Expanding Frontiers: An Alternative Investments & Machine Learning Podcast

Asset Pricing Revolution

This episode reviews an extensive systematic literature review titled "A Systematic Literature Review of Asset Pricing: Insights from AI and Big Data," authored by Zynobia Barson and colleagues from the University of Tasmania. This academic work analyzes 81 papers on AI and asset pricing, 53 on big data and asset pricing, and 24 on their combined use, employing both bibliometric and thematic analyses to map the evolution of the field. The central finding is that the integration of Artificial Intelligence (AI) and Big Data is fundamentally reshaping asset pricing by improving predictive accuracy, optimizing financial modeling, and enhancing risk management through the ability to handle complex, high-dimensional data. Specifically, the authors conclude that AI-based models are proving superior to traditional asset pricing frameworks by effectively addressing challenges like the "factor zoo" and capturing non-linear market dynamics. The paper also outlines future research directions, including exploring geographical gaps and addressing ethical considerations related to AI in finance.
 
References
Barson, Zynobia and Ahadzie, Richard Mawulawoe and Daugaard, Dan and Vespignani, Joaquin, A Systematic Literature Review of Asset Pricing: Insights from AI and Big Data (July 04, 2025). Barson, Zynobia; Ahadzie, Richard Mawulawoe; Daugaard, Daniel; Vespignani, Joaquin (2025). A Systematic Literature Review of Asset Pricing: Insights from AI  and Big Data. University of Tasmania. Preprint. https://hdl.handle.net/102.100.100/706792, Available at SSRN: https://ssrn.com/abstract=5351772 or http://dx.doi.org/10.2139/ssrn.5351772
 
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and potentially other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.
 

Virtual Land in the Metaverse: Real Estate Correlation and Portfolio Benefits

In this episode we explore the relationship between virtual land returns in the metaverse, specifically from the Decentraland platform, and the returns of physical real estate markets, approximated by equity REIT indices. Using wavelet coherence analysis on data from 2019 to 2023, the study we discuss empirically shows that the correlation between the two asset classes is generally low, suggesting potential diversification benefits for investors. However, this correlation spikes significantly during periods of acute economic turmoil such as the COVID-19 outbreak and interest rate shifts, indicating that virtual land's hedging effects may be limited during crises. Regression analysis identifies the consumer and economic climate, the price of the native cryptocurrency, and investor attention as the primary drivers of this dynamic correlation. Ultimately, the findings suggest that including virtual land can enhance risk-adjusted returns within a traditional asset portfolio, especially commercial real estate portfolios.
References
Leonhard, Heiko and Nagl, Maximilian and Schäfers, Wolfgang, Virtual land in the metaverse? Exploring the dynamic correlation with physical real estate (September 1, 2023). Available at SSRN: https://ssrn.com/abstract=4567859 or http://dx.doi.org/10.2139/ssrn.4567859
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and potentially other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

AI in Venture Capital

This episode investigates the adoption and impact of artificial intelligence (AI) within European venture capital (VC) firms, a topic previously under-researched despite AI's growing presence in finance. Based on survey data, the study we discuss reveals a significant increase in AI adoption since 2022, with screening emerging as its most common application. The research also identifies that VC firms with employees possessing strong ICT backgrounds are more likely to integrate AI. While AI has been shown to reduce due diligence time, its overall long-term benefits on VC operations remain largely inconclusive due to limited data, suggesting a need for more extensive future research.
References
Ronco, Umberto and Barontini, Roberto, Artificial Intelligence in Venture Capital Operations: An Empirical Analysis (February 15, 2025). Sant’Anna School of Advanced Studies, Institute of Management Research Paper Series_ No. 1 Winter 2025, Available at SSRN: https://ssrn.com/abstract=5164480 or http://dx.doi.org/10.2139/ssrn.5164480
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

Intelligent Internet: A Third Path for AI Development

This episode discusses a whitepaper that introduces the Intelligent Internet (II), a novel protocol designed to decentralize AI development and empower human agency. It proposes a "Third Path" by creating a "Bitcoin for the Intelligence Age," where Foundation Coins (FC) are minted only through Proof-of-Benefit (PoB), verifying societal good. The architecture includes three layers (Foundation, Culture, Personal) and is governed by principles such as Openness, Verifiable Public Benefit, and Human + Agent Dignity. The system aims to provide Universal AI (UAI) access to every individual via a sovereign II-Agent, with all knowledge anchored on auditable, open-licensed datasets through Anchor-Sets. The Intelligent Internet outlines a robust economic design, security model, and progressive governance structure, ensuring a transparent, auditable, and resilient public utility for the Intelligence Age.
 
