Agenda
Agenda
Quant Summit Europe 2023
08:00 – 08:45
Registration and refreshments
08:45 – 08:50
Risk.net welcome
Mauro Cesa is quantitative finance editor for Risk.net, based in London. He leads the team responsible for the publication of quantitative research across all brands of the division. The section of Risk.net he manages, Cutting Edge, publishes peer-reviewed papers on derivatives, asset and risk management, and commodities.
Mauro holds a degree in economics from the university of Trieste and a masters in quant finance from the University of Brescia.
08:50 – 09:00
Chairperson's welcome
Lorenzo Ravagli is an Executive Director in the Quantitative and Derivatives Strategy team, Head of European FX vol strategy, focusing on macro ideas and systematic strategies in the volatility space. Lorenzo joined J.P. Morgan in May 2018 after working as a senior Quantitative Strategist at Société Générale since 2008. He graduated from Florence University with a PhD in Theoretical Physics. He has published in top journals on physics and quantitative finance.
Risk.net Quant of the year "hall of fame"
Keynote presentations from previous winners of the Risk.net QOTY award
09:00 – 09:30
Alternatives to Deep Neural Networks in finance
- Deep Neural Networks have been widely suggested as THE way to approximate functions in math finance
- We argue that much more efficient, and much more predictable and explainable, methods can be developed for the domain of math finance given typical constraints
- We develop two high-performance methods for approximating functions, and for conditional expected value calculations (i.e., non-linear regression problem): the generalized stochastic sampling (gSS) and the functional tensor train (fTT) methods
- We briefly outline the gSS method and discuss the fTT method in some detail
Vladimir Piterbarg
Managing director, head of quantitative analytics and quantitative development
NatWest Markets
Vladimir Piterbarg is the global head of Quantitative Analytics at NatWest Markets since 2018. He held similar positions at Rokos Capital Management LLP, Barclays Capital/Barclays investment bank, and Bank of America. Vladimir Piterbarg has a PhD in Mathematics (Stochastic Calculus) from the University of Southern California. He serves as an associate editor of the Journal of Computational Finance and the Journal of Investment Strategies. Together with Leif Andersen, Vladimir Piterbarg wrote the authoritative, three-volume set of books “Interest Rate Modelling”. He published multiple papers in various areas of quantitative finance, and won Risk Magazine’s Quant of the Year award twice.
09:30 – 10:00
Customising large language models for quant finance applications
Due to the highly specialized subject matter, the leading commercial (GPT-4) and open source (Llama 2, Code Llama) LLMs are unable to provide specialized comprehension and generation suitable for quant finance applications out of the box.
In this presentation, I will describe fine-tuning and prompt engineering techniques that convert stock LLMs into specialized tools that can assist with the following model governance functions, subject to final sign-off by human analysts:
1) Validation of trade capture using LLM comprehension of trade confirmations
2) Validation of model documentation using LLM comprehension of source code
3) LLM generation of model documentation and release note drafts
Alexander Sokol is the founder, Executive Chairman, and Head of Quant Research at CompatibL, a trading and risk technology company. He is also the co-founder of Numerix, where he served as CTO from 1996 to 2003, and the co-founder of Duality Group, where he served as CTO from 2017 to 2020.
Alexander won the Quant of the Year Award in 2018 together with Leif Andersen and Michael Pykhtin, for their joint work revealing the true scale of the settlement gap risk that remains even in the presence of initial margin. Alexander’s other notable research contributions include systemic wrong-way risk (with Michael Pykhtin, Risk Magazine), joint measure models, and the local price of risk (with John Hull and Alan White, Risk Magazine), and mean reversion skew (Risk Books, 2014).
Alexander earned his BA from the Moscow Institute of Physics and Technology at the age of 18, and a PhD from the L. D. Landau Institute for Theoretical Physics at the age of 22. He was the winner of the USSR Academy of Sciences Medal for Best Student Research of the Year in 1988.
10:00 – 10:30
The probability distribution classification problem
- The probability distribution classification is one of the most challenging problems of modern quantitative finance.
- The problem can be formulated as follows: let us assume that we have two sets of samples (either ordered or not), in the most general case of unequal size, drawn from the unknown multivariate probability distributions.
