Swedroe: Debunking Gold Mythology

In mid-July 2012, Merrill Lynch added its voice to the many that were predicting gold would hit $2,000 an ounce by the end of that year. Francisco Blanch, head of global commodities research at the investment bank, said: “We think that $2,000 an ounce is sort of the right number.” Gold was then trading at about $1,577.

At about the same time, in an interview with ETF.com, money manager Peter Schiff, who has attracted much media attention with his doomsday forecasts, offered up this prediction: “I’m looking for another leg up … it’s going a lot higher. It’s hard to tell where the next move is going to take it. But it’s going thousands of dollars higher than it is now.” When asked how high, he responded: “I think a minimum of $5,000. But it could go a lot higher than that.”

These kinds of predictions almost certainly helped drive investor interest in gold. In fact, a 2011 Gallup poll found that 34 percent of Americans said gold was the best long-term investment, far more than those who chose real estate, stocks or bonds.

The question is: Were, and are, individuals investing in gold for the right reasons? Well, one reason for investor interest in gold is the belief that it’s a great inflation hedge. Another is that it provides a hedge against currency risk. And a third is that gold can act as a haven of safety in bad times. Are these valid reasons?

The Evidence

In their June 2012 study, “The Golden Dilemma,” Claude Erb and Campbell Harvey examined these issues. In terms of being a currency hedge, they found that the change in the real price of gold seems to be largely independent of the change in currency values. In other words, gold is not a good hedge of currency risk.

As for gold serving as a safe haven, meaning that it’s stable during bear markets in stocks, Erb and Harvey found gold wasn’t quite the excellent hedge some might think. It turns out that 17 percent of monthly stock returns fall into the category in which gold is dropping at the same time stocks have negative returns. If gold acted as a true safe haven, then we would expect very few, if any, such observations. Still, 83 percent of the time on the right side isn’t a bad record.

In terms of gold’s value as an inflation hedge, the following example should help provide an answer. On Jan. 21, 1980, the price of gold hit a then-record high of $850. On March 19, 2002, gold was trading at $293, below where it was 20 years earlier.

Note that the inflation rate for the period from 1980 through 2001 was 3.9 percent. Thus, its loss in real purchasing power was about 85 percent. How can gold be an inflation hedge when, over the course of 22 years, it loses 85 percent in real terms?

As further evidence of gold’s inflation hedging abilities, Goldman Sach’s “2013 Outlook” contained the following finding: In the post-World War II era, in 60 percent of the episodes when inflation surprised to the upside, gold underperformed inflation. That said, gold has been a good hedge of inflation over the very long run (such as a century). Unfortunately, that’s a much longer investment horizon than that of most investors.

Updated Findings
In August 2015, Erb and Harvey updated their paper. They begin by examining the argument that gold is an inflation hedge, or what they call a “golden constant.”

The authors explain: “One way to think about the golden constant perspective is as a collection of statements that assert that: 1) over a very long period of time the purchasing power of gold remains largely the same; 2) in the long run, inflation is a fundamental driver of the price of gold; 3) deviations in the price of gold relative to inflation will be corrected; and 4) in the long run, the real return from owning gold is zero.”

The study covered the period beginning in January 1975. The authors found that, over the period, the average real price of gold is 3.46 times the U.S. Consumer Price Index (CPI). In June 2015, the CPI level was 237.8. Multiplying gold’s average real price by the current CPI (3.46 x 237.8) delivers a price of approximately $825. This represents what the nominal price of gold should be today—if we assume the real price of gold is constant.

Of course, over time, prices have strayed far from the golden constant. And, as Erb and Harvey note, the golden constant isn’t a fact, just a hypothesis.

But if your reason for buying gold is that it’s an inflation hedge, your expectation should be that gold will revert to its golden constant over time. And despite haven fallen from its peak of almost $1,900 in September 2011 to stand at about $1,100 as I write this, it’s still about 20 percent above the golden constant.

When Gold Deviates From Its Constant
Erb and Harvey then asked: If the golden constant provides a guide to the value of gold, what typically happens when the price of gold is above or below its golden constant value? They found that the high real price of gold has been about 8.73, the low real price of gold has been about 1.47, and that the current real price of gold is about 4.63.

The charts they present show that while there is a tendency to revert to the golden constant, the price of gold can vary greatly from the golden constant, and stay well above or below the constant for a long time. And as they note, there is no way of knowing if the “future high and low real prices of gold may be more or less extreme than in the past.”

