The paper analyses the role of loan guarantee programmes following the COVID-19 outbreak in alleviating firm distress as well as their broader impacts on productivity via reallocation, relying on a simulation model and econometric estimations. The simulation exercise relies on a simple cash-flow accounting model, a large dataset reporting balance sheets of firms located in 14 countries and granular data on the magnitude of the COVID-19 shock. Our findings suggest that i) the COVID-19 shock had the potential to seriously distort market selection; and ii) policy actions corrected up to 30% of the inefficiency of market selection in the short-term, shielding many high productive firms from distress and supporting zombie firms only to a limited extent. The econometric exercise, based on historical data and standard models of dynamic allocative efficiency, examines how loan guarantees may shape the efficiency through which resources are allocated across firms of different productivity levels over the medium-term. Results suggest that, over the 2007-2018 period, increases in large-scale loan guarantee schemes were associated with weaker reallocation of credit and labour from low to high productivity firms. However, these effects are found to be more benign in intangible-intensive sectors and even positive for smaller scale programmes. Overall, engineering an effective exit strategy from these schemes, preserving their benefits while reducing their drawbacks through a gradual and state contingent phasing out, is critical to foster the recovery of the corporate sector. Further, monitoring debt overhang risks and facilitating firms’ entry and digital diffusion are relevant complementary challenges to address once COVID-19 related support is withdrawn.