Quant Insights Conference:
Portfolio Management in Quant Finance

Fat Tailed Diversification: Entropies, Tail Covariances et al

There are infinitely many ways to orthogonalize a basket. In the presence of fat tails, though, orthogonalization is only a part of the equation. We will discuss general approaches to the other, missing part of this equation, and a specific realization of such approaches using a power-law estimate for marginal entropies. This results in a parallel diagonalization of the usual covariance matrix, and a specific slice of the mixed moment tensor, termed the tail covariance matrix. We will show some examples, and discuss general ways forward.

Dr. Jan Rosenzweig, Portfolio Manager, Pine Tree Market Neutral

Jan has been working in the financial markets for close to 20 years, as a quant, structurer, trader and portfolio manager. He worked at Credit Suisse, Rabobank, HSH Nordbank, IV Capital, Brancherose and Pine Tree. He has a PhD in applied maths from Cambridge and BSc in applied maths and computer science from Zagreb. Jan is based in London.

How AI is Used to Generate Alpha in Investing

AI is widely used in many professions including medicine as AI can help in detecting patterns. How is AI used in a low signal to noise environment like investing to generate Apha?

Renee Yao, Founder, Neo Ivy Capital Management

Renee Yao began her career on Wall Street at Citadel, LLC. Later she left Citadel to join WorldQuant, a Millennium Management Portfolio Manager, as a Trader with discretionary trading authority for her own designated account. Subsequently she joined a prop trading firm where she managed multiple statistical arbitrage portfolios totaling several hundred millions USD in gross positions. In early 2015, Renee set up her own firm, Neo Ivy Capital Management, a quantitative hedge fund manager that focuses on trading liquid, publicly traded equity securities via artificial intelligence strategies. Renee was named as “Tomorrow’s Titan” by The Hedge Fund Journal in 2019 and “100 People Transforming Business” by Business Insider in 2020. Renee holds a M.A in Statistics from Columbia University.

Modeling Volatility Risk-Premia

This talk will cover:

• What is volatility risk-premia
• Reconciliation between arbitrage-free valuation and risk-premia
• How to infer risk-premia from option prices
• Examples using options on Bitcoin and Ether
• Implications for portfolio construction and risk management

Dr. Artur Sepp, Head of Quantitative Strategies, Clearstar Labs

Artur Sepp is the head of quantitative strategies and a portfolio manager at Clearstar Labs in Zurich – a family office specializing in systematic strategies for crypto assets and DeFi. Artur has over 15 years of experience in financial markets including heading quant research at a crypto manager and a systematic hedge fund, leading development of front-office quant strategies and derivatives at private (Julius Baer) and investment banking (Merrill Lynch/BofA). Artur has a PhD in Mathematical Statistics from University of Tartu, an MSc in Industrial Engineering and Management Sciences from Northwestern University, and a BA cum laude in Mathematical Economics from Tallinn University of Technology. His expertise covers quantitative investing and asset allocation, quantitative modelling of derivative securities, machine learning and data science, and blockchain applications with DeFi. He is the author and co-author of several research articles on quantitative finance published in key journals. Artur won Risk Magazine’s Quant of the Year Award (2024).

Hierarchical Minimum Variance Portfolios

We reveal a mathematical connection between Hierarchical Risk Parity (HRP) and minimum variance portfolios, thus establishing a bridge between the Machine Learning approach advocated by Lopez de Prado and more traditional Modern Portfolio Theory (MPT) going back to Markowitz.

Peter Cotton, Chief Data Scientist, ExodusPoint Capital Management

Peter Cotton is Chief Data Scientist at ExodusPoint Capital Management. He was named one of the Top 10 AI minds on Wall Street and also the Top 10 Quants on Wall Street. He created two companies, Benchmark Solutions and Microprediction LLC and has held quantitative leadership positions at buy and sell side firms including Morgan Stanley, JP Morgan Chase and Intech Investments.

The Markets are Not Normal So, When Building Strategies, We Should Do the Right Thing!

The daily returns of stock markets are not normally distributed and not homoskedastic. This is well known to analysts but, despite that, the quantitative finance community continues to model them as if they are. Not only is this empirically unsound, it is also leads to asset allocation procedures that are excessively risky and underperform more simple approaches. This underperformance is entirely consistent with the true fat-tailed nature of actual returns.

