Risk, Data and The Cold Facts
Dec 10, 2024
The Kickoff
Happy Tuesday to all the data-driven minds! Wrap up 2024 with a burst of insight from the Tradar — take a closer look at our exclusive chat with quants from Al Ramz, catch up on data trial inefficiencies from recent reports, and find your next favourite python library by Goldmans Sachs - all in this edition.
The Compass
Here's a rundown of what you can find in this edition:
Exciting updates from our team on all fronts
Exclusive insights from our interview with Giulio Occhionero, Michele Bogliardi and Zoubida Charif Khalifi from Al Ramz PJSC.
Eye-opening report from Neudata highlighting the need for better data trial efficiencies
Highlights from the latest global news after an eventful year
A deeper look at CVaR and EVaR
A Goldman Sachs Python Library
Something hilariously shareable for your Office Christmas Party
Insider Trading
This month at Quanted, we’ve made meaningful strides across multiple fronts, paving the way for a strong finish to the year. Internally, we’ve expanded our capabilities by onboarding new data partners, enabling us to cover a wider range of earnings reports and additional markets. We’ve also introduced task queuing on the platform, so users can now set up multiple tests to run overnight and evaluate the results first thing in the morning—helping quants maximise their productivity even further.
This month, we are also excited to welcome Austin Burkett to our team. From his strategic leadership at LSEG and Thomson Reuters to his board work with BattleFin, Austin has consistently demonstrated an innovative approach to data and technology. His track record of driving revenue growth, developing cutting-edge data services, and building strategic partnerships reflects exactly the kind of forward-thinking expertise we value and we are confident that his expertise will propel us into new realms of success.
Externally, we’re gaining strong traction with our partners, users, and fundraising efforts. Our angel investment round is ongoing and will close on December 20, with our VC round launching immediately after and continuing through the end of the financial year.
Completing the Tenity program's 13th batch was another significant milestone this month. Participating alongside a cohort of leading fintech and insurtech startups, pitching to over 200 industry professionals and investors, as well as participating in a competitive presentation hosted by LGT Bank were among our top highlights. These engagements provided new connections, insight into market trends, and opportunities for us to showcase our value proposition to a diverse and influential audience.
As we head into the new year, we’re energised by the momentum we’ve built and are excited for what lies ahead for 2025.
The Tradewinds
Expert Exchange
We recently sat down with three exceptional members of the quantitative team at Al Ramz PJSC, each bringing a unique blend of expertise and innovation to systematic trading and research.
Giulio Occhionero, Senior Vice President and Head of Quantitative Research and Development, leads the team in designing and deploying cutting-edge systematic strategies for proprietary trading and market making. With over 20 years in the financial industry, Giulio's background spans applied mathematics, stochastic processes, and computational finance. He has published novel research on forecasting equities using Boltzmann equations and applies this expertise through advanced tools and platforms like C#, Python, and Azure to drive innovation in trading automation and quant research.
Michele Bogliardi, AVP and Head of Quantitative Research, is a seasoned quant developer and algorithms innovator with over 20 years of experience. Michele specialises in crafting high-performing trading strategies, leveraging AI, machine learning, and deep expertise in delta-neutral hedging, high-frequency trading, and time series forecasting. His results speak for themselves, with Sharpe ratios above 2 and annualised PnL exceeding 30%. With a background in advanced physics and a career that spans roles in quantitative research and technical leadership, he combines rigorous analytics with practical implementation across global markets.
Zoubida Charif Khalifi, a Quantitative Researcher at Al Ramz PJSC, represents the next generation of talent in quantitative finance. A Financial Engineering graduate with experience in applying deep learning and reinforcement learning to options pricing and hedging, Zoubida brings a fresh perspective to systematic trading strategies. Her prior roles in equity research, data analysis, and algorithm development reflect her versatility and passion for financial innovation.
We unpacked the lessons from their career journeys and their thoughts on what lies ahead for the quant industry.
Reflecting on your careers in quant research, how has the industry evolved, and what has been the most memorable moment for you so far?
The evolution of quant research has brought new methods and tools but has also raised questions about the role of traditional mathematical approaches. Giulio notes, “The industry has shifted significantly into abandoning traditional science in favour of ML/AI tools, often without realising how the historical approach, based on calculus and deductions, yields formulas rather than mere statistical deductions bound to specific datasets.” His most memorable moment came when he derived equations that combined diffusion and concentration, unlocking a new perspective on stochastic processes.
Michele, reflecting on over two decades in the field, recalls the significance of the subprime mortgage crisis. “The breakout strategies I designed proved their resilience during heightened volatility,” he says. This experience inspired him to explore new methods, such as extending pair trading to multiple assets, taking advantage of expanded asset classes and advancements in technology.
Zoubida, earlier in her career, has found inspiration in bridging traditional mathematics and AI. “I co-authored a paper on integro-differential Boltzmann equations, which taught me the value of rigorous mathematics, but I’ve also seen how techniques like neural networks can address market frictions,” she says. The most exciting part of her journey has been exploring how these two approaches complement each other to solve complex problems.
What were the 3 key skills you honed during your career that elevated your performance most?
Across their experiences, coding emerges as a vital skill that each credits for their success. “Writing code for trading systems from scratch ensures flexibility and transparency,” Michele emphasises. “It allows you to tailor solutions to specific goals rather than relying on generic, off-the-shelf tools.” Zoubida adds that programming is a continuous journey: “Developing industry-standard code in compiled languages, like C#, bridges the gap between theory and practical implementation.”
Adaptability and curiosity are also essential. Michele explains, “The trading landscape has evolved significantly, requiring constant learning—whether it’s adopting new languages or embracing entirely new frameworks.” Zoubida agrees, emphasising the importance of critical thinking: “Quant research demands questioning how new insights fit into existing models and being open to building on them as the field evolves.”
Originality and patience are other standouts as critical attributes. “Creating unique strategies is what sets you apart,” Michele notes. “For example, leveraging synthetic data or targeting less-explored markets enhances the probability of uncovering inefficiencies.” Zoubida adds, “Meaningful results often require extended periods of research, testing, and refinement. Patience is key to navigating these long cycles.”
Could you provide a couple of technical insights or innovations in recent quant research that you believe would be particularly interesting to financial professionals?
