The desk closed the week with two very different reads on luxury demand, a Swiss June print that ran hotter than most of us on the floor expected, and a gold tape that dipped under a round number before clawing most of it back. Here is what actually moved and what it means for anyone buying or selling inventory into the fall.

Two houses, two stories

Richemont set the tone earlier in the month. Fiscal first-quarter sales for the three months ended June 30 came in at EUR 6.33 billion, up 20% at constant exchange rates against a consensus that sat closer to EUR 5.90 billion. That is a wide beat by the standards of this group, and it was the jewelry maisons that carried it, growing 24% while watchmaking added 8% over the same stretch. Respectable on the watch side, but plainly the smaller engine of the two. When a house that size beats the street by roughly EUR 400 million, it is not a rounding error, it is a signal about where the money is still moving.

Then LVMH reported first-half numbers on the 27th and told a quieter story. Organic revenue across the group rose 2% for the half, with the second quarter a shade better at 3%. Jewelry was the standout there too, up 9% organically on the back of Tiffany and Bulgari. But the combined watches and jewelry segment landed at EUR 5.15 billion for the half, down 5% as reported and 3% organic once currency washes out. Same six weeks of the calendar, two very different tapes.

Put the two side by side and you get the split every dealer I know has been trading around all year. The top of the jewelry market is still absorbing product, while the broad watch middle is flat to soft. Richemont's jewelry houses grew 24% and are pulling one direction; LVMH's reported watch and jewelry line is dragging the other. Notice that both houses named jewelry as the winner and both leaned on named-stone brands to do it. That is not a coincidence, it is the same buyer showing up at the high end of two different groups. The maison-level detail matters more than the group headline.

Swiss June ran hot

The Federation of the Swiss Watch Industry gave the watch side a genuine lift. June exports rose 11.2% to nearly CHF 2.4 billion, and the U.S. Did even better, up 12.7% to CHF 349 million for the month. That is the kind of print that gets phones ringing at the Bay Area shows and the IWJG tables, because it tells the retail side that sell-through has not collapsed the way the spring numbers threatened.

Read the price bands underneath and the picture sharpens. Watches under CHF 200 rose 9.9% and the CHF 200 to 500 band jumped 54.1%, while the CHF 500 to 3,000 middle fell 4.7% and everything over CHF 3,000 climbed 14.2%. That is a barbell, not a broad recovery. The cheap end and the genuinely expensive end are working. The mid-tier that fills most independent cases is still bleeding. Any dealer reading the 11.2% headline as a green light for CHF 1,500 to CHF 2,500 stock is reading the wrong line.

The catch is the half. First-half exports still finished down 0.7% at CHF 12.8 billion, and the U.S. Was off 14.8% year over year across those six months. One strong June does not undo a soft spring, and the price-segment data underneath the headline is doing something worth studying, which I unpack in the watch column. For now, take the June number as a real bounce, not a trend you can commit fall inventory against.

The secondary tape agrees, barely

Resale backs the same cautious story. The WatchCharts Overall Market Index gained 1.5% in Q2, with Patek Philippe up 2.2%, Audemars Piguet up 1.5%, and Rolex up 1.0%. Positive, but these are single-digit quarterly moves, not the double-digit runs of a few years back. And June itself was flat: the overall index printed at minus 0.1%, Patek eked out 0.4%, Rolex slipped 0.2%. So the primary-market export bounce and the secondary-market grind are not the same signal, and the floor should not treat them as one.

The trophy end tells you where the conviction still lives. At Phillips Geneva Watch Auction XXIII in June, a Patek Philippe Ref. 2523 'Polychrome Two-crown World-time' hammered at CHF 7,961,000, and a F.P. Journe Chronometre a Resonance 'Souscription No. 18' brought CHF 4,875,500. Seventeen F.P. Journe lots realised nearly US$29.2 million at Phillips New York the same month. When the middle is soft, the money concentrates at the top of the book, the same pattern the Richemont and LVMH jewelry lines are showing.

Gold blinked under $4,000

Metals gave everyone a scare. During the week ended July 17, spot gold broke below $4,000 for the first time since last autumn, handing the metals desk its worst five sessions in six weeks. By the close on the 27th the metal had recovered to $4,090.16 an ounce, most of the drop retraced but the round number now proven as something other than a floor.

For anyone carrying gold-case watches or scrap on the bench, that sub-$4,000 wick was a useful reminder that the metal is not a one-way trade this summer. When the same shops are quoting stronger Swiss gold-case demand out of June, a wobbling spot price complicates every margin conversation. The break did not just dent sentiment, it reset how buyers think about the round-number level they had been treating as support. I walk through what the recovery does and does not tell you in the bullion note.

Diamonds kept resetting

The stone side stayed on its back foot. De Beers realized an average of $105 per carat in the second quarter, down 32% year over year, even as first-half production rose 46% to 14.914 million carats. That is more rough coming out of the ground into a weaker price, which is the opposite of the discipline the market keeps asking for. Second-quarter rough sales came in at 7.1 million carats, down from 7.6 million a year earlier.

The July sight showed how far the producer will go to move goods. Prices on certain smaller stones were cut by as much as 50%, and the client list came into the sight shorn by at least a third. When a house that has spent a century defending price is slicing half off the small goods and thinning the room it sells to, that is not a promotional gesture, it is a supply chain trying to find a level buyers will actually take.

Lab-grown kept pressing from the other direction. The fair price on a 1-carat stone is down to $770 and off 26.3% over the past year, with the cheapest mainstream-quality stone available at $578 and equivalent D/VVS2 goods listed as low as $305 direct. Set that against a natural 1-carat D/VVS2 still retailing around $4,600 in the U.S., and the spread is now the entire counter conversation for the fall. I lay out both sides of that spread in the stones column.

The retail read

Down at store level the tape is harder. Signet declared a $0.35 dividend with a July 24 ex-date, but the capital return sits alongside a restructuring that includes roughly 100 store closures and the shuttering of the James Allen banner. A retailer paying a dividend while closing a hundred doors and folding an online nameplate is telling you it would rather shrink to something profitable than chase volume into a soft counter. And the wider backdrop is not helping: as of July 22, WARN Act filings for 2026 had reached 2,954 notices covering 270,641 employees across 44 states.

There is a policy wrinkle too. A higher U.K. Tax is set to take effect in August, applying to loose polished diamonds and to precious and base-metal jewelry alike. Anyone moving stones or finished pieces through Britain into the fall needs that in the cost model now, not in September. I break down the U.K. Change and the Signet restructuring together in this week's industry column.

What the floor is watching

Three things carry into next week. First, whether June's Swiss strength shows up in July order books or fades as a shipment-timing quirk, and whether it reaches the mid-price band or stays stuck at the barbell ends. Second, whether gold uses $4,090 as a base or retests the $4,000 line it just broke. Third, whether Richemont's jewelry momentum is a Richemont story or an industry one once the smaller houses report their quarters.

The honest read is that luxury is not one market right now. High jewelry is firm, Swiss watches had a good month against a weak half, and the metal that supports half the inventory on any bench just proved it can trade with a three-handle. The question for the fall buy is which signal you weight most, and here is the one I would put in front of any dealer before they write a check: a natural 1-carat D/VVS2 at $4,600 against a lab equivalent at $305 direct. Which of those two numbers do you think moves toward the other by year-end?