Jump to content

Statistical Study Of Grading Labs.


Recommended Posts

IDEX magazine made an interesting statistical study of labs and some of the information in their grading reports. Unlike some other studies floating around, their sample set is HUGE encompassing 12 million stones and roughly 80% of the global diamond trade (by $ value).


The entire article is here:



Here’s a few interesting tidbits.


GIA is by far the dominant player with roughly 2/3 of all stones having GIA documentation. They dominated every single category except SI3 clarity grades and Ideal cuts (which are grades that they don’t issue)


AGS was less than 2% of the total and 70% of those show ‘Ideal’ make.


Only 0.4% of the GIA grades stones showed ‘fair’ cutting. 0.0% were graded as ‘poor’.


Just over half of the GIA graded stones were round, roughly the same ratio as with the other labs. Of the GIA stones that included a cut grade, 78% were graded VG or better.


6 labs, GIA, EGL-USA, IGI, EGL-International, AGS and HRD hold 99.7% of the lab business. I listed them in order ranked by volume. No other lab was statistically significant.


It’s interesting reading.



Link to comment
Share on other sites

I think more interesting would have been how the labs actually compare in grading color and clarity on a larger scale. However they were just using lists like alot of the online consumers do, so no telling how acurately graded the diamonds were in their statistics since they weren't seeing the actual diamonds. I thought it was funny that they had GIA grading ideal cut diamonds as .2% since they have no such grade. :angry:




Also if you were a diamond dealer and had a large diamond, borderline I1-SI2 which lab would you send it to for a grade?

My stats would tell me if the stone was a strong I1ish grade it would go to EGL to get a higher grade for clarity, and most probably color as well. :)

If I had an ideal cut hearts and arrows diamond I would probably send it to AGS to get the coveted ideal cut.

It would be an interesting statistic to ask various cutters why they picked the lab that they did for grading a certain diamond.

My brain would tell me, it's all about money. How to get the most for my money if I were a diamond cutter.

Edited by jan
Link to comment
Share on other sites

Thank you for the link Neil, I found this article very interesting....... my observations:


1) I was surprised on the market share breakdown of GIA, EGL-USA, AGS, EGL-Int'l, etc... I didn't expect GIA's market share to be so big. I assumed EGL-USA was in the 8-10% range and EGL-Int'l slightly less. I didn't realize it wasn't even 1/2 that. Same with AGS, I figured their market share to be at least 3-4%, but it's not even 1/2 that as well.


2) Since when is emerald shape and oval shape popular? Maybe in 3 stone rings. I always assumed asscher was #3, ... it wasn't even #4, 5, 6, 7 or 8???? Wow, I had no idea.


3) I found it interesting there were more I1 graded diamonds then SI3, when in nature of course that's not possible, given the curve of the graph. To me that is obvious proof right there that many SI3 graded diamonds are I1. Later on down in the study it gives market share of labs based on color, clarity, etc... it woudl be very interesting to crunch some #s on those, based on the whole average, and from there you could determine what is natural, and what is unnatural if you know what I mean.


Like if you look at EGL-Int'l compared to GIA, they grade H-I-J color diamonds almost twice as much as GIA. Same thing with SI2, more then double GIA percentage wise. This obviously shows the bias for the transitional grades like H-I-J and SI2, etc... where you're borderline white and yellow tinted... or borderline eye clean and not clean. That right there shows how EGL-Int'ls grading is stretched. Same could be applied to the other labs as well.

Edited by Adylon
Link to comment
Share on other sites

This is a statistical analysis of labs, not of diamonds. I agree that a study of the stones themselves would be interesting as well but to quote one of the great philosophers of our generation, “you can’t always get what you wantâ€.



That said, there are still some interesting things in here and some that I think are important.


I think it’s fair to assume that pretty much all of these stones included were submitted to the lab by well informed and self-interested insiders. That is to say, the selection of the lab was based on the hope of realizing the highest prices in the marketplace for the 'certified' stones and that they are well aware of their options. It’s no coincidence that 70% of the AGS graded stones are ideal cuts. The dealers chose to send stones to AGS for that reason and I would venture to guess that the remaining 30% included a fair number where they were hoping for an ideal grade but didn’t quite make it for whatever reason.


