A decade ago, Gordon Brown, then the British Chancellor of the Exchequer, declared that Her Majesty's government would sell almost 400 tonnes of the central bank's gold reserves.
The timing of the decision - marking the trough of the bear cycle - was perfect in a way only governments can manage. Britain sold its gold holdings for well less than $US300 an ounce, raising £3.5 billion. Today, Brown's liquidation would raise almost £20 billion ($29.6 billion).
If governments are good contra indicators of when to buy and sell gold, what should one make of the fact that central banks - the world's single largest gold depositories - are buying once again?
What started in 2010 as a trickle has since become a flood.
Central banks bought 77 tonnes of gold in 2010; last year they bought 456 tonnes.
China, the world's largest producer, is by all accounts buying every ounce it can grab. Does this indicate what many suspect - that gold is a huge asset bubble?
Central banks have embarked on the biggest gold purchase in almost 50 years because no one has ever defaulted on gold and people have lost faith in fiat currencies and government debt.
But here's the paradox: If sufficient investors absorb that belief and act on it, the price of gold should rise beyond the point where the very protection being sought evaporates.

Investment driven
There's no sure way to tell whether we've reached that point already.
Unlike other commodities, demand for gold is investment-driven, which makes the price fickle.
There is, however, a clue about the sustainability of prices in the profitability of those that mine it.
Iron ore, coal and oil markets have all reached bubble territory in the past few years.
In each case, the vital clue that it wasn't different this time was the super profits being generated. When oil reached nearly $US150 a barrel in 2009, for example, Exxon Mobil's return on equity rose (ROE) to almost 40 per cent. Just a few years earlier it was 12 per cent.
Three years ago, Exxon made twice as much profit as the next-most profitable company in the world, also an oil producer. Then, as it became irresistibly profitable to pump more supply, prices wilted.
It's a similar story for big miners transformed by booming commodity prices. BHP Billiton's return on equity rose from less than 10 per cent in 2002 to almost 50 per cent at the height of the boom.
As its profits soared, so did the temptation to increase output.
Every boom thus sows the seeds of its own demise. Can we say that of gold now? At more than $1500 an ounce, gold prices are near record highs. If this is bubble territory, we should expect gold producers to be making stupendous profits.
And yet they aren't. Of the four largest global gold miners - Barrick, Newmont, Goldcorp and Newcrest - just one generates return on equity of more than 10 per cent.
Barrick's ROE, at 19 per cent, reflects the liberal use of debt rather than extraordinary performance. No other big gold miner generates ROE in double figures.
If there are no super profits, can there really be a bubble?

Smaller miners
Although we can't know for sure, it's unlikely. Big increases in output and lavish new production facilities are simply absent in the gold industry. Most big miners will add only incremental production, if they do so at all. The big gold miners aren't behaving as if gold is in bubble territory. Moreover, big falls in the price of gold would result in cuts to profitability and eventually, supply, which should restore higher prices.
Low profits also point to another culprit of high prices: higher costs.
We estimate the marginal cost of production has risen from about $300 an ounce a decade ago to more than $1000 an ounce today.
Although there are good reasons for higher prices, gold is still speculative. Higher prices rely on central banks continuing to expand their balance sheets.
That's a pretty good bet, but it's no certainty.
The world's biggest gold miners are custodians of terrific assets. Newmont, for example, holds the marvellous Carlin Trend all to itself, a formation that has yielded about 50 million ounces of gold.
Yet for all these riches, costs are high and returns meagre.
Above all, replacing millions of ounces of production each year is a real challenge. Rather than stump up cash for the biggest producers, investors who want gold exposure are better off looking for smaller miners with lower costs and growing production.
Nathan Bell is the research director at Intelligent Investor, intelligentinvestor.com.au. This article contains general investment advice only (under AFSL 282288).