UnRisk Insight

Will Technology Outsmart Us?

2012-02-09T08:33:00.000-08:00

Does Complexity Economics need complex simulation systems to gain insight and even beyond, complex control systems to being "managed"? A global controller is in contradiction to complexity.
I see a system as complex, if it contained subsystems of co-evolution.

Traditional economic models see an interplay between countless market participants who acting more or less rational bringing the markets towards a balance and external impacts mostly based on events as well as rules and technology developments.
In short, financial instruments taking their realistic values would "freeze" - the market would settle down - without external "shocks".
Do they? Complex, constructive learning plays an important role in such complex systems leading to an internal dynamics even with the absence of external shocks.

M Buchanan went into details in his post Minority Games. Really exciting.

But let me step back a little. Will there be a co-evolution of human and artificial intelligence or will technology just take over? In other words, will we, will we need to stay competitive?

I have a background in factory automation and in this field feed back loops are still used to control machine tools, robots, transportation and storage systems. There will be a better way of interaction based on simple local intelligence.

What about an economic agent system?

Let me escape from the feasibility question.
Can competition in principle mean "fewer" losses?
Humans build networks to enhance knowledge, interact across cultures, ask great questions, engage opposite views, .... they co-operate.

Therefore I do not see the future spectacled, say, in the Matrix films.
What interacting with technology teaches us will be important and consequently, we need to design systems strictly bottom up.

Open For Open Innovation?

2012-01-26T04:14:00.000-08:00

In a first view Open Innovation sounds like a generalization of the Open Source idea. Free intellectual property rights?
The term was promoted by Chesbrough, H. (2003), Open Innovation: the New Imperative for Creating and Profiting from Technology. Harvard Business School Press.

To my opinion ist is gaining and sharing knowledge, involving other parties when developing new products and technologies and combining internal and external ideas in the flow from research to packaging.

As a spin-off of distinguished mathematical research institutions the UnRisk Consortium lives the open innovation principles from the beginning.

Having decided to unleashing the programming power behind UnRisk we stretch this principle to customers and partners, who want to work with us on the basis of know-how packages.

This is already part of our business model since we have released UnRisk-Q. With the All-New UnRisk realization it will be further perfected.

UnRisk-Q enables quant developers to increase the effectiveness and efficiency of their innovations. They concentrate on the creation of their innovations relying on white-box tools from UnRisk - the same that are integrated in widely used UnRisk solutions at banks, capital management firms, insurances, ..

In short, we, at UnRisk, are here for innovation - we partner on innovative organizational and licensing arrangements for better technical and economic feasibility of our partners projects - transferring better tools for advanced valuation and risk management processes.

The Internet Of Things

2012-01-16T04:39:00.000-08:00

The internet of things (first mentioned 1999) refers to the representation of uniquely identifiable objects in the internet. Autonomous and intelligent entities shall interoperate online. Alliances are built to promote this idea.

In June, 2009 I was asking myself in The Internet Of whether domain specific The Internet Of ... instantiations would emerge ... like the Internet Of Finance?

In The Internet Gets Physical, The NY Times, Steve Lohr summarizes that low-cost sensors, clever software and more powerful computers are are opening the door to new uses in energy conservation, transportation health care and food distribution.
This might include "soft sensors" for all kind of intelligent analysis including sentiment analysis.

In finance we have massive market data, computational knowledge, we could introduce more intelligent data analysis, we have enormous computing muscles and we have the internet.

But I still do not see an initiative for The Internet Of Finance. For the management of systemic risk, identifying anomalies and unexpected market shifts, the transitions from swarm into avalanche behavior ...

UnRisk FACTORY 3.1 Is Released

2012-01-12T02:21:00.000-08:00

UnRisk took UnRisk FACTORY 3.1 to financial institutions.

It completes the UnRisk VaR Universe with Montecarlo VaR for all instrument types. The VaR Universe is integral part of the UnRisk FACTORY Capital Manager used for both advanced investment functions and risk management processes.

In May-08 UnRisk has releases UnRisk FACTORY. Now, UnRisk FACTORY 3.1 is the 8th release.
Customers enjoy faster time-to-insight, integrated and automated valuation and data management.

