Comparing PCGS & NGC Population Statistics

As a dealer, I hear a lot of comments about how PCGS and NGC grade rare gold coins. I thought it would be interesting to compare the population statistics for two commonly traded series, Charlotte half eagles and Dahlonega half eagles, using recent published population figures from the PCGS and NGC databases. Before I get into the numbers themselves, I think a few background tidbits are necessary. I chose Charlotte and Dahlonega half eagles because these are two branch mint series that do not have a lot of problematical issues that are extremely hard to grade (unlike, say, Dahlonega quarter eagles which are especially hard to grade). Also, the market accords relatively similar value levels to Charlotte and Dahlonega half eagles in either services’ holder (unlike, say, high quality Mercury Dimes which are clearly more valuable in PCGS holders). Finally, I chose these two series because they have comparatively high numbers of coins that have been graded, which makes the population sample we are looking at more relevant than more esoteric series that have had few coins graded.

A few more quick points. The Charlotte half eagle series consists of twenty-four coins, including the 1842-C Small Date and Large Date. For both services, I included only these twenty-four issues. For NGC coins only, I also included 1850-C and 1854-C which were designated by that service as “Weak C.” PCGS does not make this differentiation. The Dahlonega half eagle series consists of twenty-six coins, including both varieties of 1842-D, the 1846-D/D and the 1848-D/D. I also included coins designated by NGC as “Weak D.”

I. Charlotte Half Eagles

As of June 2007 PCGS had graded a grand total of 2,626 Charlotte half eagles in all grades. A breakdown of these is as follows:

Very Fine and lower, 636 (24.21% of the total graded)

Extremely Fine, 889 (33.85% of the total graded)

About Uncirculated, 930 (35.41% of the total graded)

Uncirculated, 165 (6.28% of the total graded)

As of June 2007 NGC had graded a grand total of 2,877 Charlotte half eagles in all grades. A breakdown of these is as follows:

Very Fine and lower, 288 (10.21% of the total graded)

Extremely Fine, 750 (26.06% of the total graded)

About Uncirculated, 1506 (52.34% of the total graded)

Uncirculated, 321 (11.15% of the total graded)

Before I analyze these numbers, I think there are a few very important points to make. Both PCGS and NGC have an inherent flaw with their population figures: these numbers are inflated (often severely) by resubmissions. PCGS does a recently good job of clearing the deadwood off their report and they offer submitters a “bounty” for each used coin insert that ensures that a decent number of labels will be returned. NGC, unfortunately, does not offer a bounty and this discourages certain large submitters from returning their old inserts. When I look at the NGC population figures for Charlotte half eagles, what strikes me is the large number of coins graded AU55 and higher. I think these numbers are greatly inflated due to resubmissions.

So, what do I deduct from these numbers? First of all, I am struck by the nearly equal number of total coins graded by PCGS and NGC; 2,626 for the former and 2,877 for the latter. I would have predicted that the total number would have been much higher for NGC and much lower for PCGS. Secondly, I find it very interesting that PCGS has graded around 42% of all the Charlotte half eagles submitted to them in AU and higher grades while NGC has graded slightly over 63% in AU and higher. I find it very hard to believe that over six in ten of all Charlotte half eagles grade AU50 and better, even factoring in gradeflation. One final statistic that I think is very interesting is that NGC has graded nearly twice as many Charlotte half eagles in Uncirculated than PCGS. Even factoring in the inflated population figures at NGC due to submitters not returning duplicate tags, I am still intrigued by this disparity.

II. Dahlonega Half Eagles

As of June 2007 PCGS had graded a grand total of 3,355 Dahlonega half eagles in all grades. A breakdown of these is as follows:

Very Fine and lower, 749 (22.32% of the total graded)

Extremely Fine, 990 (29.50% of the total graded)

About Uncirculated, 1329 (39.61% of the total graded)

Uncirculated, 261 (7.77% of the total graded)

As of June 2007 NGC had graded a grand total of 3,266 Dahlonega half eagles in all grades. A breakdown of these is as follows:

Very Fine and lower, 313 (9.58% of the total graded)

Extremely Fine, 781 (23.91% of the total graded)

About Uncirculated, 1824 (55.84% of the total graded)

Uncirculated, 338 (10.34% of the total graded)

In looking at these two sets of numbers, there are two areas where great disparity can be quickly noted: with coins graded VF and lower and with coins graded AU. What accounts for this?

In regards to the lower graded coins, my guess is that there are two major reasons. The first is that PCGS tends to be a bit more generous than NGC in terms of what they will or will not encapsulate in this grade range. PCGS will often net grade a lower quality Dahlonega half eagle while NGC will tend to either not grade such a coin or place it in an NCS holder. The second reason is that these lower grade coins tend to appeal more towards pure collectors than investors or speculators and these individuals often prefer to have their coins in PCGS holders.

