Down 15%, and the Floor Held | Global Signal™ — Bullion Intelligence
Gold fell below $4,400 — its lowest in two months. The $4,390 floor held through three tests. Paper and physical are telling different stories.
Gold dropped below $4,400 on Thursday, its weakest level in two months and roughly 15% off where it sat when the Iran conflict began in late February. If you only watched the screen, you’d conclude the bullion trade was unwinding.
The pulse underneath says otherwise. That $4,390 level got tested three separate times this week and held every time, with the dips absorbed by institutional desks, by physical buyers taking delivery, and by sovereigns who don’t make decisions based on what oil did overnight. Premiums on delivered metal stayed firm while the paper price slid, which is the kind of thing that only happens when the people buying the actual bars have a different time horizon than the people trading the contracts.
That split is what I want to spend this issue on, because it explains almost everything about how gold is behaving right now. The paper price is being pushed around by the dollar and by rate expectations. The physical price is anchored to something slower and more stubborn: the worry that long-dated government debt is no longer the safe foundation it used to be, and the sovereign-level diversification that worry produces. Those two forces have come apart this week, and when they do, the divergence usually tells you more than either price alone.
There’s also a wrinkle this week that hasn’t shown up in any prior Bullion Intelligence issue, and it cuts against the comfortable one-way story we’ve been telling about central banks. More on that below.
Opening Signal
The thing worth understanding this week is why a falling gold price and a strengthening case for owning gold are happening at the same time.
When the US and Iran traded fresh blows near the Strait of Hormuz this week, crude and the dollar both jumped. A firmer dollar plus oil-driven inflation pressure gives the Fed every reason to stay higher-for-longer, and higher real yields are poison for an asset that pays you nothing to hold it. Gold is the purest version of that kind of asset, so it took the hit. None of that changes the reason people hold gold in the first place, which is exactly why the physical market shrugged. The floor at $4,390 held three times. Premiums didn’t crack. For anyone buying with a multi-week horizon rather than a multi-hour one, a selloff driven entirely by the rate channel looks like a discount, not a warning.
Executive Signal
The headline development is that the divergence between paper and physical finally became something you can measure rather than just describe. Gold’s screen price fell under $4,400 while delivered-metal premiums held. In silver the same dynamic is sharper and more dangerous for anyone short the physical: COMEX registered inventory now covers only a low-double-digit percentage of open interest, and Shanghai continues to trade at a premium to Western markets, which tells you where the metal is actually tight.
The genuinely new piece is what’s happening with central banks, and it complicates the story we’ve been running. For the first time this cycle, the sovereign flow has gone two-way. Russia has been selling gold to help fund the war in Ukraine. Turkey is reportedly selling or borrowing against its reserves to prop up a Lira that hit yet another record low against the dollar this week, its eleventh since the Iran war started. Both countries have grown their gold holdings more than fivefold by weight since 2000, so if this turns into sustained selling it’s real supply. It doesn’t kill the accumulation thesis, but it does add a condition that wasn’t there a month ago. Sovereign gold is a reserve asset, and reserve assets get sold when a government is desperate enough.
The third thing worth flagging is how violently the professional forecasters now disagree with each other. Reuters polled 30 analysts and landed on a median 2026 gold forecast of $4,746.50, the highest in the history of that poll going back to 2012. But the median papers over a spread that runs from the low $4,000s up to JPMorgan’s roughly $6,300 bull framing, and in silver the range is almost comical, from UBS at $80 to Bank of America’s ratio-compression scenario topping out at $309. When the best-resourced desks on the Street are this far apart, the spread itself is the information. It’s the fingerprint of a market that genuinely doesn’t know what comes next, and that uncertainty is its own argument for holding the hedge.
For positioning, the map doesn’t change much: stay overweight physical metal and the quality miners, and treat weeks like this one as a chance to add rather than a reason to trim.
Key Signals at a Glance
Gold fell under $4,400 Thursday, a two-month low and down about 15% since the Iran conflict began. The $4,390 floor held through three intraday tests, each absorbed by physical and institutional buying.
The paper price is running on rates and the dollar while physical premiums held firm through the drop. The gap between the two is the signal worth watching.
Central bank flow went two-way for the first time this cycle, with Russia selling to fund Ukraine and Turkey reportedly selling or borrowing against reserves to defend the Lira.
