Trend Following Is Changing Faster Than Ever | Systematic Investor | Ep.409

Episode Overview

• This episode explores critical structural evolutions in public markets, focusing on the launch of single-stock futures (SSFs) by major exchanges like the CME and their implications for systematic investors.
• It examines the shifting market regime from index-level macro trends to high-dispersion, idiosyncratic equity environments where individual stock correlations are at historic lows.
• The discussion breaks down the market microstructure impacts of retail-driven 0DTE (Zero Day-to-Expiration) options, which force institutional market makers into volatility-dampening hedging patterns.
• The conversation highlights how quantitative researchers can leverage alternative data to exploit "narrative momentum" and transient thematic risk factors arising from the news cycle.

Key Concepts

• Single-Stock Futures and Democratized Short Selling: The introduction of single-stock futures (SSFs) on major exchanges provides a highly capital-efficient, operationally simple way to take directional short positions. This bypasses the complex, costly, and fragmented stock loan market traditionally managed through prime brokerage relationships.
• Market Dispersion and the Idiosyncratic Regime: Equity markets are transitioning to a low-correlation, high-dispersion state. Rather than moving together based on macro index beta, stock prices are diverging based on company-specific fundamentals, creating a favorable landscape for relative value, stock-picking, and single-name systematic trend strategies.
• 0DTE Options and the Anti-Trend Feedback Loop: Massive retail volume in 0DTE options (primarily selling at-the-money contracts) puts institutional market makers in structural long-gamma positions. To remain delta-neutral, market makers must buy index dips and sell rallies, which systematically dampens intraday realized volatility and creates a strong mean-reverting environment.
• Commodity Term-Structure Dynamics: In physical commodity markets (such as agriculture or natural gas), trend-following models must account for location on the curve. Front-month contracts are highly sensitive to immediate physical demand and storage limits, while back-month contracts are driven by structural, long-term supply and crop-cycle expectations.
• Volatility-Normalized Tick Size: As stock index prices rise significantly over decades while absolute tick sizes remain static, the economically normalized tick size shrinks. This relative reduction in minimum price increments introduces micro-market noise and reduces short-term autocorrelation, making high-frequency trend following more difficult.
• Thematic Risk and Narrative Momentum: Beyond traditional industry classifications and style factors, markets are influenced by transient, news-driven "themes" (e.g., the AI boom). Because investors take time to fully digest and verify the complex implications of these emerging stories, the market initially under-reacts, generating exploitable narrative-driven momentum over medium-term horizons.

Quotes

• At 0:01:56 - "The CME has announced they are going to launch single-stock futures on fifty-odd leading US stocks... This is really exciting because as a practitioner, one of the things that is really important is thinking about the practical implications of trading... The ability to short a stock is not very easy for a retail investor, but if you have a futures contract, suddenly it becomes possible because you don't have to worry about shorting and borrowing." - Explains how financial product innovation lowers structural barriers to short-selling and improves execution efficiency.
• At 0:05:35 - "Being short a stock can be lethal because there is unlimited downside if you are short and there is a takeover or whatever that might be." - Warns of the asymmetric tail risks inherent in individual equity short positions, emphasizing the need for robust risk management.
• At 0:11:00 - "All this money which is being thrown at AI is thrown at chips that will last maybe for two or three years, and then they become old and obsolete... if it does go wrong and AI doesn't materialize itself, basically there's going to be zero recovery." - Draws a parallel between the 2001 telecom fiber-optic overbuild and the current AI hardware-centric CapEx boom, noting the higher obsolescence risk of modern silicon.
• At 0:17:16 - "In markets such as natural gas and agricultural commodities, neighboring expiries can have materially different behavior and even different risk characteristics." - Introduces the complexity of term-structure modeling in physical commodities, where seasonality and storage constraints break price continuity.
• At 0:19:34 - "The front of the market is really dominated by demand because we've already produced... what can surprise us is demand shocks. As you go further out on the curve, suddenly you're moving to the new crop—you really don't know what's going to be the supply." - Explains how risk factors shift along a commodity curve, transitioning from short-term demand-driven volatility at the front end to supply-driven production uncertainty in back-month contracts.
• At 0:28:43 - "The market seems to look like they are becoming more idiosyncratic... Correlation is at an all-time low, meaning idiosyncratic risk is at an all-time high." - Identifies a critical macro regime shift where individual stock dispersion dominates index-level beta, altering the opportunity set for systematic trend strategies.
• At 0:31:14 - "What happens if the price starts going up? The market maker is long a lot of delta, and they hedge it by shorting the future... that reduces the price, so that pulls back the price to the strike. To me, that is anti-trend." - Deconstructs the mechanical feedback loop of retail 0DTE option selling, explaining how market maker delta-hedging dampens intraday trends and drives mean reversion.
• At 0:31:15 - "With the zero day-to-expiration [options], retail has been selling options at-the-money near expiry... and that means that creates a very strong positive gamma for the market makers... If the price starts going up, the market maker is long a lot of delta, and they hedge it by selling the future..." - Outlines the structural setup of 0DTE options and how they alter intraday equity market microstructure.
• At 0:35:15 - "The S&P has changed from about 1,000 in 1995 to about [4,800]... so the tick size in the equity space has become much smaller." - Explains the concept of volatility-normalized tick size, demonstrating how massive price growth effectively shrinks the minimum price step relative to overall value, increasing micro-market noise.
• At 0:40:37 - "Themes are another way of explaining the volatility... they are risk factors but things that are transient and they are driven by news." - Defines thematic risk and highlights how temporary, news-driven narratives drive co-movements among stocks that static industry classifications fail to capture.
• At 0:46:33 - "What he finds is that there is under-reaction... it takes time for investors [to react]. And we are very familiar with that, we as trend followers... in many ways, the whole point of that is that in the absence of complete information in the market, the fact that the price has gone up actually makes you evaluate your own valuation." - Outlines the psychological and informational inefficiencies behind narrative momentum, where slow information diffusion generates investable medium-term trends.

Takeaways

• Differentiate trend models by curve location in physical commodities: Do not treat physical commodity contract months as identical; use distinct modeling parameters for front-month contracts (driven by immediate storage and demand pressures) versus back-month contracts (driven by supply and future crop cycles).
• Hedge against rapid asset obsolescence in tech CapEx: Account for the rapid physical depreciation of highly specialized AI hardware (2–3 year lifespan) when assessing the long-term risk profiles of hardware tech companies.
• Pivot systematic strategies from index-level to single-stock trend following: Take advantage of historically low stock-to-stock correlation by deploying capital-efficient relative value or single-stock trend models rather than relying on heavily suppressed index-level trends.
• Adapt intraday models to account for 0DTE mean-reverting flows: Adjust short-term trading algorithms to expect mean reversion and "buy-the-dip" patterns around heavy option strike concentrations, as market makers dynamically hedge their long-gamma portfolios.
• Quantify news cycles to capture narrative momentum: Monitor and track the rise of transient media themes using alternative data to exploit market under-reaction to complex, slow-moving narrative developments.
• Prepare NLP systems for a fragmented information environment: Redesign natural language processing models to handle highly personalized, AI-generated news feeds as the market transitions away from a single, shared database of objective public records.