Personal tools
User menu

Events/Seminars/2009

From Thalesians

Jump to: navigation, search

Contents

Thalesian Seminar (London) — Numerical Algorithms Group — Monte Carlo Simulation and its Efficient Implementation

Numerical Algorithms Group

Date and Time

7:30 p.m. on Tuesday, 15th December, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11994542/

Abstract

Monte Carlo simulation is one of the most important numerical methods in financial derivative pricing and risk management. Due to the increasing sophistication of exotic derivative models, Monte Carlo becomes the method of choice for numerical implementations because of its flexibility in high-dimensional problems. However, the method of discretization of the underlying stochastic differential equation (SDE) has a significant effect on convergence. In addition the choice of computing platform and the exploitation of parallelism offer further efficiency gains. We consider here the effect of higher order discretization methods together with the possibilities opened up by the advent of programmable graphics processing units (GPUs) on the overall performance of Monte Carlo and quasi-Monte Carlo methods.

Speaker

Robert Tong received a PhD from the University of Bristol (UK) in the area of applied mathematics and scientific computing, following a first degree in mathematics. His research project was prompted by the need to improve the modeling and design of coastal structures after a major failure.

He followed this with the post of Research Fellow in Applied Mathematics at the University of Birmingham, where his focus was on the role of numerical software and mathematical modeling in the context of the failure of structures due to extreme events.

His interest in numerical software led naturally to his joining NAG to work on the development of their libraries. While there he has worked on data approximation methods, including wavelets and radial basis functions, in addition to applications in finance.

Kai Zhang is pursuing a PhD in mathematical finance with Dr. Nick Webber from Warwick Business School. His research focuses on novel derivative pricing methods and efficient model implementations.

He joined NAG as an intern at the end the second year of his study. His duties are supplying Monte Carlo simulation components, developing example financial applications, writing technical reports for clients and collaborating with leading researchers and practitioners to incorporate new methods in the library.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Resources


Thalesian Seminar (London) — Prof. Svetlozar (Zari) T. Rachev — Market Crashes and Modeling Volatile Markets

Svetlozar Rachev

Date and Time

7:30 p.m. on Wednesday, 2nd December, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11622834/

Abstract

We discuss:

  • econometric models for volatile markets in large dimension, exhibiting volatility clustering and heavy-tails — ARMA-GARCH models with stable and tempered stable innovations;
  • copula dependencies and factor models;
  • forecasting market downturns;
  • tail risk and performance measures;
  • large scale portfolio optimization.

This is a joint work with my co-authors Aaron Kim, Stoyan Stoyanov, Boryana Racheva-Iotova, Michele-Leonardo Bianchi and Frank Fabozzi.

Speaker

Zari Rachev is a co-founder and President of BRAVO Risk Management Group — originator of the Cognity methodology, which was acquired by FinAnalytica where he serves as Chief Scientist. Rachev holds Chair-Professorship in Statistics, Econometrics and Mathematical Finance at University of Karlsruhe, and is the author of 12 books and over 300 published articles on finance, econometrics, statistics and actuarial science. At University of California at Santa Barbara, he founded the Ph.D.program in mathematical and empirical finance. Rachev holds PhD (1979) and Doctor of Science (1986) degrees from Moscow University and Russian Academy of Sciences. Rachev's scientific work lies at the core of Cognity's newer and more accurate methodologies in risk management and portfolio analysis.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Rachev_20091202.ppt

Resources


Thalesian Seminar (London) — Dr. Peter P. Carr — Local Variance Gamma

Peter Carr

Date and Time

7:30 p.m. on Monday, 30th November, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11893766/

Abstract

Suppose that one can observe European option prices at discrete strikes and at one or more maturities. Here we assume that the risk-neutral process for the underlying futures price is a pure jump Markov martingale operating in continuous time with a continuous state space. When the given option prices are arbitrage free and on a regular grid, we show how to construct a piecewise time and space-homogeneous process that meets them. Our construction leads to partial differential difference equations (PDDE's), which permit both explicit calibration and fast numerical valuation. The construction marries ideas from local volatility and variance gamma and is christened the local variance gamma model.

