Gold is trading the level of real rates, not the direction
The old inverse correlation with TIPS yields has not broken. It has changed its input from the change in yields to the level of yields — and that distinction matters for 2027.
Ravi Deshmukh Head of Research
Every desk has a chart of gold against ten-year TIPS yields, and every desk has spent two years explaining why it no longer works. We think the chart still works; the mistake is regressing price on the change in yields rather than on the level. When real yields move violently, gold responds to the shock. When they simply sit at an elevated plateau, gold responds to the plateau's duration.
Run the regression on the level of the ten-year real yield over rolling three-year windows and the fit is around 0.7 for the current regime, materially better than the one-month change specification. The economic logic is not mysterious. A real yield is the opportunity cost of holding a zero-coupon, storage-cost-bearing asset. What matters to an allocator is not whether that cost rose 15bp this month, but whether it is 1.8% or 0.4% for the next three years.
The second leg of the argument is official-sector demand, which has become less price-sensitive than the private market. Central banks bought at a record pace through a period in which real yields were rising sharply — an outcome that is hard to reconcile with a purely rate-driven model. Reserve diversification is a slow, policy-driven flow, and it does not stop because the front-end reprices.
Put together, this gives a fairly narrow range for gold. In a world where real yields fall toward 1%, our fair value estimate is $2,650/oz. In a world where they hold near 2% for another year, it is $2,320/oz. The current price sits close to the midpoint, which is exactly what a market with no strong directional conviction should look like.
We would therefore treat gold as a portfolio hedge rather than a directional trade, sized to the volatility it is meant to offset rather than to a price target. The interesting asymmetry is not in the metal; it is in the miners, where all-in sustaining costs have been guided higher twice this year and the equity market has not yet marked the difference between cost inflation and margin expansion.
Disclaimer
This note is a work of fiction published by a fictional firm. It is provided for demonstration only, does not constitute investment research or advice, and must not be relied upon. No representation is made that any price, balance or forecast here is accurate, complete or current. Meridian Data Limited and North Holdings Group accept no liability for any use of this material. Recipients should take independent professional advice before taking, or refraining from, any action on the basis of this note.