C(omp)++ | A Portal for the Computational Finance Community: derivatives

Software is an indispensable part of most businesses. In the case of computational finance the main challenge is to propose new financial models, define mathematical structures and algorithms for them and then map these algorithms to code. Some of the questions that arise are:

How do we define a repeatable process for modelling derivatives in C++ or C#?

How do the financial, mathematical and software models link up?

What are the requirements for our resulting software products in terms of maintainability, efficiency, functionality and reliability?

These questions are more easily posed than answered for a number of reasons; there are a number of challenges to be resolved. First, the ideal development team consists of quant analysts, developers and other professionals with knowledge of finance, mathematics and numerical analysis. Second, the members of the team need to have a common language (or even several languages) that they can use to promote interoperability. Finally, we need to take account of constraints such as time-to-market, project schedule and budget as well as the requirements on the resulting software product.
The techniques that we can use to promote computational finance as a multidisciplinary activity are:

Good mathematical and numerical models that unambiguously define derivatives products

Using the de facto modeling language UML (Unified Modeling Language)

Using proven Design and System Patterns

Using multi-paradigm programming languages that support the object-oriented and generic programming models

I shall be speaking about these software-related issues at the free evening seminar in the City of London on June 26 2007 at Charing Cross. More information and registration (you must register in order to attend), see http://www.datasimfinancial.com/course_detail.php?courseId=13

For queries, please contact Ilona Hooft Graafland at ilona@datasim.nl

Let me know if you have any questions.

Daniel