The Diamond Buyer’s Handbook

I Foreword: The Gem We Have No Reason to Sell You

Anna, our FGAA-qualified gemmologist, on why a coloured-stone house is a useful place to read about diamonds — and why every number in this volume carries a source and a date.

I am a coloured-stone gemmologist. Most of what crosses my bench in Sydney began as rough from the same island gravels my family has worked for three generations in the Ceylon gem fields since the 1960s, and if you ask me what I love, I will tell you about a sapphire that has never seen the inside of an oven. So you might reasonably wonder why I have spent the last month writing you a diamond handbook.

The answer is that a diamond is the one gem I have nothing riding on. When a jeweller whose rent is paid by diamond margin tells you that cut matters most, that VS2 is the sweet spot and that fluorescence is a flaw, some of that is sound gemmology and some of it is the inventory talking — and from where you are standing there is no way to tell which is which. I can say the same true things without the hand on your sleeve. More usefully, I can say the things that are quietly expensive to say when diamonds are how you eat.

There are five of them, and they run through this book like a seam. The first is that nobody tells you what your diamond will be worth the morning after you buy it. The second is that the trade prices stones in brackets rather than on a smooth curve, which makes a 0.90 carat the rational purchase and a 1.00 carat a decision about the number rather than the stone. The third is that grading is skilled human opinion, not measurement, and that the industry had a public reckoning about exactly that within living memory. The fourth is that a fine sapphire will give you visible colour from across a room for a fraction of the money and will take a knock better than a diamond will. And the fifth is that the tradition being honoured when you spend two months of income on a ring is an advertising campaign — the slogan written in 1947, the two-months figure added by De Beers advertising in 1980s America — which is a perfectly good reason to do it, but it is not an ancient one.

None of that makes this an anti-diamond book. I would be a poor gemmologist if it were. A diamond is the hardest natural material on earth, it is essentially maintenance-free, it holds a polish that will outlive everyone who ever wears it, and it is, to our knowledge and within the world jewellery market, the one gemstone with a genuinely standardised global grading language — which is why, again to our knowledge, a well-papered white diamond remains the most readily resaleable gemstone in that market, even at the sobering percentages you will meet in Chapter IX. When somebody sits in front of me and says they want a diamond, I do not try to talk them into a sapphire. I try to make sure they buy a good one, at a sane price, with the right paperwork, and that they know precisely what they are paying extra for.

That is really the whole purpose of this volume. Not to change your mind, but to make sure the mind you have is working with real numbers. A diamond bought knowingly at a fair price is a lovely thing. A diamond bought because a certificate said IF and a salesperson said investment is a costly education. The distance between those two purchases is about thirty-five minutes of reading, which is roughly what you have in front of you.

A word on how this book is built, because it matters. Every date, figure, percentage and standard in the pages that follow is traceable to a named source, and I have put the names in the prose rather than hiding them in footnotes: the Gemological Institute of America, the Rapaport Price List, the International Gemological Institute, the Federal Trade Commission, De Beers Group, the National Council of Jewellery Valuers, the analysts Paul Zimnisky and Edahn Golan. Where a figure is a trade estimate rather than a published measurement, I say so in the sentence. Where the trade genuinely disagrees with itself, I show you both sides rather than picking the one that suits me. And where a number has a date attached — most of them do, because this market moves — I have given you the date, so that in three years you can tell how stale it has become.

One last thing before we start. I have written this for the person standing in a shop on a Saturday morning with a phone full of open tabs, feeling slightly outgunned. You are not outgunned. The whole system of grades and letters and brackets was invented in the 1940s and 1950s as a teaching tool, and it can be learned in an afternoon. By the end of Chapter V you will know more about diamond pricing than most of the people who will try to sell you one, and you will be asking questions they are not used to being asked. That is exactly the outcome I am after.

— Anna, FGAA-qualified gemmologist
Ministry of Gems, Sydney

II What a Diamond Is, and Who Wrote the Rules

Pure carbon in the cubic system, a legal definition rewritten in 2018, and a grading language invented in a Los Angeles classroom — the vocabulary you are about to buy in is barely eighty years old.

Start with the mineral, because everything else is commentary. A diamond is a mineral consisting essentially of pure carbon crystallised in the isometric — cubic — system. That phrasing is not mine and it is not decorative; it is the working definition in the United States Federal Trade Commission's Jewelry Guides, and in July 2018 the FTC did something to it that reverberated through the entire industry. It deleted the word natural.

Before 2018 the legal definition of a diamond began with the word natural, and a stone grown in a reactor was therefore, in law, something else. After 2018 a laboratory-grown stone is legally a diamond, full stop, because it is chemically and structurally the same mineral. What the FTC kept — and this is the part that protects you — is the disclosure requirement. A laboratory-grown diamond must always be clearly disclosed as laboratory-grown; the unqualified word diamond applied to a grown stone is deceptive. The FTC also moved away from recommending the word synthetic, on the reasonable ground that consumers hear synthetic and think fake, and a grown diamond is not fake. It is manufactured. Those are different words and they mean different things, and we will spend Chapter VII on the difference.

