Keeping up with the Jones's.....the effects of a decade of gradeflation
tradedollarnut
Posts: 20,205 ✭✭✭✭✭
Way back in 2006/7, Keoj did a study on Registry Sets and came up with a great way of ranking them: average grade differential over condition census. At the time, the Seated Dollar set was extremely highly rated at 1.74 census differential. Shortly thereafter, I received an offer that I couldn't refuse and sold the set.
A year ago I bought it back and have added many coins at great expense to the set. I also paid about $10k in grading fees for a review of some of the premium quality coins. Here is a summary of the changes:
1841, 1842, 1843, 1857, 1860, 1860-O, 1868, 1870, 1873 regrade +
1846-O, 1855, bought a plus coin
1851, 1852, 1853, 1859-S bought a higher grade coin
The net result of this activity is an increase in the average grade of the set by about 0.2 of a point (10 grades divided by 44 coins). HOWEVER, IN THE MEANTIME THE CENSUS GRADE AVERAGE HAS RISEN BY 0.62 OF A POINT. This means that despite spending over $1.5M on upgrades to the set and $10k in regrades, it has lost nearly half a point differential to census average because of gradeflation. Sigh
Comments
The person who has the hardest time keeping up with the Jones' is Mr. Jones.
TDN, help me understand this cenus differential.
What numbers do you use to calculate it.
Take the grade average of the PCGS condition census (top 10 coins) and subtract it from the grade of the coin in the set. This number serves to show how great a Registry set is - for instance a top of the Registry set could have all top pops yet have a zero condition census differential. That would be a very good but not that difficult to put together set.
I don't play at this level, but I figured out a while ago that chasing points in the registry game was foolish. It really has nothing to do with collecting nice coins. In Peace dollars, competing sets were passing mine up by simply playing the regrade game while I was working to acquire nicer and better coins. Many of the coins in competing sets were inferior by my own standards, and even when someone put a higher number on the label, it didn't seem to magically improve the coin.
https://youtube.com/watch?v=KOO5S4vxi0o
Liberty: Parent of Science & Industry
I hear rumblings of gradeflation all the time, but then seeing it shown in a quantifiable manner makes it seem so much more. But, that just demonstrates the need to have a good eye and solid advisors. Together, sets can still be created where only the best coins are included and the market recognizes that upon sale time.
Andrew Blinkiewicz-Heritage
As collectors have noted for many years, grades are a moving target. Unfortunately the "art" of grading ends up trumping the "science" of grading even when it is obvious that the standards have been watered down.
Years ago I bought a 1799 silver dollar from an older collector. He though that the coin was an AU-50. I told him that it was an EF-45. We agreed to send the coin to ANACS for grading. Whatever grade ANACS assigned to it would be the basis for the price I would pay him. The coin came back in an old small sized ANACS holder (This was when the service belonged to the ANA) as an EF-45. Twelve or thirteen years later, I cracked out the coin to have it graded by NGC. They graded it AU-55. This was circa 2003,
In a similar situation I have an 1800 Bolender 20 silver dollar that is now tied as the finest known for the variety. It was Bolender's plate coin and Bolender's discovery coin for the variety. It is the plate coin in his book. When the coin was last sold at public auction in 1976, the grade was EF-45. Today it's in an NGC AU-58. It was also graded by NGC circa 2003.
All of this points out to us that grade-flation has been around for a long time. My solution is not worry about it or the registries because I've come to conclusion that keeping up with the registries is like a dog chasing its tail. Unless you have a set that so far above everyone else that they can't catch you, grade-flation or some other factor is going to keep you always chasing after an impossible goal. I'm more interested in the history and the beauty of my collection than numerical grades.
Numerical grades count to a point when it comes time to sell. Even then they might not be the ultimate factor. Practical examples of that are the sometimes unwarranted premiums some people pay for gold CAC stickers.
I think I got it.
So in other words the #2 - #10 coins are getting grade inflated faster than the pop 1 coins in your set.
Correct!
I already know there are people who have more money than I do.
No reason to prove it in the Registry.
So how much of that grade inflation in the PCGS census do you think is caused by NGC coins being crossed? In Pl/DMPL Morgan's I believe that a large portion of top pop coins that were in a NGC slabs have been crossed, if they could. This gave the appearance of larger increases in some dates pop's on the PCGS side of the fence than what would have occurred by grade inflation alone.