Dr. Graham Giller, CEO, Giller Investments LLC

Graham is a senior data scientist and financial researcher, with over 25 years’ experience as one of Wall Street’s original Data Scientists. He has a doctorate from Oxford University in Experimental Elementary Particle Physics, where his field of research was statistical cosmic ray astronomy which featured large scale computer-based data analytics in an era where that was uncommon. He joined Morgan Stanley in London in 1994, where he worked on derivatives pricing and then was an early member of the now famous Process Driven Trading unit (PDT) run by Peter Muller in New York. Within PDT he managed the Futures Group, developing and managing systematic trading systems for financial futures and futures-options. He also worked on theoretical models of optimal trading strategy. In 2000, he founded his own investment fund, Giller Investments (New York), LLC, as a Commodity Pool Operator/Commodity Trading Advisor and was involved in various quantitative proprietary trading ventures over the next 12 years. In 2012—2013 he built an innovative platform for geospatially aware bidding in Google ad auctions for an internet marketing agency and in 2013 was recruited to Bloomberg LP to lead the data science effort within the Global Data division. Graham joined J.P. Morgan in 2015 as Chief Data Scientist, New Product Development, and he was appointed Head of Data Science Research in 2016. He joined Deutsche Bank as Head of Primary Research to create the unit that incorporated alternative data into analyst research content, leaving that role in 2019. He is the author of Electronics: principles and applications (1992), Adventures in Financial Data Science (2020) and Essays on Trading Strategy (2023). For the last two years he has written a regular column in WILMOTT Magazine.

The Power of Data and Quantitative Approaches in Discretionary Investing

In this talk, we delve into the growing significance of data-driven and quantitative approaches within discretionary investing. We will explore some use cases that highlight the value of incorporating these methodologies.

Dr. Pamela Saliba, Investment Manager, Pictet Asset Management

Pamela joined Pictet Asset Management in 2019 and is currently an Investment Manager in the Thematic Equities team. She was previously a quantitative analyst within the Pictet Quest equities fund where she focused on signal generation via analyzing financial and alternative datasets, back testing and portfolio construction. Pamela holds a Master in Probability and Finance (El-Karoui masters) from University Paris-VI, and earned her Ph.D. in Applied Mathematics from Ecole Polytechnique de Paris in 2019. During her PhD, she worked in the Surveillance department at the French regulator AMF. Her PhD is in the area of market microstructure, with a particular focus on high-frequency data analysis, market impact and order book modelling.

Adaptive Diversification

Compound returns can be enhanced with higher alpha, lower volatility, or positive skewness. Alpha is fickle but dynamic risk management can be robust. Risk signals extracted from option market prices provide excellent forward-looking information that also separates upside risk from downside risk. These risk signals then guide allocation among broad asset classes. Adaptive diversification uses this crowd-sourced information in option prices, without trading in the options themselves, to diversify downside risk but maintain targeted exposure to upside risk. Compared to conventional static portfolios such as 60/40 and to volatility-managed portfolios such as risk parity, an adaptive diversification approach can offer better convexity, lower drawdown, and enhanced compound returns.

Dr. Philip Maymin, Director of Asset Allocation Strategies, Janus Henderson Investors

Philip Maymin is Director of Asset Allocation Strategies at Janus Henderson Investors, a position he has held since joining the firm in 2022. Working closely with the portfolio managers and researchers, he represents the firm’s asset allocation and adaptive strategies to clients and prospective clients. Prior to Janus Henderson, Philip was an insight partner and head of insight product at Essentia Analytics from 2019. He served as research partner and academic advisor from 2009 to 2018 at Gerstein Fisher. He was co-founder and portfolio manager at Maymin Capital Management LLC from 2005 to 2007. Before that, he was a portfolio manager at Ellington Management Group LLC from 2000. Philip began his career as a strategist at Long-Term Capital Management LP in 1996. His experience also includes serving as a professor at NYU Tandon, the University of Bridgeport, and Fairfield Dolan. Philip received a BA degree in computer science and an MS degree in applied mathematics from Harvard University. He earned a PhD in finance from the University of Chicago, Booth School of Business. He has 28 years of financial industry experience.

Worrying About Alpha

Quantitative investment strategies have emerged as an alternative to more traditional discretionary type of investing. At the very heart of every quantitative investment research lies a backtest, i.e. simulated past performance of the strategy. But what if your strategy does not perform in live trading the way it used to in simulation? I will talk about some (among many) things that one should worry about when devising quantitative strategies with a particular focus on overfitting and alpha decay and how to deal with those.

Dr. Adam Rej, Head of Macro Research, Capital Fund Management

Adam Rej is Head of Macro Research at Capital Fund Management.

Dr. Randeep Gug, CQF Institute, Managing Director

Dr. Randeep Gug is the Managing Director of the CQF Institute and a lecturer on the Certificate in Quantitative Finance (CQF). Prior to joining Fitch Learning, Randeep worked in a variety of roles. He spent five years working in the Equities division at Salomon Smith Barney and later traded futures and options on the Indian National Stock Exchange (NSE). More recently he has spent time teaching mathematics at all levels. He is a qualified teacher, holds a 1st class honours degree and a PhD for research in semiconductor physics. He is a CQF Alumnus, achieving a distinction on the programme and his current interests are based around improving and promoting the teaching and learning of Quant Finance.