NGC Weighting System-Is This How it's Done?
TomB
Posts: 21,342 ✭✭✭✭✭
Over the past few weeks there has been quite a bit of debate and conversation over how NGC might weigh the coins in a Registry set. I was curious to see how it might work and so decided to conduct a very simple and limited experiment to see if I could predict where my set would fall on the NGC weighted slider scales that they use. I'll go through why I chose the set I chose and then what happened. Mercury dimes were chosen because there were very few sets registered that had coins in them (only two at the time) and I also happen to have a small number of NGC and PCGS slabbed Mercs that I could place into a set. Having decided on what set type to use (Mercs) I then decided to list only a single coin in the set and to predict what my score would be before it was listed. The coin I chose was a '21 MS65FB and I predicted that my set would be given a score of 26.7%. Upon addition of this coin to my set I received a score of 27% which is essentially indistinguishable from my prediction. How was this done? Well, it appears that there is a very heavy component in the sets as to the value of each coin in the set. All I ended up doing was dividing my set value by the top set's value and then multiplying by 100% to put it in percent form. That's it. I used the values from Numismedia (NM) and Coin Universe (CU) and averaged the percentages that I obtained. I don't know how well I am explaining this so I will put it out in stepwise form-
1)Total the value of your collection as listed in NM and CU.
2)Total the value of another collection, presumably the number one set, as listed in NM and CU.
3)Divide the value of your collection by the value of the other collection.
4)Do this for both the values obtained by NM and CU, you will obtain two values.
5)Average the two values and then multiply by 100% to obtain your NGC sliding score.
My set had values of $4,160 (NM) and $4,550 (CU). I then compared it to the number two set which had values of $7,074 (NM) and $7,000 (CU). I didn't use Ken's number one set as a reference because he had a coin listed that had no stated value on both sites ('29-S MS67). Anyway, the second set had had a rating of 43%, so, I did the following-
{[(4,160/7,074)+(4,550/7,000)]/2}*100%=NGC slider score
By this method I figured my set was worth 61.9% of the second rated set which had a value of 43%, therefore, .619*43%=26.7%. After this value was obtained Ken was kind enough to remove some of his coins from his set and we continued to caluculate the relative values of the sets as coins were added and removed. The values obtained were always within 1.5% of the NGC sliding score.
In short, it looks like your set score is the ratio of your set's published value divided by that of the top set. I don't know what price guides NGC uses and don't know if there is another component to their formula, however, this gives an excellent first approximation. As a warning, the less value your set has the more likely there is to obtain a large error as the price guides may not agree on price. Also, the more data points (coins) that you have the more refined the number you should obtain. This may not work well at all for truly modern pieces as there is no satisfactory price guide that I am aware of. At least one other set, Lincoln cents, have been tested and they behaved in the same manner as Mercs. Have fun.
1)Total the value of your collection as listed in NM and CU.
2)Total the value of another collection, presumably the number one set, as listed in NM and CU.
3)Divide the value of your collection by the value of the other collection.
4)Do this for both the values obtained by NM and CU, you will obtain two values.
5)Average the two values and then multiply by 100% to obtain your NGC sliding score.
My set had values of $4,160 (NM) and $4,550 (CU). I then compared it to the number two set which had values of $7,074 (NM) and $7,000 (CU). I didn't use Ken's number one set as a reference because he had a coin listed that had no stated value on both sites ('29-S MS67). Anyway, the second set had had a rating of 43%, so, I did the following-
{[(4,160/7,074)+(4,550/7,000)]/2}*100%=NGC slider score
By this method I figured my set was worth 61.9% of the second rated set which had a value of 43%, therefore, .619*43%=26.7%. After this value was obtained Ken was kind enough to remove some of his coins from his set and we continued to caluculate the relative values of the sets as coins were added and removed. The values obtained were always within 1.5% of the NGC sliding score.
In short, it looks like your set score is the ratio of your set's published value divided by that of the top set. I don't know what price guides NGC uses and don't know if there is another component to their formula, however, this gives an excellent first approximation. As a warning, the less value your set has the more likely there is to obtain a large error as the price guides may not agree on price. Also, the more data points (coins) that you have the more refined the number you should obtain. This may not work well at all for truly modern pieces as there is no satisfactory price guide that I am aware of. At least one other set, Lincoln cents, have been tested and they behaved in the same manner as Mercs. Have fun.
0
Comments
Great detective work. I have a feeling that the market value of each coin in each grade plays a heavy role in the NGC system. My main curiousity still lies on the modern side. Where do they figure out those prices, since there are no accurate published price guides.
Keith
Because it's based on price..... A "one" coin wonder will out rank a complete set.
In God We Trust.... all others pay in Gold and Silver!
I think you need to add 25% to the price if it is a NGC * coin.
In God We Trust.... all others pay in Gold and Silver!
Greg
I tried this on my Liberty Nickel set versus the highest rated on the NGC registry using CU prices. Sure enough, the price ratio mine/#1 worked out very close to my score (about 79%). This is interesting because here we are looking at complete or nearly complete sets. One issue raised is whether NGC and PCGS coins trade at the same price for the same grade. Number 1 is about 1/2 PCGS and 1/2 NGC, with many of the high grades being NGC. My experience is that they don't in Lib Nickels, with NGC trading at a discount. Of course, I don't expect NGC to price there coins below PCGS, but it does raise an issue with using generic market prices across holders. An alternative could be Blue Sheet bids. Use these two as a test. Dealers, assuming we both have fairly graded coins, which set would you think is worth more? I would view them as closer to toss up, but I'm not sure how NGC 67s trade, and those make up a chunck of #1s value.
Greg
Still need to test the NGC * coins, then figure out the modern issues.
In God We Trust.... all others pay in Gold and Silver!
In honor of the memory of Cpl. Michael E. Thompson
Registry 1909-1958 Proof Lincolns
Which could explain why NGC doesn't allow low-grade coins to be included in their registry sets.
New collectors, please educate yourself before spending money on coins; there are people who believe that using numismatic knowledge to rip the naïve is what this hobby is all about.