Leu Auction - Could someone explain the bidding action ?

[Tejas]

I think the allegation of auction fraud is based on a misconception or misinterpretation of the bidding history. I think some people think that the winning bid is the amount that the high bidder has put in. Instead, the winning bid is the amount that is one increment above the runner up. 

I find it highly improbable that an auction house like Leu would risk their reputation (their existence really) with these kind of illegal manipulations. It doesn’t have much of an effect on their income and involves existential risks. Only the owner of the coins (the consignor) would have a real incentive to engage in this kind of activity. 

[Tejas]

One more observation … I usually attend the Leu room auctions and I also have been to some of the NAC auctions, where you have the chance to speak to other bidders and especially their agents during the breaks. I think many people here would be surprised how many collectors there are who absolutely don’t care if a coin cost them CHF 500,- or CHF 5000,-. I remember sitting next to private bidders (not agents) who spend 200’000 in one auction. 

[Heliodromus]

3 hours ago, Tejas said:

I think the allegation of auction fraud is based on a misconception or misinterpretation of the bidding history. I think some people think that the winning bid is the amount that the high bidder has put in. Instead, the winning bid is the amount that is one increment above the runner up. 

Yes - no way of knowing the winner's high bid unless it was yourself (as has been the case with some of those folk on CT).

As a non-winner the only observable irregularity is when the bid order of such a non-winning snipe in these situations is more often W, U rather than U, W. (W = winner, U = underbidder).

As far as auction house reputation, you would think any appearance of impropriety would be something to be avoided, but note also their non-response to the community when they were recently selling coins reported as stolen (as claimed by Michael Gasvoda), or look at Roma (and the apparent lack of impact to their business as a result), or look at what happened to Lanz ... The temptation to engage in shady dealings in the ancient coin business, starting with "unprovenanced" coins must be high, and I guess it's a slippery ethical slope ...

 

Edited December 16, 2023 by Heliodromus

[Heliodromus]

3 hours ago, Tejas said:

One more observation … I usually attend the Leu room auctions and I also have been to some of the NAC auctions, where you have the chance to speak to other bidders and especially their agents during the breaks. I think many people here would be surprised how many collectors there are who absolutely don’t care if a coin cost them CHF 500,- or CHF 5000,-. I remember sitting next to private bidders (not agents) who spend 200’000 in one auction. 

Sure, but I'm not sure what the relevance is to this discussion? This is not about the high bidder, or about people making nuclear bids. I think you probably understand that, but prefer to muddy the waters for some reason. I'd imagine you also realize that someone like Andrew McCabe understands what a bid maximum is, even though it seems you want to suggest otherwise.

 

[Tejas]

1 hour ago, Heliodromus said:

Sure, but I'm not sure what the relevance is to this discussion?

Well, its simple actually, when analyzing auction behaviour we ususally assume that economic agents are "rational" in the sense that they are price sensitive and seek to minimize the price they have to pay. While this is the case for most bidders, there is a seizable group of collectors who don't care much about the price. They mostly care about the outdome. Hence, they bid any amount necessary to get the coin they want. If you have two or more of these bidders competing for the same coin one should not be surprised to see seemingly irrational bidding behaviour, i.e. big price jumps and unexpectedly high prices.

Edited December 16, 2023 by Tejas

[Heliodromus]

14 minutes ago, Tejas said:

one should not be surprised to see seemingly irrational bidding behaviour, i.e. big price jumps and unexpectedly high prices.

Sure, but that is not what we're discussing. We're discussing a very specific bidding pattern, underlined in my above post, that seems to entail the underbidder being aware of the high bidders maximum (to which only the high bidder and bidding platform/company are privvy).

We're beating a dead horse here.

[Tejas]

3 hours ago, Heliodromus said:

Yes - no way of knowing the winner's high bid unless it was yourself (as has been the case with some of those folk on CT).

As a non-winner the only observable irregularity is when the bid order of such a non-winning snipe in these situations is more often W, U rather than U, W. (W = winner, U = underbidder).

I'm not saying that these kind of auction manipulations are out of the question. I think they are, in the case of Leu both unlikely and unproven. 

I added something like 20 to 50 bids in the Leu auction and given that valuations of different bidders are usually not completely dissimilar it is not surprising that I am sometimes bid up to the maximum limit. This is unfortunate when it happens and there is a psychological bias for people to note these random occurrances (random losses) more than normal bids and since many people are adverse to believing in random misfortune they are quick to look for secret machinations that have brought that on them. 

Your second statement above I don't understand. If there are two or more people entering bids in the last seconds any bid order may result. During the last Leu auction had many occurances where I bid up the winning bid without actually winning the lot, even when I entered what seemed to be way too high bids.

Edited December 16, 2023 by Tejas

[Heliodromus]

3 minutes ago, Tejas said:

If there are two or more people entering bids in the last seconds any bid order may result.

Exactly.

So we should not expect to see any pattern of winning related to final seconds bid order.

