World Business StrategiesServing the Global Financial Community since 2000

Main Conference Day 1: Thursday 25th September

08.00 – 09.00: Registration and Morning Welcome Coffee

Morning Stream Chair:

Christopher Kantos:

Managing Director and Head of Quantitative Research, Alexandria Technology

AI / LLMs / ML Stream

09.00 – 09.45: “LLMs for Time Series: an Application for Single Stocks and Statistical Arbitrage”

Recently, LLMs (Large Language Models) have been adapted for time series prediction with significant success in pattern recognition. However, the common belief is that these models are not suitable for predicting financial market returns, which are known to be almost random. We aim to challenge this misconception through a counterexample. Specifically, we utilized the Chronos model from Ansari et al. (2024) and tested both pretrained configurations and fine-tuned supervised forecasts on the largest American single stocks using data from Guijarro-Ordonnez et al. (2022). We constructed a long/short portfolio, and the performance simulation indicates that LLMs can in reality handle time series that are nearly indistinguishable from noise, demonstrating an ability to identify inefficiencies amidst randomness and generate alpha. Finally, we compared these results with those of specialized models and smaller deep learning models, highlighting significant room for improvement in LLM performance to further enhance their predictive capabilities.

Sébastien Valeyre:

Portfolio Manager, Machina Capital

AI / LLMs / ML Stream

09.45 – 10.30: Financial Reasoning Agents: In-Context Reinforcement Learning and Test-Time Compute

Large Language Models (LLMs) are increasingly being integrated with reinforcement learning (RL) to push the boundaries of generalist AI agents. In finance, where real-time decision-making is critical, test-time compute efficiency plays a pivotal role in ensuring models can adapt dynamically to evolving market conditions. In-context reinforcement learning (ICRL) is emerging as a transformative approach, enabling LLMs to learn and refine on the fly without explicit fine-tuning. ICRL enhances adaptability in trading, risk assessment, and portfolio optimization. This paradigm shift moves us closer to AI agents capable of robust decision-making, paving the way for more autonomous and generalizable systems in high-stakes applications.

Nicole Königstein:

Chief Data Scientist, Head of AI & Quant Research, Wyden Capital AG

10.30 – 11.00: Morning Break and Networking Opportunities

AI / LLMs / ML Stream

11.00 – 11.45: Pricing Derivatives with a Generative Adversarial Network Approach.

  • Motivation: Deep Learning Approximations with Generative ML
  • Taxonomy of Generative Methods
  • An Introduction to GANs, Conditional GANs and Regression GANs
  • Theoretical Principles for Training GANs
  • Wasserstein GANs and the Earth-Mover Distance
  • Applications

Youssef Elouerkhaoui:

Managing Director, Global Head of Markets Quantitative Analysis, Citi

AI / LLMs / ML Stream

11.45 – 12.30: Morphing Distributions using Gaussian Mixtures and Optimal Transport

Jörg Kienitz:

Independent Consultant, Adjunct Prof (UCT), Assistant Prof (BUW)

12.30 – 13.45: Lunch

Afternoon Stream Chair: TBC

AI / LLMs / ML Stream

13.45 – 14.30: Filtered not Mixed: Stochastic Filtering-Based Online Gating for Mixture of Large Language Models

We propose MoE-F – a formalized mechanism for combining N pre-trained Large Language Models (LLMs) for online time-series prediction by adaptively forecasting the best weighting of LLM predictions at every time step. Our mechanism leverages the conditional information in each expert’s running performance to forecast the best combination of LLMs for predicting the time series in its next step. Diverging from static (learned) Mixture of Experts (MoE) methods, our approach employs time-adaptive stochastic filtering techniques to combine experts. By framing the expert selection problem as a finite state-space, continuous-time Hidden Markov model (HMM), we can leverage the Wohman-Shiryaev filter. Our approach first constructs N parallel filters corresponding to each of the N individual LLMs. Each filter proposes its best combination of LLMs, given the information that they have access to. Subsequently, the N filter outputs are optimally aggregated to maximize their robust predictive power, and this update is computed efficiently via a closed-form expression, generating our ensemble predictor.