Now the vocabulary. Everybody uses the four Cs as though they were handed down on tablets, and they are in fact a mid-twentieth-century teaching device with a documented author. Robert M. Shipley founded the Gemological Institute of America on 15 February 1931, and in the early 1940s he coined the four Cs — colour, cut, clarity and carat weight — as a classroom mnemonic to help his students remember what to assess. It was a memory aid for gemmology pupils. It has since become the organising grammar of the global diamond trade, which is quite a career for a mnemonic.

The letters came a little later. Richard T. Liddicoat and his colleagues at GIA built the D-to-Z colour scale and the clarity scale, and first taught them in 1953. My favourite detail in the whole of diamond history is why the colour scale begins at D. There is no C, B or A. The market at the time was already awash in grading vocabulary of the A, AA and AAA sort, applied by sellers to their own stones with the rigour you would expect. Liddicoat's team wanted a scale nobody could claim they were already using, so they started at a letter no seller would ever brag about. It was a deliberate act of vocabulary hygiene, and it worked so completely that seventy-odd years later D means the top of the world.

Carat is older and stranger. The metric carat is exactly 200 milligrams — a fifth of a gram — subdivided into 100 points, so that a 0.75 carat stone is a seventy-five pointer in trade shorthand. The United States adopted the metric carat on 1 July 1913, which is documented in the National Institute of Standards and Technology's own circulars, and the word itself descends from carob seeds, which early gem traders used as counterweights on their balances because the seeds are unusually consistent in mass. Every time somebody says one carat they are, at several removes, invoking a Mediterranean tree.

Here is the part of the system that nobody explains at the counter, and it changes how you should read a certificate. GIA colour grading is performed with the diamond table-down, under controlled lighting, against a set of master comparison stones of known colour. Table-down. That is the opposite of how any human being has ever looked at a diamond in a ring. The grader is deliberately using the most revealing possible viewing geometry, because the job is to sort stones into consistent bins, not to predict how the stone will look at your sister's wedding. GIA's own published research on the D-to-Z system is explicit about the controlled conditions. Face-up and mounted, those carefully separated bins compress dramatically — which is why a G and a D in two identical settings will defeat almost anyone, and why a great deal of money changes hands every day for a difference that exists most clearly in a laboratory.

So hold two ideas at once. The 4Cs are genuinely excellent. They gave the world a common language for a product that was previously sold on adjectives, they are consistently applied by serious laboratories, and they make it possible to compare a stone in Antwerp with a stone in Sydney without either party seeing the other. They are also a laboratory sorting system that was never designed to tell you what a stone looks like on a hand, and the market has quietly converted them into a pricing ladder on which several rungs are invisible to the naked eye. Learning which rungs are which is one of the most profitable things you can do before you buy. That is the next three chapters.

If a term in this book is unfamiliar, our gem lexicon defines it in plain language; it covers diamond and coloured-stone vocabulary alike, and it is free like everything else in this library.

III Cut: The C That Is Actually King

GIA has graded cut on round brilliants since 2005 and, as of 2026 and to our knowledge, on no other shape at all — which leaves the certificate silent on the most important C for every oval, pear and emerald cut ever sold.

Every honest guide tells you that cut matters most, and every honest guide is right. Colour and clarity describe what the stone is; cut describes what the stone does. A diamond has no colour of its own worth speaking of and no pattern worth speaking of — everything you actually see when you look at one is light that went in, bounced around inside, and came back out at you. The cutter decides whether that happens well. A superbly cut I/SI1 will outshine a lazily cut D/IF in virtually any room either of them ever enters, and the second stone will have cost several times more.

GIA launched its cut grading system for round brilliants in 2005, and the amount of work behind it is worth knowing: roughly fifteen years of research, computer modelling of tens of thousands of proportion combinations, and more than 70,000 observations of real diamonds by real human observers. The output is a five-step scale — Excellent, Very Good, Good, Fair, Poor — that predicts three things. Brightness is the white light returned to your eye. Fire is dispersion, the splitting of white light into spectral colours. Scintillation is the pattern of flashes and dark areas as the stone or the viewer moves. Get those three balanced and you have a diamond that behaves beautifully in candlelight, in office fluorescents and in the flat grey light of a Sydney winter afternoon.

Now the two things about that grade which almost nobody says at the counter.

The first: as of 2026, and to our knowledge, GIA issues a cut grade for exactly one shape — the round brilliant. Not ovals. Not pears, cushions, marquises, radiants, princesses or emerald cuts. On a report for any fancy shape, the cut grade line simply is not there. GIA states this plainly in its own consumer material, and yet a great many buyers walk out with an oval believing the certificate has vouched for how it was cut. It has vouched for the colour, the clarity, the weight, the measurements, the polish and the symmetry — and it has said nothing whatsoever about the single attribute that determines whether the stone sparkles or sits there. Fancy shapes are where the most beautiful and the most disappointing diamonds both live, and the paperwork will not help you tell them apart.

The second: even on rounds, where the grade exists, it is a prediction rather than a measurement. GIA assesses brightness, fire and scintillation through a proportion-based predictive model — it does not put your individual diamond on an instrument and measure the light coming out of it. GIA's own research publications describe the modelling approach. The practical consequence is that two stones can both say Excellent and look visibly different side by side, and buyers who have compared them know it. Excellent is a bracket, not a point.