Driving the latest and generic technology enables us to support HPC on minimalist infrastructure and offer UnRisk for unprecedented low cost of ownership.

Optimal Risk

2012-01-04T07:58:00.000-08:00

Between Xmas and New Year, I took a few days off for cross country skiing in the heart of the Bohemian Forest.
It seems that I took the optimal risk.

I am over 65 and taking this in account, I selected the right skating trails, lengths and duration. Challenging steepness, skating long and swiftly enough to have a lot of fun and add to fitness incrementally, but avoiding hazardous, exhausting actions - so I never fell down, nor suffered I from a total fatigue of joints, ...
I run as often as possible in winter time, for 40 years now, so I have some experience in cross-country skiing risk management.
Maybe the same time as the very first quants have with quantitative financial risk management, wondering how many would care about optimal risk? ;)

Agenda 2012 - The All-New UnRisk

2011-12-23T03:25:00.000-08:00

We achieved our objectives described in the Agenda 2011.
We made the box blazingly fast and even whiter. And we continuously unleashed our innovations empowering quant developers to use the same programming power, we use to swiftly develop solutions, like the UnRisk FACTORY and FACTORY Capital Manager.
UnRisk engines are now at version 5 and the FACTORY at version 3. The VaR universe runs atop the UnRik-Q and the FACTORY providing information cubes resulting from time travels of portfolios and simulation for the most advanced risk management processes.
UnRisk goes cross platform.
And we have added tools for a tighter integration of UnRisk-Q and the FACTORY, linking UnRisk-Q to the FACTORY data base.
This enables system integrators to exploit our combined valuation and data management and transform results into insight with unprecedented low efforts.

In 2012 we will go far beyond this.
This year I describe the Agenda from a common view of the core team: Andreas, Michael and Michael, Sascha and myself. It will not be detailed, but I will keep you informed during the year.

THE ALL NEW UnRisk.
What does this mean? Is it to remove weak points? A simple optimization task?

Not at all. I like Wittgenstein's Ladder in Tractatus Logico-Philosophicus: to go beyond you must throw away the ladder you climbed up (but you need to have climbed up).
In our business the ladder, a pattern, can be at the algorithms and implementations level, but also the design and programming language level.

OpenCL driven over Grids
We have transferred the most sophisticated numerical schemes from solving complex technical problems to finance and in thousands of practical tests, we have proven their advantages over the most widely used methods to solve financial PDEs, PIDEs, .. in accuracy, speed, stability and robustness.
With the new computing muscles we get new opportunities of solving. We can apply simpler methods for multiple usage.

We were early adopters of this future technologies and we are now able to re-implement the code base optimized and platform-agnostic for heterogenous architectures - in OpenCL - and verify new methods and implementations with our most advanced solvers running in the traditional architectures.

A Unified Description Layer in C++
Having integrated our optimized C++ engines into Mathematica, we have designed financial objects and operations to be described in a declarative, domain specific language, exploiting Mathematica's functional programming paradigm and extending its language into the universe of derivatives, portfolios, scenarios, models, methods, schedules, events, .... In not so rare cases, we develop UnRisk engine features in UnRisk. And quants develop their customized solutions quickly.
This symbolic layer made it easy to parallelize and make applications web enabled.

Having gained deep insight into the power of such structures, we will introduce a new layer with a unified description of the financial objects and their manipulation - in C++.
The design of the interface of this layer will be generic in the sense that interfaces to any other programming languages can be built with unprecedented low effort.

This all-new UnRisk Core will provide all bank-proof instruments, models and methods, organized orthogonally by the UnRisk systems. And much more.
In addition to the existing programming front-end a programming layer will be provided that makes it a programming Chameleon. Quant developers, and ourselves, will rely on the same unified description and accessing the identical blazingly fast pricing and calibration engines on traditional or new computing muscles.
A new intelligence in instrument-model-method combinations will allow the even quicker introduction of new (hybrid) deal types.

2011 - Unleashing The Programming Power Behind UnRisk

2011-12-06T02:02:00.001-08:00

2011 was the year of UnRisk-Q - the culmination of the co-evolutionary development of our bank-proof UnRisk Engines and the UnRisk FACTORY, offered to quant developers.