Not necessarily. Coins #2 - #10 in 2007 could be different coins. Obviously the OP would know better than me, but these could be crossovers, newly graded, etc. It is an interesting comparison, but ultimately hard to draw too much in the way of conclusions without more data.
And does this in any way effect the prices of the the top pops? For example, it would not be surprising that your MS66 doesn't upgrade but meanwhile two more MS64s get graded and push two MS63s out of the ranking. The MS66 doesn't change in value because of that. MS65's upgrading to borderline MS66s could affect the perceived value of your MS66, but your condition census methodology doesn't necessarily show that, at least if I'm understanding your methods correctly.
I certainly believe that grade-inflation is a very real and serious problem. Even so, the remarks in the original post of this thread are very ambiguous with respect to the extent of grade-inflation.
Before March 2010, only whole number grades were assigned by PCGS. The introduction of so called plus grades means that fractional "+" grades are assigned to some coins. David Hall then said that plus grades should be assigned to coins that grade from X.7 to X.9. For registry set purposes, however, a coin with a plus grade is counted as X.5. So, a coin that is PCGS graded as MS-65+ in a set registry will be computed as grading "65.5." ( I am not commenting on these rules: I am just recounting them here.)
The fact that some coins that were graded 65 before March 2010 were PCGS graded as 65+ after March 2010 refers more to a change in grading terminology, rather than to grade-inflation. If many coin that were PCGS graded MS-65 were later PCGS graded as MS-66, that would be indicative of grade-inflation.
TDN:
Is TDN saying that he replaced these dates with different coins? Upgrades involving different coins are not necessarily indicative of grade-inflation.
I provide at least two examples of grade-inflation in a recent review of an auction:
Healthy Prices Realized for FUN Platinum Night Auction Coins
I discuss reasons for grade-inflation and possible consequences in this article:
How will Coin Collectors Interpret Certified Coin Grades in the Future?
1000Lake:
Yes, I agree; crossovers would have to be factored into a sound analysis. Suppose that a coin that was earlier graded 67 elsewhere is graded 66 at PCGS, according to the formula cited by TDN, such an event would cause grade-inflation.
TDN said, with capital letters:
This is fascinating. Failed attempts at upgrades distort this data, if all labels ("inserts") are not returned. It is not a secret that crackout artists really do 'crack' PCGS holders to remove coins and seek higher grades. If a crackout artist tries a few times to get an upgrade and fails, then this "census grade average" can be distorted and may provide a false impression.
Suppose, for a particular date, there are five coins listed as being PCGS graded as MS-63, two as 64, two as 65 and one as 66. Suppose that the owner of a 64 tries for an upgrade through five submissions to PCGS and does not return the labels ("inserts"). The new top ten coin "census" would be seven at 64, two at 65 and one at 66, and the "census grade average" would increase, yet there would not have been any grade-inflation.
Population data should not be taken as literal factual data; it must be interpreted by someone who understands the games and has spent hours thinking about how to interpret various slices of such data. It is especially important to track the pedigrees of individual coins.
http://www.pcgs.com/News/Old-Pcgs-Inserts-Raise-Cash-For-New-Ana-Program
This is an excellent point. TDN could just crack out and resubmit one his top pops and really get that census grade average to shoot up!
I do find the question of gradeflation to be very interesting. I think few would deny it has occurred to some degree. However, how much and the exact impact is difficult to aggregate.
Look at the 1866 issue for an example.
The census shows 2 in 65, 4 in 66, and 2 in 67.
The 2 67's are the 4 66's. The 2 65's are the same coin.
The 3 highest grade coins have a combined pop of 8.
My 1866 Philly Mint Set
This is interesting and I will take it all in and digest the info. Thanks for posting this.
