Zongxia Liang and Ming Ma. Department of Mathematical Sciences, Tsinghua University, Beijing , China

Size: px
Start display at page:

Download "Zongxia Liang and Ming Ma. Department of Mathematical Sciences, Tsinghua University, Beijing , China"

Transcription

1 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE IN INCOMPLETE MARKETS Zongxia Liang and Ming Ma Deparmen of Mahemaical Science, Tinghua Univeriy, Beijing 184, China Abrac. In hi paper, we udy he imple invemen problem of an incomplee marke, where he expeced uiliy of erminal wealh i maximized under a ochaic inere rae and a ock invemen. Unle iuing a rolling bond o hedge he inere rae rik, he incompleene caue he following difficul: he variey of equivalen maringale meaure confue he radiional maringale approach. Thu we aim a clarifying he maringale approach in he incomplee marke, whoe eence i finding he opimal neural meaure over all admiible meaure, which i exacly a minimax meaure. The five-ep maringale approach i ummarized by he coniency heorem and he ufficien condiion. Indeed, he coniency heorem allow u o uing maringale approach dynamically. Baed on hee udie, he opimal raegie are derived explicily under CRRA and CARA uiliie. Finally, numerical illuraion reveal he behavior of opimal premium, paricularly, he ime and capial effec on he behavior. MSC(1): 91G1, 91G3, 93E. Keyword: Maringale approach; Incomplee marke; Opimal neural meaure; Minimax meaure; Premium ad curve; Sochaic inere rae. 1. Inroducion Wih he variey of rik in financial marke, more and more derivaive are iuing o hedge rik. Bu here are only everal developed derivaive due o he complexiy of rik or he policy uperviion. Thu he compleene of marke doe no hold and ome incomplee marke have o be conidered. We inveigae he imple opimizaion ha he expeced uiliy of erminal wealh i uppoed o be maximized under a ochaic inere rae and a ock invemen. The only conrollable hing i he fracion of he oal wealh inveed in ock. I i well-known ha if a rolling bond i iued o hedge he inere rae rik, hen he marke urn o be complee and here are wo developed Correponding auhor. zliang@mah.inghua.edu.cn(z.liang), maming9@gmail.com(m.ma) 1

2 ZONGXIA LIANG AND MING MA mehod o ue. One i he dynamic programming, which i propoed by Meron[16](1971) and uppored by he heory of Hamilon-Jacobi- Bellman equaion. The oher one i he maringale approach developed by Cox and Huang (1989)[3] baed on he heory of Lagrange muliplier, which uccefully olve an opimal conumpion-porfolio problem under hyperbolic abolue rik averion uiliie. Thee wo mehod boh have developed heorie for ime-conien problem in complee marke. Bu in incomplee marke, he HJB equaion derived from dynamic programming come o a higher-dimenional parially differenial equaion, whoe explici oluion can hardly be found. The concep of vicoiy oluion may be ued o olve a HJB equaion in an incomplee marke(cf. Duffie e al.(1997)[4]). The core of maringale approach i rewriing he problem o be adaped o he heory of Lagrange muliplier by an equivalen maringale meaure. Bu here are infinie equivalen maringale meaure in an incomplee marke and i i no clear which one hould be choen. The main arge of hi paper i o clarify he maringale approach in incomplee marke, by which we pick ou an equivalen maringale meaure called opimal neural meaure. The udy in Secion 3 ell u ha uing maringale approach in incomplee marke i equal o finding he opimal neural meaure from a lo of admiible meaure over Q defined by (3.1) below. If he inere rae i conan, hi kind of opic ha been widely dicued a he opimal crierion for he e of all maringale meaure in an incomplee marke. According o prior reearche, people uually udy hi opic baed on he following wo perpecive o find he appropriae candidae meaure. The fir perpecive i o define a diance from maringale meaure o he original meaure, which i uppoed o be minimized. For inance, Cizár(1975)[] ue he I-divergence or Kullback-Leibler informaion a a diance o find he I-projecion of he original meaure on ome meaure e. Delbaen and Schachermayer(1996)[5] inroduce he variance-opimal maringale meaure which minimize he L norm of deniy dq. The relaive enropy of Q wih repec o P i defined by dp Frielli()[6] and lead u o he minimal enropy maringale meaure. Along hee hough, he minimax maringale meaure i preened

3 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 3 by Goll & Rüchendorf(1)[8] and Bellini & Frielli()[1], which exacly minimize he f-divergence diance. Grandi()[7] generalize he minimal maringale meaure for quai-lef-coninue proce wih bounded jump via he definiion of minimal Hellinger maringale meaure. The econd perpecive i o conider a family of pricing or opimizaion problem wih a given uiliy, which correpond o a family of admiible meaure. For inveor wih differen uiliie, i i reaonable ha he choen meaure are differen. Karaza e al.(1991)[11] firly dicu he uiliy maximizaion and i dual problem for geing an opimal invemen raegy in an incomplee marke, where hey elec an opimal meaure by vanihing hoe ficiiou ock. Kallen(1999)[1] give a crieria of meaure o hedge rik in incomplee marke by maximizing expeced local uiliy. Indeed, hee kind of choen meaure have cloed relaionhip wih above minimax meaure, which ha been revealed by Goll & Rchendorf(1)[8]. Under he framework of opimal crierion, he opimal neural meaure in hi paper i a differen minimax meaure, which maximize he expeced uiliy of wealh bu no he deflaed wealh. If he inere rae i conan, our problem i imilar o he reearch of Karaza e al.(1991)[11]. If in addiion he uiliy i CARA, he expeced uiliy of wealh i proporional o he expeced uiliy of deflaed wealh, hu he udy of Goll & Rüchendorf(1)[8] operae. Bu conidering a ochaic inere rae, he problem become complex and non-linear and hu ome inereing effec appear. More preciely, he rik premium correponding o he opimal neural meaure may be ochaic and relaed o he inananeou ae a well a mauriie. Becaue of hi, he ime and capial boh have ignifican impac on he behavior of rik premium. The re of hi paper i organized a follow. Secion formulae he original opimizaion problem, where he dynamic of ae and he admiible conrol e are rigorouly defined. In Secion 3, we eablih he five-ep maringale approach in incomplee marke uppored by ome heorem. The expanded pace and he opimal neural meaure are preened for olving he original opimizaion. Baed on he five-ep

4 4 ZONGXIA LIANG AND MING MA approach, we calculae wo cae in Secion 4 where CRRA and CARA uiliie are conidered. The numerical illuraion of wo cae are boh diplayed in Secion 5, according o which we reveal he behavior of opimal premium. Paricularly, we ummarize he ime and capial effec on he behavior.. Problem decripion The invemen problem we conidered in hi paper i maximizing he uiliy of erminal wealh conrolled by invemen raegy, i.e., max π V[,T] E[U(X T)], (.1) where he uiliy funcion U( ) i uually concave and ricly increaing due o he rik averion. Firly, we decribe he wealh proce X } in a financial marke. The financial marke i compoed of wo ae: a ock andacah wih a ochaic inere rae. The probabiliy pace (Ω,F,F,P) conruced in he financial marke i generaed by wo independen Brownian moion W } and B }. The equipped filraion F = F } i defined by F = σ(w u,b u ) : u } and F = σ( F ). Now we characerize he dynamic of ae. The inere rae i ochaic and decribed by Vaicek model: dr = a( r r )d+σ r db, (.) where he parameer a, r and σ r are all poiive. The ock proce S } i a geomeric Brownian moion aifying he following SDE: ds S = (r +ν)d+σ 1 dw +σ db, (.3) where he inrinic volailiy σ 1 i poiive and ν repreen he oal rik premium. The correlaed volailiy σ i uually negaive becaue people rend o depoi when inere rae become higher. Under he aumpion of elf-finance, he wealh proce can be uniquely deermined by he adaped raegy proce π }, where π i he fracion of he oal wealh inveed in ock a ime. Thu, he dynamic

5 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 5 of wealh proce X } i ds dx =(1 π )X r d+π X S =(r +π ν)x d+π X σ 1 dw +π X σ db, X =x. (.4) Secondly, we inroduce he admiible raegy e V[, T]. A raegy π i called an admiible raegy if (i) π L F (,T;R), i.e., π i a F } -adapedr-valued proce and E π() d} < ; (ii) here exi an unique oluion of equaion (.4), denoed by X( ); (iii) X( ) aifie he wealh conrain, i.e., X( ). Propoiion.1. Le L F (Ω;C([,T];R)) be he e of all coninuou procee in L F (,T;R). Then we have L F(Ω;C([,T];R)) V[,T] L F(,T;R). Proof. IiobviouhaV[,T] L F (,T;R)byhedefiniionofV[,T]. So we ju need o how, for any π L F (Ω;C([,T];R)), here exi an unique oluion of equaion (.4) and i i non-negaive. Becaue π i coninuou on [,T], i i bounded and he coefficien in SDE (.4) are all Lipchiz coninuou. By Theorem 6.3 in chaper 1 of Yong and Zhou(1999)[18], he uniquene and exience of rong oluion X( ) boh hold. Finally, X( ) i non-negaive due o he propery of geomeric Brownian moion. Corollary.. Suppoe XT i he opimal erminal wealh o he opimizaion problem: max E[U(X T)], X T V (.5) where V = X T : π L F (,T;R).. X T = x + (r +π ν)x d+ π X ( σ1 dw +σ db ) }, (.6) and π i he correponding opimal raegy of X in (.6) and coninuou. Then he π maximize he original problem (.1). We omi he proof becaue i can be eaily verified uing Propoiion.1. Thi corollary ell u ha if he opimal invemen raegy of he

6 6 ZONGXIA LIANG AND MING MA erminal wealh problem (.5) i coninuou, hen he oluion of (.5) lead u o he oluion of he original problem (.1). Forunaely, he condiion of coninuiy in corollary. uually hold hank o he feedback conrol. Furhermore, he ranformaion form (.1) o (.5) i crucial becaue maringale approach ju work for hi kind of erminal wealh opimizaion. 3. Five-ep maringale approach Becaue he number of ochaic ae i le han he number of Brownian moion, he maringale meaure in he finical marke i no unique. We uually call hi kind of marke a incomplee marke. I i well-known he maringale meaure i crucial in maringale approach and influence he budge conrain direcly. In complee marke, he uniquene of equivalen maringale meaure guaranee he efficiency of he maringale approach. Once he equivalen maringale meaure i divere, here exi differen o-called opimal raegie under differen meaure. Hereo, a meaure leading u o he real opimal raegy i demanded Neural meaure and expanded pace. According o he hiorical daa of ock, premium ν can be eimaed by he aiical propery of deflaed log-reurn. Thu, any meaure accorded wih he oal premium can be choen a a maringale meaure. Significanly, he premium correponding o he equivalen maringale meaure can be ochaic. The e Q conained all admiible meaure i decribed accuraely a follow: Q= Q Λ : dqλ dp =exp( 1 [λ 1 (u) +λ (u) ]du λ 1 (u)dw u } ) λ (u)db u, Λ=(λ1,λ ) L F (,T;R ).. σ 1 λ 1 +σ λ ν, (3.1) where Λ i he ochaic premium correponding o he meaure Q Λ. For any equivalen maringale meaure in Q, he correponding expanded pace of erminal wealh i V(Λ) = X T : E QΛ [exp( r u du)x T ] = x }. (3.)