A short trip down the memory lane through 2011:

THIS DAYS IN UnRisk

14/15-Mar-11 Frankfurt - UnRisk on transtec - Feel The Heat
11-Apr-11 UnRisk VaR Universe released
30-May-11 UnRisk 5 released
Jun-11 London, Paris, Zurich, Frankfurt - Accelerating Quantitative Analytics in cooperation with Wolfram and NVIDIA
23-Aug-11 UnRisk FACTORY Capital Manager launched and taken to Schoellerbank Invest AG
21-Sep-11 UnRisk FACTORY 3 released
20-Oct-11 IDRC partners with UnRisk

But this year was also the start of a comprehensive refactoring: UnRisk Goes Cross Platform. And this means a radical innovation to removing the irrational fear of new architectures. A critical part of the overhaul of the core base of UnRisk Engines: core algorithms are uncompromisingly re-implemented in OpenCL that will be driven over the grid - supporting mixed architectures with one code.

2012, UnRisk for hybrid and cross platform architectures will be releases to support blazing and cool business. But algorithmic re-design also brought new generic foundations were born, enabling us to introduce open hybrid product valuation and analytics. UnRisk-Q will be extended to link the UnRisk FACTORY with any risk related system.

In 2012 we will convince additional quant developers that it will strengthen their situation drastically when building their know how packages atop UnRisk.
We will continuously produce big systems for the small by devoting our UnRisk FACTORY Capital Manager to small and medium sized capital management groups.
And unleash our innovations to financial consulting and services institutions.

I will post the UnRisk Agenda 2012 soon.

Aaron Brown - Red-Blooded Risk

2011-11-29T04:01:00.001-08:00

I am not a book reviewer. But there are article writers who I follow. One of them Aaron Brown, who is a frequent writer in the Wilmott magazine. He is an original outside-the-box thinker and it is insightful and entertaining how he touches risk in a general sense and in financial practice. I wrote about his pointed view on VaR in VaR of the jungle.

I just read his brand-new book Red-Blooded Risk and I highly recommend it: must-read for risk takers and managers. It addresses risk from an interesting set of angles and it motivates and guides how to thinking about risk in an intuitive way - don't mis-interprete them, but don't forget the figures, I hear him say and use more approaches and models to get insight.

Are We a Good Company? Is This a Luxury?

2011-11-17T02:12:00.001-08:00

Inspired by the HBR November 2011 issue's cover story What Great Companies Do Differently I thought, are we good? What is the meaning of "good"?

I remember, in traditional theories the notion dominated: good companies transform something into big margins. In our case it would be knowledge? Are we a machine that generates money from knowledge?

When I worked in the factory automation business it seemed quite simple. Good companies had sophisticated products represented in sophisticated blueprints and industrial practices of highly automated discrete manufacturing, adopting Japan's quality-driven production schemes. Innovations in technical computing were geometric modeling, kinematics and dynamics of mechanisms in order to program and control flexible manufacturing systems of CNC machine tools, robots and automated vehicles, chaotic stores, ...

When economy of scale became important and companies strived to become world enterprises the metrics of goodness was: shareholder value. Multinational tech and pharmaceutical industries dominated the lists of business leaders. Innovative computing meant enterprise-wide, integrated systems to optimizing planning and operations.

With the emergence of the Silicon Valley innovative company became the synonym for good company. Not so much for blueprint innovation, but for what I call: naive innovation.
The innovation atop of universal machines that are not mere tools. Computers and their open networks. Not surprisingly the acceleration of the innovative spiral led to bursting bubbles, the .com bubble the most spectacular. But it developed a broadband infrastructure that enabled the emergence of "social networks" and what have you today.

In the petabyte and petaflop age, the scientific method - have a hypothesis and test it - might become outdated and replaced by brute force methodologies and data analysis and presentation techniques.

In this century we had crises that seemed to have clear cause but it turned out that they are layered and overlapped.
As one of the consequences it is recommended that good companies produce social as well as economic values?

In computing, we have open source movements, public grid initiatives and cloud computing (virtualization), ... driving innovation.

Today, as public anger against the financial industry grows, what social values can be produced by us technology providers?
What is social when contributing to the quantitive approach to managing risk?