How wide a bracket? Here I have to be careful, because GIA has never published the figure. The commonly cited trade estimate is that roughly 55 per cent of GIA-graded round brilliants receive the Excellent cut grade — that number circulates widely in trade education, including at The Diamond Pro, but it is an estimate rather than a laboratory-published statistic, and I flag it as such. Whatever the true share, the trade behaves as though Excellent is a broad church rather than an elite, and its own commercial behaviour proves the point. If Excellent really meant the top, nobody would have built products to sit above it.

They have. The American Gem Society's 0-to-10 Ideal system, which was always tighter than the GIA scale at the top, survives the merger of AGS Laboratories into GIA at the end of 2022 as the AGS Ideal digital supplement to a GIA report, priced at US$25 — a separate document, in other words, whose entire reason for existing is to distinguish among stones GIA has already called Excellent. GCAL markets an 8X designation which the company describes as passing only around one per cent of stones; that figure comes from promotional material rather than an independent audit, so treat it as a marketing claim with a plausible direction rather than a verified rate. The pattern is what matters. When an industry builds three tiers above the top grade, the top grade is not the top.

Which brings us to depth, where carat weight goes to hide. A round brilliant with a total depth much beyond about 64 per cent is carrying weight down in its pavilion, below the girdle, where you cannot see it. You paid for that weight by the carat and it contributes nothing to face-up size; worse, deep pavilions leak light out the bottom instead of returning it. Run the numbers and the effect is startling. A well-cut 0.90 carat round measures around 6.4 millimetres across the top, while a poorly cut, over-deep 1.00 carat can come in near 6.1 millimetres. The smaller stone looks bigger, performs better and costs, on the figures published in trade size guides such as RockHer's, something like 15 to 25 per cent less. You are being invited to pay a premium for a number that makes the diamond visibly worse.

So what do you actually do? For a round, insist on a GIA Excellent cut grade and then look at the stone or a high-resolution 360-degree video, ideally against something dark so you can see the contrast pattern rather than a wash of white. Ask for total depth and table percentages and be suspicious of anything much past about 64 per cent depth. For a fancy shape, accept that the certificate will not do this job and do it yourself: ask for the millimetre measurements and work out the length-to-width ratio, ask for the depth percentage, and look at video for the bow-tie — that dark bar across the middle of ovals, pears and marquises, which is a cutting artefact, not a feature. Every one of these shapes has good and bad examples in the same weight, colour and clarity, at the same price, and the only instrument that separates them is your eye.

If you would rather not do that alone, that is what we are for; you can talk to our gemmologist about a stone you are considering, including one you found somewhere else.

IV Clarity and Colour: The Two Cs You Cannot See

Eleven clarity grades judged at ten times magnification and twenty-three colour letters judged face-down against master stones — then a mounted diamond that quietly collapses most of those distinctions the moment it goes on a hand.

Clarity and colour are where a great deal of money changes hands for differences nobody in the room can see. They are also, for exactly that reason, where a buyer who understands the ground can save several thousand dollars without surrendering a single thing they would ever have noticed. This is the chapter I would read twice.

Start with the definition, because the definition is the whole argument. The GIA clarity scale has eleven grades arranged in six categories — Flawless, Internally Flawless, VVS1 and VVS2, VS1 and VS2, SI1 and SI2, and then I1 through I3 — and every one of those grades is defined by what a skilled grader can see at ten times magnification under controlled conditions. Ten times. Not what you see across a table, not what your mother-in-law sees, not what the light in a restaurant reveals. A trained professional, with a loupe, looking for trouble. Once you have absorbed that sentence, the clarity ladder stops being a quality scale and becomes what it actually is: a rarity scale, calibrated to an instrument.

The eye-clean line, and where it honestly sits

The line that matters to you is not on the certificate at all. It is the eye-clean line — the point above which no unaided human eye can find an inclusion at normal viewing distance. IGI publishes that round brilliants of VS2 and above are reliably eye-clean, and the working trade consensus, which I have found consistent with what crosses my own bench, puts VS2 at roughly 85 to 95 per cent eye-clean and SI1 at about half. Read that last figure again. Around half of all SI1 diamonds are visually indistinguishable from a VVS1 in a ring, and they sell for a fraction of the price. The other half are not, which is the entire catch: SI1 is not a grade you buy, it is a grade you inspect.

Above VS2, on a mounted stone, you are buying paperwork. Flawless, Internally Flawless, VVS1, VVS2 and VS1 cannot reliably be separated from one another by an unaided eye. That premium is real and it is honestly earned — those stones genuinely are rarer, and rarity is a legitimate thing to pay for if you know that is what you are doing. It simply is not an appearance premium, and anybody who implies otherwise is selling. Two exceptions push you upward. Step cuts — emerald and Asscher — have long open tables and few facets to break up the view, so they show inclusions far more readily and deserve a grade higher than you would accept in a brilliant. And any stone much beyond three carats magnifies its own inclusions by sheer real estate. For everything else, ask for a high-resolution video, ask precisely where the inclusion sits, and remember that an inclusion under a prong is an inclusion that has been dealt with.

Twenty-three letters, and the four or five that matter

Colour runs D to Z, and as we saw in Chapter II it is graded table-down against master comparison stones under controlled lighting — the deliberate opposite of how any diamond is ever worn. Face-up and set, those carefully separated bins compress. D through J all read as white in a ring. The distinction between a D and a G, which can be many thousands of dollars, is invisible in a mounted stone to virtually every observer, and I include trained ones looking casually. Again: D is rarer. D is not whiter to your eye once it is on a hand.