[Tejas]

5 minutes ago, Heliodromus said:

We're discussing a very specific bidding pattern, underlined in my above post, that seems to entail the underbidder being aware of the high bidders maximum (to which only the high bidder and bidding platform/company are privvy)

I'm aware of that. I'm just pointing out that this "very specific bidding pattern" may in fact not be so specific, but a random occurance that results from specific characteristics of timed auctions and the auction psychology of collectors.

But you are right, we are beating a dead horse. I have alerted somebody I know at Leu to this discussion. I'm curious what he will answer. 

[Tejas]

9 minutes ago, Heliodromus said:

Exactly.

So we should not expect to see any pattern of winning related to final seconds bid order.

Agreed, if the bid order W, U were to occur significantly more often than U,W this could be an indication of manipulation. I guess somebody has to analyse the data in a statistically proper way.

[Heliodromus]

3 minutes ago, Tejas said:

I'm curious what he will answer. 

It's rather predictable what the answer will be, if any. However in the previous CoinTalk discussion they never cared to comment.

 

[Tejas]

Audiatur et altera pars 🙂

[El Cazador]

5 hours ago, Heliodromus said:

Yes - no way of knowing the winner's high bid unless it was yourself (as has been the case with some of those folk on CT).

As a non-winner the only observable irregularity is when the bid order of such a non-winning snipe in these situations is more often W, U rather than U, W. (W = winner, U = underbidder).

As far as auction house reputation, you would think any appearance of impropriety would be something to be avoided, but note also their non-response to the community when they were recently selling coins reported as stolen (as claimed by Michael Gasvoda), or look at Roma (and the apparent lack of impact to their business as a result), or look at what happened to Lanz ... The temptation to engage in shady dealings in the ancient coin business, starting with "unprovenanced" coins must be high, and I guess it's a slippery ethical slope ...

 

Feels like a speculation here: “he said, she said”…

[DonnaML]

12 hours ago, Heliodromus said:

look at Roma (and the apparent lack of impact to their business as a result)

Well, I did notice that for the first time in some years Roma (and specifically Richard Beale) are absent from the roster of dealers at the upcoming NYINC in January. Considering that this year's NYINC is where he was arrested -- and that as far as I know he hasn't yet been sentenced! -- I can't say I'm surprised. But I imagine that like most major dealers, Roma would ordinarily make a good deal of money at NYINC. So what happened certainly must have had some business impact. Not to mention that Roma's plans to open an office in the US appear to have gone by the wayside.

Edited December 17, 2023 by DonnaML

[Heliodromus]

21 hours ago, Tejas said:

Agreed, if the bid order W, U were to occur significantly more often than U,W this could be an indication of manipulation. I guess somebody has to analyse the data in a statistically proper way.

Well, all auctions are archived there on the Leu web site, including per-lot bid histories and lot ending times.

For people who have bought lots in these auctions, your bids, including your maximum bid as well as hammer price are there too under "My Leu".

If you care to check, then I suggest:

1) Focus on coins where you think you paid a high price (or check all if not too many).

2) Ignore coins where there was no major price gap between top two bids (winner + underbidder) and the rest of the field

3) For remaining lots, how many times was the underbidder a last seconds snipe that came in AFTER your winning bid? How many times did this underbidder read your mind and manage to come in 1-2 bid increments under your (should be secret) maximimum bid?

Draw your own conclusions.

Incidentally, auction house shill bidding was discussed in the most recent Aaron Berk Ancient Coin Podcast # 37 ("perl of wisdom" at end of podcast). Sadly he didn't name names.

 

[Tejas]

I did go through my  38 bids and 16 wins in the last Leu auction. This included only 4 more expensive coins. Nothing suspicious there. The only suspicious thing was how amazingly cheap the other 12 coins were. 

[Heliodromus]

5 hours ago, Tejas said:

I did go through my  38 bids and 16 wins in the last Leu auction. This included only 4 more expensive coins. Nothing suspicious there. The only suspicious thing was how amazingly cheap the other 12 coins were. 

For me was exact opposite - of four coins with this winner + underbidder well above the field pattern, all four(!) were cases where underbidder followed me, pushed to maximum, and came in VERY close to end of auction (1 sec, 2 sec, 3 sec, 4 sec).

I will say that I've won other coins from Leu where there was nothing unseemly happening - normal organic bidding, or me bidding high on a low priced coin without any suspicious underbidder (CHF 100 to win at CHF 35). Still, 100% suspicious cases on these other higher priced 4 is rather damning.

 

Edited December 17, 2023 by Heliodromus

[Hrefn]

I realize this is a tiny sample but I just won this from Leu, for a hammer of 35 euro.  My max bid was 75 euro.  

[Tejas]

My contact at Leu (the head of numismatics) replied this morning. Obviously, they reject the allegations of "chill bidding" as being completely false. He said that at the last auction, more than 2000 bidders participated with large numbers of bids being entered in the last seconds. Like me, he also pointed to the psychological bias of being "mysteriously" bid up. As @Heliodromus said the answer was in line with expectations. The point is, however, I know him personally and I have no doubt that what he says is correct.