Blanka Horvath:

Associate Professor in Mathematical and Computational Finance, University of Oxford

AI / LLMs / ML Stream

14.30 – 15.15: Large Language Models in Financial Services: FAIR -A Framework for Implementation, Risk Mitigation, and Remediation

Miquel Noguer Alonso:

Co – Founder and Chief Science Officer, Artificial Intelligence Finance Institute – AIFI

15.15 – 15.45: Afternoon Break and Networking Opportunities

AI / LLMs / ML Stream

15.45 – 16.30: Machine Learning with Smart Sampling

  • The strengths (and weaknesses) of traditional Machine Learning
  • Smart Sampling Machine Learning:
    • How to do it
    • benefits (and pitfalls)
  • Numerical results & applications.

Ignacio Ruiz:

MoCaX Intelligence
AI / LLMs / ML Stream

16.30 – 17.15: The Psychology of LLMs

  • AI and humans perform surprisingly alike in classic behavioral psychology experiments
  • These experiments show that AI shares many cognitive biases of humans and is prone to human-like errors of logical reasoning and recall
  • We will examine the reasons for this surprising and perhaps even shocking finding, some superficial and some profound
  • Our analysis will lead to concrete, practical techniques for avoiding these biases and increasing the reliability of AI in business settings

Alexander Sokol:

Executive Chairman and Head of Quant Research, CompatibL

17.15 – 18.00: QFC AI Panel

  • In capital markets, can AI be trusted for:
    • Knowledgebase and document repository search and ranking of results
    • Internal or regulatory document generation with human review and approval
    • Fully automated or human-supervised data extraction from unstructured text
    • AI is not to be trusted for anything
  • Which AI model family you found to be most effective in your AI applications – GPT, LLAMA, Mistral, Gemini, Anthropic?
  • Why many banks are not yet comfortable with cloud-based AI even as they widely use the cloud for their most sensitive data?
  • When can we expect to see tangible benefits from using AI – now, next year, in five years, or never?

Moderator:

Alexander Sokol:

Executive Chairman and Head of Quant Research, CompatibL

Nicole Königstein:

Chief Data Scientist, Head of AI & Quant Research, Wyden Capital AG

Ignacio Ruiz:

MoCaX Intelligence

Gala Dinner: Thursday 25th September, 20.00 til late.

Lo Stand Florio Restaurant – The Gala Dinner is complimentary for all conference delegates.

The Florio Stand, commissioned by Vincenzo Florio Jr. to Ernesto Basile in the early 1900s, was built during the Belle Epoque, a fine example of Art Nouveau architecture with incursions into Moorish art. The original idea was to create a large space for sports, entertainment, entertainment and socializing.

The Gala Dinner is complimentary for all conference delegates.

Menu:

Primo Piatto – First Course

Busiate alla Norma con melanzane fritte, salsa di pomodoro, ricotta salata e basilico
Busiate alla Norma with fried aubergines, tomato sauce, salted ricotta cheese and basil

Secondo Piatto – Second Course

Filetto di spigola al pan di agrumi siciliani, patate mascotte al rosmarino e pomodoro gratinato
Seabass fillet with Sicilian citrus breading, rosemary-scented mascotte potatoes and gratinated tomato

Dessert

Cannolo con ricotta home made
Cannolo with ricotta cheese home made

Bevande – Beverage

Mineral and sparkling water
Vino Tasca D’Almerita, Sallier de la Tour, Grillo
Amaro Florio

Contact Information:

08.00 – 09.00: Registration and Morning Welcome Coffee

Morning Stream Chair:

Peter Jaeckel:

Independent financial mathematics and analytics consultant. OTC Analytics

Volatility / Options / Monte Carlo Stream

09.00 – 09.45: ‘A New Class of Tractable Local Vol Models’

Dominique Bang: 

Managing Director, FICC Quantitative Modelling Lead, Bank Of America

Volatility / Options / Monte Carlo Stream

09.45 – 10.30: Next Gen Finite Difference Grids

Stability and accuracy of fd solution: myths, facts and orthodoxy
Discrete consistency: backwards, forwards, Dupire fd solution, monte carlo.

  • local volatility and absence of arbitrage

Dividends.

