Economists propose the “precautionary augment” to measure how uninsurable risk raises expected marginal utility, reshaping saving incentives, risk-free rates and asset prices.

Economists have long noted a basic implication of uncertainty: for a “prudent” decision-maker, additional risk tends to make an extra dollar more valuable at the margin. In a new theoretical contribution, Miles S. Kimball and Vincent A. Mastantuno sharpen this idea by introducing a precise yardstick for how much risk changes the marginal payoff from saving and related financial quantities.

The authors call this metric the precautionary augment. Formally, it is the proportional increase in expected marginal utility triggered by the presence of risk. Framed this way, the concept does more than restate precautionary saving intuitions. It links uncertainty directly to three cornerstone objects in asset pricing and intertemporal choice: the marginal value of saving, the risk-free rate in the classic Lucas tree setting, and equilibrium asset prices.

A key part of the setup is a market environment that is rich in derivatives—so many state-contingent securities are available that markets appear “complete” at first glance. Yet the authors focus on a realistic limitation: even with extensive trading in market-state securities, investors may still face idiosyncratic risks that cannot be fully insured. In that case, the precautionary augment becomes a way to quantify how these uninsurable personal shocks feed into saving incentives and prices, even when aggregate risks are heavily traded.

Building on the precautionary augment, Kimball and Mastantuno define a comparative trait they term vigilance. One person is more vigilant than another if their precautionary augment is larger for every “unfair” risk—risks that worsen the individual’s prospects in a way that is not actuarially neutral. In practical terms, a more vigilant agent experiences a bigger rise in the marginal value of resources when confronted with downside-skewed or otherwise unfavorable uncertainty.

The paper then derives conditions under which one agent can be said to be unambiguously more vigilant than another, and it compares these conditions across preference structures. To make these comparisons operational, the authors show how to apply what they call the Diffidence Theorem using a geometric approach—turning an abstract ordering of attitudes toward risk into something that can be checked through the shape and relationships implied by the underlying utility functions.

An additional payoff of the framework is that it clarifies a closely related notion: risk vulnerability, a classic concept in the precautionary saving literature. The authors show that risk vulnerability corresponds to declining vigilance—that is, circumstances in which adding background risk reduces (rather than increases) the precautionary augment associated with other risks. With that equivalence in hand, several established results about risk vulnerability follow naturally as corollaries of the new vigilance-based analysis.

The authors note that the views presented are their own and do not necessarily represent those of the National Bureau of Economic Research (NBER). The NBER also highlights that, alongside working papers, it circulates research through outlets including The NBER Reporter, The NBER Digest, and topic-focused bulletins on health, Alzheimer’s economics, and entrepreneurship, as well as research spotlights, lectures, and panel videos.

Why it matters

Households and investors often cannot fully insure personal risks such as job loss, health shocks, or business income volatility, even in sophisticated financial markets. By tying those uninsurable risks to saving incentives, the risk-free rate, and asset prices through a single measurable concept, the paper offers a unified way to think about how “background risk” can ripple into macro-finance outcomes. What to watch next is how these theoretical rankings of vigilance can be mapped to empirical measures of preferences and used in quantitative asset-pricing or policy analysis.