Gold's recent meteoric rise above $3,400 might continue even further. But I know of no way to have much confidence that it will.

That's because none of the many theories advanced to explain gold's rise are objectively provable.

While you might find the lack of proof to be discouraging, Claude Erb contends that investors perennially face this dilemma. Erb is a former commodities trader at TCW (formerly Trust Company of the West) and co-author, with Duke University professor Campbell Harvey, of a 2012 study that coincided with a prior occasion when gold was soaring. That study concluded that gold's price would most likely fall, which it did - shedding more than 40% in nominal terms over the subsequent five years, and even more than that in inflation-adjusted terms.

In an interview, Erb told me that "investing is about believing in what is unprovable. There is no objectively provable rationale for why gold is trading where it is today."

While he wouldn't be surprised if gold suffered a similar fate in coming years as it did after that 2012 study, he was quick to say there's no way of knowing that for sure.

Erb is advancing an investment philosophy sometimes referred to as "epistemic humility" (Jason Zweig's description of the late Charlie Munger's approach), or what Howard Marks, co-founder and co-chair of Oaktree Capital Management, calls the "illusion of knowledge."

As you'll see from what follows, there are reasons to be cautious about each of the many popular theories about how gold will behave in the coming months and years.

Reversion to the mean

This was at the core of Erb and Harvey's argument in 2012 that gold would struggle. One helpful way of thinking about gold, they said, was to assume that the ratio of its price to the consumer-price index would not exhibit any long-term trends. That meant that whenever that ratio was well above the historical average, the ratio subsequently would decline and thereby cause gold's returns to be mediocre - or worse.

This argument certainly held true in 2012, as you can see in the accompanying chart, just as it did in 1980 and 2022. And the ratio is currently at a record high of over 10, versus a 50-year average of around 4.

Nevertheless, the statistical significance of this argument is weak. Although there is an inverse correlation between changes to the gold/CPI ratio over the trailing and subsequent 10 years, that correlation can explain only 7% of that subsequent 10-year change (as measured by a statistic known as the r-squared). That means that more than 90% of gold's relationship to inflation over a 10-year period has nothing to do with what it was over the trailing decade.

Inflation hedge

As revealed by this discussion of the reversion to the mean argument, gold also leaves much to be desired as an inflation hedge. Though Erb and Harvey found some evidence that gold is a good inflation hedge over very long periods (when measured in decades or longer), gold's real, or inflation-adjusted, price is very volatile over the short and intermediate terms.

In an email, Harvey wrote that this result is exactly what you would expect "when using an asset with 15% volatility [gold] to hedge something that has less than 2% volatility [the CPI]." Given those relative volatilities, it's a foregone conclusion that "the hedge will be unreliable," he said.

Hedge against stock bear markets

Another theory that some are advancing to explain gold's strength is that it's a good hedge against a major decline on Wall Street. And while gold has advanced during some equity bear markets, it has declined during others. In the 12 bear markets since 1980 in the calendar maintained by Ned Davis Research, gold rose in half of them and fell in the other half.

(I focused only on the decades beginning in 1980, since gold didn't begin freely trading in the U.S. until the mid-1970s, and gold's price rise in the latter half of that decade presumably represented the impact of pent-up demand.)

Hedge against dollar devaluation

Another theory that has some plausibility is that gold is a good hedge against a lower U.S. dollar. Once again, however, there is only weak historical evidence for this theory. Since 1980, the correlation between monthly changes in gold and the U.S. dollar index DXY has an r-squared of just 7%, which means that changes in the dollar's value vis-à-vis major foreign currencies are able to explain just 7% of contemporaneous changes in the price of gold.

Hedge against the dollar losing its reserve-currency status

Another plausible theory for gold's rise is that investors, especially foreign central banks, are beginning to question the dollar's status as the world's primary reserve currency. There is even less data on which to test this theory than the other theories mentioned above, however, since concern about the dollar's reserve currency status is such a recent phenomenon.

Some date this concern to Western governments' confiscation of Russian central-bank funds following that country's invasion of Ukraine in 2022. Others date the concern more recently, to the so-called Mar A Lago Accord, which openly speculated about forcing foreign central banks to swap their holdings of U.S. Treasurys for century bonds that might pay no interest. If that were to happen, it would be the functional equivalent of default by the U.S. government.

That would certainly scare foreign central banks, of course, and presumably would increase their interest in gold. But there is little evidence that foreigners are fleeing U.S. Treasurys en masse, according to Joseph Kalish, chief global macroeconomic strategist at Ned Davis Research. Earlier this week, for example, he pointed out that there was "record foreign demand for the 10-year [Treasury] at the latest auction." Nor are there "any real signs of financial stress" in the U.S. Treasury market, as judged by the Bloomberg Government Securities Liquidity Indexes and the Office of Financial Research Financial Stress Index.

What about data that suggest that China has significantly increased its holdings of gold? Erb believes that such data may very well be unreliable, since there is no way to independently verify the data, which comes from the Chinese government.

"Since there is no reliable information on the gold holdings of countries that are antagonistic to the U.S., there is no way to measure dedollarization," Erb wrote in an email. And "without an ability to measure dedollarization, there is no way to improve our understanding of the impact of dedollarization on the price of gold other than to believe that dedollarization could impact the price of gold."

The bottom line? Betting on a gold bull market is and always has been speculative. That seems even more true today.