The integration of traditional mathematical models with modern machine learning techniques has been a major focus. “There’s been a realization that ML and AI aren’t magic solutions,” Michele explains. “They need to be grounded in robust scientific models and domain-specific knowledge to be effective.” This shift, he notes, has led to hybrid systems that combine mathematical rigor with AI’s adaptability, creating trading strategies that are both interpretable and robust.
Zoubida shares a related innovation in derivatives pricing: “In our research, we developed a novel probability distribution function for options at maturity. This promises more accurate pricing and potential alpha opportunities while also deepening our understanding of risk measures like conditional Value-at-Risk.” Giulio adds his perspective on the need for a new approach in derivatives theory, saying, “It’s time to describe derivatives by probability densities that evolve over time rather than focusing solely on pricing.”
These insights reflect an industry-wide movement toward transparency and scientific grounding in quantitative finance.
What upcoming innovations in predictive analytics excite you the most, and how do you think they will impact the market?
Giulio points to non-linear periodicities as a promising development: “These periodicities, expressed through elliptic functions, could significantly enhance our ability to predict market cycles by capturing patterns beyond traditional approaches.”
Michele sees quantum computing as a transformative tool for handling complexity: “It allows us to process vast datasets and solve equations classical systems can’t handle, opening new possibilities for strategy development.” However, he notes that adoption may initially be limited to firms with the resources to explore this frontier.
Zoubida highlights explainable AI (XAI) as a key step forward: “By making models more transparent, we can address debugging challenges, meet compliance standards, and increase trust in AI-driven decisions, especially in areas like asset management and trading.”
Together, these advancements reflect a shift toward more precise, scalable, and transparent predictive systems in finance.
How do you see the reliance on external data sources for alpha generation changing the relationship between quants and their model development process?
“The problem of data is terribly underestimated in the whole industry,” Giulio notes, pointing out that poor-quality data can derail even the strongest strategies. His view underscores the importance of not just acquiring external data but ensuring its reliability.
Michele agrees on the opportunities and challenges external data presents. He observes that datasets like macroeconomic indicators or social media sentiment can reveal previously unseen inefficiencies, but warns, “The assumption that the introduction of new exogenous historical series will automatically lead to better trading systems should be made cum grano salis.” Without careful validation, the risk of spurious correlations misleading models is high, a lesson that history has repeatedly shown.
Zoubida adds that working with these new data streams has transformed the way quants collaborate with other experts. “Quants are increasingly collaborating with data scientists and domain experts to preprocess and interpret diverse data streams, such as ESG data and satellite imagery,” she explains. This collaborative shift has made the development process more dynamic, requiring strong data governance and advanced techniques to integrate these insights effectively.
As Michele notes, explainable AI (XAI) plays a pivotal role here, helping quants and their teams understand and trust the models’ use of external data. With these tools, quants can strike a balance between exploring new data sources and maintaining robust, actionable insights.
What is the next major project or initiative you’re working on, and how do you see it improving the domain?
Giulio is focused on exploring the probability densities of options throughout their lifespans. “Calls and puts aren’t normally distributed, and their true forms reveal fascinating expressions,” he explains. This work aims to provide deeper insights into derivatives beyond traditional pricing models.
Zoubida builds on this idea with her ongoing research into a novel probability distribution function (PDF) for options. “We’re looking at how this new PDF can generate alpha signals and improve risk management strategies for contingent claims. By understanding these instruments better, we can help traders manage their risk more effectively once positions are established,” she says.
Both aim to bring fresh perspectives to option pricing and hedging, offering traders tools grounded in a deeper understanding of the mathematics behind these instruments.
Do you have any last words of advice or insights for our audience?
Giulio shared a piece of timeless advice for young quants: "Focus on tough subjects to prepare for this world. You always have time for the easy stuff later."
Numbers & Narratives
Alternative Data Spending: A Story of Growth and Gaps
The alternative data market is growing, with 95% of buyers planning to maintain or increase budgets in 2025, according to Neudata's 2024 Future of Alternative Data Report. On average, firms spend $1.6 million annually across 20 datasets, but stark disparities exist: top spenders (over $5M annually) subscribe to 43 datasets, while smaller-budget buyers (<$250K) manage only 8 datasets.
The report notes that buyers are gravitating toward emerging markets, transactional data, and niche datasets. This reflects a pivot away from saturated data categories toward sources offering untapped potential. The strong adoption of AI-driven tools, with nearly half of vendors offering AI-trainable datasets, signals a maturing ecosystem where data isn’t just acquired but immediately integrated into scalable workflows.
While confidence in alternative data remains high, challenges persist:
Pricing: High costs remain the most cited barrier, disproportionately affecting smaller funds.
Low Trial Conversion Rates: Smaller-budget buyers report limited success converting trials into subscriptions, with 26% failing to adopt any trialed datasets.
Quality Gaps: Discrepancies between trialed datasets and their final implementation were noted by 47% of buyers, underscoring misaligned expectations.
These issues from the report suggests a pressing need for better trial processes and clearer value delivery. The frustration of testing countless datasets, only to uncover a handful worth pursuing, is all too familiar. If only there were a platform that streamlined data testing and pinpointed high-impact datasets for quants. Oh wait, there is. ;)
Market Pulse
Bitcoin Reserve Debate
The growing discourse on national Bitcoin reserves reveals a collision of innovation, geopolitics, and market volatility. The U.S. plan to allocate $76 billion to Bitcoin as a hedge against inflation positions the asset as a macroeconomic tool, but its volatility challenges traditional reserve principles, demanding sophisticated risk models to stabilise its role in portfolios. Russia’s proposal for a Bitcoin reserve underscores its geopolitical appeal, using crypto to bypass sanctions and stabilise trade, potentially setting a precedent for other sanctioned nations. Critics like Former US Treasury Secretary Larry Summers argue that Bitcoin’s speculative nature and lack of yield make it unsuitable for reserves, raising questions about sovereign credit risk and bond yields. These shifts introduce opportunities and risks: Bitcoin’s price may stabilise over time if institutional adoption increases, but its current volatility requires recalibration of models sensitive to supply-demand shocks and non-linear correlations. As national treasuries explore Bitcoin, its trajectory could redefine asset correlations, future liquidity assumptions, as well as hedging strategies, underscoring the evolving interplay between innovation and fiscal pragmatism.