- Can we say with the desired degree of confidence whether these samples were drawn from the same probability distribution or not?
- This problem has many direct applications to the practical use cases such as detection of structural breaks, graduation testing, monitoring of alpha decay, time series analysis, etc.
Alexei is Quantitative Research & Development Lead at ADIA.
Formerly a Managing Director and Global Head of Data Analytics at Standard Chartered Bank, Alexei is responsible for providing data analytics services to Corporate, Commercial and Institutional Banking division of Standard Chartered Bank.
He joined Standard Chartered Bank in 2010 from Barclays Capital where he managed a model development team within Credit Risk Analytics. Prior to joining Barclays Capital in 2004, he was a senior quantitative analyst at Dresdner Bank in Frankfurt.
Alexei holds MSc in Theoretical Nuclear Physics from the University of Kiev and PhD in Mathematical Physics from the Institute for Mathematics, National Academy of Sciences of Ukraine.
He was the recipient of the 2019 Quant of the Year award from Risk magazine.
10:30 – 11:00
Statistics and calibration for rough volatility: misconceptions and optimal procedures
Rough volatility models have gained large interest in the financial engineering community in the recent years. The goal of this talk is to provide an accurate statistical analysis of such models, with minimax speeds of convergence, optimal procedures and central limit theorems. This enables us to study financial data properly in the rough volatility paradigm, with a rigorous statistician's perspective. This is joint work with Carsten Chong, Marc Hoffmann, Yanghui Li and Gregoire Szymanski.
Mathieu Rosenbaum is a full professor at École Polytechnique, where he holds the chair “Analytics and Models for Regulation”
and is co-head of the quantitative finance (El Karoui) master program. His research mainly focuses on statistical finance problems, regulatory issues and risk management of derivatives.
He published more than 80 articles on these subjects and supervised about 20 PhD students.
He is notably a renowned experts on the quantitative analysis of market microstructure and high frequency trading.
Mathieu Rosenbaum is also at the origin (with Jim Gatheral and Thibault Jaisson) of the development of rough volatility models.
He is one of the editors in chief of the journal “Market Microstructure and Liquidity“ and is associate editor for 10 other journals.
Furthermore, he received the Europlace Award for Best Young Researcher in Finance in 2014, the European Research Council Grant in 2016, the Louis Bachelier prize in 2020 and the Quant of the Year award in 2021.
11:00 – 11:30
Morning networking break
11:30 – 17:45
Stream
Breakout 1: Pricing, market and liquidity risk and capital optimisation
Stream 1 Chairperson
Signature trading
Presentation
-
We represent a trading strategy as a linear functional applied to the signature of a path (henceforth “Sig-Trading”) to encode the evolution of past time-series observations into the optimisation problem.
-
We derive a concise formulation of the dynamic mean-variance criterion alongside an explicit solution in our setting, which naturally incorporates a drawdown control in the optimal strategy over a finite time horizon and draw parallels between classical portfolio strategies and Sig-Trading strategies and explain how the latter leads to a pathwise extension of the classical setting via the “Signature Efficient Frontier”.
-
Finally, we give explicit examples when trading under an exogenous signal as well as examples for momentum and pair-trading strategies, demonstrated both on synthetic and market data.
Blanka Horvath is a Lecturer in Financial Mathematics at King’s College London as well as a Honorary Lecturer at Imperial College London and a researcher at The Alan Turing Institute, where she is co-lead of the Machine Learning in Finance theme. Blanka holds a PhD in Financial Mathematics from ETH Zurich, a postgraduate Diplom in pure Mathematics from the University of Bonn and an MSc in Economics from the University of Hong Kong. In her latest research she focusses on non-Markovian models of nancial markets such as Rough Volatility models as well as modern DNN- based market generators. Prior to her position at King’s College, Blanka worked at JP Morgan on the re nements of the Deep Hedging programme and the development of generative market simulation models. Her work on DNN-based calibration of Rough Volatility models was awarded the Rising Star Award 2020 of Risk magazine.
