Odds on Open

Former Goldman Sachs head of natural gas trading John Knorring delivers a lecture on commodities trading from first principles, adapted from the one he gives to Princeton financial engineering students. Starting from one equation, supply minus demand equals the change in inventory, he explains how storage constraints drive price, why crude oil went negative in 2020, and why US natural gas is so volatile. Topics include Henry Hub futures, physical delivery and arbitrage, basis risk, the storability of coal, oil, gas, and power, and how weather shocks can double nat gas prices in a week. Essential viewing for energy traders and quants.Knorring then covers the shale revolution and Aubrey McClendon, how low prices built demand through power burn, LNG exports, and AI data centers, and the Goldman trading framework: idea generation, trade expression, execution, and money management. He walks through his bullish silver trade, from the structural deficit and gold-silver ratio to buying futures and cheap upside calls, and what he got wrong at $120. The episode closes with market trends, including position sizing lessons from Amaranth, the solar duck curve, the US LNG export wave, and why physical gas traders will outlast automated derivatives desks, plus timeless career advice for Wall Street.

David Orr returns to Odds on Open to unpack how he sustains a 50% annualized return and a 2.1 Sharpe ratio by stripping the multi-manager model down to raw alpha generation. We deconstruct his firm’s AI-driven risk architecture, detailing how proprietary factor models recalculate custom basket correlations in real-time to manage tail risk across a 400% levered portfolio. David also breaks down his personal discretionary process, explaining his strict two-hour cap on single-name equity research to eliminate sunk-cost bias, the fatal mechanics of shorting revenue growth, and why institutional pod shops structurally sacrifice top-tier returns to optimize for AUM fee harvesting.Beyond portfolio construction, we explore the unit economics of scaling a high-performance fund through unconventional talent sourcing. David explains how he evaluates pure live track records to recruit non-consensus PMs, utilizing the fund’s long/short vehicle to offer massive tax deferral advantages for incoming high-frequency traders. We also examine his structural framework for the AI trade—specifically why hyperscalers offer superior downside protection and take-or-pay ROIC compared to hardware manufacturers—and his thesis on why unplugging from consensus narrative feeds is required to isolate variant views in shifting market regimes.

Geoffrey Williams spent more than a decade at Goldman Sachs, where he traded structured products through the 2008 financial crisis and later co-ran the CMBS trading business. In this episode of Odds on Open, he takes Ethan Kho inside the Goldman mortgage desk as Bear Stearns and Lehman collapsed. He explains how banks sold tens of billions of super senior protection on ABS CDOs for 10 basis points a year, and what it was like to trade TABX and ABS CDOs as AAA paper repriced from basis points to points upfront. He also breaks down how conduit CMBS is structured and tranched, why commercial real estate moves in cycles across malls, office, and life science, and how post-GFC capital rules pushed securitized credit risk off dealer balance sheets and onto hedge funds.Geoffrey now runs 1543 Capital LP, an opportunistic structured finance hedge fund, with a co-CIO who came from Brevan Howard. The fund focuses on end-of-life CMBS, residential solar ABS, and other dislocated credit. He walks through how he underwrites a CMBS bond loan by loan, when desktop tools like Google Earth are enough and when you need to see the property, and how he sizes positions to survive unknown unknowns while targeting mid-teens unlevered returns. The conversation also covers hedging macro tail risk from deficits, rates, and AI disruption, how populist politics and the proposed California wealth tax shape his real estate exposure, luck versus skill in finance careers, thinking in bets as an entrepreneur, and the biggest mistake structured credit PMs make: forgetting that credit markets move in step functions.

Checkout Flux 4.0 here: https://www.flux.live/flux4/index.html Chris Camillo and Tom Costello go head-to-head on Odds on Open's first every debate episode. They answer a question every trader eventually confronts: who is actually better positioned to take risk — a concentrated retail investor with no LPs, mandates, or career risk, or an institutional hedge fund manager with a two-and-twenty product to protect? Chris, who compounded $20K into roughly $80 million over 18 years using his "social arb" strategy, argues that a 50–70% drawdown is rational inside a properly bucketed, levered, long-horizon risk account. Tom, a former Tudor, Moore, and Caxton PM who now runs a market-neutral fund targeting ~20% annual returns with a 1.4% peak drawdown, counters with the Pareto distribution of hedge fund and day-trader returns, the beta-versus-alpha logic behind institutional portfolio construction, and why allocators refuse to pay for beta they can get on margin themselves. Along the way they debate 3x leveraged S&P ETFs and the hidden roll and theta costs inside them, "too big to fail" AI concentration, the Situational Awareness blowup, crypto's 98% correlation trap, and why overfit backtests stopped landing in Tom's inbox.The second half is a live stress test of social arb as an edge: Chris walks through the Beacon Roofing hailstorm trade built on Google Trends, the Hunger Games and Stranger Things trades built on conversational data, and the e.l.f. Cosmetics trade triggered by a single Jeffree Star video that no sell-side cosmetics analyst had seen. Tom critiques it as an institutional money manager — alpha decay, crowding, discretionary-desk constraints, and whether Citadel plus LLMs will systematize TikTok sentiment before retail's information asymmetry is arbitraged away. If you care about risk management, leverage, position sizing, uncorrelated returns, alternative data, market microstructure, and where retail traders still hold a structural advantage over multi-manager pod shops, this episode is for you.