7 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 7 The budge conrain in (3.) mean he erminal wealh can generae a dicouned wealh proce e rudu X E QΛ [e rudu X T F ], which i a Q Λ -maringale wih mean x. The following heorem ell he relaionhip beween he elf-finance pace and he expanded pace. Theorem 3.1. V = Λ:Q Λ Q} V(Λ). Proof. Firly, we how V V(Λ) for any Q Λ Q. For X T V we have he following: X T =x + =x + =x + (r +π ν)x d+ r X d+ r X d+ π X (σ 1 dw +σ db ) π X (σ 1 (dw +λ 1 ()d)+σ (db +λ ()d)) π X (σ 1 d W +σ d B ), where W and B are Brownian moion under meaure Q Λ. Uing Iô formula, he dicouned erminal wealh can alo be expreed a a ochaic inegral. Thu V = X T : π L F(,T;R).. e rudu X T = x + } π e rudu X (σ 1 d W +σ d B ), i.e., X T aifie he budge conrain and o X T V(Λ). (3.3) Secondly, how V Λ:Q Λ Q} V(Λ). By maringale repreenaion heorem(cf. Chaper 3 of Karaza and Shreve(1)[13]), we can rewrie V(Λ) a he follow: V(Λ) = X T : I 1,I L F (,T;R).. e } T rudu X T =x + I 1 ()σ 1 d W +I ()σ d B = X T : I 1,I L F(,T;R).. e rudu X T =x + [I ()ν+(i 1 () I ())σ 1 λ 1 ()]d+i 1 ()σ 1 dw +I ()σ db }, (3.4) wherei i (i = 1,)ianadapedochaicproceobainedbymaringale decompoiion and we call i maringale coefficien in hi paper. Thu for any X T Λ:Q Λ Q}, Λa and Λ b L F (,T;R ) here exi I a 1,Ia,Ib 1,Ib

8 8 ZONGXIA LIANG AND MING MA L F (,T;R) aifying = = e rudu X T x [I a ()ν +(I a 1() I a ())σ 1 λ a 1()]d+I a 1()σ 1 dw +I a ()σ db [I b ()ν +(I b 1() I b ())σ 1 λ b 1()]d+I b 1()σ 1 dw +I b ()σ db. Comparing he Iô inegral and he Lebegue inegral in he la wo equaliie we have I a 1 = I b 1, I a = I b and I a 1 = I a, repecively. Thu, if we denoe I I a 1 = Ib 1 = Ia = Ib and π So X T V. = = e rudu X T x e I() rudu X, hen I()νd+I()σ 1 dw +I()σ db π e rudu X (νd+σ 1 dw +σ db ). 3.. Opimizaion in expanded pace. Before propoing he opimal neural meaure, we pecify he radiional four-ep maringale approach wih a given meaure. Fixing an equivalen maringale meaure, maringale approach in incomplee marke ha no difference from complee marke, which bae on he heory of Lagrange muliplier developed by Cox and Huang[3](1989). Reader can alo refer o Karaza and Shreve [1](1998) for he deail and we ummarize hee reul by Propoiion 3.. For a fixed equivalen maringale meaure Q Λ and any iniial ime, denoe he Lagrangian funcion by [ ] L(,x,r,y,X T ) = E [U(X T ) F ]+y x E(H Λ (,T)X T F ),(3.5) where H Λ (,T) = exp ( r udu )( ) dq Λ dp exp ( r udu )( dq Λ dp = exp r u du 1 λ 1 (u) +λ (u) du ) T λ 1 (u)dw u +λ (u)db u }. If here exi X T V(Λ) and F -meaurable random variable y aifying (i) X T maximize L(,x,r,y,X T) over V(Λ), (ii) budge conrain hold, i.e., y [x E(H Λ (,T)X T F )] =,

9 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 9 hen XT i he unique opimal oluion o he problem in expanded pace V(Λ), i.e., E[U(XT ) F ] = max E[U(X T ) F ]. X T V(Λ) The Propoiion 3. enure he efficiency of maringale approach, which exacly i a rong dualiy heorem in he ochaic conrol. The proof abou opimaliy i imilar o he deermined programming, and inereed reader can refer o Luenberger and Ye(8)[15]. The proof abou uniquene i an applicaion of Jenen inequaliy hank o he well propery of he uiliy funcion. Now we can li he radiional maringale approach a a four-ep approach. Given an equivalen maringale meaure Q Λ, maringale approach o opimal wealh wih iniial x and r can be realized by ep(i) ge he opimal wealh X T (y ) a a funcion of y, which maximize he Lagrangian funcion L(,x,r,y,X T ); ep(ii) ubiue X T (y ) ino budge conrain and ge he opimal Lagrange muliplier y ; ep(iii) for any [,T], calculae he explici form of dicouned wealh a ime : e rudu X E QΛ [e rudu X T F ]; ep(iv) differeniaee rudu X andhengei1(;,x,r ),I(;,x,r ) a ime [,T]. Indeed, he fir wo ep calculae he opimal erminal wealh and Lagrange muliplier while he la wo ep realize he maringale decompoiion o ge maringale coefficien in (3.4). Bu in pracice, i i complexoexprei i (i = 1,)aeveryime. Sohefollowingconiency heorem devoe o improving he la wo ep. Theorem 3.3. (Coniency heorem) Regarding ep(iii) and ep(iv) a he definiion of maringale coefficien I i (;,x,r )( ) for i = 1,, we have where I i(;,x,r ) = e rudu I i(;,x (;,x,r ),r ), r = r + e rudu X (;,x,r )=x + a( r r )d+ σ r db, I 1(u;,x,r )σ 1 d W u +I (u;,x,r )σ d B u.

10 1 ZONGXIA LIANG AND MING MA Proof. We firly prove X (T;,x,r ) = X (T;,X (;,x,r ),r ), (3.6) which mean he global opimum i he local opimum. By he definiion of opimal wealh, we know E[U(X (T;,x,r ))] E[U(Y 1 )], Y 1 V(Λ), E[U(X (T;,X (;,x,r ),r )) F ] E[U(Y ) F ] fory V(Λ;,X (;,x,r ),r ),wherev(λ;,x (;,x,r ),r )iimilar o V(Λ) in (3.4), and defined by V(Λ;,X (;,x,r ),r ) X T : I 1,I L F (,T;R).. e T r udu X T = X (;,x,r )+ I 1 (u)σ 1 d W u +I (u)σ d B u }. Chooing Y 1 = X (T;,X (;,x,r ),r ) and Y = X (T;,x,r ), he equaliy (3.6) can be proved by he uniquene of X (T;,x,r ) and he following inequaliie: E[U(X (T;,x,r ))] E[U(X (T;,X (;,x,r ),r ))] =E[E[U(X (T;,X (;,x,r ),r )) F ]] E[E[U(X (T;,x,r )) F ]] =E[U(X (T;,x,r ))]. Secondly, reropecing (3.4), he relaionhip beween X T and I i (i = 1, ) provide e rudu X (T;,x,r )=x + and e rudu X (T;,X (;,x,r ),r )) =X (;,x,r )+ I 1 (u;,x,r )σ 1 d W u +I (u;,x,r )σ d B u, e rudu I 1 (u;,x (;,x,r ),r ))σ 1 d W u +e rudu I (u;,x (;,x,r ),r ))σ d B u. Thu he inegrand a ime aifie I i (;,x,r ) = e rudu I i (;,X (;,x,r ),r ) for i = 1,. The proof hen i complee.

11 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 11 Remark 3.4. Coniency heorem provide anoher approach o I i (;,x,r )(i = 1,, > ), which can replace he la wo ep of radiional maringale approach. For any [, T], differeniaing E QΛ [e rudu X T F ] w.r. ime can ge I i (;,X,r )(i = 1,), furhermore, I i(;,x,r )(i = 1,) i obained by he relaionhip in Theorem 3.3. I i no obviou ha hi new approach become impler, bu i really doe ince I i(;,x,r )(i = 1,) ha complex form for >. Wih a higher perpecive, he maringale approach can be realized dynamically according o he coniency heorem Opimal neural meaure. Afer inroducing he expanded pace wih repec o differen equivalen maringale meaure, we are inereed in he relaionhip beween opimizaion on he original pace and he expanded pace. By Theorem 3.1, he following inequaliy hold: max E[U(X T)] max E[U(X T)], Λ Λ : Q Λ Q}. X T V X T V(Λ) A conequen key queion i wheher he equaliy can hold a ome Λ, and wheher hi Λ make XT in V. If hee wo queion have poiive anwer, we can olve he original opimizaion by he expanded pace V(Λ ). The following definiion i propoed o realize hi idea. Definiion 3.1. A equivalen maringale meaure Q Λ Q i called an opimal neural meaure in Q, if here exi an opimal erminal wealh X T and an opimal premium Λ L F (,T;R ) aifying (i) X T V; (ii) X T maximize E[U(X T)] over V(Λ ). Propoiion 3.5. Le V(Λ) max XT V(Λ)E[U(X T )] be a he value funcion on Q. Then he opimal neural meaure Λ reache he minimum of V( ). Proof. Suppoe X T i he opimal erminal wealh correponding o Λ. For Λ Q, by Theorem 3.1 and Definiion 3.1, we have V(Λ) max X T V E[U(X T)] E[U(X T )] = V(Λ ).

12 1 ZONGXIA LIANG AND MING MA Theorem 3.6. A ufficien and neceary condiion for Λ being he opimal neural meaure i I 1 (;,x,r ) = I (;,x,r ), [,T], where I 1 ( ) and I ( ) are maringale coefficien correponding o X T in (3.4) and X T maximize E[U(X T)] over V(Λ ). Proof. By coniency heorem 3.3, we ubiue an equivalen condiion: I 1 (;,x,r ) = I (;,x,r ), [,T]. Neceiy i obviou by (3.3) and Definiion 3.1. Sufficiency i no complicaed ye. We ju need o verify X T V under he aumpion I 1 I. Chooing π I 1(;,x,r ) in (3.3), he proof i hen complee. e rudu X The concluion above formulae he maringale approach in incomplee marke. For an opimizaion problem in an incomplee marke, we need o olve a family of problem in expanded pace parameerized by Λ, by which we can find he opimal neural meaure. Combining he radiional four ep, he Remark 3.4 and Theorem 3.6, we ummarize he maringale approach in incomplee marke a he following five-ep approach: ep(i) for any Λ and any iniial ime, ge he opimal wealh X T (y ) a a funcion of y, which maximize he Lagrangian funcion L(,x,r,y,X T ); ep(ii) ubiue X T (y ) ino budge conrain and ge opimal Lagrange muliplier y ; ep(iii) for any [,T], calculae he explici form of dicouned wealh a ime : e rudu X E QΛ [e rudu X T F ]; ep(iv) differeniae e rudu X wih repec o, hen ge he maringale coefficien: I 1(;,x,r,Λ) and I (;,x,r,λ); ep(v) findheopimalneuralmeaureλ aifyingi 1 (;,x,r,λ ) = I (;,x,r,λ ) and he opimal invemen amoun π X i I 1 (;,X,r,Λ ). Moreover, in he view of opimal crierion, he opimal neural meaure i indeed a minimax meaure over Q. Thi i accuraely explained by Corollary 3.7. If he opimal neural meaure Λ exi, hen max E[U(X T)] = max E[U(X T)] = min X T V X T V(Λ ) Λ:Q Λ Q} max E[U(X T)]. X T V(Λ)

13 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 13 Proof. The concluion can be happily achieved via combining Propoiion 3.5 wih he exience of opimal neural meaure. Anoher key poin we hould expound i he opimal neural meaure doe no exi if here i a rik oally independen of he invemen. For example, conidering an ae liabiliie managemen(alm), he deah in inurance marke are oally independen of inveable ae. Hainau and Devolder(6)[9] provide an approximaion by minimizing he econd momen, while conidering he ALM in an incomplee marke. Liang and Ma (15)[14] udy he non-hedging moraliy and alary rik in penion fund by hi approximae maringale approach. To review hee reearche in our framework, he meaure hey chooe ju minimize he difference beween I 1 and I, becaue he condiion I 1 = I in Theorem 3.6 can no hold a all. Forunaely, hough he marke i incomplee in our paper, he whole rik are embodied in he ock invemen, o we have an opporuniy o ge real opimal raegy by he opimal neural meaure. 4. Explici oluion abou raegy and meaure In hi ecion, CRRA and CARA uiliie are choen and he explici oluion are obained by a lo of complicaed calculaion. To pecific he main idea and implify he noaion, we pu ome calculaion in appendixe and omi ome Λ a upercrip if i doe no caue ambiguiy CRRA. The CRRA uiliy repreen a conan relaive rik averionwihuiliyfuncionu(x) = 1 γ xγ andwenowrealizefive-epmaringale approach under hi uiliy. For a fixed meaure Q Λ, he Lagrangian funcion defined by (3.5) a ime i [ 1 [ ] L(,x,r,y,X T ) = E ]+y γ Xγ x E(H(,T)X T F ). T F The local opimal wealh X T (y ) expreed by y i X T (y ) = [y H(,T)] 1 γ 1. Subiuing i ino he budge conrain (3.), we have: (y ) 1 γ 1 = x E[H(,T) γ γ 1 F ].