The lever nobody hands you is the metal. A diamond sits in a setting and borrows from it, so the setting decides which letters you need. In white gold or platinum, the cool metal gives a warm stone nowhere to hide, and the American Gem Society's own consumer guidance points buyers towards H or better for those metals. In yellow or rose gold the reverse happens — the metal warms everything it touches, and an I, J or even K reads as perfectly consistent with its surroundings while a D would be wasted money doing an impression of the setting. So choose your metal first, then your colour grade. Doing it the other way around, which is how it is almost always done, is how people pay a rarity premium and then set it in a metal that erases the difference.

Fluorescence: a discount you are allowed to keep

Then there is fluorescence, which is the closest thing to a free lunch the diamond market offers, and it exists because the market disagrees with the laboratory. Some diamonds glow, usually blue, under ultraviolet light, and the trade prices strong blue fluorescence as a defect — typically 3 to 15 per cent below an equivalent inert stone. GIA's 1997 study in Gems and Gemology, by Moses and colleagues, tested this directly and found that average observers, meaning the jewellery-buying public, detected no systematic effect of fluorescence on appearance; where they did notice something, strong-blue stones were judged to have better colour face-up. A GIA measurement study in 2021 reached compatible conclusions. The honest caveat sits at the extreme: a small number of very strongly fluorescent stones do look faintly hazy or oily in bright sun, and those are real. So view the individual stone in daylight before you commit. In Sydney that means indirect southern light through a window, not a table with direct sun falling across it.

Put the three together and a pattern emerges that will serve you at any counter in the country. The certificate grades the stone as an object in a laboratory. You will wear it as an object on a hand, in ordinary light, at conversational distance. Those two situations reward different things, and the gap between them is where a well-informed buyer finds their money. If you would like a second opinion on a specific stone — including one you found elsewhere — you can talk to our gemmologist, and if a term here is new to you the gem lexicon has it in plain English.

V Carat: The Cliffs the Trade Cuts To

Three-tenths of a millimetre of diameter separates a 0.90 carat round from a 1.00 carat round, and about twenty-nine per cent of price per carat separates their Rapaport brackets — this is the arithmetic the trade shapes its stones around.

Carat is the simplest of the four Cs and the one where the market behaves most strangely. A carat is exactly 200 milligrams cut into 100 points, and weight is the only C that can be established with a balance rather than an opinion. You would therefore expect price to rise smoothly with it. It does not. Price rises in steps, and between the steps there are flat stretches where you get extra stone almost for nothing. Learning where those steps fall is, in my experience, the single most useful piece of arithmetic available to a diamond buyer, and it takes about five minutes.

Where the cliffs are, and as of when

The reason for the steps is a document. The Rapaport Price List is the trade's benchmark sheet of high cash asking prices, published weekly — Thursday evenings, New York time — and it quotes diamonds not on a continuous curve but in discrete size brackets: 0.90 to 0.99, 1.00 to 1.49, 1.50 to 1.99, 2.00 to 2.99, and so on. Every stone inside a bracket is quoted off the same price per carat. Cross a boundary by a single point and the whole price per carat re-rates. Physical size, meanwhile, changes continuously and invisibly. That mismatch is the cliff.

Rapaport published its own analysis of the cliffs in January 2025, looking at G/VS2 rounds, and the numbers are worth memorising. The 1.00 to 1.49 bracket ran about 29 per cent more per carat than the 0.90 to 0.99 bracket. From there to the 1.50 to 1.99 bracket was a further jump of about 65 per cent per carat — the largest single step in the range. The 2.00 line added roughly another 52 per cent, and the 3.00 line about 56 per cent beyond that. Now hold those against the physical reality. A 0.90 carat round brilliant measures around 6.2 millimetres across the top and a 1.00 carat around 6.5. Three-tenths of one millimetre. We have never met anyone who has detected three-tenths of a millimetre on somebody else's hand.

One caution, and I want to be exact about it, because a handbook that pretends the market stands still is a handbook that ages badly. Those percentages are as at January 2025. Cliffs move. The same Rapaport analysis showed the premium structure shifting after 2022 — the 1.25 to 1.49 and 2.50 to 2.99 bands strengthening, the 0.95 to 0.99 band flat — as Indian polished production fell by around half in the second half of 2024. Quote the cliffs with a date attached, always, and ask a dealer what the current spread looks like rather than reciting mine at them in 2029.

Why so many one-carat diamonds are cut badly

Here is the consequence almost nobody explains, and it links this chapter straight back to the last one. The cutter knows exactly where the brackets are. A cutter looking at a piece of rough that will finish at 0.97 carats faces an obvious commercial temptation: leave a little more weight in the pavilion, push the total depth up a percentage point or two, and land at 1.01. The stone crosses into the higher bracket and re-rates on the whole weight. Rapaport's analysis describes this behaviour openly — manufacturers hold finished stones just above the magic weights and just below the next boundary, because the incentive to stay under a big step is enormous. Weight retention wins over cut quality more often than the trade likes to admit, which is precisely why a disproportionate number of one-carat stones are cut deep.

