[ancients] Statistics : Auction estimates vs prices realized

In a thread earlier this week, the topic of estimate prices provided by auction houses at their auctions was mentioned. Some of the opinions offered... the estimates were essentially meaningless, they were just a guess as to a reasonable sale price, they are primarily part of auction house strategy used to tempt bidders. Since it all sounded abit sinister to me, I thought I would do some investigative statistic gathering to shine some light on what seemed like a murky area.

I picked two auctions that were closed in the last few months (Pecunem 14 and Roma 4) and compared all the estimates provided against the final closing prices. I then crunched out some statistics and info that you'll find below. I know it's only two random auctions out of a gajillion, but I'm sharing the data in the hopes that some will draw either useful conclusions or interesting trivia from it. My own conclusion - I have way too much time on my hands.

----

ROMA 4
28 December 2013

Number of lots auctioned = 1076
Number of lots sold = 816
Number of lots unsold 260 (24%)

No. of lots closing under estimate = 286 (35%)
No. of lots closing at estimate = 69 (8%)
No. of lots closing over estimate = 461 (57%)

Average estimate on each lot sold = GBP 161 (USD 267)
Average final closing price of each lot sold = GBP 193 (USD 320)

Percentage by which average closing price exceeds estimate = 20%
Percentage by which average lot that closes above estimate exceeds the estimate = 61%

Breakdown for lots that closed over estimate :
1 - 50% over estimate = 61% (281/461)
51 - 100% over estimate = 22% (100/461)
101 - 200% over estimate = 12% (54/461)
201 - 400% over estimate = 4% (19/461)
>than 401% over estimate = 1% (7/461)

Worst case of overestimatitis :
Macedon, uncertain mint AR Hemiobol. Circa 450 BC.
Estimate: GBP 30
Closing Price: GBP 600

----

PECUNEM 14
2 March 2014

Number of lots auctioned = 1095
Number of lots sold = 929
Number of lots unsold = 166 (15%)

No. of lots closing under estimate = 288 (31%)
No. of lots closing at estimate = 53 (6%)
No. of lots closing over estimate = 588 (63%)

Average estimate on each lot sold = EUR 125 (USD 174)
Average final closing price of each lot sold = EUR 197 (USD 274)

Percentage by which average closing price exceeds estimate = 58%
Percentage by which average lot that closes above estimate exceeds the estimate = 90%

Breakdown for lots that closed over estimate :
1 - 50% over estimate = 45% (265/588)
51 - 100% over estimate = 24% (139/588)
101 - 200% over estimate = 16% (97/588)
201 - 400% over estimate = 11% (63/588)
>than 401% over estimate = 4% (24/588)

Worst case of overestimatitis :
MACEDON. Koinon. Pseudo-autonomous. Time of Gordian III (238-244). Ae.
Estimate: EUR 50
Closing Price: EUR 1100

 

Log in or Sign up to hide this ad.

I wonder if it is a coincidence that both coins that suffered from 'overestimatitis' are Macedonian or if there are two bidders with deep pockets that both really want the coin.

 

Awesome info Z, you show add how much they make on a auction @21%..

 

I would be curious to see what percentage of CNG and other US firms, with usually lower starting bids, have of unsold lots. I know CNG electronic auctions usually have a very low percentage of unsolds.

 

Perhaps, but lucky for us! Thanks for assembling that information-- it is very interesting.

 

Interesting indeed.

 

Absolutely awesome => those are very sweet numbers, Z (nice analysis) ...

=> thanks for verifying that I'm gonna spend at least twice as much as the initial estimate!! (oh well => dog will hunt!!)

again => great job (ummm, do you need a job crunching some mining numbers?)

Cheers, brother

Oh and "please" keep-on analyzing our future/current auctions (it is very interesting/useful)

 

As it's just 2 auctions it's obviously not a basis for serious analysis, just something I thought would be fun. How much you're going to spend relative to the estimates is still going to depend on how 'accurate' the estimates provided are, who's bidding at any given auction, and whether or not they want what you want. I would have done a CNG auction too, but I got tired after awhile.

But here's a bonus stat:

Percentage of lots that closed near their estimates (+/- 20%) :
Roma 4 = 60%
Pecunem 14 = 12%
The rest closed higher.

 

This was posted on ForumAncientCoins a while ago by a member named Lloyd Taylor:

 

From CNG's posted record of a recent auction (E320, 12 Feb 2014), what I can see is 719 lots sold out of 743 put on auction. So a mere 24 lots or 3.23% either unsold or withdrawn. Much lower than Pecunem14 (15%), or Roma4 (24%).