Speaker

Peter Carr is a Managing Director at Morgan Stanley with over 18 years of experience in the derivatives industry. He was also a finance professor for 8 years at Cornell University, after obtaining his PhD from UCLA in 1989. He is presently the Executive Director of the Math Finance program at NYU’s Courant Institute, the Treasurer of the Bachelier Finance Society, and a trustee for the Museum of Mathematics in New York. He has over 80 publications in academic and industry-oriented journals and serves as an associate editor for 8 journals related to mathematical finance. He was selected as Quant of the Year by Risk Magazine in 2003 and as Financial Engineer of the Year by IAFE/Sungard in 2010. More recently, Institutional Investor has included Dr. Carr in its annual Tech 50 for the last 3 years.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Carr_20091130.pdf

Resources


Thalesian Seminar (London) — Steve Zymler — Worst-Case Value-at-Risk of Derivative Portfolios: A Cure for Black Swans?

Steve Zymler

Date and Time

7:30 p.m. on Tuesday, 24th November, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11780122/

Abstract

Despite its popularity, VaR lacks some desirable theoretical properties.

Firstly, VaR is known to be a non-convex risk measure. As a result, VaR optimization problems usually are computationally intractable. In fact, they belong to the class of chance-constrained stochastic programs, which are notoriously difficult to solve. Secondly, VaR fails to satisfy the subadditivity property of coherent risk measures. Thus, the VaR of a portfolio can exceed the weighted sum of the VaRs of its constituents. In other words, VaR may penalize diversification. Finally, the computation of VaR requires precise knowledge of the joint probability distribution of the asset returns, which is rarely available in practice.

Usually the investor has some distributional information about the asset returns, such as the first- and second-order moments. A common mistake is then to assume that the asset returns are normally distributed. This "Normal-VaR" approximation is appealing as it is a convex function (at least for most threshold values) and is therefore amenable to convex optimization techniques. However, the normality assumption can be extremely dangerous as it rarely holds in reality. When the true distribution is non Gaussian the implications can be disastrous for Normal-VaR optimized portfolios since the risk may be severely underestimated (Black Swans).

Recently, we developed new Worst-Case VaR models for derivative portfolios. These models assume that only certain elements of information are known about the distribution of the underlying asset returns (such as means, covariance matrix, support, etc) but are otherwise unknown. We take into account the set of all distributions that match the known elements of information about the distribution (even those with high skewness, kurtosis, etc). The Worst-Case VaR is then computed as the VaR under the worst-case probability distribution in this set. This method essentially immunizes us against distributional model risk by optimizing the portfolio over the (possibly infinite) set of matching distributions. As more information is provided about the distribution, Worst-Case VaR becomes less conservative.

It turns out that these worst-case portfolio optimization models can be solved using extremely efficient convex optimization algorithms. Worst-Case VaR is a convex risk measure and is therefore also subadditive. Our numerical out-of-sample Worst-Case VaR optimization backtests are very encouraging.

Speaker

Steve Zymler is a PhD student at the Optimization Group of Imperial College London. His research focusses on decision making under uncertainty. In particular, he investigates novel robust risk models which do not make strong assumptions about the probability distributions of the underlying random variables. Steve is also actively involved in developing software solutions that are able to compute as well as minimize the risk of large scale derivative portfolios. These software packages can be used for automated trading and risk management.

Prior to commencing his PhD, Steve graduated from Imperial College with a first class honours master's degree in Computer Science. Steve has received numerous awards, such as the Microsoft Prize for best master thesis (2006), and was voted amongst the top three students at the British SET Awards (Science, Engineering & Technology Student of the Year 2006). He was also a finalist at British Computing Society (BCS) national programming competition (2003) and the Flemish Mathematics Olympiad (2002).