I think the declining GIA market share by color and the corresponding increase at EGL-USA is another example of this. The dealers clearly have a preference to send lower color stones to EGL-USA despite the fact that GIA diamonds are reported to sell for 14-21% premium over similarly graded stones elsewhere. I think it’s fair to say that both the buyers and sellers in this marketplace are well aware of the relevant facts and the only way to explain this behavior is the difference is the differences in the grading standards. A clever statistician with a bit of calculus should even be able to work out where these numbers cross and deduce what the ‘market’ feels are the difference in grading standards at the various points. Markets are often remarkably smart about this sort of thing.


I too had noticed the 0.2% of GIA ‘ideals’. Pretty funny. It surely is the result of bad reporting by the dealers and there may be other cases of this as well. I also noticed that GIA issues more ‘Poor’ cuts than anyone. That’s pretty impressive given that it’s a total of 0.0% of their output.



Link to comment
Share on other sites

2) Since when is emerald shape and oval shape popular? Maybe in 3 stone rings. I always assumed asscher was #3, ... it wasn't even #4, 5, 6, 7 or 8???? Wow, I had no idea.


I find that data particularly suspect, given that GIA (and probably others) do not use proprietary names such as princess, radiant, Asscher, etc. So is a "square modified brilliant" a princess? A radiant? Something else? Considering the sketchiness of the cut plans provided by GIA, how did they assess that? Given the volume of reports examined (12 million!), I would rule out manual inspection and categorisation.

Edited by davidelevi
Link to comment
Share on other sites

IDEX is a big virtual trading network, sort of like the database here, where the dealers list their inventory for sale and other dealers will consider buying it. They’re correct when they say they’re HUGE although the claim that they include 80% of the global supply of over ½ carat stones seems a bit presumptuous. They do include all of the major producers of ‘new’ stones and they make it rather easy to the dealers list stones so there’s really no downside for dealers to list their production. It’s a pretty good service.


The name ‘princess’ is attached by IDEX and the dealers, not GIA. Dealers upload stones into categories and the listing dealers are the ones who choose which category to use for each particular stone. ‘Square brilliant’ isn’t one of the choices and I think it’s reasonable to expect that dealers who have square brilliant and modified square brilliant cut stones will upload them to the ‘princess’ category although, technically, they could just as easily call them hearts or marquise if they wanted. This also explains the GIA-Ideal cuts. They got that way because 0.2% of the dealers chose to list them that way. The folks at IDEX provide no filter at all beyond defining the basic categories. The dealers are trying to sell their wares and it's all about repeat business for them so there's a pretty strong incentive to try to communicate clearly to their customers what they have for sale. It does no one any good to be deliberately listing things in the wrong categories. The system also leaves the opportunity for outright errors where some data entry clerk simply typed the wrong thing and no one downstream checked. I wouldn’t toss out the whole dataset over that but there is definitely a margin for error to deal with.


‘Asscher’ is a slightly more difficult example in that there are some borderline cases where one dealer might be describing a stone as an asscher while another might list that same stone as an emerald cut. GIA calls most of them ‘square emerald cuts’ and none of them ‘asscher’ on the reports after all.


My suspicion is that the apparent popularity of emerald cuts has to do with non-US markets. My understanding is that they are much more popular in Europe and the Middle East than they are here. Bear in mind we’re still only talking about a 5% market share.


I don't understand your question about the sketchiness of the GIA cut plans. Roughly half the stones from GIA have no cut grade assigned and although there may be a few pre-2006 stones listed, I think it's a fair guess that the vast majority of these are because GIA will only assign a cut grade to round brilliants, which account for about half of the stones in the dataset. As with the GIA-ideals, it's certainly possible that there are some dealer assigned grades here where someone called a heart shape 'excellent' or some such thing but I would be surprised if there's a whole lot of this. It would piss off the buyers and abuse of it would get them kicked off of IDEX.



Edited by denverappraiser
Link to comment
Share on other sites

The name ‘princess’ is attached by IDEX and the dealers, not GIA.


Neil, my question/doubt was due to my ignorance of how the data were collected. Not knowing of the dealer's involvement, the only way I could see for IDEX to classify the cut was through the cut plan info on the grading report.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Create New...