References
Intelligent Internet Whitepaper July 24, 2025
by Emad Mostaque
https://webstatics.ii.inc/Intelligent-Internet-Whitepaper.pdf
 
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

Synthetic Data and Hedge Fund Replication

This episode explores and academic paper on the replication of hedge fund strategies using publicly available data and machine learning techniques, specifically autoencoders for dimension reduction and Generative Adversarial Networks (GANs) for synthesizing additional data. The author aims to demonstrate that such replicated portfolios can outperform traditional hedge fund returns after accounting for fees and transaction costs, thereby questioning the efficiency of current hedge fund performance. The research systematically evaluates different replication methodologies ultimately highlighting the superior performance and lower turnover achieved by the autoencoder-based strategies, especially when augmented with synthetically generated data. It presents a new way to benchmark hedge fund performance and potentially offers investors a more efficient alternative to direct hedge fund investment.
References
Shen, Kaiwen, Do You Really Need to Pay 2/20? Hedge Fund Strategy Replication via Machine Learning (October 10, 2022). Available at SSRN: https://ssrn.com/abstract=4243861 or http://dx.doi.org/10.2139/ssrn.4243861
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

Banks as Synthetic Hedge Funds

This episode explores how deposit-taking institutions, exemplified by Silicon Valley Bank (SVB), are transforming into "synthetic hedge funds". It examines SVB's hybrid business model, which combined on-balance-sheet "private equity-style banking" with off-balance-sheet "hedge fund-like trading strategies". The analysis highlights how SVB's reliance on "factor-based models" and "premature hedging exits" exposed it to significant interest rate and liquidity risks, ultimately leading to its collapse. The paper discussed argues that traditional regulatory frameworks are ill-equipped to address the complexities and systemic risks introduced by banks engaging in such "synthetic financial strategies," advocating for a reassessment of oversight to ensure financial stability in this evolving landscape.
Reference
Saeidinezhad, Elham, Banks as Synthetic Hedge Funds  (December 02, 2024). Available at SSRN: https://ssrn.com/abstract=5041554 or http://dx.doi.org/10.2139/ssrn.5041554
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

Hedge Fund Return Persistence and Performance

In this episode, we review two academic papers investigating various aspects of hedge fund performance and investment strategies. One relatively new study primarily examines how macroeconomic factors and the inclusion of hedge fund strategies impact portfolio performance for risk-averse investors, particularly focusing on out-of-sample predictability and risk-adjusted returns. It highlights that integrating hedge funds and considering macro-driven patterns can significantly enhance economic value, even though traditional measures like Sharpe ratios may not always reflect this fully due to higher-order moments like skewness and kurtosis. The other paper provides more background with a comprehensive survey of literature on hedge fund performance up until 2004, detailing various biases in hedge fund databases (e.g., survivorship, instant history, selection) and discussing different performance measurement methodologies, including traditional and adjusted Sharpe ratios, and multi-factor models that account for their unique non-linear exposures and time-varying risk profiles.
References
Magnani, Monia, Does Macroeconomic Predictability Enhance the Economic Value of Hedge Funds to Risk-Averse Investors?  (October 15, 2024). BAFFI Centre Research Paper No. 232, Available at SSRN: https://ssrn.com/abstract=4988114 or http://dx.doi.org/10.2139/ssrn.4988114
Géhin, Walter, A Survey of the Literature on Hedge Fund Performance (October 2004). Available at SSRN: https://ssrn.com/abstract=626441 or http://dx.doi.org/10.2139/ssrn.626441
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.
 

