The bridge between
investment banking and
8th & 9th April 2019
Venue: Frankfurt School of Finance & Management, Adickesallee 32-34, 60322 Frankfurt, Germany
MathFinance hosts the annual Conference in Frankfurt which is tailored to the European finance community. Providing cutting-edge research and brand new practical applications, the conference is intended for practitioners in the areas of trading, quantitative or derivative research, risk and asset management, insurance as well as for academics studying or researching in the field of financial mathematics.
As always, we expect around 100 delegates both from the academia and the industry. This ensures a unique networking opportunity which should not be missed. A blend of world renowned speakers ensure that a variety of topics and issues of immediate importance are covered.
This event is a must for everyone in the quantitative financial industry.
Follow us on Twitter
We would like to thank our sponsors:
MathFinance Conference 2019 is supported by:
Dr. Marcus Becker
Arbitrage and Non-linear Taxes
Incorporating a progressive or convex income tax into valuation problems raises the question of the appropriate tax rate to use in common valuation formulas. We apply arbitrage theory in a riskless, as well as risky, (multi-period) economy to answer this question. It turns out that the appropriate tax rate depends on the marginal tax rate of the investor’s initial endowment.
With tax liabilities that are a convex function of the tax base, we identify a new kind of arbitrage: trading strategies where the gains from trades remain unchanged if this strategy is applied multiple times. We call these strategies ‘bounded’ arbitrage opportunities. Going beyond earlier research, we are able to give a complete characterization based on properties of the tax liability function as to whether bounded as well as unbounded arbitrage opportunities will exist.
Marcus Becker is a Quant in Financial Risk at Deloitte’s Risk Advisory division, working on IFRS Hedge Accounting automatisation processes and machine learning algorithms for Robo Advisory. Before, he was working as a Risk Manager in the Traded Credit and Wholesale Analytics team at HSBC Düsseldorf steering the counterparty credit risks for OTC derivatives. In 2016, he obtained his Ph.D. from Freie Universität Berlin working on general arbitrage theory with non-linear taxation. Marcus has a strong background in stochastics and mathematical finance. His main research interests focus on wealth allocation systems and re-distribution processes by optimal tax designs.
Dr. Mark Beinker
Derivative pricing: A pattern-matching problem?
Before computers became a handy tool available to everyone everywhere, pricing of financial options was mainly a matter of experience, a pattern matching task performed by human neural networks. This changed drastically with the advent of the Black-Scholes equation an the availability of computer power which enables everyone to price options in no time. Today, the pricing of derivatives has become a more complex, sophisticated exercise and enormous computer power is required to calculate all the needed portfolio-based risk figures (e.g. XVA). In this case study, we examine if the application of deep neural networks, which allow pricing of complex derivatives in constant time, could be helpful.
Dr. Mark Beinker is partner at d-fine GmbH and head of d-fine’s derivative valuation consulting business. With more than 20 years of experience in this field, he has been responsible for projects covering all aspects related to derivative pricing, including specification, implementation and testing of new and old pricing models, selection of third party tools, estimation of risk figures and valuation adjustments, and the design and specification of IT infrastructure and business process required for efficient valuation and processing of financial derivatives. He is also responsible for d-fine’s bespoke pricing library MoCo and tools based thereon. Before he joined d-fine he worked as a manager at Arthur Andersen. He holds a PhD in theoretical particle physics.
Prof. Dr. Griselda Deelstra
Université libre de Bruxelles
Mutivariate FX models with jumps: Triangles, Quantos and implied correlation
We propose an integrated model of the joint dynamics of FX rates and asset prices for the pricing of FX derivatives, including Quanto products; the model is based on a multivariate construction for Lévy processes which proves to be analytically tractable. The approach allows for simultaneous calibration to market volatility surfaces of currency triangles, and also gives access to market consistent information on dependence between the relevant variables. A successful joint calibration to real market data is presented for the particular case of the Variance Gamma process.