Margin modelling approaches for optimisation and capital efficiencies
Presentation
- Overview of margin methodologies
- The objective of capital optimization and use case
- A case study of the margin optimization
- Future possibilities of capital optimization
Claire Liu is a Director within the Clearing and Post-Trade Services Division at CME Group. She joined the quantitative risk management group at CME Clearing in 2014. She holds a master’s in mathematical finance degree from Illinois Institute of Technology. She is a Chartered Financial Analyst charterholder and a certified Financial Risk Manager. She likes playing the violin during her leisure time.
Bridging the gap risk reloaded: a comprehensive methodology for wrong way risk and leveraged exposures
Presentation
- How dangerous is the interplay between WWR, leverage and collateralised exposures?
- How to generalise MC frameworks in the presence of WWR?
- And what about non-MC based frameworks such as stress testing and SA-CCR?
Fabrizio Anfuso, Senior Technical Specialist, PRA, BANK OF ENGLAND
Fabrizio is a leading expert in developing complex risk analytics, quantitative modelling and financial regulations. He has an extensive track-record of heading quant teams onshore and offshore, as well as of taking part in firm-wide programs, such as IMM, BCBS-IOSCO Margin Requirements and IBOR transition.
In his present and previous roles, Fabrizio has gained a comprehensive knowledge of the full model development cycle, including the model design, the validation of model performance, the IT implementation and the attainment of regulatory compliance.
His main areas of expertise are Counterparty credit risk, Monte Carlo simulations, Internal Models for the trading book (IMM and IMA), derivatives pricing, CCPs & collateral modelling, Initial Margin methodologies and regulatory capital.
Fabrizio is chairing the master’s course in Counterparty Credit Risk of the ETH / University of Zurich and taught a number of advanced professional trainings in topics such as CCR, capital management and Initial Margin methodologies.
As part of his academic activities, he has authored numerous research articles in the fields of quantitative finance and condensed matter physics. Fabrizio holds a Ph.D. in Theoretical Physics from the Chalmers University of Technology (Gothenburg, Sweden).
Networking lunch
Pricing volatility swaps using Machine Learning
Presentation
- Vol swaps, popular instruments without clear pricing formulas
- Deploying machine learning techniques for pricing vol swaps
- Pure data driven pricing explored and tested
Wim Schoutens is a quantitative finance professor at the University of Leuven, Belgium.
He has extensive practical experience of model implementation and validation. He is well known for his consulting work with the banking industry and national and supra-national institutions. He is an independent expert advisor to the European Commission, has worked for the IMF and is the author of several books on quantitative finance.
His latest books, co-authored with Dilip Madan, are about the brand new theory of conic finance.
He is also a member of different editorial Boards of international finance journals. Wim is also a founding partner of RiskConcile, a fintech company with roots within the University of Leuven.
He likes arbitrages, politically incorrect statements and making jam.
Pricing derivatives with a variational auto-encoder approach
Presentation
- Motivation: Generative vs Discriminative ML approaches
- Mathematical Introduction to Variational Auto-Encoders
- Conditional Variational Auto-Encoders
- Numerical Implementation
- Applications
Youssef Elouerkhaoui is a Managing Director and the global head of markets quantitative analysis at Citi. His group supports all modelling and product development activities across businesses. He is also in charge of CVA, Funding and Regulatory Capital for his businesses. Prior to this, he was a Director in the Fixed Income Derivatives Quantitative Research Group at UBS, where he was in charge of model development for the Structured Credit Desk. Before joining UBS, Youssef was a Quantitative Research Analyst at Credit Lyonnais supporting Interest Rates Exotics. He is a graduate of Ecole Centrale Paris and he holds a PhD in Mathematics from Paris-Dauphine University.
Time changes, Fourier transforms and the joint calibration to the S&P500/VIX smiles
Presentation
- We develop a model based on time changed Lévy processes to tackle the 'joint calibration problem', i.e. reproducing the joint S&P500/VIX implied volatility smiles and the VIX futures prices
- The calibration procedure uses efficient Fourier based pricing schemes.
- We focus on one specification of the proposed general setting which uses discontinuous processes only.
- Results show satisfactory performances in solving the joint calibration problem, and therefore confirm that also the class of affine processes can provide a workable fit.
