PORTFOLIO OPTIMIZATION WHEN ASSET RETURNS HAVE THE GAUSSIAN MIXTURE DISTRIBUTION

Size: px
Start display at page:

Download "PORTFOLIO OPTIMIZATION WHEN ASSET RETURNS HAVE THE GAUSSIAN MIXTURE DISTRIBUTION"

Transcription

1 PORTFOLIO OPTIMIZATION WHEN ASSET RETURNS HAVE THE GAUSSIAN MIXTURE DISTRIBUTION IAN BUCKLEY, GUSTAVO COMEZAÑA, BEN DJERROUD, AND LUIS SECO Abstact. Potfolios of assets whose etuns have the Gaussian mixtue distibution ae optimized in the static setting to find potfolio weights and efficient fonties using the pobability of outpefoming a taget etun and Hodges modified Shape atio objective functions. The sensitivities of optimal potfolio weights to the pobability of the maket being in the distessed egime ae shown to give valuable diagnostic infomation. A two-stage optimization pocedue is pesented in which the high-dimensional non-linea optimization poblem can be decomposed into a elated quadatic pogamming poblem, coupled to a lowe-dimensional non-linea poblem. Contents 1. Intoduction 2 2. Evidence of covaiance egimes 3 3. Gaussian mixtue distibution Definitions and identities Visualizing the GM distibution 5 4. Potfolio optimization Pobability of shotfall as a isk measue Hodges atio Lowe patial moments Optimization poblems fo diffeent objectives Investment oppotunity set Numeical examples Altenative algoithm fo solving the non-linea poblem Conclusions 17 Acknowledgment 17 Appendix A. Gaussian mixtue distibution 18 A.1. Definitions 18 A.2. Moments 18 A.3. Linea combinations of andom vaiables with the GM distibution 20 Refeences 20 Date: Tuesday, Febuay 18, 2003 at 20:32 and, in evised fom,... Key wods and phases. Mixtue of nomals distibution, Gaussian mixtue distibution, potfolio optimization, maket distess, hedge fund potfolio, Shape atio, efficient fontie, Hodges modified Shape atio, exponential utility, coelation switching, egime switching, asset allocation, commodity tading adviso, pobability of shotfall, distess sensitivities. This wok was completed with the suppot of... Scholaship. 1

2 2 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO 1. Intoduction In this aticle Makowitz mean-vaiance potfolio theoy [12], the foundation fo single-peiod investment theoy, is genealized to descibe potfolios of assets whose etuns ae descibed by the (finite) Gaussian mixtue (GM) (altenatively mixtue of nomals) distibution. Whilst the assets in the univese could be of the conventional vaiety, such as equities o bonds, ou pimay goal is to develop a famewok which lends itself to the management of potfolios of hedge funds o even fo optimally combining the ecommendations fom a goup of commodity tading advisos (CTAs). That is to say, we seek an appoach suitable fo finding an optimal fund of funds. Because of the infinite vaiety of hedge fund and CTA stategies and the speed at which a given fund s composition can be changed, the altenative assets that we descibe ae not expected to behave like conventional assets such as individual equities, bonds o even long-only funds such as index tacke funds and exchange taded funds (ETFs). Indeed we need to be pepaed fo thei pices to be less pedictable, moe volatile and to have moe exotic distibutions. The new appoach is ideal fo an industial setting, poviding consideable additional flexibility ove and above a standad Makowitz appoach, with only a modest incease in complexity. The assumption that asset etuns have the multivaiate Gaussian distibution is a easonable fist appoximation to eality and gives ise to tactable theoies. Many theoies foming the foundations of mathematical finance adopt this conjectue, including Black-Scholes-Meton option picing theoy, Makowitz potfolio theoy, and the CAPM and APT equity picing models. Howeve, it is well-known that fo assets, both in the conventional sense of equities and bonds, but also in a boade sense, fo example in the fom of county o secto-based equity o bond indices, and even moe so fo altenative investments such as hedge funds and CTAs, the situation is moe complex. The pupose of the genealization descibed in this pape is to addess two well-known limitations with the assumption that asset etuns obey the multivaiate Gaussian distibution with constant paametes ove time: The skewed (asymmetic aound the mean) and leptokutotic (moe kutotic o fat-tailed than a Gaussian distibution) natue of maginal pobability density functions (pdfs) The asymmetic coelation (o coelation beakdown) phenomenon, which descibes the tendency fo the coelations between asset etuns to be dependent on the pevailing diection of the maket. Typically coelations ae lage in a bea maket than a bull maket. The fist point efes to the univaiate distibutions fo etuns that ae obseved if assets ae consideed one at a time. The second point descibes effects that can only be obseved when the etuns to multiple assets ae investigated togethe. One way to captue the dependence stuctue of multiple andom vaiables in a isk management setting is by using copulas [22], [7], [9], [13], [18]. Copulas descibe that pat of the shape of the pdf that cannot be descibed by the maginal distibutions. A (finite) Gaussian mixtue distibution, as descibed in this pape, can be used to appoximate a geneal multivaiate distibution, as equivalently expessed eithe as a pdf o decomposed into a seies of univaiate maginal distibutions and a copula.

3 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 3 In fact, stictly the tem asymmetic coelation descibes a poposed explanation fo a phenomenon, athe than the undelying, fundamental phenomenon itself. The latte is simply that the iso-pobability contous of the multivaiate pdf fo asset etuns ae less symmetic than the ellipsoidal contous of the multivaiate nomal distibution. Apat fom those distibutions with pdfs with ellipsoidal iso-pobability sufaces, such as the Gaussian and multivaiate t, fo all othe multivaiate distibutions the coelation matix does not povide an adequate summay of the dependence stuctue of the constituent isks. Consideing multiple egimes within which the odinay constant paamete Gaussian assumptions pevail successfully gives ise to a model that eflects eality bette than a standad Gaussian model, without having to depat too adically fom it. Howeve, it is cetainly not the only way to constuct a model with non-ellipsoidal iso-pobability contous. Indeed altenative assumptions to achieve the same ends may ende the concept of coelation edundant and statements about its evolution ove time, meaningless. Apat fom the distibutional assumptions, the second fundamental diffeence between the novel appoach descibed in this pape (the GM appoach) and the Makowitz mean-vaiance appoach, is that we adopt a diffeent objective function. This is essential in ode to get diffeent optimal potfolio weights. When we efe to the GM appoach, the use of an altenative objective will be implicit. A dawback of using a moe exotic objective than the vaiance is that the esultant optimization poblem is not a linea-quadatic pogam (LQP), as it is when vaiance is the chosen isk measue. Howeve, the nonlinea optimization poblem with multiple local extema that eplaces it can be solved eliably and quickly with commonly available outines, at least with a modeate numbe of assets (i.e. less than ten). The plan fo the pape is as follows. Evidence of covaiance egimes ove time is given in Section 2, including a summay of existing esults fom the liteatue fo conventional assets and oiginal esults fo altenative investments. We ae intoduced to the GM distibution in Section 3 in which we find definitions of GM distibuted andom vaiables, estimation issues, and some identities concening moments and linea combinations of GM vaiables. In Section 4 we take the pobability of shotfall objective, assume GM distibuted asset etuns and conside a theoy based on these ingedients. Key theoetical and numeical esults ae descibed. Section 5 contains ou conclusions. 2. Evidence of covaiance egimes Duing the aftemath of the 1987 maket cash, a deficiency in the isk models based on the multivaiate nomal distibution eceived inceased attention: a simultaneous downwad movement in all the makets of the wold was a moe fequent occuence than the models pedicted when calibated using asset etuns obseved duing tanquil peiods. Divesifying amongst diffeent assets o makets was less effective at educing isk than many paticipants had hitheto believed. Impotant investigations of the contagion phenomenon include [8], [4], [20]. In common with thei conventional asset countepats, altenative assets exhibit the coelation beakdown phenomenon. As evidence we pesent Figue 1, which shows the coelation matices between hedge fund etuns in tanquil and distessed egimes. Duing tanquil peiods coelations ae small, wheeas duing peiods of maket distess, the asset etuns become highly coelated, with off-diagonal coelation values close to one.

4 4 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Figue 1. Ba chats to show typical values of coelations between the etuns of a goup of eight hedge funds duing tanquil (left) and distessed (ight) peiods. We note in passing that because finding potfolio weights in a mean-vaiance setting is tantamount to inveting the covaiance matix, then eos in the optimal potfolio weights will be most sensitive to eos in the covaiance matix when the latte is closest to being singula (having a deteminant close to zeo). The covaiance matix will be moe singula in the distessed egime than in the tanquil egime. 3. Gaussian mixtue distibution The Gaussian mixtue distibution is selected fom the ange of paametic altenatives to the nomal distibution fo its tactability: calculations using it often closely esemble those using the nomal distibution. Whilst thee ae many univaiate paametic pobability distibutions, (e.g. hypebolic, t, genealized beta, α-stable) the list fo multivaiate distibutions is shote (e.g. t, α-stable). The GM distibution has been used befoe in the field of finance, mostly in its univaiate guise fo the estimation of Value at Risk (VaR). [16] develop a model fo estimating VaR in which the use is fee to choose any pobability distibution fo the daily changes in each maket vaiable and employ the univaiate mixtue of nomals distibution as an example. In the same field, [28] assumes pobability distibutions fo each of the paametes descibing the mixtue of nomals and uses a Bayesian updating scheme; and [26] uses a quasi-bayesian maximum likelihood estimation pocedue. The cuent RiskMetics TM methodology uses GM with a mixtue of two nomal distibutions. Moe ecently GM models have been used [19] to model futues makets and fo potfolio isk management and by [10] fo cedit isk. [27] develops an efficient analytical Monte Calo method fo geneating changes in asset pices using a multivaiate mixtue of nomal distibutions with abitay covaiance matix. [25] descibe computational tools fo the calculation of VaR and othe moe sophisticated isk measues such as shotfall, Max-VaR,