Griselda Deelstra is Professor in Stochastic Finance in the Department of Mathematics at the Université libre de Bruxelles (ULB), where she is the director of the Actuarial Sciences Group. Griselda Deelstra’s research interests focus on stochastic modelling in finance and insurance, in particular on risk management and derivative pricing. She holds a PhD from the Vrije Universiteit Brussel (1996). In her career, she had teaching/research positions at ENSAE and CREST, UGhent and VUB.
Prof. Dr. Wolfgang Härdle
Chair Professor of Statistics at the School of Business and Economics
Humboldt University of Berlin
Pricing Cryptocurrency Options: the Case of CRIX and Bitcoin
Wolfgang Härdle did 1982 his Dr. rer. nat. in Mathematics at Universität Heidelberg and 1988 his Habilitation at Universität Bonn. He is Ladislaus von Bortkieviecz chair professor of statistics at the School of Business and Economics, Humboldt-Universität zu Berlin. He is director of C.A.S.E. – Center for Applied Statistics & Economics. He leads the Collaborative Research Center “Economic Risk” and the International Research Training Group (together with WISE, Xiamen University) „High dimensional, non stationary time series“. His research focuses on dimension reduction techniques, computational statistics and quantitative finance. He has published 34 books and more than 250 papers in top statistical, econometrics and finance journals. He is one of the “Highly cited Scientist” according to the Institute or Scientific Information.
Ph. D. Candidate
Humboldt University of Berlin
Realized Cryptocurrency Volatility Forecasting with Jumps: An Empirical Study
In this paper, to investigate the realized volatility process of the highly volatile cryptocurrency assets, we employ threshold realized variance method to separate jump component apart from continuous process. Jumps appear more frequently and significant comparing with other assets, we can find jumps bring more information, despite that the jump process doesn’t improve explanation power. The change of one-day lagged threshold jump component has significant positive impact on the change of future realized volatility, this suggest that investors are likely to find cryptocurrency market more volatile after jumps happened one day before. However, the un-threshold jump component is negatively correlated to the future realized volatility significantly. We then evaluate the economic values among a variety of forecasting results under a simple utility-based framework.
Junjie Hu is a Ph.D. candidate at Ladislausvon Bortkiewicz chair of statistics, Humboldt-University in Berlin. He holds a M.Sc. in Finance from Sun Yat-sen University, China. His current research interests are time series modeling and forecasting for financial markets, computational statistics and applied machine learning.
Prof. Dr. Karel in ’t Hout
Associate Professor Mathematics and Computer Science
University of Antwerp
Numerical Valuation of Bermudan Basket Options via Partial Differential Equations
We study the efficient numerical valuation via partial differential equations of Bermudan basket options with a large number of underlying assets. To deal with the high-dimensionality of the problem, we combine the principal component analysis from Reisinger & Wittum (2007) with modern alternating direction implicit (ADI) schemes, see e.g. in ‘t Hout & Welfert (2009). The convergence of this combined analytical-numerical approach is investigated in detail, and ample numerical experiments are presented illustrating the high performance it attains. This is joint work with Jacob Snoeijer (U Antwerp).
Karel in ’t Hout is Associate Professor in the Department of Mathematics and Computer Science at University of Antwerp, specializing in the analysis and development of numerical methods for time-dependent partial differential equations with applications to finance. He has previously held positions as Visiting Professor at Arizona State University, Visiting Professor at Boise State University and Researcher at Leiden University and University of Auckland. Karel has also spent time in the industry, working as quantitative analyst at ABN Amro, Amsterdam. He holds a PhD in Mathematics from Leiden University.
Dr. Antoine Jacquier
Senior Lecturer in Mathematics / Director MSc Mathematics and Finance
Imperial College London
VIX Options in Rough Volatility Models
We discuss the pricing and hedging of volatility options in some rough volatility models. First, we develop efficient Monte Carlo methods and asymptotic approximations for computing option prices and hedge ratios in models where log-volatility follows a Gaussian Volterra process. While providing a good fit for European options, these models are unable to reproduce the VIX option smile observed in the market, and are thus not suitable for VIX products. To accommodate these, we introduce the class of modulated Volterra processes, and show that they successfully capture the VIX smile. Joint work with Blanka Horvath and Peter Tankov.
