This study documents the liquidity-supplying behavior of proprietary algorithmic traders during the abrupt and sustained market decline caused by the COVID-19 outbreak. The findings suggest that these endogenous liquidity providers reduced their supply of liquidity during sustained market stress that lasted several days. Proprietary algorithmic traders showed a greater propensity to trade via market orders, reduced the fraction of contrarian trades, and reduced their share of order book depth compared to other traders during the in-COVID period. Our work provides the first direct evidence of the behavior of proprietary algorithmic traders during the pandemic.