5 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 5 conditional VaR and conditional isk measues that aim to take account of the heteoskedastic stuctue of time seies. This pape descibes the static, single peiod setting only, in which distibutions of andom vaiables ae sufficient to specify the model. In this setting the key idea is that we build an exotic distibution by mixing simple ones, namely copies of the nomal distibution. Ou only assumption is that ove an inteval of time, the etuns to the assets in the univese ae descibed by the multivaiate GM distibution. It is unnecessay to make futhe assumptions about the natue of the asset pice evolution duing this inteval. Howeve, we do maintain an inteest in the dynamic case, i.e. the multiple peiod discete o continuous time setting, because we wish to motivate the use of the GM distibution and we pefe to constuct static models that extend natually to the dynamic case. Thee is a gowing body of wok in which exotic (asset etun) stochastic pocesses have been constucted by mixing simple ones. Pocesses fo asset etuns may be constucted fom unconditional o conditional distibutions. As an example of the latte case, by mixing autoegessive pocesses such as ARCH and GARCH, pocesses can be constucted that can account fo both the heteoscedastic and leptokutic natue of financial time seies. See [23] and [11]. GM distibutions can aise natually as the level of cetain stochastic pocesses at a point in time, conditional on the level at an ealie time e.g. Makov (egime) switching models and jump pocesses. Regime switching models descibe pocesses in which paametes of a continuous diffusion pocess may change discontinuously accoding to the ealized stochastic path though an associated Makov chain. We conclude that GM distibutions ae bette motivated and less contived than fist impessions might suggest. Recent applications of egime switching asset etun pocesses include: in the field of Meton-style option picing theoy, [17], and in potfolio management [5] (CAPM ) and [2], [1] (intenational divesification). Mixtue distibutions have the appeal that by adding togethe a sufficient numbe of component distibutions, any multivaiate distibution may be appoximated to abitay accuacy. With an infinite numbe of contibutions, any distibution can be econstucted exactly. As fa as estimation is concened, a disadvantage of using the GM distibution is that the log-likelihood function does not have a global minimum. A esolution to this poblem exploed in [14] is to use a modified log likelihood function. Because of the use of the GM distibution and othe mixtue distibutions in image pocessing, clusteing, and unsupevised leaning a host of estimation techniques have been developed to addess this poblem [6]. When using the GM distibution to model asset etuns, [27] employs the EM algoithm Definitions and identities. In Appendix A the definition of the GM distibution and vaious identities involving it ae pesented Visualizing the GM distibution. In Figue 2, in fou contou plots of pdfs, we see how two bivaiate Gaussian distibutions (top ow) can be added to yield a GM distibution (bottom, left). Note the potential fo highly non-elliptic iso-pobability contous when the GM distibution is used. Fo compaison a bivaiate nomal with the same sample means, vaiances and covaiances is included (bottom, ight). If the GM distibution given wee used to descibe the etuns to two assets, the lobe pointing down and to the left of the figue would descibe the

6 6 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Figue 2. Contou plots of pobability density functions. The top ow contains two bivaiate Gaussian distibutions - potentially fo the tanquil (left) and distessed (ight) egimes. The bottom ollustates the composite Gaussian mixtue distibution obtained by mixing the two distibutions fom the top ow (left) and a bivaiate nomal distibution with the same means and vaiancecovaiance matix as the composite (ight). popensity of the two assets to decline shaply togethe (the asymmetic coelation phenomenon). The Gaussian distibution, with its elliptic contous, is clealy unable to captue this featue. These examples illustate the two asset case. With thee assets the contous fo the Gaussian distibution ae thee-dimensional ellipsoids (athe than ellipses in two dimensions), and fo the GM distibution the contous ae complicated lobed sufaces embedded in a thee-dimensional space. 4. Potfolio optimization In ou famewok, the n Gaussian contibutions to the GM distibution ae associated with n egimes. In all the numeical examples, we take n = 2, and call the two egimes the tanquil and distessed egimes Pobability of shotfall as a isk measue. When vaiance is used as the optimization objective, the efficient fontie is independent of the distibution fo

7 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 7 the asset etuns. Theefoe, to achieve non-tivial esults in the GM setting, we adopt altenative, non-quadatic objective functions. One with the advantage of being intuitive fo pactitiones is the pobability of shotfall below a taget etun (PoS). When etuns have a nomal distibution, minimizing the PoS is equivalent to maximizing the out-pefomance Shape atio (atio of diffeence between ealized etun and taget etun in the numeato to volatility in the denominato). Theefoe adopting the GM appoach will not epesent a significant depatue fom cuent pactice fo many investos. Note that the PoS is not the same as the VaR: the fome is the pobability beyond a given point on the distibution (e.g. a taget etun of 18.2%), the latte is the point on the distibution such that the pobability of being beyond that point is equal to a given value (e.g. 1% o 5%). In the GM setting of this pape the pobability of shotfall objective is the pobability that a univaiate mixtue of nomals andom vaiable, with egime means µ i = µ i.θ and egime vaiances σ 2 i = θ.v i.θ, exceeds the taget etun k. Fom the expession fo the univaiate mixtue of nomals CDF Eqn. A.1.2, above, it can be shown that: Poposition The pobability that the potfolio etun falls shot of the taget k is: ( ) µ (4.1.1) F k (θ) = Φ i.θ k θ.v i.θ 4.2. Hodges atio. To addess the paadoxes inheent in using the Shape atio [24] as a measue fo anking the desiability of payoff distibutions, Hodges [15] intoduced a measue of isk fo pefomance anking based on the exponential utility function (4.2.1) U(w) = e λw, which is intuitive and is a genealization of the Shape atio, educing to it fo nomally distibuted etuns. We efe to Hodges modified Shape atio as the Hodges atio (HR). Because the HR uses a utility function with constant absolute isk avesion, the composition of the optimal potfolio is independent of the coefficient of isk toleance and so without loss of geneality this can be taken to be one. It educes to the Shape atio fo nomally distibuted etuns and yet can be applied to any etun distibution. It is compatible with stochastic dominance: an investment oppotunity that outpefoms anothe in evey state of the wold necessaily has a highe HR. This is one of the popeties of a coheent isk measue [3]. Madan and McPhail [21] point out an appaent limitation with the HR, namely that it sometimes pevesely consides distibutions with lage negative skew to be desiable investments, because the appoach effectively shots the isky asset. It tuns out that this poblem is simply esolved by scaling the utility function by the sign of the weight in the isky assets, ξ. In this pape, fo all sensible paamete values ξ > 0, so this efinement was unnecessay Lowe patial moments. Patial moments of a andom vaiable ae its moments ove a sub-inteval of the domain of the pdf. In paticula, the lowe and uppe nth-patial moments (LPM, UPM) ae the nth ode moments fo the andom vaiable when it takes nonzeo values below and above a given value, espectively.

8 8 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Necessaily, as a statistic it summaises the popeties of the distibution only to one side of the paamete. The LPM is a good candidate fo a isk measue because it only consides those states in which etun on an asset is below a pe-specified taget ate. To eal individuals a downside isk measue such as this coesponds moe closely to thei concept of isk than the ubiquitous vaiance isk measue. We pesent some useful identities: (4.3.1) (4.3.2) (4.3.3) (4.3.4) (4.3.5) a a a (a x) φ(x) dx = a Φ(a) + φ(a) = a Φ(x) dx (a x) 2 φ(x) dx = ( 1 + a 2) Φ(a) + a φ(a) a x σ φ µ,σ (x) dx = a µ σ Φ µ,σ (a) + σ φ µ,σ (a) ( 1 a σ 2 (a x) 2 φ µ,σ (x) dx = (a µ) φ µ,σ (a) ) (a µ)2 σ 2 Φ µ,σ (a) 4.4. Optimization poblems fo diffeent objectives. In each case µ T is the taget potfolio expected etun. Makowitz: (4.4.1) min θ θ.v S.θ s.t. θ.1 = 1 µ S.θ µ T whee µ S = n µ i and V S ae the sample mean and sample vaiancecovaiance matix, espectively. Note that the latte is not equal to the weighted sum of the egime vaiances. Instead it is calculated using the identity in Equation A.2.5 o equivalently A.2.6. Moments of GM distibutions ae the sums of the component distibution moments. Howeve, fo cental moments, such as the vaiance, the expessions ae moe complicated. Shape Ratio: (4.4.2) µ max S.θ k θ V S s.t. θ.1 = 1 µ S.θ µ T

9 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 9 (4.4.3) (4.4.4) (4.4.5) (4.4.6) Pobability of shotfall: Hodges atio: max θ,ξ max θ ( ) µ Φ i.θ k θ.v i.θ s.t. θ.1 = 1 µ i.θ µ nx T i. exp λ ξ µ i.θ 1 2 λ2 ξ 2.V s.t. θ.1 = 1 µ i.θ µ T whee ξ is the weight in the isky assets and λ is the paamete fo the exponential utility function. The emainde, 1 ξ is invested in a isk-fee bank account, taken to have zeo etun fo simplicity. The above optimization poblem is clealy equivalent to a simila one in which ξ = 1 and the budget constaint is emoved. Howeve, in the numeical examples, both the total weight in isky assets vaiable, ξ, and the budget constaint, ae etained so that the nomalised weights of the individual assets within the isky asset potion can be compaed with the esults fom othe objective functions. E.g., see 6. Note that the exponential tem in the objective can be deived by obseving that if the andom vaiable X has mean µ and vaiance V then ( ( E[U(ξX)] = exp λ ξ µ 1 )) 2 λ2 ξ 2 V whee the utility function U(x) = e λ x. Lowe patial moment, ode n = 1: ( ) k µi max Φ µi,σ θ σ i (k) + σ i φ µi,σ i (k) i s.t. θ.1 = 1 µ i µ T whee µ i = µ i.θ and σ i = θ.v i.θ Lowe patial moment, ode n = 2: ( ) max (k µ i ) φ µi,σ i (k) (k µ i) 2 Φ θ σ 2 µi,σ i (k) i s.t. θ.1 = 1 µ i µ T

