How do you determine split values from the Grey Sheet?

loanshark

Just got a new subscription to the Coin Dealer Grey Sheet for Christmas and have a question.

How do you determine the values on split grade coins? For example, the 1895-S Morgan in XF-40 shows it to be worth $395 while the next level of AU-50 shows it to be worth $1,325.

My NGC graded coin is a XF-45. Would you split the difference to come up with an approximate value?

Loan Shark

ccex

I think it's too dangerous, and too simple to just take the average of the prices in the adjacent grades. I try to picture a graph of prices according to grade, and interpolate a value along the curve. Often there are exponential curves, where the value of the midpoint is much closer to the lower published value than to the next one up.

Your example of an 1895-S Morgan shows relatively minor price increases through the circulated grades. The first big jump is between XF-40 and AU-50. If the XF-45 is really nice, you might interpolate a graysheet bid of $500 to $600. But I wouldn't bank on finding many dealers who would buy the coin at much above XF-40 bid, unless they felt they could sell it at an AU price. I don't think too many people spend much time with advanced mathematical formulas to determine the price, which seems to be determined more by skepticism, eye appeal, and a match of wits between buyer and seller, with perhaps a bit of help from a grading service.

DennisH

I just sold a PCGS-45 1895-S on eBay for $793, which is right what PCGS-45s have been bringing lately. Can't say what an NGC-45 would bring, though.

loanshark

Thanks for the help guys.

Is there a publication that will also show the split grade values? I have several like this.

Loan Shark

DennisH

There's no price guide I'm aware of for split grades. Unless you follow them and know what they're bringing, you can leave a lot of money on the table selling split grades too cheap.

I've been a fan of split grade better dates ever since I was building my Morgan set. Back in the mid to late-90s dealers would typically price split grades very close to the lower grade published in the Grey Sheet. Today, that has changed dramatically. Now most of those coins sell for significantly closer to halfway money, if not more.

Coinhusker1

<< <i>I think it's too dangerous, and too simple to just take the average of the prices in the adjacent grades. I try to picture a graph of prices according to grade, and interpolate a value along the curve. Often there are exponential curves, where the value of the midpoint is much closer to the lower published value than to the next one up. >>



Yes, based on all the data points by grade from the coin's value, derive the equation which produces the best fit, perhaps using a least squares method.

Actually what ccex said makes sense --- the price will be likely be closer to the 40 price (even closer than an exponential formula would indicate), but will be heavily driven by the coin itself (eye appeal, under/over grading).

FrankHKP7

To add to what was stated above, you could make a graph in Excel and it would plot the curves for you. Plug in the numbers and graph it!

Frank

dizzlecc

The value depends on what kind of loan rate discount and how many fees you will waive. Coins qualify as collateral, right?

Many of the coins graded under ms63 are variable. It depends on how nice the coin is, what you paid for it and what someone else is willing to pay.

It seems the key date coins sell closer to the next grade up and the more common date coins sell closer to the lower grade.

au58

The rule of thumb that I use is to add one third of the price differential to the price at the lower grade.
Using the example above provided by DennisH., the difference between $1325 (AU50) and $395 (XF40) is $930. One-third of $930 is $310.
$310 + $395 is $705. His coin sold for $793.
It's not perfect, but it helps.

Merc

Try pricing a Vf35. There is little to no difference between a VF20 and 25, but a 35 has much more detail.

loanshark

I like the one third idea. I used it in the price I have paid for some recent split grades and it comes pretty close. Also close to some recent sales I have been watching.

Thanks guys,

Loan Shark

morgannut2

If it's any help and you're serious, auction values commonly follow a log-geometric curve except at the high and low ends.
You won't get the curve to fit unless you convert to logs. An exponential curve is a close approximation too. But since
you are really graphing human perceptions of value, the easiest thing is to average each grade in the historical file
of sales at Heritage Auction History. You will find that bidding is irrational!!