SciComp Inc. - SciFinance

SciComp adds advanced Stochastic Local Volatility calibration

Tue, 10 Apr 2012 18:27:08 -0700

The SLV Calibrator is based on the original Ren & Madan paper and follows the implementation as described in the Bloomberg paper "Stochastic Local Volatility" by Tataru & Fisher. The SLV Calibrator takes any calibrated LV surface that matches vanillas. Then, using a non-linear Fokker-Planck equation, one adds a SV component and for any given set of SV parameters computes a new "leveraged local volatility surface" that still matches the vanillas, while accommodating SV. The SLV Calibrator then applies to this PDE solution a Levenberg-Marquardt optimizer and finds the (time bucketed) SV parameters that yield a maximally flat leveraged local volatility surface. At this point you have the pure LV model (the original LV surface) and the pure SV model (the SV parameters that yield a nearly flat leveraged LV surface). Both models fit the vanillas. The SLV Calibrator then solves for a mixing fraction the mixes the two models together either by a) calibrating to selected exotics, or b) using historical data.

SciComp adds cutting edge commodity models.

Mon, 12 Mar 2012 16:29:20 -0500

SciComp is pleased to announce a new research paper "Putting Smiles Back to The Futures" that describes a practical method to extend the classical log-normal Gabillon model to incorporate volatility smiles; and a forward market model for pricing options on Forward Freight Agreements (FFA) with a term structure of FFA volatility.

London Seminar: GPU-Accelerated Derivative Pricing Models

Thu, 29 Sep 2011 19:48:00 -0500

Seminars on GPU-Accelerated Derivative Pricing Models
London, October 17, 2011
Presented by SciComp Inc., NVIDIA Corporation, Dell Inc., and Microsoft Corporation.

Sign up today!

SciComp Speeds Derivatives Performance with Support for New NVIDIA Hardware and Software and Enhancements for CDO Pricing and Commodity Futures

Tue, 26 Oct 2010 18:47:26 -0500

Financial institutions are turning to graphics processing unit (GPU) computing for real economic and performance benefits. Fast and accurate derivatives pricing model development and accelerated execution speeds are crucial for today's derivatives marketplace. SciComp Inc. has enhanced SciFinance®, its flagship derivatives pricing software, to help quantitative developers further shorten Monte Carlo derivatives pricing model development time and create models with faster execution speeds. SciFinance® now features support for NVIDIA® Tesla&trade 20-series GPUs and CUDA™ 3.0.

New Enhancements in SciFinance

Mon, 04 Oct 2010 16:37:29 -0500

SciFinance now automatically generates GPU-enabled pricing & risk model source code that runs up to 50-300x faster than serial code using a new NVIDIA Fermi-class Tesla 20-Series GPU.

SciFinance^® is a derivatives pricing and risk model development tool that automatically generates C/C++ and GPU-enabled source code from concise, high-level model specifications. No parallel computing or CUDA programming expertise is required.

SciFinance's automatic, GPU-enabled Monte Carlo pricing model source code generation capabilities have been significantly extended in the latest release. This includes:

Support for CUDA 3.0.
- At a keystroke, users can recompile for CUDA 3.0 for instant speedups.
Support for NVIDIA Fermi-class GPU hardware (Tesla 20-Series).
- These new GPUs feature ECC memory, 448 streaming processor cores, and a 7X increase in double precision performance compared Tesla 10-series GPUs.
Astounding accelerations: typically between 50X-300X faster than serial code.

New models for single-tranche CDO pricing and commodity futures

The Stein stochastic recovery model for single-tranche CDO pricing has been implemented. It is about 1000 times faster than the Krekel grid method.

Serial: 5.05 millisec per tranche-scenario
- Running the STCDO serial stochastic recovery code with 10,000 scenarios on a Windows XP machine with 2.4GHz CPU, the total serial execution time is about 606 seconds, for an average of 5.05 millisec per tranche-scenario.
GPU-enabled: 30.6 microsec per tranche-scenario, 165X faster than serial code
- Re-synthesizing and running on a single Fermi-class Tesla 20-Series GPU, execution time is reduced from 606 sec to 3.67 sec (30.6 microsec per tranche-scenario) for an acceleration factor of about 165X over serial code. Thus, realistic portfolio risk computations for several thousand tranches and 50,000 scenarios could execute in O(~1hr).

A new market model for commodity futures with storage costs has been added to the commodities example suite. Storage costs are characterized by the contango limits of the futures curve. A no-arbitrage argument is applied to determine the relation between different futures. The model is applied to price commodity derivatives of several kinds.