Apply to Onyx’s trading event here: https://www.onyxcapitalgroup.com/uni-studentsIn this episode of Odds on Open, former Two Sigma quant Omer Cedar joins host Ethan to deconstruct how top-tier quantitative hedge funds systematically aggregate discretionary signals to isolate pure alpha. Cedar reveals the inner workings of institutional alpha capture programs, detailing how premier multi-manager pods map the information propagation curve and exploit crowded consensus sentiment to capitalize on structural mispricings. The conversation provides a rigorous, finance-native breakdown of market microstructure, analyzing the statistical deltas between asset pricing and expert variant perceptions during high-surprise macro and corporate catalysts. For portfolio managers, quantitative researchers, and buy-side analysts, this discussion offers a masterclass in data validation, situational weighting, and the mechanics of separating idiosyncratic returns from passive factor premiums.The dialogue transitions into the future of market structure, exploring how the proliferation of generative AI and autonomous digital analysts will reshape liquidity and market efficiency across equities, commodities, and secondary private markets. Cedar delivers a framework-first outlook on the scaling hedge fund ecosystem, explaining how large language models alter competitive advantage by shifting the alpha premium from commoditized data crunching to proprietary context curation. Crucially, he exposes the most persistent behavioral pitfalls observed across sophisticated institutional desks, specifically unpacking how the conservatism bias hampers optimal sizing during initial portfolio construction. Whether you are an asset allocator evaluating systematic strategies or an MFE student analyzing modern trading frameworks, this episode delivers actionable insights into balancing algorithmic risk management with human judgment.00:00 Intro00:01:14 Why systematic quant models require discretionary human judgment00:06:27 A message from Onyx00:07:06 How Two Sigma engineered an institutional alpha capture pipeline00:13:25 Why extreme consensus sentiment creates contrarian trading opportunities00:18:53 Aligning buy-side and sell-side incentives through informational edge00:25:36 How to extract alpha from the information propagation curve00:34:54 Navigating analyst mean reversion and situational weighting00:39:08 How generative AI and digital analysts reshape alpha capture00:48:13 Why a fragmented hedge fund ecosystem ensures market efficiency00:56:17 How modern LLMs accelerate data validation and market entropy01:03:47 Overcoming the conservatism bias in initial portfolio construction

Apply to Onyx’s Junior Tech Graduate Scheme here: https://verichain.io/apply/0aa1debe-ac6c-452f-9421-da6cbf4a3e8cIn this episode of Odds on Open, we sit down with Victor Haghani, co-founder of Long-Term Capital Management (LTCM) and founder of Elm Wealth, to dissect why institutional alpha frequently breaks down during the position sizing phase. While standard market commentary focuses heavily on asset selection, Haghani establishes that optimizing your risk-adjusted return is an independent, non-zero-sum discipline that dictates long-term survival. We explore the structural friction between expected value and compound return, the misapplication of the Kelly criterion by sophisticated Wall Street portfolio managers, and how treating variance as an internal financial fee reshapes quantitative portfolio construction and risk management.The discussion shifts to edge verification through Haghani’s famous "Crystal Ball" experiment, analyzing how elite macro traders and advanced LLMs process information asymmetry against historical market regimes. Designed for hedge fund analysts, quants, allocators, and advanced finance students, this section provides a rigorous framework for isolating compensated systematic risk from uncompensated idiosyncratic risk. We close with an actionable breakdown of how practitioners should mathematically model and discount their own human capital, offering a definitive blueprint for maximizing lifetime smooth capital accumulation without succumbing to high-volatility ruin.00:00 Intro01:27 LTCM: Why sizing matters more than selection03:00 Expected value vs. risk-adjusted value in portfolios05:15 Why sophisticated investors struggle with sizing bets08:20 The zero-sum reality of beating the market09:30 A message from Onyx10:35 Why most firms lack a risk-adjusted return rubric12:55 Risk as an internal "fee" in portfolio construction16:30 The math of sizing concentrated stock positions20:50 The hidden danger of high-volatility wealth22:20 "How to become a billionaire" is the wrong question27:25 Testing the "Crystal Ball" hypothesis with Wall Street Journal data34:55 How LLMs perform at macro trading games40:35 Can individual investors generate alpha sustainably?47:40 Solving for optimal sizing at Elm Wealth50:25 Risk limits for young investors and human capital57:55 How to estimate the value of your human capital1:02:45 Why changing minds on investing is nearly impossible1:07:35 The most critical factor for a stable wealth curve