Dr Jacquier is a Senior Lecturer in the Department of Mathematics at Imperial College London. His research focuses on volatility modelling, with a special emphasis on rough volatility and applications of asymptotic methods in finance. He holds a PhD in Mathematics from Imperial College London, has co-edited a book on Asymptotic Methods in Finance, and has published more than 30 papers in mathematical finance and applied probability.
Machine Learning for Factor Investing
Style investing helps to construct portfolios that deliver positive long-term returns and have a low correlation with the market. We study the benefits of use of various machine learning techniques on the portfolio construction and “style” selection stages. The goal of the first stage is to find an optimal combination of the asset characteristics, which has a superior predictive power of the future returns. On the second stage we use macro economic parameters to study the “style” factor rotation to construct a composite portfolio that is comprised of individual factor-based strategies.
Vadim Kanofyev is a Quantitative Researcher at Bloomberg L.P. His research interests include quantitative asset allocation, algorithmic trading strategies and derivatives pricing. He has an extensive experience in numerical computing, applied machine learning and financial econometrics. Vadim holds a Master’s degree in Economics from the University of Pennsylvania.
Managing Director CIO Multi Asset Europe
Allianz Global Investors
Current Challenges and Developments of the Investment Industry
Ingo R. Mainert is Managing Director and CIO Multi Asset Europe at Allianz Global Investors. He is a member of the European Executive Committee, the management board of Allianz Global Investors GmbH and the Global Policy Council.
Mr Mainert has worked in the asset management industry since 1994. Before the merger between Cominvest Asset Management GmbH and Allianz Global Investors, Mr Mainert held various managerial positions at Cominvest Asset Management GmbH from 2001 to 2010, and spent the last spending four years as Managing Director and Chief Investment Officer.
Mr Mainert worked for Commerzbank AG since 1988 in various roles including: Head of Asset Management Private Banking, Head of Fixed Income/Currencies and Head of Equity Strategy Germany.
Mr Mainert is deputy chairman of DVFA – Society of Investment Professionals in Germany and a member of the Issuer Market Advisory Committee (IMAC) of the Deutsche Börse. He is also a representative of the European Fund and Asset Management Association (EFAMA) in the Bond Market Contact Group (BMCG) of the European Central Bank (ECB) and an honorary associate member of the Objections Committee at Federal Financial Supervisory Authority (Bundesanstalt für Finanzdienstleistungsaufsicht – BaFin).
Ingo Mainert obtained his degree in business administration from Johann Wolfgang Goethe-University, Frankfurt/Main, and completed his Certified Investment Analyst/DVFA qualification.
Unicredit Bank AG
Estimation of Volatility Surfaces via Functional Approximation and Machine Learning
Volatility surfaces are crucial for the pricing and hedging of a huge number of financial instruments, yet they cannot always be obtained reliably from financial markets for various reasons. In this talk we show how to predict volatility surfaces for equity underlyings in scenarios where a price estimation of the underlying is available. This is accomplished by leveraging a newly improved functional approximation algorithm in combination with supervised machine learning. This talk gives a formal introduction to the theoretical concepts and showcases real-world applications from the banking industry.
Maximilian Mair is working as a quantitative analyst at UniCredit Bank AG. His area of work includes mathematical finance, data analytics and machine learning applications in the context of equity derivatives. He holds a diploma in mathematics from the LMU University in Munich and is a member of the research team of Kathrin Glau at the chair for mathematical finance at TUM University Munich.
Dr. Jacopo Mancin
Volatility Swaps: PDE Pricing Improvements for LSV frameworks
We study one dimensional PDE pricing techniques for volatility swaps that are able to closely replicate Local Stochastic Volatility Monte Carlo prices also in the context of fairly oscillating volatility term structure, while sensibly improving the computational performance. Our analysis is based on Foreign Exchange (FX) volatility swaps, but can be equivalently applied to other asset classes.
