肖鴻民, 劉愛玲, 何 艷
常數(shù)比例投資下延遲索賠風(fēng)險(xiǎn)模型的漸近破產(chǎn)概率
肖鴻民, 劉愛玲, 何 艷
(西北師范大學(xué)數(shù)學(xué)與統(tǒng)計(jì)學(xué)院,甘肅蘭州730070)
研究一類帶投資的延遲索賠更新風(fēng)險(xiǎn)模型的漸近破產(chǎn)概率,其中允許保險(xiǎn)公司將其資產(chǎn)按常數(shù)比例投資于滿足幾何布朗運(yùn)動(dòng)的股票市場(chǎng),其余部分投資于非負(fù)利率的債券市場(chǎng),假設(shè)主索賠額和延遲索賠額序列各自為負(fù)相依同分布且屬于重尾分布族L∩D族隨機(jī)變量序列的情形下,根據(jù)Ito公式,給出保險(xiǎn)公司資產(chǎn)的表達(dá)式,最終得到有限時(shí)間的破產(chǎn)概率.
延遲索賠風(fēng)險(xiǎn)模型;負(fù)相依;L∩D族;幾何布朗運(yùn)動(dòng);破產(chǎn)概率
在現(xiàn)實(shí)生活中,保險(xiǎn)公司常常遇到延遲索賠情況,即在主索賠發(fā)生后,在某個(gè)不定時(shí)間還會(huì)產(chǎn)生由此引起的附加索賠,即延遲索賠.例如,當(dāng)一起車禍發(fā)生后,擔(dān)保人不僅要賠付車的損失,而且如果買了第三方保險(xiǎn),擔(dān)保者在隨機(jī)延遲的一段時(shí)間后還要為第三方進(jìn)行賠付,針對(duì)這類情況,H.Waters等[1]提出了帶延遲索賠的風(fēng)險(xiǎn)模型,國(guó)內(nèi)外諸多學(xué)者對(duì)該類模型產(chǎn)生了濃厚的興趣并進(jìn)行了相應(yīng)的研究.例如,K.C.Yuen等[2]研究了一類帶有復(fù)合二項(xiàng)延遲風(fēng)險(xiǎn)模型的有限時(shí)間破產(chǎn)概率,K.C.Yuen等[3]運(yùn)用鞅的方法研究了連續(xù)時(shí)間的絕對(duì)破產(chǎn)概率,J.H.Xie等[4]討論了延遲索賠風(fēng)險(xiǎn)模型的總股息的期望貼現(xiàn),C.Zhao等[5]考慮延遲更新風(fēng)險(xiǎn)模型直到破產(chǎn)的聯(lián)合密度索賠表達(dá)式,文獻(xiàn)[6-10]對(duì)延遲更新風(fēng)險(xiǎn)模型進(jìn)行了進(jìn)一步的分析.
近幾年來,隨著金融市場(chǎng)的發(fā)展以及保險(xiǎn)市場(chǎng)業(yè)務(wù)競(jìng)爭(zhēng)日趨激烈,帶投資的風(fēng)險(xiǎn)模型成為當(dāng)今金融數(shù)學(xué)研究的熱點(diǎn)問題之一.投資利潤(rùn)已經(jīng)成為現(xiàn)代保險(xiǎn)公司利潤(rùn)中重要的組成部分.Q.Tang[11]考慮了投資于常數(shù)利率下的更新風(fēng)險(xiǎn)模型的漸近破產(chǎn)概率.緊接著,保險(xiǎn)公司可以將其盈余投資于滿足幾何布朗運(yùn)動(dòng)的股票市場(chǎng)時(shí),L.Wei[12]研究了在此情形下更新風(fēng)險(xiǎn)模型的漸近破產(chǎn)概率,其中索賠服從重尾分布.C.Hipp等[13]分析了帶布朗運(yùn)動(dòng)擾動(dòng)的復(fù)合泊松風(fēng)險(xiǎn)過程下風(fēng)險(xiǎn)投資的最優(yōu)策略.進(jìn)一步地,J.Caier等[14]研究了保險(xiǎn)公司可以將其盈余的常數(shù)比例部分投資在滿足幾何布朗運(yùn)動(dòng)的股票市場(chǎng),剩余資產(chǎn)投資于常利率的國(guó)債市場(chǎng)的問題,用不同的方法得到了最終破產(chǎn)概率的相似結(jié)果.更進(jìn)一步地文獻(xiàn)[15-17]在索賠額兩兩擬漸近獨(dú)立假設(shè)下得到了破產(chǎn)概率的漸近關(guān)系.