That is how you end up with the situation set out in Chapter III: a well-cut 0.90 carat facing up at around 6.4 millimetres, and a poorly cut, over-deep 1.00 carat facing up nearer 6.1. The lighter stone looks bigger, performs better, and on the figures in trade size guides such as RockHer's costs something like 15 to 25 per cent less. You have paid a premium for a number on a certificate and received a visibly worse diamond in exchange. It is not fraud and nobody has lied to you. It is simply what happens when a pricing convention and a physical object are allowed to drift apart.

What to do with this

Shop the shoulders. For a target of about one carat, look at 0.90 to 0.96 and insist on an Excellent cut with sane depth; you will hold a stone that faces up as large as most one-carat rounds for meaningfully less money. If your heart is set on going bigger, staying under 1.50 avoids the steepest cliff in the range — though note the caution above: Rapaport's January 2025 analysis showed the 1.25 to 1.49 band itself strengthening, so ask a dealer where the shoulder currently sits. Ask for millimetre measurements on every stone you consider and compare those rather than the weights, because millimetres are what other people see. And if the number itself matters to you — if it needs to say one carat because that is what the moment calls for — then buy one carat with your eyes open. That is a perfectly reasonable thing to want. It is only a bad purchase when nobody told you it was a choice.

One last thing about Rap, because retailers occasionally produce it as evidence. The Rapaport list is a sheet of high asking prices for well-cut stones without medium or strong fluorescence. Real stones trade at negotiated discounts to it, and consumers do not buy at Rap. A price presented as a percentage below Rap is a starting point in a negotiation, not proof of a bargain, and treating it as the latter is theatre.

VI The Paperwork: Labs, Trust and the Lesson of 2014

Grading is skilled human opinion under a ten-times loupe rather than measurement — and in 2014 the trade learned in public what happens when opinion can be shopped for.

A grading report looks like a measurement. It has a serial number, a laser inscription, a plotted diagram and a set of figures to two decimal places, and the whole design of the document invites you to read it as physics. Some of it is. The carat weight is weighed. The measurements are measured. The proportions are calculated. But the two lines that move the price most — colour and clarity — are the considered opinions of trained human beings working under ten times magnification and standardised lighting, comparing your stone against reference stones and against their own long experience. That is a serious, disciplined, and genuinely valuable thing. It is not a thermometer reading.

Opinion, carefully made — and calibrated differently in different rooms

Because it is opinion, laboratories calibrate differently. GIA, IGI, HRD Antwerp and GCAL all use GIA-derived nomenclature, so a report from any of them speaks the same language of letters and abbreviations, but speaking the same language is not the same as drawing the lines in the same place. The honest position on how they compare is less satisfying than either side of the argument would like. No blind, peer-reviewed cross-laboratory study exists. The cross-submission samples that circulate in trade discussion are small — one widely referenced set of 29 stones found IGI stricter than GIA on 12, looser on 7 and identical on 10 — and they show inconsistency in both directions rather than a fixed offset. IGI, for its part, publishes that its grading is among the strictest in the world.

So here is what I will actually say, and it is a statement about market perception rather than an accusation about anyone's work. For a natural diamond of real value, GIA paper is the conservative benchmark and it maximises resale credibility — and resale credibility matters, because the secondary market is what will eventually price your stone whether you intend to sell or not. Buy on GIA paper and you have removed one argument from a future conversation. That is the whole of the case, and it does not require anybody to be doing anything wrong.

The lesson of 2014

The reason the trade cares so much about which laboratory is because it has seen the alternative. In 2014 RapNet, the Rapaport trading network, delisted all EGL grading reports from its platform, effective 1 October, following investigations and litigation concerning grading standards. Martin Rapaport put a number on the consumer harm at the time, estimating that buyers had purchased more than a billion dollars of overgraded diamonds. A class action against EGL International followed in December of the same year. We report the delisting and the litigation as trade history on the record; the point here is about structure rather than about any one laboratory. The trade press — JCK and National Jeweler among them — covered the episode extensively, and it remains the reference point whenever anyone asks why the choice of laboratory is not a formality.

The mechanism is worth understanding because it is structural rather than exotic. Grading is paid for by the seller. If two laboratories are known to draw the clarity line in slightly different places, a seller with a borderline stone has an obvious incentive about where to send it, and that incentive operates one grade at a time, quietly, at scale, with no individual act that looks like misconduct. That is what the trade means by grade-shopping. It is also why a consumer comparing two stones must compare like paper with like paper. A G/VS1 on one laboratory's report and a G/VS1 on another's are not automatically the same object, and the cheaper one is not automatically the bargain.

The practical hierarchy as it stands in 2026

GIA is the benchmark for natural stones and absorbed AGS Laboratories at the end of 2022, taking the AGS Ideal system with it as a paid digital supplement. HRD Antwerp works to broadly similar standards, though stones on HRD paper often trade at a discount to GIA paper — a market fact rather than a quality judgement. IGI is dominant in laboratory-grown certification, where it certifies roughly 65 per cent of the world's output, and is strong and improving in natural stones; it listed on the Indian exchanges on 13 December 2024 in a ₹4,225 crore, Blackstone-backed offering that closed 22 per cent up on debut. I mention the listing not as criticism but as context a reader deserves: a certification business whose volume growth is tied to laboratory-grown production is now answerable to public shareholders, and knowing that is part of reading its paper intelligently. GCAL, now Sarine-owned, issues guarantee-backed reports with light-performance analysis, which is a different and useful proposition. Anything outside that group appearing on a valuation deserves questions.

