T-Space at The University of Toronto Libraries >
School of Graduate Studies - Theses >
Please use this identifier to cite or link to this item:
|Title: ||Accelerating an Analytical Approach to Collateralized Debt Obligation Pricing|
|Authors: ||Gupta, Dharmendra|
|Advisor: ||Chow, Paul|
|Department: ||Electrical and Computer Engineering|
|Keywords: ||FPGA Application|
|Issue Date: ||19-Jan-2010|
|Abstract: ||In recent years, financial simulations have gotten computationally intensive due to larger portfolio sizes, and an increased demand to perform real-time risk analysis. In this paper, we propose a hardware implementation that uses a recursive analytical method to price the Collateralized Debt Obligations. A novel convolution approach based on FIFOs for storage is implemented for the recursive convolution. It is also used to address one of the main drawbacks of the analytical approach. The FIFO-based convolution approach is compared against two different convolution approaches outperforming them with a much smaller memory usage.
The CDO core designed with the FIFO-based convolution method is implemented and tested on a Virtex-5 FPGA and compared against a C implementation, running on a 2.8GHz Intel Processor, resulting in a 41-fold speed up. A brief comparison against a Monte Carlo based hardware implementation for structured instruments yields mixed results.|
|Appears in Collections:||Master|
The Edward S. Rogers Sr. Department of Electrical & Computer Engineering - Master theses
This item is licensed under a Creative Commons License
Items in T-Space are protected by copyright, with all rights reserved, unless otherwise indicated.