What's Next for Stocks?
The 2025 stock market outlook reflects a complex interplay of risks and opportunities requiring adaptive strategies. The U.S. market, buoyed by AI-led structural changes and deregulation, could sustain growth, but elevated valuations expose it to potential sharp corrections, particularly as consumer spending momentum fades and labor market softness intensifies. Resilience may emerge in sectors driven by AI monetisation and private credit expansion, underscoring the shift from traditional financing models to long-duration asset investments. Additionally, the evolving role of Bitcoin as a diversifier, with low correlation to traditional markets, highlights opportunities in alternative assets that may offset equity volatility. Contrasting this optimism, Europe faces subdued equity prospects despite opportunities in bonds , while the UK stands out with undervalued sectors like housing benefiting from favourable macroeconomic shifts. For quants, these dynamics stress the need for models that capture the volatility inherent in geopolitical risks and market corrections, while leveraging growth drivers like AI infrastructure and regional economic divergence.
Navigational Nudges
In trading and portfolio management, it's the unexpected, not the routine, that tests our strategies. Conditional Value at Risk (CVaR) and Entropic Value at Risk (EVaR) help illuminate these blind spots by quantifying severe potential losses. Here’s a guide to making these tools work for you:
Tail Risks with CVaR
Quantify beyond typical limits.
→ Use CVaR to evaluate the average losses once the VaR threshold (e.g., 95% or 99%) is exceeded. For example, if VaR at 95% is $10 million, CVaR might reveal an average loss of $15 million, indicating more severe potential outcomes.Refine CVaR Adjustments for Non-Normal Markets
Adjust risk metrics to real-world data.
→ Use the Cornish-Fisher expansion to recalibrate CVaR, accounting for skewness and kurtosis. For a portfolio with significant asymmetric risks, these adjustments can shift the CVaR estimate substantially, affecting risk assessment and capital allocation decisions.Strategic Portfolio Optimisation Using CVaR
Balance risk with performance.
→ In portfolio optimisation, apply CVaR as a constraint to limit potential losses in adverse scenarios. For example, setting a CVaR limit at $12 million can help in structuring a portfolio that maximizes returns while controlling for worst-case losses.Implement EVaR for Robust Upper-Bound Estimation
Plan for the extreme.→ Utilise EVaR to define the upper loss limit within a 99.5% confidence interval, crucial for stress testing and regulatory compliance. For a high-volatility asset, EVaR might estimate a worst-case scenario loss at $20 million, compared to a CVaR of $15 million, offering a more conservative risk stance.
Continual Adaptation of Risk Models
The only constant in markets is change.
→ Regular updates to CVaR and EVaR models are necessary to reflect current market conditions. For instance, recalibrating the risk models quarterly with updated market data can reveal shifts in risk exposure, prompting timely adjustments in the fund’s risk strategy.
By following these steps, trading desks and investment management firms can refine their risk management frameworks, making them not only compliant with regulatory standards but also robust enough to handle the certainty of uncertainty in financial markets.
The Knowledge Buffet
🐍 GS-Quant Python Library 🐍
This Python toolkit from Goldman Sachs excels in building and testing sophisticated financial models, with strong capabilities in time-series forecasting, portfolio optimisation, and risk management. Its integration with both Goldman's proprietary data and external data sources allows for versatile strategy development and analysis. Whether you're pricing derivatives or automating trading strategies, GS-Quant offers robust tools, definitely worth testing to enhance your financial modeling.
The Closing Bell
If you're interested in learning more about how our tool can fit into your workflows schedule an intro call with our founders by clicking the button below, or reach out directly at sales@quanted.com to schedule a meeting.
Finance Fun Corner
A lawyer calls his wealthy art collector client with some news, "Paul, I have some good news and some bad news." Eager to hear about his investments in art, Paul says, "Give me the good news first!" The lawyer reveals, "Your wife spent $1,500 on two pictures that she believes will bring in $15-20 million." Overjoyed, Paul exclaims, "What could possibly be the bad news then?" The lawyer replies, "The pictures are of you and your secretary."
The Kickoff
Happy Tuesday to all the data-driven minds! Wrap up 2024 with a burst of insight from the Tradar — take a closer look at our exclusive chat with quants from Al Ramz, catch up on data trial inefficiencies from recent reports, and find your next favourite python library by Goldmans Sachs - all in this edition.
The Compass
Here's a rundown of what you can find in this edition:
Exciting updates from our team on all fronts
Exclusive insights from our interview with Giulio Occhionero, Michele Bogliardi and Zoubida Charif Khalifi from Al Ramz PJSC.
Eye-opening report from Neudata highlighting the need for better data trial efficiencies
Highlights from the latest global news after an eventful year
A deeper look at CVaR and EVaR
A Goldman Sachs Python Library
Something hilariously shareable for your Office Christmas Party
Insider Trading
This month at Quanted, we’ve made meaningful strides across multiple fronts, paving the way for a strong finish to the year. Internally, we’ve expanded our capabilities by onboarding new data partners, enabling us to cover a wider range of earnings reports and additional markets. We’ve also introduced task queuing on the platform, so users can now set up multiple tests to run overnight and evaluate the results first thing in the morning—helping quants maximise their productivity even further.
This month, we are also excited to welcome Austin Burkett to our team. From his strategic leadership at LSEG and Thomson Reuters to his board work with BattleFin, Austin has consistently demonstrated an innovative approach to data and technology. His track record of driving revenue growth, developing cutting-edge data services, and building strategic partnerships reflects exactly the kind of forward-thinking expertise we value and we are confident that his expertise will propel us into new realms of success.
Externally, we’re gaining strong traction with our partners, users, and fundraising efforts. Our angel investment round is ongoing and will close on December 20, with our VC round launching immediately after and continuing through the end of the financial year.
Completing the Tenity program's 13th batch was another significant milestone this month. Participating alongside a cohort of leading fintech and insurtech startups, pitching to over 200 industry professionals and investors, as well as participating in a competitive presentation hosted by LGT Bank were among our top highlights. These engagements provided new connections, insight into market trends, and opportunities for us to showcase our value proposition to a diverse and influential audience.