14 14 ZONGXIA LIANG AND MING MA Thu we ge he opimal wealh: X T (y ) = x E[H(,T) γ γ 1 F ] [H(,T)] 1 γ 1. So far he fir wo ep have been finihed, bu he la wo ep will be more complicaed. e rudu X E Q [exp r u du } XT F ] =E[H(,T)exp r u du } XT F ] =E[H(,T)exp r u du } = E[H(,T) γ γ 1 F ] [H(,T)] 1 x E[H(,T) γ γ 1 F ] [H(,)] 1 γ 1 exp r u du } E x = E[H(,T) γ γ 1 F ] exp x exp γ γ 1 A(T )r +C(T ; r u du } [H(,)] 1 γ 1 γ γ 1 )}, γ 1 F ] [[H(,T)] γ γ 1 F ] where A( ) i a deermined funcion and C( ) i a funcion relaed o Λ and γ γ 1. Their expreion are boh in appendix A. Uing Iô formula, he differenial of e rudu X a ime = i: ( X 1 γ 1 λ 1()d W 1 γ 1 λ ()d B + γ γ 1 A(T )σ rd B where B and W are boh Brownian moion under Q Λ. So we ge he following explici expreion: Leing I 1 (;,x,r ) = X λ 1 () σ 1 (γ 1), λ ()+γa(t )σ r I (;,x,r ) = X. σ (γ 1) λ 1 () = νσ 1 σ 1 σ σ r γa(t ), σ1 +σ λ () = νσ +σ 1 σ 1 σ r γa(t ), σ1 +σ ), (4.1) i i eay o verify Q (λ 1,λ ) Q and Λ = (λ 1,λ ) i he opimal neural meaure by Theorem 3.6. By ep(v), he correponding opimal raegy

15 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 15 in elf-financial pace i π = ν σ σ r γa(t ) (1 γ)(σ 1 +σ ). (4.) I i noeworhy Λ become conan once σ r =, which repreen he inere rae i no ochaic. So, while conidering a ochaic inere rae, he rik premium behave exraordinarily and really deerve a deep udy. 4.. CARA. The CARA uiliy repreen a conan abolue rik averion wih uiliy funcion U(x) = 1 α e αx. For a fixed meaure Q Λ, he Lagrangian funcion a ime i L(,x,r,y,X T ) = E [ 1 [ ] α exp( αx T) F ]+y x E(H(,T)X T F ). (4.3) The local opimal wealh X T (y ) expreed by y i X T(y ) = 1 α ln[y H(,T)]. By he budge conrain (3.), he opimal Lagrange muliplier aifie Thu he opimal wealh i ln(y ) = αx +E[H(,T)ln(H(,T)) F ]. E[H(,T) F ] X T (y ) = x + 1 α E[H(,T)ln(H(,T)) F ] E[H(,T) F ] 1 α ln(h(,t)). We ill omi ome ubcrip Λ in nex wo ep. Following he reul in appendix A and B, we have e rudu X E Q[ exp r u du } ] XT F =E [ H(,T)exp r u du } 1 α ln(y ) 1 α ln(h(,)) 1 α ln(h(,t))} ] F =exp r u du } 1 α ln(y ) 1 α ln(h(,))} E [ ] H(,T) F 1 α ( r u du)e [ ] H(,T)ln(H(,T)) F

16 16 ZONGXIA LIANG AND MING MA =exp r u du } 1 α ln(y ) 1 α ln(h(,))} exp A(T )r +C(T ) } 1 α exp r u du } exp A(T )r +C(T ) } (A(T )r +F(T )) = 1 α exp r u du } exp A(T )r +C(T ) } ln(y ) +ln(h(,))+a(t )r +F(T ) }. Uing Iô formula, we ge he differenial of e rudu X a ime = : exp [ A(T )r +C(T ) ] ( λ1 () α [ λ () + αa(t )σ r ) d W +exp [ A(T )r +C(T ) ] x exp(a(t )r +C(T )) 1 ] A(T )σr d B. α Acually, ince e rudu X i a Q Λ maringale, i i ufficien o focu on he differenial of Brownian moion in above calculaion, which ake a huge implificaion. Then he maringale coefficien are I 1 (;,x,r ) = λ 1()exp[A(T )r +C(T )], σ 1 α I (;,x,r ) = exp[a(t )r +C(T )]A(T )σ r σ [ λ () + αa(t )σ r x exp(a(t )r +C(T )) 1 α ]. To make he ufficien condiion hold, we wan I 1 = I, which equal o λ () = νσ A(T )σ r σ1 [αx exp( A(T )r C Λ (T )) 1]. σ1 +σ (4.4) I mu be emphaized ha (4.4) i no an explici formulaion becaue C Λ (T ) i dependen on he informaion of Λ afer ime. Thu, he opimal pair (X,Λ ) hould aify he following equaion:

17 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 17 X T =x + (r +π ν)x d+ π X (σ 1 dw +σ db ), ( )( ) ν X π =exp A(T )r +C Λ σ λ () (T ), ασ 1 λ () = νσ A(T )σ r σ1 [αx exp( A(T )r C Λ (T )) 1], σ1 +σ C Λ (T ) = (A(T )+T ) r + 1 [σ r A(T u)] du [ ( ) ]) F +ln (E Q(λ 1,(λ σra(t ))) exp σ r A(T u)λ (u)du. (4.5) Thi family of forward-backward equaion expree a complex relaionhip beween he opimal wealh and he opimal neural meaure. Though i i hard o ge an explici oluion, (4.5) ill decribe he opimal neural meaure compleely. In appendix C, we provide an algorihm o approach he opimal premium which i ued in he numerical illuraion. 5. Numerical illuraion Thi ecion how he numerical illuraion of he opimal neural meaure a well a he opimal raegie. An opimal neural meaure canbeoallydeerminedbyanopimalpremiumproceλ = (λ 1,λ ),o we diplay he opimal premium proce inead. The aniheic raegy i derived from a conan premium, which i exacly he opimal premium under a conan inere rae. The fir reul in hi ecion i ha he opimal premium proce ecure a beer wealh proce in differen ene under differen uiliie. We conider a peron wih uni wealh a iniial ime = and he invemen horizon i year, i.e., T =. For he ochaic inere rae, we uppoe he iniial rae r i %, he peed of adjumen a i 1.7%, he revering mean r i 3.88%, and he volailiy σ r i 1.75%. For he ock proce, we uppoe he oal rik premium ν i 5.35%, he inrinic volailiy σ 1 i 15.4% and he correlaed volailiy σ i 1% CRRA. We chooe he parameer of CRRA uiliy: γ = 5% and denoe X1 a he correponding wealh proce deermined by opimal

18 18 ZONGXIA LIANG AND MING MA raegy (4.). X repreen he wealh proce deermined by following aniheic raegy ˆπ wih repec o a conan premium: ˆπ = λ 1(), [,T], σ 1 (γ 1) λ 1 λ 1(T) = νσ 1. σ1 +σ I i remarkable he opimal invemen proporion under CRRA uiliy i uually a proce independen o wealh, which mean he inveor hould keep he invemen proporion no maer how he marke behave. Thi appearance i exacly caued by he relaive averion. Moreover, he relaive averion i ill refleced in he behavior of opimal wealh. The Figure 1 compare he differen behavior of opimal wealh X1 and aniheic wealh X, where he relaive exce reurn X1 X and X he abolue exce reurn X1 X are depiced repecively in he lef and righ graph. I evidence X1 gain a anding relaive exce reurn regarding X a a benchmark, bu X1 behave wore in he ene of abolue exce reurn. Thi phenomenon i reaonable once noing he uiliy i conan relaive rik avere o ha we ill realize an inveor wih CRRA uiliy ju concenrae on he relaive exce reurn. Afer 1.5 x 1 4 average of relaive exce wealh 1.5 (X1 X)./X ime(year) average of abolue exce wealh X1 X ime(year) Figure 1. Comparion beween opimal wealh and anihei under CRRA hi comparion, we udy he opimal premium under CRRA. By (4.1), Λ i a deermined funcion o we draw i direcly in he Figure. There are hree ignifican appearance in hi figure. Fir one i he ock premium λ 1 keep poiive while inere rae premium λ i negaive. Thi reflec a pychology ha people wih CRRA like inere rae rik and hae ock rik. Acually, ock rik reduce he

19 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 19 abiliy of growh bu he inere rae rik promoe he growh due o he convexiy of exponenial funcion. Second one i λ 1 i an order of magniude larger han λ, which ell u he influence of ock rik i greaer han inere rae rik. Thi coincide wih he relaionhip beween he inrinic volailiy and he correlaed volailiy. The la one i λ 1 and λ are boh decreaing in he coure of ime. Thi how a rik-avere inveor urn o be relaively fearle while approaching he erminal ime, becaue he foregone concluion can be hardly changed. We call hi phenomenon a ime effec. premium of ock rik λ 1.3 λ premium of inere rae rik ime(year) ime(year) Figure. The opimal premium under CRRA 5.. CARA. We chooe he parameer α = 5%. Our ulimae aim i o imulae he opimal pair (premium, raegy and wealh) in (4.5), bu he expecaion under opimal neural meaure in FBSDE i hardly o handle. So an approximae pair i ued in hi imulaion, which i deailed calculaed in appendix C. In hi par, X1 i he approximae opimal wealh inroduced in he appendix and X i he aniheic wealh proce deermined by a imilar-formed raegy ˆπ wih a conan premium. ˆπ X = ν σ rσ A(T ) α(σ1 +σ ) exp [ A(T )r +C(T ) ] + σ rσ A(T ) X σ }}, 1 +σ }} conan correcion C(T ) = A(T u)a( r σ rλ (u) )+.5A(T u) σ a rdu, λ λ (T) = νσ. σ1 +σ I i worhwhile o noe ha he opimal invemen amoun under CRRA uiliy i he um of wo componen. The conan erm dominae when