First, we deliver know-how packages - the financial business is risky enough, no need to add technological black-box risk.
Second, with UnRisk-Q we unleash our own development environment - why not enable creative developers to transform their knowledge into solutions without reinventing foundations?
Third, our licensing and pricing is bottom up, not market clearance - UnRisk position is highest possible value at the lowest possible price.

Is this luxury?

No. We are a comparatively small outfit.
We were lucky to chose generic technologies and tools that enable us to develop quickly and cost-effective.
We also work on the most complicated technical problems (combustion engines, chemical reactors, metallurgical processes, observatory, ...) that require similar mathematical schemes and approaches as quantitative finance and use the synergy.

And we are and will remain an independent boutique without ambitions for uncontrollable growth. We rather built an enterprise without borders.

It is not that we do not want to make money, but it is as essential to earn recognition by our blend of competences in financial engineering, technologies, models, methods and implementations.

On Time Travel

2011-11-16T06:46:00.001-08:00

Time travels raise a lot of questions (like the grandfather paradox) - political regimes banned time-travels-in-culture, because they were "irreverent" - science says it can only go forward.

Portfolios can travel back in time. Providing historic market data are available portfolios could be valued due to "real" market data scenarios of the past, assuming that they are valued with the present models and schemes. You might be interested in return-risk profiles that your current portfolio would have had in certain periods of the history. Maybe your interest is focussed on the most turbulent periods, where markets showed unexpected anomalies. You could do a broad variety of portfolio across scenario time ride.

You know that the historical behavior and risk does not tell you anything about future risk, in the sense of forecast, but it creates insight, what might happen. And especially when driving through stress periods you might get information on the sanity of your portfolio with respect to certain risk factors.

The historical VaR, as an example, is calculated due to market date scenarios of the past. Due to the time series of given risk factors the historical change of these risk factors are calculated. By applying these historical changes to today's risk factors, the historical scenarios are generated. Then each individual instrument is valued under each historical scenario - the difference between these scenario values and today's value are the scenario deltas that are sorted and so on.

The output is a cube spanning VaRs across risk factors and instruments in the portfolio. In the UnRisk VaR Universe also the deltas are accessible for further testing and analysis. So the traveling in the past is in a hyper space of information for a better understanding of sources and impact of risk factors. The portfolio behaves like a chameleon with changing the contribution of instruments to the portfolio VaR - it is not always the same who behave badly.

Hopefully really we are not traveling back in time, because the future is just to horrifying.

We Did it Right by Doing it Wrong?

2011-11-11T04:20:00.000-08:00

Technical and economic feasibility do not always go as a couple. Since we launched UnRisk, we made decisions that might have seemed questionable at the time they were made. Design principles, algorithmic innovation, transparency and openness ... do not always pay back immediately.

REINVENT THE FOUNDATION - when we entered into the quantitative finance market, tree-based and Montecarlo methods were widely propagated.
We transferred numerical schemes from complex technical system solving to finance:

Methods for the stable solution of partial differential equations. Keywords here are finite elements, asymptotic math approaches, and techniques for handling convection-dominated equations (upwinding, streamline diffusion).
Methods for robust calibration. As the parameter identification problems in finance are typically ill posed, one should know about the regularization techniques which make ill-conditioned problems tractable. If the forward problem has no closed-form solution, adjoint method techniques are used.
FFT and wavelet techniques.
and clearly (Quasi)Montecarlo with Longstaff Schwartz ...

The choice of the right solvers is the very kernel. They enabled us to extend the coverage of deal types unprecedented quickly. In risk management, where numbers are crunched many millions of time, they must be accurate and robust.

And even now, having enormous computing muscles, allowing for algorithmic simplification they are indispensable benchmarks in model validation.

USE MIXED LANGUAGE CONCEPTS - when we started, not only solvers were written in C++. The learning and interaction effort for teams is minimal, when having a one-language development environment. We decided to integrate our number crunchers into Mathematica, not common in financial circles at that time. We exploited two major principles of Mathematica:

a declarative programming environment
link technologies that allow the integration of data bases and other languages, like Java, Python, ...