As we head into the new year, we’re energised by the momentum we’ve built and are excited for what lies ahead for 2025.
The Tradewinds
Expert Exchange
We recently sat down with three exceptional members of the quantitative team at Al Ramz PJSC, each bringing a unique blend of expertise and innovation to systematic trading and research.
Giulio Occhionero, Senior Vice President and Head of Quantitative Research and Development, leads the team in designing and deploying cutting-edge systematic strategies for proprietary trading and market making. With over 20 years in the financial industry, Giulio's background spans applied mathematics, stochastic processes, and computational finance. He has published novel research on forecasting equities using Boltzmann equations and applies this expertise through advanced tools and platforms like C#, Python, and Azure to drive innovation in trading automation and quant research.
Michele Bogliardi, AVP and Head of Quantitative Research, is a seasoned quant developer and algorithms innovator with over 20 years of experience. Michele specialises in crafting high-performing trading strategies, leveraging AI, machine learning, and deep expertise in delta-neutral hedging, high-frequency trading, and time series forecasting. His results speak for themselves, with Sharpe ratios above 2 and annualised PnL exceeding 30%. With a background in advanced physics and a career that spans roles in quantitative research and technical leadership, he combines rigorous analytics with practical implementation across global markets.
Zoubida Charif Khalifi, a Quantitative Researcher at Al Ramz PJSC, represents the next generation of talent in quantitative finance. A Financial Engineering graduate with experience in applying deep learning and reinforcement learning to options pricing and hedging, Zoubida brings a fresh perspective to systematic trading strategies. Her prior roles in equity research, data analysis, and algorithm development reflect her versatility and passion for financial innovation.
We unpacked the lessons from their career journeys and their thoughts on what lies ahead for the quant industry.
Reflecting on your careers in quant research, how has the industry evolved, and what has been the most memorable moment for you so far?
The evolution of quant research has brought new methods and tools but has also raised questions about the role of traditional mathematical approaches. Giulio notes, “The industry has shifted significantly into abandoning traditional science in favour of ML/AI tools, often without realising how the historical approach, based on calculus and deductions, yields formulas rather than mere statistical deductions bound to specific datasets.” His most memorable moment came when he derived equations that combined diffusion and concentration, unlocking a new perspective on stochastic processes.
Michele, reflecting on over two decades in the field, recalls the significance of the subprime mortgage crisis. “The breakout strategies I designed proved their resilience during heightened volatility,” he says. This experience inspired him to explore new methods, such as extending pair trading to multiple assets, taking advantage of expanded asset classes and advancements in technology.
Zoubida, earlier in her career, has found inspiration in bridging traditional mathematics and AI. “I co-authored a paper on integro-differential Boltzmann equations, which taught me the value of rigorous mathematics, but I’ve also seen how techniques like neural networks can address market frictions,” she says. The most exciting part of her journey has been exploring how these two approaches complement each other to solve complex problems.
What were the 3 key skills you honed during your career that elevated your performance most?
Across their experiences, coding emerges as a vital skill that each credits for their success. “Writing code for trading systems from scratch ensures flexibility and transparency,” Michele emphasises. “It allows you to tailor solutions to specific goals rather than relying on generic, off-the-shelf tools.” Zoubida adds that programming is a continuous journey: “Developing industry-standard code in compiled languages, like C#, bridges the gap between theory and practical implementation.”
Adaptability and curiosity are also essential. Michele explains, “The trading landscape has evolved significantly, requiring constant learning—whether it’s adopting new languages or embracing entirely new frameworks.” Zoubida agrees, emphasising the importance of critical thinking: “Quant research demands questioning how new insights fit into existing models and being open to building on them as the field evolves.”
Originality and patience are other standouts as critical attributes. “Creating unique strategies is what sets you apart,” Michele notes. “For example, leveraging synthetic data or targeting less-explored markets enhances the probability of uncovering inefficiencies.” Zoubida adds, “Meaningful results often require extended periods of research, testing, and refinement. Patience is key to navigating these long cycles.”
Could you provide a couple of technical insights or innovations in recent quant research that you believe would be particularly interesting to financial professionals?
The integration of traditional mathematical models with modern machine learning techniques has been a major focus. “There’s been a realization that ML and AI aren’t magic solutions,” Michele explains. “They need to be grounded in robust scientific models and domain-specific knowledge to be effective.” This shift, he notes, has led to hybrid systems that combine mathematical rigor with AI’s adaptability, creating trading strategies that are both interpretable and robust.
Zoubida shares a related innovation in derivatives pricing: “In our research, we developed a novel probability distribution function for options at maturity. This promises more accurate pricing and potential alpha opportunities while also deepening our understanding of risk measures like conditional Value-at-Risk.” Giulio adds his perspective on the need for a new approach in derivatives theory, saying, “It’s time to describe derivatives by probability densities that evolve over time rather than focusing solely on pricing.”
These insights reflect an industry-wide movement toward transparency and scientific grounding in quantitative finance.
What upcoming innovations in predictive analytics excite you the most, and how do you think they will impact the market?
Giulio points to non-linear periodicities as a promising development: “These periodicities, expressed through elliptic functions, could significantly enhance our ability to predict market cycles by capturing patterns beyond traditional approaches.”
Michele sees quantum computing as a transformative tool for handling complexity: “It allows us to process vast datasets and solve equations classical systems can’t handle, opening new possibilities for strategy development.” However, he notes that adoption may initially be limited to firms with the resources to explore this frontier.
Zoubida highlights explainable AI (XAI) as a key step forward: “By making models more transparent, we can address debugging challenges, meet compliance standards, and increase trust in AI-driven decisions, especially in areas like asset management and trading.”
Together, these advancements reflect a shift toward more precise, scalable, and transparent predictive systems in finance.
How do you see the reliance on external data sources for alpha generation changing the relationship between quants and their model development process?
“The problem of data is terribly underestimated in the whole industry,” Giulio notes, pointing out that poor-quality data can derail even the strongest strategies. His view underscores the importance of not just acquiring external data but ensuring its reliability.
