán)重.信心水平是BL模型中的一個重要的變量,信心水平越高,最優(yōu)權(quán)重越接近投資者個人的觀點;信心水平越低,最優(yōu)權(quán)重越接近市場權(quán)重.
通過參考國內(nèi)外多篇文獻(xiàn),發(fā)現(xiàn)很少有論文研究BL模型與信心水平的關(guān)系.實證論文中,2011 年賈慧[9]用ARMAGARCH 模型預(yù)測BL模型的投資者觀點收益向量,同時較早地在國內(nèi)結(jié)合信心水平進(jìn)行BL模型實證.該文結(jié)合10%、50%、80%三種信心水平進(jìn)行分析.2012年,王楠溪[10]用ARMAEGARCH模型預(yù)測BL模型的投資者觀點收益向量,同時結(jié)合20%、50%、90%三種信心水平進(jìn)行分析.這些極少數(shù)研究過BL模型資產(chǎn)收益與信心水平關(guān)系的實證論文,其選取的信心水平也都很少,甚至還不對稱.Idzorek等的理論性論文中也沒有給出最優(yōu)權(quán)重與信心水平的直接關(guān)系式.
本文在理論上通過推導(dǎo)得出了最優(yōu)權(quán)重與信心水平的直接關(guān)系式.在各資產(chǎn)收益不相關(guān)假設(shè)下,對具有單一絕對觀點的投資者,本文將上述關(guān)系式化簡,得出各資產(chǎn)的BL模型最優(yōu)權(quán)重與信心水平的簡化表達(dá)式.借助于此,本文詳細(xì)分析投資者在沒有市場觀點、擁有內(nèi)幕信息時,其所持各投資品權(quán)重的特點.并結(jié)合“烏龍指”作為實例,研究0~100%的所有信心水平,畫出信心水平與投資品權(quán)重和投資收益率的曲線,在各資產(chǎn)收益不相關(guān)假設(shè)下,得出投資者投資組合最優(yōu)權(quán)重與信心水平的關(guān)系式.
內(nèi)幕消息泛指所有非公開的公司消息,通常只有部份內(nèi)部人士能獲得.投資者對某一特定資產(chǎn)有內(nèi)幕消息時,本文求出了他投資該資產(chǎn)的最優(yōu)權(quán)重確切表達(dá)式.若該投資者認(rèn)為其超額收益率為正數(shù),則會較大幅度地做多該資產(chǎn);若該投資者認(rèn)為其超額收益率為負(fù)數(shù),則會較大幅度地做空該資產(chǎn).同時,投資者對沒有掌握內(nèi)幕消息的其他資產(chǎn)進(jìn)行投資的頭寸不變.
6 結(jié) 論
本文在理論上通過推導(dǎo)得出了最優(yōu)權(quán)重與信心水平的直接關(guān)系式.在各資產(chǎn)收益不相關(guān)假設(shè)下,對具有單一絕對觀點的投資者,本文將上述關(guān)系式化簡,得出各資產(chǎn)的BL模型最優(yōu)權(quán)重與信心水平的簡化表達(dá)式.借助于此,通過理論和實證相結(jié)合,本文詳細(xì)分析投資者在不同信心水平下,其所持各投資品權(quán)重的特點.endprint
參考文獻(xiàn)
[1] F BLACK, R LITTERMAN. Asset alloeation: Combining investor views with market equilibrium [J]. The Journal of Fixed Income, 1991, 1(2): 7-18.
[2] F BLACK, R LITTERMAN. Global portfolio optimization [J]. The Financial Analysts Journal,1992, 48(5): 28-43.
[3] H MARKOWITZ. Portfolio selection [J]. Journal of Finance, 1952, 7(1): 77-91.
[4] Guangliang HE, R LITTERMAN. The intuition behind BlackLitterman model portfolios[R]. Investment Management Research, New York:Goldman Sachs,1999.
[5] T IDZOREK. A stepbystep guide to the BlackLitterman model[R]. Chicago, Illlinois: Ibbotson Associates,2002.
[6] T IDZOREK. A stepbystep guide to the BlackLitterman model: Incorporating userspecified confidence levels [R].Chicago, Illlinois: Ibbotson Associates,2005.
[7] C MANKERT. The BlackLitterman modelmathematical and behavioral finance approaches towards its use in practice[D]. Sweden:Royal Institute of Technology, School of Industral Engineering and Management.Department of Industrial Economics and Management, 2006.
[8] 溫琪,陳敏,梁斌. 基于BlackLitterman框架的資產(chǎn)配置策略研究[J]. 數(shù)理統(tǒng)計與管理, 2011, 30(4): 741-751.
[9] 賈慧. BlackLitterman 模型在中國股票市場資產(chǎn)配置中的應(yīng)用研究[D].西安:西北大學(xué)經(jīng)濟(jì)學(xué)院數(shù)量經(jīng)濟(jì)系,2011.
[10]王楠溪. BlackLitterman模型在中國市場中的應(yīng)用——考慮非對稱投資者觀點的擴(kuò)展模型[D]. 北京:北京大學(xué)金融數(shù)學(xué)系,2012.
[11]S C NORRBIN, K C CHAN, P LAI. Are stock and bond prices collinear in the long run[J]. International Review of Economics & Finance, Elsevier, 1997, 6(2) : 193-201.