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Zymler_20091124.pdf


Thalesian Seminar (London) — Prof. Kevin Parrott — PDE Methods for Option Pricing under Jump Diffusion Processes

Kevin Parrott

Date and Time

7:30 p.m. on Wednesday, 18th November, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11778739/

Abstract

The PDE approach to Option Pricing under jump-diffusion processes (useful for American style contracts) leads to the need to solve numerically a partial integro-differential equation. Merton’s jump-diffusion model is not too hard but probability distributions such as Variance-Gamma and CGMY produce singularities within the integrals and require more care. A number of approaches can be used and this talk describes in detail the use of finite differences with a coordinate transformation. These transformations permit a highly refined mesh in localised regions of interest and, since they have a uniform mesh in the transformed coordinate, allow the use of Richardson extrapolation. Continuously sampled Asian options have a nice structure with this approach and can be solved accurately using a semi-Lagrange time integration approach that should speed up efficiently on parallel hardware.

Speaker

Kevin Parrott is a Professor at the School of Computing and Mathematical Sciences in the University of Greenwich. He is also a visiting lecturer on the Part-time Postgraduate Diploma in Mathematical Finance at Oxford (http://www.maths.ox.ac.uk/prospective-students/graduate/courses/finance/part-time). Kevin has served as a research follow and UCINA coordinator at the University of Oxford for 13 years and has a PhD in computational astrophysics from Manchester University. He is a member of the Society for Industrial mathematics and a regular speaker at academic computational finance events.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Parrott_20091118.pdf

Resources


Thalesian Seminar (London) — Saeed Amen — The intraday impact of economic data releases on the Australian dollar

Saeed Amen

Date and Time

7:30 p.m. on Wednesday, 21st October, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11478789/

Abstract

In this presentation we quantify the impact of both US and Australian economic data releases on the Australian dollar in terms of both intraday volatility and direction. We create an ordering of importance of economic data releases based on intraday volatility. We also investigate the relationship between economic data surprises and spot displacement.

Speaker

Saeed is the founder of Cuemacro and is a co-founder of the Thalesians. Over the past decade, Saeed Amen has developed systematic trading strategies at major investment banks including Lehman Brothers and Nomura. Independently, he is also a systematic FX trader, running a proprietary trading book trading liquid G10 FX, which has had a Sharpe ratio over 1.5 since 2013. He is also the author of Trading Thalesians: What the ancient world can teach us about trading today (Palgrave Macmillan).

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Amen_20091021.pdfrestricted access; please contact Saeed dot Amen who happens to be at Nomura dot com

Resources

  • Saeed Amen's home page: http://www.saeedamen.com/ — here in addition to finance you will find some of Saeed's excellent photos (he is also a photographer!)

Thalesian Seminar (London) — Dr. Aly Kassam — Implementing High Frequency Trading Algorithms Ten Times Faster

Aly Kassam

Date and Time

7:30 p.m. on Wednesday, 19th August, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/11056717/

Abstract

This presentation will discuss some of my investigations into the use of a set of visual programming tools to develop and deploy high frequency program trading systems. The tools are well developed and used in other engineering disciplines, but as yet relatively new for finance. In particular we will look at a new design paradigm — Model Based Trading — together with automatic HDL and C-code generation from models and deployment to specific hardware including DSP boards.

Speaker

Aly Kassam has had a long working relationship with The MathWorks Ltd — the creators of MATLAB — where he is currently a Principal Application Engineer for Computational Finance. Prior to this he worked for a couple of years as a program trader at Barclays Capital in London. He has a degree in Astronomy and Physics along with a Masters in Nonlinear Dynamics and Chaos from UCL in London, and a Doctorate in Numerical Methods for PDEs from the Computing Laboratory at Oxford University.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

To be published here.

Resources


Thalesian Seminar (London) — Dr. Dan Crisan — Solving Backward Stochastic Differential Equations using Cubature Methods

Dan Crisan

Date and Time

7:30 p.m. on Wednesday, 22nd July, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10642489/

Abstract

In the last decade, a new class of numerical methods for approximating distributions of solutions of SDEs have been introduced by Kusuoka, Lyons, Ninomiya and Victoir. These methods are based on the work of Kusuoka and Stroock who established refined gradient upper bounds for the associated semigroup using Malliavin Calculus techiniques. In this talk, I will present an application of these methods to the numerical solution of Backward SDEs together with applications to option pricing. The talk is based on joint work with S. Ghazali and K. Manolarakis.