Michele agrees on the opportunities and challenges external data presents. He observes that datasets like macroeconomic indicators or social media sentiment can reveal previously unseen inefficiencies, but warns, “The assumption that the introduction of new exogenous historical series will automatically lead to better trading systems should be made cum grano salis.” Without careful validation, the risk of spurious correlations misleading models is high, a lesson that history has repeatedly shown.
Zoubida adds that working with these new data streams has transformed the way quants collaborate with other experts. “Quants are increasingly collaborating with data scientists and domain experts to preprocess and interpret diverse data streams, such as ESG data and satellite imagery,” she explains. This collaborative shift has made the development process more dynamic, requiring strong data governance and advanced techniques to integrate these insights effectively.
As Michele notes, explainable AI (XAI) plays a pivotal role here, helping quants and their teams understand and trust the models’ use of external data. With these tools, quants can strike a balance between exploring new data sources and maintaining robust, actionable insights.
What is the next major project or initiative you’re working on, and how do you see it improving the domain?
Giulio is focused on exploring the probability densities of options throughout their lifespans. “Calls and puts aren’t normally distributed, and their true forms reveal fascinating expressions,” he explains. This work aims to provide deeper insights into derivatives beyond traditional pricing models.
Zoubida builds on this idea with her ongoing research into a novel probability distribution function (PDF) for options. “We’re looking at how this new PDF can generate alpha signals and improve risk management strategies for contingent claims. By understanding these instruments better, we can help traders manage their risk more effectively once positions are established,” she says.
Both aim to bring fresh perspectives to option pricing and hedging, offering traders tools grounded in a deeper understanding of the mathematics behind these instruments.
Do you have any last words of advice or insights for our audience?
Giulio shared a piece of timeless advice for young quants: "Focus on tough subjects to prepare for this world. You always have time for the easy stuff later."
Numbers & Narratives
Alternative Data Spending: A Story of Growth and Gaps
The alternative data market is growing, with 95% of buyers planning to maintain or increase budgets in 2025, according to Neudata's 2024 Future of Alternative Data Report. On average, firms spend $1.6 million annually across 20 datasets, but stark disparities exist: top spenders (over $5M annually) subscribe to 43 datasets, while smaller-budget buyers (<$250K) manage only 8 datasets.
The report notes that buyers are gravitating toward emerging markets, transactional data, and niche datasets. This reflects a pivot away from saturated data categories toward sources offering untapped potential. The strong adoption of AI-driven tools, with nearly half of vendors offering AI-trainable datasets, signals a maturing ecosystem where data isn’t just acquired but immediately integrated into scalable workflows.
While confidence in alternative data remains high, challenges persist:
Pricing: High costs remain the most cited barrier, disproportionately affecting smaller funds.
Low Trial Conversion Rates: Smaller-budget buyers report limited success converting trials into subscriptions, with 26% failing to adopt any trialed datasets.
Quality Gaps: Discrepancies between trialed datasets and their final implementation were noted by 47% of buyers, underscoring misaligned expectations.
These issues from the report suggests a pressing need for better trial processes and clearer value delivery. The frustration of testing countless datasets, only to uncover a handful worth pursuing, is all too familiar. If only there were a platform that streamlined data testing and pinpointed high-impact datasets for quants. Oh wait, there is. ;)
Market Pulse
Bitcoin Reserve Debate
The growing discourse on national Bitcoin reserves reveals a collision of innovation, geopolitics, and market volatility. The U.S. plan to allocate $76 billion to Bitcoin as a hedge against inflation positions the asset as a macroeconomic tool, but its volatility challenges traditional reserve principles, demanding sophisticated risk models to stabilise its role in portfolios. Russia’s proposal for a Bitcoin reserve underscores its geopolitical appeal, using crypto to bypass sanctions and stabilise trade, potentially setting a precedent for other sanctioned nations. Critics like Former US Treasury Secretary Larry Summers argue that Bitcoin’s speculative nature and lack of yield make it unsuitable for reserves, raising questions about sovereign credit risk and bond yields. These shifts introduce opportunities and risks: Bitcoin’s price may stabilise over time if institutional adoption increases, but its current volatility requires recalibration of models sensitive to supply-demand shocks and non-linear correlations. As national treasuries explore Bitcoin, its trajectory could redefine asset correlations, future liquidity assumptions, as well as hedging strategies, underscoring the evolving interplay between innovation and fiscal pragmatism.
What's Next for Stocks?
The 2025 stock market outlook reflects a complex interplay of risks and opportunities requiring adaptive strategies. The U.S. market, buoyed by AI-led structural changes and deregulation, could sustain growth, but elevated valuations expose it to potential sharp corrections, particularly as consumer spending momentum fades and labor market softness intensifies. Resilience may emerge in sectors driven by AI monetisation and private credit expansion, underscoring the shift from traditional financing models to long-duration asset investments. Additionally, the evolving role of Bitcoin as a diversifier, with low correlation to traditional markets, highlights opportunities in alternative assets that may offset equity volatility. Contrasting this optimism, Europe faces subdued equity prospects despite opportunities in bonds , while the UK stands out with undervalued sectors like housing benefiting from favourable macroeconomic shifts. For quants, these dynamics stress the need for models that capture the volatility inherent in geopolitical risks and market corrections, while leveraging growth drivers like AI infrastructure and regional economic divergence.
Navigational Nudges
In trading and portfolio management, it's the unexpected, not the routine, that tests our strategies. Conditional Value at Risk (CVaR) and Entropic Value at Risk (EVaR) help illuminate these blind spots by quantifying severe potential losses. Here’s a guide to making these tools work for you:
Tail Risks with CVaR
Quantify beyond typical limits.
→ Use CVaR to evaluate the average losses once the VaR threshold (e.g., 95% or 99%) is exceeded. For example, if VaR at 95% is $10 million, CVaR might reveal an average loss of $15 million, indicating more severe potential outcomes.Refine CVaR Adjustments for Non-Normal Markets
Adjust risk metrics to real-world data.
→ Use the Cornish-Fisher expansion to recalibrate CVaR, accounting for skewness and kurtosis. For a portfolio with significant asymmetric risks, these adjustments can shift the CVaR estimate substantially, affecting risk assessment and capital allocation decisions.Strategic Portfolio Optimisation Using CVaR
Balance risk with performance.