How can the great disparity between NGC and PCGS for AU grade Dahlonega half eagles be explained? I think there are two important things to consider. The first is that the NGC populations for Dahlonega half eagles graded AU55 and (especially) AU58 are hugely inflated by resubmissions. If NGC were to clean-up their populations figures, I think the number of AU coins would be reduced by at least 200-300+. The second reason is probably due to the fact that the NGC grading line for AU Dahlonega half eagles is a bit looser than PCGS’. In my opinion, a number of AU50 Dahlonega half eagles graded by NGC would not qualify as such at PCGS.

The most important thing to remember about these numbers is that they are subject to any number of interpretations. If you are pro-NGC, you will form your own conclusions while if you are pro-PCGS you will, no doubt, reach another conclusion.

Misc. Reader Questions

was recently asked a few interesting questions by a reader and thought it might be interesting to answer them. In no order, here they are: Q: In auction catalog descriptions, are there certain giveaways that lead you to think the cataloger doesn’t like the coin?

A: A good auction cataloger is like a good journalist: impartial and able to state facts without imposing personal prejudices. That said, catalogers are only human and it is possible to “read between the lines” and figure out subtle hints that catalogers don’t like the coins they are describing. I’ve always thought a good clue was when the cataloger talks about every possible thing except the coin itself. In other words, he mentions the coin’s rarity, its history and its contextual significance but “forgets” to mention what the coin itself looks like. There are other little clues like describing a circulated coin as “bright” or referring to coloration as “unusual” or “smoky” or “hazy.” I’d also beware of instances when a readily visible mark isn’t mentioned. If a mark is mentioned, it generally isn’t that bad. If it is “overlooked” than it is often so bad that the cataloger can’t find a way to nicely mention it.

Q: Why are different gold coins different colors?

A: There are actually two different answers to this question: the classic, “old school” answer and the “today’s market” answer. In the classic sense, a gold coin is a certain color because of the alloy used in its production. As an example, early Charlotte and Dahlonega gold often used Appalachian gold which had a considerable amount of silver included in the alloy. As a result, these coins often display a green-gold color. San Francisco coins used gold from California which had a high copper content which gave the gold a reddish or rose-gold tint. Unfortunately, few circulated gold coins remain that have not been dipped or which have not been recolored. Today’s color gradations are often more a result of the chemical that was used in recoloring the surfaces. As an example, many gold coins are seen with an unnatural tangerine-orange hue which is the result of a certain type of chemical used to recolor the surfaces.

Q: Why are NGC gold coin populations so much higher than PCGS populations?

A: I think there are two answers. The first and more politically correct is that NGC simply grades more rare date gold coins than PCGS does. Many of the major submitters of these coins send the majority of their coins to NGC and, thus, more have been graded. This is especially true with Liberty Head issues. For 20th century gold coinage, PCGS has actually graded a fairly comparable number of coins and, in the case of certain generic issues, they have actually graded more than NGC. I notice a huge difference between populations on rare dates in AU58. I think the reason for this is the combination of the fact that NGC is looser with this grade than PCGS and PCGS has done a better job, especially of late, in cleaning up their population report.

Do you have any interesting questions? Please send me an email and, if possible, I will answer them at some point in the future.

Are Population Reports Useful For The Collector?

Both PCGS and NGC publish listings of the coins that they have graded. The PCGS report comes out on a monthly basis while NGC's report is issued quarterly. At one point in time, I found these to be extremely useful, both as a buyer and seller of rare coins. Today, I dislike them and, more often than not, have to make excuses regarding their numbers when I sell coins. What went wrong with these reports and why are they no longer a good tool for the coin collector?

There are many flaws inherent with the PCGS and NGC population listings. The first is their reliance on dealers to maintain the integrity of the data.

When a coin is submitted to PCGS or NGC, it is registered in a database. In a perfect numismatic world, each unique coin would be counted just once. But, as even the most casual collector realizes, grading is subjective and a tiny increase in grade can mean a huge increase in value. If I purchase a coin in a Mint State-63 holder and I think it is very high end, the chances are good that I will remove it from the old holder in an attempt to get a Mint State-64 grade, especially if the financial rewards are great.

Each PCGS and NGC holder contains a "tag" or an insert that lists the date, grade and unique serial number of a coin. Once a coin has been removed from its holder it is the responsibility of the dealer (or collector) to send the tag back to PCGS or NGC. When the tags are returned they are deleted from the original database.

But this creates a number of problems. What if the dealer loses the tag? What if he does not care about the population figures and throws the tag(s) away? Suddenly the population figures lose much of their value.

Let me give you an example. I recently bought an 1891 double eagle. This is a very rare coin in all grades with an original mintage of just 1,390 business strikes and an estimated surviving population of four dozen pieces. After a few submissions, it attained an AU-58 grade from NGC. I was very excited about this coin and prepared to add it to my website. And then came a total shock...

When I looked in the October 2001 NGC population report, I was astonished to find that the population for AU-58 1891 double eagles was twenty-two (with another two higher). PCGS had, as of February 2002, graded seven with none better. This means that the two services had graded twenty-nine examples in AU-58. Suddenly, my very rare 1891 double eagle seemed a lot less rare. Or was it?