20 ZONGXIA LIANG AND MING MA wealh amoun i le or he erminal i upcoming. The correcion erm dominae ju a he beginning of invemen wih a bigger amoun of wealh. I ell u a CARA inveor i ap o keep a conan invemen amoun no maer how he marke behave. Furhermore, hi abolue rik averion i ill refleced in he opimaliy of X1. The Figure 3 compare he differen behavior of he opimal wealh X1 and he aniheic wealh X, where he relaive exce reurn X1 X X and he abolue exce reurn X1 X are depiced. I evidence X1 gain a anding abolue exce reurn regarding X a a benchmark, bu X1 canno win X in he ene of relaive exce reurn. So we aer an inveor wih CARA uiliy ju concenrae on he abolue exce reurn. average of relaive exce wealh (X1 X)./X ime(year) average of abolue exce wealh X1 X ime(year) Figure 3. Comparion beween opimal wealh and anihei under CARA Now we udy he opimal premium. By (4.4), Λ i a ochaic proce o we can only draw i average behavior in he Figure 4 o udy i. The paen of opimal premium under CARA ha ome difference from CRRA condiion, hough ome explicaion in CRRA condiion i ill valid. We ju emphaize he premium ad curve now. The ad curve i a ynheical appearance caued by ime effec and capial effec ogeher. The ime effec ha already menioned in CRRA condiion a inveor urn o be fearle while approaching he erminal ime. The capial effec do no appear in CRRA condiion becaue inveor wih relaive averion conider he relaive growh bu no he abolue amoun. When inveor have abundan wealh, he abolue exce reurn can be grealy gained or lo. So, a an abolue rik-avere inveor

21 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 1 wih more abundan wealh, i rik premium become higher. Thi exacly explain a phenomenon ha ome rich men refue he invemen in ock by he pychology of abolue rik averion. Following hi logic, he average premium urn o be higher a he beginning of invemen ince he expeced wealh gain. The decreaing of premium ad curve a erminal ime i dominaed by ime effec of coure average premium of ock rik λ 1 average premium of inere rae rik λ ime(year) ime(year) Figure 4. The premium ad curve under CARA 6. Concluion In hi paper, we conider he imple invemen problem in an incomplee marke. The invemen proporion in ock i he only hing ha an inveor can decide, which mean he number of invemen i le han he rik and hu he marke i incomplee. The invemen arge i maximizing he expeced uiliy of erminal wealh. The uual wo mehod, HJB equaion by dynamic programming and maringale approach, boh mee difficulie in incomplee marke, repecively. The five ep of maringale approach in incomplee marke are diplayed in our paper by defining an opimal neural meaure. The coniency heorem i proved o improve he ep(iii) and ep(iv), which admi u uing maringale approach dynamically. Baed on above udy, we implemen he maringale approach under CRRA and CARA uiliie. The opimal premium under CRRA uiliy i deerminiic indeed, and an explici opimal raegy i obained. Under CARA uiliy, he opimal premium i really ochaic and we can only ue a family of equaion o decribe i. In order o gain a deep knowledge of opimal neural meaure, we make numerical illuraion

22 ZONGXIA LIANG AND MING MA a he epilog of reearch. In fac, an equivalen maringale meaure repreen a rik premium, which i he minimum amoun of money by which an inveor i willing o ake a rik. The udy of opimal premium curve embodie he pychological change during he invemen period. Under CRRA uiliy, he opimal premium i decreaing which mean a rik-avere inveor urn o be fearle while approaching he erminal ime. We call hi phenomenon he ime effec in behavioral finance. There appear a capial effec a well a he ime effec in he CARA uiliy condiion. The capial effec decribe an abolue rikavere inveor become more pruden when he own abundan wealh. Combining he ime and capial effec, he average opimal premium preen a a concave curve, and we call i premium ad curve. 7. Appendix Appendix A: Calculaion of E[H(,T) γ F ] Firly, we give an explici expreion of r u du relaed o he SDE (.). Thank o he propery of Ornein-Uhlenbeck proce, r u = e a(u ) r + r(1 e a(u ) )+ Inegraing by par we have u σ r e a(u v) db v, u >. r u du = A(T )r (A(T )+T ) r +σ r A(T u)db u, u >, where A(x) = 1 a (e ax 1). Now we ar he following calculaion baed on hee preparaion: E[H(,T) γ F ] = E exp [ γ( r u )du γ } ] γ λ 1 (u)dw u +λ (u)db F u =E exp γ λ 1 (u) +λ (u) du [γa(t )r γ(a(t )+T ) r +γσ r A(T u)db u ] } F λ 1 (u) +λ (u) du γ λ 1 (u)dw u +λ (u)db u

23 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 3 =E Q(γλ 1,γ(λ σra(t ))) exp γ λ 1 (u) +λ (u) du+ 1 ] } F +[γλ 1 (u)] du [ γa(t )r γ(a(t )+T ) r [ =exp γa(t )r γ(a(t )+T ) r + 1 [ E Q(γλ 1,γ(λ exp σra(t ))) ] } F γ σ r A(T u)λ (u)du } exp γa(t )r +C Λ (T ;γ). [γσ r A(T u) γλ (u)] γ γ (λ 1 (u) +λ (u) ) ] [γσ r A(T u)] du We emphaize C( ) i relaed o he equivalen maringale meaure Λ and parameer γ a C Λ (T ;γ) = γ(a(t )+T ) r + 1 [γσ r A(T u)] du [ ( +ln (E Q(γλ 1,γ(λ γ γ σra(t ))) exp (λ 1 (u) +λ (u) ) ) ]) F γ σ r A(T u)λ (u)du. And when γ = 1, we uually omi γ in C Λ (T ;γ) for implificaion. Appendix B: Calculaion of E[H(,T)ln(H(,T)) F ] Referring o Lemma 8.6. in Okendal(13)[17] we have E[H(,T)ln(H(,T)) F ] = exp A(T )r +C Λ (T ;1) } E R [ln(h(,t)) F ], where R i a new probabiliy meaure uch ha dr dp o appendix A, we have = H(,T). Similar E R[ ] ln(h(,t)) F } H(,T) =E E[(H(,T)) F ] ln(h(,t)) F

24 4 ZONGXIA LIANG AND MING MA ( =E 1 exp λ 1 (u)dw u + [σ r A(T u) λ (u)]db u 1 ) ( [λ (u) σ r A(T u)] du exp σ r A(T u)λ (u)du (E Q(λ 1,(λ σra(t ))) [ exp [ ( = (E Q(λ 1,(λ σra(t ))) exp ( E Q(λ 1,(λ [exp σra(t ))) λ 1 (u) du ( ) ]) 1 F σ r A(T u)λ (u)du ln(h(,t)) } F σ r A(T u)λ (u)du ) F ] ) 1 ) σ r A(T u)λ (u)du ln(h(,t)) ] F [ ( =A(T )r + (E Q(λ 1,(λ σra(t ))) exp σ r A(T u)λ (u)du ) ] ) 1 F ( E Q(λ 1,(λ exp σra(t ))) 1 + )( σ r A(T u)λ (u)du (A(T )+T ) r λ 1 (u) +λ (u) du+ λ 1 (u) +[λ (u) σ r A(T u)] du ) } F σ r A(T u) λ (u)dbu Q λ 1 (u)dwu Q [ ( =A(T )r + (E Q(λ 1,(λ σra(t ))) exp σ r A(T u)λ (u)du ) ] ) 1 F ( E Q(λ 1,(λ exp σra(t ))) + 1 λ 1 (u) du+ A(T )r +F Λ (T ). )( σ r A(T u)λ (u)du (A(T )+T ) r ) } F [λ (u) σ r A(T u)].5λ (u) du We emphaize F( ) i relaed o he equivalen maringale meaure Λ oo and i doe no have a clear form unle Λ i a deermined funcion. [ ( F Λ (T ) = (E Q(λ 1,(λ σra(t ))) exp σ r A(T u)λ (u)du ) ] ) 1 F ( E Q(λ 1,(λ exp σra(t ))) + 1 λ 1 (u) du+ )( σ r A(T u)λ (u)du (A(T )+T ) r ) } F [λ (u) σ r A(T u)].5λ (u) du. ) Appendix C: Approximae opimal pair under CARA

25 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 5 Propoiion 7.1. Suppoe (Y 1,Y ) are wo dimenional normal diribuion wih mean (µ 1,µ ) and covariance marix ( ) σ Σ = 1 ρσ 1 σ ρσ 1 σ σ. Then E[exp(Y 1 )Y ] = exp(µ 1 + σ 1 )(µ ρσ 1 σ ). Proof. E[exp(Y 1 )Y ] = E E[Y Y 1 ]exp(y 1 ) } =E µ exp(y 1 )+ρ σ σ 1 (Y 1 µ 1 )exp(y 1 ) } =(µ ρ σ σ 1 µ 1 )E[exp(Y 1 )]+ρ σ σ 1 E[Y 1 exp(y 1 )] =(µ ρ σ σ 1 µ 1 )exp(µ 1 + σ 1 )+ρσ σ 1 exp(µ 1 + σ 1 )(µ 1 σ 1) =exp(µ 1 + σ 1 )(µ ρσ 1 σ ). Equaion (4.5) are forward-backward equaion wih a ochaic riple pair (X,Λ,π). We ry o implify hem by he following approximaion: [ ( ) ]} F ln E Q(λ 1,(λ σra(t ))) exp σ r A(T u)λ (u)du σ r A(T u)e[λ (u) F ]du. Acually, above approximaion ignore he meaure ranformaion form P o Q (λ 1,(λ σ ra(t ))), and he volailiy of ochaic λ. Thi can be rue when Λ i minucule wih mall volailiy. We ue a new noaion λ (u;) o replace E[λ (u) F ]. Baing on hi ubiuion, he equaion (4.5) change ino X T = x + X π = exp (r +π ν)x d+ A(T )r +C Λ (T ) π X (σ 1 dw +σ db ), }( ν σ λ (;) λ (u;) = νσ A(T u)σ r σ1(αex u exp( A(T u)r u C Λ (T u)) F } 1), σ1+σ 1 C Λ (T )= (A(T )+T ) r+ [σ ra(t u)] σ r A(T u)λ (u;) } du. ασ 1 ),

26 6 ZONGXIA LIANG AND MING MA Thi family of equaion deermine he X1 in numerical illuraion and we ju need o calculae E[X exp( A(T )r C(T )) F ] now. Subiuing he opimal invemen amoun X π ino he dynamic of wealh, we ge dx =X r d+x π (σ 1 d W +σ d B ) ν σr σ A(T ) =X r d+ exp(a(t )r α(σ1 +σ) +C(T )) + σ } rσ A(T )X (σ σ1 1 d W +σ d B ). +σ By he heory of general one-dimenional linear equaion(cf.chaper 5 of Karaza and Shreve(1)[13]), he explici oluion of X can be expreed by where Thu, X = Z Y, [ Y = X + Z 1σ r σ A(T u)(ν σ r σ A(T u)) ] u α(σ1 +σ) expa(t u)r u +C(T u)}du + Zu 1 ν σ r σ A(T u) α(σ1 +σ ) expa(t u)r u + C(T u)}(σ 1 d W u +σ d B u ), Z = exp (r u.5 [σ rσ A(T u)] )du σ1 +σ } σ r σ A(T u) + (σ σ1 +σ 1 d W u +σ d B u ). } E X exp( A(T )r C(T )) F } =E exp( A(T )r C(T )+ln(z ))Y F E[exp(U )Y F ] =exp E[U F ]+Var[U F ]/ } E[Y F ] Cov[U,Y F ] }, where he la-econd equaliy come from a definiion and he la equaliy i guaraneed by Propoiion 7.1. To calculae he expecaion and