10 10 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Figue 3. Scatte plot of potfolios in the investment oppotunity set, ovelaid on the contou plot of the pobability of shotfall objective function. The axes ae the out-pefomance Shape atios in the tanquil (x) and distessed (y) egimes. The GM appoach optimal potfolio is maked in the top ighthand cone, on the efficient fontie. Typically, the potfolios that ae optimal with espect to the mean-vaiance objectives in the tanquil and distessed egimes will be sub-optimal with espect to the pobability of shotfall objective used in the GM appoach. This is a thee asset example Investment oppotunity set. Figue 3 shows the investment oppotunity set (IOS) in the 3-asset case, ovelaid on the contou plot fo the PoS objective function, in tanquil and distessed egime out-pefomance Shape atio space. The investment oppotunity set is the set of all attainable points in a given space that may be eached by constucting potfolios of the assets in a given univese. Typically vaiance and etun ae taken as the dimensions of a space to exploe, but any statistics may be used. The GM appoach optimal potfolio is maked in the top ighthand cone, on the efficient fontie. Compae this with Figue 4, which shows the IOSs fo the tanquil and distessed egimes supeimposed onto the same plot. The axes ae the potfolio means and vaiances in the two egimes. Clealy the potfolio that maximizes the PoS objective, will be suboptimal with espect to the mean-vaiance efficient fontie in each of the component egimes Numeical examples. We pesent a thee asset example with five objectives: Vaiance Pobability of shotfall Hodges atio Lowe patial fist moment

11 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 11 Figue 4. Investment oppotunity sets fo the tanquil and distessed egimes supeimposed onto the same plot. The axes ae the potfolio mean and vaiance. Typically the GM appoach optimal potfolio will be sub-optimal with espect to both the tanquil and distessed mean-vaiance objectives. This is a thee asset example. Lowe patial second moment All esults ae pesented gaphically. Paamete values ae: (4.6.1) (4.6.2) (4.6.3) µ T = σ T = ρ T = µ D = σ D = ρ D = Figue 5 shows 2-dimensional pojections of the GM pdf as contou plots. The contous indicate iso-pobability cuves. The pictues ae slices in the sense that the suppessed dimension has asset weight zeo. Figue 6 shows the optimal potfolio weights along the fontie. All heights on the plot ae optimal objectives fo a given taget expected etun. Figue 7 shows the objective values fo all five objectives based on optimal potfolio weights obtained by using a given, single objective. Fo this example, all optimal potfolio weights ae simila and the five plots coespond closely. Figue 8 shows the dependence of the optimal potfolio weights on the mixing paamete w. Note that pue tanquil (distessed) egime is on the ight (left) of the figue. Theefoe assets with positive (negative) slope on this diagam, i.e., those with a tendency to have small (big) positions in the distessed egime, ae good fo insuing against (speculating on) the maket enteing a distessed state.

12 12 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Figue 5. Contou plots of 2d pojections of 3d asset etun pobability density function Figue 9 shows the dependence of the ate of change of asset weight with espect to mixing paamete along the efficient fontie, i.e., fo diffeent values of desied expected etun Altenative algoithm fo solving the non-linea poblem. The PoS objective gives ise to the non-linea optimization poblem (Equation 4.4.6) of dimension equal to the numbe of assets in the univese, m. This can be solved using standad non-linea pogam (NLP) optimizes, but such tools ae unable to exploit the close esemblance between the non-linea poblem and a linea-quadatic pogam (LQP). An altenative appoach is to ecognize that the optimal potfolio weights, fo a given taget etun µ T, will be embedded in the 2(n 1)-dimensional solution hype-suface of a elated LQP, whee n is the numbe of egimes. The latte is that of minimizing a linea sum of the egime vaiances subject to the linea constaint that a linea sum of the egime expected etuns be less than the etun taget. An NLP is still equied to find the optimal solution, but the dimensionality of the poblem is educed fom m to 2(n 1). Typically the fome exceeds the latte. E.g., if thee ae thity assets in the poblem and two egimes. We define some notation: Linea and quadatic functions of the potfolio weight vecto, θ and the Shape atio, fo each egime i: (4.7.1) (4.7.2) (4.7.3) L i (θ) = µ i.θ Q i (θ) = θ T V i.θ x i (θ) = L i(θ) k Qi (θ)

13 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 13 i. Vaiance ii. Shape Ratio iii. Pobability iv. Hodges MSR v. LPM1 vi. LPM2 Figue 6. Optimal asset weights along the efficient fontie Objective functions, one non-linea, one quadatic in θ: (4.7.4) f α (θ) = α i Φ(x i (θ)) (4.7.5) g β (θ) = whee α i, β i 0 i. β i Q i (θ)) The non-linea and linea-quadatic pogammes ae: Definition NLP(1) (4.7.6) whee α.1 = 1. max f α (θ) θ s.t. θ.1 = 1 α i L i (θ) p Solutions to NLP(1) will be denoted by θ α,p.

14 14 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO Obj i. Vaiance Obj ii. Shape Ratio Obj iii. Pobability Obj iv. Hodges MSR Obj v. LPM1 Obj vi. LPM2 Figue 7. Efficient fonties fo diffeent objectives Definition LQP(2) (4.7.7) min g β (θ) θ s.t. θ.1 = 1 whee β.1 = 1 L i (θ) q i Solutions to LQP(2) will be denoted by θ β,q. The set of solutions fo all possible paamete values ae denoted fo the NLP and LQP espectively as: (4.7.8) Θ NLP = {θ α,p} α R n,α.1=1,p R Θ LQP = {θ β,q} β R n,β.1=1,q R n,q.1=1 Poposition All solutions to NLP(1) ae solutions of LQP(2). That is to say, thee always exist paamete values of β, q in LQP(2), such that the solution θ β,q coesponds to the solution θ α,p to NLP(1), fo all paamete values α, p. (4.7.9) Θ NLP Θ LQP Poof. The poof is by contadiction. The stategy is to conside a solution of NLP(1) that will necessaily satisfy the constaints of LQP(2), and hypothesize the existence of anothe potfolio, which also satisfies the constaints of LQP(2), but

15 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 15 i. Vaiance ii. Shape Ratio l l iii. Pobability iv. Hodges MSR l l v. LPM1 vi. LPM2 l l Figue 8. Dependence of optimal weights on the mixing paamete which gives a smalle objective fo LQP(2) (i.e., is moe optimal) and show that this leads to a contadiction. Given values fo the paametes fo NLP(1), choose specific values fo the LQP(2) paametes: (4.7.10) (4.7.11) (4.7.12) (4.7.13) q i = L i (θ α,p) β i = 1 f α θ N Q i α,p = 1 f α x i θ N x i Q i α,p = 1 2N α iφ(x i ) x i θ Q i α,p whee N is a nomalisation facto to ensue that β.1 = 1. Fo the LQP objective to be convex, we equie that β i 0, and this holds because f α (x) is inceasing in 1 x, whilst Q is deceasing in Q. Due to the choice of the chosen LQP paametes, θ α,p satisfies the constaints of the LQP poblem constaints. What is less obvious is whethe θ α,p is the optimal solution to the poblem.

16 16 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO DS i. Vaiance DS ii. Shape Ratio DS iii. Pobability DS iv. LPM1 DS v. LPM2 Figue 9. Distess sensitivities along the efficient fontie If θ α,p is not the solution to LQP(2), then thee exists a θ = θ α,p +ɛ that gives a smalle value fo the LQP (constained) objective, g β (θ ), with paamete values β(α, p), (objective paametes) and q(α, p) (constaint paametes), than θ α,p. (4.7.14) g β (θ ) < g β (θ α,p ) Taylo s expansion on NLP(1): (4.7.15) f α (θ ) f α (θ α,p ) + = f α (θ α,p ) + n,m i,j=1 n,m i,j=1 ( fα L i + f ) α Q i ɛ j L i θ j Q i θ j θ α,p f α Q i Q i θ j θ α,p ɛ j The tem involving L i vanishes because θ and θ α,p both satisfy the linea constaints of LQP(2) by assumption and design, espectively. Theefoe, (4.7.16) m j=1 L i θ j θ α,p ɛ j = 0.

