If after a sale a diamond is found to have a color or clarity grade that differs from the way the seller represented it, the dollars at stake can be substantial. The grading process is subjective, so a discrepancy in grades from two different appraisers might be expected. But such a variance can take a toll on your profits—and on your reputation.

The industry allows a certain tolerance for error. For example, in color and clarity grading, the accepted margin of error is one grade. But this can mean a tremendous difference in price—and the difference is magnified if the diamond in question is large.

In a 2-ct. round diamond with VS1 clarity, the difference in price between G color and H color is $1,400 wholesale ($700 per carat). For 3-ct. stones with the same grade difference, the price differential is $3,900 wholesale ($1,300 per carat). If you sell a diamond as a G VS1, what happens when one appraiser says it is H VS1 and another declares it to be F VS2? The customer is often confused and may demand a refund if the grades come out lower than the way you represented them. Yet calling a G VS1 an H VS1 or an F VS2 is acceptable by industry standards.

Large discrepancies are unacceptable to the industry as well as the consumer and can result in lawsuits. A review of past cases shows that jewelers can’t use the subjectivity excuse for misrepresenting diamond grades. Discrepancies of one or two grades may be arguably acceptable, but three or more grades will probably be looked at by the courts as intent to deceive.

Customers know the GIA system. Most consumers today are aware of the Gemological Institute of America diamond grading scale (even if they don’t fully understand it). Diamond sales presentations usually include GIA grading notations. GIA has made its system readily available, and it’s become the most widely accepted grading scale, used by laboratories all over the world.

Color grading is based on visual comparison of the stone in question with master diamonds that have been graded by GIA. Clarity is based on a set of rules the experienced grader follows. New instruments may potentially reduce subjectivity (see box on p. 89), but an absolute scale does not exist.

If you are using the GIA grading scale to represent diamonds, you’d better be prepared to defend how you used the system. Did you use GIA graded master diamonds, a colorimeter, cubic zirconia master sets, or a guess? Is your clarity grade consistent with GIA’s clearly defined terms? If the stone has a large inclusion visible under 10x magnification, for example, it would be hard to argue that it fits the definition of VVS: “… minute inclusions that are difficult for even a skilled grader to locate under 10x.”

GIA teaches grading tolerance. GIA’s diamond grading course provides some tolerance for error. Even in the final exam, students are not expected to get every grade exactly right. However, the tolerance allowed is small. In the exam, students assess cut proportions and grade color and clarity of diamonds; their answers are compared against grades given by GIA laboratory and education staff. Errors require deductions from a perfect score of 100. A margin of error of up to three grades is acceptable, with more points deducted for grades that are further off. A student who is four grades off will fail.

If we have standards for passing the GIA exam, why shouldn’t we require jewelers and appraisers also to meet these standards? I have been consulting on a pending legal case in which the jeweler’s representation of grades on a diamond differed from a subsequent GIA report by a total of seven grades. When the jeweler refused to give a refund, the customer sought legal advice.

Loose vs. mounted stones. In most labs, the lingo used in discussing color grading of loose stones describes tolerance to within half a grade, or even less. For example, a diamond will be described as a “high G,” which means it’s getting very close to an F grade, but the stone could not possibly be considered an H. In essence, the only choices in this scenario are F and G—a total range of two grades, with a margin of error of only half a grade.

There is greater tolerance for error in grading mounted diamonds. GIA and other major laboratories don’t grade mounted diamonds, but GIA’s diamond grading course includes a lesson on this topic. The course materials state that “You cannot grade mounted diamonds as accurately as loose stones because the color of the metal influences the appearance of the stone, and the mounting keeps you from getting a clear pavilion view of the color.… Consequently, most jewelers feel that for mounted rounds, assigning a range of three or four possible color grades is the most reasonable approach. Since color grading mounted fancy shapes is especially challenging, many report an even broader range.… Assign a range of color grades: if the stone looks like a J, record the grade as I-K; if the grade is elusive, consider a range of four grades.”

Some have tried to stretch the interpretation by looking at the range H-J as a three-grade tolerance. But in reality, it’s only a one-grade tolerance: “I” is used as the valuing grade, with the tolerance one grade on either side. Especially for an appraiser who sees a diamond for the first time in a mounting, grading is very tricky, and there will be times when even a skilled grader will miss by two grades or more. The jeweler, however, can’t use the mounting as an excuse. Someone at some time graded the diamond loose—if not the jeweler, then the supplier or a laboratory. A jeweler cannot “hide behind” the mounting.

