How do you define "RARE" in numismatics?

Not "relatively rare"... but as what I've called the formula, "relative rarity"... maybe I should call my ratio formula something else like "extant ratio".

 

Log in or Sign up to hide this ad.

Sorry, what is your formula, I don't see it anywhere. What is the denominator of the extant ratio?

If it's demand, doesn't that just give you price? Price is a function of supply and demand, your extant ratio (supply/demand) is also a function of supply and demand. Why calculate some new function with a huge error bar on demand when we already have a large market that provides price discovery and is exactly a function of the things you are interested in.

I.e. if there are 10 known coins and they are selling for $300 in decent shape doesn't that tell you how large the demand for that coins is, compared to a S vdb where there are probably 10,000s of thousands extant and they sell for ~$1,000 per coin. Can't you infer from those two pieces of information the concept you are interested in? That is a serious questions, because the best I can tell that is what you are interested in no?

 

This is exactly as you said iPen, "As I define it". So what is the purpose of this discussion? Have you become one of the great thinkers of mankind? Where are you going with this, but most importantly, why?

 

In many instances, the number in existence isn't meaningful because collectors don't want many or most of them. It doesn't change the definition or how many exist, but its probably what actually matters to most collectors of a particular coin or series. Reading the most recent posts by the OP, I think phrasing the topic in this context instead asking for definitions of rarity might have been more useful.

 

Yes, I agree lol. I should have phrased it much better at the onset. I was inundating myself with various questions out of curiosity, and wrote that question ahead of it all.

I don't think you have to be a great thinker of mankind to apply a ratio lol. The purpose stemmed from many examples such as when:

Mintage : # of active buyers for a given coin/type

 

My head hurts..........I don't own nuthin' 'rare' anyway.

 

You are basically looking for a number that relates supply and demand. Why is price not what you are looking for? To the extent you want to compare your ratio of one coin to another, what information would you gain over just comparing the price of the coins given their known rarity?

You can find the rarity of a coin, and you can find the price of the coin. Those are both knowable with a decent amount of accuracy, and provide the relationship you are looking for. On the other hand, the ratio you want to calculate requires you to know the number of active collectors, a much more difficult, and fluid, thing to measure. Why re-invent the wheel when the market is doing the heavy lifting of signaling demand when it does price discovery for you?

 

I'm one of the people arguing against his proposed formula, but I can think of at least one purpose...it could help to determine over-valued vs under-valued items. For example, if one determines that there are 3000 prospective buyers for an item with approximately 200 known examples (1797 half dollar)--a ratio of 15 to 1 and there are the same 3000 prospective buyers for a coin with about 1500 examples still in existance (1815 half) a ratio of only 2 to 1 then the 1797 should cost about 7.5 times the 1815.

Of course, this isn't true. A 1797 half will cost about 80,000 dollars in VF. A VF 1815, on the other hand will only run you about 5 grand. Sounds to me like either the 1797 is vastly overpriced or the 1815 is underpriced.

 

For one, prices are not a linear function of demand like that, I've done regressions on price and always find an exponential relationship. Second, how do you determine the demand? The market provides that information via price, but independent of price, there is no way to find that signal. Third, the relationship between demand and supply regardless of it it's price or done with some other calculation, should behave the same way, because they are both a function of the same underlying relationship. I would posit that to the extent this ratio provides a different signal than the price signal it's because the estimated demand had more error, not because you found an undervalued coin.

The overview here, price is a function of supply and demand. Extant ratio is a function of supply and demand. Because both metrics are entirely governed by the same two variables they should both behave in the same way when compared across coins. But, for the price function you can look at the market and find the price. You can look at research and find the supply. Given that we know with a high degree of certaintaty two of the value, we can get a pretty clear picture of the demand. In the extant ratio, we only know supply and have to guess at demand. So it will always be a much less accurate attempt at providing the same information that price already does.

 

And to a large extent I agree with you. Thus the fact that I have argued against his theory all through the thread. However, I can see where he feels that his theory could be helpful.

 

In which instances do you believe the supply estimates are reasonably accurate? I look at PCGS Coin Facts and see some that appear reasonable and others that do not. My review isn't scientific but in most instances, I don't believe those who make the estimates are in a position to really know.

As examples, I'd say the 1794 dollar at least in MS are accurate, which is eight confirmed according to the owner of the SP-66 and 10 according to Coin Facts which leaves two potential others out there. John Amato's book on the 1796-1797 half dollars I believe itemizes each coin which are known according to "common knowledge". The book on the Peter Getz Washington patterns identifies 22 1792 silver small eagle halves and 56 of the copper/bronze.

On one of the series I collect, Gilboy provides estimates for the pillar coinage. It's not a US series but I believe his claims are way off base. If his estimates were remotely accurate, he should have been able to find decent plate coins for practically every single entry except all of the scarcest considering he must have had access to the best collections. I acknowledge the lower price level creates a larger margin of error but don't see how even this reconciles to his claims.

