Abstract. Quasi-Newton methods accelerate gradient methods for minimizing a function by approximating the inverse Hessian matrix of the function. Several papers in recent literature have dealt with the generation of classes of approximating matrices as a function of a scalar parameter. This paper derives necessary and sufficient conditions on the range of one such parameter to guarantee stability of the method. It further shows that the parameter effects only the length, not the direction, of the search vector at each step, and uses this result to derive several computational algorithms. The algorithms are evaluated on a series of test problems.
We propose a non-symmetric copula to model the evolution of electricity and gas prices by a bivariate non-Gaussian autoregressive process. We identify the marginal dynamics as driven by normal inverse Gaussian processes, estimating them from a series of observed UK electricity and gas spot data. We estimate the copula by modeling the difference between the empirical copula and the independent copula. We then simulate the joint process and price options written on the spark spread. We find that option prices are significantly influenced by the copula and the marginal distributions, along with the seasonality of the underlying prices.Mathematical finance, Copulas, Derivative pricing models, Asymmetry, Empirical time series analysis, Energy derivatives, Levy process, Numerical simulation,
We propose a quasi-Monte Carlo (qMC) algorithm to simulate variates from the normal inverse Gaussian (NIG) distribution. The algorithm is based on a Monte Carlo technique found in Rydberg [13], and is based on sampling three independent uniform variables. We apply the algorithm to three problems appearing in finance. First, we consider the valuation of plain vanilla call options and Asian options. The next application considers the problem of deriving implied parameters for the underlying asset dynamics based on observed option prices. We employ our proposed algorithm together with the Newton Method, and show how we can find the scale parameter of the NIG-distribution of the logreturns in case of a call or an Asian option. We also provide an extensive error analysis for this method. Finally we study the calculation of Value-at-Risk for a portfolio of nonlinear products where the returns are modeled by NIG random variables.
The rows and columns of an arbitrary coefficient matrix of large numerical problems can often be permuted so that substantial time can be saved in computations. For example, if a large linear programming problem has a suitable block-angular structure, one of the time-saving decomposition algorithms can be used. This article presents a systematic method for effecting such a block-angular permutation. An example and the results of manipulations of matrices with more than 300 rows and 2500 columns are shown.
Abstract. Bond duration in its basic deterministic form is a concept well understood. Its meaning in the context of a yield curve on a stochastic path is less well developed. We extend the basic idea to a stochastic setting. More precisely, we introduce the concept of stochastic duration as a Malliavin derivative in the direction of a stochastic yield surface modeled by the Musiela equation. Further, using this concept we also propose a mathematical framework for the construction of immunization strategies (or delta hedges) of portfolios of interest-ratesensitive securities with respect to the fluctuation of the whole yield surface.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.