Jacopo is a quant at Barclays, which he joined in March 2017. His main focus is PDE pricing in LSV and Hybrid models for FX. Prior to that, he earned a PhD in Financial Mathematics at the LMU University in Munich, working on Model Uncertainty under the supervision of Prof. Dr. F. Biagini.
Panelist: Crypto Exchanges and Custody Services
Adrian has over 24 years of experience in financial markets, with a primary focus on currency management. He has worked for Sal. Oppenheim Jr. & Cie. Frankfurt/Zurich, Lehman Brothers International, London, Commerzbank AG and Deutsche Bank Frankfurt. He is a founding partner in the family office and asset management advisor company Resilience AG in Switzerland, founded in 2010. He also participated in Mio AG, a blockchain start-up in 2018. Adrian is a member of the supervisory board of Mathfinance AG. He has a strong understanding of equity and currency valuation and markets. Adrian graduated in 1995 from Goethe University Frankfurt with a Diploma in Business Administration.
Dr. Daniel Oeltz
Pailab – audit, revisioning and analysis of machine learning in finance
The free availability of sophisticated machine learning packages has led the ground for a widespread application of machine learning ideas to classical problems of quantitative finance. But not only the heavily regulated financial industry is confronted with some challenges: How to make the development audit safe, i.e. how to version and track changes of different models. How to analyze changes between different versions and finally how to deploy these algorithms into a production setting?
In this talk we discuss methods to tackle these questions in a very general setting based on the open source machine learning environment pailab. Here, after an introduction to the overall topic we will focus on model interpretability as one ingredient to the solution of the aforementioned problems. For ease of presentation, we choose a simple but illustrative use case: The approximation of a call price function for the Scott-Chesney stochastic volatility model using neural networks.
Daniel is managing partner at RIVACON and one of its co-founders. He worked as a quantitative analyst at Sal. Oppenheim Jr. and Cie in the front office, as quantitative risk controller at Talanx Asset Management, and as head of quantitative analysis at Macquarie Group Germany. In these different roles, he dealt intensively with model development and risk management in the context of equity, currency, interest rate and commodity derivatives.
Daniel holds a degree in Mathematics from the University of Bonn and received his doctorate at the Institute of Scientific Computing and Numerical Simulation at Bonn.
Dr. Alla Petukhina
Humboldt University of Berlin
Portfolio Optimization with CoVar in Cryptocurrency Markets
Recently, a new approach for portfolio optimization with risk measured by Conditional Value-At-Risk (CoVaR) was suggested (Zalewska A., 2018). The approach is based on the stress event of chosen asset being equal to the opposite of its value-at-risk level, under the normality assumption. The current research investigates characteristics of the CoVaR oprimized portfolios applied to the universe of cryptocurrencies and stock indices serve as stress event assets. We test in- and out-of-sample performance compared with other widely used risk-based portfolio optimization rules, such as such as Mean-variance model (MV), Risk-parity (ERC) and Maximization diversification (MD). The empirical results show high potential of crypto-currencies for portfolio hedging and improve the risk-return profile of portfolios. Performance of CoVaR portfolio is strongly dependent on the level of stress event of the core asset.
Alla Petukhina is a research assistant at LvB chair of statistics. Alla Petukhina holds a M.Sc. in economics from the Ural state university, Russia. He received her Ph.D. degree in statistics and econometrics from Humboldt University of Berlin in 2018. Her research interests are focused on asset allocation strategies and risk modelling for high-dimensional portfolios, investment strategies in crypto-currencies market.
Prof. Rolf Poulsen
University of Copenhagen
The Fed Isn’t Federal – And Other Odd Things in Finance
Things nobody told you about quantitative finance. Some for good reason.
Rolf Poulsen is a professor of Mathematical Finance at the Dept. of Math. Sciences at the University of Copenhagen. His main research interest is quantitative methods for pricing and hedging of derivatives. He will talk about exchange rate markets at length to all who will listen – and some who won’t.
