Coin populations: To what extent are these two assumptions true?

RYK

I am working on a project to determine the survival populations of coins in all grades in my series of interest. To what extent do you think that the following two statements are true:

1. Auction appearances of individual rare coins are proportional to their relative rarity, within a given series.

2. Grading events (ie. pop report numbers) of individual rare coins are proportional to their relative rarity, within a given series.

Any other comments or suggestions are appreciated.

Longacre

I would love to hear what the true experts think. But I think both statements are generally false, and there are too many intangibles that go into whether a coin appears at auction, and how many times a rarity gets submitted for a grading event. Again, I would really like to hear what others have to say. This is interesting.

291fifth

The existence of a hoard in the hands of "hold 'em till death" individual would negate your assumptions. There may be more such hoards than most collectors realize.

Russ

3. Values are based on the proximal relationship of existing price guide dynamics.

Russ, NCNE

RYK

3. Values are based on the proximal relationship of existing price guide dynamics.

I was just about to say that...

mgoodm3

Problem you have with auction reports in your series is that some of the coins have a low enough population and don't show up commonly enough to provide an accurate statistical sample. Probably the same problem with grading events.

RYK

Problem you have with auction reports in your series is that some of the coins have a low enough population and don't show up commonly enough to provide an accurate statistical sample. Probably the same problem with grading events

Agree, and I am coming to the conclusion that a key date might be held longer and tighter than a common date. I will be working on this over the weekend and will report some preliminary observations for discussion next week.

ColonialCoinUnion

If you want any sort of accurate estimates of extent populations, I would suggest that the only way to do that with any degree of accuracy is to trace plated examples of a given coin type in old auction catalogs, reference books, dealers listing, museum holdings, etc. and to build a photo census of these items.

I've been doing this for years for colonials and I have a reasonably good assessment now of what exists and in what grade. It took me countless (I mean really, really countless) hours to get there but I have no doubt it could be done in a relatively specialized area such as yours. I would not suggest that someone try it for Lincolns, however.

Based on my experience, I can tell you that the number of auction appearances of coins are wildly skewed by the same coins re-appearing mutliple times even though catalogers through the years have identified them as different specimens, at different grades, or, either intentionally or unintentionally with faulty information, have misdescribed them, etc. And as you know, often pictures of the same item look different in different appearances and are only identifiable as the same item after very careful research.

My advice also would be to ignore the pop reports altogether. These are too susceptible to manipulation through resubmittals and prone to errors in data entry. Perhaps it is more of a problem in colonials, but I am aware of a number of cases of a coin submitted, miscatergorized and never corrected leaving, for example, a pop 1 finest known in the wrong coin# holder thus rendering the pop reports pretty meaningless.

cladking

<< <i>I am working on a project to determine the survival populations of coins in all grades in my series of interest. To what extent do you think that the following two statements are true:

1. Auction appearances of individual rare coins are proportional to their relative rarity, within a given series.

2. Grading events (ie. pop report numbers) of individual rare coins are proportional to their relative rarity, within a given series.

Any other comments or suggestions are appreciated. >>




Good luck.

1. Yes, but minus a constant that increases with rarity. Very rare coins tend to be a little
less likely to cross the auction block unless they are worth a great deal of money.

2. Yes, if you can exclude ranges where the coin is worth much more in slightly higher grades.
There are probably a few coins which have a substantial number of liners as well but I don't
know which these would be.

Getting good estimates of the existence of any coin is a daunting task after collectors get in-
volved and affect the attrition rates. Coins are often hidden from view for many decades and
an individual specimen may trade repeatedly over a short period of time. Many collectors horde
specific coins. While collectors will often help with determining availability, others will provide no
help at all or might even try to confuse the issue.

I've always suspected that the best bet is to simply study the coin from the time of issue and
make inferences from attrition, assumed events, and the known facts. There's a lot of insight
to be gleaned just from the price guides alone.

fc

There's a lot of insight to be gleaned just from the price guides alone.

There is a lot to be said for that statement.
I do think the populations at NGC and PCGS will be helpful though
to get an idea of where the coin stands grade wise. As in, what
is the finest known by a grading company or maybe knowing that
finding an example in AU50 is outstanding.

Good thread, i will think about it and post more when at home.

Boom

Population is not etched in stone, never can or will be. Everytime a coin is sent in for regrade, cracked out and grades higher, lower or the same, when collectors free them from their plastic tombs to place them in Dansco or other type albums, population is thrown completely out of kilter.

