Jakob Sidenius - Academia.edu (original) (raw)
Papers by Jakob Sidenius
We study LIBOR market models with a number of factors ranging from one to ten. The emphasis throu... more We study LIBOR market models with a number of factors ranging from one to ten. The emphasis throughout is on the implications for term structure modelling and (exotic) securities pricing rather than on technical issues. The study falls into two parts. In the first part the models are calibrated to cap and swaption prices for several major markets and we observe that-perhaps contrary to expectations-the overall market fit is independent of the number of factors. Closer investigation shows that the number of factors affect the implicit calendar time dependence of the volatility function; models with a high number of factors allow a more stationary volatility function than do models with few factors. In the second part of the study we investigate the implications for exoctics pricing of the number of factors in the model. Here we find-as expected in view of our findings regarding the time-dependence of the volatility function-a very strong sensitivity of exotics prices to the number of factors. This finding has far-reaching implications for derivatives pricing in practice.
Http Dx Doi Org 10 3905 Jod 1998 407997, Feb 22, 2009
Options involving two barriers, including the case of rebates, are analyzed and analytic valuatio... more Options involving two barriers, including the case of rebates, are analyzed and analytic valuation formulas given. The analysis is self-contained and intuitive, and relies only on de®ning properties of Brownian sample paths. The connection with results obtained by Laplace transforms is made explicit.
Models with systematic factors are popular in the modeling of CDOs, mainly due to their simplicit... more Models with systematic factors are popular in the modeling of CDOs, mainly due to their simplicity and tractability. In this small note we provide a general framework which we use to survey a number of CDO models that have appeared in the literature so far. We suggest extensions and also briefly discuss a select number of issues with factor models, ranging from calibration against CDO market data (i.e. base corre- lation skews) to credit spread hedging and maturity extrapolation. We highlight a number of inherent limitations of factor models, and also discuss certain idiosyncra- cies of popular model-independent approaches to computation of spread hedges.
Physics Letters B, 1988
ABSTRACT The covariant loop calculus provides an efficient technique for computing explicit expre... more ABSTRACT The covariant loop calculus provides an efficient technique for computing explicit expressions for the density on moduli space corresponding to arbitrary (bosonic string) loop diagrams. Since modular invariance is not manifest, however, we carry out a detailed comparison with known explicit two- and three-loop results derived using analytic geometry (one loop is known to be okay). We establish identity to “high” order in some moduli and exactly in others. Agreement is found as a result of various non-trivial cancellations, in part related to number theory. We feel our results provide very strong support for the correctness of the covariant loop calculus approach.
Physics Letters B, 1986
ABSTRACT An N-reggeon vertex is constructed with the inclusion of the contribution of the ghost c... more ABSTRACT An N-reggeon vertex is constructed with the inclusion of the contribution of the ghost coordinates. It is shown that it is projective and BRST invariant.
Physics Letters B, 1988
ABSTRACT We carry out a comparison between vertices proposed by various authors by discussing the... more ABSTRACT We carry out a comparison between vertices proposed by various authors by discussing the multi-loop vertices for the bosonic string and for the NS-string in terms of oscillator modes. Address after 1 September 1988: Institute of Theoretical Physics, Chalmers Institute of Technology, S-412 96 Göteborg, Sweden.
Nuclear Physics B, 1989
A previously developed formalism for the bosonic string is extended to the Neveu-Schwarz-Ramond s... more A previously developed formalism for the bosonic string is extended to the Neveu-Schwarz-Ramond string using 2-d superspace techniques throughout. Three-string vertices for NS and R-strings are constructed, sewing rules developed, and the technique of quasi-superconformal modes is set up for constructing the measure on supermoduli space. Symmetries, such as superconformal invariance and BRST invariance, are guaranteed ab initio. Picture changing and bosonization are avoided. Examples are given. The formalism should allow a superstring loop calculus based on supermoduli. Results concerning the ensuing super-Schottky description are given.
Nuclear Physics B, 1987
ABSTRACT The BRST-invariant N-reggeon vertex (for the bosonic string) previously given by us in t... more ABSTRACT The BRST-invariant N-reggeon vertex (for the bosonic string) previously given by us in the operator formulation is considered in more detail. In particular we present a direct derivation from the string path integral. Several crucial symmetry properties found a posteriori before, become a priori clearer in this formulation. A number of delicate points related to zero modes, cut off procedures and normal ordering prescriptions are treated in some detail. The old technique of letting the string field acquire a small dimension 1/2ε --> 0+ is found especially elegant.