A report is not a valuation, and a valuation is not a price

This confusion costs Australians more money than any grading dispute. No GIA, IGI, HRD or GCAL report states a price; they describe the stone and stop. An insurance valuation is an entirely different instrument, and in this country the National Council of Jewellery Valuers is the professional body that governs the practice. An NCJV insurance replacement valuation states the retail cost of replacing the item new — including labour, retail margin and taxes — which is by design the highest legitimate figure that can be written about your ring, because its purpose is to make an insurer's payout sufficient to actually replace it. It is not an estimate of what anyone would pay you. NCJV valuers document second-hand replacement value as a separate figure, and it is considerably lower. When somebody arrives at a jeweller with a valuation certificate expecting that number in cash, this is the gap they have walked into, and Chapter IX puts figures on it.

Our own practice, for what it is worth, is that every stone we sell is assessed in-house by our FGAA-qualified gemmologist and travels with a Digital Gemstone Passport recording what it is, where it came from and what has been done to it — which you or anybody you later pass the stone to can verify independently. That is our answer to the same problem the laboratories are trying to solve, applied to coloured stones where the paperwork culture is younger and thinner.

VII Lab-Grown, Melee and What You Are Actually Buying

A laboratory-grown diamond is legally a diamond, visually a diamond and commercially a different animal — plus the small stones in your halo that nobody certifies and few sellers have screened.

Let me state our position before the numbers, because the numbers get pointed. A laboratory-grown diamond is a legitimate product. It is a real diamond by the legal definition, it is beautiful, it is honestly made, and a buyer who chooses one with clear eyes has made a perfectly good decision. What we object to is not the stone. It is the small number of sellers who let a customer believe they are buying something rare when they are buying something manufactured, and the slightly larger number who never mention what happens to the price afterwards. Disclosure is the whole of the ethics here, and the law agrees.

What the law says, and what an instrument can see

Since the Federal Trade Commission removed the word natural from its definition in July 2018, a laboratory-grown stone is legally a diamond. What the FTC retained is the disclosure obligation: a grown stone must always be clearly disclosed as laboratory-grown, and using the unqualified word diamond for one is deceptive. The Commission also stepped back from recommending the word synthetic, on the sensible ground that consumers hear synthetic and think fake — and a grown diamond is not fake, it is manufactured.

You will also hear that nobody can tell the difference. That claim is true in one setting and false in every other. Across a dinner table, with an unaided eye, a laboratory-grown diamond and a natural one are indistinguishable — same mineral, same optics, same behaviour in light. In a laboratory they are separated reliably, because their growth histories leave different traces. GIA's iD100 screening instrument works down to 0.005 carat and functions on mounted stones as well as loose ones. So "no one can tell" is a statement about restaurants, not about the market you will eventually meet if you ever try to sell.

The price story, with dates

Here is the part that decides how much you should spend. Wholesale laboratory-grown prices are down roughly 90 to 95 per cent from 2015 levels on Paul Zimnisky's figures. The analyst Edahn Golan logged a further 26 per cent wholesale decline across 2025 alone, moderating to about 13 to 14 per cent year-on-year by the second quarter of 2026 — moderating, note, not reversing. De Beers launched its own laboratory-grown brand, Lightbox, in 2018 at a flat US$800 per carat, and announced its closure in May 2025 after the category's price collapse; the business had lost US$101.3 million in 2023. De Beers' own stated conclusion was that natural diamonds belong in jewellery and synthetics belong in industry, which is exactly the conclusion you would expect that company to reach, and it is also supported by its own financial statements.

The grading institutions have responded to the same reality. From 1 October 2025 GIA stopped applying the 4Cs to laboratory-grown diamonds altogether, issuing simplified Premium and Standard descriptors instead, on the reasoning that more than 95 per cent of grown stones fall into one narrow colour and clarity band. A continuum scale exists to sort a natural population that varies; when a manufacturing process reliably lands in the same place, the scale stops carrying information. That change is a genuine service to buyers, because full 4Cs printed on a grown stone made a manufactured product look like a graded rarity.

And buyers have voted with real money in the other direction, which is equally worth reporting. The Knot's 2025 Real Weddings survey of more than ten thousand American couples found that 61 per cent of engagement-ring centre stones bought that year were laboratory-grown — up 239 per cent since 2020 — with an average grown centre stone of 1.9 carats and an average ring spend of US$4,600. A well-cut one-carat G to H, VS grown stone typically retails around US$700 to US$1,500 in 2026 against several times that for a natural equivalent; I flag those retail figures as market aggregation rather than a published index, though the direction is not in doubt. If size per dollar is what you want, that trade is available and it is honest. Spend what you would be comfortable never seeing again, buy it to wear rather than to hold, and you will not be disappointed by anything. We go into this at proper length in The Natural vs Lab-Grown Handbook.