Example ... Collector thinks his 66 is a 67 and cracks it out/ coin comes back lower, higher ... whatever grade ... that pop goes up and the original stays the same. You now have 2 coins in population when in actuality there is only One, the one it was and the one it now is. If it comes back a 6 then it appears a new 6 has been made. Same as if it comes back a 5 or a 7 because it still shows as a 6 because MOST collectors do not report the changes.

With that, even so, I have personally reported changes in grade yet still population of said coins were not changed. Then there are the crossovers. A coin will show as being in Census or population with one or the other service and added to the new one. Again, it now looks like there are two when there is actually one.

The whole population/ census thing is a giant fiasco. I personally know of a Registry participant who owns a very high ranking set that buys certified coins, removes them and places them in his "Raw" collection. This transpires all the time. Mix in the submit/ re-submit, re-submit. re-submit till you quit collector/ Dealer and things get really hopeless.

The ONLY sure fire way for even a temporary fix would be to have every certified coin in population sent in and recorded as proof positive. No coin produced - certification # deleted and pop adjusted accordingly. Obviously, I place NO STOCK in the credibility of population/census.

JMHO, of course.

DeusExMachina

RYK, great post. I've been thinking about this a lot recently. In fact this site, www.dahlonegagold.com, is what started it all. They run down each "D" coin and then estimate the number of survivors. If you look at the 1839-D $2.50, for example, it has an estimated survival rate around 1.6% (225/13674). How did they come up 225? And if that is true, these coins are extremely rare.

Unfortunately, I have no insights for you. I am just interested in the outcome.

MrBreeze

I started a thread a month ago asking the same things. Concerning your statements, I will address those at the end of my short, cautionary tale. The populations are nearly impossible to figure out without a great deal of assumption. It will take a number of types of information just to determine a relative guess. I am undergoing the same process of determining "real" populations in the Seated Quarters series. There is a relationship between rarity and price, but demand, not rarity, is more of a portend when referencing a price guide. In the seated quarter series, there are some very expensive coins. They are well-known as rarities. The price for a decent PCGS slabbed example of these coins is multiple thousands, increasing exponentially in advance of grade. But their market price, while loosely based on "assumed" rarity, is more a function of demand and the numismatic lore associated with the coin. With that, there are also numerous variables(intangibles, as Longacre states) that go into auction prices and price guides, in general, to fully realize a modifier or formula to obtain your intended answer. Experience(as CCU states) is often one of the most valuable pieces of information. It does not provide a direct answer, but it does probably reinforce what you already know. I ask myself this question when I find myself studying and crunching the reference numbers too much. Do I really need a black and white answer to know what I already know from the simple perspective of availability? I have not found any evidence to contradict my original assumptions concerning populations of seated quarters. I will continue to look for some "hard" numbers, but that probably is a function of my enjoyment of research and mathematical calculation, and not any sane person's actions.




<< <i>1. Auction appearances of individual rare coins are proportional to their relative rarity, within a given series. >>



I think, to a large degree, yes, with exceptions. I have found that to be true in seated quarters. The only caveat to that is the lack of demand for certain coins actually helps keep their auction appearances down. The seated quarter series seems really focused on certain coins at auction right now. But, when an uncommon date comes out, even if it is not considered by any metric to be rare (other than the fact that you don't see it that much at auction), it gets strong bids. So , while your statement may be a bit too concrete, it will work on some level (relative rarity, within a given series)




<< <i>2. Grading events (ie. pop report numbers) of individual rare coins are proportional to their relative rarity, within a given series. >>



Again, as in statement 1, you use the term, "relative rarity, within a given series." You are asking for a relative value not to other coins, but against other coins of the same make and model, for lack of a better term. I would again have to agree with this statement, with a few reservations. The key dates have many years of marketing and misperception behind them. That equates to a low level foundational demand that is hard to remove. So, you must account for that demand in population reports. How you do that, I don't know, as I am working on it myself. Also, although I would not like to say it, the self-fulfilling prophecy of, price guide begets crackout begets price guide, has added populations to coins which do not exist. Someone made a very good point on another thread about a coin which seemed woefully off the mark, as far as population goes. When the dealer who had 13,000(?) crackouts sent them back to PCGS, there were somewhere near 100 examples of this one date, which, specialists knew were rare, but pop reports had greatly skewed.

