Not sure whether it makes sense to average all returns and call it "negative expected returns" across the board. As with any game that mixes luck & skill (like poker), trading has a variety of expected returns. My guess (based on poker) is that a large part of people are small losers or break-even (in poker because of rake, in trading because of trading fees), some are big losers (in poker called whales), some are small winners (regs), and some are big winners (truly good players / traders).

It's not easy figuring out in the beginning which bucket you fall into, but if you accurately track your poker / trades over time, then with a large enough sample size (assuming proper risk management of course) you should be able to figure it out.