Laura Ballotta is a Professor of Mathematical Finance at the Faculty of Finance of Bayes Business School (formerly Cass). She works in the areas of quantitative finance and risk management. She has written on topics including stochastic modelling for financial valuation and risk management, numerical methods aimed at supporting financial applications, and the interplay between finance and insurance.
Capital valuation adjustments
Presentation
This presentation aims to review the rationale for the Capital Valuation Adjustment (KVA):
- Should derivatives values be adjusted for capital costs? If so in what circumstances and by how much?
- How does KVA fit in with our understanding of “value”?
- How does KVA depend on capital levels and how can we determine the return-on-equity?
- What is the correct KVA for counterparty credit risk?
Matthias Arnsdorf is a Managing Director and the global head of the Counterparty Credit Risk Quantitative Research team at J.P. Morgan. His responsibilities include the development of credit exposure and funding models for valuation, risk management as well as credit risk capital. He has published and presented numerous articles on credit risk and credit derivatives.
Matthias started his career in finance in 2002 working in credit derivatives quantitative research. Prior to this he spent two years as a post-doctoral researcher at the Niels Bohr Institute in Copenhagen. Matthias holds a PhD in Quantum Gravity from Imperial College London.
Afternoon networking break
Survival in the dark
Presentation
- How to estimate and leverage correctly dark venues liquidity in execution algorithm?
- Two challenges: under-estimation caused by partial information of own fills and feedback loop induced by orders depending on the estimations
- We propose an adaptive Bayesian approach inspired by survival analysis combined with a multi-armed bandit solution to address both issues
- With our approach, we compute fill expectations tailored to the order characteristics.
Jaya Nagaradjasarma is the head of EMEA Equities Algo Quant Research for the Equity eTrading line of business at Bank of America London. She and her team create quantitative solutions that leverage data science, AI, and machine learning to optimise equities execution algorithms. Before joining Bank of America, Jaya worked in quantitative roles for automated trading desks in several asset classes on the buy side. Her earlier career includes roles at Susquehanna International Group in Dublin, Man AHL in Oxford, and Antares Technologies in Paris. Jaya received her master’s degree from Ecole Nationale de la Statistique et de l'Administration Economique, her DEA in Probability and Finance from Pierre et Marie University (DEA El Karoui), and her Ph.D. in Statistics from the London School of Economics. Jaya currently resides in London.
Damien Besombes
Vice president, quantitative strategies and data group, EMEA equities execution
Bank of America
Damien Besombes is a Vice President in the Quantitative Strategies and Data Group for EMEA Equities Execution at BofA Securities. Damien has 6 years of experience designing and building execution algorithms. He also previously worked as a Software Developer at the bank and as a Quant Developer in Energy trading at TotalEnergies. Damien graduated from Supélec in France.
Designing generative models for finance: a structured approach
Presentation
- Generative models for market scenario simulation
- The model validation problem for GANs and other black box models
- Designing bespoke learning algorithms for financial applications
- An example : TailGAN
Prof. Rama Cont holds the Chair of Mathematical Finance at Imperial College London and is director of the CFM-Imperial Institute of Quantitative Finance since 2012, after previous appointments at Ecole Polytechnique (France), Columbia University (New York) and Sorbonne (Paris).
His research in finance has focused on modeling of extreme market risks: market discontinuities and breakdowns, liquidity risk, endogenous risk and systemic risk. His 2006 paper on ‘model risk', an early reference on the topic, was the first to propose a quantitative approach to model risk.
Cont has served as a consultant to the BIS, the European Central Bank, the New York Federal Reserve, Norges Bank, the US Commodity Futures Commission (CFTC), the US Office of Financial Research, the IMF and a dozen major CCPs in Europe, Asia, the US and Latin America.
He was awarded the Louis Bachelier Prize by the French Academy of Sciences in 2010 and the Royal Society Award for Excellence in Interdisciplinary Research in 2017 for his research on mathematical modeling in finance.
Explicit option pricing and volatility surface modelling
Presentation
- Fully analytical and empirically sound option pricing formulae exist;
- They are supported by computationally elementary processes;
- Volatility surfaces can be explicitly determined, and calibration performed analytically.
