Signed small trade turnover (SSTT) measures temporary uninformed buy or sell pressure that is initiated by small trades in the same direction. Using Nasdaq OMX Helsinki tick-by-tick trade data with known investor category, we confirm that SSTT is a robust proxy for uninformed trading. In the short term (1-3 months), stocks with a high proportion of signed small trades outperform, but in the medium (4-6 months) and long term (7-36 months), SSTT by individual investors has a negative correlation to stock returns, while SSTT by institutional investors and foreign nominees is not related to stock returns. Systematic trading behavior appears to better explain the excess return generated by the low SSTT portfolio relative to the high SSTT portfolio when compared to traditional risk factors in the CAPM and Fama-French models. In the aggregate, small trades are noise in the spirit of Kyle (1985) and Black (1986), but small trade behavior impacts the performance of individuals and not the performance of institutions. Large trades by both individuals and institutions perform better in the intermediate and long term, which lends some credibility to the practice of classifying larger trades as informed, but our results show that these trades are not always by institutions.