2400 CGC books on ebay....

[thecomicsheet]

I do include reserve auctions that receive bids in our statistics. I had a long involved discussion about this with the editors at CBG. Our final consensus was to leave them in since, in the absence of a reserve, that IS the price that the book would have traded at. In other words it does represent a fair market value for the book at that moment.

 

If the reserve on an auction is $4000, the second-highest bid is $2000, and the first-highest bid is $3500, then it looks from your statistics as if the most someone was willing to pay was $2100--which is far from the truth. Including items which don't meet reserve throws figures into the statistics that are lower than the actual highest price bid. I can't imagine a situation where those numbers wouldn't bias the mean or median price of a set of pricing data points--rather dramatically in the case of some of the rarer, more expensive comics where only a few people have the disposable income at the time to bid.

 

Perhaps a better way to say it in this example is that the reported auction close is at $2100 whether or not there is a reserve price set. It does NOT close at the highest price that someone is willing to pay regardless of a reserve. That's not the way eBay auctions work.

 

The reported price on eBay is the sales price (in the absence of a reserve or if reserve is met), or what would have been the sales price in the absence of a reserve.

 

The example you gave is a good illustration. (I don't remember what the minimum increment is at the $2,000 level. Let's just use $100 for our example. Now suppose you have the highest bidder who bids $3500 and the second highest bidder bids $2000. You have 4 cases:

 

1) No reserve. The auction closes at $2,100 (Even though the high bidder was willing to go $3500)

 

2) Reserve set at $4000. The auction still closes at $2,100 although no sale takes place (Even though the high bidder was willing to go $3500). This is the price the auction would have closed at in the absence of a reserve - just the same as 1).

 

3) Reserve is set higher than $2,000 but less than or equal to $3500. Auction closes at the reserve price. (This can inflate the final price versus 1 or 2 IF the reserve is above $2100.) In essence the seller takes over the second highest bidders roll with the reserve price.

 

4) Reserve is set at or below $2,000. Auction closes at $2100 (Again even though the high bidder was willing to go $3500). The reserve has no effect on the price since 2 bidders went above the reserve.

 

In NO case does the presence of a reserve cause a LOWER closing price to be reported than the price that would be reported without a reserve.

 

Remember, in eBay auctions, the auction does NOT close at the highest price someone is willing to pay. It closes at the SECOND HIGHEST PRICE someone is willing to pay PLUS THE MINIMUM INCREMENT at that price level or the highest bid whichever is lower.

 

Hope that helps. I had to think about this a lot a year and a half ago before I finally got it straight in my head.

[fantastic_four]

That's a pretty good enumeration of the possible outcomes of an auction, but I think I need a summary of what your point is--how does the reporting of reserve prices not bias the data? Are you inferring that scenario #1 balances out scenario #2?

[thecomicsheet]

That's a pretty good enumeration of the possible outcomes of an auction, but I think I need a summary of what your point is--how does the reporting of reserve prices not bias the data? Are you inferring that scenario #1 balances out scenario #2?

 

My point is trying to arrive at the best representation of the 'market clearing' or 'fair market value' price for a book on eBay. In scenarios 1 and 2 the 'market clearing' price is $2,100. And the presence or absence of a reserve in those 2 cases had NO effect on the reported price.

In scenario 4 there is also no effect on the reported price in the presence or absence of the reserve.

It is only in scenario 3 where the presence of a reserve can have an upward bias on the final price by, in effect, letting the seller enter into the auction with a 'shill' bid set at the reserve price.

 

Note that many people are under the mistaken impression that reserve auctions cause reported prices to be lower than they would be in the absence of a reserve. The exact opposite is the case. Reserve auctions, because of the reasons outlined in scenario 3 above, have a slight UPWARD bias in the final reported price of an auction.

[joe_collector]

That's a pretty good enumeration of the possible outcomes of an auction, but I think I need a summary of what your point is--how does the reporting of reserve prices not bias the data?

 

The only way I could understand your point, would be to eliminate ALL reserve auction data, both Not Met and Met.

 

Is that what you're saying?

 

I could actually agree with that point, as anytime you have a Reserve, it can influence the data both up and down.

[thecomicsheet]

That's a pretty good enumeration of the possible outcomes of an auction, but I think I need a summary of what your point is--how does the reporting of reserve prices not bias the data?

 

The only way I could understand your point, would be to eliminate ALL reserve auction data, both Not Met and Met.

 

Is that what you're saying?

 

I could actually agree with that point, as anytime you have a Reserve, it can influence the data both up and down.