COMEX registered silver inventory covers only a low-double-digit share of open interest, and Shanghai trades at a premium to the West, the classic setup for a delivery squeeze.
Forecaster dispersion hit record extremes: a Reuters median of $4,746.50 for gold against individual targets reaching $6,300, and silver calls ranging from $80 to $309.
The real positioning map starts below →
Conviction map, named vehicles for each thesis, forward scenarios with confidence tiers, and the Watch Triggers for the weeks ahead — in the Premium Subscription. Premium subscribers see this on publish day. Free subscribers receive it 7 days later.
Market Breakdown — Premium
This Week’s Pulse
Gold is sitting around $4,430 to $4,450 after dipping to roughly $4,390 intraday Thursday, down about 15% from where it traded when the Iran conflict started. Silver held up better on the week at $74 to $77. The gold/silver ratio is near 59.85, which is worth a note because it’s actually backed up from the 55 we flagged in late May, so the compression we wrote about then has partially unwound. The 10-year is near 4.50%, the dollar is firm on renewed Hormuz uncertainty, and oil whipsawed: Brent bounced back above $95 and WTI near $92 after a sharp selloff on peace-deal hopes reversed when fresh US strikes hit Iranian drone sites near the strait. Through all of it, premiums on physical metal stayed firm.
The Paper-Physical Mechanic
If there’s one concept to take away this week, it’s why the screen price and the delivered-metal price have pulled apart. Paper gold (futures, ETF shares, unallocated positions) moves on the dollar and on rate expectations, because when yields rise the opportunity cost of holding something that pays no interest rises with them. The person taking physical delivery of a bar is playing a completely different game. They aren’t positioning for the June Fed meeting; they’re making a years-long bet that paper claims on gold are worth less than the gold itself. This week the paper crowd sold and the physical crowd didn’t, and the metal that left Western exchanges mostly went into vaults and reserves it won’t come back out of soon.
Silver’s version of this is more acute and worth watching closely. With COMEX registered inventory covering only a sliver of open interest, the paper market has effectively written far more claims than there’s metal on hand to settle. Add the persistent Shanghai premium and you have a structure where even a modest pickup in delivery demand could force a disorderly repricing, because there simply isn’t enough registered metal to go around if too many holders ask for it at once.
The Iran War as Bullion’s Paradox
Here’s the part that trips up anyone relying on the old playbook. In past cycles, a flare-up in the Middle East sent gold higher in the same session. This time it’s been the opposite, and this week was a clean example: conflict escalated and gold fell. The reason is that the war reaches markets mainly through oil and the dollar rather than through pure fear-buying. Higher oil feeds inflation, inflation keeps the Fed restrictive, a restrictive Fed lifts the dollar and real yields, and that combination sits on gold. So the same geopolitics that should support the metal end up suppressing its paper price through the rate channel, even as the deeper reasons to own it get stronger.
Macro Undercurrents — Premium
The most important shift beneath the surface this week is the two-way central bank flow, because it forces an honest revision to a story we’ve been telling cleanly for weeks. The structural buyers (China, Poland, the emerging-market sovereigns) are still mechanical and largely price-insensitive. But alongside them you now have stressed sellers, and that wasn’t true a month ago. Russia is liquidating gold to fund Ukraine, and Turkey appears to be doing the same or borrowing against its hoard to defend the currency. The takeaway for anyone holding bullion isn’t that the thesis is broken, it’s that the thesis now has a condition attached: sovereign gold gets sold when a government runs out of better options, and knowing that is the difference between a position you understand and one you’re just hoping holds.
The forecaster dispersion deserves to be read as a signal in its own right rather than dismissed as noise. In a settled market, analyst targets bunch together. When they scatter from the low $4,000s to $6,300 on gold and from $80 to $309 on silver, what they’re collectively admitting is that the outcome hinges on a few binary questions nobody can handicap well: whether the Iran war resolves, whether the Fed holds or actually hikes, and whether sovereign stress selling swamps the structural bid. That kind of spread tends to show up precisely when a market is between regimes, and those transitions are usually where the asymmetric money gets made.
