A family of density expansions for Lévy-type processes (Q2258531)

The authors consider scalar Lévy-type processes with sufficiently smooth coefficients which include local volatility models with or without jumps of various kinds extensively investigated in the past. In particular, earlier results and methods of \textit{S. Pagliarani} et al. [SIAM J. Financ. Math. 4, 265--296 (2013; Zbl 1285.60084)] that were obtained for state-independent Lévy measures and no-default situations are extended and generalized here to include state-dependent Lévy measures and the local default intensity case. No small diffusion or small jump size/intensity assumptions as in [\textit{E. Benhamou} et al., Finance Stoch. 13, No. 4, 563--589 (2009; Zbl 1195.91153)] are made in this paper. The authors derive a family of asymptotic expansions (with respect to an arbitrary basis) for the transition densities of Lévy-type processes as well as European-style derivative prices and defaultable bond prices. Pseudo-differential calculus techniques are used to provide explicit formulae for the Fourier transform of every term in the transition density and option pricing expansions. For models with Gaussian-type jumps, respective inverse Fourier transforms are explicitly computed, and pointwise error estimates for the transition densities are provided. Numerical examples that illustrate the versatility and effectiveness of the methods include CEV-like processes and processes with Gaussian or Variance-Gamma Lévy measure.