 

My goal with my database and in my reporting of the market is to help arrive at 'market clearing prices' on eBay. To that end, reserve auctions make up a significant percentage of eBay auctions, especially on high dollar books. So I feel that including reserve auction data is very important.

 

And to repeat again, reserve auctions do NOT result in lower reported prices than non-reserve auctions. In most cases there is NO difference from the price that would be reported in the absence of a reserve. The only exception to the same price being reported is a slight UPWARD bias in prices that happens when there is a single bidder that exceeds the reserve. While this latter case does happen in a percent or two of auctions, the effect is usually fairly small.

 

This explanation of course only takes into account the mechanics of how bidding is done on eBay. It cannot explain any psychological effects that reserves may have on bidding. Some people believe that reserves have a depressing effect on prices as it 'turns off' some prospective bidders. Others believe that 'thrill bidders' may bid a little higher than normal if they know or believe their bid is still coming in below the reserve. I have no way of knowing if either effect is large.

 

(Does anyone remember if a seller can lower or eliminate a reserve on an auction after it has received bids? If so, I think thrill bidding would be very risky behavior indeed.)

 

(Whew! My original 2 cent explanation has turned into a nickels worth! )

[fantastic_four]

My goal with my database and in my reporting of the market is to help arrive at 'market clearing prices' on eBay. To that end, reserve auctions make up a significant percentage of eBay auctions, especially on high dollar books. So I feel that including reserve auction data is very important.

 

I appreciate your explanations...I'm not trying to put you on the spot or anything, and if I already have, I apologize...I'm just curious about the topic and I'm pretty sure GPAnalysis has chosen to toss out unmet reserves, so I'm trying to understand the alternative viewpoint. We can take this to PM or e-mail if you prefer.

 

Let me increase the numbers a bit to hopefully make the point clearer. If Metropolis listed an Action #1 CGC 9.0 unrestored on e-bay, and the bidding went up to $150,000 with reserve not met, what good is it for us to have that $150,000 price if in reality Geppi threw in a $750,000 bid but still couldn't meet the reserve? It seems misleading to me--especially if I know phantom numbers like this run throughout the dataset. The only way I can think of to deal with it is to toss unmet reserves out entirely, or to report unmet reserves in a different statistical category.

 

 

And to repeat again, reserve auctions do NOT result in lower reported prices than non-reserve auctions.

 

I still can't see this point...reporting the last price on unmet reserves biases the means/medians. Whether you're talking about an E-Bay auction with a reserve which was met, or a dealer who has set a fixed price on a book, there will always be people who were willing to pay more than the price they actually paid. Reporting unmet reserves is almost exactly the same as reporting the highest offer someone has made to a dealer...but if someone comes into a shop and offers a dealer $5 for a book with a $500 price tag and that's the only offer he's ever gotten...what use is mixing in that $5 top bid with the actual sales data if it's less than the dealer would have accepted anyway? Including unmet reserves mixes lowball offers in with the rest of the real sales. Of course, with E-Bay, it takes a second underbidder for us to see a little bit more about how far the top bidder was willing to go...but if another guy in the local comic book shop makes a follow-up offer of $15 to the first guy's offer of $5...it's still just as useless as the first guy's bid.

 

I can see reporting unmet reserves by themselves, but not tossed in with the actual sales data. If E-Bay's system would SHOW you what the maximum bid was, then I could see your point...but it hides that number from you. Because E-Bay protects the maximum amount a bidder was willing to pay, the number they do report is fictitious, and therefore, it's less useful than actual final prices.

[thecomicsheet]

My goal with my database and in my reporting of the market is to help arrive at 'market clearing prices' on eBay. To that end, reserve auctions make up a significant percentage of eBay auctions, especially on high dollar books. So I feel that including reserve auction data is very important.

 

I appreciate your explanations...I'm not trying to put you on the spot or anything, and if I already have, I apologize...I'm just curious about the topic and I'm pretty sure GPAnalysis has chosen to toss out unmet reserves, so I'm trying to understand the alternative viewpoint. We can take this to PM or e-mail if you prefer.

 

Let me increase the numbers a bit to hopefully make the point clearer. If Metropolis listed an Action #1 CGC 9.0 unrestored on e-bay, and the bidding went up to $150,000 with reserve not met, what good is it for us to have that $150,000 price if in reality Geppi threw in a $750,000 bid but still couldn't meet the reserve? It seems misleading to me--especially if I know phantom numbers like this run throughout the dataset. The only way I can think of to deal with it is to toss unmet reserves out entirely, or to report unmet reserves in a different statistical category.

