Gold continues its free fall

(MoneyWatch) Gold is still heading south, but it may still make sense for your portfolio. You just have to view it correctly and be vigilant about your exposure, as I've been saying all along.

On Sept. 5, 2011, gold closed at an all-time high of $1,895, a cumulative rise of 581 percent from its close of $278 at the end of 2001. Despite that meteoric rise, or perhaps because of it, if there was one thing many investors and prognosticators were sure of, it was that the Federal Reserve's easy monetary policy would lead to much higher prices. In fact, a 2011 Gallup poll found that 34 percent of Americans said gold was the best long-term investment, much more so than real estate, stocks or bonds.

Yet yesterday, gold prices closed at $1,278, dropping 7 percent from the previous day to its lowest closing price in more than two years. That's a fall of $617 and 33 percent from their peak.

It's hard to blame investors for getting caught up in gold's rise, especially when they hear predictions such as these:

• In 2008, Peter Schiff predicted that gold would hit $2,000 by 2009 and $5,000 by 2013.
• In July 2012, Merrill Lynch predicted gold would hit $2,000 an ounce by the end of the year.
• In January 2013, at the annual Barron's roundtable, Bill Gross stated: "Investors can choose between artificially priced financial assets or real assets like oil and gold, or to be really safe, cash. ... My first recommendation is GLD -- the SPDR Gold Trust."

Readers of my blog know that while my crystal ball is always cloudy, we did provide several warnings that there were strong reasons to believe that gold's price could just as easily collapse as rise. Here are some of those warnings.

Jan. 14 -- Gold closed the previous week at $1,658. We warned: "The marginal cost of production (or the cost of producing one more unit than currently being produced) is around $750 per ounce, or less than half its current price. The laws of economics state that prices tend to move toward the marginal cost of production over the long term. After all, producers will produce more and more if there's profit to be had. Such a gap between the cost to produce gold and the price is a pretty big incentive to up the production. All else being equal, prices should eventually move toward the marginal cost of production as supply increases. This is especially true for gold, since unlike other commodities, nearly all the gold that has been produced is still in circulation and potentially available for sale."

Feb. 18 -- Gold closed the previous week at $1,612. We provided reasons why it was possible that the gold rally might be over: "They [smart investors] know the risks are there that if we do have a shift in monetary regimes to a tightening of policy to prevent the inflation we all fear gold could easily repeat its performance of 1980-2002. In other words, it could collapse even faster than it rose. It would be like pricking the proverbial bubble." And we noted that "[m]arkets look forward. They anticipate, moving prices even before events occur."

April 23 -- Gold closed the previous week at $1,406. We again cautioned that the fall in the price of gold might be just the beginning by providing this reminder: "For those who know their history, the drop is reminiscent of what happened to gold after hitting its peak close of $850 on Jan. 21, 1980. By April 3, 1980, it had dropped all the way to $486. It recovered to $590 by year's end, but closed the following year at just $398 and languished for the next 20 years. In real terms, the price of gold fell about 85 percent from its peak before bottoming out in 2002."

I point this out not to say I called gold's quick fall. The price of gold could just as easily continue its downfall. Research by Claude Erb and Campbell Harvey shows that the price of gold tends to revert to a real return of zero. This suggests that $800 is a reasonable target when applying the metric of zero real return, which gold has provided for centuries. That happens to coincide pretty nicely with Goldman Sachs' estimate of production costs. (Such a return depends on the Fed's unwinding its easy monetary policy before unexpected inflation rears its ugly head. Right now, the market seems to be anticipating low inflation, as the breakeven rate on 10-year Treasury inflation-protected securities is only about 2 percent.)

On the other hand, it's possible that there's validity to the claim that gold is underowned by investors (such as central banks) and that even a small shift in demand could lead to a significant rise in the real price of gold, assuming supply is inelastic (a big assumption). Keep in mind that almost all the gold ever produced is available for sale. And as Dimensional Fund Advisors' Weston Wellington pointed out: "It's also conceivable that a significant real price increase would encourage development of electrochemical extraction of the estimated 8 million tons of gold contained in the world's oceans, dwarfing the existing gold supply."

Since we obviously can't predict the price of gold, the proper way to view such an investment would be in the context of its effect in a portfolio. My advice on investing in gold remains the same. If you invest in gold because you value the diversification benefits gold can provide -- hedging the risks created by loose monetary policy and some geopolitical events -- it's appropriate to consider having a small part of your portfolio in gold as part of your asset-allocation plan. Just be sure that you are a disciplined rebalancer, not one who gets caught up in the noise and emotions the noise can create. That's the only way you get gold's diversification benefits. Otherwise, you'll end up like most investors who are "returns chasers" -- causing them to buy high and sell low.

On the other hand, I prefer that allocation be made to a broader based commodity index such as the DJ-UBS commodity index. A broader index provides a much better hedge against inflation.