→ In portfolio optimisation, apply CVaR as a constraint to limit potential losses in adverse scenarios. For example, setting a CVaR limit at $12 million can help in structuring a portfolio that maximizes returns while controlling for worst-case losses.Implement EVaR for Robust Upper-Bound Estimation
Plan for the extreme.→ Utilise EVaR to define the upper loss limit within a 99.5% confidence interval, crucial for stress testing and regulatory compliance. For a high-volatility asset, EVaR might estimate a worst-case scenario loss at $20 million, compared to a CVaR of $15 million, offering a more conservative risk stance.
Continual Adaptation of Risk Models
The only constant in markets is change.
→ Regular updates to CVaR and EVaR models are necessary to reflect current market conditions. For instance, recalibrating the risk models quarterly with updated market data can reveal shifts in risk exposure, prompting timely adjustments in the fund’s risk strategy.
By following these steps, trading desks and investment management firms can refine their risk management frameworks, making them not only compliant with regulatory standards but also robust enough to handle the certainty of uncertainty in financial markets.
The Knowledge Buffet
🐍 GS-Quant Python Library 🐍
This Python toolkit from Goldman Sachs excels in building and testing sophisticated financial models, with strong capabilities in time-series forecasting, portfolio optimisation, and risk management. Its integration with both Goldman's proprietary data and external data sources allows for versatile strategy development and analysis. Whether you're pricing derivatives or automating trading strategies, GS-Quant offers robust tools, definitely worth testing to enhance your financial modeling.
The Closing Bell
If you're interested in learning more about how our tool can fit into your workflows schedule an intro call with our founders by clicking the button below, or reach out directly at sales@quanted.com to schedule a meeting.
Finance Fun Corner
A lawyer calls his wealthy art collector client with some news, "Paul, I have some good news and some bad news." Eager to hear about his investments in art, Paul says, "Give me the good news first!" The lawyer reveals, "Your wife spent $1,500 on two pictures that she believes will bring in $15-20 million." Overjoyed, Paul exclaims, "What could possibly be the bad news then?" The lawyer replies, "The pictures are of you and your secretary."
The Kickoff
Happy Tuesday to all the data-driven minds! Wrap up 2024 with a burst of insight from the Tradar — take a closer look at our exclusive chat with quants from Al Ramz, catch up on data trial inefficiencies from recent reports, and find your next favourite python library by Goldmans Sachs - all in this edition.
The Compass
Here's a rundown of what you can find in this edition:
Exciting updates from our team on all fronts
Exclusive insights from our interview with Giulio Occhionero, Michele Bogliardi and Zoubida Charif Khalifi from Al Ramz PJSC.
Eye-opening report from Neudata highlighting the need for better data trial efficiencies
Highlights from the latest global news after an eventful year
A deeper look at CVaR and EVaR
A Goldman Sachs Python Library
Something hilariously shareable for your Office Christmas Party
Insider Trading
This month at Quanted, we’ve made meaningful strides across multiple fronts, paving the way for a strong finish to the year. Internally, we’ve expanded our capabilities by onboarding new data partners, enabling us to cover a wider range of earnings reports and additional markets. We’ve also introduced task queuing on the platform, so users can now set up multiple tests to run overnight and evaluate the results first thing in the morning—helping quants maximise their productivity even further.
This month, we are also excited to welcome Austin Burkett to our team. From his strategic leadership at LSEG and Thomson Reuters to his board work with BattleFin, Austin has consistently demonstrated an innovative approach to data and technology. His track record of driving revenue growth, developing cutting-edge data services, and building strategic partnerships reflects exactly the kind of forward-thinking expertise we value and we are confident that his expertise will propel us into new realms of success.
Externally, we’re gaining strong traction with our partners, users, and fundraising efforts. Our angel investment round is ongoing and will close on December 20, with our VC round launching immediately after and continuing through the end of the financial year.
Completing the Tenity program's 13th batch was another significant milestone this month. Participating alongside a cohort of leading fintech and insurtech startups, pitching to over 200 industry professionals and investors, as well as participating in a competitive presentation hosted by LGT Bank were among our top highlights. These engagements provided new connections, insight into market trends, and opportunities for us to showcase our value proposition to a diverse and influential audience.
As we head into the new year, we’re energised by the momentum we’ve built and are excited for what lies ahead for 2025.
The Tradewinds
Expert Exchange
We recently sat down with three exceptional members of the quantitative team at Al Ramz PJSC, each bringing a unique blend of expertise and innovation to systematic trading and research.
Giulio Occhionero, Senior Vice President and Head of Quantitative Research and Development, leads the team in designing and deploying cutting-edge systematic strategies for proprietary trading and market making. With over 20 years in the financial industry, Giulio's background spans applied mathematics, stochastic processes, and computational finance. He has published novel research on forecasting equities using Boltzmann equations and applies this expertise through advanced tools and platforms like C#, Python, and Azure to drive innovation in trading automation and quant research.
Michele Bogliardi, AVP and Head of Quantitative Research, is a seasoned quant developer and algorithms innovator with over 20 years of experience. Michele specialises in crafting high-performing trading strategies, leveraging AI, machine learning, and deep expertise in delta-neutral hedging, high-frequency trading, and time series forecasting. His results speak for themselves, with Sharpe ratios above 2 and annualised PnL exceeding 30%. With a background in advanced physics and a career that spans roles in quantitative research and technical leadership, he combines rigorous analytics with practical implementation across global markets.
Zoubida Charif Khalifi, a Quantitative Researcher at Al Ramz PJSC, represents the next generation of talent in quantitative finance. A Financial Engineering graduate with experience in applying deep learning and reinforcement learning to options pricing and hedging, Zoubida brings a fresh perspective to systematic trading strategies. Her prior roles in equity research, data analysis, and algorithm development reflect her versatility and passion for financial innovation.
We unpacked the lessons from their career journeys and their thoughts on what lies ahead for the quant industry.
Reflecting on your careers in quant research, how has the industry evolved, and what has been the most memorable moment for you so far?