 

Don't worry about putting me on the spot. I rather enjoy the topic. I'll keep talking about it as long as people are interested.

 

I'm at work and I only have time to respond to this part of your post right now.

 

Let's examine a couple of variations on the example you gave.

 

Variation 1:

Let's suppose a reserve was set at $1,000,000, let's suppose the second highest bidder bid $150,000, and the highest bidder bid $750,000. I don't know what eBay's increment is at $150,000 (I don't get to bid at that rarified level ) but let's just assume it is $5,000 for the sake of this example.

What eBay reports:

1. The high bid is reported as $155,000. (The $750,000 is hidden from us.)

2. The auction is reported as 'reserve not met'. No sale takes place.

 

Variation 2:

Let's suppose there is no reserve, let's suppose the second highest bidder bid $150,000, and the highest bidder bid $750,000..

What eBay reports:

1. The high bid is reported as $155,000. (The $750,000 is hidden from us.)

2. The book sold for $155,000.

 

Note in both cases the bidders bid the same ($150,000 and $750,000).

Note in both cases the final bid reported by eBay is the same: $155,000.

Note in both cases the actual highest bid is the same ($750,000) AND that the actual highest bid is hidden from us.

The fact that there was a reserve in this case had absolutely no effect on the final bid reported by eBay.

 

At that moment on eBay the 'fair market value' of the book ON EBAY was $155,000. The ASKING or RESERVE price someone puts on something has NO bearing on it's worth in the marketplace. Something is worth only what it can be sold for in the marketplace. If I put a $1,000,000 reserve on my beautiful Incredible Hulk #181 in 9.2, I'm not going to get it, the book isn't worth it. (Unabashed plug - please bid on my Incredible Hulk #181 in 9.2 - it ends this evening - my reserve is LOTs less than $1,000,000 I promise )

 

So, now suppose I am lucky enough to own a copy of Action #1 in 9.0 condition. Let's suppose the only auction in recent history of the same book ended as above with a final reported bid by eBay of $155,000. It doesn't matter to me if it was a regular auction or a 'reserve not met' auction. I have a reasonable expectation that the 'fair market value' of my book ON EBAY is about $155,000.

 

How would I play this in MY auction: If I had lots of time, I would hope for the huge gap between highest and second highest bidder like we have in the example auction. I would probably list it with a reserve some $$$ above $155,000, hope the highest bidder crosses that boundary, and sell it at a higher price. If noone hit it, I would drop the reserve and try again later. Repeat until sold. (I believe some sellers follow exactly this strategy - giving up listing fees and not selling very often, but trying to snag THE highest bid with high reserve prices.)

 

If I didn't have lots of time or would be happy with $155,000, I would probably put a reserve at about $150,000 and hope to snag an equivalent series of bids.

 

I'll try to respond to more of your post as time permits and interest continues. Got to get back to work.

[fantastic_four]

So, now suppose I am lucky enough to own a copy of Action #1 in 9.0 condition. Let's suppose the only auction in recent history of the same book ended as above with a final reported bid by eBay of $155,000. It doesn't matter to me if it was a regular auction or a 'reserve not met' auction. I have a reasonable expectation that the 'fair market value' of my book ON EBAY is about $155,000.

 

But if that Action #1 CGC 9.0 were really put up on E-Bay with NO reserve, there would have been more bidders! It is extremely unlikely it would ever sell for that $155,000 total without the reserve. Maybe Bill Hughes called Metropolis on behalf of Jay Parrino, asked what the reserve was, and when he heard it was $1,000,000, he decided there was no use in bidding because Parrino wouldn't go that high. In some cases, the seller tells bidders freely what the reserve is via e-mail--or even right in the auction--and if that reserve is too high, people might just not bid because they know they're not going to pay what the reserve is.

 

I just can't see that phantom reported price on a reserve auction being useful--and I would be really ticked if an insurance adjuster used those numbers to place a value on an item I owned but somehow lost. Most of the time when I see reserve prices not being met, the final visible price is around 50% to 80% of market value, but every once in a while, I do see a final price that's like 20% or 30% of market. That price simply isn't useful because you don't know what the last bidder's bid was or what the final price would have been if it was a no-reserve auction.

[joe_collector]

I appreciate your explanations...I'm not trying to put you on the spot or anything, and if I already have, I apologize...I'm just curious about the topic and I'm pretty sure GPAnalysis has chosen to toss out unmet reserves,

 

Personally, the only way to determine a "close to" true market value is to either include ALL reserve auctions, or delete ALL reserve auctions.

