Mathematical models like the Kelly Criterion are only as good as their inputs. Historical data, such as a stock market's return, isn't a fixed 'true' value but rather one random outcome from a distribution of possibilities. Using this single data point as a precise input leads to overconfidence and overallocation of capital.
Related Insights
Ken Griffin is skeptical of AI's role in long-term investing. He argues that since AI models are trained on historical data, they excel at static problems. However, investing requires predicting a future that may not resemble the past—a dynamic, forward-looking task where these models inherently struggle.
Top tennis players like Rafael Nadal win only ~55% of total points but triumph by winning the *important* ones. This analogy illustrates that successful investing isn't about being right every time. It's about consistently tilting small odds in your favor across many bets, like a casino, to ensure long-term success.
Post-WWII, economists pursued mathematical rigor by modeling human behavior as perfectly rational (i.e., 'maximizing'). This was a convenient simplification for building models, not an accurate depiction of how people actually make decisions, which are often messy and imperfect.
Allocate more capital to businesses with a highly predictable future (a narrow "cone of uncertainty"), like Costco. Less predictable, high-upside bets should be smaller positions, as their future has a wider range of possible outcomes. Conviction and certainty should drive allocation size.
People justify high-risk strategies by retroactively fitting themselves into a successful subgroup (e.g., 'Yes, most investors fail, but *smart* ones succeed, and I am smart'). This is 'hindsight gerrymandering'—using a trait like 'smartness,' which can only be proven after the fact, to create a biased sample and rationalize the risk.
Judging investment skill requires observing performance through both bull and bear markets. A fixed period, like 5 or 10 years, can be misleading if it only captures one type of environment, often rewarding mere risk tolerance rather than genuine ability.
Contrary to popular belief, economists don't assume perfect rationality because they think people are flawless calculators. It's a simplifying assumption that makes models mathematically tractable. The goal is often to establish a theoretical benchmark, not to accurately describe psychological reality.
Moving from science to investing requires a critical mindset shift. Science seeks objective, repeatable truths, while investing involves making judgments about an unknowable future. Successful investors must use quantitative models as guides for judgment, not as sources of definitive answers.
Quoting G.K. Chesterton, Antti Ilmanen highlights that markets are "nearly reasonable, but not quite." This creates a trap for purely logical investors, as the market's perceived precision is obvious, but its underlying randomness is hidden. This underscores the need for deep humility when forecasting financial markets.
Absolute truths are rare in complex systems like markets. A more pragmatic approach is to find guiding principles—like "buy assets for less than they're worth"—that are generally effective over the long term, even if they underperform in specific periods. This framework balances conviction with flexibility.