17 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 17 Similaly Taylo s expansion on LQP(2): (4.7.17) g β (θ ) g β (θ α,p ) + = g β (θ α,p ) + = g β (θ α,p ) 1 N n,m i,j=1 n,m i,j=1 g β Q i Q i θ j θ α,p ɛ j β i Q i θ j θ α,p ɛ j n,m i,j=1 f α Q i Q i θ j θ α,p ɛ j suggesting that n,m f α Q i θ i,j=1 Q i θ j ɛ j > 0 and theefoe that α,p (4.7.18) f α (θ ) > f α (θ α,p ) which contadicts the optimality of θ α,p to NLP(1), which was ou initial assumption. 5. Conclusions We ae cetain that the GM appoach, namely the assumption of a multivaiate finite Gaussian mixtue distibution fo asset etuns used in conjunction with a pobability of shotfall objective, will be useful fo a whole ange of potfolio management applications. It is only slightly hade to implement than standad Makowitz, with many featues in common (e.g. we etain the use of covaiance matices). Howeve, it is moe flexible because of its ability to handle non-elliptic etun distibutions. Because it is intuitive, the technique is unlikely to face esistance fom pactitiones aleady familia with mean-vaiance appoaches. With two egimes, the objective function does not possess moe than two maxima, so ou numeical examples have been obust and quick to solve. We have compaed the GM and mean-vaiance appoaches. The obvious questions to ask ae whethe the two appoach give diffeent optimal weights fom one anothe, and if so, whethe holding the GM weights will give impoved pefomance using measues pefeed by pactitiones. The esponse to both questions is in the affimative. The optimal weights ae significantly diffeent between the appoaches. The optimal weights fo the GM appoach will by definition bette seve an investo seeking to minimize the pobability of shotfall in an envionment with multiple egimes. Because the GM distibution is a bette model fo eality than the Gaussian distibution, we believe that the GM appoach will do a bette job fo managing potfolios in the eal wold. Acknowledgment. IB would like to thank M. Gey Salkin and Pof. Nicos Chistofides of the CQF, Impeial College fo funding.

18 18 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO A.1. Definitions. Appendix A. Gaussian mixtue distibution Definition A.1.1. A (scala) andom vaiable Z has the univaiate GM distibution if its pobability density function f Z (z) is of the fom (A.1.1) f Z (z) = f Xi (z) = φ( z µ i ) σ i whee the andom vaiables X i ae nomally distibuted with nomal pobability density functions φ Xi (x) = φ( z µ i σ i ), φ(z) is the standad nomal pobability density function and the weights sum to one. The andom vaiables X i have means µ i and vaiances σi 2. The finite sum is ove the desied numbe of nomal components to combine, n. Remak A.1.2. The cumulative distibution function is tivially: (A.1.2) F Z (z) = Φ( z µ i ) σ i whee Φ(z) is the standad nomal cumulative density function. We will made use of this obsevation in Section 4.1 when we define the PoS objective. Similaly, Definition A.1.3. A vecto andom vaiable Z has the multivaiate GM distibution if its pobability density function f Z (z ) is of the fom (A.1.3) f Z (z ) = f X (i)(z ) = φ µ (i),v (i)(z ) whee the (vecto) andom vaiables X (i) ae multivaiate nomally distibuted with pobability density functions φ µ (i),v (i)(z ), and the weights sum to one. The vecto andom vaiable X (i) has mean µ (i) and vaiance-covaiance matix V (i). E.g. if we take the ath and bth components of X (i), thei covaiance is the element (a, b) of V (i) ; i.e. Cov(X a (i), X (i) b ) = V (i) ab. Also E[X(i) a ] = µ (i) a. Remak A.1.4. Note that by definition Cov(X a (i), X (j) b ) = 0 fo i j. Remak A.1.5. In the numeical expeiments descibed late, the mixtue distibution contains two nomal components, descibing asset etuns unde tanquil and distessed conditions. We shall efe to the weights as egime weights. A.2. Moments. The mean of a andom vaiable with the mixtue distibution is simply expessed as a linea combination of the means of the component nomal distibutions. Poposition A.2.1. The expectation of a function f of a andom vaiable with the GM distibution can be expessed in tems of the expectations of functions of the component nomally distibuted vaiables: (A.2.1) E[f(Z )] = E[f(X (i) )]

19 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT 19 Remak A.2.2. In paticula, (A.2.2) E[Z ] = µ (i) N.B. This is a potential souce of confusion given that it is not tue in geneal that Z = n X (i). The vaiance depends not only on the vaiances of the components, but also on the diffeences between the means of the components. Poposition A.2.3. The vaiance of a andom vaiable with the (univaiate) GM distibution can be expessed in tems of the expectations and vaiances of the component nomally distibuted vaiables: (A.2.3) Va[Z a ] = = Va[X (i) a ] + (σ (i) a ) 2 + n,n i,j<i n,n i,j<i w j (E[X a (i) ] E[X a (j) ]) 2 w j (µ (i) a µ (j) a ) 2 Remak A.2.4. If we pemit the vaiance and expectation opeatos to thead ove the components of the vecto aguments f(x ) a := f(x a ), this can be witten in altenative vecto fom as (A.2.4) Va[Z ] = = Va[X (i) ] + σ (i)2 + n,n i,j<i n,n i,j<i w j (E[X (i) ] E[X (j) ]) 2 w j (µ (i) µ (j) ) 2 The vaiance esult, above is a special case of the following esult fo the covaiance: Poposition A.2.5. The covaiance between two elements of a vecto andom vaiable with the (multivaiate) GM distibution can be expessed in tems of the expectations of functions of the component nomally distibuted vaiables: (A.2.5) Cov[Z a, Z b ] = = n,n Cov[X a (i), X (i) b ]+ i,j<i w j (E[X (i) a V (i) ab n,n + i,j<i ] E[X (j) a w j (µ (i) a ])(E[X (i) b ] E[X(j) b ]) µ (j) a )(µ (i) b µ (j) b ) Remak A.2.6. In matix notation, whee W ab := Cov[Z a, Z b ], n,n (A.2.6) W = V (i) + w j (µ (i) µ (j) ).(µ (i) µ (j) ) whee indicates matix tanspose. i,j<i

20 20 I. BUCKLEY, G. COMEZAÑA, B. DJERROUD, AND L. SECO A.3. Linea combinations of andom vaiables with the GM distibution. Poposition A.3.1. Linea combinations of andom vaiables with the (multivaiate) GM distibution will themselves have a (univaiate) mixtue of nomals distibution. In paticula, the (scala) andom vaiable Y = m a=1 θ az a whee the m-vecto andom vaiable Z has the multivaiate GM distibution and θ is an m-vecto of eal coefficients, has pobability density function (A.3.1) f Y (y) = φ µi, σ i 2(y) whee µ i = µ i.θ and σ 2 i = θ.v i.θ. Simila identities may be found in [27]. Refeences 1. Andew Ang and Geet Bekaet, How do egimes affect asset allocation?, Online at Apil , Intenational asset allocation with egime shifts, On-line, Apil F. Delbaen J. Ebe Atzne, P. and D. Heath, Thinking coheently, RISK 10 (1997), no. 11, K. Koedijk Campbell, R. and P. Kofman, Inceased coelation in bea makets, Financial Analyst Jounal (2002), Loenzo Capiello and Tom A. Feanley, Intenational capm with egime switching gach paametes, On-line at July David Dowe, David dowe s clusteing, mixtue modelling and unsupevised leaning page, On-line at dld/mixtue.modelling.page.html, Staumann D Embechts P, McNeil AJ, Coelation: pitfalls and altenatives, RISK (1999), C. R. Havey Eb, C. B. and T. E. Viskanta, Foecasting intenational equity coelations, Financial Analysts Jounal (1994), E. Fees and E. Valdez, Undestanding elationships using copula, Noth Ameican Actuaial Jounal (1998), no. 2, McNeil A.J. Nyfele M. Fey, R., Copulas and cedit models, RISK (2001), A. Gaflund and B. Nilsson, Dynamic potfolio selection: The elevance of switching egimes and investment hoizon, htm, Mach Makowitz H., Potfolio selection, Jounal of Finance 7 (1952), C.N. Haas, On modeling coelated andom vaiables in isk assessment, Risk Analysis 19 (1999), James Hamilton, A quasi-bayesian appoach to estimating paametes fo mixtues of nomal distibutions, Jounal of Business and Economic Statistics 9 (1991), no. 1, S Hodges, A genealization of the shape atio and its application to valuation bounds and isk measues, Univesity of Wawick Financial Options Reseach Cente, FORC Pe-Pint 98/88, Apil John Hull and Alan White, Value at isk when daily changes in maket vaiables ae not nomally distibuted, Jounal of Deivatives 5 (1998), no. 3, Matin Sola John Diffill, Tualay Kenc, Meton-style option picing unde egime switching, On-line at Januay S.A. Klugman and R. Pasa, Fitting bivaiate loss distibution with copulas, Insuance : Mathematics and Economics 24 (1999), Chiaz Labidi and Thiey An, Revisiting the finite mixtue of gaussian distibutions with applications to futues makets, Computing in Economics and Finance (2000), no F. Longin and B. Solnik, Coelation stuctue of intenational equity makets duing extemely volatile peiods, Jounal of Finance 56 (2001), no. 2,

21 PORTFOLIO OPTIMIZATION IN A GAUSSIAN MIXTURE ENVIRONMENT Dilip B. Madan and Gavin S. McPhail, Investing in skews, June Y. Malevegne and D. Sonette, Testing the gaussian copula hypothesis fo financial assets dependences, id=291140, Novembe D. Pelletie, Regime switching fo dynamic coelations, Mach W.F. Shape, The shape atio, The Jounal of Potfolio Management Fall (1994). 25. A. Suáez and S. Caillo, Computational tools fo the analysis of maket isk, Computing in Economics and Finance (2000), no Subu Venkataaman, Value at isk fo a mixtue of nomal distibutions: The use of quasibayesian estimation techniques, Economic Pespectives (1997). 27. Jin Wang, Modeling and geneating daily changes in maket vaiables using a multivaiate mixtue of nomal distibutions, On-line at jwang/pape/mixnomal.pdf, Januay Pete Zangai, An impoved methodology fo measuing va, RiskMetics Monito 2Q (1996). addess: i.buckley@impeial.ac.uk Cente fo Quantitative Finance,, Impeial College,, London,, SW7 2BX,, UK addess: gustavo@sigmanalysis.com addess: ben d@achilles.math.utoonto.ca Sigmanalysis, Toonto, Canada addess: seco@math.toonto.edu RiskLab,, Univesity of Toonto,, Toonto,, Canada

