The core variable currently missing from most pricing models is precisely the geopolitical risk premium. For a long time, gold and the US dollar have maintained a stable inverse correlation. However, this traditional linear relationship now appears to be breaking down, with the magnitude of their fluctuations becoming imbalanced. Over the past full annual trading cycle (260 trading days), the ratio of the gold price increase to the US dollar's decline reached a staggering 7.2 times; on a six-month basis, this multiple escalated to 31 times. Does this significant deviation in sensitivity and non-linear appreciation imply that the traditional modeling methods for anchoring gold prices have become obsolete? Dhaval Joshi, Head of Counterpoint Research Services at global investment consultancy BCA Research, stated in an interview conducted before the recent sharp correction in gold and silver prices that the divergence between precious metal prices and the US dollar's trajectory had reached extreme levels, and a reversal of such extreme conditions was to be expected. "Comparing the volatility of gold and the US dollar, gold tends to rise when the dollar weakens. However, upon closer examination of the specific data, it appears that the recent rise in gold has exceeded reasonable bounds. The relationship is not a simple one-to-one correlation. Theoretically, gold's increase is typically greater when the dollar falls by 1%, which aligns with historical patterns. Yet, even so, gold's relative gains currently seem excessively front-run. Furthermore, I believe the upward momentum in other precious metals is even more pronounced," he said.
The "Omitted Variable" in Gold Models Regarding the failure of gold pricing models, Patrick Zweifel, Chief Economist at Pictet Asset Management, also concurred in a recent interview that gold's sensitivity to the US Dollar Index and real interest rates has significantly declined. According to BCA Research's calculations, over a full annual trading cycle, the US Dollar Index retreated from 107.96 to 97.42, a decline of only 9.76%; meanwhile, the gold price surged from $2,875 per ounce to $4,905 on February 3rd, an increase exceeding 70%. Using a six-month observation window, this divergence is even more pronounced: the US Dollar Index experienced a slight 1.38% decline, while the gold price skyrocketed by 42.84%, resulting in a multiple of 31 times between their movements. As an econometrician, Zweifel stated that when faced with the failure of such linear models, researchers typically have two potential approaches. One is to perform linear regression based on a "Rolling Window," observing how factor responses change over time; the other is to turn to more complex nonlinear models. However, in his view, many complex nonlinear or nonparametric models are essentially statistical "tricks." He explained that such models often just complicate "moving averages"; their core still relies on historically effective factors for backtesting and frequently adjusts weights to adapt to the market, but this often masks the fundamental issue of "Omitted Variables." Zweifel further elaborated that the core variable currently missing from most pricing models is precisely the geopolitical risk premium. While nonlinear models might capture the growth of such premiums, they cannot quantify the specific extent to which the gold price is overvalued, nor can they indicate future price direction. The reason lies in the high sensitivity of nonlinear models to price fluctuations; once the gold price changes, the model's conclusions adjust accordingly. Their nature is similar to moving averages, lacking stable forward-looking capability. He admitted, "We do use such models as well; for instance, we employ semi-parametric methods when studying the 'Taylor Rule,' which works well in certain areas. But for gold pricing, it doesn't truly solve the problem."
The Risk Game: Bullish or Bearish? How will the gold price trend evolve next? Market views are not entirely consistent. Recently, several major banks have raised their price forecasts. Societe Generale and Deutsche Bank believe the gold price could reach $6,000 per ounce this year. Morgan Stanley expects the gold price to climb to $5,700 in the second half of the year, while Goldman Sachs predicts a price of $5,400 by December. UBS, in an investment insight released at the end of January, also stated that gold's sustained strong performance has demonstrated its value as a portfolio hedge and a diversification asset. Despite a price increase of over 90% in the past 13 months, the bank still finds the investment thesis for gold continuously attractive. The report stated: "Declining real yields and persistent macroeconomic uncertainty should support gold's appeal. If political or financial risks escalate sharply, the gold price could climb to $5,400 per ounce. For investors favoring physical gold, it is recommended to allocate a mid-single-digit percentage gold position within a diversified portfolio." However, Citibank cautioned that while gold allocation is supported by multiple overlapping geopolitical and economic risks, approximately half of these risks could subside later this year. The report analyzed that a potential Trump administration is expected to push the US economy towards a "Goldilocks" state during the mid-term election year of 2026, while de-escalation in the Ukraine-Russia and Iran situations implies a potential significant retreat of risks from current levels. Furthermore, Citibank stated that if Kevin Warsh, nominated by US President Trump, is confirmed as Fed Chair, it would reinforce the bank's long-standing baseline judgment that the Federal Reserve will maintain political independence, which could constitute another medium-term bearish factor for the gold price.
Comments