Gompers, Gornall, Kaplan, and Strebulaev survey over 1,000 institutional and corporate venture capitalists at more than 900 firms on how COVID-19 affected their decisions, comparing responses to a similarly structured 2016 survey the same authors previously analyzed. VCs reported investing at 71% of their normal pace during the pandemic and expected to recover to 81% of normal over the following year, while devoting more time to guiding existing portfolio companies through the crisis, though the authors found little change in how VCs split their time between portfolio support and sourcing new deals. Portfolio impact was mixed rather than uniformly negative: 52% of portfolio companies were positively affected or unaffected by the pandemic, 38% were negatively affected, and 10% were severely negatively affected, and VCs expected only a small overall hit to fund performance, a 1.6 percentage point drag on IRR and a 0.07 reduction in MOIC, with only modest differences between institutional and corporate VC responses.

Founders navigating a future crisis get a genuinely reassuring benchmark from VCs themselves, since even during the acute early phase of a global pandemic, investors expected only a modest overall hit to fund performance and reported the majority of portfolio companies as unaffected or positively affected, contradicting the panic that often dominates headlines during a downturn. The finding that VCs devoted more time to guiding portfolio companies without meaningfully cutting time spent on new deal sourcing is useful for founders trying to gauge whether their existing investor will remain available for support during a crisis, and whether new fundraising conversations are likely to keep moving even during turbulent conditions. Because the study directly compares 2016 baseline responses against the pandemic period using the same survey instrument and largely the same authors, researchers get an unusually clean, methodologically consistent dataset for studying how a specific macro shock changes venture capitalist behavior, rather than relying on differently designed surveys that are harder to compare.