Risk Responses to Dynamic Asset Thresholds

A recurring theme in development economics is that risk affects individual production, consumption, exchange, and investment behaviors in ways that ultimately shape income and wealth distributions. Arrow's Decreasing Absolute Risk Aversion conjecture implies that the poor prefer low return, low risk activities, while the rich more quickly adopt higher return, higher risk activities. The resulting divergence in microlevel growth rates may create a risk-aversion-induced poverty trap. This risk aversion-to-asset dynamics logic has fueled decades of research. In this thought piece, we casually explore the possibility of the opposite causality: might underlying patterns of asset dynamics affect risk-related behaviors? Suppose (a) that asset dynamics in a particular context are non-convex for reasons unrelated to risk and (b) that individuals accurately perceive the location and severity of key dynamic thresholds. Should not we expect individuals to adjust their behavior, including their risk responses, near these thresholds accordingly? We hypothesize that, relative to static risk preferences as commonly captured by the concavity of contemporaneous utility function, the rational adjustment involves greater risk avoidance just above the dynamic asset threshold and greater risk taking just below it. After a brief literature review, we sketch out a conceptual model and discuss suggestive empirical evidence before concluding with a few thoughts on possible extensions of this line of research and its relevance to policy-making