Awhile back, I asked what the next bubble would be. The answer from Harper's magazine (thanks, Louis!) was clean energy investments. That may be true (or may not -- since it hasn't happened yet). But it's possible that we'll see a bubble before that eventuality -- in gold. This might be financed by the Fed's expansionary monetary policies (the efforts to save the U.S. credit markets) while spurred by fears of inflation.
Assume that a gold bubble has started, i.e., that recent price upsurges do not just represent a short-term blip. If anyone wants to wants to buy and hold gold for long periods as a hedge against inflation, it's then too late, since prices are too high.
The exception to this outrageous assertion would be if the end of the "long period" is (by coincidence) either at the peak of the bubble or some time after the bubble pops when gold prices start rising again. (Gold prices should be measured in real terms, corrected for the ability of gold to buy actual goods & services, by the way. It's only when real gold prices rise that they're a successful investment.) As the bubble grows, the "too late" verdict will become more and more true.
On the other hand, if anyone wants to speculate by buying now (when prices are low compared to the future peak) and selling right before or at the peak price, that actually encourages the bubble, by increasing demand.
We can tell if a bubble is actually happening if we start hearing about how "gold is a perfect hedge" or "you can't lose by investing in gold" or "gold prices have nowhere to go but up." These kinds of statements reflect the hopes of the aforementioned speculators -- and their later efforts to ensure that their speculation pays off by driving up gold prices further.
Of course, it's a big gambling game. The gold-bugs who win at the peak (buying low and then selling high) have to find others who are willing to buy at the high price. (These are called the "greater fools.") Note that the winners' loot corresponds to the losses of those left holding the bag: as the bubble pops, the sellers drive prices down, imposing capital losses on the late buyers.
(In addition, the "house" wins to the extent that there are brokerage fees in the gold market.)
Of course, gold pays no interest or dividends. That makes the popping worse, since the only way to win in a bubble is to grab for all the capital gains one can.
J.M. Keynes's "betting on a beauty contest" theory of bubbles doesn't add anything to this discussion (because, unlike corporate stocks & bonds, gold is a well-known asset). But his "snap" analogy does say something. It's more familiar to U.S. residents as the "musical chairs" analogy.
As people begin to think that the bubble is reaching its peak size, their fear that the music will stop heightens. However, because everyone wants to avoid sitting down until a second before the music stops, they hold on to their gold. Then, for some reason -- the scary sight of an advertisment for beer projected on the Moon? -- some people start sitting down. This causes a panicked rush to the chairs. In other words, everyone tries to sell, driving gold prices down steeply.