當(dāng)索賠額服從重尾分布時(shí),前期也進(jìn)行了一些研究.文獻(xiàn)[18]考慮的是一類帶常數(shù)利息力的延遲索賠風(fēng)險(xiǎn)模型的有限時(shí)間破產(chǎn)概率,文獻(xiàn)[19]研究了負(fù)相依賠付下延遲索賠風(fēng)險(xiǎn)模型的漸近表達(dá)式.在本文中將常數(shù)比例投資模型應(yīng)用到延遲索賠風(fēng)險(xiǎn)中,當(dāng)索賠次數(shù)滿足更新過程,索賠額滿足L∩D且負(fù)相依同分布時(shí),給出常數(shù)比例投資下的資產(chǎn)表達(dá)式,得到有限時(shí)間的漸近破產(chǎn)概率.本文的結(jié)果豐富了延遲索賠風(fēng)險(xiǎn)模型的研究并對(duì)于保險(xiǎn)公司的實(shí)際運(yùn)營(yíng)具有很好的現(xiàn)實(shí)指導(dǎo)意義.
定義1 對(duì)于任一非負(fù)隨機(jī)變量X,如果矩母函數(shù)
則稱X(或稱其分布函數(shù)F)屬于重尾分布,記為K;相應(yīng)地,如果存在r>0,使得MX(r)<∞,則稱X(或稱其分布F)屬于輕尾分布,記為Kc.
下面給出幾類重要的重尾分布族及負(fù)相依的概念,它們對(duì)本文主要結(jié)果的討論是必要的.
1)稱一個(gè)分布F(x)屬于S族,如果對(duì)任意的n≥2(或等價(jià)地,對(duì)n=2),F(xiàn)滿足
2)稱一個(gè)分布F(x)屬于D族,如果對(duì)任意固定的0<y<1(或等價(jià)地,對(duì)),F(xiàn)滿足
3)稱一個(gè)分布F(x)屬于L族,如果對(duì)任意的y>0(或等價(jià)地,對(duì)y=1),分布F滿足
上述重尾分布族有如下關(guān)系:
定義2 對(duì)每個(gè)n=1,2,…,n和所有的x1,x2,…,xn,若滿足
則稱隨機(jī)變量序列{x1,x2,…,xn}下負(fù)相依,記為L(zhǎng)ND;
則稱隨機(jī)變量序列{x1,x2,…,xn}上負(fù)相依,記為UND,若上述2個(gè)條件都滿足,則稱隨機(jī)變量序列x1,x2,…,xn是負(fù)相依,記為ND.
下面介紹延遲索賠模型,U(t)為盈余過程,
其中:
1)u(u≥0)是保險(xiǎn)公司的初始資金;
2)c(c>0)為單位時(shí)間收取的保費(fèi);Si是發(fā)生第i次主索賠的時(shí)刻,Xi為發(fā)生第i次主索賠的索賠額;
3)Yi為發(fā)生第i次主索賠等待Ti時(shí)間后發(fā)生的延遲索賠額.
本文假設(shè)保險(xiǎn)公司不僅投資于利率為r的國(guó)債市場(chǎng),而且投資于股票市場(chǎng),股票價(jià)格為過程S(t),由幾何布朗運(yùn)動(dòng)定義如下:其中,a,b∈R為常數(shù),W(t)是標(biāo)準(zhǔn)布朗運(yùn)動(dòng)且獨(dú)立于U(t).
如果在時(shí)刻 t,保險(xiǎn)公式有資產(chǎn) Y(t),投資kY(t-)(0≤k≤1)在股票市場(chǎng),投資(1-k) Y(t-)在債券市場(chǎng)(利息力為r),關(guān)于資產(chǎn)Y(t)的隨機(jī)微分方程為
整理一下,資產(chǎn)過程Y(t)滿足下列跳擴(kuò)散方程
根據(jù)模型(1),由Ito公式[20],可以得出Y(t)的表達(dá)式如下,此外為了方便記
跟往常一樣,有限時(shí)間破產(chǎn)概率可以表示為
終極破產(chǎn)概率表示為
為方便起見記
本文的研究在以下假設(shè)條件下進(jìn)行:
(A1)更新記數(shù)過程N(yùn)(t)有有限的更新函數(shù)m(t)=E(N(t));
(A2)主索賠額{Xi,i=1,2,…,n}負(fù)相依于X,它們的共同分布為F(·);延遲索賠額{Yi,i=1,2,…,n}負(fù)相依于Y,共同分布為G(·);延遲賠付間隔{Ti,i=1,2,…,n}負(fù)相依于 T,共同分布為H(·);
(A3)假定隨機(jī)變量序列{Xi,i=1,2,…,n},{Yi,i=1,2,…,n},{N(t),t≥0},{Ti,i=1,2,…,n}之間是相互獨(dú)立的.