Chapter 3 Savings, Present Value and Ricardian Equivalence

Chapter 3 Savings, Present Value and Ricardian Equivalence Chapte 3 Savings, Pesent Value and Ricadian Equivalence Chapte Oveview In the pevious chapte we studied the decision of households to supply hous to the labo maket. This decision was a static decision,

More information

Ilona V. Tregub, ScD., Professor

Ilona V. Tregub, ScD., Professor Investment Potfolio Fomation fo the Pension Fund of Russia Ilona V. egub, ScD., Pofesso Mathematical Modeling of Economic Pocesses Depatment he Financial Univesity unde the Govenment of the Russian Fedeation

More information

An Introduction to Omega

An Introduction to Omega An Intoduction to Omega Con Keating and William F. Shadwick These distibutions have the same mean and vaiance. Ae you indiffeent to thei isk-ewad chaacteistics? The Finance Development Cente 2002 1 Fom

More information

AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM

AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM Main Golub Faculty of Electical Engineeing and Computing, Univesity of Zageb Depatment of Electonics, Micoelectonics,

More information

Questions & Answers Chapter 10 Software Reliability Prediction, Allocation and Demonstration Testing

Questions & Answers Chapter 10 Software Reliability Prediction, Allocation and Demonstration Testing M13914 Questions & Answes Chapte 10 Softwae Reliability Pediction, Allocation and Demonstation Testing 1. Homewok: How to deive the fomula of failue ate estimate. λ = χ α,+ t When the failue times follow

More information

Risk Sensitive Portfolio Management With Cox-Ingersoll-Ross Interest Rates: the HJB Equation

Risk Sensitive Portfolio Management With Cox-Ingersoll-Ross Interest Rates: the HJB Equation Risk Sensitive Potfolio Management With Cox-Ingesoll-Ross Inteest Rates: the HJB Equation Tomasz R. Bielecki Depatment of Mathematics, The Notheasten Illinois Univesity 55 Noth St. Louis Avenue, Chicago,

More information

INITIAL MARGIN CALCULATION ON DERIVATIVE MARKETS OPTION VALUATION FORMULAS

INITIAL MARGIN CALCULATION ON DERIVATIVE MARKETS OPTION VALUATION FORMULAS INITIAL MARGIN CALCULATION ON DERIVATIVE MARKETS OPTION VALUATION FORMULAS Vesion:.0 Date: June 0 Disclaime This document is solely intended as infomation fo cleaing membes and othes who ae inteested in

More information

Financial Planning and Risk-return profiles

Financial Planning and Risk-return profiles Financial Planning and Risk-etun pofiles Stefan Gaf, Alexande Kling und Jochen Russ Pepint Seies: 2010-16 Fakultät fü Mathematik und Witschaftswissenschaften UNIERSITÄT ULM Financial Planning and Risk-etun

More information

How Much Should a Firm Borrow. Effect of tax shields. Capital Structure Theory. Capital Structure & Corporate Taxes

How Much Should a Firm Borrow. Effect of tax shields. Capital Structure Theory. Capital Structure & Corporate Taxes How Much Should a Fim Boow Chapte 19 Capital Stuctue & Copoate Taxes Financial Risk - Risk to shaeholdes esulting fom the use of debt. Financial Leveage - Incease in the vaiability of shaeholde etuns that

More information

Software Engineering and Development

Software Engineering and Development I T H E A 67 Softwae Engineeing and Development SOFTWARE DEVELOPMENT PROCESS DYNAMICS MODELING AS STATE MACHINE Leonid Lyubchyk, Vasyl Soloshchuk Abstact: Softwae development pocess modeling is gaining

More information

STUDENT RESPONSE TO ANNUITY FORMULA DERIVATION

STUDENT RESPONSE TO ANNUITY FORMULA DERIVATION Page 1 STUDENT RESPONSE TO ANNUITY FORMULA DERIVATION C. Alan Blaylock, Hendeson State Univesity ABSTRACT This pape pesents an intuitive appoach to deiving annuity fomulas fo classoom use and attempts

More information

ON THE (Q, R) POLICY IN PRODUCTION-INVENTORY SYSTEMS

ON THE (Q, R) POLICY IN PRODUCTION-INVENTORY SYSTEMS ON THE R POLICY IN PRODUCTION-INVENTORY SYSTEMS Saifallah Benjaafa and Joon-Seok Kim Depatment of Mechanical Engineeing Univesity of Minnesota Minneapolis MN 55455 Abstact We conside a poduction-inventoy

More information

Questions for Review. By buying bonds This period you save s, next period you get s(1+r)

Questions for Review. By buying bonds This period you save s, next period you get s(1+r) MACROECONOMICS 2006 Week 5 Semina Questions Questions fo Review 1. How do consumes save in the two-peiod model? By buying bonds This peiod you save s, next peiod you get s() 2. What is the slope of a consume

More information

The transport performance evaluation system building of logistics enterprises

The transport performance evaluation system building of logistics enterprises Jounal of Industial Engineeing and Management JIEM, 213 6(4): 194-114 Online ISSN: 213-953 Pint ISSN: 213-8423 http://dx.doi.og/1.3926/jiem.784 The tanspot pefomance evaluation system building of logistics

More information

Exam #1 Review Answers

Exam #1 Review Answers xam #1 Review Answes 1. Given the following pobability distibution, calculate the expected etun, vaiance and standad deviation fo Secuity J. State Pob (R) 1 0.2 10% 2 0.6 15 3 0.2 20 xpected etun = 0.2*10%

More information

Channel selection in e-commerce age: A strategic analysis of co-op advertising models

Channel selection in e-commerce age: A strategic analysis of co-op advertising models Jounal of Industial Engineeing and Management JIEM, 013 6(1):89-103 Online ISSN: 013-0953 Pint ISSN: 013-843 http://dx.doi.og/10.396/jiem.664 Channel selection in e-commece age: A stategic analysis of

More information

9:6.4 Sample Questions/Requests for Managing Underwriter Candidates

9:6.4 Sample Questions/Requests for Managing Underwriter Candidates 9:6.4 INITIAL PUBLIC OFFERINGS 9:6.4 Sample Questions/Requests fo Managing Undewite Candidates Recent IPO Expeience Please povide a list of all completed o withdawn IPOs in which you fim has paticipated

More information

Strategic Asset Allocation and the Role of Alternative Investments

Strategic Asset Allocation and the Role of Alternative Investments Stategic Asset Allocation and the Role of Altenative Investments DOUGLAS CUMMING *, LARS HELGE HAß, DENIS SCHWEIZER Abstact We intoduce a famewok fo stategic asset allocation with altenative investments.

More information

An Analysis of Manufacturer Benefits under Vendor Managed Systems

An Analysis of Manufacturer Benefits under Vendor Managed Systems An Analysis of Manufactue Benefits unde Vendo Managed Systems Seçil Savaşaneil Depatment of Industial Engineeing, Middle East Technical Univesity, 06531, Ankaa, TURKEY secil@ie.metu.edu.t Nesim Ekip 1

More information

Trading Volume and Serial Correlation in Stock Returns in Pakistan. Abstract

Trading Volume and Serial Correlation in Stock Returns in Pakistan. Abstract Tading Volume and Seial Coelation in Stock Retuns in Pakistan Khalid Mustafa Assistant Pofesso Depatment of Economics, Univesity of Kaachi e-mail: khalidku@yahoo.com and Mohammed Nishat Pofesso and Chaiman,

More information

THE CARLO ALBERTO NOTEBOOKS

THE CARLO ALBERTO NOTEBOOKS THE CARLO ALBERTO NOTEBOOKS Mean-vaiance inefficiency of CRRA and CARA utility functions fo potfolio selection in defined contibution pension schemes Woking Pape No. 108 Mach 2009 Revised, Septembe 2009)

More information

Efficient Redundancy Techniques for Latency Reduction in Cloud Systems

Efficient Redundancy Techniques for Latency Reduction in Cloud Systems Efficient Redundancy Techniques fo Latency Reduction in Cloud Systems 1 Gaui Joshi, Emina Soljanin, and Gegoy Wonell Abstact In cloud computing systems, assigning a task to multiple seves and waiting fo

More information

Carter-Penrose diagrams and black holes

Carter-Penrose diagrams and black holes Cate-Penose diagams and black holes Ewa Felinska The basic intoduction to the method of building Penose diagams has been pesented, stating with obtaining a Penose diagam fom Minkowski space. An example

More information

Valuation of Floating Rate Bonds 1

Valuation of Floating Rate Bonds 1 Valuation of Floating Rate onds 1 Joge uz Lopez us 316: Deivative Secuities his note explains how to value plain vanilla floating ate bonds. he pupose of this note is to link the concepts that you leaned

More information

Optimal Capital Structure with Endogenous Bankruptcy:

Optimal Capital Structure with Endogenous Bankruptcy: Univesity of Pisa Ph.D. Pogam in Mathematics fo Economic Decisions Leonado Fibonacci School cotutelle with Institut de Mathématique de Toulouse Ph.D. Dissetation Optimal Capital Stuctue with Endogenous

More information

arxiv:1110.2612v1 [q-fin.st] 12 Oct 2011

arxiv:1110.2612v1 [q-fin.st] 12 Oct 2011 Maket inefficiency identified by both single and multiple cuency tends T.Toká 1, and D. Hováth 1, 1 Sos Reseach a.s., Stojáenská 3, 040 01 Košice, Slovak Republic Abstact axiv:1110.2612v1 [q-fin.st] 12

