Leigh Drogen – Quantamental Approach to Crypto at Starkiller Capital (Manager Meeting, EP.32)

Episode Summary

Executive Summary: Lee Drogan traces his path from behavioral economics and hedge fund quant work to building Estimize and then Starkiller Capital. He argues crypto is best approached as a volatile, early-stage technology market where momentum, on-chain fundamentals, code/audit review, and disciplined risk management can capture upside while limiting catastrophic drawdowns. He is skeptical of long/short equities and of crypto narratives that overclaim certainty.

Main Topics: Early career and behavioral edge in quant investing (Priority: 5/5): Drogan describes moving from international relations and behavioral economics into a hedge fund internship, where a test of his problem-solving style revealed the behavioral fit that made him successful in quant long/short investing. Retail vs. institutional market experience (Priority: 4/5): He contrasts the rigid process and emotional discipline of an institutional quant shop with the chaotic, entrepreneurial, sentiment-rich environment at StockTwits, which taught him to operate without structure. Estimize and the crowdsourced estimates thesis (Priority: 5/5): Drogan explains how Estimize emerged from observing unstructured earnings estimates on StockTwits and aiming to build a structured, representative crowdsourced expectations dataset to improve forecasting and alpha generation. What hedge funds get wrong with alternative data (Priority: 5/5): He argues many hedge funds misuse data by hunting for evidence to support existing theses instead of letting data generate ideas, especially in discretionary long/short equity shops. Why crypto was a natural next step (Priority: 5/5): Drogan’s crypto thesis evolved from owning Bitcoin out of curiosity to seeing Ethereum and broader crypto as a technological adoption cycle with large momentum effects and inefficiencies suitable for systematic trading. Starkiller Capital’s process: momentum plus fundamentals (Priority: 5/5): He details a hybrid crypto process: cross-sectional and time-series momentum for asset selection and beta control, combined with on-chain fundamentals, code/audit diligence, team analysis, and liquidity filters to avoid frauds and weak projects. Risk management, portfolio construction, and outlook (Priority: 4/5): He emphasizes reducing exposure during rollovers, hedging with liquid futures, re-entering tactically after drawdowns, and believes crypto can scale into a major financial infrastructure layer while hedge funds in equities face tougher economics.

Key Arguments: Behavioral disposition matters as much as raw technical skill in quantitative investing; his box-folding interview test showed he was suited to admit uncertainty and seek help rather than force a bad process. StockTwits taught him how markets behave in unstructured, sentiment-driven environments, while Geller Capital taught him regimented process and emotional self-monitoring; both were essential to founding Estimize. Estimize’s premise was that structured, crowdsourced earnings expectations could be more representative and more accurate than traditional sell-side estimates when properly controlled. Many hedge funds, especially discretionary long/short equity shops, reverse-engineer data to support existing views instead of using data to build a factor dashboard and generate ideas objectively. In crypto, nobody truly knows which technology will win long term; therefore the best approach is to treat assets like venture-style experiments and let adoption, fundamentals, and momentum guide exposure. Crypto markets exhibit unusually strong momentum and trend behavior because liquidity is uneven, leverage is high, and there is little intrinsic-value anchoring; this makes CTA-style models particularly effective. A hybrid approach works best: use momentum models to capture broad adoption cycles and on-chain fundamentals/code reviews/team diligence to avoid frauds, rugs, and structurally weak assets. Risk must be managed by reducing beta as trends roll over, hedging with liquid futures, and accepting that bottoms are hard to catch exactly; the goal is to be close, not perfect. He is skeptical of “crypto as global reserve currency” and similar maximalist narratives, but bullish on crypto as infrastructure for transparent, programmable financial activity and better risk management over time. Long/short equity is increasingly difficult and crowded, while crypto still offers a sizable inefficiency set and enough liquidity for a meaningful hedge fund business, even if not at the scale of the largest equity platforms.

Data Points: Hedge fund internship timing: Summer between sophomore and junior year of college - When Drogan entered Geller Capital and began learning quant long/short investing Age when he started first asset management firm: 22 years old - He launched a small firm after Geller Capital shut down Participation rate for Estimize’s early outreach: 25-30% - Approximate share of hedge funds/asset managers that agreed to participate in the new crowdsourced estimates platform Crypto universe today: ~350-400 coins - Assets Starkiller Capital can meaningfully trade given liquidity constraints Liquid market concentration: Top 10-15 coins have a lot of liquidity - He notes liquidity falls off quickly beyond the top names Drawdowns in crypto: 70-80% - Typical pain point for holders of spot crypto through bear markets Potential additional decline after a major drawdown: Another 50% after being down 60% - Illustrates why bottom-fishing is difficult in crypto bear markets Typical re-entry window: 1-2 weeks - How Starkiller steps back into risk when models and conditions improve Outsized long-term opportunity: 5x, 10x, 100x - He frames crypto investments as adoption-cycle bets rather than 40% equity-style returns Time horizon for blockchain infrastructure thesis: 10, 15, 20, 30 years - He sees crypto’s broader financial infrastructure impact as a long-duration theme Acquisition timing: April of last year - Estimize was acquired, enabling him to start Starkiller Capital fully AlphaSense event timing: October 6th through 8th, 2025 - Promo for Alpha Summit 2025 mentioned in the ad reads AlphaSense source count: Over 500 million premium sources - Marketing claim about the research platform AlphaSense expert calls: Over 200,000 expert calls - Marketing claim about the research platform

Pivotal Quotes: "I hired you specifically because of your behavioral disposition and the fact that it fits exactly what we do here." — David Geller (as recounted by Lee Drogan): Explaining why Drogan was selected for quant long/short work after the cardboard-box interview test "Most of what's actually happening is a PM says, I have a specific hypothesis, and he asks his data team go out and find me a piece of data that supports this hypothesis." — Lee Drogan: Critique of how many hedge funds misuse alternative data and analysis "The goal in crypto is simply don't be the midwit." — Lee Drogan: His shorthand for avoiding overly skeptical conventional thinking that misses asymmetric winners

Implications: For investors, the episode argues for systematic discipline in crypto, not narrative chasing: pair momentum with real-time fundamentals and risk controls. For hedge funds, it highlights the limits of discretionary stock picking and the growing attractiveness of crypto’s inefficiencies.