定理1 在模型(1)滿足條件(A1)~(A3)下,如果F,G∈L∩D,且珔G=o(珔F),那么直到時(shí)刻T的有限時(shí)間破產(chǎn)概率為
特別地,如果N(·)是參數(shù)為λ的齊次Poisson過程,那么
為完成定理1的證明,需如下引理.
引理1[19]在條件(A2)和(A3)下,若F,G∈L∩D,且珔G=o(珔F),則{Xi+Yi,i≥1}是負(fù)相依的,且對(duì) i≥1,M(Xi,Yi,Si,Wi)∈L∩D.
引理2[19]考慮更新風(fēng)險(xiǎn)模型(1),在定理1的條件下,對(duì)任意給定的正整數(shù)n0有
令
則
由上述,可將有限時(shí)間破產(chǎn)概率記為
則
要求的是有限時(shí)間破產(chǎn)概率,假設(shè)存在N<∞,使得0<珘C(t)<N,
所以
類似地有
如果證明了
根據(jù)L族的性質(zhì)有
下面證明
對(duì)任意正整數(shù)N,和t∈(0,T]有
先考慮I2(u,t,N),根據(jù)文獻(xiàn)[22]中3.20式,對(duì)于M∈D,正常數(shù)p、h、c和D,對(duì)任一的有
上述第六式可由文獻(xiàn)[23]得到.
由文獻(xiàn)[24]的引理3.2,可知當(dāng)N→∞時(shí),
從而對(duì)u>0,
對(duì)I1(u,t,N),利用引理2,對(duì)所有的t∈(0,T]有
而對(duì)于I11((u,t,N),顯然有
上述第三式可由文獻(xiàn)[21]推論1得到.
對(duì)于I12(u,t,N)和I2(u,t,N)的處理方法類似有
而對(duì)所有的u>0有
成立,從而(6)式成立.
注1 定理1表明,帶投資的延遲索賠且主索賠計(jì)數(shù)過程為更新過程的風(fēng)險(xiǎn)模型中,當(dāng)主索賠與延遲索賠均屬于重尾分布L∩D族時(shí),保險(xiǎn)公司的極端行為與索賠額分布的尾部特性有關(guān).這一結(jié)果對(duì)保險(xiǎn)公司的平穩(wěn)運(yùn)行和風(fēng)險(xiǎn)估計(jì)具有借鑒意義.
[1]WATERS H,PAPATRIANDAFYLOU A.Ruin probabilities allowing for delay in claims settlement[J].Insurance:Math Econ,1985,4:113-122.
[2]YUEN K C,GUO J Y.Ruin probabilities for time-correlated claims in the compound binomial model[J].Insurance:Math Econ,2001,29(1),47-57.
[3]YUEN K C,GUO J Y,NG K W.On ultimate ruin in a delayed-claims risk model[J].J Appl Probability,2005,42:163-174.
[4]XIE J H,ZOU W.Expected present value of total dividends in a delayed claims risk model under stochastic interest rates[J].Insurance:Math Econ,2010,46:415-422.
[5]ZHAO C,ZHANG C.Joint density of the number of claims until ruin and the time to ruin in the delayed renewal risk model with Erlang(n)claims[J].Compu Appl Math,2013,244(1):102-144.
[6]DENG C,ZHOU J,DENG Y.The Gerber-Shiu discounted penalty function in a delayed renewal risk model with multi-layer dividend strategy[J].Statistics and Probability Letters,2012,82:1648-1656.
[7]ZOU W,XIE J.On the Gerber-Shiu discounted penalty function in a risk model with delayed claims[J].Korean Statistical Society,2012,41:387-397.
[8]CHEN Y,HUANG Y,ZHANG W.Asymptotic ruin probabilities for proportional investment under interest force with dominatedly-varying-tailed claims[J].Korean Statistical Society,2012,41:87-95.
