Yesterday, I published a self-audit of a blog post I wrote 3 years ago (“War in Ukraine: The Long-Term Impact”, August 16, 2022).
What emerges in hindsight: I systematically undervalued adaptive capacity of political and market systems when existential stakes are present. Cynicism about inertia was a bias — understandable given decades of EU under-delivery! — but wrong in the face of shock. Conversely, I over-indexed on the uniformity of sanction logic, assuming all frozen-asset cases would rhyme. And I under-credited opportunism in the Global South, mistaking it for passivity. Ironically, this does align with my operating principles: “Default to small, true signals. Ignore hype”: In 2022, the signals I weighted most (bureaucratic inertia, sanctions rigidity) turned out less true than the latent capacity for strategic acceleration.
If I were to treat my Ukraine forecast + 3-year audit as a “live-fire exercise” in foresight, a few discipline shifts stand out for next time I frame directional consequences under tension.
1. Separate Structural vs. Flow Variables
I bundled structural inertia (EU politics, NATO spending) with flow variables (commodity flows, reserve allocations): Structural factors (bureaucracy, procurement lags) move slower, while flow variables (oil cargoes, FX reserves, central bank buying) can swing in months.
Next Time: I will keep those categories distinct in my next investment thesis. That prevents overgeneralizing “all things are slow.”
2. Map Sequencing, Not Just Outcomes
I called for a “clean break on Russian energy.” What happened was sequenced decoupling: First oil, then pipeline gas, then LNG.
Next Time: I will sketch first-, second-, third-order decoupling paths. Even if the end state is correct, the path matters for timing and capital allocation.
3. Stress-Test Biases Explicitly
Cynicism was my bias. I assumed sclerosis was the dominant force.
Next Time: I’ll use my mentors and partners and LLMs to pressure-test my biases: What does my cynicism make me blind to? Where might adaptive capacity show up faster than I expect?
4. Weigh Opportunism as Much as Fear
I framed EM neutrality as fear of repercussions. It turned out to be opportunism.
Next Time: I will ask, who benefits if the system fractures? and assume rational exploitation, not passivity.
5. Distinguish “Signal” from “Analogy”
I extrapolated Russia freezes to cannabis, LatAm, blockchains. It’s the correct principle, but wrong analogies.
Next Time: After drawing an analogy, I must ask is this truly the same mechanism, or just thematically rhyming? If the mechanics differ (foreign policy vs. regulatory enforcement), I have to down-weight the analogy.
6. Explicitly Frame Denominator Effects
Many asset allocators track KPIs and ratios. But system shocks can create denominator effects — like a seemingly over-exposure in Venture Capital when public markets have a Black Friday. That creates opportunity (for a short time). My LNG miss was partly because oil fell faster, making LNG share look bigger. Something I didn’t take advantage of in my asset allocation strategy.
Next Time: I will test for denominator effects: “If X collapses fast, does Y look artificially larger?” This avoids mis-reading ratios as directional trends.
7. Build Reversibility into Foresight
Some of my misses (sanctions rigidity, global trade collapse) were low-reversibility in my framing — too absolute. It made for a good blog post, but not for a good investment decision. Howard Marks would demand to be precise about what’s permanent vs what’s cyclical.
Next Time: I will phrase forecasts with “path-dependent reversibility”: e.g., sanctions rigidity until enforcement fatigue, leads to carve-outs, leads to re-tightening.