Nuclear Physics B - NUCL PHYS B, 1988
C0var1ant N-5tr1n9 amp11tude5 f0r the c105ed 6050n1c 5tr1n9 are ana1y2ed w1th empha515 0n the re1... more C0var1ant N-5tr1n9 amp11tude5 f0r the c105ed 6050n1c 5tr1n9 are ana1y2ed w1th empha515 0n the re1at10n 6etween 9h05t 2er0 m0de5 and 1nte9rat10n mea5ure 0ver K06a-N1e15en 11ke var1a61e5 (here we f1nd that a m0d1f1cat10n 0f the re5u1t5 0f ref5. [1,18] 15 needed). Fact0r12at10n and c0rrect 8R57-c0h0m0109y pr0pert1e5 are e5ta6115hed. Mu1t1-100p amp11tude5 f0r ar61trary externa1 5tate5 are c0n5tructed 6y f0rmu1at1n9 5ew1n9 ru1e5 9enera1121n9 the 01d treatment 0f re99e0n5 t0 1nc1ude 9h05t5. 1n part1cu1ar the 9h05t 2er0 m0de5 enc0de and 9enerate the mea5ure 0n m0du11 5pace 1n a way 51m11ar t0 1dea5 pr0p05ed 6y Mart1nec [2]. Exp11c1t N-tachy0n mu1t1-100p re5u1t5 a9ree w1th re5u1t5 0f ref. [30] 06ta1ned u51n9 51m11ar, 6ut techn1ca11y 4u1te d1fferent meth0d5, and man1fe5t1y 5h0w 8e1av1n-Kn12hn1k 51n9u1ar1t1e5 when 5urface5 de9enerate.
Financial Analysts Journal, 1996
International Journal of Theoretical …, 2008
Abstract: We present the SPA framework, a novel approach to the modeling of the dynamics of portf... more Abstract: We present the SPA framework, a novel approach to the modeling of the dynamics of portfolio default losses. In this framework, models are specified by a two-layer process. The first layer models the dynamics of portfolio loss distributions in the absence of information ...
Journal of Credit Risk Volume, 2004
The Journal of Derivatives, 1998
Page 1. Page 2. tract equals the number of periods in the time span covered, then both the auto-a... more Page 1. Page 2. tract equals the number of periods in the time span covered, then both the auto-and chooser-flex cap are equivalent to the ordinary cap covering the same time span. Also, from the fact that the holder of a chooser ...
We study LIBOR market models with a number of factors ranging from one to ten. The emphasis throu... more We study LIBOR market models with a number of factors ranging from one to ten. The emphasis throughout is on the implications for term structure modelling and (exotic) securities pricing rather than on technical issues. The study falls into two parts. In the first part the models are calibrated to cap and swaption prices for several major markets and we observe that-perhaps contrary to expectations-the overall market fit is independent of the number of factors. Closer investigation shows that the number of factors affect the implicit calendar time dependence of the volatility function; models with a high number of factors allow a more stationary volatility function than do models with few factors. In the second part of the study we investigate the implications for exoctics pricing of the number of factors in the model. Here we find-as expected in view of our findings regarding the time-dependence of the volatility function-a very strong sensitivity of exotics prices to the number of factors. This finding has far-reaching implications for derivatives pricing in practice.
Http Dx Doi Org 10 3905 Jod 1998 407997, Feb 22, 2009
Options involving two barriers, including the case of rebates, are analyzed and analytic valuatio... more Options involving two barriers, including the case of rebates, are analyzed and analytic valuation formulas given. The analysis is self-contained and intuitive, and relies only on de®ning properties of Brownian sample paths. The connection with results obtained by Laplace transforms is made explicit.
Models with systematic factors are popular in the modeling of CDOs, mainly due to their simplicit... more Models with systematic factors are popular in the modeling of CDOs, mainly due to their simplicity and tractability. In this small note we provide a general framework which we use to survey a number of CDO models that have appeared in the literature so far. We suggest extensions and also briefly discuss a select number of issues with factor models, ranging from calibration against CDO market data (i.e. base corre- lation skews) to credit spread hedging and maturity extrapolation. We highlight a number of inherent limitations of factor models, and also discuss certain idiosyncra- cies of popular model-independent approaches to computation of spread hedges.
Physics Letters B, 1988
ABSTRACT The covariant loop calculus provides an efficient technique for computing explicit expre... more ABSTRACT The covariant loop calculus provides an efficient technique for computing explicit expressions for the density on moduli space corresponding to arbitrary (bosonic string) loop diagrams. Since modular invariance is not manifest, however, we carry out a detailed comparison with known explicit two- and three-loop results derived using analytic geometry (one loop is known to be okay). We establish identity to “high” order in some moduli and exactly in others. Agreement is found as a result of various non-trivial cancellations, in part related to number theory. We feel our results provide very strong support for the correctness of the covariant loop calculus approach.