It’s also worth sitting with Goldman’s “sticky positions” point, because it’s the cleanest explanation for why the floor keeps holding. Goldman’s argument is that gold bought as insurance against fiscal-sustainability risk behaves differently from gold bought around a one-off event like an election. The event-driven buyer sells once the event passes; the fiscal-risk buyer has no event to wait for, because the underlying problem isn’t resolving this year. That’s why the marginal holder right now is patient capital rather than fast money, and patient capital is what held $4,390 three times this week.
On silver, the industrial story is quietly more nuanced than the headlines suggest. Everyone points at solar, but solar’s silver intensity per panel is actually falling as manufacturers thrift the metal to cut costs, so photovoltaic demand is set to ease even while total solar capacity grows. The more durable demand is coming from data centers, AI hardware, automotive electronics, and 5G, which are picking up the slack. And the deficit persists mostly because of supply, not demand: silver is largely a byproduct of copper and zinc mining, so it can’t scale up just because the price says it should. That supply inelasticity is what makes the shortage stubborn.
Smart Money — Premium
Underneath the stress selling, the structural sovereign bid is still very much intact. JPMorgan is modeling roughly 755 tonnes of central bank gold buying for 2026, which is well above the 400-to-500-tonne norm that prevailed before 2022, even if it’s a step down from the 1,000-plus tonne years that just passed. JPMorgan’s own read is that the step-down is mechanical rather than a loss of appetite, because at prices north of $4,000 a central bank needs fewer tonnes to hit the same target share of reserves. The price-insensitive buyer hasn’t left; they just don’t need as much metal to accomplish the same goal. That’s the demand floor sitting beneath the Russia and Turkey selling.
The behavior of the physical crowd this week told its own story. A floor that holds through three tests while delivered premiums stay firm means the buyer at these levels is taking metal off the market, not flipping paper. Physically-backed silver vehicles pulled meaningful supply out of float last year, and China’s silver imports ran at an eight-year high early in 2026. That accumulation is happening quietly underneath all the screen volatility, which is historically the pattern that precedes paper and physical snapping back together to the upside.
The risk to respect in the near term is the leverage in the paper market. The same thin COMEX structure that makes an upside squeeze possible also makes downside cascades possible, and silver’s earlier crash from its January high was driven far more by leveraged liquidation and margin hikes than by anything fundamental. Expect more of that kind of volatility, disconnected from the physical reality. It’s why sizing rather than conviction is usually the thing that gets a bullion position into trouble.
Conviction Map — Premium
Overweight — physical gold and silver in allocated form, the royalty and streaming names, and quality primary silver producers. The paper selloff strengthens rather than weakens the case for physical, and the divergence is an accumulation window.
Tactical — use the weakness to add. The $4,300 to $4,450 zone in gold and $70 to $77 in silver are reasonable accumulation references, and a selloff that holds the physical floor is a higher-confidence entry rather than a lower one.
Underweight — leveraged paper positions, unallocated accounts where you don’t actually control the metal, and miners with weak balance sheets that can’t ride out a long paper drawdown.
Hedges — physical metal is itself the hedge here against the debasement and fiscal-sustainability scenarios, so hold the structural allocation regardless of the weekly print, and keep some cash ready for further weakness.
Portfolio Playbook — Premium
The cleanest ways to express the thesis, grouped by role, with the emphasis this week tilting toward physical vehicles and delivery optionality given the divergence theme.
Physical and core exposure:
IAU (iShares Gold Trust) — low-fee gold for a core brokerage allocation
SIVR (abrdn Physical Silver Shares) — physically-backed silver at a lower fee than the largest silver ETF
PSLV (Sprott Physical Silver Trust) — fully allocated, redeemable physical silver, the cleanest bridge from paper to metal if you want delivery optionality
Royalty and streaming, for lower-risk leverage:
FNV (Franco-Nevada) — the largest gold royalty, diversified, with a balance sheet that survives drawdowns
WPM (Wheaton Precious Metals) — silver-weighted royalty leverage
RGLD (Royal Gold) — a focused royalty name with disciplined capital allocation
Producers and broad exposure, for operating leverage:
PAAS (Pan American Silver) — a quality primary silver producer with real leverage to a physical repricing
AEM (Agnico Eagle) — a premier low-cost gold producer with balance-sheet strength
SIL (Global X Silver Miners ETF) — a diversified silver-miner basket for exposure to the silver-leadership theme without single-name risk
The way to use the divergence is simple enough: a paper selloff that holds the physical floor is a place to add. The royalty names give you lower-risk exposure that survives the drawdowns, while the producers and SIL give you operating leverage when paper and physical eventually reconverge. Size for the volatility, because the leverage in futures means the screen can move hard without anything in the physical story actually changing.