  • stochastic volatility
  • jumps
  • american options
  • bids and offers of listed options
  • monte-carlo simulation, likelihood ratio tricks and adjoint differentiation

Jesper Andreasen: 

Head of Quantitative Analytics, Verition Fund Management LLC

10.30 – 11.00: Morning Break and Networking Opportunities

Volatility / Options / Monte Carlo Stream

11.00 – 11.45: The Arbitrage Free Autoencoder

Brian Norsk Huge:

Head of Financial Modeling, Trafigura

Volatility / Options / Monte Carlo Stream

11.45 – 12.30: “A General Approach to Statistical Arbitrage”

Bruno Dupire:

Head of Quantitative Research, Bloomberg

12.30 – 13.45: Lunch

Afternoon Stream Chair:

Nikolai Nowaczyk:

Quantitative Analytics, Director, NatWest Group

Volatility / Options / Monte Carlo Stream

13.45 – 14.30: Towards the lower bound for Bermudans

  • An achievable lower bound for a Bermudan is not always the max European
  • Finding a robust (model-free) lower bound for Bermudans is surprisingly difficult
  • It is important for designing “Bermudan discount” adjustment models prevalent in the industry for the last 30+ years
  • We make some progress towards this goal

Vladimir Piterbarg:

MD, Head of Quantitative Analytics and Quantitative Development, NatWest Markets

Volatility / Options / Monte Carlo Stream

14.30 – 15.15: Discovering fundamental financial laws: Universal regimes across full rates spectrum

  • Universal regimes for nominal, real and inflation rates
  • Universal regimes across economic periods and across markets
  • Fundamentals behind the universal regimes. Critical role of Central Banks. Tenor dimension
  • Universal regimes across P and Q universes
  • Expanding Fisher equation from rate levels to its volatilities: is knowing two out of three enough?

Maria Makarova:

Assistant Vice President Quantitative Analyst, BNP Paribas

Vladimir Chorniy:

Managing Director, Head of Risk Model Fundamentals and Research Lab, Senior Technical Lead, BNP Paribas

15.15 – 15.45: Afternoon Break and Networking Opportunities

Volatility / Options / Monte Carlo Stream

15.45 – 16.30: Approaches to Learning Basis Functions for Least-Squares Monte Carlo Methods

  • LSMC for future values, sensis, conditional percentiles to support XVA, risk and capital projections Active research into moving from regression to neural networks, other modern specifications
  • Solves the basis specification problem, particularly onerous in higher dimensions
  • More expensive than simple regression as we’re learning the basis (extracting features) at calc time
  • Stagger the problem by pre-learning the basis? Something between fully-online and fully-offline methods?

Andrew McClelland: 

Director, Quantitative Research, Numerix

Volatility / Options / Monte Carlo Stream

16.30 – 17.15: A Hair for A Square

Subtitle: Standard Error and Goodness of Fit in American Monte Carlo

Authors: Andrew Greene, Nikolai Nowaczyk

– review statistical framework for model validation of regression pricing

– explore advanced metrics using hairy simulation

Nikolai Nowaczyk:

Quantitative Analytics, Director, NatWest Group

Andrew Greene:

Director, CVA Quantitative Analyst, NatWest Markets
All Streams

17.15 – 18.00: QFC AI Panel

  • In capital markets, can AI be trusted for:
    • Knowledgebase and document repository search and ranking of results
    • Internal or regulatory document generation with human review and approval
    • Fully automated or human-supervised data extraction from unstructured text
    • AI is not to be trusted for anything
  • Which AI model family you found to be most effective in your AI applications – GPT, LLAMA, Mistral, Gemini, Anthropic?
  • Why many banks are not yet comfortable with cloud-based AI even as they widely use the cloud for their most sensitive data?
  • When can we expect to see tangible benefits from using AI – now, next year, in five years, or never?

Moderator:

Alexander Sokol:

Executive Chairman and Head of Quant Research, CompatibL

Nicole Königstein:

Chief Data Scientist, Head of AI & Quant Research, Wyden Capital AG

Ignacio Ruiz:

MoCaX Intelligence

Gala Dinner: Thursday 25th September, 20.00 til late.

Lo Stand Florio Restaurant – The Gala Dinner is complimentary for all conference delegates.

The Florio Stand, commissioned by Vincenzo Florio Jr. to Ernesto Basile in the early 1900s, was built during the Belle Epoque, a fine example of Art Nouveau architecture with incursions into Moorish art. The original idea was to create a large space for sports, entertainment, entertainment and socializing.

