Thalesians Ltd

Paul Bilokon, PhD, Imperial College London discusses the limitations of Harry Markowitz's 1952 Portfolio Optimization Model. He delves into issues such as non-stationarity in financial time series and proposes modern AI-based solutions like reinforcement learning and multi-armed bandits to address these challenges. Explore the impact of AI on the quantitative finance workforce, the critical need for explainable AI and more! Dive in here 👉 http://spr.ly/6046QPg4p #QuantMinds #portfoliooptimization #AI #quantfinance

Looking forward to the Thalesians Ltd Christmas Dinner on Wednesday 18 Dec at 6pm GMT at the iconic City University Club in London! Paul Bilokon, PhD will be giving a very exciting talk on "Optimal allocation of resources in nonstationary conditions". Sign up for the dinner/talk below at https://lnkd.in/dtEiwKd5

After a long time, as I have been ill for years, a new paper on Interpretability in deep learning for the volatility smile-to-parameters map in the Rough Heston model. This is a continuation and an upgrade of the first paper of 2021 on normal Heston to rough volatility. Deep learning is everywhere now, it entered even the Nobel award in physics and chemistry this year, but the black box feature of this technique remains the elephant in the room in many cases. The initial response of the scientific community has been introducing local and global interpretability methods. In this paper we apply LIME, DeepLift, LRP and Shapley Values (cooperative game theory) to interpret the Deep Neural Network mapping the volatility smile to the rough Heston model parameters. My hope is that this research will encourage more contributions in this field of interpretability, which is still in its infancy. Joint work with Bo Yuan (Cambridge), Jack Jacquier (Imperial), and Nicola Pede (Mediobanca), whom I thank all for their collaboration and insight. SSRN: https://lnkd.in/eq4Zpwsh ArXiv: https://lnkd.in/ej_Nxgcx Previous paper on interpretability for normal Heston: SSRN: https://lnkd.in/dNY-NtN ArXiv: https://lnkd.in/e4_-dXHD

Dear colleagues: It is a great pleasure for me to share with you the publication of the second edition of our book Malliavin Calculus in Finance: Theory and Practice:  https://lnkd.in/d4wYFht5 In this new edition, we comment some recent results on the implied and local volatilities, and we add a new chapter on the behaviour of the Bachelier implied volatility. Thanks a lot to my coauthor David García Lorite and our editors Callum Fraser and Mansi Kabra. Working with them has been an enriching and very nice experience.

How to Perform RAG with the New Open-Source Library 'aisuite' for Financial Analysis? 👉 Recently, I explored how to streamline financial analysis using a combination of the new open-source library "aisuite", LlamaParse from LlamaIndex, a vector-based approach, and powerful LLMs like GPT-4o and Claude Sonnet 3.5. Here's a breakdown: 👉 The Workflow: 1- Parsing Financial PDFs with LlamaParse: I used LlamaParse to convert Amazon’s 2023 10-K report into a structured format for analysis. 2- Chunking for Retrieval: Using a SentenceSplitter, I divided the report into chunks optimized for embedding and retrieval. 3- Building a Vector Index: I created a vector store using the local embedding model BAAI/bge-small-en-v1.5, ensuring full control. 4- Querying Financial Data with RAG: Questions like "What was the total lease cost for 2023?" and "What was the net income in 2023?" were answered by combining context retrieval and language model synthesis. 5- Comparing LLM Outputs: Using aisuite, I ran the queries through both GPT-4o and Claude for robust, multi-perspective answers. ▪️ Claude: Total lease cost: $18.918 billion. ▪️ GPT-4o: Net income: $30.437 billion. 👉 Why It Matters: With RAG workflows powered by "aisuite", we can automate the tedious process of sifting through dense financial reports, and comparing results from different LLMs. This approach not only accelerates analysis but also ensures accuracy, empowering analysts to focus on decision-making. 👉 This is just the beginning for 'aisuite.' Will they add AI Agents implementation? If so, I’d be very excited to test it. — Enjoy this? ♻️ Repost it and share it with your network. Want to learn more about building AI-powered tools? Let's connect! Drop your thoughts in the comments below 👇