The evolution of quant research has brought new methods and tools but has also raised questions about the role of traditional mathematical approaches. Giulio notes, “The industry has shifted significantly into abandoning traditional science in favour of ML/AI tools, often without realising how the historical approach, based on calculus and deductions, yields formulas rather than mere statistical deductions bound to specific datasets.” His most memorable moment came when he derived equations that combined diffusion and concentration, unlocking a new perspective on stochastic processes.
Michele, reflecting on over two decades in the field, recalls the significance of the subprime mortgage crisis. “The breakout strategies I designed proved their resilience during heightened volatility,” he says. This experience inspired him to explore new methods, such as extending pair trading to multiple assets, taking advantage of expanded asset classes and advancements in technology.
Zoubida, earlier in her career, has found inspiration in bridging traditional mathematics and AI. “I co-authored a paper on integro-differential Boltzmann equations, which taught me the value of rigorous mathematics, but I’ve also seen how techniques like neural networks can address market frictions,” she says. The most exciting part of her journey has been exploring how these two approaches complement each other to solve complex problems.
What were the 3 key skills you honed during your career that elevated your performance most?
Across their experiences, coding emerges as a vital skill that each credits for their success. “Writing code for trading systems from scratch ensures flexibility and transparency,” Michele emphasises. “It allows you to tailor solutions to specific goals rather than relying on generic, off-the-shelf tools.” Zoubida adds that programming is a continuous journey: “Developing industry-standard code in compiled languages, like C#, bridges the gap between theory and practical implementation.”
Adaptability and curiosity are also essential. Michele explains, “The trading landscape has evolved significantly, requiring constant learning—whether it’s adopting new languages or embracing entirely new frameworks.” Zoubida agrees, emphasising the importance of critical thinking: “Quant research demands questioning how new insights fit into existing models and being open to building on them as the field evolves.”
Originality and patience are other standouts as critical attributes. “Creating unique strategies is what sets you apart,” Michele notes. “For example, leveraging synthetic data or targeting less-explored markets enhances the probability of uncovering inefficiencies.” Zoubida adds, “Meaningful results often require extended periods of research, testing, and refinement. Patience is key to navigating these long cycles.”
Could you provide a couple of technical insights or innovations in recent quant research that you believe would be particularly interesting to financial professionals?
The integration of traditional mathematical models with modern machine learning techniques has been a major focus. “There’s been a realization that ML and AI aren’t magic solutions,” Michele explains. “They need to be grounded in robust scientific models and domain-specific knowledge to be effective.” This shift, he notes, has led to hybrid systems that combine mathematical rigor with AI’s adaptability, creating trading strategies that are both interpretable and robust.
Zoubida shares a related innovation in derivatives pricing: “In our research, we developed a novel probability distribution function for options at maturity. This promises more accurate pricing and potential alpha opportunities while also deepening our understanding of risk measures like conditional Value-at-Risk.” Giulio adds his perspective on the need for a new approach in derivatives theory, saying, “It’s time to describe derivatives by probability densities that evolve over time rather than focusing solely on pricing.”
These insights reflect an industry-wide movement toward transparency and scientific grounding in quantitative finance.
What upcoming innovations in predictive analytics excite you the most, and how do you think they will impact the market?
Giulio points to non-linear periodicities as a promising development: “These periodicities, expressed through elliptic functions, could significantly enhance our ability to predict market cycles by capturing patterns beyond traditional approaches.”
Michele sees quantum computing as a transformative tool for handling complexity: “It allows us to process vast datasets and solve equations classical systems can’t handle, opening new possibilities for strategy development.” However, he notes that adoption may initially be limited to firms with the resources to explore this frontier.
Zoubida highlights explainable AI (XAI) as a key step forward: “By making models more transparent, we can address debugging challenges, meet compliance standards, and increase trust in AI-driven decisions, especially in areas like asset management and trading.”
Together, these advancements reflect a shift toward more precise, scalable, and transparent predictive systems in finance.
How do you see the reliance on external data sources for alpha generation changing the relationship between quants and their model development process?
“The problem of data is terribly underestimated in the whole industry,” Giulio notes, pointing out that poor-quality data can derail even the strongest strategies. His view underscores the importance of not just acquiring external data but ensuring its reliability.
Michele agrees on the opportunities and challenges external data presents. He observes that datasets like macroeconomic indicators or social media sentiment can reveal previously unseen inefficiencies, but warns, “The assumption that the introduction of new exogenous historical series will automatically lead to better trading systems should be made cum grano salis.” Without careful validation, the risk of spurious correlations misleading models is high, a lesson that history has repeatedly shown.
Zoubida adds that working with these new data streams has transformed the way quants collaborate with other experts. “Quants are increasingly collaborating with data scientists and domain experts to preprocess and interpret diverse data streams, such as ESG data and satellite imagery,” she explains. This collaborative shift has made the development process more dynamic, requiring strong data governance and advanced techniques to integrate these insights effectively.
As Michele notes, explainable AI (XAI) plays a pivotal role here, helping quants and their teams understand and trust the models’ use of external data. With these tools, quants can strike a balance between exploring new data sources and maintaining robust, actionable insights.
What is the next major project or initiative you’re working on, and how do you see it improving the domain?
Giulio is focused on exploring the probability densities of options throughout their lifespans. “Calls and puts aren’t normally distributed, and their true forms reveal fascinating expressions,” he explains. This work aims to provide deeper insights into derivatives beyond traditional pricing models.
Zoubida builds on this idea with her ongoing research into a novel probability distribution function (PDF) for options. “We’re looking at how this new PDF can generate alpha signals and improve risk management strategies for contingent claims. By understanding these instruments better, we can help traders manage their risk more effectively once positions are established,” she says.
Both aim to bring fresh perspectives to option pricing and hedging, offering traders tools grounded in a deeper understanding of the mathematics behind these instruments.
Do you have any last words of advice or insights for our audience?
Giulio shared a piece of timeless advice for young quants: "Focus on tough subjects to prepare for this world. You always have time for the easy stuff later."
Numbers & Narratives
Alternative Data Spending: A Story of Growth and Gaps
The alternative data market is growing, with 95% of buyers planning to maintain or increase budgets in 2025, according to Neudata's 2024 Future of Alternative Data Report. On average, firms spend $1.6 million annually across 20 datasets, but stark disparities exist: top spenders (over $5M annually) subscribe to 43 datasets, while smaller-budget buyers (<$250K) manage only 8 datasets.