More information

VISCOSITY OF BIO-DIESEL FUELS

VISCOSITY OF BIO-DIESEL FUELS VISCOSITY OF BIO-DIESEL FUELS One of the key assumptions fo ideal gases is that the motion of a given paticle is independent of any othe paticles in the system. With this assumption in place, one can use

More information

Saving and Investing for Early Retirement: A Theoretical Analysis

Saving and Investing for Early Retirement: A Theoretical Analysis Saving and Investing fo Ealy Retiement: A Theoetical Analysis Emmanuel Fahi MIT Stavos Panageas Whaton Fist Vesion: Mach, 23 This Vesion: Januay, 25 E. Fahi: MIT Depatment of Economics, 5 Memoial Dive,

More information

Semipartial (Part) and Partial Correlation

Semipartial (Part) and Partial Correlation Semipatial (Pat) and Patial Coelation his discussion boows heavily fom Applied Multiple egession/coelation Analysis fo the Behavioal Sciences, by Jacob and Paticia Cohen (975 edition; thee is also an updated

More information

Tracking/Fusion and Deghosting with Doppler Frequency from Two Passive Acoustic Sensors

Tracking/Fusion and Deghosting with Doppler Frequency from Two Passive Acoustic Sensors Tacking/Fusion and Deghosting with Dopple Fequency fom Two Passive Acoustic Sensos Rong Yang, Gee Wah Ng DSO National Laboatoies 2 Science Pak Dive Singapoe 11823 Emails: yong@dso.og.sg, ngeewah@dso.og.sg

More information

Comparing Availability of Various Rack Power Redundancy Configurations

Comparing Availability of Various Rack Power Redundancy Configurations Compaing Availability of Vaious Rack Powe Redundancy Configuations By Victo Avela White Pape #48 Executive Summay Tansfe switches and dual-path powe distibution to IT equipment ae used to enhance the availability

More information

Concept and Experiences on using a Wiki-based System for Software-related Seminar Papers

Concept and Experiences on using a Wiki-based System for Software-related Seminar Papers Concept and Expeiences on using a Wiki-based System fo Softwae-elated Semina Papes Dominik Fanke and Stefan Kowalewski RWTH Aachen Univesity, 52074 Aachen, Gemany, {fanke, kowalewski}@embedded.wth-aachen.de,

More information

The Predictive Power of Dividend Yields for Stock Returns: Risk Pricing or Mispricing?

The Predictive Power of Dividend Yields for Stock Returns: Risk Pricing or Mispricing? The Pedictive Powe of Dividend Yields fo Stock Retuns: Risk Picing o Mispicing? Glenn Boyle Depatment of Economics and Finance Univesity of Cantebuy Yanhui Li Depatment of Economics and Finance Univesity

More information

Continuous Compounding and Annualization

Continuous Compounding and Annualization Continuous Compounding and Annualization Philip A. Viton Januay 11, 2006 Contents 1 Intoduction 1 2 Continuous Compounding 2 3 Pesent Value with Continuous Compounding 4 4 Annualization 5 5 A Special Poblem

More information

Data Center Demand Response: Avoiding the Coincident Peak via Workload Shifting and Local Generation

Data Center Demand Response: Avoiding the Coincident Peak via Workload Shifting and Local Generation (213) 1 28 Data Cente Demand Response: Avoiding the Coincident Peak via Wokload Shifting and Local Geneation Zhenhua Liu 1, Adam Wieman 1, Yuan Chen 2, Benjamin Razon 1, Niangjun Chen 1 1 Califonia Institute

More information

MATHEMATICAL SIMULATION OF MASS SPECTRUM

MATHEMATICAL SIMULATION OF MASS SPECTRUM MATHEMATICA SIMUATION OF MASS SPECTUM.Beánek, J.Knížek, Z. Pulpán 3, M. Hubálek 4, V. Novák Univesity of South Bohemia, Ceske Budejovice, Chales Univesity, Hadec Kalove, 3 Univesity of Hadec Kalove, Hadec

More information

The Supply of Loanable Funds: A Comment on the Misconception and Its Implications

The Supply of Loanable Funds: A Comment on the Misconception and Its Implications JOURNL OF ECONOMICS ND FINNCE EDUCTION Volume 7 Numbe 2 Winte 2008 39 The Supply of Loanable Funds: Comment on the Misconception and Its Implications. Wahhab Khandke and mena Khandke* STRCT Recently Fields-Hat

More information

MULTIPLE SOLUTIONS OF THE PRESCRIBED MEAN CURVATURE EQUATION

MULTIPLE SOLUTIONS OF THE PRESCRIBED MEAN CURVATURE EQUATION MULTIPLE SOLUTIONS OF THE PRESCRIBED MEAN CURVATURE EQUATION K.C. CHANG AND TAN ZHANG In memoy of Pofesso S.S. Chen Abstact. We combine heat flow method with Mose theoy, supe- and subsolution method with

More information

Reduced Pattern Training Based on Task Decomposition Using Pattern Distributor

Reduced Pattern Training Based on Task Decomposition Using Pattern Distributor > PNN05-P762 < Reduced Patten Taining Based on Task Decomposition Using Patten Distibuto Sheng-Uei Guan, Chunyu Bao, and TseNgee Neo Abstact Task Decomposition with Patten Distibuto (PD) is a new task

More information

Comparing Availability of Various Rack Power Redundancy Configurations

Comparing Availability of Various Rack Power Redundancy Configurations Compaing Availability of Vaious Rack Powe Redundancy Configuations White Pape 48 Revision by Victo Avela > Executive summay Tansfe switches and dual-path powe distibution to IT equipment ae used to enhance

More information

Converting knowledge Into Practice

Converting knowledge Into Practice Conveting knowledge Into Pactice Boke Nightmae srs Tend Ride By Vladimi Ribakov Ceato of Pips Caie 20 of June 2010 2 0 1 0 C o p y i g h t s V l a d i m i R i b a k o v 1 Disclaime and Risk Wanings Tading

More information

A Capacitated Commodity Trading Model with Market Power

A Capacitated Commodity Trading Model with Market Power A Capacitated Commodity Tading Model with Maket Powe Victo Matínez-de-Albéniz Josep Maia Vendell Simón IESE Business School, Univesity of Navaa, Av. Peason 1, 08034 Bacelona, Spain VAlbeniz@iese.edu JMVendell@iese.edu

More information

Nontrivial lower bounds for the least common multiple of some finite sequences of integers

Nontrivial lower bounds for the least common multiple of some finite sequences of integers J. Numbe Theoy, 15 (007), p. 393-411. Nontivial lowe bounds fo the least common multiple of some finite sequences of integes Bai FARHI bai.fahi@gmail.com Abstact We pesent hee a method which allows to

More information

METHODOLOGICAL APPROACH TO STRATEGIC PERFORMANCE OPTIMIZATION

METHODOLOGICAL APPROACH TO STRATEGIC PERFORMANCE OPTIMIZATION ETHODOOGICA APPOACH TO STATEGIC PEFOANCE OPTIIZATION ao Hell * Stjepan Vidačić ** Željo Gaača *** eceived: 4. 07. 2009 Peliminay communication Accepted: 5. 0. 2009 UDC 65.02.4 This pape pesents a matix

More information

Promised Lead-Time Contracts Under Asymmetric Information

Promised Lead-Time Contracts Under Asymmetric Information OPERATIONS RESEARCH Vol. 56, No. 4, July August 28, pp. 898 915 issn 3-364X eissn 1526-5463 8 564 898 infoms doi 1.1287/ope.18.514 28 INFORMS Pomised Lead-Time Contacts Unde Asymmetic Infomation Holly

More information

A framework for the selection of enterprise resource planning (ERP) system based on fuzzy decision making methods

A framework for the selection of enterprise resource planning (ERP) system based on fuzzy decision making methods A famewok fo the selection of entepise esouce planning (ERP) system based on fuzzy decision making methods Omid Golshan Tafti M.s student in Industial Management, Univesity of Yazd Omidgolshan87@yahoo.com

More information

HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING

HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING U.P.B. Sci. Bull., Seies C, Vol. 77, Iss. 2, 2015 ISSN 2286-3540 HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING Roxana MARCU 1, Dan POPESCU 2, Iulian DANILĂ 3 A high numbe of infomation systems ae available

More information

est using the formula I = Prt, where I is the interest earned, P is the principal, r is the interest rate, and t is the time in years.

est using the formula I = Prt, where I is the interest earned, P is the principal, r is the interest rate, and t is the time in years. 9.2 Inteest Objectives 1. Undestand the simple inteest fomula. 2. Use the compound inteest fomula to find futue value. 3. Solve the compound inteest fomula fo diffeent unknowns, such as the pesent value,

More information

Financing Terms in the EOQ Model

Financing Terms in the EOQ Model Financing Tems in the EOQ Model Habone W. Stuat, J. Columbia Business School New Yok, NY 1007 hws7@columbia.edu August 6, 004 1 Intoduction This note discusses two tems that ae often omitted fom the standad

More information

Physics 235 Chapter 5. Chapter 5 Gravitation

Physics 235 Chapter 5. Chapter 5 Gravitation Chapte 5 Gavitation In this Chapte we will eview the popeties of the gavitational foce. The gavitational foce has been discussed in geat detail in you intoductoy physics couses, and we will pimaily focus

More information

Modeling and Verifying a Price Model for Congestion Control in Computer Networks Using PROMELA/SPIN

Modeling and Verifying a Price Model for Congestion Control in Computer Networks Using PROMELA/SPIN Modeling and Veifying a Pice Model fo Congestion Contol in Compute Netwoks Using PROMELA/SPIN Clement Yuen and Wei Tjioe Depatment of Compute Science Univesity of Toonto 1 King s College Road, Toonto,