Lorenzo is a Lecturer in Mathematics at the University of Bologna. He was previously appointed at the University of Parma and holds a Post-Doc from LMU Munich. Before joining academia, he worked for the Italian Pension Regulator (COVIP) and took part in the EIOPA-Occupational Pension Committees for the revision of the European Union IORP directive on European Pension Funds.
His research interests are the modelling of the implied volatility surface, applications of jump processes in finance, time changed processes, tempered stable laws, fractional calculus and anomalous diffusions and their applications to finance. He is also interested in economic and mathematical models for public pension systems.
Lorenzo holds an MSc in Mathematics and Finance from Imperial College London and one in Pure Mathematics with a major in Geometry from the University of Roma Tre. His PhD in Mathematics was obtained from UCL London.
Stream
Breakout 2: Data science-driven modelling and computational methods
Stream 2 Chairperson
Stefan Zohren is deputy director of the Oxford-Man Institute of Quantitative Finance and an associate professor at the Department of Engineering Science at the University of Oxford. He is a fellow of the Turing Institute, the UK’s national institute for artificial intelligence (AI) and data science. Zohren’s research is focused on machine learning in finance, including deep learning, reinforcement learning, network and natural language processing approaches, and early use cases of quantum computing. He works with Man Group on commercial research projects and is a frequent speaker on AI in finance, representing the Oxford-Man Institute at academic conferences, industry panels and corporate events.
Validation of internal expected shortfall models: comparing risk predictions and risk realizations
- Ridge backtest: unique prudential backtest for ES
- Unveiling monetary discrepancy between predicted and realized risk
- New paradigm: managing realized risk, not predicted risk
- ES/VaR joint elicitability and related model selection methods
- Potential relevance for FRTB Basel III regulation
Carlo Acerbi is a quantitative financial risk management researcher and professional, author of relevant contributions in the field of banking regulation (2002 coherent definition of ES ; 2019 ES backtesting), asset management liquidity regulation (2013 MSCI LiquidityMetrics) and stress testing (2016 MSCI ST best practices), among others.
He received a PhD in Theoretical Physics (1998, ISAS-SISSA, Trieste, Italy). He served for leading institutions in the financial risk industry (Banca Intesa, Abaxbank, McKinsey, RiskMetrics-MSCI, Banque Pictet). He teaches Advanced Derivatives at Bocconi University, Milan, Advanced Risk Management Topics at EPFL Lausanne (from Q4 2023) and he's Honorary Professor at Corvinus University, Budapest.
Cross-impact in equity markets
Presentation
-
The empirical finding that market movements in stock prices may be correlated with the order flow of other stocks has led to the notion of "cross impact" and has prompted the development of multivariate models of market impact.
-
Empirical evidence suggest that cross-impact terms are unstable and change sign randomly over time, which poses a problem for their interpretation and use.
-
We show that the observed correlations between the returns of an asset and the order flow imbalance (OFI) of other assets have a simpler explanation in terms of common components in order flow across stocks.
-
This commonality in order flow arises naturally from multi-asset trading strategies such as index or ETF portfolios. We provide empirical evidence from order flow and price changes of NASDAQ-100 stocks to support this explanation.
-
Our results show the main determinants of impact to be each stock’s own OFI and the common component of OFI across stocks. Additional ‘cross-impact’ terms account for less than 1% of total impact.
Francesco Capponi, PhD, is a member of BlackRock’s Trading Research team.
Francesco is responsible for trading research in both Equities and FX, with an emphasis on market microstructure, transaction cost analysis, and execution research.
Prior to joining BlackRock, he worked at Barclays as a quantitative researcher within the quantitative portfolio construction team.
He holds an MPhil in Economics from the University of Cambridge, and a PhD in Mathematical Finance from Imperial College London.
The cost of misspecifying price impact
Presentation
- What happens when trades are based on an incorrect price impact model?