The stones nobody certifies

Now the part of your ring that no guide covers, because it is not glamorous. Melee are the small diamonds — under 0.20 carat, below roughly 3.75 millimetres — that make up halos, pavé bands and side stones, and the trade sorts them by size into stars at 1.2 to 1.7 millimetres, fulls at 1.7 to 2.6 and pointers at 2.7 to 3.75. Melee is almost never individually certified. There is no report on the forty small stones in your halo, which means that a claim like F/VS melee printed on a receipt is an unverifiable assertion rather than a graded fact. It may well be accurate. You simply have no way to check it, and neither does the retailer in most cases.

Two things follow. First, undisclosed laboratory-grown melee mixed into natural parcels is a recognised problem in the supply chain, and serious manufacturers screen their parcels with instruments such as the GIA iD100 or the De Beers AMS2. Ask the question — has this melee been screened, and is it natural or laboratory-grown — and notice how the answer arrives. A house that screens will tell you immediately; most retailers cannot answer at all, which is information in itself. Second, for how the finished ring actually looks, consistency of colour across the melee matters far more than any grade claimed for it. A halo of well-matched I colour reads beautifully; a halo of mixed stones, one of which is visibly warmer than its neighbours, reads as cheap no matter what the paperwork says. Look at the small stones in daylight, side by side, before you look at anything else. Our own treatment and disclosure practice is set out in the treatment register, and it applies to the small stones as much as the centre.

VIII Diamond or Sapphire: The Comparison Nobody Neutral Makes

Diamond wins hardness by a margin larger than the whole rest of the scale; sapphire wins toughness because it has no cleavage at all — the honest trade-off, from a house that sells both.

This is the chapter where you would ordinarily expect me to sell you a sapphire. I am not going to, partly because it would be dishonest and partly because the comparison is genuinely more interesting than either trade lets on. Diamond sellers tell you diamond is harder, which is true, and stop. Sapphire sellers tell you sapphire is tougher, which is also true, and stop. Both halves are real, they measure different failure modes, and the useful answer only appears when you put them together.

Hardness: diamond wins by more than the scale suggests

Diamond is Mohs 10. Corundum — the mineral family that gives us sapphire and ruby — is Mohs 9. That looks like one step, and this is where nearly everybody is misled, because the Mohs scale is ordinal rather than linear. It ranks minerals in order of which scratches which; it does not say by how much. Put the same materials on the Knoop scale, which does measure, and the picture changes completely. Diamond comes in at roughly 8,000 to 10,000 on the Knoop scale against corundum's 1,800 to 2,000. That single step from 9 to 10 is a jump of more than three hundred per cent — and in absolute Knoop points it is a wider gap than the whole distance from talc up to sapphire beneath it. So when a diamond seller says the hardness difference is not a small thing, they are right, and the Mohs number badly understates their case.

What that buys you in daily life is scratch immunity. Only a diamond scratches a diamond. A sapphire is scratched by diamond and by other corundum, and by essentially nothing else you will meet. The practical threshold is quartz, which sits at Mohs 7 and is the abrasive component of ordinary household dust — this is why softer stones dull over years of wear without anything dramatic ever happening to them. Both diamond and sapphire sit comfortably above that line. Over a lifetime, though, a hard-worn sapphire will develop faint abrasion on its facet edges, and the trade rule of thumb is a re-polish somewhere around every seven to ten years of genuinely hard daily wear. I flag that interval as trade lore rather than measured data, because as far as I am aware nobody has published a proper study of it. A diamond will not need that service at all.

Toughness: the half the diamond trade skips

Hardness and toughness are not the same property, and GIA teaches the distinction explicitly in its own durability guidance. Hardness resists scratching. Toughness resists breaking. Diamond has perfect cleavage in four directions — planes of relatively weak atomic bonding along which the crystal will part cleanly if struck the right way. That is the property that makes diamond cuttable in the first place, and it is also why a knock against a doorframe at exactly the wrong angle can chip a diamond despite it being the hardest natural material on earth. Chipped girdles are a routine insurance claim. Corundum has no cleavage whatsoever, and its toughness is rated excellent; a sapphire absorbs an impact that would take a corner off a diamond.

So the honest summary is this. If your worry is a stone going dull and tired-looking after twenty years, diamond is the more forgiving choice. If your worry is a hard knock on a kitchen bench or a car door, sapphire is. Neither of those is a reason to reject the other stone, because both pass the test for a ring worn every day for a lifetime — which is precisely why durability should not be the argument that decides this. It gets used as one because it sounds objective.

What the money actually buys

The real difference is where the quality lives. In a diamond, quality lives under a loupe: the distinctions you are paying for in clarity and colour are, as Chapter IV set out, largely invisible on a hand. In a sapphire, quality is bodycolour — it is the first thing anyone sees, from across a room, in any light, and it does not fade. That changes what your money does. A fine one-carat blue sapphire typically retails somewhere in the hundreds to low thousands of US dollars, with top unheated Ceylon material well above that, against something in the range of US$5,000 to US$30,000 for a one-carat natural diamond of jewellery quality. I flag both ranges as retail aggregation rather than index data, but the order of magnitude is not controversial. For the price of a modest diamond you can own an exceptional sapphire, and exceptional is visible at conversational distance in a way that VVS1 never will be.

The counterweight, and I want it stated as plainly as the rest: a diamond is essentially maintenance-free, and it is — to our knowledge, and within the world jewellery market — the one gem with a standardised global grading language and the most readily resaleable gemstone there is, even at the sobering recovery rates in the next chapter. Coloured stones have no equivalent liquidity. If what you want is a stone that most jewellers can price against a published benchmark, that is a diamond. If what you want is colour, presence and a stone that nobody else in the room is wearing, that is a sapphire, and you will spend a fraction of the money getting it.