The GIA course information on grading mounted diamonds for clarity can be confusing when taken at face value. Here is what the course materials say: “Grading the clarity of a mounted diamond is tricky if the mounting makes it hard to both find inclusions and judge their severity—and any mounting limits your viewing angles, some more than others. Prongs often conceal blemishes and inclusions, and a bezel hides the girdle.…When assigning a clarity grade, remember to make your first call as you see it and then assign a range of three or four clarity grades downward.”

This text requires further interpretation. The information is intended to be overly cautious because one could conceivably miss a dangerous feather under a prong, for example. Face-up, the diamond may look like a VVS or possibly even an IF, but the unseen feather could lower the grade substantially. This rule of three possible grades lower should probably be reserved for the highest grades only and even then under extreme conditions. No experienced grader would ever grade a diamond SI1 in the mounting and discover later that out of the mounting, the grade could be as low as I2 or I3.

For loose diamonds, there is less tolerance in clarity grading, just as there is in color grading. JCK gemstone editor Gary Roskin, a former employee of GIA’s Gem Trade Laboratory, uses terms such as “high, solid, low” to help define where in the range of clarity a particular diamond should be. GTL considers the clarity grade to be accurate to within half a grade. Let’s say the grader determines the grade to be a low SI1. Another grader might call it a solid SI1 on the upper end or a high SI2 on the lower end. VS2 or I1 should not be choices! While some stores or appraisers may split grade clarity for borderline stones that are difficult to call, GIA and major laboratories provide single grades only.

A few words about “SI3.” GIA does not endorse the term “SI3.” It was introduced by Philadelphia appraiser David Atlas to differentiate between some eye-imperfect SI2s and I1s. When GIA started using its grading system, not many SI2 diamonds were sold by jewelers, and I1 was almost taboo. As consumer acceptance of lower grades kept rising, many “good I1” diamonds were showing up, but they were difficult to sell with the stigma of the I1 name.

While GIA doesn’t endorse the SI3 grade and doesn’t appear to have any plans to do so in the near future, some diamond dealers like the idea. Sheldon Kwiat of Kwiat, Roisen, Ferman in New York is among them. His logic is that there is a VVS1 and a VVS2 grade for diamonds, and there is almost no difference between them and little difference in demand for these two grades. However, there is a great range of what SI2 and I1 diamonds can look like microscopically, and we separate them by only one grade. The debate is sure to continue.

Soothing the customer. When a discrepancy does arise, the jeweler has several options, but only a few are prudent. The jeweler can declare, “Those are my grades and I am sticking to them, so there is nothing more I can do.” This may seem hard to believe, but many jewelers make this claim. The customer may accept this explanation, but chances are he will never return to your store or refer a friend. Even if you strongly believe you are right and the others are wrong, be careful not to show an ego. The customer does not understand discrepancies and generally expects a “guarantee.”

The wise approach is to explain that grading is subjective and that the store strives to grade as accurately as possible. If a question of the grade exists, the diamond can be sent to an independent laboratory of the customer’s choice. If the grade comes back different from what the store represented it to be, the store will make a replacement. As long as the customer is aware of the potential for a grade to be different and of the store’s unquestioned willingness to make good on the grade, the sale and the customer can be saved. An important rule of marketing is to admit your mistakes openly. The customer will trust you more than if you claim you are never wrong.

One more issue might arise for the jeweler or appraiser. Occasionally, we might disagree with a grading report from a laboratory, or a second opinion may differ from the first report. These situations are bound to happen. The best advice is to relate to the customer that grading is subjective and diamonds are graded by people, not machines. The normal process is to have two or more graders look at each diamond, and they do not always agree. Therefore, grades can vary slightly. We hope they will at least be close. If a two-grade discrepancy occurs, perhaps you should show a different diamond to the customer. While not everyone agrees with every grade on every diamond graded by the GIA laboratory, GIA is still the lab of the highest authority in the industry.

Richard B. Drucker, G.G., is president of Gemworld International and publisher of The Guide, a pricing periodical he began in 1982. An international gemstone consultant, he has published numerous books on the jewelry industry.

How Grading Differences Affect Pricing

Everyone knows that diamond grading is subjective. According to gemological convention, an evaluation can be off one grade and still fall within acceptable tolerances. But how do such differences affect the buying and selling of diamonds?

