Todos los libros

Profesional

Apps Acerca de Coach Iniciar sesión Empezar a leer

Quantitative Finance · Glosario

¿Qué es Randomised quasi-Monte Carlo, Brownian bridge construction?

También llamado: randomised quasi-Monte Carlo · Brownian bridge construction

Definition 26.10 Quantitative Methods · Capítulo 26 — Monte Carlo

Randomised quasi-Monte Carlo randomises a low-discrepancy set so that each point is uniform while the set keeps its structure, and estimates the error from independent replicates. The kit uses Owen’s nested uniform scrambling (1995): each binary digit of each coordinate is flipped by a random bit attached to the node of the binary tree that its leading digits define. The Brownian bridge construction assigns the first coordinate to WTW_T, the next to WT/2W_{T/2} given its ends, and so on down the levels, so that the first coordinates, where low-discrepancy points are best, carry most of the variance of the path.

256 points in the unit square: independent Philox uniforms and the first two coordinates of Owen-scrambled Sobol points. Data: the chapter’s tutorial, seeded.
Figure 26.2. 256 points in the unit square: independent Philox uniforms and the first two coordinates of Owen-scrambled Sobol points. Data: the chapter’s tutorial, seeded.
Standard error of the average-price call (64 fixings) against the number of paths: plain Monte Carlo and the control variate (slope -1/2), and scrambled Sobol points (16 independent scrambles) with increments in time order, with the Brownian bridge construction, and with the bridge and the control variate. Data: the chapter’s tutorial, seeded.
Figure 26.3. Standard error of the average-price call (64 fixings) against the number of paths: plain Monte Carlo and the control variate (slope −12-\frac12), and scrambled Sobol points (16 independent scrambles) with increments in time order, with the Brownian bridge construction, and with the bridge and the control variate. Data: the chapter’s tutorial, seeded.
Leer en el capítulo →