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb, Ricko
Bad transactions with : nobody to date
I have been frustrated with this issue as well. My biggest complaint is that while I would like to upgrade many of the AU50, 53, and 55 coins in my set, that were purchased back when the grading was tighter, the new higher graded coins are not any nicer than the current coins in my set. I won't purchase a lesser coin for the set just to have a higher numeric grade and I won't resubmit all of my current coins to max my set out. I like the OGH holders and the Series holders with the gaskets and don't really want to pay the extra fees just like what TDN mentioned. So now I am stuck attempting to upgrade coins but not being able to find coins that are both reflective of improvements in both numeric value and actual quality. The upgrades are much fewer and far between and as a result I become less and less active in the hobby.
Seated Dollar Collection
Never thought of looking at it that way - interesting.
You need an old fashioned showdown of the top 3 sets at the ANA to actually compare coins, not grades.
When it was done with IHC's years ago, the difference in quality between sets with comparable grades was quite revealing.
This.
As TDN noted, this was a mental exercise that I did about 10 years ago for the Gobrecht Journal. Its purpose was the following:
1) I got tired of people trying to compare sets of different series against one another. By using the PCGS Census average, it was a way to compare sets. To pick on TDN, which is a better set...his Trade Dollar Set or his Seated Dollar set. I also did a couple of quick comparisons to other top (and I mean the very top) collections in other series.
2) It was a way to get rid of the PCGS "weighting". The weight is not a bad concept but it's a little inconsistent between series. A weight of 3 in one series is not the same as a 3 in another series.
3) It was an excuse to be a coin nerd.
On the 1st two, it holds up pretty well but there are other methods that I looked at. This was the simplest. Little did I know that TDN would use it to make a grade inflation argument.
To answer the question that I posed before, which is the better set.....back in 2004, at that time, his Seated Dollar set edged out the Trade Dollar set for a higher number (1.74 vs 1.54 points over census average respectively) but this is what i wrote: "With respect to what set is “better”, the Legend Seated Dollar set or the Legend Trade Dollar set, I’m going to take a neutral stance. Both sets are outstanding achievements and are arguably the best sets ever assembled in their respective series. On a per coin basis, the Trade Dollar set has a phenomenal percentage of finest graded sole coins in the set. However, the Seated Dollar set is a much larger collecting challenge with more than twice the number of coins required to complete the set. There are also many coins within the set that are prohibitively rare in any grade (the 70-S and 73-CC as examples) and to have a coin, much less the best coin for the date and mintmark, represents a significant achievement."
Still feel that way!
Keoj
If a coin wouldn't cross at the same grade before because it wasn't quite nice enough but now does - that too is grade inflation
Say no more! I'm fully on-board!
What part, if any, does "crack out, resubmit", "crack out, resubmit" ... play in the issue of grade inflation? Humans will make different judgements on different days, and if the $$$ to be made by even a one point upgrade are significant, well, "crack out, resubmit" will go on and on for some (many?) coins.
I think everyone would agree with that. Though I'm not sure if this a response to a particular point or just a general statement that could/would explain part of the increased condition census.
Sorry - was responding to this:
So how much of that grade inflation in the PCGS census do you think is caused by NGC coins being crossed?
I'm thinking with plus grades added where before there were none, gradeflation is inevitable...
Bruce, for the coins which received the (+) grade --- was that available when they were graded and placed in the holder they were in prior to your regrade?? if not, then I don't think it's fair to characterize that as grade-flation
A agree with Analyst and Nic - the largest component of the increase of the Top 10 Census grade average
over time is likely to be from regrades on the same coins.
To restate Nic's example:
In 2007 the top 10 PCGS Census (aka ever graded) may have been:
7 x MS-64, 1 x MS-65, 2 x MS-66: average 64.5 for 10 different coins.
In 2017 the top 10 PCGS ever graded are:
2 x MS-64, 2 x MS-65, 4 x MS-66, 2 x MS-67: average 65.6 for 5 different coins; the increase is 1.1 .
Actual grade inflation for the top 3 coins was:
2007: 1 x MS-65, 2 x MS-66: average 65.67
2017: 1 x MS-65, 2 x MS-67: average 66.33 ; difference = 0.67
But if we average it over the top 10 2007 coins, the average grade inflation is 0.2 .
I agree with this
I do know that I was responsible for one event - downgrading an NGC66 1857 to PCGS65
I DO the registry but I do not play the registry game.
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb, Ricko
Bad transactions with : nobody to date
As Analyst noted, plus grades are also a factor which increases the "top 10" average.
