The problem with that sort of thinking is that it relies on you being able to time when when market will go from being irrational in one direction to being irrational in another.
Lets take the dot-com boom as an example. There were people saying that dot-com companies were overvalued in 1997, when the Asian currency crisis hit. Those people were, in retrospect, probably right. However, the next two years made them look like complete fools, as the bubble maintained exponential growth for another two years before completely collapsing.
Same thing with China. Yeah, the bubble might have another couple of years left. If you get in now and get out when the crash hits you'll make a modest profit. But, if you mistime the crash, you could stand to take huge losses, as the value of your currency investments falls far below the price you paid for them.
The original argument was that there was no point in investing, because if you'd invested 10-15 years ago, you'd have made a lot of money. That argument is wrong for obvious reasons, which I will spell out anyway; just because something did well in the past is no reason to think you'll lose money on it in the future.
Your argument is that if you invest in something and it goes down in value, you lose money. That's true but not relevant.
Eli's point seems to be that it's possible to make money on an investment after it's had an explosive growth stage; the obvious truth in counter to the initial falsehood, and in no way addressed by your comment.
I think that the point is that after the explosive growth stage, if something is a bubble you end up in a (possibly) short period before the bust. While you can make money investing during the period, the risk is that much greater that you will miscalculate when the bust will happen and take heavy losses. So:
1. Explosive growth is over, you can't make money anymore.
2. That's not true, you *can* still make money.
3. But attempting to make money at this stage is riskier.
Lets take the dot-com boom as an example. There were people saying that dot-com companies were overvalued in 1997, when the Asian currency crisis hit. Those people were, in retrospect, probably right. However, the next two years made them look like complete fools, as the bubble maintained exponential growth for another two years before completely collapsing.
Same thing with China. Yeah, the bubble might have another couple of years left. If you get in now and get out when the crash hits you'll make a modest profit. But, if you mistime the crash, you could stand to take huge losses, as the value of your currency investments falls far below the price you paid for them.