The Gala Dinner is complimentary for all conference delegates.

Menu:

Primo Piatto – First Course

Busiate alla Norma con melanzane fritte, salsa di pomodoro, ricotta salata e basilico
Busiate alla Norma with fried aubergines, tomato sauce, salted ricotta cheese and basil

Secondo Piatto – Second Course

Filetto di spigola al pan di agrumi siciliani, patate mascotte al rosmarino e pomodoro gratinato
Seabass fillet with Sicilian citrus breading, rosemary-scented mascotte potatoes and gratinated tomato

Dessert

Cannolo con ricotta home made
Cannolo with ricotta cheese home made

Bevande – Beverage

Mineral and sparkling water
Vino Tasca D’Almerita, Sallier de la Tour, Grillo
Amaro Florio

Contact Information:

08.00 – 09.00: Registration and Morning Welcome Coffee

Morning Stream Chair: TBC

Modelling / Quantum / Trading Stream

09.00 -09.45: The Paradigm Shift in Yield Curve Modeling

Andrei Lyashenko:

Head of Market Risk and Pricing Models, Quantitative Risk Management (QRM), Inc.

Modelling / Quantum / Trading Stream

09.45 -10.30: Auto-encoding term-structure models – An arbitrage-free low-dimensionality interest rate model

Alexander Sokol:

Executive Chairman and Head of Quant Research, CompatibL

10.30 – 11.00: Morning Break and Networking Opportunities

Modelling / Quantum / Trading Stream

11.00 – 11.45: Curve Building, Autoencoders v PCA

Ivan Saroka:

Senior Quantitative Analyst, Schonfeld

Modelling / Quantum / Trading Stream

11.45 – 12.30: Forecasting inflation and creating inflation based macro systematic trading strategies

In this talk we shall discuss an approach to forecasting inflation in both DM and EM using machine learning models, contrasting it to more traditional approaches, alongside a discussion of the types of data required. We shall also be discussing how such an approach has performed historically. Later, we discuss ways we can utilise such inflation forecasts within the systematic trading strategies for macro assets such as FX, and how they have performed in a live environment.

Saeed Amen

Turnleaf Analytics / Cuemacro / Visiting Lecturer at QMUL

12.30 – 13.45: Lunch

Afternoon Stream Chair: TBC

Modelling / Quantum / Trading Stream

13.45 – 14.30: Advances in Quantum Machine Learning

Quantum Machine Learning (QML) is an exciting area of quantum computing research that promises to be the first to deliver tangible quantum advantage and quantum utility since many of its algorithms are resistant to some types of noise and do not require large fault-tolerant quantum computers. QML is ideally suited to conducting experiments on the Noisy Intermediate-Scale Quantum (NISQ) computers. We are investigating what is behind the power of QML models and present the latest generation of QML algorithms with applications in quantitative finance.

Alexei Kondratyev:

Research Fellow: ADIA Lab and Visiting Professor: Imperial College London

Modelling / Quantum / Trading Stream

14.30 – 15.15: Quantum-Inspired Algorithms for Enhanced Cross-Sectional Dispersion Trading

There are areas in Finance where computational power has traditionally been the limiting factor — cross-sectional dispersion (CSD) is one of them. Quantum-inspired algorithms offer a breakthrough: they solve complex combinatorial optimization problems efficiently on classical hardware, removing the need to wait for quantum computing advancements.

Financial assets and portfolios exhibit rich qualitative and quantitative structures, forming complex clusters and hierarchies. Previously, unsupervised learning techniques were used to approximate this complexity. Now, leveraging quantum-inspired methods developed over decades at the Centro Superior de Investigaciones Científicas (CSIC) in Spain by the founders of the startup Inspiration-Q, it is possible to fully capture and monitor the entire clustering and hierarchical structure across asset features.

This approach enables, for the first time, an exact solution for market cross-sectional dispersion — where previously only approximations were possible.

Alejandro Rodríguez Domínguez:

Head of Quantitative Research & Analysis, Miraltabank

15.15 – 15.45: Afternoon Break and Networking Opportunities

Modelling / Quantum / Trading Stream

15.45 – 16.30: Topic to be confirmed.