Speaker

Dan Crisan is a Professor in Mathematics at Imperial College London. His expertise lies in the area of Stochastic Analysis with applications in Engineering and Finance. His current research is on developing high-order numerical algorithms for solving stochastic differential equations, approximating schemes for backward SDEs and particle methods for nonlinear filtering. His book, Fundamentals of Stochastic Filtering appeared at Springer Verlag at the begining of the year and he is currently involved in editing an advanced handbook on Nonlinear Filtering to be published by the Oxford University Press. Dr. Crisan is a member of the editorial board of the Journal of Mathematics and Computation. He is also actively involved in teaching. Among numerous other courses, he has taught stochastic filtering, numerical stochastics, and measure-valued processes at Imperial College; applied probability, and stochastic calculus and applications at Cambridge University.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Crisan_20090722.pdf

Resources


Thalesian Seminar (London) — Dr. Rene Reinbacher — Markovian Projection, Heston Model and Pricing European Basket Options with Smile

Rene Reinbacher

Date and Time

7:30 p.m. on Tuesday, 7th July, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10731842/

Abstract

Following an idea of Piterbarg, we explain how Markovian projection can be used to derive an effective pricing model with smile for European options on a basket. The basket will contain n currencies, each admitting smile and calibrated to the Heston model. This implies that the process for the basket has 2n underlying Brownian motions. Markovian projection generates an effective model driven only by two Brownian motions (a shifted Heston model) representing the basket.

Along the course we give some introduction to the Heston model and develop a calibration method for the correlation matrix between the Heston models of various currencies.

Speaker

Rene Reinbacher holds a Masters degree in Mathematics and a PhD in Physics from the University of Pennsylvania. He was a Postdoctoral Researcher at Rutgers University and Harvard University.

Rene's academic interests include string theory and algebraic geometry. In particular, he was working on the application of algebraic-geometric methods in string theory to study the existence and properties of solutions to nonlinear PDE’s on compact spaces via Kobayashi correspondence, and the numerical methods for explicit construction of these solutions. He used recent theoretical results and numerical methods in C++ to obtain the first known solutions of Ricci flat metrics and Hermitian Yang-Mills equations on compact Calabi-Yau threefolds.

Rene has worked at Lehman Brothers and Barclays Capital as an FX and DCRM front office quant. He is currently working for Barclays Capital.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Reinbacher_20090707.pdf

Resources


Thalesian Seminar (London) — Dr. Adrian Zymolka — Constraint Attribution: Mastering Constraints for Better Portfolio Construction

Adrian Zymolka

Date and Time

7:30 p.m. on Wednesday, 17th June, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10513811/

Abstract

Using constraints in portfolio construction causes deviations from the classical mean-variance portfolio, which are referred to as unrealized alpha, opportunity costs, or implementation inefficiency. Various measures like the transfer coefficient attempt to quantify such effects.

Axioma developed a novel methodology — called Constraint Attribution — to consistently allocate these deviations to individual constraints. Besides identifying the largest obstacles for increasing implementation efficiency, such attribution gives portfolio managers deeper insights into and thus greater control over their portfolio construction process. This talk introduces the methodology and discusses results for exemplary strategies.

Speaker

Dr. Adrian Zymolka has a decade of experience in working with clients of Axioma’s sophisticated products in the financial industry. With his mathematical and optimization background, he is passionate about making sure that any users have maximum benefit from the portfolio construction, risk management and performance attribution software tools as well as content products like risk models and factor libraries.

From 2015 on, Adrian heads Axioma's new Frankfurt office servicing German-speaking Europe. In 2010, he moved to New York to lead the Client Services team in the US. Adrian joined Axioma in 2007 in the London office as Director of Client Services Europe, offering training, consulting services and mathematical support to users.

Prior to Axioma, Adrian was a research assistant at the Zuse Institute Berlin (ZIB) in the Optimization department headed by Prof. Martin Grötschel. During his Ph.D. time, he developed optimization methods for highly complex problems in the area of telecommunication network design. Besides leading and participating in various industrial projects, he also carried out a strong research record. In 2006, he joined atesio, a ZIB spin-off company, where he worked as optimization developer and consultant.