The percentage unsold for their most recent auction (E322) is about the same : 3% (25/824).

 

Zumbly, I would think that if a coin sells for more than the estimate the estimate was an "underestimate," so your worst cases were of "underestimatitus."

 

Ah, you're definitely right about the estimates in those cases coming in 'under'. I guess I was thinking on the coins themselves closing 'over'.

 

I had a little (more) time to kill, so I did the most recent CNG eauction as well.
Note: where the 2 previous auctions mentioned in the OP had starting prices 20% below their estimates, almost all the lots in this auction had starting prices 40% below their estimates.

CNG E-322
12 March 2014

No. of lots auctioned = 824
No. of lots sold = 799
No. of lots unsold = 25 (3%)

No. of lots closing under estimate = 228 (29%)
No. of lots closing at estimate = 36 (4%)*
No. of lots closing over estimate = 535 (67%)

*No. of lots closing +/-20% of estimate = 269 (34%)

Average estimate on each lot sold = USD 191
Average final closing price of each lot sold = USD 314

Percentage by which average closing price exceeds estimate = 64%
Percentage by which average lot that closes above estimate exceeds the estimate = 95%

Breakdown for lots that closed over estimate :
1 - 50% over estimate = 43% (230/535)
51 - 100% over estimate = 28% (159/535)
101 - 200% over estimate = 16% (88/535)
201 - 400% over estimate = 9% (47/535)
>than 401% over estimate = 2% (11/535)

Top 3 Overachieving Coins :

#1 - L150, KINGS of MACEDON. Philip III Arrhidaios. 323-317 BC. AR Drachm
Estimate $100
Closing Price $1000

#2 - L539, Hadrian. AD 117-138. AR Denarius. Alexandria reverse.
Estimate $100
Closing Price $950

#3 - L277, IONIA, Uncertain. Circa 625-600 BC. EL Myshemihekte
Estimate $300
Closing Pricce $2600

 

"Overachievers". Perfect descriptor.

You know we're going to expect you to parse the data for each auction now, right?

 

I'd love to, but I may be slightly busy going forward. I mean, check out my to-do list :

1) Write business plan for Coin Auction Data Analysis Service.
2) Start talking to small, select group of suckers potential investors.
3) Get in touch with John Anthony to do a bundle with his Coin Approval Service.

 

What?! John didn't tell me there was another partner. Now we have to split the megaprofits three ways? Grrr.

-----------------------------Press Release--------------------------------
​

Certified Ancient Coin Acceptance™ Corporation proudly announces their long-awaited ancient coin verification program. Protect your investment now with a genuine CACA sticker.

​

Get more bang for your Boeotian, more value for your Vabalathus!

Here's why:

• Proven increased resale value*
• Peace of mind knowing your coin has been expertly assessed**
• Easy-open slab***
• Low cost!

Basic package, $24.99†

• Basic cleaning with a spit rag
• Easy-open slab
• Best-guess attribution
• Green CACA sticker

Optional upgrades:

• Deluxe cleaning with sulphuric acid, add $19.99 (plus $100 for hazardous waste disposal)
• Gold CACA sticker indicating Overachiever status (hammer price >20% of estimate), add $9.99
• Smiley sticker to indicate just how happy you are with your purchase, add $4.99 (free if you "like" us on Facebook)
• "Better" attribution (wildwinds search), add $19.99
• Supply your own attribution, add $24.99
• Bar code on label, add $4.99
• Unique bar code on label, add $9.99

* When shown the same Gallienus with and without a CACA slab and sticker, my mother was more impressed with the CACA. She thought the sticker was pretty.
** Applies to Nabataeans only.
*** Easier to replace the coin with a coin of your choosing.
† plus excise, duty, handling, shipping, VAT where applicable, and a 21.5% buyer's fee.
Offer void where prohibited by law. And as always, ask your doctor if CACA is right for you.

 

Ahahaha TIF, that's hilarious!! (I shot coffee through my snout!!)

Oh, but I think you may have forgotten to add-in one thing:

"If symptoms last for more than four hours, consult your physician"

 

Fantastic effort on your analysis, Z-bro

=> please keep-up the great work!!

 

Don't miss out on your free sticker (with purchase). "Like" CACA on Facebook!

 

Hmmm? ... man, that is a very cool and prestigious sticker (it is a very tempting offer) ... but sadly, that would mean I'd have to sign-up for Facebook ...

Oh, and as I've already told you => I have enough addictions without adding-in "Creepy stalker-guy on Facebook!!"