As someone who is knowledgeable when it comes to gold coins, I know that there are not twenty-nine AU-58 1891 double eagles in existence. In fact, there are probably no more than a dozen accurately graded 1891 double eagles in all AU grades combined. So what happened to distort the data?

An NGC or PCGS AU-55 1891 double eagle is worth approximately $10,000-12,000. An AU-58 is worth $14,000-17,000. If a coin is a "tweener" (i.e., it is right between a 55 and a 58 in terms of quality) there is considerable motivation for it to be resubmitted. In MS-60 the same coin is worth $30,000-35,000. When one considers how infinitesimal an amount of difference in quality there is between most AU-58 and MS-60 coins (or even AU-55 and MS-60, to be honest), is it any wonder why an especially nice 1891 double eagle might be resubmitted ten, fifteen or even twenty times?

The dealer who owns a coin such as an AU-58 1891 double eagle can be profiled (fairly or unfairly) as a reasonably wealthy individual with reasonably sophisticated taste. In other words, he is likely to know the difference between a "58" and a "60" and have deep enough pockets to tie the coin up for three months while it is submitted and resubmitted. Rare coins with big price spreads are, therefore, more likely to have inaccurate populations then common coins or scarcer coins with minimal price spreads between one grade and the next.

Unfortunately, many of the leading coin submitters are not what I would describe as "avid numismatists." For them, coins are a business and submissions are a major part of their day-to-day operations. This means that any of the major wholesale submitters are exactly the type least likely to care if populations are accurate. I have confronted a number of these individuals about distorting coin populations and their reactions are always similar: they could care less.

The population reports have other inherent flaws. PCGS, as an example, has graded over 7 million coins as of February 2002. Managing a database of this size is a huge task. And the person managing this database is most likely to be a data entry clerk making a few dollars over minimum wage, not a numismatist. How are they going to know that a certain population figure is obviously inflated? Many savvy professional numismatists would have trouble making these judgement calls.

Are the population figures of 2002 totally worthless? Not really. I have long maintained that the true value of these numbers is comparative as opposed to quantitative. Let's say, for example, that you look at the population figures for Type One double eagles. It makes no sense to compare an 1853 Philadelphia double eagle to an 1853-O in terms of the total number graded. One--or both--of these issues could have a population that is distorted by at least 30-40% due to resubmissions. But it makes some sense to compare the number of 1853 double eagles graded in AU-55 to the number of 1853-O double eagles graded in AU-55.

When examining population figures, here are a few things to remember:

High-end coins are more likely to have inflated populations than low-grade coins. As an example, a coin that grades AU-50 is less likely to be resubmitted than an AU-58. This is due to the fact that the AU-50 probably isn't a very nice coin and the value spread between this and the next grade up is much less than between an AU-58 and an MS-61.

Many low-end examples of scarcer to rare coins do not get submitted. As an example, if I owned a Fine-15 1851-O half eagle, I would just as soon sell it unencapsulated as slab it. If I did this, it would never be counted in a population report. But this coin clearly exists, as do no-grades, damaged coins and other "misfits." The total number of coins graded for an issue may represent only a percentage of those actually available in the market.

Be aware of population figures at both services. As an example, I currently own a PCGS AU-58 1852-D half eagle. The February 2002 population at PCGS is four in AU-58 with three better. The current NGC population of this issue is twenty in AU-58 with eleven better. This means one of two things. Either NGC grades this issue much more loosely than PCGS or the NGC population is severely inflated due to resubmissions. Given the fact that this coin is worth $5,000-6,000 in AU-58 and $8,000-12,000 in MS-60 to MS-61, I would suggest that the latter is more likely the case.

Only a small fraction of modern coins have been resubmitted. Never purchase a modern coin based solely on population figures as there are tens of millions (or more) of most issues waiting to be graded.

With well over 10 million coins submitted between the two services, mistakes do exist on the population report. As an example, NGC shows two 1883-CC half eagles graded MS-64. I have never seen or heard of an 1883-CC half eagle in any grade higher than MS-62 (let alone two of them) and am inclined to believe that these represent data entry errors. Similarly, I know coins that exist that for whatever reason have never made it onto the PCGS or NGC reports.

What can you as a collector do to improve the accuracy of the population reports? I would suggest the following:

Compile lists of inaccurate numbers and send them to PCGS and NGC. But, you need facts to change a population. Merely telling one of the services "I think the number of 1847-O quarter eagles graded in Mint State-62 seems a little high" will not work. You need to be able to prove your assertion.

Urge your dealer (or dealers) to be careful in managing PCGS and NGC inserts. Encourage him to turn in his tags if he does not already do so. If he is resubmitting a coin of yours for an upgrade, make certain the tags are redeemed.

Tell PCGS and NGC that the population figures are important to you and you would like them to become accurate. Both services are sensitive to the needs of collectors.

Realize that PCGS and NGC figures are inaccurate and make your purchasing decisions accordingly. Never pass on a good coin because of the population figures in one--or both-of the reports.