Was just wondering....does anyone know?
I believe it depends on the nature of the EA, but in most cases yes. If it's a quality backtest and looks back at a large amount of data (at least a few years) and shows a steadily declining equity curve, chances are it won't be profitable in forward tests either.
It certainly cannot be a good omen.
What about if you reverse trade order logic. Maybe then line will go up and you will get profit. It could work.
I found that NOT to be the case in most EAs I've been testing so far. If an EA is generating loss it usually will do that even if you reverse trade logic. I don't remember ever testing an EA that would benefit from it and turn into profit maker.