27 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 7 covariance of (U,Y ), we decompoe hem repecively ino he following: U = A(T )r C(T )+ r u.5 [σ } rσ A(T u)] du σ1 +σ σ r σ A(T u) + (σ σ1 +σ 1 d W u +σ d B u ) ) = A(T ) (e a( ) r +(1 e a( ) ) r+σ r e a( u) db u C(T )+ r u.5 [σ } rσ A(T u)] du σ1 +σ σ r σ A(T u) + (σ σ1 1 d W u +σ d B u ) +σ = A(T )(e a( ) r +(1 e a( ) ) r) A(T )σ r e a( u) db u C(T ) + r u + σ rσ A(T u)ν σ1 +σ + σ r σ A(T u) (σ σ1 1 dw u +σ db u ) +σ.5 [σ rσ A(T u)] σ 1 +σ } du = A(T )(e a( ) r +(1 e a( ) ) r) A(T )σ r e a( u) db u C(T )+ r(a( )+ ) A( )r σ r A( u)db u σr σ A(T u)ν +.5 [σ } rσ A(T u)] du σ1 +σ σ1 +σ σ r σ A(T u) + (σ σ1 +σ 1 dw u +σ db u ) =(A(T ) A(T )+ ) r A(T )r C(T ) σr σ A(T u)ν +.5 [σ } rσ A(T u)] du σ1 +σ σ1 +σ σ r σ A(T u) + σ σ1 1 dw u +σ + A(T )σ r e a( u) σ r A( u)+ σ rσ A(T u) σ σ1 +σ }db u, Y = X + Zu 1 σ r σ A(T u)(ν σ r σ A(T u)) α(σ1 +σ) exp A(T u)r u +C(T u) }} du + Zu 1 ν σ r σ A(T u) α(σ1 +σ ) exp A(T u)r u +C(T u) } (σ 1 d W u +σ d B u )

28 8 ZONGXIA LIANG AND MING MA = X + Z 1 (ν σ r σ A(T u)) u α(σ1 +σ ) expa(t u)r u +C(T u)}du+ Z 1 ν σ r σ A(T u) u α(σ1 +σ ) exp A(T u)r u +C(T u) } (σ 1 dw u +σ db u ). By he properie of Gauian diribuion and maringale momen equaliie we ge E[U F ] =(A(T ) A(T )+ ) r A(T )r σr σ A(T u)ν +.5 [σ rσ A(T u)] σ1 +σ σ1 +σ Var[U F ] = E[Y F ] =X + =X + = = Cov[U,Y F ] } du A(T u)a( r σ rλ (u;) )+.5A(T u) σr a } du, [A(T )σ r e a( u) +σ r A( u) σ rσ A(T u) σ σ1 +σ ] +[ σ rσ A(T u) σ σ1 1 ] }du, +σ (ν σ r σ A(T u)) α(σ 1 +σ ) E Z 1 u exp[a(t u)r u +C(T u)] F } du (ν σ r σ A(T u)) α(σ1 +σ ) exp E[U u F ]+Var[U u F ]/ } du, [ σ rσ A(T u) σ1 +σ ν σ rσ A(T u) α(σ 1 +σ ) σ1 +A(T )σ r σ e a( u) σ r σ A( u)+ σ rσ A(T u) σ σ1 +σ ] E[Z 1 u exp(a(t u)r u +C(T u)) F ]du σ r σ [A(T u) A( u)] ν σ rσ A(T u) α(σ 1 +σ ) exp E[U u F ]+Var[U u F ]/ } du, where we have ued he following equaliy in above calculaion. E[exp( U u ) F ] =E[Z 1 u exp(a(t u)r u +C(T u)) F ] =exp E[U u F ]+Var[U u F ]/}. Reference [1] Bellini, F., & Frielli, M. (). On he exience of minimax maringale meaure. Mahemaical Finance, 1(1): 1-1.

29 MARTINGALE APPROACH AND OPTIMAL NEUTRAL MEASURE 9 [] Cizár, I. (1975). I-divergence geomery of probabiliy diribuion and minimizaion problem. The Annal of Probabiliy, 3: [3] Cox, J. C., Huang, C. F., Opimal conumpion and porfolio policie when ae price follow a diffuion proce. Journal of economic heory 49(1): [4] Duffie, D., Fleming, W., Soner, H. M., & Zariphopoulou, T. (1997). Hedging in incomplee marke wih HARA uiliy. Journal of Economic Dynamic and Conrol, 1(4): [5] Delbaen, F., & Schachermayer, W.(1997). The variance-opimal maringale meaure for coninuou procee. Bernoulli, 1: [6] Frielli, M. (). The minimal enropy maringale meaure and he valuaion problem in incomplee marke. Mahemaical Finance, 1(1): [7] Grandi, P. (). On maringale meaure for ochaic procee wih independen incremen. Theory of Probabiliy & I Applicaion, 44(1): [8] Goll, T., & Rchendorf, L. (1). Minimax and minimal diance maringale meaure and heir relaionhip o porfolio opimizaion. Finance and Sochaic, 5(4): [9] Hainau, D., Devolder, P., 6. A maringale approach applied o he managemen of life inurance. Iniue of Acuarial Science, Univeri Caholique de Louvain, WP6, 1. [1] Kallen, J. (1999). A uiliy maximizaion approach o hedging in incomplee marke. Mahemaical Mehod of Operaion Reearch, 5(): [11] Karaza, I., Lehoczky, J. P., Shreve, S. E., & Xu, G. L. (1991). Maringale and dualiy mehod for uiliy maximizaion in an incomplee marke. SIAM Journal on Conrol and Opimizaion, 9(3):7-73. [1] Karaza, I., Shreve, S. E., Mehod of Mahemaical Finance, Vol. 39, Springer. [13] Karaza, I., & Shreve, S. (1). Brownian moion and ochaic calculu (Vol. 113). Springer Science & Buine Media. [14] Liang, Z., & Ma, M. (15). Opimal dynamic ae allocaion of penion fund in moraliy and alary rik framework. Inurance: Mahemaic and Economic,64:151C161. [15] Luenberger, D. G., & Ye, Y. (8). Linear and nonlinear programming (Vol. 116). Springer Science & Buine Media. [16] Meron,R.C., Opimum conumpion and porfolio rule in a coninuouime model. Journal of Economic Theory, 3(4): [17] Okendal, B. (13). Sochaic differenial equaion: an inroducion wih applicaion. Springer Science & Buine Media.

30 3 ZONGXIA LIANG AND MING MA [18] Yong, J., & Zhou, X. Y. (1999). Sochaic conrol: Hamilonian yem and HJB equaion (Vol. 43). Springer Science & Buine Media.

Macroeconomics 1. Ali Shourideh. Final Exam

Macroeconomics 1. Ali Shourideh. Final Exam 4780 - Macroeconomic 1 Ali Shourideh Final Exam Problem 1. A Model of On-he-Job Search Conider he following verion of he McCall earch model ha allow for on-he-job-earch. In paricular, uppoe ha ime i coninuou

More information

ANSWERS TO ODD NUMBERED EXERCISES IN CHAPTER

ANSWERS TO ODD NUMBERED EXERCISES IN CHAPTER John Riley 6 December 200 NWER TO ODD NUMBERED EXERCIE IN CHPTER 7 ecion 7 Exercie 7-: m m uppoe ˆ, m=,, M (a For M = 2, i i eay o how ha I implie I From I, for any probabiliy vecor ( p, p 2, 2 2 ˆ ( p,

More information

Fractional Ornstein-Uhlenbeck Bridge

Fractional Ornstein-Uhlenbeck Bridge WDS'1 Proceeding of Conribued Paper, Par I, 21 26, 21. ISBN 978-8-7378-139-2 MATFYZPRESS Fracional Ornein-Uhlenbeck Bridge J. Janák Charle Univeriy, Faculy of Mahemaic and Phyic, Prague, Czech Republic.

More information

, the. L and the L. x x. max. i n. It is easy to show that these two norms satisfy the following relation: x x n x = (17.3) max

, the. L and the L. x x. max. i n. It is easy to show that these two norms satisfy the following relation: x x n x = (17.3) max ecure 8 7. Sabiliy Analyi For an n dimenional vecor R n, he and he vecor norm are defined a: = T = i n i (7.) I i eay o how ha hee wo norm aify he following relaion: n (7.) If a vecor i ime-dependen, hen

More information

Stability in Distribution for Backward Uncertain Differential Equation

Stability in Distribution for Backward Uncertain Differential Equation Sabiliy in Diribuion for Backward Uncerain Differenial Equaion Yuhong Sheng 1, Dan A. Ralecu 2 1. College of Mahemaical and Syem Science, Xinjiang Univeriy, Urumqi 8346, China heng-yh12@mail.inghua.edu.cn

More information

Notes on cointegration of real interest rates and real exchange rates. ρ (2)

Notes on cointegration of real interest rates and real exchange rates. ρ (2) Noe on coinegraion of real inere rae and real exchange rae Charle ngel, Univeriy of Wiconin Le me ar wih he obervaion ha while he lieraure (mo prominenly Meee and Rogoff (988) and dion and Paul (993))

More information

Introduction to SLE Lecture Notes

Introduction to SLE Lecture Notes Inroducion o SLE Lecure Noe May 13, 16 - The goal of hi ecion i o find a ufficien condiion of λ for he hull K o be generaed by a imple cure. I urn ou if λ 1 < 4 hen K i generaed by a imple curve. We will

More information

Lecture 26. Lucas and Stokey: Optimal Monetary and Fiscal Policy in an Economy without Capital (JME 1983) t t

Lecture 26. Lucas and Stokey: Optimal Monetary and Fiscal Policy in an Economy without Capital (JME 1983) t t Lecure 6. Luca and Sokey: Opimal Moneary and Fical Policy in an Economy wihou Capial (JME 983. A argued in Kydland and Preco (JPE 977, Opimal governmen policy i likely o be ime inconien. Fiher (JEDC 98

More information

Utility maximization in incomplete markets

Utility maximization in incomplete markets Uiliy maximizaion in incomplee markes Marcel Ladkau 27.1.29 Conens 1 Inroducion and general seings 2 1.1 Marke model....................................... 2 1.2 Trading sraegy.....................................

More information

EECE 301 Signals & Systems Prof. Mark Fowler

EECE 301 Signals & Systems Prof. Mark Fowler EECE 31 Signal & Syem Prof. Mark Fowler Noe Se #27 C-T Syem: Laplace Tranform Power Tool for yem analyi Reading Aignmen: Secion 6.1 6.3 of Kamen and Heck 1/18 Coure Flow Diagram The arrow here how concepual

More information

NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY

NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY NECESSARY AND SUFFICIENT CONDITIONS FOR LATENT SEPARABILITY Ian Crawford THE INSTITUTE FOR FISCAL STUDIES DEPARTMENT OF ECONOMICS, UCL cemmap working paper CWP02/04 Neceary and Sufficien Condiion for Laen

More information

Discussion Session 2 Constant Acceleration/Relative Motion Week 03

Discussion Session 2 Constant Acceleration/Relative Motion Week 03 PHYS 100 Dicuion Seion Conan Acceleraion/Relaive Moion Week 03 The Plan Today you will work wih your group explore he idea of reference frame (i.e. relaive moion) and moion wih conan acceleraion. You ll

More information

Suggested Solutions to Midterm Exam Econ 511b (Part I), Spring 2004

Suggested Solutions to Midterm Exam Econ 511b (Part I), Spring 2004 Suggeed Soluion o Miderm Exam Econ 511b (Par I), Spring 2004 1. Conider a compeiive equilibrium neoclaical growh model populaed by idenical conumer whoe preference over conumpion ream are given by P β

More information

An Introduction to Backward Stochastic Differential Equations (BSDEs) PIMS Summer School 2016 in Mathematical Finance.

