One of gold's most prominent bulls, John Paulson, the asset manager who made more than $1 billion betting on the housing downturn, is tarnished.

Bloomberg reported that Paulson's $900 million gold fund is down 26% through the beginning of March, after falling 25% last year. The fund has been hurt by the price of gold falling to around $1,600 off its all-time high of more than $1,900, which it hit in Sept. 2011. Paulson told clients that "his Gold Fund would beat his other strategies over five years because the metal was the best hedge against inflation and currency debasement as countries pump money into their economies."

Paulson echoes comments from Ray Dalio, the man behind the $140 billion hedge fund Bridgewater Associates. Dalio told Barron's in March 2011, "Currency devaluations are good for stocks, good for commodities, and good for gold."

The price of gold has fallen off those highs and, so best we can tell, another economic crisis isn't happening soon, meaning investors are less likely to flock toward the most prominent bearish investment. But gold's effectiveness as just that -- protection against the worst economic and financial distress -- is also under attack.

History
One of the most common arguments to invest in gold is inflation. Credit Suisse Senior Advisor Robert Parker told CNBC last month, "To get back into a bull market on gold we need inflation." Although there are different ways to interpret gold as an "inflation hedge," the simplest version is that changes in the consumer price index should drive changes in the price of gold: if prices (in dollars) rise 10%, then the price of gold should rise 10% as well, giving gold a constant purchasing power.

The historical record, however, suggests little relationship between the gold price and inflation. In a recent paper titled "The Golden Dilemma," Claude Erb and Campbell Harvey found that "over 1, 5, 10, 15 and 20 year investment horizons the variation in the nominal and real returns of gold has not been driven by realized inflation."

Looking at the price of gold and the consumer price index since 1975, when private ownership and trade of gold was reallowed in the U.S., they find that in March 2012 (when they last did the calculations), the price of gold "should" have been $780 an ounce, but the actual price was closer to $1,600. (It was right around $1,600 this morning.)

When it comes to whether gold can hedge unexpected inflation, the authors find "effectively no correlation" between the real price of gold and future inflation; any relationship is driven by 1980, when the real price of gold more than doubled and hit its all-time peak of around $2,440 (2013 dollars) and inflation was up above 12%.

After 1980, gold sharply declines and then starts moving sideways until 2003, while average annual inflation was around 4%. Even for 10-year periods, gold did little to nothing to hedge inflation, expected or unexpected.

Gold as a hedge, gold as a safe haven
What about as a currency hedge? Dalio and Paulson both reference quantitative easing and money printing as reasons to invest in gold, but, again, the historical record doesn't give much reason to hold gold.

Harvey and Erb found that the price of gold tends to move together in different countries and in different currencies: For eight major currencies, gold was high in the early 1980s, low in the 1990s, and high right now. And while they find a slight relationship between declines in the dollar exchange rate across eight major currencies, they're on the order of 1.5% in dollar depreciation against a given major currency for every 10% increase in the price of gold -- far from the strong connection you'd expect in a true hedge against a precipitous fall in the dollar.

Another popular argument for having some exposure to gold is that it is a safe haven or hedge against extreme financial and economic stress. This argument is most plausible based on recent experience -- from Sept. 2007 to Sept. 2011, the price of gold rose from $743 to more than $1,940.

But this is just one observation. The price of gold declined during the Asian financial crisis and subsequent market drop in Oct. 1997. Following the Russian debt default, gold declined again, and gold went up just $9 per ounce during the 2001 recession.

Harvey and Erb studied the monthly total returns of the S&P 500 and gold and found that 17% of monthly stock market negative returns were matched by negative gold returns. In the last year, gold has declined 4.8% while the S&P is up 13.5%. Just because gold is poorly correlated with U.S. equities does not make it a safe haven.

The roundabout case for purchasing gold, part 1
If Paulson and other gold bulls are wrong about why to own gold and gold-related assets, could gold still rise in the long run? Yes, for two reasons:

  1. If emerging market central banks decide to increase their gold holdings to the levels of their developed market counterparts.
  2. If it becomes conventional wisdom that the typical "market portfolio" should have a target gold allocation.

Central banks' net purchases of gold in 2012 were $29 billion, compared to $23 billion in 2011. Central banks' net purchases only turned positive in 2010. Rich countries and the IMF hold an overwhelming portion of existing central bank reserves (the U.S. has more than 8,100 tons, compared to just over 2,600 tons for Brazil, India, Russia, and China combined), but developing and middle-income nations make up a huge portion of new purchases.

For years, gold had been piling up in rich countries' central bank reserves as a leftover form the gold standard and Bretton Woods. They responded by managing the sale of their reserve gold so that the price would not fall too quickly. Switzerland and France, for example, have significantly decreased their gold holdings while Russia, China, and India have substantially added to theirs.

If this pattern -- rich-country gold reserves holding steady or falling slightly and developing countries' reserves rapidly expanding -- continues, then price-insensitive central banks might drive up the price of gold, or at least put a floor on it, for reasons that have nothing to do with the traditional investment rationale.

"As long as some central banks are insensitive to the real price they pay for gold, the possible move into gold could drive the real price of gold much higher," write Harvey and Erb.

The roundabout case for purchasing gold, part 2
According to the World Gold Council, in 2012 investors made up 36% ($82 billion out of $236 billion) of world gold purchases, the rest coming from jewelry (44%), technology (8%), and central bank purchases (12%).

Over the past decade, gold investors have not acted like typical asset buyers: They tend to buy more gold when the price is higher. Harvey and Erb have found that between 2001 and 2011, gold has had positive price elasticity for investors -- for every 10% the price has gone up, investor demand has increased 9.8%.

It seems like investors around the world are chasing momentum, believing that as the price of gold starts to go up, it will continue to do so. Which leads to the second reason the price of gold might stay elevated or continue to rise: It's underowned.

Dalio told Barron's, "If you ask any central bank, any sovereign wealth fund, any individual what percentage of their portfolio is in gold in relationship to financial assets, you'll find it to be a very small percentage."

If Dalio, one of the world's most successful investors, gets other asset managers to imitate him, gold could become a standard part of the market portfolio. Basic finance theory suggests that an investor wants to hold a mix of assets that reflects the market as a whole. So if the sum total of financial assets is 54% stocks and 46% bonds, the "market portfolio" is 54/46.

According to Harvey and Erb, using World Gold Council data, the actual mix right now is more like 53% stocks/45% bonds/2% gold, but it is likely that very few investors own that mix of assets.

If pension funds and endowments decide to pursue an asset mix that's more weighted toward gold, they will start buying up a limited amount of gold available to investors. This "would probably result in much higher nominal gold prices," because momentum-riding gold investors would chase the higher prices and price-insensitive central banks would keep on buying, argue Harvey and Erb.

That's because the gold supply hasn't been able to keep up with demand -- thus the high prices. In 2001, gold mine production was 2,600 metric tons. In 2012, it was 2,800, leading to the price going up six times (scraping and recycling round out the gold supply). There's just not much more mining companies can do to get gold out of the ground.

Gold doesn't do what it's supposed to -- it's really just shiny metal. But if central banks keep on buying it and investors believe in it, it might start to act like something more valuable. But that's only in the long run; in the short run, John Paulson's gold fund could still be dead.

For more on gold, see:

Matthew Zeitlin does not own shares of any companies mentioned.