Approximating the Laplace transform of the sum of dependent lognormals

preprint2015arXivOpen access

Abstract

Let (X1,…,Xn)(X_1, \dots, X_n) be multivariate normal, with mean vector μ\boldsymbolμ and covariance matrix Σ\boldsymbolΣ, and Sn=eX1+⋯+eXnS_n=\mathrm{e}^{X_1}+\cdots+\mathrm{e}^{X_n}. The Laplace transform L(θ)=Ee−θSn∝∫exp⁡{−hθ(x)} dx{\cal L}(θ)=\mathbb{E}\mathrm{e}^{-θS_n} \propto \int \exp\{-h_θ(\boldsymbol{x})\} \,\mathrm{d} \boldsymbol{x} is represented as L~(θ)I(θ)\tilde{\cal L}(θ)I(θ), where L~(θ)\tilde{\cal L}(θ) is given in closed-form and I(θ)I(θ) is the error factor (≈1\approx 1). We obtain L~(θ)\tilde{\cal L}(θ) by replacing hθ(x)h_θ(\boldsymbol{x}) with a second order Taylor expansion around its minimiser x∗\boldsymbol{x}^*. An algorithm for calculating the asymptotic expansion of x∗\boldsymbol{x}^* is presented, and it is shown that I(θ)→1I(θ)\to 1 as θ→∞θ\to\infty. A variety of numerical methods for evaluating I(θ)I(θ) are discussed, including Monte Carlo with importance sampling and quasi-Monte Carlo. Numerical examples (including Laplace transform inversion for the density of SnS_n) are also given.

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.