We created a domain specific language enabling us to develop parts of UnRisk in UnRisk and built our UnRisk FACTORY that combines valuation and data management controlled by Java services and SQL, and webbed-it by web services, Java server pages, ...

This house stands stable, although no implementation stone is left unturned (UnRisk Goes Cross Platform) right now. Driving OpenCL over grids.

UNLEASHING THE INNOVATION - making UnRisk-Q, our own development environment, available to quant developers we reinvented our business. It was not an obvious decision. But to maintain it without compromise, we have established the UnRisk Academy completing the tools to know-how packages,

We want to do where we are good at. And serve customers first, whilst driving generic technologies. We rather want to partner and run a company without borders than growing by high pressure strategies ....

UnRisk goes Cross Platform

2011-11-09T04:12:00.000-08:00

UnRisk-Q – Removing the irrational fear of new architectures

2001, when UnRisk started, quant finance practitioners in banks utilized almost exclusively Wintel platforms. Consequently, UnRisk was optimized for Wintel.

But now, we use MacBook Pro with 8 cores and an AMD GPU for presentations and workshops. Our customers utilize, say, DELL servers with, say, 128 CPUs and four NVIDIA Tesla C 20 series GPU cards. My iPad is a powerful little machine, as well as Android based tablets..

More than a year ago, we decided to overhaul the UnRisk engines code base to make it compatible with other target operating systems than Windows and to support modern processor architectures and utilize the latest computing muscles. See The UnRisk Agenda 2011

The first outcome of this refactoring effort will be a Linux and OS X version - UnRisk-Q version 5.1 to be released quite soon. But our spade-work embraces the radical redesign of core algorithms to be implemented in OpenCL. Yes, this is a step back into programming in the small, but the benefit of platform-independence pays back.

This blazingly fast one-for-all code will run on current and future technologies. Transparent to users it will utilize the power of the CPU cores, combined with massive parallel architectures. The level of parallelization granularity is defined by the architecture. OpenCL code is driven over the grid.

Double benefit from a high level of platform independence

Quant developers, UnRisk-Q users, enjoy the double benefit of the blazingly fast engines that can be programmatically manipulated in a declarative financial language framework.

There are two great principles in UnRisk's platform independence. The first is that the messy plumbing of the utilization of hybrid platforms is automated. The second is that most that can be done in UnRisk-Q can be done optimized without change - from a small device to a large HPC farm.

Turning On The Innovation Spiral

2011-11-02T07:35:00.000-07:00

Last week we conducted another development workshop with Solventis.

Solventis uses the UnRisk FACTORY as a valuation and risk engine producing massive information from portfolio across scenario and VaR calculations automatically.
That information is then transformed by distinguished analysis and test routines into special insight.
All together offered as SaaS/web platform to their customers for better decision support.
I wrote about the development partnership in A Barcelona Cloud That Rains Insight.

Solventis' experience and eminent practical approaches to managing investment and risk for so many different market participants are amazing.

At the workshop we concluded that all FACTORY objects - market data, models, scenarios, valuation parameters, instruments - and information evident over the whole history in the FACTORY data base shall be made available within UnRisk-Q.
UnRisk-Q will be linked tighter to the UnRisk FACTORY by extending it with a package that imports UnRisk FACTORY objects and results. Those results are not limited to valuation and risk factors, but interim results, like VaR deltas will be made available.

This will enable Solventis to transforming their distinguished knowledge into further sophisticated products and services atop both the UnRisk FACTORY and UnRisk-Q (web enabled with webUnRisk).

Remember, UnRisk FACTORY and UnRisk-Q use identical valuation models and methods and calibration schemes as well as portfolio and scenario constructors, because the FACTORY-based engine is built with UnRisk-Q.

Solventis will receive the package in advance. It will then be integrated part of UnRisk-Q 5.1 and later.