Adrian holds a Ph.D. in Mathematics (Optimization) from the Technical University in Berlin and a Master in Mathematics from Philipps University in Marburg, Germany.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Zymolka_20090617.pdf

Resources


Thalesian Seminar (London) — Dr. Patrick Burns — Using Random Portfolios with R

Patrick Burns

Date and Time

7:30 p.m. on Wednesday, 3rd June, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10199379/

Abstract

Random portfolios sample from the set of portfolios that obey a given set of constraints. R is a language that was designed for data analysis and graphics. This talk will discuss some applications of random portfolios, including performance measurement and testing trading strategies. Then it will provide an introduction to R for those who are unfamiliar with it. Finally, a demonstration of some operations in R on random portfolios will be given.

Speaker

In 2002 Patrick Burns founded Burns Statistics, which focuses on consulting and software for asset management. Prior to that he spent 4 years at Citigroup in London in the Equity Research and Equity departments where he worked on quantitative models for trading and risk measurement. Before entering finance Patrick was a lead developer of S-PLUS in its early days. He has a PhD in Statistics from the University of Washington in Seattle.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Burns_20090603.pdf

Resources


Thalesian Seminar (London) — Dr. David Bellot — An Introduction to Probabilistic Decision Support Systems for Automatic Trading and Financial Series Analysis

David Bellot

Date and Time

7:30 p.m. on Wednesday, 20th May, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10259088/

Abstract

A recent evolution in modeling, reasoning and decision-making under uncertainty, has been achieved through the use of graphical models and decision graphs. As a result of the convergence of artificial intelligence and statistics, they ease the natural specification of problem domains, and bring a wealth of new algorithms for automatic construction of models, efficient computations of posteriors and query answering. They are mainly used for the purposes of analysis, simulation, prediction and diagnosis in a variety of fields.

Speaker

David Bellot is interested in Statistical Machine Learning, especially in Probabilistic Graphical Models and Bayesian reasoning with applications on time series analysis, knowledge representation and natural language processing. He applies his background on various decision-support problems in radar tracking, medical monitoring and more recently in the analysis of financial assets. He has a BSc and MSc in Computer Science and a PhD in Artificial Intelligence. He has worked as a researcher in several organizations like INRIA, France and the University of California, Berkeley. Currently David is focused on financial applications of probabilistic decision-support systems.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Bellot_20090520.pdf

Resources


Thalesian Seminar (London) — Gernot Ziegler and Thomas Bradley — CUDA — GPU Computing for Financial Applications

Gernot Ziegler

QUESTIONNAIRE

Our speakers would appreciate if you could submit the following (optional) questionnaire with your registration for this event:

  • Position/job title:
  • Brief description of your role: (e.g. exposure to quantitative models, asset classes, front office/middle office)
  • Institution:
  • Number of years experience in the finance industry:
  • What types of computationally intensive processes for financial applications do you run and/or depend on? (e.g. Monte Carlo simulations for derivative pricing)
  • Have you had exposure to numerical methods and high performance computing?
  • Have you been involved or are you aware of any projects to investigate using CUDA for any of these applications within your institution?
  • What are your reservations, if any, about using CUDA to speed up financial applications and reduce hardware costs?

Date and Time

7:30 p.m. on Tuesday, 28th April, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: http://www.meetup.com/thalesians/calendar/10199345/

Abstract

With the world's first teraflop many-core processor, NVIDIA® Tesla computing solutions enable the necessary transition to energy efficient parallel computing power. With 240 cores per processor and a standard C compiler that simplifies application development, Tesla scales to solve the world's most important computing challenges more quickly and accurately. The C for CUDA programming environment simplifies many-core programming and enhances performance by offloading computationally-intensive activities from the CPU to the GPU. It enables developers to utilise NVIDIA GPUs to resolve the most complex computation-intensive challenges in financial analysis, such as real-time pricing, value-at-risk and complex Monte Carlo simulations.

Speaker

Gernot Ziegler (MSc/civ.ing.) is an Austrian engineer with an MSc degree in Computer Science and Engineering from Linköping University, Sweden. He pursued his PhD studies at the Max-Planck-Institute for Informatics in Saarbrücken, Germany, where he specialized in GPU algorithms for computer vision and data-parallel algorithms for spatial data structures.