Dr. Adil Reghai
Real Time management with local stochastic volatility
Prof. Dr. Thorsten Schmidt
Professor for Mathematical Stochastics
University of Freiburg
Generalized arbitrage in the sense considered here corresponds to trading strategies which yield positive gains on average in a class of scenarios rather than almost surely. The relevant scenarios or market states are specified via a sigma-algebra and so this notion contains classical arbitrages as a special case. It also covers the notion of statistical arbitrage introduced in Bondarenko (2003). Relaxing these notions further we introduce generalized profitable strategies which include also static or semi-static strategies. We show that even under standard no-arbitrage (NA) there may exist generalized gain strategies yielding positive gains on average. In the first part of the paper we characterize these generalized no-arbitrage notions. In the second part of the paper we explicitly construct profitable generalized strategies and study their performance on simulated data and on market data. These strategies, albeit simple in nature, show a surprising performance being profitable on average with little remaining risk. This is joint work with Christian Rein and Ludger Rüschendorf.
Thorsten Schmidt is Professor for Mathematical Stochastics at University Freiburg (successor of Ernst Eberlein). Prior to this he was professor for Mathematical Finance at Chemnitz University of Technology since 2008, held a replacement Professorship from Technical University Munich in 2008 and was Associate Professor at University of Leipzig from 2004 on. His Ph.D. he obtained from University in Giessen in 2003 on credit risk with infinite dimensional models. Besides his interests in Mathematical Finance, in particular interest rates, credit risk and energy markets, he has a strong background in statistics and probability theory. His research focusses on topics in mathematical finance and the theory and application of stochastic processes. This includes credit risky markets, interest rate markets, dynamic term structure models, insurance mathematics, energy markets and related fields.
Dr. Martin Simon
Head of Equity and Equity Derivatives Valuation
Deka Investment GmbH
Stock Price Bubbles – An Option-based Indicator
In this talk we are going to discuss an option-based mathematical indicator for stock price bubbles. The first introductory part recaps the strict local martingale theory for modeling asset price bubbles and its implications for pricing contingent claims. In the second part we present a novel forward-looking indicator based on the information content of bid and ask market quotes for exchange-traded plain vanilla options. Our construction is motivated by a recent theoretical result by A. Jacquier and M. Keller-Ressel proving that bubbles can be identified from the asymptotic behavior of the implied volatility surface. However, in practice, the resulting inverse parameter identification problem is ill-posed and we adopt a statistical perspective in order to cope with this ill-posedness and to quantify the indicator’s inherent uncertainty. Finally, we provide real-market tests of the proposed indicator with focus on tech stocks addressing increasing concerns about a tech bubble 2.0. The talk is based on joint work with Lassi Roininen, Petteri Piiroinen and Tobias Schoden.
Martin Simon works for the German asset management company Deka Investment where he is head of equity and equity derivatives valuation. His research interests focus on numerical analysis, high performance computing, mathematical modeling and uncertainty quantification with applications in pricing and hedging of equity derivatives, asset allocation and risk management. He holds a doctoral degree in applied mathematics from the University of Mainz.
Practical and quantitative issues to be solved trading and risk managing crypto assets As a matter of fact, the Crypto Asset markets is a new asset market and attracts a lot of investors. In the last two years, we have seen a lot of new entrants in the market: crypto currency exchanges, security token platforms, stable coins, established financial services institutions that are launching crypto products and services. Crypto Assets are meanwhile impossible to ignore; even Amazon, Google, Alibaba, Starbucks etc. seem to think Crypto Assets are worth paying attention to. Whilst Crypto Assets may came to stay, they have been object of mixed reactions and expectations through the last years. Some investors saw cryptos as a unique way to diversify their portfolios and avoid some of the downsides of traditional bonds, stocks and commodities. Those investors rather disregard negative comparisons to the dot-com era by pointing to the Nasdaq Composite’s recovery 15 years later and the enormous potential of cryptos and blockchain on the society. They also note that Bitcoin has rebounded from past crashes of similar magnitude. Others are concerned about market manipulation, money laundering, excessive hype, security flaws, tighter regulation and slower-than-anticipated adoption by Wall Street. However, a soundstanding market for Crypto Currencies, Futures&Options and other Crypto Derivatives has not been built up yet with regards to regulations, transparency, liquidity, reliable valuation and pricing tools. Since Crypto Assets are still immature, they will become less volatile as they mature. First steps towards pricing and margining of Crypto Derivatives and how to measure the risk of Crypto Assets can now be presented. Sophisticated statistical methods and analysis classified cryptocurrencies as a new asset class with unique features in the tails of the distribution and confirm the potential of crypto-currencies for portfolio hedging and improve the risk-return profile of portfolios.