An Introduction to Backward Stochastic Differential Equations (BSDEs) PIMS Summer School 2016 in Mathematical Finance. 1 An Inroducion o Backward Sochasic Differenial Equaions (BSDEs) PIMS Summer School 2016 in Mahemaical Finance June 25, 2016 Chrisoph Frei cfrei@ualbera.ca This inroducion is based on Touzi [14], Bouchard

More information

Randomized Perfect Bipartite Matching

Randomized Perfect Bipartite Matching Inenive Algorihm Lecure 24 Randomized Perfec Biparie Maching Lecurer: Daniel A. Spielman April 9, 208 24. Inroducion We explain a randomized algorihm by Ahih Goel, Michael Kapralov and Sanjeev Khanna for

More information

Laplace Transform. Inverse Laplace Transform. e st f(t)dt. (2)

Laplace Transform. Inverse Laplace Transform. e st f(t)dt. (2) Laplace Tranform Maoud Malek The Laplace ranform i an inegral ranform named in honor of mahemaician and aronomer Pierre-Simon Laplace, who ued he ranform in hi work on probabiliy heory. I i a powerful

More information

CHAPTER 7: SECOND-ORDER CIRCUITS

CHAPTER 7: SECOND-ORDER CIRCUITS EEE5: CI RCUI T THEORY CHAPTER 7: SECOND-ORDER CIRCUITS 7. Inroducion Thi chaper conider circui wih wo orage elemen. Known a econd-order circui becaue heir repone are decribed by differenial equaion ha

More information

Research Article Existence and Uniqueness of Solutions for a Class of Nonlinear Stochastic Differential Equations

Research Article Existence and Uniqueness of Solutions for a Class of Nonlinear Stochastic Differential Equations Hindawi Publihing Corporaion Abrac and Applied Analyi Volume 03, Aricle ID 56809, 7 page hp://dx.doi.org/0.55/03/56809 Reearch Aricle Exience and Uniquene of Soluion for a Cla of Nonlinear Sochaic Differenial

More information

Backward Stochastic Differential Equations and Applications in Finance

Backward Stochastic Differential Equations and Applications in Finance Backward Sochaic Differenial Equaion and Applicaion in Finance Ying Hu Augu 1, 213 1 Inroducion The aim of hi hor cae i o preen he baic heory of BSDE and o give ome applicaion in 2 differen domain: mahemaical

More information

Explicit form of global solution to stochastic logistic differential equation and related topics

Explicit form of global solution to stochastic logistic differential equation and related topics SAISICS, OPIMIZAION AND INFOMAION COMPUING Sa., Opim. Inf. Compu., Vol. 5, March 17, pp 58 64. Publihed online in Inernaional Academic Pre (www.iapre.org) Explici form of global oluion o ochaic logiic

More information

Algorithmic Discrete Mathematics 6. Exercise Sheet

Algorithmic Discrete Mathematics 6. Exercise Sheet Algorihmic Dicree Mahemaic. Exercie Shee Deparmen of Mahemaic SS 0 PD Dr. Ulf Lorenz 7. and 8. Juni 0 Dipl.-Mah. David Meffer Verion of June, 0 Groupwork Exercie G (Heap-Sor) Ue Heap-Sor wih a min-heap

More information

ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS

ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS Elec. Comm. in Probab. 3 (998) 65 74 ELECTRONIC COMMUNICATIONS in PROBABILITY ESTIMATES FOR THE DERIVATIVE OF DIFFUSION SEMIGROUPS L.A. RINCON Deparmen of Mahemaic Univeriy of Wale Swanea Singleon Par

More information

To become more mathematically correct, Circuit equations are Algebraic Differential equations. from KVL, KCL from the constitutive relationship

To become more mathematically correct, Circuit equations are Algebraic Differential equations. from KVL, KCL from the constitutive relationship Laplace Tranform (Lin & DeCarlo: Ch 3) ENSC30 Elecric Circui II The Laplace ranform i an inegral ranformaion. I ranform: f ( ) F( ) ime variable complex variable From Euler > Lagrange > Laplace. Hence,

More information

Chapter 2. First Order Scalar Equations

Chapter 2. First Order Scalar Equations Chaper. Firs Order Scalar Equaions We sar our sudy of differenial equaions in he same way he pioneers in his field did. We show paricular echniques o solve paricular ypes of firs order differenial equaions.

More information

News-generated dependence and optimal portfolios for n stocks in a market of Barndor -Nielsen and Shephard type.

News-generated dependence and optimal portfolios for n stocks in a market of Barndor -Nielsen and Shephard type. New-generaed dependence and opimal porfolio for n ock in a marke of Barndor -Nielen and Shephard ype. Carl Lindberg Deparmen of Mahemaical Saiic Chalmer Univeriy of Technology and Göeborg Univeriy Göeborg,

More information

The Residual Graph. 12 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm

The Residual Graph. 12 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm Augmening Pah Algorihm Greedy-algorihm: ar wih f (e) = everywhere find an - pah wih f (e) < c(e) on every edge augmen flow along he pah repea a long a poible The Reidual Graph From he graph G = (V, E,

More information

Generalized Orlicz Spaces and Wasserstein Distances for Convex-Concave Scale Functions

Generalized Orlicz Spaces and Wasserstein Distances for Convex-Concave Scale Functions Generalized Orlicz Space and Waerein Diance for Convex-Concave Scale Funcion Karl-Theodor Surm Abrac Given a ricly increaing, coninuou funcion ϑ : R + R +, baed on he co funcional ϑ (d(x, y dq(x, y, we

More information

Mathematische Annalen

Mathematische Annalen Mah. Ann. 39, 33 339 (997) Mahemaiche Annalen c Springer-Verlag 997 Inegraion by par in loop pace Elon P. Hu Deparmen of Mahemaic, Norhweern Univeriy, Evanon, IL 628, USA (e-mail: elon@@mah.nwu.edu) Received:

More information

6.8 Laplace Transform: General Formulas

6.8 Laplace Transform: General Formulas 48 HAP. 6 Laplace Tranform 6.8 Laplace Tranform: General Formula Formula Name, ommen Sec. F() l{ f ()} e f () d f () l {F()} Definiion of Tranform Invere Tranform 6. l{af () bg()} al{f ()} bl{g()} Lineariy

More information

The Residual Graph. 11 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm

The Residual Graph. 11 Augmenting Path Algorithms. Augmenting Path Algorithm. Augmenting Path Algorithm Augmening Pah Algorihm Greedy-algorihm: ar wih f (e) = everywhere find an - pah wih f (e) < c(e) on every edge augmen flow along he pah repea a long a poible The Reidual Graph From he graph G = (V, E,

More information

u(t) Figure 1. Open loop control system

u(t) Figure 1. Open loop control system Open loop conrol v cloed loop feedbac conrol The nex wo figure preen he rucure of open loop and feedbac conrol yem Figure how an open loop conrol yem whoe funcion i o caue he oupu y o follow he reference

More information

Essential Microeconomics : OPTIMAL CONTROL 1. Consider the following class of optimization problems

Essential Microeconomics : OPTIMAL CONTROL 1. Consider the following class of optimization problems Essenial Microeconomics -- 6.5: OPIMAL CONROL Consider he following class of opimizaion problems Max{ U( k, x) + U+ ( k+ ) k+ k F( k, x)}. { x, k+ } = In he language of conrol heory, he vecor k is he vecor

More information

Hedging strategy for unit-linked life insurance contracts in stochastic volatility models

Hedging strategy for unit-linked life insurance contracts in stochastic volatility models Hedging raegy for uni-linked life inurance conrac in ochaic volailiy model Wei Wang Ningbo Univeriy Deparmen of Mahemaic Feng Hua Sree 818, Ningbo Ciy China wangwei@nbu.edu.cn Linyi Qian Ea China Normal

More information

Chapter 7: Inverse-Response Systems

Chapter 7: Inverse-Response Systems Chaper 7: Invere-Repone Syem Normal Syem Invere-Repone Syem Baic Sar ou in he wrong direcion End up in he original eady-ae gain value Two or more yem wih differen magniude and cale in parallel Main yem

More information

2. VECTORS. R Vectors are denoted by bold-face characters such as R, V, etc. The magnitude of a vector, such as R, is denoted as R, R, V

2. VECTORS. R Vectors are denoted by bold-face characters such as R, V, etc. The magnitude of a vector, such as R, is denoted as R, R, V ME 352 VETS 2. VETS Vecor algebra form he mahemaical foundaion for kinemaic and dnamic. Geomer of moion i a he hear of boh he kinemaic and dnamic of mechanical em. Vecor anali i he imehonored ool for decribing

More information

Exponential Sawtooth

Exponential Sawtooth ECPE 36 HOMEWORK 3: PROPERTIES OF THE FOURIER TRANSFORM SOLUTION. Exponenial Sawooh: The eaie way o do hi problem i o look a he Fourier ranform of a ingle exponenial funcion, () = exp( )u(). From he able

More information

Introduction to Congestion Games

Introduction to Congestion Games Algorihmic Game Theory, Summer 2017 Inroducion o Congeion Game Lecure 1 (5 page) Inrucor: Thoma Keelheim In hi lecure, we ge o know congeion game, which will be our running example for many concep in game

More information

Rough Paths and its Applications in Machine Learning

Rough Paths and its Applications in Machine Learning Pah ignaure Machine learning applicaion Rough Pah and i Applicaion in Machine Learning July 20, 2017 Rough Pah and i Applicaion in Machine Learning Pah ignaure Machine learning applicaion Hiory and moivaion

More information

18 Extensions of Maximum Flow

18 Extensions of Maximum Flow Who are you?" aid Lunkwill, riing angrily from hi ea. Wha do you wan?" I am Majikhie!" announced he older one. And I demand ha I am Vroomfondel!" houed he younger one. Majikhie urned on Vroomfondel. I

More information

arxiv: v7 [q-fin.pm] 20 Mar 2019

arxiv: v7 [q-fin.pm] 20 Mar 2019 Porfolio choice, porfolio liquidaion, and porfolio raniion under drif uncerainy arxiv:161107843v7 [q-finpm] 20 Mar 2019 Alexi Bimuh, Olivier Guéan, Jiang Pu Abrac hi paper preen everal model addreing opimal

More information

Network Flows: Introduction & Maximum Flow

Network Flows: Introduction & Maximum Flow CSC 373 - lgorihm Deign, nalyi, and Complexiy Summer 2016 Lalla Mouaadid Nework Flow: Inroducion & Maximum Flow We now urn our aenion o anoher powerful algorihmic echnique: Local Search. In a local earch

More information

Problem Set If all directed edges in a network have distinct capacities, then there is a unique maximum flow.

