AI, Risk, and the Future of Systematic Investing | Systematic Investor | Ep.387

Episode Overview

• A structural shift in global markets: The era of "US Exceptionalism" is fading as investors re-evaluate the risks of US dollar dominance and equity concentration, driving capital toward global diversification and alternative assets.
• The "Trust Gap" in the age of AI: While AI revolutionizes data analysis and efficiency, it paradoxically increases the premium on human relationships in asset management, as allocators are unlikely to trust algorithms with capital allocation decisions.
• Hedge Funds vs. Private Equity: A divergence is occurring where liquid hedge funds (specifically trend followers) are positioned to thrive in a high-volatility, high-rate environment, while Private Equity faces headwinds from valuation challenges and the end of cheap money.
• The failure of "Liquid Alts" products: The episode critiques the retail liquid alternatives industry, arguing that most products are designed by sales teams rather than investors, leading to high fees, complexity, and poor real-world returns compared to the strategies they mimic.

Key Concepts

• Trend Following Alpha Sources: True alpha in trend following occurs when managers are "early, contrarian, and right." This happens when "smart money" (local experts) continues to buy an asset even as prices rise, signaling a fundamental value shift rather than just market noise.
• Risk Allocation over Market Selection: In volatile or "flat" years, returns are often driven less by picking the right markets and more by how risk is allocated. Dynamic volatility sizing is critical; static position sizing can lead to vastly different (and often poorer) outcomes during market corrections.
• The "Whipsaw" of Short-Term Models: Short-term trading models often act as expensive risk management tools. They tend to react to noise (buying high, selling low) and reduce long-term returns. Long-term models that can withstand volatility are generally required to capture major market moves.
• "Sold, Not Bought" Engineering: Many financial products are built backward: marketing teams identify a sales opportunity (e.g., "we need a Liquid Alt fund"), and investment teams are told to build it. This contrasts with successful funds that launch because an investor identifies a genuine structural opportunity first.
• The "Trust Gap" Thesis: As AI generates hyper-realistic content and commoditizes analysis, the value of human verification increases. In high-stakes industries like asset management, AI will be a tool for efficiency, but human connection will remain the gatekeeper for trust and capital flows.
• The "Beta 0.2 / Cash + 5%" Fallacy: Investors should be wary of liquid products promising high returns with low stock market correlation. Mathematically, achieving these returns without high leverage or massive alpha is nearly impossible; products promising this often rely on fragile backtests that fail in live trading.

Quotes

• At 0:03:30 - "There's something very, very profound in this... watching what's happening in the markets and how people are trying to figure this out, I think it is going to be disruption galore." - Andrew Beer predicting massive industry disruption due to the acceleration of AI.
• At 0:04:06 - "The asset management business is a people business. I don't think asset allocators are going to say, 'You know what, let's have AI pick the next hedge fund for me.' I think it's too risky." - Andrew Beer explaining why human relationships remain the bedrock of capital allocation.
• At 0:07:33 - "Global markets really did change in 2023, and the perception with investors is that the period of US exceptionalism is maybe over." - Tom Wrobel identifying a macro shift in how global investors view US asset dominance.
• At 0:15:13 - "The return characteristics of private equity essentially being some version of leveraged equity with... the wonderful feature that you don't have to mark things to market." - Andrew Beer critiquing the Private Equity model's reliance on smoothing volatility.
• At 0:22:42 - "CTAs generate alpha when they are early, contrarian, and right." - Andrew Beer defining the specific conditions required for trend followers to outperform the broader market.
• At 0:27:15 - "It was really the allocation of risk last year that was the key thing... usually we see... a much stronger preference for longer-term models outperforming." - Tom Wrobel discussing how managing risk exposure was more critical than just signal generation.
• At 0:31:18 - "Our view is [short-term models] are a risk management tool. And they're often a costly risk management tool... if you're going to allocate a meaningful amount of money to something that has a zero Sharpe ratio over time... no thanks." - Andrew Beer critiquing trading signals that reduce volatility but destroy returns.
• At 0:39:56 - "The valuable trends for CTAs are when... fair market value goes to 11 [from 10]... and the people with local knowledge like it more at 11 than they liked it at 10." - Andrew Beer explaining the difference between market noise and fundamental repricing.
• At 0:45:58 - "The returns over 15 years... is between 2 and 3% and the fee structures [are] on average about 200 basis points... This is worse than throwing darts." - Andrew Beer on the statistical failure of the liquid alts category.
• At 0:48:15 - "The guys who are building the products are the equivalent of the salesman on the showroom floor... designing a car for you because he thinks you'll buy it, and if it's a lousy car in three years, it's not his issue." - Andrew Beer illustrating the misalignment in financial product design.
• At 0:52:54 - "People really don't talk much about returns. They talk about modeling and research and data and innovations... And I kept asking these questions like, 'What's the realized Sharpe ratio of the strategy over what period of time?'" - Andrew Beer observing that complexity is often used to mask poor performance.
• At 0:57:35 - "G10 currency carry does not have a long-term Sharpe ratio of 1.2. I'm sorry, that's a backtest." - Andrew Beer highlighting the dangerous gap between marketing simulations and reality.

Takeaways

• Diversify beyond the US Dollar: Scrutinize your portfolio's reliance on US exceptionalism. Consider how currency risk affects your real returns and look for assets that do not correlate perfectly with the USD.
• Focus on realized returns, not complexity: When evaluating quantitative strategies, ignore the "black box" complexity of the model and focus on the realized Sharpe ratio. Complexity often masks poor performance; simple, robust strategies often win long-term.
• Avoid short-term trend models for growth: If your goal is capital appreciation, favor long-term trend following strategies. Short-term models are effective for reducing volatility but act as a drag on total returns over time.
• Be skeptical of "Sales-First" products: When buying a liquid alternative fund (ETF/Mutual Fund), ask if it was built to solve an investment problem or to fill a marketing gap. Avoid products that seem designed purely to capture current market hype.
• Look for "unsexy" markets: Don't limit investment to major indices or tech stocks. Significant alpha in trend following often comes from niche markets like cocoa, coffee, and minor metals where structural supply/demand imbalances create durable trends.
• Rethink Private Equity allocations: Be cautious with Private Equity in a high-rate environment. Its past success relied heavily on cheap leverage and volatility smoothing, conditions that may not persist. Liquid alternatives may offer better opportunities in the current cycle.
• Verify "Risk Premia" claims: Treat any product labeled "Alternative Risk Premia" as an active trading strategy with execution risk, not a guaranteed yield. If a product promises high returns with very low volatility and low fees, assume the backtest is flawed.