Spot silver, which set an intraday record high of $121.79 per ounce earlier this year, has now fallen to around $57 per ounce. Since the end of January, spot silver prices have dropped by more than half, while spot gold prices have corrected by about 25% over the same period. Why has silver fallen harder than gold?
Different pricing anchors
Yuan Zheng, a precious metals researcher at Galaxy Futures, told China News Service that the core reason silver has fallen harder than gold lies in their different pricing anchors. Yuan noted that about 90% of gold's value is supported by its monetary attributes, safe-haven demand, and central bank reserves, with industrial demand accounting for a very low share. In contrast, roughly half of silver's value comes from industrial demand in photovoltaics, electronics, and new energy. When macroeconomic expectations reverse, gold mainly bears a correction in its financial attributes, while silver must simultaneously bear the drag from its industrial attributes and a correction in its financial attributes. Li Gang, research director at the China Foreign Exchange Investment Research Institute, analyzed that silver's rise to around $121 per ounce in January 2026 carried clear characteristics of a short squeeze and speculative premium. Subsequently, with a sustained high-interest-rate and strong-dollar environment, industrial demand—especially silver for photovoltaics—fell by about 19% year-on-year, compounded by a concentrated unwinding of leveraged positions, naturally amplifying the decline. "More critically, it is the market size. The silver market is only about one-tenth the size of the gold market, with weaker liquidity and historical volatility about 1.7 times that of gold. The same amount of capital flowing in or out has a greater impact on silver," Yuan said. He mentioned that after the January nomination of the Federal Reserve chair triggered hawkish trades and rising expectations of a stronger dollar, speculative long positions in silver that had poured in through leveraged ETFs and futures were forced to liquidate. Leveraged ETFs, to maintain a fixed leverage ratio, sold early during the decline; futures margin calls triggered concentrated liquidations, and the two combined to form a chain reaction of "fall—sell—fall further." Gold, meanwhile, has central bank strategic buying as a support below, while silver lacks strategic buyers of a comparable scale. Li Gang added that historical patterns show that during precious metals corrections, silver's decline is often 1.5 to 2 times that of gold. Market data also confirms the speed at which funds are exiting. According to the U.S. Commodity Futures Trading Commission (CFTC)持仓 report, as of the week ending October 13, speculative net long positions in COMEX silver futures had fallen to 5,735 contracts, down about 54% from 12,632 contracts a month earlier. Last Wednesday (October 14), COMEX silver futures fell 6.1% in a single day, while gold fell only 1.7% over the same period.
Long and short divergence persists
Currently, there is clear divergence between bulls and bears on silver. The bearish side argues that high-interest-rate suppression remains unresolved and the weak trend is hard to reverse; the bullish side emphasizes six consecutive years of supply shortages and AI demand providing a floor. Behind the divergence, the supply-demand figures themselves are clear. According to the World Silver Survey 2026 released by the Silver Institute, the global silver market is expected to be in a deficit of 46.3 million ounces in 2026, 15% wider than the 40.3 million ounces in 2025, marking the sixth consecutive year of supply shortfall, with mine supply nearly flat. However, a Goldman Sachs commodities report released on October 8 mentioned that CFTC data showed that in the week ending September 29, managed funds, other institutions, and non-reportable accounts combined sold a net roughly $1.6 billion in silver futures, the largest single-week sell-off since February this year; about $800 million came from long liquidation, and another $800 million from newly established short positions. Over a longer horizon, overall speculative net longs shrank from a January peak of about $24 billion to about $12 billion, a decline of nearly half, corroborating the halving in price. Yuan's judgment is that silver is currently in a wide-ranging oscillation pattern of "interest rate ceiling suppressing upside space, shortage floor constraining downside depth," rather than the starting point of a one-way trend. Yuan believes that how long high interest rates can hold determines the upper limit of silver's oscillation range. Shortages and AI demand mean that after sharp drops, silver prices often see corrective rebounds; but whether rebounds can persist depends on whether total industrial demand can return to positive growth after the photovoltaic drag bottoms out. Yuan pointed out that a true directional opportunity requires waiting for one of two signals: first, a clear shift in the Federal Reserve's policy path toward easing; second, growth in silver use from AI and data centers beginning to fill the gap left by photovoltaics in aggregate terms. Before that, consecutive shortages are better understood as a reason "it won't fall deeply" rather than a reason "it's about to surge," which is closer to the current pricing structure. Li Gang believes that in the medium term, as long as the supply side remains tight and new demand from AI, data centers, power grids, and electric vehicles gradually materializes, the price center still has support. Li Gang stated that the gold-silver ratio has rebounded from an extreme low in January to above 70, indicating that silver has clearly weakened relative to gold. If the interest rate environment eases later, silver's elasticity may again exceed gold's; if the macro environment remains tight, the oscillation and bottoming period will be prolonged. Not all institutions hold an oscillation view. Mike McGlone, senior commodity strategist at Bloomberg Intelligence, is more cautious in a research note, believing silver may test $50 in the short term, be trapped in a $50–$84 range for years to come, with a 10-year mean reversion price of about $31.
The shift period is not short
According to the World Silver Survey 2026 by the Silver Institute, silver demand in the photovoltaic sector is expected to fall about 19% year-on-year in 2026 to about 151 million ounces, the largest single-year decline on record. Photovoltaic silver demand is retreating, while AI and data center demand is advancing, but Yuan previously pointed out that data center demand has a "small base and its supplementary effect is not yet significant." The alternating stage between the photovoltaic ebb and the AI handover is the so-called "shift period." How long this period lasts directly relates to silver's pricing logic. Li Gang expects the shift period to last at least until 2027, or even longer. He believes that "de-silverization" in photovoltaics is a technological trend, with declining silver consumption per unit combined with a slowing installation pace, hard to reverse quickly in the short term. AI, data centers, electric vehicles, and grid upgrades are indeed forming new increments, but the current volume is not yet enough to fully offset the photovoltaic decline. Li Gang pointed out that the impact of this situation on silver pricing is direct: the slowdown in net industrial demand growth in the short term weakens the immediate support of the previous narrative of "strong industry + persistent deficit," making prices more easily dominated by interest rates, the dollar, and speculative sentiment. In the medium to long term, as long as the total deficit remains and the irreplaceable portion of new demand continues to grow, the price bottom will gradually solidify. Yuan pointed out that the mainstream manufacturers' silver consumption per watt has dropped from 5mg/W to within 3mg/W, and the inertia of process optimization continues, with total annual photovoltaic silver consumption in 2027 expected to fall by more than 20% from its peak. Yuan believes the real variable appears in 2027 to 2028: replacement technologies such as silver-coated copper and pure copper paste are expected to form significant increments in photovoltaic module mass production only by then, meaning the "technical reduction" in photovoltaic silver use will continue through at least the next two years. Yuan pointed out that the AI and data center handover is clearly misaligned in timing with the photovoltaic ebb. Data center silver use is expected to exceed 42.3 million ounces in 2026 and surpass 10% of electronics silver use in 2027, with growth not weak. But Yuan believes the volume gap remains large—by 2030, AI-related silver consumption as a share of total global silver demand will still be less than 4%. The essence of the shift period is that the old engine is decelerating while the new engine is accelerating, and the power gap between the two may reach its maximum around 2027 to 2028.
Comments