IDRC, A Speciality Management Consulting, Partners With UnRisk

2011-10-20T04:11:00.000-07:00

IDRC is a professional services and consulting expert in Data Analysis and Quantitative Modeling and Research, based in Frankfurt, London, Moscow and Mumbai. Their consulting is packed as two broad formats: ProcessAnalytics and Quant Research Center.
Both these get direct involvement of IDRC directors, IDRC advisor network, IDRC,s own analytics standby facilities and affiliates. Under the Quant Research Center, IDRC goes on to develop client's dedicated team, systems and infrastructure.
IDRC has developed Quant Incubator for Evaluserve, and an enterprise-wide quant support desk for Bank of America and has executed structuring, derivatives pricing and risk quantification assignments for top banks and buy-side firms globally, including Citigroup, RBS and Barclays in London, Kotak goup and IL&FS in India.

UnRisk offers a selected range of products that enable financial institutions to focus on in-time decision support for valuing and managing risk from the single deal type to the large and diversified portfolio. For entereprise-wide pricing and analytics strategies all UnRisk products access the same blazingly fast but accurate and robust valuation engines that avoid cross model & method risk that usually becomes horrible in interplay.

"With this synergistic affiliation with UnRisk, IDRC will now offer its buy-side and corporate clients, a cost-efficient access to UnRisk's cutting edge valuation, risk & analytics platforms, and UnRisk-powered service solutions. Thus we expect to capture a huge untapped demand of cutting-edge models and analytics, packaged in client-centric cost efficient service framework.

UnRisk's platforms will enable IDRC to develop and integrate solutions with unparalleled performance and with unprecedented low effort. IDRC expects to cater effectively to highly demanded services of OTC derivatives and structured trade valuation and risk to clients, bringing further client-acknowledgement to UnRisk’s product range from India, Russia, UK and Germany", summarizes, Amit Batra, Director of IDRC, "The concurrence of the two firms, on synergy and trust based business philosophy and passion for quality and innovation, makes me confident that this affiliation will thrive and bring excellent cash returns", he adds.

I agree, we package know how partnering with the most talented individuals and groups for service and business development on and of groundbreaking technology platforms.

EMH Is An Illuminating Idea - But Not Reality

2011-10-12T06:43:00.000-07:00

Recently, John Kay received attention on his essay The Map Is Not The Territory.
He spells out methodological critiques of economic theory in general and rational expectations in particular (interestingly enough 2011 Nobel prize winners Sargent and Sims - received the price for their research having centered around the rational expectation framework).
In the core of Kay's critique: the EMH is an illumination idea, but it is not "reality"

INET forwarded Kay's paper to a handful of economists ...
Roger Guesnerie
Paul Davidson
Michael Woodford

Guesnerie: The optimism embedded in the EMH has been one of the sources of the recent turmoil.

And I also found Crazy Economics Models in The Physics of Finance a blog emphasizing on positive feed backs that often drive markets away from equilibrium.

It is a complex world and I question myself whether economic models should'nt be calibrated and re-calibrated to real market behavior more frequently? If market participants have rational expectations about the future and the current behavior can only model an approximation, frequent re-calibration (inverting) can help to understand better and predict a little further?

Model Validation - New or Used?

2011-10-04T05:33:00.000-07:00

Validate models qualitatively and quantitatively that are used by the front-office.
Is it enough to assess, whether a model or methodology is fit for its purpose?
No. What is the subdomain where it is? Can it be easily extended to become more adequate covering a broader variety of deal types (extrapolation is dangerous)?
Not surprisingly the quality assessment of model risk is not so different from any assessment of a mathematical problem solving process.

is a model describing the problem applicable in principle?
is it correct in describing the reality?
can it be transformed in a form that can be solved in the expected accuracy?
are there numerical schemes that approximate the model adequate?
are the inverse problems of calibration solved correctly?
are the implementations adequate, precise, stable and robust?
are the data right and curated?

In short, models need to be theoretical sound, mathematical adequate, correctly calibrated and solved and error-free implemented.

It is obvious that model validation is easier when the assessing models, solvers and implementation are organized orthogonal.

Models need to be consistent with market practices and contractual terms and it is vital to know, how they impact variations in model parameters, say, under stress conditions.

And because process-through simulation across many factors is important, they provide the most comprehensive assessment of model risk, it is obvious that speed matters.

We all have observed that models can be calibrated perfectly to price a set of liquid instruments, but mis-price exotics. The fit is so well, the parameters obtained exhibited good time stability and robustness, why is the price from different volatility models so different? This raises the question of the required frequency of re-calibration and the impact of the models.

