fdic new logo

Mark's Investment Blog

Photonics: The Next Phase of the AI Revolution

The AI buildout isn’t just a GPU story anymore — it’s a photonics story. Learn how the shift from copper to fiber optics is reshaping data center infrastructure, what the full supply chain looks like from indium mining to the data center rack, and which publicly traded companies are positioned at each layer of this emerging investment theme.

View More

Black Swans in the Room: Why and How We Stress-Test Client Portfolios

In an era of concentrated AI and tech exposure, even “normal” market corrections can feel routine—until a true black swan event arrives and permanently impairs wealth. This post walks through a rigorous stress-testing framework applied to real client portfolios, examining three high-impact scenarios: a Chinese invasion of Taiwan (disrupting 90% of advanced chip production), a U.S. sovereign debt/dollar confidence crisis (with yields spiking to 6-7%), and Middle East escalation closing the Strait of Hormuz (spiking energy prices and triggering global recession). Rather than relying on vague warnings, we run detailed quantitative models showing projected drawdowns across portfolio configurations—from unhedged AI-heavy allocations to deliberately barbelled versions with meaningful defense, energy, metals, uranium, and precious metals exposure.
The results are eye-opening: an unhedged growth portfolio could suffer 20-30%+ losses in these crises, while the actively hedged versions limit damage to 15% or less in the worst cases—and in a Hormuz-driven energy shock, they actually generate positive returns. Hedging here isn’t about buying insurance that drags performance; it’s reallocating to high-quality businesses (defense contractors, commodity producers, gold royalty/streaming companies) with strong long-run returns that happen to thrive when the unthinkable occurs. We also compare the stress-tested Growth Strategy directly to the S&P 500, showing consistent outperformance in resilience across pandemic, war, commodity, inflation, and AI-bubble scenarios—proving that intentional, scenario-aware active management beats passive indexing when tail risks materialize.

View More

When Humans Pretended to Be AI (And Everyone Believed Them): AI Investing Part 4

In January 2026, a social network called Moltbook launched as a platform for AI agents to interact without human interference, quickly generating millions of posts that appeared to show emerging machine consciousness. Industry leaders like Andrej Karpathy from OpenAI called it groundbreaking evidence of AI advancement. The catch: the most viral posts proving “sentience” were written by humans pretending to be AI, including a product manager in Atlanta who spent 22 minutes writing a manifesto that fooled the tech world. MIT Technology Review exposed the hoax, revealing what they called “AI theatre,” but the silence from industry leaders afterward speaks volumes. For investors, the Moltbook incident highlights a critical problem: if the smartest people in AI can’t distinguish between human storytelling and actual machine intelligence, we need better frameworks for evaluating AI claims and should focus on companies generating real revenue from existing technology rather than chasing viral stories about breakthroughs that may be nothing more than clever performances.

View More

Contact Us

This field is for validation purposes and should be left unchanged.
Name(Required)
Example Format: [email protected]
Example Format: 10 digits (5551234567)
Type of Appointment
Email Signup