Frank has been a Managing Partner of WEPEX Consulting for Financial Markets since 2010 acting as Head of WEPEX Competency Technology & Digitalization (DLT, Blockchain, Crypto Assets, RPA, Intelligent Automation, Artificial Intelligence, Big Data, Digital Banking, Platform Economy). He is co-founder and Managing Partner of WEPEX micobo Digital & Blockchain and Global Crypto Excellence GmbH. After a degree in business administration (Diplom-Kaufmann, Business Administration Graduate) and postgraduate degrees (M.Phil., CFA), doctoral studies (Drs.rer.pol.) and lectureships followed at various universities (RWTH Aachen University, FOM University, Frankfurt School of Finance and Management, Maastricht School of Management, University St. Gallen, University of Derby, Osnabrueck). Starting his career with an apprenticeship as banker (Bankkaufmann) in Germany, Frank has 30 years of international professional experience in the financial industry: Consulting and management positions at renowned international banks, exchanges and asset managers in Germany, Europe, US, Near/Middle East and Asia (Deutsche Bank, WestLB, Arthur Andersen / Accenture).
Dr. Niels Wesselhöft
Humboldt University of Berlin
Separating the asset universe from cryptocurrencies
The aim of this paper is to find the proximal genus and the specific difference (genus proximum et differentia specifica) for the daily time series of cryptocurrencies returns, compared to classical asset returns.
In this sense, a daily time series of asset returns (either crypto or classical assets) can be characterized by a multidimensional vector with statistical components like volatility, skewness, kurtosis, tail probability, quantiles, expected shortfall or fractal dimension.
By using dimension reduction (Factor Analysis) and classification models (Support Vector Machines) for a representative sample of cryptocurrencies, stocks, exchange rates and commodities, we are able to classify cryptocurrencies as a new asset class with unique features in the tails of the distribution.
The DLT Markets AG
Panelist: Crypto Exchanges and Custody Services
Prof. Dr. Uwe Wystup
FX Options Greeks Unlimited
Black-Scholes delta, smile delta, model delta, premium-included delta, inverse delta, forward delta, traders’ gamma, Black-Scholes vega, traders’ vega, vanna, volga, volunga, vanunga, aega, rega, sega, revga, bufga, enoughga.
Uwe Wystup is managing director of MathFinance AG. Before, he has actively worked in FX derivatives trading as Financial Engineer, Global Structured Risk Manager and Advisor since 1992, including Citibank, UBS, Sal. Oppenheim and Commerzbank. He is one of the few hybrids in the world working in the intersection of the derivates market and academic research.
Uwe earned his PhD in mathematical finance from Carnegie Mellon University, is currently Professor of Financial Option Price Modeling and Foreign Exchange Derivatives at University of Antwerp and Honorary Professor of Quantitative Finance at Frankfurt School of Finance & Management.
Together with his team at MathFinance he provides independent (re-)structuring, valuation, model validation and expert witness services.
His first book Foreign Exchange Risk was published in 2002, quickly became the market standard and has also been translated into Mandarin. His second book FX and Structured Products appeared in 2006 with a fully updated and expanded second edition in 2017. Many of his papers appeared in scientific journals.
Enjoyable atmosphere, lots of networking, expert speakers, way to learn developments in the industryArtur Sepp
The MathFinance conference provides an excellent environment to learn about recent developments and networking with leading experts from both industry and academiaMartin Simon
The conference is a great opportunity to meet interesting people and develop new ideas on recent market trends. Special thanks to the organizers, they did a very good jobEugen Tiganu