For such a complex task, model validation teams might want to be equipped with a technology platform including pricing and calibration engines and a domain specific declarative programming environment for defining higher level valuation and simulation tasks. They need to run scenarios and cross-model tests independent from the front office model selections and valuation practice.
They need to come as know-how package - UnRisk-Q.

Did Our Training Change In Response To The Crisis?

2011-10-03T08:38:00.000-07:00

Yes it did. In 2009 we launched the UnRisk Academy that extended our product use training with courses providing full explanation on quantitative theories, mathematical approaches and critical implementations.
Markets had shifted to new regimes characterized by almost unimaginable anomalies and even familiar models and valuation techniques became unmasked as unreliable.

The courses cover 6 modules in 17 days. They have been intensively used by quants and risk professionals who want to avoid any methodological and technological risk and traps and consequently they benefit from finding the box even whiter.

Academy courses will be customized for capital manages for whom we offer the UnRisk FACTORY Capital Manager solution to be used for both advanced investment functions and risk management processes.

UnRisk FACTORY 3 Is Released

2011-09-21T01:08:00.000-07:00

21-Sep-11, UnRisk took UnRisk FACTORY 3 and UnRisk FACTORY Capital Manager 3 to financial institutions.

In May-08 UnRisk FACTORY was released. Version 3 is its 7th release since. It supports now all deal types, models and utilities that have beed added to UnRisk 5 - recently released, including but not limited to inflation linked products, all-in-one pricing functions for callable/putable FX linked products, and new contract and scheduling features. New features on instrumen and scenario grouping. More valuation automation. A complete new online help.

The VaR Universe has been vastly enhanced. It produces now VaR cubes across a vast variety of risk factors for portfolios with all pruduct type under the historic and parametric methodology (Monte Carlo VaR will come soon).

The VaR Univese optional in the UnRisk FACTORY is part of UnRisk FACTORY Capital Manager.

A fast growing number of banks, capital management firms, insurance, .. enjoy the following benefits from using UnRisk FACTORY: Faster Time-to-Insight - High Accuracy - Complete Evidence - Wide Access - Low Cost of Ownership

more: Release Announcement

The Age of Hyperspecialization?

2011-09-16T05:02:00.000-07:00

After the Think It. Build It post, the article in HBR, Aug-11 came in my mind: The Age of Hyperspecialization, the authors say that modern technological advances will transform the workplace by dividing work into smaller tasks performed by ever more specialized workers. And, they summarize, this will create quality, speed, and cost advantages for employers.
It is true that labor becomes more knowledge based and communication and computer support advances. But does this mean that the division of labor accelerates?

Yes, we use an integrated project management environment and workbench enabling distributed work, but we also rely on the creativity and innovativeness of teams recognizing and and discussing patterns and create new things together.

Think It. Build It.

2011-09-15T05:32:00.000-07:00

I often ask myself, do quants need to know all traps and dirty tricks solving multi-factor PDEs numerically or with asymptotic math approaches, solving inverse problems, like model calibration, exploiting, say, CUDA over the grid optimally?
Isn't their strength: financial engineering, structuring, model validation, risk and product controlling methodologies? Isn't a quant's challenge to provide rationales for decisions and accelerate time to insight?

Quants in their high-level work enjoy UnRisk-Q's bank-proof pricing and calibration engines, computing blazingly fast programmatically manipulated from a declarative financial and mathematical programming environment empowering quants to build what they think in a style they would describe the problem.

See: Documentation Center

Click on deal types, VaR, parallelization, ... and look for examples.

Distinguishing Between Complicated and Complex Matters

2011-09-13T08:53:00.000-07:00

My skeleton is complicated, I am complex. A convertible with reset, call / put features and soft calls is complicated, how to trade it is complex.
Complicated is when something contains many intricately combined building blocks, hard to figure out how it works.
Complex is something that acts as a system and it has properties and behaves in a way that are not obvious. We can understand how it works, but usually not predict what it does.

How complicated a thing might be, we can hope that we find a hard science approach, a scientific method to model its properties and behavior. This is what we do with derivatives and portfolio analytics. We apply computational finance emphasizing on blazingly fast model solvers, calibration engines in UnRisk.

