Rapaport published its monthly price release from New York today, and it opens: "Diamond prices declined in October as inventories rose in categories that had seen strong gains earlier this year." The sharpest fall was at the small end. Its index "fell 5.5% for 0.30-carat diamonds and 2.3% for 0.50-carat goods after peaking in late September", while "the 1-carat index dropped 0.9%, and the 3-carat index rose 0.9%".
If you are choosing a champagne diamond this week, the first thing to know about that is which stones the index counts. It counts the diamond on the other side of our price comparison. It does not count yours.
The price indexes the trade quotes are built on colorless-to-near-colorless material, this one included, so the honest answer to whether champagne prices are falling is that there is no index of these stones to read. What exists is listings. Across the 2,699 stones indexed here the median asking price was $1,382 on the 2026-11-02 data snapshot, with listings running from $91 to $243,430 — asking prices on one snapshot, not records of completed sales, and one snapshot cannot show a direction.
Three months, by size
The figures as the three monthly releases reported them — August on 2 September, September on 6 October, October today:
| Size | August 2026 | September 2026 | October 2026 |
|---|---|---|---|
| 0.30 ct | +2% | +3.2% | −5.5% |
| 0.50 ct | +2.5% | +2.5% | −2.3% |
| 1 ct | +0.5% | +0.3% | −0.9% |
| 3 ct | −0.4% | +1.5% | +0.9% |
The small end did the rising and has now done the falling: Rapaport notes that "the 0.30-carat index is still up 8.8% for the year to date", down from 15.1% after September. The 3-carat index is the one that held. We covered the start of that run on 9 September, in Rapaport's 1-carat index rose for the first time in 15 months, which closed on the condition that the small-round recovery held. It did not.
The reason given is supply rather than sentiment. "The number of 0.30-carat rounds in RAPI qualities on Rapaport Trade grew 165% between June 1 and November 1." The weekly market comment of 29 October put it the same way from the trading floor — sentiment weaker as Indian inventory builds and prices drop in 0.30 to 0.69 ct rounds — and the rough side was quiet behind it: De Beers held prices at a thinly traded October sight, reported on 2 November, with buying slowed because "Indian factories prepared to close for 15 to 30 days over the Diwali festival, which occurs November 8". Today's release also notes the impending sale of Anglo American's 85% stake in De Beers; the sight report of 2 November says those terms are still being negotiated.
The side of the card this reaches
RAPI is "the average asking price of the 10% lowest-priced round diamonds in each of the top 25 quality categories (D-H, IF-VS2, GIA-graded, RapSpec-A3 and better) offered for sale on Rapaport Trade". Those limits are why none of it is a proxy for a champagne diamond: the color window stops at H, and the white-scale stones here start at K with a brown notation printed beside the letter, while the fancy-scale browns carry no letter at all. The scale itself is GIA's D-to-Z, and what "faint brown" means covers the notation.
What the window does cover is our comparator. Every card here carries a struck-through price, and the code builds it from live listings for a natural GIA-graded G–H diamond of the same shape, the same clarity with no widening to an adjacent grade, the same cut tier on GIA's Excellent, Very Good and Good scale, and a carat weight inside a window drawn so it never crosses a trade threshold where price per carat steps up — then takes the lower quartile of those matched prices per carat. G and H sit inside RAPI's D-to-H window, so the index and the benchmark describe overlapping material, one at wholesale and one at retail. The method and its limits are on the prices page and in how our near-colorless price benchmark works.
Read directionally, a sustained fall in small near-colorless asking prices softens the comparator on a sub-carat card, which would narrow the gap shown against it even if the champagne stone's own price never moved. Two qualifications sit on that. RAPI is a trade platform's asking prices in a clarity window that stops at VS2, while our benchmark is drawn from retail listings and matches each stone's own clarity, including the SI and I grades the index excludes. And a month is a month: the 0.30-carat index is still higher than it started the year.
What this piece cannot tell you is whether our own discount has moved. The index here is rebuilt as a snapshot, today's figures are today's, and setting them beside a figure from an earlier build would compare two different populations of stones. A claim about a trend in the champagne discount needs a matched series of benchmark prices per carat band, dated, which this site does not publish.
What our index shows — data snapshot 2026-11-02, counted 3 November 2026
2,699 unique verified natural brown diamonds, deduplicated by GIA report number, across 4,780 retailer listings at 9 retailers, with 1,112 stones listed at more than one retailer. By GIA wording: 2,155 Faint Brown, 137 Very Light Brown, 18 Light Brown, 337 Fancy Brown, 52 Fancy Light Brown. Median asking price $1,382, range $91 to $243,430. 2,329 stones sit below their matched near-colorless G–H benchmark, at a median discount of 45%. Retailer asking prices on one snapshot, not records of completed sales.
The small end was already the weak place to look for a saving
That matters more than the month, because the two effects point the same way. The brown notation is worth least on small stones. On the eleven matched color-and-carat cells measured on 2 September 2026 and set out in the champagne diamond price guide, every cell at 1.50 carats and over ran between 26% and 31% below stones of the same color letter with no brown wording, while two sub-carat cells ran the other way, the brown stones asking slightly more. Champagne diamond carat weight works through why: warmth that is hard to see face-up does little against weight and clarity.
So a saving quoted on a sub-carat card is the softest number on this site for two separate reasons — the notation does least work at that size, and the comparator it is measured against is the part of the near-colorless market that just fell. At one carat and above, both arguments get stronger. Round is where the choice is widest, at 1,401 of the 2,699 stones on the snapshot.
The comparison that does not depend on the market
One figure here owes nothing to any index. Of the 2,699 stones, 1,112 are listed by more than one retailer, and because listings are matched by GIA report number those are the same physical diamond at several stores at several asking prices. Nothing about October changes the gap between two prices for one stone, which is the subject of why the same champagne diamond costs different amounts. champ. holds no inventory: it reads the report behind each listing, collapses the duplicates into one card and sends you to the retailer.
What would make this a champagne price story
Two things, neither available today. Several months of direction rather than one month either side of a September peak — Rapaport's next monthly release covers November. And a dated series of our own benchmark figures, built on a fixed population, so the question of whether the discount on these stones is widening or narrowing could be answered with data instead of a mechanism.