- We derive tractable formulas for these misspecification costs and illustrate them on proprietary trading data
- The misspecification costs are naturally asymmetric
- Underestimating impact concavity or impact decay shrinks profits, but overestimating concavity or impact decay can even turn profits into losses
Johannes is the Head of the Mathematical Finance Section at Imperial College London, where he also directs the CFM-Imperial Institute on Quantitative Finance. Before his appointment at Imperial, Johannes held faculty positions at Carnegie Mellon University, the University of Michigan and ETH Zurich. His research focuses on the impact of "market frictions” such as trading costs or asymmetric information on optimal trading strategies.
Networking lunch
Trading the vol-of-vol risk premium
Presentation
- The notion of vol-of-vol risk premium
- Impact of term-structure, smile and fluctuations
- Widespread implications for investors
- Two realistic constructs with plain vanillas
- Variance vs. Vol swap: vol-of-vol sensitivity
- Statistical properties of short vol-of-vol trades
Lorenzo Ravagli is an Executive Director in the Quantitative and Derivatives Strategy team, Head of European FX vol strategy, focusing on macro ideas and systematic strategies in the volatility space. Lorenzo joined J.P. Morgan in May 2018 after working as a senior Quantitative Strategist at Société Générale since 2008. He graduated from Florence University with a PhD in Theoretical Physics. He has published in top journals on physics and quantitative finance.
Darwinian model risk, reverse stress testing, and HVA
Presentation
- Darwinian", adverse selection of models leading to short-and medium-term gains but long term (big) losses
- Darwinian model risk is long-term and directional, hence likely to stay unnoticed from traditional risk systems, which are focused on shorter-term moments of order two and beyond.
- Long-term, large-scale simulations reveal the consequences of using various models in extreme scenarios.
- Darwinian model risk can be managed through HVA (hedging valuation adjustment) reserves
- These reserves need to be risk-adjusted, so not only HVA, but also contributions to the KVA of the bank, which in our case studies happen to be very material and greater than the HVA itself.
- Based on joint works with Claudio Albanese, Cyril Benezet, Dounia Essaket, and Stefano Iabichino).
Stéphane Crépey is Professor of Mathematics at Université Paris Cité, Laboratoire de Probabilités, Statistique et Modélisation (LPSM), Team Mathematical Finance and Numerical Probability. His research interests are counterparty credit risk, XVA analysis, central counterparties, quantitative reverse stress tests; Model risk and uncertainty quantification; Backward stochastic differential equations (reflected, anticipated, in combination with progressive enlargement of filtration,...); Machine learning for finance (learning conditional expectations and risk measures, calibration by neural networks or Gaussian processes, anomaly detection,...). He is the author of numerous research papers published in journals including Annals of Probability, Finance and Stochastics, Mathematical Finance, Stochastic Processes and their Applications or Risk Magazine. He wrote two books: Financial Modeling: A Backward Stochastic Differential Equations Perspective (S. Crépey, Springer Finance Textbook Series, 2013) and Counterparty Risk and Funding, a Tale of Two Puzzles (S. Crépey, T. Bielecki and D. Brigo, Chapman & Hall/CRC Financial Mathematics Series, 2014). He is a member of the scientific council of the French financial markets authority (AMF). Stéphane graduated from ENSAE ParisTech and holds a PhD in differential games and mathematical finance from Ecole Polytechnique and INRIA Sophia Antipolis.
Generative AI for modelling financial markets at the atomic level
Presentation
- We develop a generative model of realistic order flow in financial markets through a first end-to-end model that generates tokenized limit order book (LOB) messages similar to tokenization in large language models (LLM)
- Out-of-sample results show promising performance in approximating the data distribution, as evidenced by low model perplexity and mid-price returns calculated from the generated order flow exhibit a significant correlation with the data, indicating impressive conditional forecast performance
- Due to the granularity of generated data, and the accuracy of the model, it offers new application areas for future work beyond forecasting, e.g. acting as a world model in high-frequency financial reinforcement learning applications.
Stefan Zohren is deputy director of the Oxford-Man Institute of Quantitative Finance and an associate professor at the Department of Engineering Science at the University of Oxford. He is a fellow of the Turing Institute, the UK’s national institute for artificial intelligence (AI) and data science. Zohren’s research is focused on machine learning in finance, including deep learning, reinforcement learning, network and natural language processing approaches, and early use cases of quantum computing. He works with Man Group on commercial research projects and is a frequent speaker on AI in finance, representing the Oxford-Man Institute at academic conferences, industry panels and corporate events.