Thomas Bradley MEng(Hons) MIEE graduated with a first-class MEng degree in Computer Systems Engineering from the University of Bristol in 2000, having also completed the final year of the Diplôme d’Ingénieur at l’École Nationale Supérieure de Télécommunications in Brest, France. He worked as processor architect for video encoding processors at STMicroelectronics before moving to ClearSpeed Technology plc to lead architecture development for general purpose parallel processors. Since then he has specialised in High Performance Computing software development at ClearSpeed and now at NVIDIA.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_NVIDIA_20090428.pdf

Resources


Thalesian Seminar (London) — Prof. Berç Rustem and Steve Zymler — Robustness in Investment Decisions

Berc Rustem

Date and Time

7:30 p.m. on Wednesday, 25th March, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: N/A

Abstract

Recognizing that robust decision making is vital in risk management, we provide concepts for computing the best decision in view of the worst-case scenario. The main tool used is minimax, which ensures robust policies with guaranteed optimal performance that will improve further if the worst case is not realized.

Critically, worst-case design addresses not only Armageddon-type uncertainty. Indeed, the determination of the worst case becomes nontrivial when faced with numerous — possibly infinite — and reasonably likely rival scenarios. Optimality does not depend on any single scenario but on all the scenarios under consideration. Worst-case optimal decisions provide guaranteed optimal performance for systems operating within the specified scenario range characterising the uncertainty.

The noninferiority of minimax solutions — which also offer the possibility of multiple maxima — ensures this optimality. Worst-case design is not intended to necessarily replace expected value optimization when the underlying uncertainty is stochastic. However, wise decision making requires the justification of policies based on expected value optimization in view of the worst-case scenario. Conversely, the cost of the assured performance provided by robust worst-case decision making needs to be evaluated relative to optimal expected values.

Robustness is thus intended to relax the inherent inaccuracy of forecasts. This applies to single as well as multi-period portfolio decision problems. We begin by introducing a generalisation of the standard Markowitz framework to multiple/rival return and risk scenarios. The optimal investment decision is evaluated in view of the worst-case scenario. The multi-period extension is considered to model non-liquid assets and the effect of transaction costs on the optimality of final performance. Robustness is not restricted to simply the mean-variance framework. Indeed, we also consider optimisation of multi-period conditional value at risk in view of rival representations of the future (in terms of scenario trees).

Speaker

Berç Rustem (FIMA, CMath) (Professor of Computational Methods in Operations Research, Imperial College, London) is editor of Automatica, co-editor of Computational Management Science, Advisory Editor of Journal of Economic Dynamics & Control (JEDC), and on the editorial board of Computational Economics, Journal of Global Optimization, Cybernetics, and several book series. He was President of the Society for Computational Economics (2000-2002), editor of JEDC (1987-2002), chair of the IFAC Technical Committee "Computation in Economics & Finance" (1992-2000), and chair of the IPC IFAC Symposium "Computation in Economics & Finance" (2001). He also was the Principal Investigator in various research projects that led to the development of optimisation software for nonlinear economic models (supplied to HM Treasury), and financial and engineering risk management software. He has authored over 150 journal and conference publications, edited many journal special issues (including JEDC, Parallel Computing), and 4 book volumes (published by MacMillan and Kluwer). He is the author of 3 research monographs on optimisation algorithms, multiple-objective decision making and min-max robust design (Springer-Verlag, Wiley & Sons, Princeton University Press). He is the author of "Projection Methods in Contrained Optimisation and Applications to Optimal Policy Decisions", "Algorithms for Nonlinear Programming", and "Algorithms for Worst-Case Design and Applications to Risk Management".

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Rustem_20090325.pdf

Resources


Thalesian Seminar (London) — Prof. Claudio Albanese — Interest Rate Derivatives and GPU Computing

Claudio Albanese

Date and Time

7:30 p.m. on Wednesday, 4th March, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: N/A

Abstract

GPU coprocessors give access to extraordinary computational power as long as one can design financial algorithms to take advantage of the hardware features. The talk elaborates on the challenges and solutions in the context of interest rate exotics. I outline algorithms for calibration, pricing and risk management and discuss the implementation of lattice and Monte Carlo methods.

