<i>L</i> ^<i>p</i> -solutions of backward stochastic differential equations with time-delayed generators (Q7305919)

This article tackles the problem of backward stochastic differential equations with delays in the generator, focusing on the case where the input data are only integrable to the \(p\)-th power (\(p>1\)). The key novelty is the derivation of refined estimates that make the smallness conditions on the Lipschitz constant \(K\) and the terminal time \(T\) fully explicit. To prove well-posedness, the authors utilize two separate strategies: a truncation/approximation method and a contraction-mapping (Picard) argument. The latter sharpens earlier work by simplifying the proof and providing a clearer convergence criterion. Crucially, by covering the range \(1 < p < 2\), the paper removes the standard square-integrability requirement. This relaxation accommodates important practical scenarios in finance and insurance where coefficients such as interest rates or risk premiums are inherently unbounded.