More information

PAN STABILITY TESTING OF DC CIRCUITS USING VARIATIONAL METHODS XVIII - SPETO - 1995. pod patronatem. Summary

PAN STABILITY TESTING OF DC CIRCUITS USING VARIATIONAL METHODS XVIII - SPETO - 1995. pod patronatem. Summary PCE SEMINIUM Z PODSTW ELEKTOTECHNIKI I TEOII OBWODÓW 8 - TH SEMIN ON FUNDMENTLS OF ELECTOTECHNICS ND CICUIT THEOY ZDENĚK BIOLEK SPŠE OŽNO P.., CZECH EPUBLIC DLIBO BIOLEK MILITY CDEMY, BNO, CZECH EPUBLIC

More information

College Enrollment, Dropouts and Option Value of Education

College Enrollment, Dropouts and Option Value of Education College Enollment, Dopouts and Option Value of Education Ozdagli, Ali Tachte, Nicholas y Febuay 5, 2008 Abstact Psychic costs ae the most impotant component of the papes that ae tying to match empiical

More information

Experimentation under Uninsurable Idiosyncratic Risk: An Application to Entrepreneurial Survival

Experimentation under Uninsurable Idiosyncratic Risk: An Application to Entrepreneurial Survival Expeimentation unde Uninsuable Idiosyncatic Risk: An Application to Entepeneuial Suvival Jianjun Miao and Neng Wang May 28, 2007 Abstact We popose an analytically tactable continuous-time model of expeimentation

More information

Mean-Reverting-Ebit-Based Stock Option Evaluation: Theory and Practice

Mean-Reverting-Ebit-Based Stock Option Evaluation: Theory and Practice Jounal of Applied Finance & aning, vol. 3, no. 5, 03, 35-36 ISSN: 79-6580 pint vesion, 79-6599 online Scienpess Ltd, 03 Mean-Reveting-bit-ased Stoc Option valuation: Theoy and Pactice Hassan l Ibami Abstact

More information

How To Find The Optimal Stategy For Buying Life Insuance

How To Find The Optimal Stategy For Buying Life Insuance Life Insuance Puchasing to Reach a Bequest Ehan Bayakta Depatment of Mathematics, Univesity of Michigan Ann Abo, Michigan, USA, 48109 S. David Pomislow Depatment of Mathematics, Yok Univesity Toonto, Ontaio,

More information

2 r2 θ = r2 t. (3.59) The equal area law is the statement that the term in parentheses,

2 r2 θ = r2 t. (3.59) The equal area law is the statement that the term in parentheses, 3.4. KEPLER S LAWS 145 3.4 Keple s laws You ae familia with the idea that one can solve some mechanics poblems using only consevation of enegy and (linea) momentum. Thus, some of what we see as objects

More information

Contingent capital with repeated interconversion between debt and equity

Contingent capital with repeated interconversion between debt and equity Contingent capital with epeated inteconvesion between debt and equity Zhaojun Yang 1, Zhiming Zhao School of Finance and Statistics, Hunan Univesity, Changsha 410079, China Abstact We develop a new type

More information

STABILITY ANALYSIS IN MILLING BASED ON OPERATIONAL MODAL DATA 1. INTRODUCTION

STABILITY ANALYSIS IN MILLING BASED ON OPERATIONAL MODAL DATA 1. INTRODUCTION Jounal of Machine Engineeing, Vol. 11, No. 4, 211 Batosz POWALKA 1 Macin CHODZKO 1 Kzysztof JEMIELNIAK 2 milling, chatte, opeational modal analysis STABILITY ANALYSIS IN MILLING BASED ON OPERATIONAL MODAL

More information

Do Bonds Span the Fixed Income Markets? Theory and Evidence for Unspanned Stochastic Volatility

Do Bonds Span the Fixed Income Markets? Theory and Evidence for Unspanned Stochastic Volatility Do Bonds Span the Fied Income Makets? Theoy and Evidence fo Unspanned Stochastic olatility PIERRE COLLIN-DUFRESNE and ROBERT S. GOLDSTEIN July, 00 ABSTRACT Most tem stuctue models assume bond makets ae

More information

Spirotechnics! September 7, 2011. Amanda Zeringue, Michael Spannuth and Amanda Zeringue Dierential Geometry Project

Spirotechnics! September 7, 2011. Amanda Zeringue, Michael Spannuth and Amanda Zeringue Dierential Geometry Project Spiotechnics! Septembe 7, 2011 Amanda Zeingue, Michael Spannuth and Amanda Zeingue Dieential Geomety Poject 1 The Beginning The geneal consensus of ou goup began with one thought: Spiogaphs ae awesome.

More information

Chris J. Skinner The probability of identification: applying ideas from forensic statistics to disclosure risk assessment

Chris J. Skinner The probability of identification: applying ideas from forensic statistics to disclosure risk assessment Chis J. Skinne The pobability of identification: applying ideas fom foensic statistics to disclosue isk assessment Aticle (Accepted vesion) (Refeeed) Oiginal citation: Skinne, Chis J. (2007) The pobability

More information

The Role of Gravity in Orbital Motion

The Role of Gravity in Orbital Motion ! The Role of Gavity in Obital Motion Pat of: Inquiy Science with Datmouth Developed by: Chistophe Caoll, Depatment of Physics & Astonomy, Datmouth College Adapted fom: How Gavity Affects Obits (Ohio State

More information

Uncertain Version Control in Open Collaborative Editing of Tree-Structured Documents

Uncertain Version Control in Open Collaborative Editing of Tree-Structured Documents Uncetain Vesion Contol in Open Collaboative Editing of Tee-Stuctued Documents M. Lamine Ba Institut Mines Télécom; Télécom PaisTech; LTCI Pais, Fance mouhamadou.ba@ telecom-paistech.f Talel Abdessalem

More information

Seshadri constants and surfaces of minimal degree

Seshadri constants and surfaces of minimal degree Seshadi constants and sufaces of minimal degee Wioletta Syzdek and Tomasz Szembeg Septembe 29, 2007 Abstact In [] we showed that if the multiple point Seshadi constants of an ample line bundle on a smooth

More information

Effect of Contention Window on the Performance of IEEE 802.11 WLANs

Effect of Contention Window on the Performance of IEEE 802.11 WLANs Effect of Contention Window on the Pefomance of IEEE 82.11 WLANs Yunli Chen and Dhama P. Agawal Cente fo Distibuted and Mobile Computing, Depatment of ECECS Univesity of Cincinnati, OH 45221-3 {ychen,

More information

Controlling the Money Supply: Bond Purchases in the Open Market

Controlling the Money Supply: Bond Purchases in the Open Market Money Supply By the Bank of Canada and Inteest Rate Detemination Open Opeations and Monetay Tansmission Mechanism The Cental Bank conducts monetay policy Bank of Canada is Canada's cental bank supevises

More information

An Efficient Group Key Agreement Protocol for Ad hoc Networks

An Efficient Group Key Agreement Protocol for Ad hoc Networks An Efficient Goup Key Ageement Potocol fo Ad hoc Netwoks Daniel Augot, Raghav haska, Valéie Issany and Daniele Sacchetti INRIA Rocquencout 78153 Le Chesnay Fance {Daniel.Augot, Raghav.haska, Valéie.Issany,

More information

Strength Analysis and Optimization Design about the key parts of the Robot

Strength Analysis and Optimization Design about the key parts of the Robot Intenational Jounal of Reseach in Engineeing and Science (IJRES) ISSN (Online): 2320-9364, ISSN (Pint): 2320-9356 www.ijes.og Volume 3 Issue 3 ǁ Mach 2015 ǁ PP.25-29 Stength Analysis and Optimization Design

More information

Supply chain information sharing in a macro prediction market

Supply chain information sharing in a macro prediction market Decision Suppot Systems 42 (2006) 944 958 www.elsevie.com/locate/dss Supply chain infomation shaing in a maco pediction maket Zhiling Guo a,, Fang Fang b, Andew B. Whinston c a Depatment of Infomation

More information

Insurance Pricing under Ambiguity

Insurance Pricing under Ambiguity Insuance Picing unde Ambiguity Alois Pichle a,b, a Univesity of Vienna, Austia. Depatment of Statistics and Opeations Reseach b Actuay. Membe of the Austian Actuaial Association Abstact Stating fom the

More information

CONCEPT OF TIME AND VALUE OFMONEY. Simple and Compound interest

CONCEPT OF TIME AND VALUE OFMONEY. Simple and Compound interest CONCEPT OF TIME AND VALUE OFMONEY Simple and Compound inteest What is the futue value of shs 10,000 invested today to ean an inteest of 12% pe annum inteest payable fo 10 yeas and is compounded; a. Annually

More information

Electricity transmission network optimization model of supply and demand the case in Taiwan electricity transmission system

Electricity transmission network optimization model of supply and demand the case in Taiwan electricity transmission system Electicity tansmission netwok optimization model of supply and demand the case in Taiwan electicity tansmission system Miao-Sheng Chen a Chien-Liang Wang b,c, Sheng-Chuan Wang d,e a Taichung Banch Gaduate

More information

Model-based clustering of longitudinal data. McNicholas, Paul D.; Murphy, Thomas Brendan. Canadian Journal of Statistics, 38 (1): 153-168

Model-based clustering of longitudinal data. McNicholas, Paul D.; Murphy, Thomas Brendan. Canadian Journal of Statistics, 38 (1): 153-168 Povided by the autho(s) and Univesity College Dublin Libay in accodance with publishe policies Please cite the published vesion when available Title Model-based clusteing of longitudinal data Autho(s)

More information

Towards Realizing a Low Cost and Highly Available Datacenter Power Infrastructure

Towards Realizing a Low Cost and Highly Available Datacenter Power Infrastructure Towads Realizing a Low Cost and Highly Available Datacente Powe Infastuctue Siam Govindan, Di Wang, Lydia Chen, Anand Sivasubamaniam, and Bhuvan Ugaonka The Pennsylvania State Univesity. IBM Reseach Zuich

More information

Firstmark Credit Union Commercial Loan Department

Firstmark Credit Union Commercial Loan Department Fistmak Cedit Union Commecial Loan Depatment Thank you fo consideing Fistmak Cedit Union as a tusted souce to meet the needs of you business. Fistmak Cedit Union offes a wide aay of business loans and

More information

Define What Type of Trader Are you?

