Alan Greenspan joins the chorus.... Mainland China is in the middle of an equities bubble.
How long can bubbles last? In principle, for a long time. The reason is that it is very difficult to make money from knowing that the market represents a bubble, even if you turn out to be correct.
At the risk of displaying my ignorance of the nuances of financial markets, I'll develop this argument further. First a couple of disclaimers: I'm a macroeconomist, NOT a financial advisor. What follows is intended to stimulate discussion, NOT as investment advice. OK- on with the argument.....
Suppose Greenspan, Li Ka-shing, and Yetman :-) are all correct. To profit from this, one could short mainland equities, and generate income from their inevitable decline. But shorting equities is costly. If you're correct, you make a large profit; if you're wrong, and equity prices continue to rise, you loose 100% of the capital you spent shorting them. In the meantime, you've also lost out on the continuing share price appreciation by being absent from the market. So even though you will eventually be correct, you might have lost so much potential gain that you would have been better off staying in the market!
The result of this is that investors who believe that the market is a bubble might be better off keeping their positions while buying partial insurance against a market correction by shorting the equities, rather than pulling their investments out entirely. Thus investors are biased towards making bigger bets on continued market growth than might be optimal. So the bubble continues for longer than it should, and when it crashes, it falls further.
If I'm right, bubbles are more persistent and more economically damaging because of an underlying asymmetry in share price instruments: it's easier to make money from an expanding market than a contracting one. All we need is access to instruments that allow investors to benefit from a declining market more efficiently, reducing their incentives to bet against a market fall.
Consider, for example, "inverse shares" whose value moved in opposition to the market. If the share price increased by 2%, the "inverse share" would loose 2% of it's value. They'd provide an efficient way for the median investor to effectively purchase an entire portfolio of short positions that are currently only available to large, institutional investors.
I'm sure there are a million reasons this wouldn't work, and I'm looking forward to reading why in the comments!
(See also my earlier posts on bubbles here and here).