Supply chains have evolved into ever more complex networks of buyers and sellers over time. A key property of these networks is the interconnectedness of firms. This property is a double-edged sword – while firms benefit from participating in such networks, their very interconnectedness to other firms also exposes them to additional risks. These risks include exogenous shocks brought about by natural disasters and pandemics. Todo and Inoue (2021) provide a review of the empirical evidence of how supply chain participation and diversification affect firms when they experience such shocks (directly or indirectly). The lessons drawn from the literature make sense – reducing the degree of supply chain concentration – whether through increasing substitutability and diversification – can improve the robustness and resilience of supply chain firms. An intriguing finding highlighted by Todo and Inoue (2021) is the impact of geographical diversification of supply chains on the robustness and resilience of firms in the short term and medium/long term. That locally concentrated supply chains are superior to more diffused ones in the medium/longer term seems to suggest that agglomeration economies could be an important element. There are hints in the paper, but the relationship between agglomeration economies and supply chain networks is not fully fleshed out. It is an essential point as agglomeration economies are not just a high concentration of local supply chains. This leads to the question of the micro-foundations of supply chain networks. Todo and Inoue (2021) have been pioneers in using agent-based models to simulate the propagation of shocks from natural disasters through supply chain networks in Japan (Inoue & Todo, 2019a, 2019b, 2020). The one-off impacts of these shocks have been studied using models in which firms are embedded in networks (calibrated with real data) and follow fixed rules on production, inventory as well as sale. These are essentially percolation models which are very useful in estimating the immediate damage from natural disasters. Modeling and simulating longer term effects of shocks will require more malleable firm behavior. Otherwise, the impact of natural disasters might be overestimated. An even more ambitious approach will require the modeling of entry, exit, growth (including through merger) of firms within a network framework. Without more dynamic micro-foundations, it will be difficult for the theoretical literature to catch up to the empirical literature on how firms embedded in supply chains adjust after experiencing shocks. Another point worth highlighting is the conceptualization of risks associated with exogenous shocks such as natural disasters and pandemics. Natural disasters are multidimensional – some aspects can be predicted generally (location – earthquake fault lines, general timing – seasonal typhoon) but not others (frequency and intensity). In terms of micro-foundation, how will/should firms optimize their supply chains given the difficulties involved in assessing these risks? Is there firm-level heterogeneity in these matters? If yes, in what sense? Todo and Inoue's (2021) policy recommendations follow directly from their analyses – greater geographical diversification (away from greater Tokyo and China but more toward developed countries) and state support for Japanese firms abroad in terms of subsidizing the fixed cost of relocation and information provision. These recommendations are likely to be debated in terms of the appropriate boundaries between the market and the state. A salient and critical policy decision during this period has been how to strike a balance between human cost and economic cost. The study by Inoue et al. (2020) is a good start but needs to incorporate the spread of Covid-19 that accompanies the lifting of lockdowns. Both have spatial dimensions embedded in networks that can be simulated. Finally, another important type of Covid-19 policy is state support for firms in the form of wage subsidies, low-interest loans or loan repayment deferment. How should such support be targeted? Given that large firms tend to have greater access to financial resources (internal and external) and small firms account for a larger share of total employment, it would seem sensible to channel more aid to smaller firms. However, the heavy-tailed distribution of supply chain network connections may point to a counterintuitive approach that pays more attention to larger firms with hub status.