An interview with Neil Clive Fowler has just been published in the Thalesians Magazine. Neil talks about Women in Quant Finance, careers, LLMs, Quantitative Developer Certificate (QDC), The Machine Learning Institute Certificate in Finance (MLI) and more. https://lnkd.in/e_zhKAvC

🎉 Quant DevNight: A Night of Quantitative Insights, Innovation Inspiration and International Collaboration! 🎉 #FinTech_Dev_Night #3 (https://lnkd.in/esj-FaW4) is in the books, and I'm truly impressed by the amazing team at Man Group and humbled by the energy and the unique perspectives brought by our speakers, panelists and attendees. 🙌 A heartfelt thank you to our incredible speakers and panelists for sharing their expertise and sparking thought-provoking conversations: 💡 Barry Fitzgerald, Man Group 💡James Munro, Man Group (ArcticDB) 💡Saeed Amen, Turnleaf Analytics 💡Paul Bilokon, PhD, Thalesians Ltd 💡Thomas Schmelzer, Abu Dhabi Investment Authority (ADIA) 💡Nicole Königstein, quantmate 💡Rajnish Sharma, Allianz Global Investors Your deep insights made the evening special and gave energy to continue fostering these dialogues. A personal highlight for me was meeting John O'Hara, author of the AMQP protocol, fueling RabbitMQ and many message queueing systems. 🔥 Stay tuned! We’ll be sharing key takeaways from the event, including a detailed recap of the panel discussion—don't miss it! 🙏 A special shoutout to the incredible team at Man Group who made this event possible: ✨ Abhi Thakur ✨ Elle Palmer 👏 Thank you for your endorsement of the idea and dedication! It's been a great pleasure working together, and I am excited about what to come! Looking forward to more engaging events ahead and continuing to push the boundaries of fintech innovation and collaboration together! 🚀 #quanttech #quantdev #networking #collaboration #innovation #fintech_devnight #dare2deliver #ai #ai_in_finance #london #quantminds

Announcing new open-source Python package: aisuite! This makes it easy for developers to use large language models from multiple providers. When building applications I found it a hassle to integrate with multiple providers. Aisuite lets you pick a "provider:model" just by changing one string, like openai:gpt-4o, anthropic:claude-3-5-sonnet-20241022, ollama:llama3.1:8b, etc. pip install aisuite Open-source code with instructions: https://lnkd.in/gB3AWxvh Thanks to Rohit Prsad, Kevin Solorio, Ryan Prinz, Jeff Tang and John Santerre PhD for helping build this!

There is still time to register for the final International Association for Quantitative Finance IAQF/Thalesians Ltd talk of the year in NewYork: Hybrid: Frédéric Godin - Enhancing Deep Hedging of Options hosted at Fordham University and online. Abstract: We present a dynamic hedging scheme for S&P 500 options, where rebalancing decisions are enhanced by integrating information about the implied volatility surface dynamics. The optimal hedging strategy is obtained through a deep policy gradient-type reinforcement learning algorithm, with a novel hybrid neural network architecture improving the training performance. The favorable inclusion of forward-looking information embedded in the volatility surface allows our procedure to outperform several conventional benchmarks such as practitioner and smiled-implied delta hedging procedures, both in simulation and backtesting experiments. Bio: Frédéric Godin is an Associate Professor at Concordia University (Montreal, Canada) in the Department of Mathematics and Statistics. His areas of research are financial engineering, risk management, actuarial science, reinforcement learning and energy markets. He also holds the Fellow of the Society of Actuaries (FSA) and Fellow of the Canadian Institute of Actuaries (FCIA) designations. https://lnkd.in/ebJpumUG