In this episode of Odds on Open, we deconstruct the evolution of alternative data and alpha generation with Matt Ober, former Head of Data Strategy at WorldQuant and Chief Data Scientist at Third Point. We dive deep into the institutional framework required to scale systematic trading strategies, the cultural friction of implementing quantamental processes within long-short equity pods, and the specific mechanisms used by portfolio managers to extract a variant view from massive datasets. From the factory-floor automation of PhD-led quant shops to the high-stakes risk management of activist fundamental funds, Matt reveals how market microstructure and data-driven edge define success in liquid markets.Explore the shift toward the "degenerate economy" and the rising institutional utility of prediction markets like Kalshi and Polymarket for hedging structured KPIs. We analyze the future of decision intelligence through the lens of LLMs and MCPs (Model Context Protocol), discussing how traders, analysts, and allocators can maintain differentiation in a regime of rapid alpha erosion. Whether you are optimizing portfolio construction, refining factor exposure, or seeking a competitive advantage in venture capital, this conversation provides a masterclass in leveraging social networks and information symmetry to secure uncorrelated returns.00:00 Intro01:18 The WorldQuant thesis: Mining alternative data for systematic alpha05:06 Integrating finance expertise into PhD-heavy quant factory pipelines07:00 Scaling the quant factory: Automation and the researcher pipeline09:46 Monitoring dataset performance: Risk controls and alpha decay11:07 A message from ONYX16:38 Quantamental shifts: Transitioning data strategy to fundamental funds22:29 Institutionalizing "Old Guard" firms via top-down data buy-in26:38 Networking for quants: Comparing Tulchinsky and Loeb’s sources of edge32:44 The Degenerate Economy: Prediction markets and the volatility of attention39:42 Non-consensus career bets: Why selling beta is a stickier strategy45:01 Sourcing venture alpha: Identifying exceptional founders and GTM wedges52:30 Institutionalizing prediction markets via structured KPI hedging01:00:39 How MCPs and LLMs democratize decision intelligence01:03:08 The single source of edge: Leveraging social network information

Apply here: https://onyxcapitalgroup.com/uni-studentsGeorge Livadas, Portfolio Manager and founder of Peregrine Capital, joins the show to break down the mechanics of running a concentrated, defensive long/short strategy as a solo PM. We explore how George generates alpha by systematically avoiding "hedge fund hotel" crowding, focusing instead on microstructure edges within niche sectors like non-bank financials and packaging to maintain a variant view. George details his portfolio construction framework—balancing core quality compounders with tactical value—and explains the math behind delivering equity-like returns while maintaining a beta-adjusted net exposure of approximately 35%.The conversation shifts to the evolving market regime, specifically how the dominance of multi-manager pods has created liquidity opportunities for patient, independent traders to exploit short-term data noise. George shares his technical survival guide for short selling, from managing volatility in "fraud and fad" names to recalibrating his process following the 2021 SPAC boom. We conclude with a deep dive into macro risk management, discussing how to insulate a portfolio against geopolitical tail risks and the psychological discipline required to develop a professional PM skill set without a traditional institutional pedigree.How do you stay independent in a market dominated by pods?00:00 Intro01:18 Selection criteria and sizing for defensive longs06:04 Sponsor break: Onyx Capital Group anniversary event06:42 Portfolio construction: Balancing quality and value factors09:13 How the SPAC boom changed short-side portfolio construction13:42 Why nimbleness and independent thinking generate alpha20:06 Capitalizing on the short-termism of multi-manager pods28:18 Managing macro tail risks without being wrong-footed38:26 Why defensive strategies offer an "inverse pod" return stream42:04 How to develop a professional PM skill set54:37 Circle of competence: Identifying disastrous longs and shorts