[9]YANG Y,YUEN K C.Finite-time and infinite-time ruin probabilities in atwo-dimensional delayed renewal risk model with Sarmanov dependent claims[J].J Math Anal Appl,2016,442(2):600-626.
[10]SO-YEUN K,GORDON E W.On the analysis of ruin-related quantities in the delayed renewal risk model[J].Insurance: Math Econ,2016,66:77-85.
[11]TANG Q.The finite time ruin probability of the compound Poisson model with constant interest force[J].Appl Probab,2005,42:608-619.
[12]WEI L.Ruin probability of the renewal model with risky investment and large claims[J].Science in China(Math),2009,A52(7):1539-1545.
[13]HIPP C,PLUM M.Optimal investment for insurers[J].Insurance:Math Econ,2000,27:215-228.
[14]GAIER J,GRANDITS P.Ruin probabilities and investment under interest force in the presence of regularly varying tails[J].Scand Actuarial J,2004,4:256-278.
[15]CHEN Y,NG K W.The ruin probability of the renewal model with constant interest force and negatively dependent heavytailed claims[J].Insurance:Math Econ,2007,40:415-423.
[16]YI L,CHEN Y,SU C.Approximation of the tail probability of randomly weighted sums of dependent random variables with dominated variation[J].J Math Anal Appl,2011,376(1):365-372.
[17]肖鴻民,何艷.常數(shù)比例投資下基于進(jìn)入過程風(fēng)險(xiǎn)模型的漸進(jìn)破產(chǎn)概率[J].河南師范大學(xué)學(xué)報(bào)(自然科學(xué)版),2015,43(2):14-19.
[18]肖鴻民,李紅.重尾賠付下帶常數(shù)利息力的延遲索賠風(fēng)險(xiǎn)模型的破產(chǎn)概率[J].西北師范大學(xué)學(xué)報(bào)(自然科學(xué)版),2011,47(6):17-20.
[19]肖鴻民,李紅.負(fù)相依賠付下延遲風(fēng)險(xiǎn)模型的破產(chǎn)概率[J].蘭州大學(xué)學(xué)報(bào)(自然科學(xué)版),2012,48(3),118-122.
[20]HE S,WANG J,YAN J.Semimartingale Theory and Stochastic Calculus[M].Beijing:Science Press,1992.
[21]CLINE D B H,SAMORODNITSKY G.Subexponentiality of the product ofindependent random variables[J].Stochastic Processes Appl,1994,49(1):75-98.
[22]TANG Q,TSITSIASHVILI G.Precise estimates for the ruin probability in finite horizon in a discrete-time model with heavytailed insurance and financial risks[J].Stochastic Processes and Their Applications,2003,108(2):299-325.
[23]EMBRECHTS P,KIUPPELBERG C,MIKOSCH T.Modeling Extremal Events for Insurance and Finance[M].Berlin:Springer-Verlag,1997.
[24]HAO X M,TANG Q H.A uniform asymptotic estimate for discounted aggregate claims with subexponential tails[J].Insurance: Math Econ,2008,43(1):116-120.
Asymptotic Ruin Probabilities with Delayed-claims Risk Model under Proportional Investment
XIAO Hongmin, LIU Ailing, HE Yan
(College of Mathematics and Statistics,Northwest Normal University,Lanzhou 730070,Gansu)
The asymptotic behavior of ruin probabilities is investigated in a renewal risk model for delayed claims,in which the insurance company is allowed to invest a constant fraction of its wealth in a stock market which is described by a geometric Brownian motion and the remaining wealth in a bond with nonnegative interest force.Under the assumptions that the sequences of the main and delayed claims are negatively dependent random varies with a common distribution and that the claim sizes belong to the heavy-tailed distribution class L∩D,the expression of the wealth process is derived by the Ito formula,and the finite-time ruin probabilities are obtained.
delayed-claims risk model;negatively dependent;class L∩D;geometric Brownian motion;ruin probability
O211.4
A
1001-8395(2016)05-0665-06
10.3969/j.issn.1001-8395.2016.05.009
(編輯 鄭月蓉)
2016-03-01
國(guó)家自然科學(xué)基金(71261023)
肖鴻民(1967—),女,教授,主要從事概率極限理論與保險(xiǎn)數(shù)學(xué)的研究,E-mail:xiaohm9@126.com
2010 MSC:91B30