top of page

The Cost of Resilience

Rom Avissar
21/11/26

Each month, households that face income uncertainty must decide how to divide what remains after essential spending between savings, debt repayment, and investment. Yet, the advice these households commonly receive comes in the form of fixed rules that do not consider income or market risk. In the paper, the allocation of a household’s monthly surplus across savings, debt repayment, and investment is first formulated as a two-stage stochastic program with a CVaR term. Second, it calibrates the model to three archetypes using federal microdata: income level and annual transitions come from the CPS, monthly volatility comes from bank-account data, debt and credit denial come from the SCF, and essential spending comes from the CEX. Third, this paper uses a benchmark comparison on identical income paths to evaluate the optimal allocation policy against the five heuristic policies. In doing so, it measures the expected wealth that the household forgoes in attempting to gain financial security; this is called the cost of resilience. This investigation’s central finding is that a buffer held in the form of savings reduces the expected net-worth gain of a household, while getting no additional protection from bad income and market months. Within the optimal policy, the surplus is almost never allocated to savings, and investing the entire surplus is statistically indistinguishable from it. Over two years, three of the four heuristics that hold savings give up between $4,100 and $11,400 of expected net-worth gain. Since greater caution increases the expected cost without reducing CVaR no cost of resilience frontier can be traced.

bottom of page