In "Gold: Assumptions vs Reality," Moses Kim debunks the notion that a drop in jewelry demand means the gold price will fall. As it turns out, forecasting based upon that assumption would have led to being wrong nine years out of ten (he says.) Although it may be iffy, investment demand has a stronger correlation to gold prices. Kim ascribes this disconnect to gold being one of the few goods that sees increased demand go along with increased prices.
In other words, gold is bubble-prone: that's what rising demand caused by rising prices implies. Perhaps that's why "gold" and "bubble" are put together so much these days.