Yale68

Both assumptions are clearly FALSE. Very common coins are almost never put in an auction, and they also are rarely sent for certification, so they don't appear in population stats. For example, a 1921 Morgan dollar in MS60 is more common than one in MS64, but you would never know it by looking at auction frequencies or population stats.

AgMe

<< <i>I am working on a project to determine the survival populations of coins in all grades in my series of interest. To what extent do you think that the following two statements are true:

1. Auction appearances of individual rare coins are proportional to their relative rarity, within a given series.

2. Grading events (ie. pop report numbers) of individual rare coins are proportional to their relative rarity, within a given series.

Any other comments or suggestions are appreciated. >>




The one thing that pops into my mind is early copper, "why is this slabbed? Give me a EAC grade..."

Just a thought.

RYK

Both assumptions are clearly FALSE. Very common coins are almost never put in an auction, and they also are rarely sent for certification, so they don't appear in population stats.

In the OP, I fairly explicit in that I would be analyzing and discussing rare coins. For example, can you use pop reports and auction appearances to accurately assess the relative scarcity of an 1871-CC dime and an 1862-S dime, to make a somewhat apples-to-apples comparison? I am not interested in condition at this point, just overall scarcity in all grades.

RBinTex

4. (Russ did #3) Pop reports OVERSTATE the number actually graded (duh...) BUT are NOT relative in their overstatement since rarer/more valuable coins have had more crackouts (& thus have relatively larger overstatements).

ColonialCoinUnion

<< <i>4. (Russ did #3) Pop reports OVERSTATE the number actually graded (duh...) BUT are NOT relative in their overstatement since rarer/more valuable coins have had more crackouts (& thus have relatively larger overstatements). >>



I agree but I think the problem is worse / more inconsistent than even that. One person on a mission to make one upgrade can and has ruined the accuracy of rankings for a coin. Get a few guys like that pounding away on the services and you have pops that bear little resemblance to anything, are not comparable to one another and are actually more misleading than they are useful.

Its a jungle out there.

fc

an interesting way to look at the half eagle series
is to goto the registry and look at the weights.

This is for P and S half eagles.

NoMotto - Weight

1854-S 10.00

1864-S 9.00 <- pcgs xf45 26,000 M3888

1858-S 8.00
1859-S 8.00
1861-S 8.00
1862-S 8.00
1863 8.00
1865 8.00
1866-S 8.00


--------------------------

Motto - Weight

1875 10.00

1876-S 8.00

1866 7.00 <- 2250 xf45 pcgs M6700
1866-S 7.00 <- 2900 xf40 pcgs M35000
1867 7.00 <- 1867 xf40 pcgs M6900
1867-S 7.00 <- 23500 au55 pcgs M29000

-----------------

so we are all doing this in our own ways, using different input.

any dealer in rare gold will recoginize and prob agree with
the data above. notice the civil war dates.

Russ

<< <i>Two what extent are these two assumptions true? >>



Man, I can't believe I missed that earlier.

Russ, NCNE

RYK

Me, too.

morgannut2

These assumptions aren't going to help much. The idea of a photo/image report from all old auction catelogs is a good start for rare and uncommon coins. Then you calculate a frequency of new unknown coins appearing --say over 10 or 20 year intervals, to project the unknown-unrecognised-privately traded material that will eventually show up in the market. That gives an upper time frame and population for a future given time frame, or new high price market.

By the way, the major oil companies have used this method of populations discovered vs. unknown through time, to estimate total oil field populations over a given area of exploration at the end of searching. It's accurate enough to produce an 18% profit on exploration and guessing the total census of known verses unknown, so I guess it works OK.

RYK

CCU,

The photography idea is not going to cut it. There are too many coins (up to 300 per date).

Follow-up question:

Would you expect the ratio of coins in PCGS:NGC holders to be relatively constant?

I am accumulating and analyzing data and hope to have something presentable next week.

morganbarber

I would expect most rareties to be slabbed in a higher percentage compared to their populations. What percentage of existing 16-D Mercs are slabbed? What percentage of surviving 1944-P Mercs are slabbed? I would expect PCGS and NGC population reports to overestimate percentage of 16-Ds vs. 44s.

ColonialCoinUnion

<< <i>CCU,

The photography idea is not going to cut it. There are two many coins (up to 300 per date) >>

.

I didn't say it wasn't a gargantuan task, but note that I have hundreds of photos of many specific issues. So it can be done.



<< <i>Follow-up question:

Would you expect the ratio of coins in PCGS:NGC holders to be relatively constant? >>



Not necessarily depending on the mood of a few major crack-out guys.