[12]常麗莉. 債券市場與股票市場相關(guān)性問題及對策研究[D]. 重慶:重慶大學(xué)產(chǎn)業(yè)經(jīng)濟(jì)學(xué)系,2009.
[13]Z BODIE, A KANE, A J MARCUS. Investment[M]. (8th edition)Beijing: China Machine Press, 2011.endprint
參考文獻(xiàn)
[1] F BLACK, R LITTERMAN. Asset alloeation: Combining investor views with market equilibrium [J]. The Journal of Fixed Income, 1991, 1(2): 7-18.
[2] F BLACK, R LITTERMAN. Global portfolio optimization [J]. The Financial Analysts Journal,1992, 48(5): 28-43.
[3] H MARKOWITZ. Portfolio selection [J]. Journal of Finance, 1952, 7(1): 77-91.
[4] Guangliang HE, R LITTERMAN. The intuition behind BlackLitterman model portfolios[R]. Investment Management Research, New York:Goldman Sachs,1999.
[5] T IDZOREK. A stepbystep guide to the BlackLitterman model[R]. Chicago, Illlinois: Ibbotson Associates,2002.
[6] T IDZOREK. A stepbystep guide to the BlackLitterman model: Incorporating userspecified confidence levels [R].Chicago, Illlinois: Ibbotson Associates,2005.
[7] C MANKERT. The BlackLitterman modelmathematical and behavioral finance approaches towards its use in practice[D]. Sweden:Royal Institute of Technology, School of Industral Engineering and Management.Department of Industrial Economics and Management, 2006.
[8] 溫琪,陳敏,梁斌. 基于BlackLitterman框架的資產(chǎn)配置策略研究[J]. 數(shù)理統(tǒng)計與管理, 2011, 30(4): 741-751.
[9] 賈慧. BlackLitterman 模型在中國股票市場資產(chǎn)配置中的應(yīng)用研究[D].西安:西北大學(xué)經(jīng)濟(jì)學(xué)院數(shù)量經(jīng)濟(jì)系,2011.
[10]王楠溪. BlackLitterman模型在中國市場中的應(yīng)用——考慮非對稱投資者觀點的擴(kuò)展模型[D]. 北京:北京大學(xué)金融數(shù)學(xué)系,2012.
[11]S C NORRBIN, K C CHAN, P LAI. Are stock and bond prices collinear in the long run[J]. International Review of Economics & Finance, Elsevier, 1997, 6(2) : 193-201.
[12]常麗莉. 債券市場與股票市場相關(guān)性問題及對策研究[D]. 重慶:重慶大學(xué)產(chǎn)業(yè)經(jīng)濟(jì)學(xué)系,2009.
[13]Z BODIE, A KANE, A J MARCUS. Investment[M]. (8th edition)Beijing: China Machine Press, 2011.endprint
參考文獻(xiàn)
[1] F BLACK, R LITTERMAN. Asset alloeation: Combining investor views with market equilibrium [J]. The Journal of Fixed Income, 1991, 1(2): 7-18.
[2] F BLACK, R LITTERMAN. Global portfolio optimization [J]. The Financial Analysts Journal,1992, 48(5): 28-43.
[3] H MARKOWITZ. Portfolio selection [J]. Journal of Finance, 1952, 7(1): 77-91.
[4] Guangliang HE, R LITTERMAN. The intuition behind BlackLitterman model portfolios[R]. Investment Management Research, New York:Goldman Sachs,1999.
[5] T IDZOREK. A stepbystep guide to the BlackLitterman model[R]. Chicago, Illlinois: Ibbotson Associates,2002.
[6] T IDZOREK. A stepbystep guide to the BlackLitterman model: Incorporating userspecified confidence levels [R].Chicago, Illlinois: Ibbotson Associates,2005.
[7] C MANKERT. The BlackLitterman modelmathematical and behavioral finance approaches towards its use in practice[D]. Sweden:Royal Institute of Technology, School of Industral Engineering and Management.Department of Industrial Economics and Management, 2006.
[8] 溫琪,陳敏,梁斌. 基于BlackLitterman框架的資產(chǎn)配置策略研究[J]. 數(shù)理統(tǒng)計與管理, 2011, 30(4): 741-751.
[9] 賈慧. BlackLitterman 模型在中國股票市場資產(chǎn)配置中的應(yīng)用研究[D].西安:西北大學(xué)經(jīng)濟(jì)學(xué)院數(shù)量經(jīng)濟(jì)系,2011.
[10]王楠溪. BlackLitterman模型在中國市場中的應(yīng)用——考慮非對稱投資者觀點的擴(kuò)展模型[D]. 北京:北京大學(xué)金融數(shù)學(xué)系,2012.
[11]S C NORRBIN, K C CHAN, P LAI. Are stock and bond prices collinear in the long run[J]. International Review of Economics & Finance, Elsevier, 1997, 6(2) : 193-201.
[12]常麗莉. 債券市場與股票市場相關(guān)性問題及對策研究[D]. 重慶:重慶大學(xué)產(chǎn)業(yè)經(jīng)濟(jì)學(xué)系,2009.
[13]Z BODIE, A KANE, A J MARCUS. Investment[M]. (8th edition)Beijing: China Machine Press, 2011.endprint