Markets Are Becoming Less Efficient Not More
Markets: The Inefficiency Paradox
In a recent discussion, I posited that when confronted with complex challenges, rational agents often fail to calculate the optimal solution or learn the best strategy. Instead, they settle for “satisficing” solutions. A common counter-argument suggests that the aggregate decision-making of a large group of agents could negate individual errors, leading to efficient markets as though each participant were rational.
However, this perspective overlooks a crucial aspect: individual errors are frequently correlated. This correlation arises not just because humans inherently share similar behavioral biases, but more significantly due to our strong interconnections. These connections manifest through social interactions (both physical and virtual) and are further influenced by price changes, which we often misinterpret as informative, even when they’re not.
This ties back to the concept of price impact, a topic I’ve touched on in previous posts and can be explored further at The inelastic market hypothesis: a microstructural interpretation: Quantitative Finance: Vol 22, No 10 (tandfonline.com).
We understand that when interactions among individuals reach a certain threshold, collective behavior radically diverges from what would occur if each acted independently. History is littered with evidence of this phenomenon, from mass panics and fashion fads to economic bubbles, crashes, and various crises. Often, these events seem inexplicable, as they lack significant exogenous shocks that could justify their magnitude. Yet, when one considers the potent combination of interactions and feedback loops, it becomes evident how minor incidents can escalate into major crises (see Crises and Collective Socio-Economic Phenomena: Simple Models and Challenges | Journal of Statistical Physics (springer.com) for examples).
This oversight of social effects is a glaring deficiency in classical economic models, including DSGE and efficient market hypotheses. These models are essentially blind to the quintessential traits that define our humanity: our nature as social beings, for better or worse. By ignoring these critical social dynamics, classical models fail to capture the full spectrum of human behavior and its impact on market efficiency.
This is probably even more relevant with modern social media, and perhaps why Cliff Asness argues that financial markets are less efficient than 30 years ago, cf. Markets are becoming less efficient, not more, says AQR’s Clifford Asness (ft.com)
(3) jean-philippe bouchaud | LinkedIn