Forward Scenarios — Premium
Base case — High confidence — The Iran war drags on or resolves slowly, oil and the dollar stay elevated, and the Fed holds. Paper gold churns between $4,300 and $4,800 with plenty of noise, physical premiums hold, and silver outperforms modestly enough to pull the ratio back toward the mid-50s. Structural central bank buying continues near JPMorgan’s 755-tonne pace, partly offset by Russia and Turkey. Confirms if: gold holds $4,300 through Q2, premiums stay firm, and the ratio breaks back below 56.
Reconvergence case — Medium confidence — The paper-physical gap closes upward. A COMEX or LBMA delivery squeeze, a sharp dollar reversal, or a Fed pivot signal drags paper back toward physical reality. Gold pushes toward the $5,400-to-$6,300 desk targets, silver re-rates hard given how thin COMEX coverage is, possibly toward the $100-plus calls from Citi and BofA, and the miners outperform on operating leverage. Confirms if: COMEX silver coverage keeps falling as delivery demand rises, gold holds above $4,800, and the dollar reverses below recent support.
Stress-selling case — Speculative — Sovereign selling temporarily overwhelms the structural bid. Russia and Turkey accelerate, other stressed sovereigns follow, and the supply lands on a paper market already pressured by a firm dollar and a restrictive Fed. Gold tests $4,000 to $4,200, which would be a deeper place to accumulate rather than a broken thesis, since the structural buyers would absorb metal at those levels. Confirms if: WGC data shows central bank selling accelerating, gold breaks $4,300 on rising volume, and the dollar pushes to new cycle highs.
Watch Triggers — Premium
COMEX registered silver coverage against open interest. A continued slide toward single-digit coverage is the single most important silver-specific risk to track for a delivery squeeze.
The Iran war and the status of Hormuz. A confirmed, durable reopening would ease oil and the dollar and could trigger the reconvergence case as rate-cut bets return, while continued conflict keeps the rate channel pressing on paper gold.
The next World Gold Council flow data, which is the clean read on whether the Russia and Turkey selling is a one-off or the start of a trend.
The gold/silver ratio, currently near 59.85 after backing up from 55. A move back below 56 says silver is reasserting leadership; a push above 65 would signal a broader risk-off pullback in the complex.
Physical premiums on delivered gold and silver. Premiums widening while the paper price falls is the clearest possible confirmation that the divergence is structural and the floor is real.
TL;DR — Premium
Gold fell under $4,400 this week, a two-month low and down about 15% since the Iran war began, but the $4,390 floor held three times and physical premiums never cracked. The paper price is running on rates and the dollar; the physical price is anchored to fiscal-risk hedging, and right now they’ve come apart.
The newer development is that the central bank bid went two-way for the first time this cycle, with Russia and Turkey selling under fiscal stress even as the structural buyers stay price-insensitive underneath them. Forecaster dispersion hit record extremes, which is itself the signature of a market between regimes, and silver’s thin COMEX coverage leaves room for a violent repricing.
Positioning stays overweight physical metal, the royalty and streaming names (FNV, WPM, RGLD) and quality producers (PAAS, AEM, SIL), with physical vehicles (IAU, SIVR, PSLV) for the core. A paper selloff that holds the physical floor is a window to add, not a reason to leave.
The screen broke this week. The metal didn’t.
— Written by The Global Signal Team
Global Signal™ is published for informational and educational purposes only. Nothing in this newsletter constitutes financial, investment, legal, or tax advice, nor a recommendation to buy, sell, or hold any security, asset, or strategy. All opinions are those of the author at the time of publication and are subject to change without notice. Markets involve risk, including possible loss of principal. Past performance is not indicative of future results. No client or advisory relationship is formed by reading this newsletter. Readers are solely responsible for their own decisions and should conduct independent research and consult a licensed professional before acting on any information. The author and publisher disclaim any liability for losses incurred based on this content. Full terms: https://globalsignalhq.substack.com/tos · © Global Signal™