Position Sizing When Markets Break feat. Rob Carver | Systematic Investor | Ep.386

Episode Overview

• Explores the critical mechanisms of systematic trend following, specifically focusing on how volatility targeting and position sizing protect capital during market crashes like the recent Silver drop.
• Examines the "Total Portfolio Approach" (TPA), arguing that investment decisions should be viewed through the lens of the entire portfolio's risk profile rather than as isolated bets.
• Discusses the specific role of "Crisis Alpha," explaining why traditional markets (stocks/bonds) often provide better protection during major financial panics than exotic alternative markets.
• Provides a technical deep dive into portfolio construction, including the different time horizons needed for estimating volatility versus correlation and the structural advantages of capital efficiency in futures trading.

Key Concepts

• Dynamic Position Sizing & Volatility Adjustments A cornerstone of systematic trading is that position size is not static; it fluctuates based on market volatility. As a market becomes more volatile (riskier), the system automatically reduces position size to keep the overall risk constant. This mechanical adjustment is what protected systematic traders during the silver crash; while "buy and hold" investors suffered the full drawdown, systematic models had likely already reduced exposure as volatility crept up before the collapse.

• The Liquidity Illusion in Parabolic Markets In niche markets like Silver, a parabolic price run-up attracts "tourists"—traders who don't usually operate in that space. This creates an illusion of high liquidity because volume is high on the way up. However, liquidity is dynamic. When the trend turns, buyers vanish instantly (the "cinema with one exit" analogy). This liquidity vacuum causes outsized price drops, trapping those without strict risk protocols.

• Portfolio as "Gesamtkunstwerk" (Total Work of Art) Investing should be approached via the "Total Portfolio Approach." Instead of viewing a portfolio as a collection of disconnected trades, it should be seen as a single, cohesive entity. Every decision—whether adding a new asset or adjusting a stop-loss—must be evaluated based on how it impacts the risk and return characteristics of the entire portfolio, effectively treating the portfolio as a single instrument.

• The "Crisis Alpha" Paradox & Idiosyncratic Risk In normal market conditions, diversification relies on assets being uncorrelated (moving independently). However, during major crises, correlations often converge to 1 (everything falls together). Research suggests that during these specific panic events, trend followers should actually prioritize exposure to traditional markets (like shorting stocks and going long bonds) rather than "exotic" alternatives. The massive trends in major asset classes during a crisis drive the bulk of the returns ("Crisis Alpha"), while niche markets offer less protection during liquidity events.

• Hierarchy of Predictability When building trading models, not all variables are equally predictable. Future volatility is the most predictable (high autocorrelation), followed by correlation (moderately predictable), with future returns being nearly impossible to predict (mostly noise). Therefore, robust systems use shorter lookback windows (e.g., 1 month) to catch fast-moving volatility changes, but longer windows (e.g., 6 months) for correlation estimates to avoid over-trading based on noisy data.

Quotes

• At 6:10 - "CTAs generally will have positions that are a function of two things: the strength of the trend and the volatility of the instrument you're trading." - Explaining the fundamental formula that automatically scales down risk when markets become dangerous.
• At 10:02 - "Anyone who sells on the hint of volatility spiking... the more people there are in a particular market like that, the more it will magnify the move." - Describing the feedback loop where simultaneous risk management selling can unintentionally accelerate a market crash.
• At 15:36 - "If the price starts to move adversely against you... it's like the cinema with only one fire door, right? Everyone gets jammed trying to get through the exit at the same time." - Using a vivid analogy to explain why liquidity evaporates instantly during panic selling in smaller markets.
• At 17:04 - "On Friday I lost 3% of my total account value because I have a reasonably diversified portfolio... If I was very highly concentrated in silver that would have been a much bigger number." - Illustrating the massive difference between asset-level volatility (a 26% crash) and portfolio-level risk (a 3% dip).
• At 19:20 - "For him [Bunker Hunt] was not about making a lot of money necessarily, it was about controlling... but he kept buying for borrowed money." - Contrasting emotional, leveraged pyramiding (which leads to ruin) with the dispassionate risk reduction of modern systems.
• At 32:27 - "Volatility... is much more predictable than correlation... but if you were to try and predict means [returns], you'd get nothing. You'd just get noise." - Establishing the hierarchy of financial data reliability that guides how systems should be built.
• At 33:28 - "You probably want to use something like about a 6-month rolling estimate [for correlation] versus about a 1-month rolling estimate for vol." - Providing a practical rule of thumb for calibrating the sensitivity of different risk models.
• At 39:27 - "The key thing that's interesting about these alternative and esoteric markets is... a much bigger proportion of their returns are idiosyncratic; they're not coming from these big factors." - Explaining why alternative markets are valuable for diversification in normal times, even if they are less useful during global crises.
• At 40:27 - "In a crisis... you basically want to just make fairly big bets on going... short equities, long bonds. You're not really going to need your fancy... other weird markets there to help you." - Highlighting the counter-intuitive strategy that simpler markets provide the best protection during complex financial meltdowns.

Takeaways

• Target Volatility, Not Just Returns: Implement a system where your position size automatically decreases as an asset's volatility increases. This acts as a natural brake that prevents a sudden market explosion from blowing up your account.
• Differentiate Your Lookback Windows: Do not use the same timeframe for all metrics. Use a short window (e.g., 1 month) to react quickly to volatility spikes, but a longer window (e.g., 6 months) to estimate correlations to prevent "whipsaw" trading from noise.
• Leverage Capital Efficiency: Utilize the margin efficiency of futures (where you only post ~20-30% collateral) to overlay trend following strategies onto a core portfolio (like stocks/bonds) without having to sell your base assets.
• Prioritize Diversification Over Prediction: Accept that you cannot predict which specific market will crash. Defense comes from holding uncorrelated assets so that a 26% drop in one instrument (like Silver) only results in a minor single-digit drag on the total portfolio.
• Adjust Allocation for "Crisis" vs. "Normal" Regimes: Recognize that while exotic markets provide great diversification in normal times, during a true financial crisis, you should rely on major asset classes (bonds, currencies, indices) to generate the "Crisis Alpha" needed to offset equity losses.