Software vendors and service providers ease GPU adoption.

Mon, 04 Oct 2010 16:38:26 -0500

SciComp was featured in the recently released Ovum report "Graphics Processing Units in Computational Finance: Approaching the Mainstream (Strategic Focus)."

"...As mentioned earlier, companies seeking to run applications written in proprietary code in environments that include GPUs face the challenge of rewriting them in such a way that they can take advantage of the graphics processors' parallelism. The financial sector is a case in point, as there is a marked preference for proprietary code...However, there are a number of independent software vendors (ISVs) that seek specifically to ease capital markets participants' adoption of parallel programming...SciComp's flagship product, SciFinance, delivers code synthesis technology (i.e. software that generates software) for pricing complex derivatives. In essence, this approach masks the complexity of parallel programming from the end user, while leaving them free to define the characteristics of the pricing model that they want to run on GPUs."

SciComp Inc. - Derivatives Pricing - Summer Newsletter

Wed, 08 Sep 2010 17:32:27 -0500

Read the latest news from SciComp on SciFinance, the complete solution that automates coding and delivers source code (C/C++/CUDA) for derivatives pricing and risk models.

Reval Tries to Reduce Time to Price OTC Derivatives

Wed, 08 Sep 2010 17:33:03 -0500

Reval said Tuesday said it expanded a relationship with an Austin, Texas, firm so that it can more quickly price over-the-counter derivatives.

The New York-based risk management and hedge accounting firm said it will implement SciFinance, a product of SciComp, a provider of automated pricing technology.

The code generation product allows users to develop derivatives pricing and risk models in-house. That will mean companies who use Reval's online accounting services of Reval's software will be able to cut the time it takes to value a portfolio of OTC derivatives.

Reval Speeds Up Pricing Complex Instruments in the Cloud with SciFinance

Wed, 08 Sep 2010 17:32:48 -0500

Reval uses SciFinance from SciComp to accelerate the development and valuation time for highly-structured financial instruments.

SciComp Inc. - Derivatives Pricing - Spring Newsletter

Wed, 08 Sep 2010 17:17:58 -0500

Read the latest news from SciComp on SciFinance, the complete solution that automates coding and delivers source code (C/C++/CUDA) for derivatives pricing and risk models.

Zurich Seminar on GPU-Accelerated Derivative Pricing and Risk Models

Wed, 08 Sep 2010 17:27:22 -0500

A Seminar on GPU-Accelerated Derivative Pricing and Risk Models
Zurich, Thursday, April 22, 2010
Presented by SciComp Inc. and NVIDIA Corporation

SciFinance automatically generates GPU-enabled pricing & risk model source code that runs up to 220x faster than serial code using NVIDIA GPUs.

Find out more about creating GPU-enabled pricing and risk models by signing up for this FREE seminar.

Frankfurt Seminar on GPU-Accelerated Derivative Pricing and Risk Models

Wed, 08 Sep 2010 17:27:16 -0500

A Seminar on GPU-Accelerated Derivative Pricing and Risk Models
Frankfurt am Main, Wednesday, January 20, 2010
Presented by SciComp Inc. and NVIDIA Corporation

SciFinance automatically generates GPU-enabled pricing & risk model source code that runs up to 220x faster than serial code using NVIDIA GPUs.

Find out more about creating GPU-enabled pricing and risk models by signing up for this FREE seminar.

SciComp Inc. - Derivatives Pricing - Fall Newsletter

Wed, 08 Sep 2010 17:18:08 -0500

Read the latest news from SciComp on SciFinance, the complete solution that automates coding and delivers source code (C/C++) for derivatives pricing and risk models.

SciComp enhances derivatives pricing software with v5.0 release of SciFinance.

Wed, 08 Sep 2010 17:22:44 -0500

SciComp Speeds Derivatives Performance with New Parallel Code Styles and SciXpress Features.

Webinar: Automated GPU Computing for Derivative Pricing Models

Wed, 08 Sep 2010 17:27:02 -0500

A Webinar on GPU-Accelerated Derivative Pricing and Risk Models
11:00am-12:00pm EDT, Tuesday, November 17, 2009
Presented by SciComp Inc. , NVIDIA Corporation and Dell, Inc.

SciFinance automatically generates GPU-enabled pricing & risk model source code that runs up to 220x faster than serial code using NVIDIA Tesla GPUs.

Find out more about creating GPU-enabled pricing and risk models by signing up for this free Webinar.