Define What Type of Trader Are you? Define What Type of Tade Ae you? Boke Nightmae srs Tend Ride By Vladimi Ribakov Ceato of Pips Caie 20 of June 2010 1 Disclaime and Risk Wanings Tading any financial maket involves isk. The content of this

More information

YARN PROPERTIES MEASUREMENT: AN OPTICAL APPROACH

YARN PROPERTIES MEASUREMENT: AN OPTICAL APPROACH nd INTERNATIONAL TEXTILE, CLOTHING & ESIGN CONFERENCE Magic Wold of Textiles Octobe 03 d to 06 th 004, UBROVNIK, CROATIA YARN PROPERTIES MEASUREMENT: AN OPTICAL APPROACH Jana VOBOROVA; Ashish GARG; Bohuslav

More information

Referral service and customer incentive in online retail supply Chain

Referral service and customer incentive in online retail supply Chain Refeal sevice and custome incentive in online etail supply Chain Y. G. Chen 1, W. Y. Zhang, S. Q. Yang 3, Z. J. Wang 4 and S. F. Chen 5 1,,3,4 School of Infomation Zhejiang Univesity of Finance and Economics

More information

Loyalty Rewards and Gift Card Programs: Basic Actuarial Estimation Techniques

Loyalty Rewards and Gift Card Programs: Basic Actuarial Estimation Techniques Loyalty Rewads and Gift Cad Pogams: Basic Actuaial Estimation Techniques Tim A. Gault, ACAS, MAAA, Len Llaguno, FCAS, MAAA and Matin Ménad, FCAS, MAAA Abstact In this pape we establish an actuaial famewok

More information

Coordinate Systems L. M. Kalnins, March 2009

Coordinate Systems L. M. Kalnins, March 2009 Coodinate Sstems L. M. Kalnins, Mach 2009 Pupose of a Coodinate Sstem The pupose of a coodinate sstem is to uniquel detemine the position of an object o data point in space. B space we ma liteall mean

More information

4a 4ab b 4 2 4 2 5 5 16 40 25. 5.6 10 6 (count number of places from first non-zero digit to

4a 4ab b 4 2 4 2 5 5 16 40 25. 5.6 10 6 (count number of places from first non-zero digit to . Simplify: 0 4 ( 8) 0 64 ( 8) 0 ( 8) = (Ode of opeations fom left to ight: Paenthesis, Exponents, Multiplication, Division, Addition Subtaction). Simplify: (a 4) + (a ) (a+) = a 4 + a 0 a = a 7. Evaluate

More information

Supplementary Material for EpiDiff

Supplementary Material for EpiDiff Supplementay Mateial fo EpiDiff Supplementay Text S1. Pocessing of aw chomatin modification data In ode to obtain the chomatin modification levels in each of the egions submitted by the use QDCMR module

More information

The Binomial Distribution

The Binomial Distribution The Binomial Distibution A. It would be vey tedious if, evey time we had a slightly diffeent poblem, we had to detemine the pobability distibutions fom scatch. Luckily, thee ae enough similaities between

More information

CONCEPTUAL FRAMEWORK FOR DEVELOPING AND VERIFICATION OF ATTRIBUTION MODELS. ARITHMETIC ATTRIBUTION MODELS

CONCEPTUAL FRAMEWORK FOR DEVELOPING AND VERIFICATION OF ATTRIBUTION MODELS. ARITHMETIC ATTRIBUTION MODELS CONCEPUAL FAMEOK FO DEVELOPING AND VEIFICAION OF AIBUION MODELS. AIHMEIC AIBUION MODELS Yui K. Shestopaloff, is Diecto of eseach & Deelopment at SegmentSoft Inc. He is a Docto of Sciences and has a Ph.D.

More information

UNIT CIRCLE TRIGONOMETRY

UNIT CIRCLE TRIGONOMETRY UNIT CIRCLE TRIGONOMETRY The Unit Cicle is the cicle centeed at the oigin with adius unit (hence, the unit cicle. The equation of this cicle is + =. A diagam of the unit cicle is shown below: + = - - -

More information

Statistics and Data Analysis

Statistics and Data Analysis Pape 274-25 An Extension to SAS/OR fo Decision System Suppot Ali Emouznead Highe Education Funding Council fo England, Nothavon house, Coldhabou Lane, Bistol, BS16 1QD U.K. ABSTRACT This pape exploes the

More information

Patent renewals and R&D incentives

Patent renewals and R&D incentives RAND Jounal of Economics Vol. 30, No., Summe 999 pp. 97 3 Patent enewals and R&D incentives Fancesca Conelli* and Mak Schankeman** In a model with moal hazad and asymmetic infomation, we show that it can

More information

The LCOE is defined as the energy price ($ per unit of energy output) for which the Net Present Value of the investment is zero.

The LCOE is defined as the energy price ($ per unit of energy output) for which the Net Present Value of the investment is zero. Poject Decision Metics: Levelized Cost of Enegy (LCOE) Let s etun to ou wind powe and natual gas powe plant example fom ealie in this lesson. Suppose that both powe plants wee selling electicity into the

More information

Definitions and terminology

Definitions and terminology I love the Case & Fai textbook but it is out of date with how monetay policy woks today. Please use this handout to supplement the chapte on monetay policy. The textbook assumes that the Fedeal Reseve

More information

An application of stochastic programming in solving capacity allocation and migration planning problem under uncertainty

An application of stochastic programming in solving capacity allocation and migration planning problem under uncertainty An application of stochastic pogamming in solving capacity allocation and migation planning poblem unde uncetainty Yin-Yann Chen * and Hsiao-Yao Fan Depatment of Industial Management, National Fomosa Univesity,

More information

Cloud Service Reliability: Modeling and Analysis

Cloud Service Reliability: Modeling and Analysis Cloud Sevice eliability: Modeling and Analysis Yuan-Shun Dai * a c, Bo Yang b, Jack Dongaa a, Gewei Zhang c a Innovative Computing Laboatoy, Depatment of Electical Engineeing & Compute Science, Univesity

More information

Mining Relatedness Graphs for Data Integration

Mining Relatedness Graphs for Data Integration Mining Relatedness Gaphs fo Data Integation Jeemy T. Engle (jtengle@indiana.edu) Ying Feng (yingfeng@indiana.edu) Robet L. Goldstone (goldsto@indiana.edu) Indiana Univesity Bloomington, IN. 47405 USA Abstact

More information

Liquidity and Insurance for the Unemployed*

Liquidity and Insurance for the Unemployed* Fedeal Reseve Bank of Minneapolis Reseach Depatment Staff Repot 366 Decembe 2005 Liquidity and Insuance fo the Unemployed* Robet Shime Univesity of Chicago and National Bueau of Economic Reseach Iván Wening

More information

Load Balancing in Processor Sharing Systems

Load Balancing in Processor Sharing Systems Load Balancing in ocesso Shaing Systems Eitan Altman INRIA Sophia Antipolis 2004, oute des Lucioles 06902 Sophia Antipolis, Fance altman@sophia.inia.f Utzi Ayesta LAAS-CNRS Univesité de Toulouse 7, Avenue

More information

Load Balancing in Processor Sharing Systems

Load Balancing in Processor Sharing Systems Load Balancing in ocesso Shaing Systems Eitan Altman INRIA Sophia Antipolis 2004, oute des Lucioles 06902 Sophia Antipolis, Fance altman@sophia.inia.f Utzi Ayesta LAAS-CNRS Univesité de Toulouse 7, Avenue

More information

Liquidity and Insurance for the Unemployed

Liquidity and Insurance for the Unemployed Liquidity and Insuance fo the Unemployed Robet Shime Univesity of Chicago and NBER shime@uchicago.edu Iván Wening MIT, NBER and UTDT iwening@mit.edu Fist Daft: July 15, 2003 This Vesion: Septembe 22, 2005

More information

Things to Remember. r Complete all of the sections on the Retirement Benefit Options form that apply to your request.

Things to Remember. r Complete all of the sections on the Retirement Benefit Options form that apply to your request. Retiement Benefit 1 Things to Remembe Complete all of the sections on the Retiement Benefit fom that apply to you equest. If this is an initial equest, and not a change in a cuent distibution, emembe to

More information

Vector Calculus: Are you ready? Vectors in 2D and 3D Space: Review

Vector Calculus: Are you ready? Vectors in 2D and 3D Space: Review Vecto Calculus: Ae you eady? Vectos in D and 3D Space: Review Pupose: Make cetain that you can define, and use in context, vecto tems, concepts and fomulas listed below: Section 7.-7. find the vecto defined

More information

CHAPTER 10 Aggregate Demand I

CHAPTER 10 Aggregate Demand I CHAPTR 10 Aggegate Demand I Questions fo Review 1. The Keynesian coss tells us that fiscal policy has a multiplied effect on income. The eason is that accoding to the consumption function, highe income

More information