Physics Letters B, 1986
ABSTRACT An N-reggeon vertex is constructed with the inclusion of the contribution of the ghost c... more ABSTRACT An N-reggeon vertex is constructed with the inclusion of the contribution of the ghost coordinates. It is shown that it is projective and BRST invariant.
Physics Letters B, 1988
ABSTRACT We carry out a comparison between vertices proposed by various authors by discussing the... more ABSTRACT We carry out a comparison between vertices proposed by various authors by discussing the multi-loop vertices for the bosonic string and for the NS-string in terms of oscillator modes. Address after 1 September 1988: Institute of Theoretical Physics, Chalmers Institute of Technology, S-412 96 Göteborg, Sweden.
Nuclear Physics B, 1989
A previously developed formalism for the bosonic string is extended to the Neveu-Schwarz-Ramond s... more A previously developed formalism for the bosonic string is extended to the Neveu-Schwarz-Ramond string using 2-d superspace techniques throughout. Three-string vertices for NS and R-strings are constructed, sewing rules developed, and the technique of quasi-superconformal modes is set up for constructing the measure on supermoduli space. Symmetries, such as superconformal invariance and BRST invariance, are guaranteed ab initio. Picture changing and bosonization are avoided. Examples are given. The formalism should allow a superstring loop calculus based on supermoduli. Results concerning the ensuing super-Schottky description are given.
Nuclear Physics B, 1987
ABSTRACT The BRST-invariant N-reggeon vertex (for the bosonic string) previously given by us in t... more ABSTRACT The BRST-invariant N-reggeon vertex (for the bosonic string) previously given by us in the operator formulation is considered in more detail. In particular we present a direct derivation from the string path integral. Several crucial symmetry properties found a posteriori before, become a priori clearer in this formulation. A number of delicate points related to zero modes, cut off procedures and normal ordering prescriptions are treated in some detail. The old technique of letting the string field acquire a small dimension 1/2ε --> 0+ is found especially elegant.
Nuclear Physics B - NUCL PHYS B, 1988
C0var1ant N-5tr1n9 amp11tude5 f0r the c105ed 6050n1c 5tr1n9 are ana1y2ed w1th empha515 0n the re1... more C0var1ant N-5tr1n9 amp11tude5 f0r the c105ed 6050n1c 5tr1n9 are ana1y2ed w1th empha515 0n the re1at10n 6etween 9h05t 2er0 m0de5 and 1nte9rat10n mea5ure 0ver K06a-N1e15en 11ke var1a61e5 (here we f1nd that a m0d1f1cat10n 0f the re5u1t5 0f ref5. [1,18] 15 needed). Fact0r12at10n and c0rrect 8R57-c0h0m0109y pr0pert1e5 are e5ta6115hed. Mu1t1-100p amp11tude5 f0r ar61trary externa1 5tate5 are c0n5tructed 6y f0rmu1at1n9 5ew1n9 ru1e5 9enera1121n9 the 01d treatment 0f re99e0n5 t0 1nc1ude 9h05t5. 1n part1cu1ar the 9h05t 2er0 m0de5 enc0de and 9enerate the mea5ure 0n m0du11 5pace 1n a way 51m11ar t0 1dea5 pr0p05ed 6y Mart1nec [2]. Exp11c1t N-tachy0n mu1t1-100p re5u1t5 a9ree w1th re5u1t5 0f ref. [30] 06ta1ned u51n9 51m11ar, 6ut techn1ca11y 4u1te d1fferent meth0d5, and man1fe5t1y 5h0w 8e1av1n-Kn12hn1k 51n9u1ar1t1e5 when 5urface5 de9enerate.
Financial Analysts Journal, 1996
International Journal of Theoretical …, 2008
Abstract: We present the SPA framework, a novel approach to the modeling of the dynamics of portf... more Abstract: We present the SPA framework, a novel approach to the modeling of the dynamics of portfolio default losses. In this framework, models are specified by a two-layer process. The first layer models the dynamics of portfolio loss distributions in the absence of information ...
Journal of Credit Risk Volume, 2004
The Journal of Derivatives, 1998
Page 1. Page 2. tract equals the number of periods in the time span covered, then both the auto-a... more Page 1. Page 2. tract equals the number of periods in the time span covered, then both the auto-and chooser-flex cap are equivalent to the ordinary cap covering the same time span. Also, from the fact that the holder of a chooser ...