Problem Set If all directed edges in a network have distinct capacities, then there is a unique maximum flow. CSE 202: Deign and Analyi of Algorihm Winer 2013 Problem Se 3 Inrucor: Kamalika Chaudhuri Due on: Tue. Feb 26, 2013 Inrucion For your proof, you may ue any lower bound, algorihm or daa rucure from he ex

More information

Mon Apr 2: Laplace transform and initial value problems like we studied in Chapter 5

Mon Apr 2: Laplace transform and initial value problems like we studied in Chapter 5 Mah 225-4 Week 2 April 2-6 coninue.-.3; alo cover par of.4-.5, EP 7.6 Mon Apr 2:.-.3 Laplace ranform and iniial value problem like we udied in Chaper 5 Announcemen: Warm-up Exercie: Recall, The Laplace

More information

Curvature. Institute of Lifelong Learning, University of Delhi pg. 1

Curvature. Institute of Lifelong Learning, University of Delhi pg. 1 Dicipline Coure-I Semeer-I Paper: Calculu-I Leon: Leon Developer: Chaianya Kumar College/Deparmen: Deparmen of Mahemaic, Delhi College of r and Commerce, Univeriy of Delhi Iniue of Lifelong Learning, Univeriy

More information

1 Motivation and Basic Definitions

1 Motivation and Basic Definitions CSCE : Deign and Analyi of Algorihm Noe on Max Flow Fall 20 (Baed on he preenaion in Chaper 26 of Inroducion o Algorihm, 3rd Ed. by Cormen, Leieron, Rive and Sein.) Moivaion and Baic Definiion Conider

More information

A Risk-Averse Insider and Asset Pricing in Continuous Time

A Risk-Averse Insider and Asset Pricing in Continuous Time Managemen Science and Financial Engineering Vol 9, No, May 3, pp-6 ISSN 87-43 EISSN 87-36 hp://dxdoiorg/7737/msfe39 3 KORMS A Rik-Avere Inider and Ae Pricing in oninuou Time Byung Hwa Lim Graduae School

More information

FIXED POINTS AND STABILITY IN NEUTRAL DIFFERENTIAL EQUATIONS WITH VARIABLE DELAYS

FIXED POINTS AND STABILITY IN NEUTRAL DIFFERENTIAL EQUATIONS WITH VARIABLE DELAYS PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 136, Number 3, March 28, Page 99 918 S 2-9939(7)989-2 Aricle elecronically publihed on November 3, 27 FIXED POINTS AND STABILITY IN NEUTRAL DIFFERENTIAL

More information

1 Review of Zero-Sum Games

1 Review of Zero-Sum Games COS 5: heoreical Machine Learning Lecurer: Rob Schapire Lecure #23 Scribe: Eugene Brevdo April 30, 2008 Review of Zero-Sum Games Las ime we inroduced a mahemaical model for wo player zero-sum games. Any

More information

An ergodic BSDE approach to forward entropic risk measures: representation and large-maturity behavior

An ergodic BSDE approach to forward entropic risk measures: representation and large-maturity behavior An ergodic BSDE approach o forward enropic rik meaure: repreenaion and large-mauriy behavior W. F. Chong Y. Hu G. Liang. Zariphopoulou Fir verion: January 2016; hi verion: November 30, 2016 Abrac Uing

More information

On the Exponential Operator Functions on Time Scales

On the Exponential Operator Functions on Time Scales dvance in Dynamical Syem pplicaion ISSN 973-5321, Volume 7, Number 1, pp. 57 8 (212) hp://campu.m.edu/ada On he Exponenial Operaor Funcion on Time Scale laa E. Hamza Cairo Univeriy Deparmen of Mahemaic

More information

FLAT CYCLOTOMIC POLYNOMIALS OF ORDER FOUR AND HIGHER

FLAT CYCLOTOMIC POLYNOMIALS OF ORDER FOUR AND HIGHER #A30 INTEGERS 10 (010), 357-363 FLAT CYCLOTOMIC POLYNOMIALS OF ORDER FOUR AND HIGHER Nahan Kaplan Deparmen of Mahemaic, Harvard Univeriy, Cambridge, MA nkaplan@mah.harvard.edu Received: 7/15/09, Revied:

More information

Pricing the American Option Using Itô s Formula and Optimal Stopping Theory

Pricing the American Option Using Itô s Formula and Optimal Stopping Theory U.U.D.M. Projec Repor 2014:3 Pricing he American Opion Uing Iô Formula and Opimal Sopping Theory Jona Bergröm Examenarbee i maemaik, 15 hp Handledare och examinaor: Erik Ekröm Januari 2014 Deparmen of

More information

EE Control Systems LECTURE 2

EE Control Systems LECTURE 2 Copyrigh F.L. Lewi 999 All righ reerved EE 434 - Conrol Syem LECTURE REVIEW OF LAPLACE TRANSFORM LAPLACE TRANSFORM The Laplace ranform i very ueful in analyi and deign for yem ha are linear and ime-invarian

More information

10. State Space Methods

10. State Space Methods . Sae Space Mehods. Inroducion Sae space modelling was briefly inroduced in chaper. Here more coverage is provided of sae space mehods before some of heir uses in conrol sysem design are covered in he

More information

Chapter 6. Laplace Transforms

Chapter 6. Laplace Transforms Chaper 6. Laplace Tranform Kreyzig by YHLee;45; 6- An ODE i reduced o an algebraic problem by operaional calculu. The equaion i olved by algebraic manipulaion. The reul i ranformed back for he oluion of

More information

CHAPTER 7: UNCERTAINTY

CHAPTER 7: UNCERTAINTY Eenial Microeconomic - ecion 7-3, 7-4 CHPTER 7: UNCERTINTY Fir and econd order ochaic dominance 2 Mean preerving pread 8 Condiional ochaic Dominance 0 Monoone Likelihood Raio Propery 2 Coninuou diribuion

More information

5.2 GRAPHICAL VELOCITY ANALYSIS Polygon Method

5.2 GRAPHICAL VELOCITY ANALYSIS Polygon Method ME 352 GRHICL VELCITY NLYSIS 52 GRHICL VELCITY NLYSIS olygon Mehod Velociy analyi form he hear of kinemaic and dynamic of mechanical yem Velociy analyi i uually performed following a poiion analyi; ie,

More information

ARTIFICIAL INTELLIGENCE. Markov decision processes

ARTIFICIAL INTELLIGENCE. Markov decision processes INFOB2KI 2017-2018 Urech Univeriy The Neherland ARTIFICIAL INTELLIGENCE Markov deciion procee Lecurer: Silja Renooij Thee lide are par of he INFOB2KI Coure Noe available from www.c.uu.nl/doc/vakken/b2ki/chema.hml

More information

Main Reference: Sections in CLRS.

Main Reference: Sections in CLRS. Maximum Flow Reied 09/09/200 Main Reference: Secion 26.-26. in CLRS. Inroducion Definiion Muli-Source Muli-Sink The Ford-Fulkeron Mehod Reidual Nework Augmening Pah The Max-Flow Min-Cu Theorem The Edmond-Karp

More information

Optimal Consumption and Investment Portfolio in Jump markets. Optimal Consumption and Portfolio of Investment in a Financial Market with Jumps

Optimal Consumption and Investment Portfolio in Jump markets. Optimal Consumption and Portfolio of Investment in a Financial Market with Jumps Opimal Consumpion and Invesmen Porfolio in Jump markes Opimal Consumpion and Porfolio of Invesmen in a Financial Marke wih Jumps Gan Jin Lingnan (Universiy) College, China Insiue of Economic ransformaion

More information

Inventory Analysis and Management. Multi-Period Stochastic Models: Optimality of (s, S) Policy for K-Convex Objective Functions

Inventory Analysis and Management. Multi-Period Stochastic Models: Optimality of (s, S) Policy for K-Convex Objective Functions Muli-Period Sochasic Models: Opimali of (s, S) Polic for -Convex Objecive Funcions Consider a seing similar o he N-sage newsvendor problem excep ha now here is a fixed re-ordering cos (> 0) for each (re-)order.

More information

Buckling of a structure means failure due to excessive displacements (loss of structural stiffness), and/or

Buckling of a structure means failure due to excessive displacements (loss of structural stiffness), and/or Buckling Buckling of a rucure mean failure due o exceive diplacemen (lo of rucural iffne), and/or lo of abiliy of an equilibrium configuraion of he rucure The rule of humb i ha buckling i conidered a mode

More information

EE202 Circuit Theory II

EE202 Circuit Theory II EE202 Circui Theory II 2017-2018, Spring Dr. Yılmaz KALKAN I. Inroducion & eview of Fir Order Circui (Chaper 7 of Nilon - 3 Hr. Inroducion, C and L Circui, Naural and Sep epone of Serie and Parallel L/C

More information

Physics 235 Chapter 2. Chapter 2 Newtonian Mechanics Single Particle

Physics 235 Chapter 2. Chapter 2 Newtonian Mechanics Single Particle Chaper 2 Newonian Mechanics Single Paricle In his Chaper we will review wha Newon s laws of mechanics ell us abou he moion of a single paricle. Newon s laws are only valid in suiable reference frames,

More information

BSDE Approach to Non-Zero-Sum Stochastic Differential Games of Control and Stopping

BSDE Approach to Non-Zero-Sum Stochastic Differential Games of Control and Stopping BSDE Approach o Non-Zero-Sum Sochaic Differenial Game of Conrol and Sopping Ioanni Karaza INTECH Invemen Managemen One Palmer Square, Suie 441 Princeon, NJ 8542 ik@enhanced.com Qinghua Li Saiic Deparmen,

More information

ON FRACTIONAL ORNSTEIN-UHLENBECK PROCESSES

ON FRACTIONAL ORNSTEIN-UHLENBECK PROCESSES Communicaion on Sochaic Analyi Vol. 5, No. 1 211 121-133 Serial Publicaion www.erialpublicaion.com ON FRACTIONAL ORNSTEIN-UHLENBECK PROCESSES TERHI KAARAKKA AND PAAVO SALMINEN Abrac. In hi paper we udy

More information

E β t log (C t ) + M t M t 1. = Y t + B t 1 P t. B t 0 (3) v t = P tc t M t Question 1. Find the FOC s for an optimum in the agent s problem.

E β t log (C t ) + M t M t 1. = Y t + B t 1 P t. B t 0 (3) v t = P tc t M t Question 1. Find the FOC s for an optimum in the agent s problem. Noes, M. Krause.. Problem Se 9: Exercise on FTPL Same model as in paper and lecure, only ha one-period govenmen bonds are replaced by consols, which are bonds ha pay one dollar forever. I has curren marke

More information

FULLY COUPLED FBSDE WITH BROWNIAN MOTION AND POISSON PROCESS IN STOPPING TIME DURATION

FULLY COUPLED FBSDE WITH BROWNIAN MOTION AND POISSON PROCESS IN STOPPING TIME DURATION J. Au. Mah. Soc. 74 (23), 249 266 FULLY COUPLED FBSDE WITH BROWNIAN MOTION AND POISSON PROCESS IN STOPPING TIME DURATION HEN WU (Received 7 Ocober 2; revied 18 January 22) Communicaed by V. Sefanov Abrac

More information

CHAPTER. Forced Equations and Systems { } ( ) ( ) 8.1 The Laplace Transform and Its Inverse. Transforms from the Definition.

