Tall Tools

For tall buildings you need tall cranes. This impressive crane is not tall enough to maintain the Centre Point in London. But, the principle construction of the crane looks to me configurable from parametric building blocks.

Computers give us even more flexibility for parametric design.

For the valuation and risk analytics from the single deal type to the most comprehensive portfolio, we have tuned C++ based pricing and calibration engines for accuracy, speed and robustness. Algorithms are designed to work in next generation CPU+GPU architectures.

Integrated into Mathematica 7, built-in parallelism exploits multi-core processors. An UnRisk seat comes with 4 parallel computation kernels that can be extended. Single power users can extend to gridUnRisk Local, power user groups to gridUnRisk Server.

This ENGINE combined with transaction processing and task-building-and-scheduling services, build the UnRisk FACTORY that crunch even massiver valuations automatically, and with lightning speed. Accessible from a web browser. We tie together building blocks built in C++, Java, Mathematica, SQL and HTML.

For cranes quickly customized to the purpose and enabling crane operators to monitor the crane that automatically completes all scheduled operations with full precision, from at home.

DESIGN REAL

This weekend my wife and I made a lightning-visit to London. We dived into the shopping-parades, like at Regent Street, but most important visited our favorite small spot for arts, the Serpentine Gallery at Kensington Gardens. Current exhibition: Design Real, curated by Konstantin Grcic, one of the world's leading industrial designers. As the title of the exhibition suggests, all of the selected products are 'real', mass produced items that have a practical function in everyday life. Grcic: good design admits to the deeper insight that beyond performing a purpose in a good way, the purpose itself has to be good.

My quick thoughts:

a. we at UnRisk also believe: everything from concepts, models to the concrete implementations must lead to a product for a special purpose, but be built upon clear principles.

b. Real industries outsource to independent industrial designers, even for core developments. Financial industries still hesitate ... ?

300

No, I do not recommend this "phantasy" movie about King Leonidas and his 300 Spartan Warriors.

Seing the trailer, I just remembered that the Battle of Termopylae, has been an integral part of the curriculum in my high school, and I am sure many others.

It is believed from mythology, that Leonidas and 300 (risk-trained) Spartan warriors were the dominant force of a Greek army of between 5000 and 7000 strong to hold the pass of Termopylae against the massive Persian army under Xerxes.

Risk aspects: the oracle of Delphy has warned Leonidas that eiher Sparta would be destroyed or he would lose his life. The pass was chosen with care. The Greeks were armed with long spears to prevent persians to getting close with their weapons. After they have repulsed the attacks for the first days an unexpected risk was caused by a Greek traitor, who showed the Persian a path around. Knowing that they will be outflanked, Leonidas sent away all Greek troops, because he cared about their safety to preserve them for future battles (only 900 Helots and 700 Thespians stayed).

They finally were cut down, but saved Sparta and in the following Athens and Greece.

Away from weapons, wars ... what we want to do for risk management: provide powerful technology optimized for certain usages ( a clever handful is enough ) in a minimalist infrastructure, giving in-time decision support. A system that is widely usable (even from smart mobiles) , act powerful and robust where required and collaborate with the surrounding systems. And we train risk experts.

Disrupt Or Re-Invent Ourselves?

How vulnerable is our business? As vulnerable as that of the market participants in our market segments.

For years, we were in the given-a-problem-searched-a-solution innovation business. In modeling and simulation of complex technical systems.

1997, we entered into the quantitative finance market by developing software tools for complex convertible bonds for the London trading group of an American bank. We saw many commonalities with our technical problem solving as well as abstractions. With UnRisk, we transformed ourselves into product-and-services providers, and our projects became kind of reverse innovation: developing generic technologies from the concrete local requirements of our customers.

In return, we decided to transfer our own technological and implementation benefits to our customers (Because We Are Nice ). By high-value / low-cost offerings (low-price software in minimalist infrastructure).

Having given our new UnRisk Academy a special mission, we extend this to transfer also the know-how-to-exploit-the-benefits by giving insight into models, methods and technology implementation. We are not entering the seminar business, we are entering the know-how package business. We learn from teaching and get a next level of abstraction enabling us to implement automatic model & method selection, precision control by parameter optimization, decision suport, ... And the innovative co-evolution will go on.

Work and Knowledge - Wellsprings of Our Lives

compiled from Wilhelm Reich.

Today, we have launched UnRisk Academy, a new business unit that will package and transfer know how.

Why ? From the beginning we are highly committed to serving our customers individually, whilst driving generic technology. Generic methods are already a first pre-requisite for explorative learning and consequently our product user training can be concise. In addition the document centered approach (Excel Workbooks, Mathematica Notebooks) allows for electronic exchange of use cases and makes support constructive and insightful.

And we describe our valuation techniques in Technical Papers .

But we found, because there is background, this is not enough. Knowing the background helps not only developers, but also users, to avoid any methodological and technological risk that can become horrible in interplay.

This is why our services group is built of top researchers and developers.

UnRisk Academy extends the already dynamic packages for product user training and other exploration resources with a brand-new course framework enabling quant trading and risk experts to understand the quantitative theories, mathematical schemes and critical implementation aspects in detail. Models for Underlyings, Deals Types and Cases, PDE, PIDE and their Numerical Treatment, Montecarlo and Longstaff-Schwarz, Calibration will be covered in modules, each designed for 3 days.

The courses will be live-in-classroom with practical hands-on examples. Continual support via online exchange of examples is offered. Focuses on the practical side of quantitative finance....

UnRisk Academy will tie all learning aspects of UnRsk together and complete UnRisk offerings from product/services to Know-How Packages.

The Risk Analytics Factory

In Wired Nov-09, I read under World's Biggest Digital Brains that those, with names like Blue Brain, Discover, Ranger, .. are able to handle "many" teraflops, studying the inner working of the human brain, climate simulations, .....: Number of processors up to 16.000 for 500 Teraflops.

In Finance?

You cannot predict extreme events. Studying the past will also not help to manage all risk. You need to ask What-If, Provided-Then, .. questions?!

Simulate deals across comprehensive scenarios can become heavy within a bank, not to speak about market segment, sectorial analysis, .. Such simulations are not performed, because of technological limitations.

A portfolio of 1000 moderately complex structured instruments (average valuation time 10 sec) across 10.000 scenarios would take more than 3 years on a single core process. To get the result in 10 seconds, a speed up of 10 Million was required.

Can we imagine to set up affordable risk-answering-factories with such a performance?

First, kernel sw would need to take its part, say, by doing principle component application , surrogate models. Speed-up of 1000. Still 10.000 to do.

Put part of the valuations on a Personal Supercomputer from NVIDIA (quad-core machine plus GPUs), and you can get a speed-up of approx.100.

Still 100 to do. With the help of Mathematica's symbolic parallelization techniques, we have made UnRisk grid-enabled and optimized for close to linear speed-up related to the available core processes. Provided, we take a networked pool of 128 core processes. 32 Tesla personal Supercomputers for the job?!

Luckily, we have selected the right algorithms and technologies to multiply speed-up. NVIDIA implementation to come next. And yes, it is possible.

If UnRisk was a Bike

it was most probably a YikeBike . It is rated one of the best 50 inventions of the year 2009 in Time magazine. YikeBike says: YikeBike is a statement about using smart technology to solve the problems of our increasingly congested, polluted, stressful cities.

In this case I relate it to our UnRisk-Q the quant-computing-and-developing system that packs the latest technologies into a minimalist infrastructure of a supercomputer-under-a-quants-desk. Shhh on NVIDIA Tesla. As YikeBike to be available in Q2 2010.