Quote from flat5:
The sign of the expectation of a bet is one way to look at it, but I don't think it's the common view. I play irrefutably positive expectation poker and I'll be damned if I can convince even a few people that it's not gambling.
Keeping it in quantitative terms but trying to match that up to how people "feel" about it, I think it has more to do with the magnitude of the "risk of ruin," (suitably defined) which of course still exists for positive expectation bets, and is a function not only of expectation but also the ratio of bet size and bankroll, as we call it in the gambling world. I believe you call that "money management."
I think people associate gambling with any activity with a significant risk of ruin, even if it has positive expectation. If I can bet $1 to win $100 on 1:98 odds, and have a $3 bankroll to work from, I think most people would think of that as gambling, even though the expectation of each bet is positive. Because I'm very likely to go belly up if I play the game.
By this logic, anything with risk of financial loss is gambling. So real estate development, banking, insurance, venture capital, no win/no fee legal work, running for political office etc is all "gambling". Clearly it's more complicated than that.
People view something as gambling when you are risking money to make money, with no perceived "productive" activity underlying the money-making. So real estate development is not gambling, despite being very high risk, because you are building something concrete (excuse the pun). Correctly anticipating the cards someone is holding in a game is not viewed as "productive".
Now it becomes clear why ignorant people view trading as gambling - they do not understand the productivity of liquidity provision or efficient pricing of assets. They think trading is a zero sum game. To be fair to the public, many traders and financial market participants are equally ignorant and share their view.