We sell both, and the conversation I have with people who come to us undecided usually resolves itself in about five minutes once we stop discussing hardness and start discussing what they actually want to see when they look down at their hand. If you would like to have that conversation, our Ceylon sapphires and unheated stones are the obvious place to start looking, designing a ring around either stone works the same way, and The Sapphire Buyer's Handbook is the companion volume to this one.

IX Is a Diamond an Investment? And How to Buy One Well

The plain answer, the resale numbers, the one narrow Australian exception — and then the checklist I would hand a friend walking into a jeweller on Saturday morning.

A note before anything else: this chapter is general information only. It is not financial product advice, it is not personal advice, and nothing in it takes account of your circumstances. We are gemmologists and jewellers, not licensed financial advisers. If you are thinking about any purchase as an investment, speak to someone qualified to advise you on that.

With that said, the plain answer is no. A diamond is a luxury good that holds value better than most luxury goods, which is a genuinely different thing from an asset. Here are the numbers that make the case, all of them dated.

What the resale market actually pays

Sellers of secondhand diamond jewellery typically recover 20 to 60 per cent of the original retail price, with well-documented GIA-graded stones at the top of that band and everything else beneath it. A documented resale test of a branded 0.53 carat ring bought at US$4,500 drew offers of US$1,000, then US$1,200, with a best offer of US$1,850 — about 41 per cent, and that at the upper end of the range. The gap is structural rather than anybody's misconduct. Retail prices carry markups that resale does not recover: up to around 100 per cent at traditional bricks-and-mortar retail, and about 18 per cent at the large online sellers by their own account, which is a self-reported figure I would treat as a floor rather than a fact.

The underlying asset has not been kind either. Natural rough diamond prices fell about 15 per cent in 2023 and about 18 per cent in 2024, leaving the Zimnisky global rough index roughly 40 per cent below its early-2022 peak. Average one-carat natural retail slid from around US$6,819 in May 2022 to around US$4,997 by December 2024, a fall of about 26 per cent. A store of value that loses a quarter of its retail price in thirty months is a luxury good. None of this is new, either: Edward Jay Epstein made the same argument in The Atlantic in February 1982, in an essay whose title asked whether the reader had ever tried to sell a diamond, and forty-four years later the mechanics he described are intact.

For completeness, Australia's own regulator frames the general category the same way. ASIC's Moneysmart guidance treats collectables and alternative assets, jewellery included, as hard to value, illiquid, producing no income, and costly to store and insure. That is the standard checklist, and diamonds fail most of it.

The one narrow exception, and why it does not generalise

There is a genuine exception and it happens to be Australian. The Argyle mine in Western Australia closed in November 2020, removing more than ninety per cent of world pink diamond supply at a stroke. Fancy Colour Research Foundation data shows pinks appreciating something like 23 to 39 per cent in the year after closure depending on intensity, with fifteen-year gains quoted at 310 to 413 per cent; I flag those percentages as coming from seller-adjacent sources rather than an independent index. Even taken at face value, this is a rare, certificate-gated, auction-grade niche in which stones change hands between specialists. It is not a template for a white diamond bought in a shopping centre, and anybody who cites Argyle to you while selling a colourless stone is arguing in bad faith.

Why you think you have to spend so much

One more piece of context, offered without cynicism. The copywriter Frances Gerety of the agency N.W. Ayer wrote "A Diamond Is Forever" in 1947 for De Beers; Advertising Age later named it the slogan of the twentieth century, and it has appeared in every De Beers engagement advertisement since 1948. When N.W. Ayer took the account in 1938, roughly ten per cent of American engagement rings held a diamond. By 1980 it was around eighty per cent. The months-of-salary rule came from the same place: one month in the American campaigns of the late 1930s, inflated to two months in 1980s America, and set at three months in Japan. It has no basis in etiquette or tradition anywhere on earth. Knowing that does not make the tradition less lovely — traditions are allowed to have authors — but it does mean the budget is entirely yours to set.

The checklist

So buy a diamond because you want a diamond, and buy it well. Decide the metal before the colour grade, then take H or better in white metals and I to J in yellow or rose. Insist on an Excellent cut grade for a round and check total depth is inside about 64 per cent; for any fancy shape accept that no cut grade exists and judge it yourself on millimetre measurements, length-to-width ratio and video. Take VS2 or an eye-clean SI1 you have personally verified, going one grade higher for step cuts and for anything past three carats. Shop 0.90 to 0.96 carat rather than 1.00, and stay under 1.50 rather than at it. Treat strong blue fluorescence as a discount to harvest, after viewing the stone in indirect southern daylight. Buy on GIA paper for a natural stone of real value. Ask whether the melee has been screened. And read the valuation for what it is — a replacement figure for your insurer, not a price anyone will pay you.

Do that and you will have bought better than most people in the room, at a price that reflects what you can actually see. If you would like a second pair of eyes on a specific stone, wherever you found it, talk to our gemmologist — and everything we sell travels with a Digital Gemstone Passport and free lifetime re-authentication, because we would rather you never had to take anybody's word for it, including ours.