CHAPTER. Forced Equations and Systems { } ( ) ( ) 8.1 The Laplace Transform and Its Inverse. Transforms from the Definition. CHAPTER 8 Forced Equaion and Syem 8 The aplace Tranform and I Invere Tranform from he Definiion 5 5 = b b {} 5 = 5e d = lim5 e = ( ) b {} = e d = lim e + e d b = (inegraion by par) = = = = b b ( ) ( )

More information

Linear Motion, Speed & Velocity

Linear Motion, Speed & Velocity Add Iporan Linear Moion, Speed & Velociy Page: 136 Linear Moion, Speed & Velociy NGSS Sandard: N/A MA Curriculu Fraework (2006): 1.1, 1.2 AP Phyic 1 Learning Objecive: 3.A.1.1, 3.A.1.3 Knowledge/Underanding

More information

Lecture 20: Riccati Equations and Least Squares Feedback Control

Lecture 20: Riccati Equations and Least Squares Feedback Control 34-5 LINEAR SYSTEMS Lecure : Riccai Equaions and Leas Squares Feedback Conrol 5.6.4 Sae Feedback via Riccai Equaions A recursive approach in generaing he marix-valued funcion W ( ) equaion for i for he

More information

CSC 364S Notes University of Toronto, Spring, The networks we will consider are directed graphs, where each edge has associated with it

CSC 364S Notes University of Toronto, Spring, The networks we will consider are directed graphs, where each edge has associated with it CSC 36S Noe Univeriy of Torono, Spring, 2003 Flow Algorihm The nework we will conider are direced graph, where each edge ha aociaed wih i a nonnegaive capaciy. The inuiion i ha if edge (u; v) ha capaciy

More information

as representing the flow of information over time, with

as representing the flow of information over time, with [Alraheed 4(4): April 15] ISSN: 77-9655 Scienific Journal Impac Facor: 3.449 (ISRA) Impac Facor:.114 IJSR INRNAIONAL JOURNAL OF NGINRING SCINCS & RSARCH CHNOLOGY H FINANCIAL APPLICAIONS OF RANDOM CONROL

More information

Optimal Investment, Consumption and Retirement Decision with Disutility and Borrowing Constraints

Optimal Investment, Consumption and Retirement Decision with Disutility and Borrowing Constraints Opimal Invesmen, Consumpion and Reiremen Decision wih Disuiliy and Borrowing Consrains Yong Hyun Shin Join Work wih Byung Hwa Lim(KAIST) June 29 July 3, 29 Yong Hyun Shin (KIAS) Workshop on Sochasic Analysis

More information

6.302 Feedback Systems Recitation : Phase-locked Loops Prof. Joel L. Dawson

6.302 Feedback Systems Recitation : Phase-locked Loops Prof. Joel L. Dawson 6.32 Feedback Syem Phae-locked loop are a foundaional building block for analog circui deign, paricularly for communicaion circui. They provide a good example yem for hi cla becaue hey are an excellen

More information

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation

Course Notes for EE227C (Spring 2018): Convex Optimization and Approximation Course Noes for EE7C Spring 018: Convex Opimizaion and Approximaion Insrucor: Moriz Hard Email: hard+ee7c@berkeley.edu Graduae Insrucor: Max Simchowiz Email: msimchow+ee7c@berkeley.edu Ocober 15, 018 3

More information

Modeling the Evolution of Demand Forecasts with Application to Safety Stock Analysis in Production/Distribution Systems

Modeling the Evolution of Demand Forecasts with Application to Safety Stock Analysis in Production/Distribution Systems Modeling he Evoluion of Demand oreca wih Applicaion o Safey Sock Analyi in Producion/Diribuion Syem David Heah and Peer Jackon Preened by Kai Jiang Thi ummary preenaion baed on: Heah, D.C., and P.L. Jackon.

More information

Admin MAX FLOW APPLICATIONS. Flow graph/networks. Flow constraints 4/30/13. CS lunch today Grading. in-flow = out-flow for every vertex (except s, t)

Admin MAX FLOW APPLICATIONS. Flow graph/networks. Flow constraints 4/30/13. CS lunch today Grading. in-flow = out-flow for every vertex (except s, t) /0/ dmin lunch oday rading MX LOW PPLIION 0, pring avid Kauchak low graph/nework low nework direced, weighed graph (V, ) poiive edge weigh indicaing he capaciy (generally, aume ineger) conain a ingle ource

More information

Chapter 3 Boundary Value Problem

Chapter 3 Boundary Value Problem Chaper 3 Boundary Value Problem A boundary value problem (BVP) is a problem, ypically an ODE or a PDE, which has values assigned on he physical boundary of he domain in which he problem is specified. Le

More information

Vehicle Arrival Models : Headway

Vehicle Arrival Models : Headway Chaper 12 Vehicle Arrival Models : Headway 12.1 Inroducion Modelling arrival of vehicle a secion of road is an imporan sep in raffic flow modelling. I has imporan applicaion in raffic flow simulaion where

More information

Research Article An Upper Bound on the Critical Value β Involved in the Blasius Problem

Research Article An Upper Bound on the Critical Value β Involved in the Blasius Problem Hindawi Publihing Corporaion Journal of Inequaliie and Applicaion Volume 2010, Aricle ID 960365, 6 page doi:10.1155/2010/960365 Reearch Aricle An Upper Bound on he Criical Value Involved in he Blaiu Problem

More information

Let. x y. denote a bivariate time series with zero mean.

Let. x y. denote a bivariate time series with zero mean. Linear Filer Le x y : T denoe a bivariae ime erie wih zero mean. Suppoe ha he ime erie {y : T} i conruced a follow: y a x The ime erie {y : T} i aid o be conruced from {x : T} by mean of a Linear Filer.

More information

Diebold, Chapter 7. Francis X. Diebold, Elements of Forecasting, 4th Edition (Mason, Ohio: Cengage Learning, 2006). Chapter 7. Characterizing Cycles

Diebold, Chapter 7. Francis X. Diebold, Elements of Forecasting, 4th Edition (Mason, Ohio: Cengage Learning, 2006). Chapter 7. Characterizing Cycles Diebold, Chaper 7 Francis X. Diebold, Elemens of Forecasing, 4h Ediion (Mason, Ohio: Cengage Learning, 006). Chaper 7. Characerizing Cycles Afer compleing his reading you should be able o: Define covariance

More information

Analytical Pricing of An Insurance Embedded Option: Alternative Formulas and Gaussian Approximation

Analytical Pricing of An Insurance Embedded Option: Alternative Formulas and Gaussian Approximation Journal of Informaic and Mahemaical Science Volume 3 (0), Number, pp. 87 05 RGN Publicaion hp://www.rgnpublicaion.com Analyical Pricing of An Inurance Embedded Opion: Alernaive Formula and Gauian Approximaion

More information

Network Flows UPCOPENCOURSEWARE number 34414

Network Flows UPCOPENCOURSEWARE number 34414 Nework Flow UPCOPENCOURSEWARE number Topic : F.-Javier Heredia Thi work i licened under he Creaive Common Aribuion- NonCommercial-NoDeriv. Unpored Licene. To view a copy of hi licene, vii hp://creaivecommon.org/licene/by-nc-nd/./

More information

QoS-Oriented Distributed Opportunistic Scheduling for Wireless Networks with Hybrid Links

QoS-Oriented Distributed Opportunistic Scheduling for Wireless Networks with Hybrid Links Globecom 2013 - Wirele Neworking Sympoium QoS-Oriened Diribued Opporuniic Scheduling for Wirele Nework wih Hybrid Link Wenguang Mao, Shanhan Wu, and Xudong Wang UM-SJU Join Iniue, Shanghai Jiao ong Univeriy,

More information

On the Benney Lin and Kawahara Equations

On the Benney Lin and Kawahara Equations JOURNAL OF MATEMATICAL ANALYSIS AND APPLICATIONS 11, 13115 1997 ARTICLE NO AY975438 On he BenneyLin and Kawahara Equaion A Biagioni* Deparmen of Mahemaic, UNICAMP, 1381-97, Campina, Brazil and F Linare

More information

Additional Methods for Solving DSGE Models

Additional Methods for Solving DSGE Models Addiional Mehod for Solving DSGE Model Karel Meren, Cornell Univeriy Reference King, R. G., Ploer, C. I. & Rebelo, S. T. (1988), Producion, growh and buine cycle: I. he baic neoclaical model, Journal of

More information

Price of Stability and Introduction to Mechanism Design

Price of Stability and Introduction to Mechanism Design Algorihmic Game Theory Summer 2017, Week 5 ETH Zürich Price of Sabiliy and Inroducion o Mechanim Deign Paolo Penna Thi i he lecure where we ar deigning yem which involve elfih player. Roughly peaking,

More information

A Portfolio Decomposition Formula

A Portfolio Decomposition Formula arxiv:mah/72726v1 [mah.pr] 24 Feb 27 A Porfolio Decompoiion Formula Traian A. Pirvu Deparmen of Mahemaic The Univeriy of Briih Columbia Vancouver, BC, V6T1Z2 pirvu@mah.ubc.ca Ulrich G. Haumann 1 Deparmen

More information

Physics 240: Worksheet 16 Name

Physics 240: Worksheet 16 Name Phyic 4: Workhee 16 Nae Non-unifor circular oion Each of hee proble involve non-unifor circular oion wih a conan α. (1) Obain each of he equaion of oion for non-unifor circular oion under a conan acceleraion,

More information

Flow networks. Flow Networks. A flow on a network. Flow networks. The maximum-flow problem. Introduction to Algorithms, Lecture 22 December 5, 2001

Flow networks. Flow Networks. A flow on a network. Flow networks. The maximum-flow problem. Introduction to Algorithms, Lecture 22 December 5, 2001 CS 545 Flow Nework lon Efra Slide courey of Charle Leieron wih mall change by Carola Wenk Flow nework Definiion. flow nework i a direced graph G = (V, E) wih wo diinguihed verice: a ource and a ink. Each

More information

ANALYSIS OF SOME SAFETY ASSESSMENT STANDARD ON GROUNDING SYSTEMS

ANALYSIS OF SOME SAFETY ASSESSMENT STANDARD ON GROUNDING SYSTEMS ANAYSIS OF SOME SAFETY ASSESSMENT STANDARD ON GROUNDING SYSTEMS Shang iqun, Zhang Yan, Cheng Gang School of Elecrical and Conrol Engineering, Xi an Univeriy of Science & Technology, 710054, Xi an, China,

More information

Investment and valuation under backward and forward dynamic exponential utilities in a stochastic factor model

Investment and valuation under backward and forward dynamic exponential utilities in a stochastic factor model Invemen and valuaion under backward and forward dynamic exponenial uiliie in a ochaic facor model Marek Muiela and Thaleia Zariphopoulou BNP Pariba, London and The Univeriy of Texa a uin Fir verion: January

More information

OPTIMAL INVESTMENT AND CONSUMPTION IN A BLACK SCHOLES MARKET WITH LÉVY-DRIVEN STOCHASTIC COEFFICIENTS

OPTIMAL INVESTMENT AND CONSUMPTION IN A BLACK SCHOLES MARKET WITH LÉVY-DRIVEN STOCHASTIC COEFFICIENTS The Annal of Applied Probabiliy 2008, Vol. 18, No. 3, 879 908 DOI: 10.1214/07-AAP475 Iniue of Mahemaical Saiic, 2008 OPTIMAL INVESTMENT AND CONSUMPTION IN A BLACK SCHOLES MARKET WITH LÉVY-DRIVEN STOCHASTIC

More information

STATE-SPACE MODELLING. A mass balance across the tank gives:

STATE-SPACE MODELLING. A mass balance across the tank gives: B. Lennox and N.F. Thornhill, 9, Sae Space Modelling, IChemE Process Managemen and Conrol Subjec Group Newsleer STE-SPACE MODELLING Inroducion: Over he pas decade or so here has been an ever increasing

More information

Testing for a Single Factor Model in the Multivariate State Space Framework

Testing for a Single Factor Model in the Multivariate State Space Framework esing for a Single Facor Model in he Mulivariae Sae Space Framework Chen C.-Y. M. Chiba and M. Kobayashi Inernaional Graduae School of Social Sciences Yokohama Naional Universiy Japan Faculy of Economics

More information

CONTROL SYSTEMS. Chapter 10 : State Space Response

CONTROL SYSTEMS. Chapter 10 : State Space Response CONTROL SYSTEMS Chaper : Sae Space Repone GATE Objecive & Numerical Type Soluion Queion 5 [GATE EE 99 IIT-Bombay : Mark] Conider a econd order yem whoe ae pace repreenaion i of he form A Bu. If () (),

More information

Approximation for Option Prices under Uncertain Volatility

Approximation for Option Prices under Uncertain Volatility Approximaion for Opion Price under Uncerain Volailiy Jean-Pierre Fouque Bin Ren February, 3 Abrac In hi paper, we udy he aympoic behavior of he wor cae cenario opion price a he volailiy inerval in an uncerain

More information

6. Stochastic calculus with jump processes

6. Stochastic calculus with jump processes A) Trading sraegies (1/3) Marke wih d asses S = (S 1,, S d ) A rading sraegy can be modelled wih a vecor φ describing he quaniies invesed in each asse a each insan : φ = (φ 1,, φ d ) The value a of a porfolio

More information