Dmitri Goloubentsev:

CTO, Head of Automatic Adjoint Differentiation, Matlogica

Modelling / Quantum / Trading Stream

16.30 – 17.15: “Risk-Aware Trading Portfolio Optimization”

We investigate portfolio optimization in financial markets from a trading and risk management perspective. We term this task Risk-Aware Trading Portfolio Optimization (RATPO), formulate the corresponding optimization problem, and propose an efficient Risk-Aware Trading Swarm (RATS) algorithm to solve it. The key elements of RATPO are a generic initial portfolio P, a specific set of Unique Eligible Instruments (UEIs), their combination into an Eligible Optimization Strategy (EOS), an objective function, and a set of constraints. RATS searches for an optimal EOS that, added to P, improves the objective function repecting the constraints.

RATS is a specialized Particle Swarm Optimization method that leverages the parameterization of P in terms of UEIs, enables parallel computation with a large number of particles, and is fully general with respect to specific choices of the key elements, which can be customized to encode financial knowledge and needs of traders and risk managers.
We showcase two RATPO applications involving a real trading portfolio made of hundreds of different financial instruments, an objective function combining both market risk (VaR) and profit&loss measures, constrains on market sensitivities and UEIs trading costs. In the case of small-sized EOS, RATS successfully identifies the optimal solution and demonstrates robustness with respect to hyper-parameters tuning. In the case of large-sized EOS, RATS markedly improves the portfolio objective value, optimizing risk and capital charge while respecting risk limits and preserving expected profits.

Our work bridges the gap between the implementation of effective trading strategies and compliance with stringent regulatory and economic capital requirements, allowing a better alignment of business and risk management objectives.

Marco Bianchetti:

Head of Market and Counterparty Risk IMA Methodologies, Intesa Sanpaolo

All Streams

17.15 – 18.00: QFC AI Panel

  • In capital markets, can AI be trusted for:
    • Knowledgebase and document repository search and ranking of results
    • Internal or regulatory document generation with human review and approval
    • Fully automated or human-supervised data extraction from unstructured text
    • AI is not to be trusted for anything
  • Which AI model family you found to be most effective in your AI applications – GPT, LLAMA, Mistral, Gemini, Anthropic?
  • Why many banks are not yet comfortable with cloud-based AI even as they widely use the cloud for their most sensitive data?
  • When can we expect to see tangible benefits from using AI – now, next year, in five years, or never?

Moderator:

Alexander Sokol:

Executive Chairman and Head of Quant Research, CompatibL

Nicole Königstein:

Chief Data Scientist, Head of AI & Quant Research, Wyden Capital AG

Ignacio Ruiz:

MoCaX Intelligence

Gala Dinner: Thursday 25th September, 20.00 til late.

Lo Stand Florio Restaurant – The Gala Dinner is complimentary for all conference delegates.

The Florio Stand, commissioned by Vincenzo Florio Jr. to Ernesto Basile in the early 1900s, was built during the Belle Epoque, a fine example of Art Nouveau architecture with incursions into Moorish art. The original idea was to create a large space for sports, entertainment, entertainment and socializing.

The Gala Dinner is complimentary for all conference delegates.

Menu:

Primo Piatto – First Course

Busiate alla Norma con melanzane fritte, salsa di pomodoro, ricotta salata e basilico
Busiate alla Norma with fried aubergines, tomato sauce, salted ricotta cheese and basil

Secondo Piatto – Second Course

Filetto di spigola al pan di agrumi siciliani, patate mascotte al rosmarino e pomodoro gratinato
Seabass fillet with Sicilian citrus breading, rosemary-scented mascotte potatoes and gratinated tomato

Dessert

Cannolo con ricotta home made
Cannolo with ricotta cheese home made

Bevande – Beverage

Mineral and sparkling water
Vino Tasca D’Almerita, Sallier de la Tour, Grillo
Amaro Florio

Contact Information:

  • Discount Structure
  • Super early bird discount
    20% until 6th June 2025

  • Early bird discount
    15% until 18th July 2025

  • Early bird discount
    10% until 15th August 2025

  • Special Offer
    When two colleagues attend the 3rd goes free!

  • 70% Academic Discount
    (FULL-TIME Students Only)

Download Event PDF

Please enter your details to download this PDF

Error

Event Email Reminder

Error

Also attend