Gold investors have taken it for granted that there is no mystery behind the surge in gold prices over the past year. They have no shortage of theories to explain the rally. However, MarketWatch columnist Mark Hulbert wrote, "To my knowledge, none of these theories has a statistically significant record of being able to forecast the gold price. Therefore, gold's strong performance in 2025 actually has no real explanation. In other words, it remains a mystery." This mystery should worry gold bulls. Without a convincing statistical explanation for why gold has risen so sharply, it is impossible to predict how it will perform this year. The following is a list of popular theories believed by gold bulls. Hulbert says none have consistently served as a coincident indicator, let alone a leading one. 1. Gold is a hedge against inflation. Perhaps the most common explanation for gold's moves is that it serves as an inflation hedge; its price rises when inflation heats up and falls when it cools. But as a guide to gold's short- and medium-term direction, inflation data is inadequate. Consider the 12-month change in the annual CPI rate versus the corresponding 12-month change in the price of gold. Then consider the R-squared of the correlation between these two data points. (R-squared measures the extent to which changes in one data series can explain and predict changes in another.) Based on data from the past four decades, as shown in the chart, the R-squared in this case is just 1.1%. This means that the 12-month change in the annual CPI rate explains only 1.1% of the 12-month change in gold. Unsurprisingly, this low R-squared is not statistically significant at the 95% confidence level that statisticians often use to determine if a pattern is genuine. This is tantamount to saying that changes in the annual CPI inflation rate are nearly useless as a short-term gold timing indicator. 2. Gold is a hedge against *expected* inflation. Some gold investors argue that gold does not react to changes in the actual inflation rate, which is only known after the fact, but to changes in *expected* future inflation. To test this possibility, Hulbert used the monthly forecasts from the Cleveland Fed's inflation expectations model. He found no significant correlation between these forecasts and the gold price. In fact, the explanatory power of changes in expected inflation—whether for 12 months or 10 years—was even weaker than that of the annual CPI rate. This is not to say that inflation has no relationship with gold. But history suggests this correlation exists only when looking at very long periods—decades, if not centuries. This was the finding in the study "The Golden Dilemma," co-authored by Duke University finance professor Campbell Harvey and Claude Erb, former commodities portfolio manager at TCW Group. 3. Geopolitical risk. Another popular belief is that gold hedges against geopolitical risk; its price rises when risk increases and falls when it decreases. To test this theory, Hulbert used the Geopolitical Risk (GPR) index constructed by Dario Caldara of the Federal Reserve Board's Division of International Finance and Matteo Iacoviello. They calculate the index by counting the number of articles related to adverse geopolitical events in each of 10 major newspapers each month, as a proportion of the total number of news articles. "I again came up empty. The 12-month change in the GPR index explained just 0.1% of the corresponding change in gold—so low it's barely visible on the chart above," Hulbert said. 4. Economic policy risk. A similar conclusion was reached when testing whether gold could serve as a hedge against economic risk, as measured by the Economic Policy Uncertainty (EPU) index. Once again, no correlation was found. The EPU index is measured by counting the frequency of newspaper articles that mention the economy, government regulations and policies, and uncertainty. The 12-month change in the EPU index explained just 0.9% of the corresponding change in gold prices. 5. Chinese gold purchases. Another widely circulated theory links gold's recent bull market to buying by the Chinese central bank. According to the World Gold Council, China's gold reserves have more than tripled since 2000. While this explanation sounds plausible, it provides poor guidance for gold's short- or medium-term trends. The R-squared between the 12-month change in China's gold reserves and the 12-month change in gold was a mere 0.6%. 6. Gold ETF net inflows. The data series with the highest correlation to the gold price was the 12-month net inflows into physical gold ETFs. This is not surprising, as ETF inflows cause them to buy more gold, and vice versa. However, even here, the correlation between inflows and the gold price was not statistically significant at the 95% confidence level. Coincident Indicators vs. Leading Indicators. None of these theories provide a solid coincident indicator for gold's fluctuations, and it is unsurprising that they all fail as leading indicators. This is one reason why timing the gold market is so difficult. Hulbert wrote, "This helps explain why the dozens of gold market timing strategies monitored by my performance auditing firm have significantly lagged the market. Based on all the rolling 10-year data I have since the mid-1980s, the average gold timing strategy has underperformed a buy-and-hold gold portfolio by 4.0 annualized percentage points. The conclusion? Investing in gold is risky—especially when there is no reliable basis for explaining what drives its behavior."
Comments