Homelessness in the US: Economic Impact and Solutions

This episode examines different facets of the housing market and its interconnectedness with broader economic factors. It reviews three recent SSRN papers. One source explores how contractionary monetary policy can lead to higher homeowners’ insurance prices, particularly for financially constrained insurers with interest-rate-sensitive assets, subsequently impacting home prices and mortgage applications. The other source investigates the effects of affordable housing developments (specifically, those financed by Low-Income Housing Tax Credits) on local rental markets, finding no evidence of increased rents in nearby market-rate apartments, and in some cases, even downward pressure on rents. Both highlight the complex interplay of financial mechanisms, policy interventions, and their observable effects on residential real estate.
References
Anguche, Scovia, Homelessness in the United States and its effects on the Economy (November 08, 2024). Available at SSRN: https://ssrn.com/abstract=5092765 or http://dx.doi.org/10.2139/ssrn.5092765
Damast, Dominik and Kubitza, Christian and Sørensen, Jakob Ahm, Homeowners Insurance and the Transmission of Monetary Policy (January 31, 2025). Available at SSRN: https://ssrn.com/abstract=5119139 or http://dx.doi.org/10.2139/ssrn.5119139
An, Brian and Fitzpatrick, Caleb and Jakabovics, Andrew and Orlando, Anthony W. and Rodnyansky, Seva and Voith, Richard and Zielenbach, Sean, The Effects of Affordable Housing Development on Local Rental Markets (April 03, 2025). Available at SSRN: https://ssrn.com/abstract=5204026 or http://dx.doi.org/10.2139/ssrn.5204026
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.

Quantum Machine Learning for Financial Services

This is the second Expanding Frontiers episode devoted to quantum computing for finance. It  explores the current (2024) state and future potential of Quantum Machine Learning (QML), specifically focusing on its applications within the financial services industry. We discuss various QML algorithms including Quantum Variational Classifiers, Quantum Kernel Estimation, and Quantum Neural Networks, and also touch upon quantum generative AI techniques like Quantum Transformers and Quantum Graph Neural Networks. The paper discussed identifies key financial applications for QML, such as risk management, credit scoring, fraud detection, and stock price prediction, while also outlining the promises and limitations of integrating QML into real-world financial operations. The review aims to serve as a practical guide for financial professionals and data scientists interested in understanding QML's relevance to their field.
 
References
 A Brief Review of Quantum Machine Learning for Financial Services (July 2024)
Mina Doosti, Petros Wallden, Conor Brian Hamill, Robert Hankache, Oliver Thomson Brown, Chris Heunen https://doi.org/10.48550/arXiv.2407.12618
 
Resources
Medium article: Are You Ready to Learn About Quantum Computing?
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the references listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.
 

Multimodal Financial Foundation Models - A Paper Review

This episode provides an overview of multimodal Financial Foundation Models (MFFMs), exploring their progress, potential applications, and associated challenges. It emphasizes the ubiquitous nature of multimodal financial data—including text, audio, images, and tabular information—in various financial applications like search, robo-advising, and trading. The paper review also addresses the development lifecycle of MFFMs, from pre-training to fine-tuning and alignment, while highlighting the need for robust benchmarks. Crucially, it discusses significant challenges such as data privacy, the risk of misinformation and hallucination, and the need for ethical AI readiness and governance within the financial sector.
 
References
Liu, Xiao-Yang and Cao, Yupeng and Deng, Li, Multimodal Financial Foundation Models (MFFMs): Progress, Prospects, and Challenges (May 31, 2025). Available at SSRN: https://ssrn.com/abstract=5277657 or http://dx.doi.org/10.2139/ssrn.5277657
 
Podcast Disclaimer
This podcast is an independent production and is not affiliated with or endorsed by any third-party entities unless explicitly stated. The content is for educational and informational purposes only and does not constitute financial, investment, legal, or professional advice. Listeners should consult qualified professionals before making any decisions based on this content.
This episode is based on the reference listed above and was generated using Notebook LM and other AI tools. While I have reviewed the content for accuracy, it may still contain errors, inaccuracies, or omissions. Neither the producers nor any affiliates accept liability for any damages or losses arising from the use or interpretation of this content.