Generative modelling for time series and applications to deep hedging
Presentation
- Generative model via Schrödinger Bridge
- Financial time series
- Kernel regression
- Synthetic samples
- Deep hedging
Huyên PHAM is a distinguished Professor of Mathematics at Université Paris Cité. He is also Adjunct Professor at ENSAE and was Chair of Applied Mathematics at the John Von Neumann Institute of VNU-HCM. He leads research in quantitative finance, stochastic analysis and control, machine learning, and is the author of more than 110 publications, including the monograph Continuous time Stochastic Control and Optimization with Financial Applications. He serves on the editorial boards of several international journals and is the co-editor in chief of the journal Applied Mathematics and Optimization. Prof. Pham was appointed member of the Institut Universitaire de France in 2006, awarded the Louis Bachelier prize by the French Academy of Sciences in 2007, and was a plenary speaker at the 9th World congress of the Bachelier Finance Society in 2016, and at the 6th Asian Quantitative Finance Conference in 2018.
Mini refreshments break
VolGAN: a generative model for arbitrage-free implied volatility surfaces
Presentation
- VolGAN is a generative model for arbitrage-free implied volatility surfaces
- VolGAN is trained via a customised loss function
- We include arbitrage penalisation via scenario re-weighting
- We demonstrate our approach on SPX implied volatility data and show that VolGAN is able to learn the covariance structure of the co-movements in implied volatilities and generate realistic dynamics for the VIX CBOE volatility index.
Milena Vuletić is a DPhil Student in the Centre for Doctoral Training in Mathematics of Random Systems at the University of Oxford. She is supervised by Prof. Rama Cont and Prof. Mihai Cucuringu. Her research is focused on mathematical and data-driven modelling of multi-asset markets.
Quants and the Qlimate
- Equation based weather and climate models: Navier-Stokes, continuity, energy, humidity
- Challenges and compromises: grid size, calibration, parametrisation, Poincaré
- Ensemble simulations and interpretation of the distributional statistics
- Practical application: reinsurance and catastrophe bonds
Erik Vynckier is board member of Foresters Friendly Society, general partner of InsurTech Venture Partners and chair of the Institute and Faculty of Actuaries, following a career in investment banking, insurance, asset management and the petrochemical industry. He co-founded European Union initiatives on high performance computing and big data in finance, and co-authored High-performance computing in finance and Tercentenary essays on the philosophy and science of Leibniz. Erik holds a master of business administration from the London Business School and as chemical engineer from Universiteit Gent.
Backtesting correlated quantities
Presentation
Backtesting financial models over long horizons often leads to samples that are correlated. This poses a challenge as most standard statistical tests assume their independence. In the talk we will learn various strategies how to deal with this problem and understand the impact of these choices on the discriminatory power of the test.
- arise naturally when backtesting over long horizons
- review of statistical frameworks
- discriminatory power analysis
Nikolai Nowaczyk is a Risk Management consultant who has advised more than 10 clients in over 20 projects around counterparty credit risk, xVA and model validation. He is broadly interested in classical methods of financial mathematics and statistics as well as data science and machine learning. Nikolai holds a PhD in mathematics from the University of Regensburg, has been an academic visitor to Imperial College London and likes to build bridges between academic research and practical applications.
Rough volatility: fact or artefact?
Presentation
- Is volatility rough?
- We explored the question via a model-free approach
- Our simulation study suggests roughness of realised vol is not a good indicator of the roughness of Instantaneous vol
- This suggests that the origin of the roughness observed in realised volatility lies in the estimation error.
Purba Das is a Lecturer in Financial Mathematics at King's College London. Purba's current research interests are Pathwise methods in stochastic analysis and it's application in Finance. Previously, she was a Research Assistant Professor of Mathematics at the University of Michigan (2022-2023). Purba completed her DPhil in Mathematical Institute at the University of Oxford (2018- 2022). Previously Purba received her Master’s, and Bachelor’s degree from Chennai Mathematical Institute, India.
17:45 – 18:30