Diamond wholesaler Jussi Kiuas of Timanttiala Oy (Gem Diamonds Ltd.) in Turku, Finland, purchases large quantities of loose goods and is concerned about grading standards, especially for diamonds of .10 ct. to .50 ct. in the more “common” grades, such as G-J color and VS and SI clarities.

To demonstrate that such goods are subject to grading differences, he submitted the same 10 diamonds to GIA in New York and HRD (the Diamond High Council) in Antwerp. The resulting grades were similar—mostly within tolerance—but not consistent (see table below). In fact, GIA and HRD gave identical clarity and color grades to only three diamonds.

Almost all the differences shown here can be attributed to subjective interpretation rather than differing standards, but Kiuas’s findings do reveal a dilemma. For example, the first diamond is worth $2,200 per carat ($968) as a G/SI1 and $2,600 per carat ($1,144) as a VS2, a difference of $400 per carat ($176). If you buy it as a VS2 but it’s evaluated as an SI1, that’s a 17% difference—not in your favor. Some maintain that profit margins are 15% to 20% for large purchases of loose goods in these size ranges. If so, there goes your profit.

Is this an argument for stricter grading standards? Darold Allen, former GIA Gem Trade Lab diamond grader and now a diamond wholesaler in Los Angeles, doesn’t think so. “Diamond grading is subjective,” he says. “So you bought a diamond as a VS2 and got a certificate as an SI1. How many times did you buy an SI1 and get a VS2 certificate? It all evens up in the end. I buy about 80% right-on-the-grade with GIA, 10% I get lucky, and the other 10% I don’t. But all in all, we’re even.”—Gary Roskin

	GIA	$/ct.	HRD	$/ct.
1. 0.445 ct.	G/SI1	$2,200	G/VS2	$2,600
2. 0.337 ct.	H/VS2	$2,000	H/VVS2	$2,400
3. 0.310 ct.	J/VS1	$1,600	J/VVS2	$1,700
4. 0.304 ct.	G/VS2	$2,400	H/VS2	$2,000
5. 0.287 ct.	E/SI1	$1,450	E/SI1	$1,450
6. 0.244 ct.	E/VVS2	$2,300	G/VVS1	$1,850
7. 0.226 ct.	E/VVS1	$1,500	E/VVS1	$1,500
8. 0.236 ct.	E/VS2	$1,850	F/VS1	$1,850
9. 0.214 ct.	J/VS2	$1,050	J/VS2	$1,050
10. 0.259 ct.	J/SI1	$1,100	I/SI1	$1,100

Diamond Grading Equipment

Clarity grading requires magnification. Although the Gemological Institute of America course requires only a 10x loupe, a microscope is highly recommended. The loupe must have corrected-lens optics, but these are not expensive. You can buy a corrected-lens loupe for less than $50.

Microscope models with Leica optics are considered the best. They are expensive, though; they can easily exceed $3,000. Graders in the lab will examine the diamond at highermagnification, then return to 10x magnification, and finally use the loupe to set the grade.

Color grading is done in the laboratory by comparison. The standard has always been master diamonds graded by GIA. A master set can cost several thousand dollars and is difficult and time-consuming to obtain. You send in diamonds for approval and replace the rejects until a set is formed. An alternative used by many jewelers and appraisers is a cubic zirconia master set. While these sets certainly fill a need for inexpensive comparison stones, they should be used with great caution. GIA does not grade them as masters, and CZs can change color slightly over time. This does not preclude their use entirely, but the limitations should be known and noted on any written appraisals.

There are also instruments for color grading, and many will even grade mounted diamonds if the mounting is not too restrictive. GIA currently sells the Gran colorimeter. Martin Haske, owner of the Adamas Gemological Laboratory in Brookline, Mass., has the SAS 2000. GemEx Systems, an equipment manufacturer in Mequon, Wis., has the Lambda Spec. These instruments are improving all the time. However, GIA currently does not use them in the laboratory to assign grades. GIA graders still rely on the comparison of master diamonds. One reason may be that fluorescence can influence the color readings on these machines.

As colorimeters continue to advance, we may one day see a total reliance on them for grading color. That still leaves clarity to the human grader, but even there we are seeing some advances in grading equipment. Inclusions can already be measured for size. Who knows? One day, it may be possible to do a full computer analysis in an accurate and repeatable way.