Similar idea on most US coins per Coin Facts.

 

The ratio isn't perfect and isn't the end all be all, but may be one of many ways to look at what and how one should be buying a particular coin. As you say, finding the price, whether it's too high or too low, will be much easier than calculating the ratio, given the ease of available data from what I can tell. However, price is not always static. As mentioned earlier, some mintages may be low, but prices may be high solely because it is low despite there being no demand. I mentioned earlier, too, that it nearly if not completely synonymous with supply and demand, but it looks at what the price should be so as not to over pay.

 

Yes, this is part of what I'm getting at - I wouldn't want to overpay because a few outliers skewed the price up. I always look for a deal when buying coins, and generally I like to underpay so I look at prices to see if they make sense, but it's not telling the whole story (i.e. if it's actually competitively priced), as you say.

 

With US coins, I don't consider the demand for any low mintage US coin "low". If this is the belief, I'd say the expectations are unrealistic.

 

Right, and why isn't the price static? Generally because demand fluctuates. But we don't know that the demand has fluctuated until we see the price fluctuation. There in lies the problem with using 'demand' as the denominator in any calculation, and more specifically using it to calculate 'what the price should be'. We don't know, and can't find out, the true demand, thus we can't calculate what the true 'price' should be. All we can say is given a certain price relative to another price that demand has changed.

To your point about trying to underpay for a coin and trying to find the price so as to not overpay. How does calculating your ratio, with a huge error bar on demand, give you a more accurate picture of true value for a coin than past auction estimates? We are all in your boat. There are 40 coins coming up for auction this month that I am interested in buying and have been researching. I am trying to set my price on each coin. Given that extant numbers aren't really available, and mintage numbers aren't really available either, and certainly I have no idea how to estimate 'demand', I am looking at previous auction results to set a price. I could guess at all those other numbers and try to come up with a mathematical formula, but at the end of the day previous auction results are a much better guide because the price is already a function of supply and demand.

 

I left out quoting the rest of your post, but I agree with it in general. It furthers the point, if we don't know the supply and we don't know the demand, what is the point of trying to create a ratio of those two concepts. We could guess at them, but then our ratio has a ton of error in it, why would a function that badly estimates the relationship between supply and demand be better than the price discovery the market provides based on the actual supply and demand?

To your question, for a lot of early copper there is a fairly large amount of research into both survival estimates, mintage numbers, and census/conditional census. the survival estimates are generally based on hoard finds. It's not perfect, but it's a reasonable guess. In addition, for each of the 1000+ variets of large cents there is a fairly accurate conditional census of the best coins. It's not the truth in the sense that there are coins out in the wild that are not included in the census, but the census certainly contains all coins known to quite a large collector base, which in reality represents the actual supply of coins at any given time point. I.e. the fact that 2 coins of a variety exist in some old time collection that hasn't seen the light of day in 50 years is not significant to the price I would pay for that variety today because those 2 coins do not actually become part of the supply until they enter the market.

 

Makes sense.

 

This is a best answer! I don't know how long you have been a member here but I have noticed that some posters in particular are deep thinking philosophers. No matter what your answer, they'll come up with another "What If" or "Why" to further suck others into the fog of ideas that usually have no direct answer like "What color is the sky?"

On the plus side, the posts are both entertaining and serious. It's also interesting to see how far they can run up a page count.

IMO, every KNOWLEDGEABLE, intelligent numismatist could define RARE as it applies to numismatics in one sentence. The problem comes when a philosopher tries to cloud a simple, easily determined definition with all sorts of rubbish. For example: "What color is the sky?" I can tell you when we consider the time of year, the time of day, the weather, our color acuity, the position of Mars, .

PS I shall not add my definition of rare here as this thread is too much and I'm learning a lot of the "fluff" that most never consider.

However, I will challenge @iPen to do one thing for the members who have posted here. When this thread runs its course, I challenge you to answer your own question in a very concise complex sentence that could be published in a dictionary of numismatics. Are you up for the challenge?

Then I will add my definition and hope others here do the same. Bet we'll come up with something great!

Until then, I'm

 

I really dislike that word "dreck" . We always hear people say buy what you like and can afford . But by calling a '16-D Merc that , it's like telling all those who bought a low grade key date , though not rare , coin garbage .
Now for Rarity I use 2 scales the Bust half scale with R-1 meaning over a 1000 up to R-8 which is 1-3 . I also use the Fuld scale for Civil War Tokens with R-1 = greater than 5000 (very Common) to R-10 = 1 Unique .

 

I was referring to "world" coins - select limited edition Canadians come to mind.

For example:

I'm sure there are some folks who collect them, but I'd think that the demand is relatively low compared to many of the other limited edition and standard offerings.