The trade came out of the July 4 break with gold finding its feet, De Beers preparing the most consequential sight in years, and the secondary watch market grinding higher on thin summer volume. None of it moved in a straight line, and the week left more open questions than it settled. Here are the movers that actually mattered on dealer floors.

Gold consolidates in the low $4,100s

Spot gold traded around $4,143 an ounce midday on July 6, up roughly $20 from its $4,123 close on July 2, a gain of about half a percent. That followed the prior week's rebound, when metal added close to 2.5 percent after June nonfarm payrolls printed at just 57,000 against expectations near 110,000. The soft jobs number pushed the odds of a September Federal Reserve hike down to about 50 percent on CME's FedWatch, from the 66 to 67 percent priced in before the release. Silver ran harder than gold on the news, jumping 3.7 percent in a single July 2 session as the rate debate tilted.

The July 8 minutes are the fulcrum. A hawkish read that pushes September odds back above 60 percent would reinforce the World Gold Council's $4,100 base case and put the metal's floor in play. A dovish tone that keeps odds below 50 percent supports the upside scenario and lends weight to forecasts like J.P. Morgan's $4,300 third-quarter target. The gap between those two outcomes is why the tape has been coiled and volume thin all week, with dealers unwilling to commit size before Wednesday.

By July 7 the metal had eased back toward $4,140 on a firmer dollar, with the trade parked ahead of the June FOMC minutes due July 8. Year over year gold is still up better than 25 percent, though it sits about 6 percent below the $4,408 level it touched in early June. For the bench and the counter, the practical read has not changed: scrap keeps arriving, casting-grain costs stay elevated, and retail buyers are still negotiating around a spot number most of them expected to be lower by now. I break down the setup into the minutes in our gold market note.

De Beers resets the rough

The larger structural story broke in Gaborone. De Beers opened its July sight on July 7, the first under new supply agreements that cut the sightholder roster from roughly 70 buyers to between 45 and 50. Alongside the shorter client list, the miner moved to bring book prices back toward the secondary market, where its official rough had been running some 20 to 30 percent above traded levels on goods under a carat and 5 to 15 percent above on larger stones.

Provisional cycle-five sales landed near $315 million, with chief executive Al Cook flagging that the northern summer is a seasonally quiet stretch for rough. The one-line invoicing De Beers adopted earlier this year makes clean year-on-year comparisons hard, but the direction is not in doubt: the company is trading list-price prestige for a client base it can actually keep supplied. What that does to polished margins over the next two cycles is the question every diamond desk is chewing on, and I take it apart in our industry piece on the roster cut.

The reset matters for economics as much as prestige. For years sightholders bought boxes at book values above the market and made the difference back on the spread to polished, a spread that had all but closed by early 2026. Marking the book down restores some of that margin for the clients who remain, and cutting the roster concentrates the rough among the buyers most able to move it. It is a leaner machine, and it arrives at exactly the moment the trade needs one.

Watches find their footing

The secondary watch market kept its slow repair. WatchCharts data through late June had Rolex up 6.7 percent year over year and Patek Philippe up 18.8 percent, even after both gave back a little ground in May, when Rolex slipped 2.2 percent and Patek finished essentially flat. Patek still carries the strongest retention of the big three, trading roughly 10.7 percent above authorized-dealer retail; Rolex sits near a 6.7 percent premium and Audemars Piguet is barely above parity at 0.7 percent. Back the clock up to the first quarter and the trend reads cleaner still: Rolex added 1.7 percent, Patek 3.0 percent, and Audemars Piguet 2.0 percent, with the Royal Oak up 2.7 percent and most in-production references back above retail.

Summer is thin for auctions, so the action is coming off dealer-to-dealer trade rather than headline hammers. The steel sports references that led the correction on the way down are the ones stabilizing first, while precious-metal pieces still move slowly at the shows. That split matters for anyone carrying inventory into the fall, and our full read on the secondary market covers where the bid is firming and where it is not.

Anglo keeps the exit clock running

The De Beers ownership saga stayed live in the background. Anglo American reaffirmed it intends to divest the diamond unit and expects to conclude the process within about six months, with a public listing held in reserve as a backup. Chief executive Duncan Wanblad confirmed multiple consortium bids, including one led by former De Beers boss Gareth Penny with Qatari backing, while Angola signaled interest in a 20 to 30 percent stake and Botswana, already holding 15 percent through Debswana, weighed lifting its position. Anglo took a fresh $2.3 billion writedown on the asset ahead of the sale, and De Beers reported that its average realized rough price fell 7 percent to $142 a carat in 2025.

Put the two De Beers headlines together and the shape of the trade becomes clear. A miner that is being prepared for sale is also shrinking its distribution and lowering its book to match the market. That is a company optimizing for a clean handoff, not for the long-standing habit of holding official prices above traded levels to signal confidence. Whoever ends up owning it inherits a leaner, more honestly priced business, which is arguably what a buyer wants and certainly what sightholders have been asking for.

What I am watching

On the polished side, Rapaport's list showed 0.30 to 0.49 carat rounds recovering off recent inventory lows, the clearest sign yet that the small-stone segment has found a bottom. Lab-grown continues to trade at an 80 to 90 percent discount to natural, a gap the market now treats as structural rather than temporary; our diamond desk note has the detail. Central banks, meanwhile, keep buying: Poland added 18 tonnes in May, its fourth straight month of double-digit accumulation, Singapore returned with 4 tonnes, and China extended its streak past 18 straight months to reserves near 2,322 tonnes. The World Gold Council's latest official-sector survey found 89 percent of respondents expect global reserves to rise over the next year and a record 45 percent expect their own institution to add, which is the price-insensitive bid that keeps a floor under the metal regardless of the September dot plot.

On the retail side, the independent trade keeps thinning even as the pace slows. The most recent Jewelers Board of Trade figures had United States jewelry-business closures down 23 percent year over year with openings rising, yet the active company count still ground lower to roughly 22,200 firms. Concentration at the top of the rough trade and attrition at the bottom of the retail base are the two structural currents shaping the back half of 2026.

The single number that resolves the most this week prints July 8. If the FOMC minutes read hawkish and September-hike odds climb back above 60 percent, gold's floor gets tested and the scrap math changes overnight. Below 50 percent and the metal has room to retest $4,200, with the diamond reset landing against a firmer bullion tape. Either way the trade will know by Wednesday afternoon, and the back half of July gets written from there.