 

Those that are unmet may lower the price, but those that pass many times inflate the final bid.

 

I bid on a book once, using a higher bid than I intended, forgetting there was a reserve, and ended up winning the book for 3X the next-lowest bid (I hit the insane reserve exactly).

 

I see that stuff all the time, and without a reserve, the ending price would have been much, much, much lower. I've also see the same reserve auction listed 10-20 times, with one poor sucker finally paying the piper. GPA only includes the final FredKarger-like marathon sale, no matter how many times the market refused to pay the price.

 

And what about that SW Variant reserve auction, that was obviously shilled to $10K?

 

At least with CS's methdology, both sides get factored in, not just the "Retal Priced" reserve auctions that juice the averages.

[thecomicsheet]

...That price simply isn't useful because you don't know what the last bidder's bid was or what the final price would have been if it was a no-reserve auction.

 

But that is my point. You DO know what the final price would have been if it was a no-reserve auction. It's exactly the same price that is reported on the reserve auction. You don't know what the last bidder's bid was on a no-reserve auction either.

 

There is the psychological aspect of a reserve auction being 'off-putting' to some people. I have no way to objectively measure it. I have a subjective opinion that it has a small but very real effect of turning off some prospective bidders. I can report that in my own bidding it does have some small effect. Once in a while, I just give up on a reserve auction because I get tired of bidding against the seller. For me this probably translates to 5 percent fewer bids on reserve auctions.

 

Remember what I said earlier about reserve auctions being a way for seller's to 'shill' their own auctions. EBay extorts more money from seller's by letting them buy the right to shill bid their own auctions. Get caught shilling without paying eBay their rake and you get banned. But cough up the dough and it's ok with them.

 

[fantastic_four]

Why are "retail priced" auctions not representative of what people are willing to pay? That's how books are sold everywhere OUTSIDE an auction, you just can't haggle with an ebay reserve whereas you can with a seller listing a fixed price. I don't see a difference in terms of gauging what people are willing to pay between an auction with a $500 reserve and a no-reserve auction which starts at $500.

[Warlord]

What about auctions that have a reserve and get only one bid?

 

Let's say I try to sell CGC 2.0 Valiant Turok #1. I think it's worth a lot, so I put a $250 opening

bid on it, and a $500 reserve.

 

You come along, laugh at the opening bid because it's ridiculous, but you want to watch

the auction, so you safely enter the $250 bid knowing that because there's a reserve you

are not going to win the auction. (I am assuming that eBay doesn't let the opening

bid be greater than or equal to the reserve. Is that an incorrect assumption?)

 

According to your stated rules, you would enter a value of $250 "sales price" into your DB.

Did I get that right?

 

Wouldn't you expect that the same item, in a no reserve auction, isn't even likely to get

a bid of $5 or $10 or so??? But your DB has a bogus value in it at this point.

 

Obviously the scenario I've described sounds ridiculous, but aren't you likely to be pulling

such exaggerated values into your DB on a less ridiculous, although still misleading, final "sales price"?

[joe_collector]

I don't see a difference in terms of gauging what people are willing to pay between an auction with a $500 reserve and a no-reserve auction which starts at $500.

 

Sure, but you're talking about two very different animals.

 

Got through EBay, and look at comics that start at $500 or more min bid, and then check out the reserve auctions that ended at higher than $500.

 

The numbers are not in your corner.

 

And like I said before, is the "new price" of those Star Wars variants $10K? Would you want that obviously reserve-shilled data in the system?

[joe_collector]

Obviously the scenario I've described sounds ridiculous, but aren't you likely to be pulling such exaggerated values into your DB on a less ridiculous, although still misleading, final "sales price"?

 

Now that scenario you envision is pretty out there, but check out that "Shilled Star Wars Variant" Reserve Auction that ended at $10K?

 

That's real-world, it passed reserve, and will go into GPA's database.

[thecomicsheet]

What about auctions that have a reserve and get only one bid?

 

Let's say I try to sell CGC 2.0 Valiant Turok #1. I think it's worth a lot, so I put a $250 opening

bid on it, and a $500 reserve.

 

You come along, laugh at the opening bid because it's ridiculous, but you want to watch

the auction, so you safely enter the $250 bid knowing that because there's a reserve you

are not going to win the auction. (I am assuming that eBay doesn't let the opening

bid be greater than or equal to the reserve. Is that an incorrect assumption?)

 

According to your stated rules, you would enter a value of $250 "sales price" into your DB.

Did I get that right?