The average computation could be adjusted to make 2007 and 2017 comparable, though.
3/10 coins get a plus (say 66.7 to 66.9), and 7/10 (66.0 to 66.6) do not get a plus.
So if the 66+ gets a 66.5 score, then the 66 plain need to get a 65.7867 score,
in order to maintain an overall average of 66 = .3 x 66.5 + .7 x 65.7867 , and retain comparability with 2007.
There are few enough seated dollar in plus that I would be shocked if it was more than 0.1 of the 0.6 grade inflation. I welcome someone to crunch the numbers tho
Good work on staying consistent over time; valuing the coin over the grade on the holder is the primary goal, of course.
The observation of less frequent upgrades is a classic problem that many collectors have encountered.
It seems the recommended solution is to start a new series, but of course resources are finite.
TDN,
Are any of the seated dollars in Central States some of your former coins before upgrades? Sorry to go off topic (which is an interesting one).
"Look up, old boy, and see what you get." -William Bonney.
P.S. I've always been a believer in the advice from The Temptations - "Don't Let the Joneses Get You Down":
https://youtube.com/watch?v=AmxWbEfdiZI
The only one of those that I owned was the 1867. It was nice but a bit unoriginal so when I got the chance at the Childs coin I jumped on it
The concept of comparing a coins grade to the average condition census grade in order to rate its exemplary status is great in theory. Makes sense to me.
I can't see, however, completely ignoring ATS coins. Maybe some adjustment to their grade is needed, but not ignored.
And as Nic and others point out, the pop reports first have to be adjusted for their built in misinformation regarding crossovers, regrades, and errors. Anybody collecting at the high end for the classics should hopefully be able to give this a go.
Pop numbers may be a problem , {actually they are} but I think to adjust them [fix them] would require a total **restart **of the recording/grading information and record keeping system.
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb, Ricko
Bad transactions with : nobody to date
I've been thinking more about gradeflation and I don't know. Is it just a correction of undergraded coins from periods when grading was tighter?
Although ... I have seen at least 6 coins in the
current Heritage sale that grew a grade in the last 6 months, some in the last 3 months (e.g. They were an AU55 in the last heritage sale -- normal / recent holder, not OGH, and now they are AU58).
I believe that the so called grade inflation phenomenon is simply the state of our current interpretation of the currently accepted grading system. I look at the coin, and the potential change in grade really, truly has no affect on my day to day life.
JMHO
PS People that want or need to make a living buying and selling coins may have a valid disagreement with me.
That's OK.
As an aside.............I love this forum.
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb, Ricko
Bad transactions with : nobody to date
Baley - Am I the only one who saw the humor in that Clip ??
There are a great number of coins overgraded from all of the years they have been grading. Of course no one cracks those out. This is why we only hear about those that go up.
Go to any coin show, randomly buy 100 OGH coins, crack and send. I would be shocked if half did not go down.
I dunno, but it was the first thing I thought of when I opened the thread. TDN's coins are far and away the best there are, and if the rare improvement to any one coin exists, he goes after it and gets it, even if it takes 15 years or more (as in a related TDN thread)
Does it really matter How Much better his coins are than the second, third, fourth etc place sets?
On a scale of 1 to 10, TDNs coins go to 11.
Liberty: Parent of Science & Industry
It could be exactly as surmised by TDN, but could also be due, in part at least, by higher graded coins gravitating towards the higher graded sets. Therefore, the depth of the aggressiveness in putting sets together can naturally close this gap without any gradeflation.
In honor of the memory of Cpl. Michael E. Thompson
Floating "standards" are not standards - they drift with the wind and tide until crashing on rocks or lost at sea.
THIS. The market grading principle is partly an exercise in circular reasoning. The problem, however, is that grades do not fall when the market values fall.
RMR: 'Wer, wenn ich schriee, hörte mich denn aus der Engel Ordnungen?'
CJ: 'No one!' [Ain't no angels in the coin biz]
Bongo bango ....BINGO!
The grades can CHANGE when the market falls.
Thus keeping the "market"....LOOKING.... strong.
Sheesh, I remember being blissfully and stupidly satisfied looking at my 7070.
Is it that new grades do not fall or coins are not regraded when the market falls?