In today's markets where high frequency traders (HFTs) both provide and take liquidity, what influences HFTs' liquidity provision? I argue that information asymmetry induced by liquidity-taking HFTs' use of machine-readable information is important. Applying a novel statistical approach to measure HFT activity and using a natural experiment of index inclusion, I show that liquidity-providing HFTs supply less liquidity to stocks that suffer more from this information assymmetry problem. Moreover, when markets are volatile, this information asymmetry problem becomes more severe, and HFTs supply less liquidity. I discuss implications for market-making activity in times of market stess and for HFT regulations.