Predictable liquidity properties in a Segmented, inelastic stock market

We explore the predictive capability of two investment strategies on idiosyncratic volatility, liquidity risk and liquidity commonality, by investor type. Investors are characterized as positive-feedback or contrarian once their trades are significantly associated with daily stock returns on a given month. We find that this classification has predictive power: positive-feedback traders (mainly foreign investors) tend to increase, while contrarian traders (mainly local individuals) tend to reduce, the following month's volatility and liquidity. Different investor clienteles segment the market by stock characteristics, questioning linear cross-sectional pricing. Controlling for supply inelasticity we find that share issuance/buyback datapoints tilt some of the statistics and blur the findings.