Gold’s sell-off is pricing turbulence, not calm bears

Gold is being marked down as if one hot inflation print and a louder Fed have restored a simple anti-bullion regime. We think that reading is too neat. The drop is real, but the options market is not behaving as if lower prices have made gold safer to own.

From $4,085.40 just before the first sell-off headline on Nov. 10 to $3,857.50 at Thursday, Nov. 12, 2026, low, COMEX gold fell $227.90, or 5.6% [P1][P3]. Yet Cboe’s Gold Volatility Index closed at 25.9 on Thursday, up 1.7 on the day and implying a daily move of about $63.86 an ounce [D1]. Over four weeks, GVZ rose 16.1%, from 22.3 to 25.9, while COMEX gold fell 6.8%, from $4,170.10 to $3,885.30 [D1]. That is not the signature of a market growing comfortable with a one-way bear trend. It is a market paying up for protection into the fall.

The macro story is right, but incomplete

The prevailing interpretation is easy to state and mostly correct: hotter inflation, a firmer dollar and more hawkish Fed expectations hit gold [R5][R6][R7]. Reuters reported that traders were pricing a 91% chance of a 25 basis point Fed hike in December, with a 10% chance of 50 basis points [R5]. October CPI rose 0.4% month on month against a 0.3% consensus, while core CPI rose 0.3% against a 0.2% consensus [C2][C3]. Those are genuine hawkish surprises.

Real yields make the same case. The US 10-year real yield stood at 2.78% on Nov. 10, up 0.13 percentage points on the week, with the nominal 10-year at 4.98% [D7]. That puts real yields at 86% of their one-year range, a clear headwind for a non-yielding asset [D7][R2].

So we are not arguing that the sell-off is false. We are arguing that the simple reading of it is false. The mainstream macro view assumes that lower gold means lower risk and therefore less need to own protection. The vol market is saying the opposite.

Vol and skew are warning against a neat bear story

The strongest evidence sits in options. Cboe’s Nov. 11 GLD chain showed, for the same expiry that day, the 363 call at 0.00 implied volatility against 0.52 for the 363 put, while GLD volume reached 5.9 million shares and the ETF fell 1.13% [R10]. One chain snapshot is not the whole market, but it is not how a calm, consensual downtrend is usually priced.

Cboe’s own earlier work points the same way. In March it wrote that demand for puts was increasing, GLD skew had steepened to a one-year high and puts were trading at a premium to calls [R1]. In January it said one-month GLD skew steepened on hedging demand even as the surface remained inverted beyond one month, a sign that short-dated downside hedging could coexist with longer-dated upside interest [R11]. That history matters here. Gold options can express fear of an immediate air pocket without endorsing a clean longer-run bear view.

GVZ itself reinforces that point. Thursday’s 25.9 reading is still only 24% of its one-year range from 19.6 to 46.0, so this is not panic [D1]. But 52% of the past year’s readings were lower, and the index is back near the top of its past 12 sessions after touching 24.2 on Nov. 11 [D1]. Vol is bid enough to say the market expects unstable trading, not a smooth glide lower.

Futures are being cut, but the long is not bloated

The positioning backdrop also argues against the lazy “finally the crowded long broke” story. The CFTC put total speculative net length at 179.5K contracts on Nov. 3, down 6.9K on the week and down for eight straight weeks [D4]. That leaves the net position at just 26% of its one-year range [D4]. Managed money net length was 103.6K contracts, down 1.9K on the week and only 25% of its one-year range [D5]. Over 13 weeks, hedge-fund net length fell 20.8%, from 130.8K to 103.6K, while gold fell 5.1%, from $4,152.60 to $3,940.40 [D5]. That is de-risking, not a grotesquely extended long finally being crushed.

Nor is there evidence that all investment demand has vanished. SPDR Gold Shares reported a 3.6 tonne inflow into GLD and GLDM in the latest week, even though flows over four weeks were still negative 6.5 tonnes [D3]. Over 26 weeks, flows were positive 28.4 tonnes [D3]. That does not cancel the futures liquidation. It does tell us the sell-off is uneven across holders.

If we are right, the next move need not be down

The strongest case against us is obvious. Gold fell from a Nov. 9 COMEX settlement of $4,115.00 to $3,886.30 on Nov. 12, a drop of $228.70, or 5.6% in three sessions [S1][S4]. Bloomberg reported spot at $3,908.03 on Thursday afternoon, down nearly 2% on the day, and later headlines spoke of a one-year low near $3,850 as Fed hawks took charge [H5][H6]. We should not confuse uneven repricing with a floor.

But if vol stays firm while speculative length keeps shrinking from already low levels, gold becomes more vulnerable to gaps in both directions. A modest change in the macro script can produce an outsized rebound when traders are still paying for downside cover. On Friday, Nov. 13, 2026, consensus expects US PPI at 0.4% month on month and core PPI at 0.3% [C8][C7]. The test is not only whether those numbers are soft or hot. It is whether a softer print can knock down hedging demand. If it cannot, then this is not a clean bear trend at all. It is a market still braced for another lurch.