Demand Deep Design

In Surf the Exponentials and Bank on Openness I have written about our treatment of rules to relating and responding to ideas that might change the word, as they are published in How to Spot the Future - Wired Magazine UK Edition - Jun-12 issue. Today I take Demand Deep Design. Again with the eye of a mathematician who works at the intersection points of mathematics and computer sience - applied to our quant finance developments.

First, for us this is probably the most easiest rule to understand. Studying the design of Mathematica, you understand deep design in software. And using Mathematica as a platform we have inherited the most important principles. In A short story of being lucky, we have summarized how this paid back.

Design, at its best, let users prioritise. In depth the simplification begins with the representation of any objects and actions by a unified expression model. And at the front end you have a language that is descriptive and applies principles of the language of mathematics but transformable into domain-specific forms.
UnRisk extends Mathematica into the universe of financial derivatives, structured products, portfolios, scenarios and what have you. Valuate becomes a generic interface to the valuation of the simplest to the most sophisticated deal types of equities, FX, interest rates, inflation, commodities or credit.

But to prioritise needs more. Automation, Built-In Knowledge, Multi-Strategy and Multi-Method Approaches, Evaluation and Information Management.

In our Reinventing UnRisk effort, we are lucky again to be able to combining the best practices we experienced over the years and extend them with new ideas. Without changing any of the principles.

One is to understand a term sheet of a financial instrument as a "language" that talks about financial objects and events. And we represent it in layers where the core layer provides the operational semantics in real environments of markets. It calibrates and re-calibrates models to market behavior and calculates millions of results in short time that are compared, aggregated, visualized, .. for more risk insight and better trading decisions in a multi-agent environment.

BTW, the remaining rules in the article are Look for Cross-Pollination (mathematics does exactly that), Give Points for Audacity (not in our business?), Favour the Liberators, Spend Time with Time-Wasters. Yes.