The World is Chaotic and Unmodelable

World systems that interact with human agents are mostly complex, unpredictable, and beset by chaos and irrationality. Periods of order—and relatively predictable outcomes—alternate with bouts of disorder, often triggered by insignificant reasons. We can easily be lulled into a false sense of security, only for disruptions, amplified by emotional feedback loops, to spark outsized events.

How We Think Beyond Models

We distinguish between risk, which we define as something that can be measured, modelled, and often mitigated, and uncertainty, which involves significant unknowns with poorly assessable probabilities and vaguely defined outcomes.

We approach this kind of uncertainty through imaginative scenarios and narratives (like science fiction). We stay alert to the lure of excessive hope over likely reality. Instead of relying on intuitive logic, we use backward logic: starting from the endpoint and working back to understand how it might come to pass (some generative AI models are surprisingly helpful). We also spend more time pondering upside possibilities than downside probabilities. This reflects the positively skewed nature of equity returns (where losses beyond shareholder capital are "socialized"). Nearly all fund managers who lose their jobs in bull markets do so not for picking duds, but for missing the big winners—without which outperformance is particularly challenging. Modern Portfolio Theory has sown the seeds of active fund management's decline, with its overemphasis on downside protection and failure to account for the extreme skew in long-term returns.

We also consider scenarios that can be modelled fairly accurately (e.g., Trump has a 60% chance of winning the presidential election), but where the potential outcomes remain unclear. In these cases, analysing portfolio resilience and robustness can be especially helpful.

Finally, some scenarios have clear outcomes, but assessing their likelihood is extremely difficult (e.g., will Google be fined $20 billion by a European court?). Here, we estimate likelihoods in broad bands, rank scenarios, and create lists of possibilities.

Our Competitive Edge

We devote most of our time to publicly listed equities with the potential for 3-5x returns over 3–5 years. We find that many investors rely on simple heuristics to avoid these companies—or they're barred by institutional constraints including volatility constraints, myopic valuation tools or “uninvestibilility” (I.e. Israel, defence, oil, China, ‘too hard to understand’). Taking a position in a highly ‘speculative’ stock that then flops carries real career risk. We strive to create space for fun, humility, spirit, and audacity.

We get the majority of our calls wrong. But over multiple decades the approach has worked well, across all market phases (bear/bull, growth/value), in many industries, and in many  stockmarkets around the world.

Richard Sneller, October 2025