The authors add: “The high and low real prices of gold highlight that even if there is on average a golden constant the real price of gold has strayed, and probably will stray, far from this possible central tendency. It is also possible that the future will be unlike the past.”

They thus warn that if the “real price of gold falls, the golden constant level is not a floor—a protective line in the sand that the real price of gold will not cross.” With this caveat, they did go on to examine the outlook for gold returns given where the price is relative to the golden constant.

Expected Returns To Gold

Erb and Harvey examined what happened to the return on gold when prices were above or below the golden constant. As you might expect, they found that “below average real gold prices have been followed by above average 10-year real gold returns and above average real gold prices have been followed by below average 10-year real gold returns.” Because the real price of gold is currently above its historical average, this “suggests that over the next 10 years real gold returns could be below average.”

Erb and Harvey also looked at the downside risk of owning gold. To do so, they asked the question: How low might the price of gold go if the previous low real price of gold is revisited? Given the value of the U.S. CPI for June 2015 and the previous low real price of gold, a possible low price for gold is about $350 an ounce.

This, of course, does not mean that the price of gold will immediately decline to $350 an ounce. Rather, it’s a suggestion that, given the volatile history of real gold prices, because the real price of gold once fell to 1.47, it could fall to that level again.

A Return To Highs
The authors also examined what would happen if gold went back to its previous highest real price. If that occurred, it means the price of gold would again have reached about $2,080.

Erb and Harvey then looked at what would happen to real and nominal returns on gold if we assumed inflation of 2 percent a year for the next 10 years. Why 2 percent? It’s roughly the difference between the yield on 10-year Treasurys and 10-year TIPS, as well as similar to the consensus forecast of economists gathered by the Federal Reserve Bank of Philadelphia.

They found that the golden constant value of gold would increase from $825 an ounce to $1,006 an ounce, and the “overshoot” price would rise from $350 an ounce to $427 an ounce. If, over a 10-year investment horizon, the price of gold fell from $1,096 an ounce to $1,006 an ounce, it would experience a nominal return of ‐0.9 percent per year and a real return of ‐2.8 percent per year.

If the price of gold dropped from $1,096 an ounce to its 10-year “overshoot level,” the nominal and real returns would be ‐9.0 percent per year and ‐10.8 percent per year, respectively. Keep in mind that, regardless of the future inflation rate, the real rate of return is ‐2.8 percent per year if gold falls to its golden constant fair value over the 10-year period.

Erb and Harvey concluded that, even though there is little relation between the nominal price of gold and inflation when measured over 10‐year periods, the evidence suggests that gold does hold its value over the very long run.

For example, in a prior paper, they presented historical evidence that the wage of a Roman centurion (in gold) was approximately the same as the pay earned by a U.S. Army captain today. They also showed that the price of bread (in gold) thousands of years ago is about the same as we would pay today at an upscale bakery.

Summary

The conclusion we can draw is that, while gold might protect against inflation in the very long run, 10 years is not the long run. As Erb and Harvey note: “In the shorter run, gold is a volatile investment which is capable and likely to overshoot or undershoot any notion of fair value.”

I’d add to that another insight that becomes important in the long term. While the laws of economics can be defied in the short term, history demonstrates that investors ignore them at their peril. For instance, a basic economic principle is that, over the long term, prices tend to move toward the marginal cost of production.

In its “2013 Outlook,” Goldman Sachs estimated that the marginal cost of producing gold is less than half the current price (around $750 an ounce). The financial services firm also observed that more than 80 percent of gold production costs less than $1,000 an ounce—or about 10 percent below the current price. Another important point to consider is that, unlike with other commodities, all the gold that’s ever been mined is basically available for sale today.

And, as Dimensional Fund Advisors’ Weston Wellington recently 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.” That’s a lot of supply that could potentially hit the market.

The bottom line is that, while my crystal ball always remains cloudy, based on the fundamentals and the historical evidence, there doesn’t really seem to be a case that gold is likely to provide strong investment returns, even though it has already fallen about 40 percent from its peak nominal value (and even more in real terms). Forewarned is forearmed.

If you have been considering an investment in gold—perhaps you see the 40 percent drop from its high as a buying opportunity—hopefully the information in this article will enable you to make a more informed decision.

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.