The report notes that buyers are gravitating toward emerging markets, transactional data, and niche datasets. This reflects a pivot away from saturated data categories toward sources offering untapped potential. The strong adoption of AI-driven tools, with nearly half of vendors offering AI-trainable datasets, signals a maturing ecosystem where data isn’t just acquired but immediately integrated into scalable workflows.
While confidence in alternative data remains high, challenges persist:
Pricing: High costs remain the most cited barrier, disproportionately affecting smaller funds.
Low Trial Conversion Rates: Smaller-budget buyers report limited success converting trials into subscriptions, with 26% failing to adopt any trialed datasets.
Quality Gaps: Discrepancies between trialed datasets and their final implementation were noted by 47% of buyers, underscoring misaligned expectations.
These issues from the report suggests a pressing need for better trial processes and clearer value delivery. The frustration of testing countless datasets, only to uncover a handful worth pursuing, is all too familiar. If only there were a platform that streamlined data testing and pinpointed high-impact datasets for quants. Oh wait, there is. ;)
Market Pulse
Bitcoin Reserve Debate
The growing discourse on national Bitcoin reserves reveals a collision of innovation, geopolitics, and market volatility. The U.S. plan to allocate $76 billion to Bitcoin as a hedge against inflation positions the asset as a macroeconomic tool, but its volatility challenges traditional reserve principles, demanding sophisticated risk models to stabilise its role in portfolios. Russia’s proposal for a Bitcoin reserve underscores its geopolitical appeal, using crypto to bypass sanctions and stabilise trade, potentially setting a precedent for other sanctioned nations. Critics like Former US Treasury Secretary Larry Summers argue that Bitcoin’s speculative nature and lack of yield make it unsuitable for reserves, raising questions about sovereign credit risk and bond yields. These shifts introduce opportunities and risks: Bitcoin’s price may stabilise over time if institutional adoption increases, but its current volatility requires recalibration of models sensitive to supply-demand shocks and non-linear correlations. As national treasuries explore Bitcoin, its trajectory could redefine asset correlations, future liquidity assumptions, as well as hedging strategies, underscoring the evolving interplay between innovation and fiscal pragmatism.
What's Next for Stocks?
The 2025 stock market outlook reflects a complex interplay of risks and opportunities requiring adaptive strategies. The U.S. market, buoyed by AI-led structural changes and deregulation, could sustain growth, but elevated valuations expose it to potential sharp corrections, particularly as consumer spending momentum fades and labor market softness intensifies. Resilience may emerge in sectors driven by AI monetisation and private credit expansion, underscoring the shift from traditional financing models to long-duration asset investments. Additionally, the evolving role of Bitcoin as a diversifier, with low correlation to traditional markets, highlights opportunities in alternative assets that may offset equity volatility. Contrasting this optimism, Europe faces subdued equity prospects despite opportunities in bonds , while the UK stands out with undervalued sectors like housing benefiting from favourable macroeconomic shifts. For quants, these dynamics stress the need for models that capture the volatility inherent in geopolitical risks and market corrections, while leveraging growth drivers like AI infrastructure and regional economic divergence.
Navigational Nudges
In trading and portfolio management, it's the unexpected, not the routine, that tests our strategies. Conditional Value at Risk (CVaR) and Entropic Value at Risk (EVaR) help illuminate these blind spots by quantifying severe potential losses. Here’s a guide to making these tools work for you:
Tail Risks with CVaR
Quantify beyond typical limits.
→ Use CVaR to evaluate the average losses once the VaR threshold (e.g., 95% or 99%) is exceeded. For example, if VaR at 95% is $10 million, CVaR might reveal an average loss of $15 million, indicating more severe potential outcomes.Refine CVaR Adjustments for Non-Normal Markets
Adjust risk metrics to real-world data.
→ Use the Cornish-Fisher expansion to recalibrate CVaR, accounting for skewness and kurtosis. For a portfolio with significant asymmetric risks, these adjustments can shift the CVaR estimate substantially, affecting risk assessment and capital allocation decisions.Strategic Portfolio Optimisation Using CVaR
Balance risk with performance.
→ In portfolio optimisation, apply CVaR as a constraint to limit potential losses in adverse scenarios. For example, setting a CVaR limit at $12 million can help in structuring a portfolio that maximizes returns while controlling for worst-case losses.Implement EVaR for Robust Upper-Bound Estimation
Plan for the extreme.→ Utilise EVaR to define the upper loss limit within a 99.5% confidence interval, crucial for stress testing and regulatory compliance. For a high-volatility asset, EVaR might estimate a worst-case scenario loss at $20 million, compared to a CVaR of $15 million, offering a more conservative risk stance.
Continual Adaptation of Risk Models
The only constant in markets is change.
→ Regular updates to CVaR and EVaR models are necessary to reflect current market conditions. For instance, recalibrating the risk models quarterly with updated market data can reveal shifts in risk exposure, prompting timely adjustments in the fund’s risk strategy.
By following these steps, trading desks and investment management firms can refine their risk management frameworks, making them not only compliant with regulatory standards but also robust enough to handle the certainty of uncertainty in financial markets.
The Knowledge Buffet
🐍 GS-Quant Python Library 🐍
This Python toolkit from Goldman Sachs excels in building and testing sophisticated financial models, with strong capabilities in time-series forecasting, portfolio optimisation, and risk management. Its integration with both Goldman's proprietary data and external data sources allows for versatile strategy development and analysis. Whether you're pricing derivatives or automating trading strategies, GS-Quant offers robust tools, definitely worth testing to enhance your financial modeling.
The Closing Bell
If you're interested in learning more about how our tool can fit into your workflows schedule an intro call with our founders by clicking the button below, or reach out directly at sales@quanted.com to schedule a meeting.
Finance Fun Corner
A lawyer calls his wealthy art collector client with some news, "Paul, I have some good news and some bad news." Eager to hear about his investments in art, Paul says, "Give me the good news first!" The lawyer reveals, "Your wife spent $1,500 on two pictures that she believes will bring in $15-20 million." Overjoyed, Paul exclaims, "What could possibly be the bad news then?" The lawyer replies, "The pictures are of you and your secretary."
