Answers are grounded in this paper and cite their sources.
We study admissible operator realizations for stochastic Volterra equations with regular and singular memory kernels and establish explicit approximation bounds for a fractional kernel. The nonlocal domain is not shift-invariant in general and is non-dense in the stated history space for an exponential kernel; hence the trace identity is proved using the right-shift semigroup for regular kernels. For the singular fractional kernel K(t)=tβ−1e−γt/Γ(β) with 1/2<β<1, an exponential-factor realization yields a stochastic-Fubini trace identity for globally Lipschitz equations and fractional-domain estimates for a self-adjoint multiplication generator. An explicit positive N-factor rule satisfies an L2(0,T) kernel bound of order N−(2β−1)/3; the same rate holds for the corresponding solutions in root-mean-square error under global Lipschitz assumptions. For the rough-Heston application, the characteristic transform is obtained from affine Volterra theory. Independent product-integration calculations give reproducible option prices and convergence comparisons for positive exponential kernel approximations. A stochastic finite-factor experiment examines sampling variability and discretization effects in the square-root volatility model. Fractional and spectral estimates are formulated under explicit hypotheses on an abstract analytic generator, with the corresponding compactness, commutation, and resolvent conditions imposed where required. The principal contribution is a kernel-dependent analysis of operator admissibility, together with explicit kernel and solution approximation bounds for the singular fractional case under global Lipschitz assumptions.
No takes yet. Share an insight, caveat, or question.
Aljalal et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: