1) You have to properly take into account other losing HFT traders in the sample.Quote from Tsing Tao:
----question.
----average aggressive high-speed trader....
----who is this money coming from?
2) Was it a "biased" data sample that excluded some unprofitable HFT traders in order to produce a more "sensational" profit number?
3) A "smaller" HFT trader has much less overhead than a "larger" HFT trader who has scads of programmers and expensive office space. The smaller guy can get by with a lower rate of return on capital where the larger guy can be unprofitable at the same level.