 

Wouldn't you expect that the same item, in a no reserve auction, isn't even likely to get

a bid of $5 or $10 or so??? But your DB has a bogus value in it at this point.

 

Obviously the scenario I've described sounds ridiculous, but aren't you likely to be pulling

such exaggerated values into your DB on a less ridiculous, although still misleading, final "sales price"?

 

My database reports the data as final bid, not sales price. I do include an indicator for 'reserve not met' auctions.

 

A bid of $250 is a binding contract. If you make the bid, it is a real bid (unless of course it is entirely shilled which will get you into trouble). It is also entirely possible that the seller set the reserve at the opening price and/or that the seller will end the auction early to sell to the highest bidder - be careful. Setting a watch is much safer than setting an inflated 'watch' bid. (I do confess that I occasionally put in a low 'watch' bid on some auctions myself. It's an easy way to keep track of some interesting auctions. But I never, never put in an inflated auction bid expecting to be safely under an absurd reserve price. I can't afford to p%$$ away $250 on a mistake.)

 

I do try to weed out obvious shilled bids. But they have to be really obvious for me to do so. $250 for a Turok #1 would stick out like a sore thumb.

 

I didn't have to try to figure out if the Star Wars variants auction was a shill bid situation because I don't include auctions for multiple books in my database. It's just too hard to try and impute the individual value for each book in the lot. (I don't know how GPAnalysis handles multiple auctions by the way, or if they take the same approach I do and just eliminate them.)

[JiveTurkeyMoFo]

I don't think anyone raised this point: have you tried contacting sellers on items that do not meet reserve to see if there was a sale after the fact? Often times, the high bidder and seller will come to terms post auction at an amount greater then the highest recorded bid, but lower then Reserve. Wouldn't that more accurately impact the pricing data?

 

I have to agree with James. Nowhere in this thread have you convinced me that using auctions where the reserves were not met is a true measure of market pricing.

[thecomicsheet]

I don't think anyone raised this point: have you tried contacting sellers on items that do not meet reserve to see if there was a sale after the fact? Often times, the high bidder and seller will come to terms post auction at an amount greater then the highest recorded bid, but lower then Reserve. Wouldn't that more accurately impact the pricing data?

 

I have to agree with James. Nowhere in this thread have you convinced me that using auctions where the reserves were not met is a true measure of market pricing.

 

I don't track off eBay sales data. My weekly column and my database are both for the eBay CGC market.

 

In my database there is a 'RNM' indicator on auctions where reserve is not met. It is a simple matter of using a filter to remove those auctions from the current view if you don't want to see them.

 

I'll probably have to sign off til this evening. We were snowed in this morning. Made me a couple of hours late for work. Have to work like a demon most of the rest of the day.

[JiveTurkeyMoFo]

I'll probably have to sign off til this evening. We were snowed in this morning. Made me a couple of hours late for work. Have to work like a demon most of the rest of the day.

 

See, I always thought having incliment weather was the perfect excuse to goof off.

Here in Florida, we have no such luck....the best we can do is: "Sorry Boss, but traffic was murder; glare from all that sunshine caught many drivers by surprise on the Interstate, and slowed everything to a crawl."

 

Of course, it don't work too well.

[fantastic_four]

I don't think anyone raised this point: have you tried contacting sellers on items that do not meet reserve to see if there was a sale after the fact? Often times, the high bidder and seller will come to terms post auction at an amount greater then the highest recorded bid, but lower then Reserve. Wouldn't that more accurately impact the pricing data?

 

If the software GPA and Darryl have written is doing the type of HTML parsing I think it is, then it's partially or fully automated. Contacting sellers like you suggest would either be more time-consuming or more complex than it's probably worth.

[gpanalysis]

Obviously the scenario I've described sounds ridiculous, but aren't you likely to be pulling such exaggerated values into your DB on a less ridiculous, although still misleading, final "sales price"?

 

Now that scenario you envision is pretty out there, but check out that "Shilled Star Wars Variant" Reserve Auction that ended at $10K?

 

That's real-world, it passed reserve, and will go into GPA's database.

 

Just to set the record straight: Our subscribers know that this auction result was NOT included in our data - we do not capture single auctions with multiple comic book sales (unless it doesn't make any difference in the price, as in including a crappy reading copy, etc).

 

Also, you would not be aware (although our subscribers are) that our system has the ability to show way out prices and sometimes these are flagged when there has been shill bidding, outrages BINs, etc.

 

Hi Darryl - not hijacking your interesting thread on RNM auction results, just wanted to set the record straight and clarify some things Joe Collector has mentioned here about our product - cheers.