Complexity happens on the small border between, kind of solid and liquid "aggregate" phases, or in other words order and chaos. Complex behavior often emerges from simple rules and programs (as fundamentally formulated in Stephen Wolfram's NKS)

One important segment in complex systems are those which are universal in the sense of programmability (like register machines - computers). They in general are "solid" enough to keep data persistent and fluid enough to transform signals.

Thinking about complexity in general, complexity economics, and econophysics approaches, like agent based models, for studying pricing formation and behavior of underlyers, I questioned myself whether our money system is programmable?
It is solid enough to store values and liquid enough to be the media for economic transactions.

UnRisk FACTORY or UnRisk FACTORY?

2011-09-09T04:49:00.000-07:00

Having built radical innovation from bottom up, we are able to swiftly enter into new directions and service longtime partners and friends with special attention.

Integrating our blazingly fast valuation and risk engines into transaction processing and web service environments to a portfolio across scenario simulation factory, we have created the UnRisk FACTORY. It is used in banks, private banks, capital management firms, insurance and financial services institutions.

Capital managers were UnRisk customers from the first hour in 2002.
Smaller capital management firms, buying UnRisk FACTORY in its compact EXPRESS Edition form recently drove some special functionality for better investment and risk management as well as regulatory compliance.

Today we released UnRisk FACTORY Capital Manager bundling UnRisk FACTORY with the VaR Universe and adding functionality devoted to capital management's risk benchmarking requirements.
This is the first step into a product line, which will again serve capital management needs first but drive generic technologies.

EMH <==> P=NP

2011-09-01T05:23:00.000-07:00

I recently found this: Phil Maymin at TEDxNSIT

Markets are efficient (current prices fully reflect all information available in past prices) if and only if every decision problem thats solution can be efficiently verified (NP) by a computer, can also be efficiently solved by a computer (P). In polynomial time.

In general the P versus NP complexity problem in computer science is unsolved.

The Paper. Aas expected, "If EMH then P=NP" is quite intuitive, to prove the converse Philip Maymin shows how markets can be "programmed" to to solve NP-complete problems.

So I ask myself, when the market still played the Black Scholes game and if the valuation of, say, a certain class of options was NP-complete, then that market segment was still efficient?

However, it serves the intuition that markets become increasingly inefficient as the time series lengthen or become more frequent. In an inefficient market there exist exploitable profit opportunity.

The fact that market efficiency and computational efficiency are linked suggests that it was beneficial if fast algorithms for the valuation of instruments, calibration of models, .. were widely used. They are only achievable by clever approximation.

But this is not enough, patterns in data need to be evaluated efficiently to decide the best strategies ...

In short, speed matters.

Shhh, .... UnRisk FACTORY 3 is Coming

2011-08-29T07:43:00.000-07:00

In about a month, we will take UnRisk FACTORY version 3 to financial institutions.
It will update all new deal types, like inflation linked products, generalized, all-in-one pricing functions for callable/putable FX linked products, .... contract and scheduling features, the VaR Universe, from UnRisk-Q version 5, and add new features on instrument and scenario grouping and valuation automation and a complete online help.

In total I have counted over 60 new features,.

It will come as UnRisk FACTORY for an unlimited number of over the web users and the UnRisk FACTORY EXPRESS Edition serving smaller groups of capital, asset and risk management professionals.

UnRisk or UnRisk?

2011-08-01T08:09:00.000-07:00

Serving practitioners from front offices to product and risk controlling in 10 years, we have developed a sophisticated set of engines and tools and tied them together to a valuation and data management factory the UnRisk FACTORY. It offers one key benefit to its customers: automation of portfolio across scenario analysis - by pushing buttons or running scheduled tasks.
System variables are user roles and usage definitions, deal type specifications and valuation and risk management procedures.

Having built radical innovation from bottom up, we were able to unleash them - fusing high-end numerical schemes with massive data and valuation management - as UnRisk-Q. Its declarative, domain specific language and link technologies allow for building individual quant finance systems, in the scope of the UnRisk engines. In the extreme another "UnRisk Factory" and beyond. Practically everything is variable, except the model-method combinations that have been proven as the most sophisticated in the bank practice.