Problem

From a time series of E-mini S&P 500 futures tick data,

(a) Apply the tick rule to derive the series of trade signs.

(b) Compare to the aggressor's side, as provided by the CME (FIX tag 5797). What is the accuracy of the tick rule?

(c) Select the cases where FIX tag 5797 disagrees with the tick rule.

(i) Can you see anything distinct that would explain the disagreement?

(ii) Are these disagreements associated with large price jumps? Or high cancelation rates? Or thin quoted sizes?

(iii) Are these disagreements more likely to occur during periods of high or low market activity?