Speaker

Claudio Albanese is a Visiting Professor at the Financial Mathematics Group at King's College and an independent consultant at Global Valuation Ltd. He received his doctorate in Physics from ETH Zurich, following which he held post-doctoral positions at New York University and Princeton University. He was Associate Professor in the Mathematics Department of the University of Toronto and then Professor of Mathematical Finance at Imperial College London.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Albanese_20090304.pdf

Resources


Thalesian Seminar (London) — Saeed Amen — Introduction to Foreign Exchange

Saeed Amen

Date and Time

7:30 p.m. on Wednesday, 11th February, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: N/A

Abstract

In this presentation we shall give a short introduction to the FX markets, explaining various quoting conventions and how FX is traded. We shall mostly examine FX spot and options markets, and look at which factors in general influence FX markets.

Speaker

Saeed is the founder of Cuemacro and is a co-founder of the Thalesians. Over the past decade, Saeed Amen has developed systematic trading strategies at major investment banks including Lehman Brothers and Nomura. Independently, he is also a systematic FX trader, running a proprietary trading book trading liquid G10 FX, which has had a Sharpe ratio over 1.5 since 2013. He is also the author of Trading Thalesians: What the ancient world can teach us about trading today (Palgrave Macmillan).

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Amen_20090212.ppt

Resources


Thalesian Seminar (London) — Dr. Matthew Dixon — Calibrating Spread Options using a Seasonal Commodity Forward Model

Matthew Dixon

Date and Time

7:30 p.m. on Thursday, 29th January, 2009.

Venue

Upstairs at City Pride, Canary Wharf, London, UK.

Meetup.com

You can register for this event and pay online on Meetup.com: N/A

Abstract

We describe the calibration of spread options on power and natural gas forward prices whose dynamics are described by the Borovkova & Geman (2006) commodity forward price model. This two-factor model resolves the stochastic dynamics of the average value of the forward curve and the term structure of convenience yields, the latter of which incorporates seasonal effects. Capturing the seasonal volatility term-structure from historical prices is essential for accurate forward curve construction and subsequent option pricing. An attractive feature of the seasonal forward model is that the covariance of the forward dynamics is quadratic in the volatility term-structure and can be effectively calibrated to historical data using a non-linear gradient based constrained optimization algorithm. We show the comparative effect on calender spread and heating rate option prices through resolving the full term-structure of seasonal forward volatilities.

The talk will end with a more general and interactive discussion on how term-structure calibration methodology can be modified to account for major market dislocations.

Speaker

Matthew Dixon is a Managing Director and Head of Americas at Thalesians Ltd.

He is also an Assistant Professor of Finance in the Stuart Business School at the Illinois Institute of Technology. His research focuses on the application of advanced computational techniques to financial modeling and data analysis especially where high performance and scalability are critical for practical application. Matthew's research is currently funded by Intel Corporation. He has contributed to the R package repository and published around twenty peer-reviewed technical articles. He has taught financial econometrics, derivatives, machine learning and text mining at the University of San Francisco and held visiting appointments in CS/Math at Stanford University and UC Davis.

Prior to joining academia, he has held industry appointments as a quant at banks such as Lehman Brothers, the Bank for International Settlements and Barclays Capital. He chairs the workshop on computational finance at the annual SuperComputing conference and serves on the program committee of HPC and on the editorial board of the Journal of Financial Innovation. Matthew holds a MEng in Civil Engineering from Imperial College London, a MSc in Parallel and Scientific Computation (with distinction) from the University of Reading, and a PhD in Applied Math from Imperial College London. He became a chartered financial risk manager in 2014.

Video

Click below for Flash streaming video. Alternatively download QuickTime video.

Get the Flash Player to see this player.

Slides

Thalesians_Dixon_20090129.pdf

Resources

  • This page was last modified on 25 April 2015, at 19:07.
  • This page has been accessed 8,023 times.