Edited December 18, 2023 by Tejas

[Heliodromus]

6 hours ago, Tejas said:

with large numbers of bids being entered in the last seconds

Of course, that's how fixed end-time auctions work. Most serious bidders, self included, snipe.  No surprise there. Just chaff.

6 hours ago, Tejas said:

Like me, he also pointed to the psychological bias of being "mysteriously" bid up

It's not mysterious if someone outbids you, or underbids you pushing your bid up a bit towards your maximum, etc. There are plenty of non-suspicious ways one can can get bid up. This is an auction - we get it. Again more chaff that completely avoids addressing the core issue, in a seemingly deliberate way.

These suspicions (or rather OBSERVATIONS, on behalf of experienced bidders) have been around for years against Leu specifically, not against other auction houses where - surprise, surprise - there is also sniping taking place and people getting outbid more often than winning.

If Leu were innocent and wanted to regain trust, they could easily extend the auction as long as bids are coming in and therefore avoid sniping and "hidden" maximums. Having seen such an evasive non-reply that you conveyed, I'm sure they'll do no such thing, nor address it in any other way.

Edited December 18, 2023 by Heliodromus

[Hrefn]

As numismatists, we find coins to be fascinating.  Although the degree to which other people share our interest is a question which seldom rises to the level of conscious consideration, I suspect it is a natural human tendency to assume other people are similar to us.  The manifest differences exhibited by others are curious and sometimes seem quite odd, especially when we are using ourselves as the yardstick by which we measure others.  For example, I am well aware that watching sports matches on television is a popular pastime.  Some people are sports fanatics.  My own attitude towards televised sports is utter indifference, and if I were forced to watch I would rather keenly resent the waste of my time.  

It is obvious once one thinks about it that I as the numismatist is the outlier, and the sports fan is the norm.  On any given day, tens of millions of people are watching televised sports.  Poor Leu, which casts a worldwide net, can only attract 2000 bidders to its auction.  For any lot in the auction, only a very small number of the those bidders will bid.  We know this, not least because some lots attract no bids at all.  Why is this important to this discussion?

I think it is important because the small number of bidders and bids may explain the quirky nature of the data set comprised of the auction bids and their progression.  These auctions are so small that there may be only one person with serious interest in a lot.  Once the bid rises to a certain level, the serious bidder will win, and probably at a reasonable price.  But once there are two serious bidders, who often wait till the last few seconds to enter their bids, the bidding will move substantially and erratically.  A coin might go from 100 euro to 1000 euro in one jump.  

This price action is a consequence of the very few numbers of serious bidders.  No nefarious activity on the part of the auction house is required to explain it.  As collectors, we should be glad we do not have more competition than we do.   

Addendum to address @Heliodromus’s comment above.  If an auction house were selectively and intermittently shill bidding, I am not sure how easy that would be to ascertain.  

 

 

Edited December 18, 2023 by Hrefn

[Heliodromus]

32 minutes ago, Hrefn said:

But once there are two serious bidders, who often wait till the last few seconds to enter their bids, the bidding will move substantially and erratically.  A coin might go from 100 euro to 1000 euro in one jump.  

Yes, this is understood. But in your scenario where the pack is at 100 EUR and only two serious bidders are left, one would not expect any close degree of coordination between their bids. If one of these two bidders bids 1000 EUR, then it'd be surprising if the other just so happened to choose EUR 950 or EUR 1100 (i.e. bid within one increment of the other). We would expect a much wider variation such as maybe one bidding EUR 1000 and the other EUR 750, or EUR 1300.

With final seconds bids, one is bidding blind (which is the whole point of sniping - to bid so late that no-one else has time to react to your bid). So we have the current bid at EUR 100, and these two serious bidders - one of who will win the lot - having entered their bids and mouse hovering over the "bid" button at T-10, 9, 8, ... One bids at T-6, and maybe the other at T-4 (although this would be a bit aggressive - one risks network/server delays preventing bid from getting there on time if cutting it too close)...

So, let's say bidder 1 submits their snipe at T-6, and bidder 2 at T-4. Let's say that it just so happens that they ARE a bid increment apart (hey, it CAN happen), so one of them bids EUR 950 and the other bids EUR 1000. Now, without me telling you whether it was bidder 1 who bid EUR 1000, or bidder 2, can you tell me who is likely to win the lot ?

Sadly at Leu, it seems the winner in such a scenario can be predicted.

 

Edited December 18, 2023 by Heliodromus

[Hrefn]

@Heliodromus thank you.  I see your point. If the last bid is never the winning bid, but serves only to push the price up, that would indeed be suspicious.  

Edited December 18, 2023 by Hrefn
Typo

[Heliodromus]

2 hours ago, Hrefn said:

@Heliodromus thank you.  I see your point. It the last bid is never the winning bid, but serves only to push the price up, that would indeed be suspicious.  

Yes, last bid losing and also within 1 or 2 increments of the other, and often coming in VERY (un-humanly?) close to end of auction.

Of course any such shenanigans is overlayed on top of other normal bidding activity, and it seems that low priced lots are perhaps excluded. Maybe there are other conditions too, such as certain well-known customers being excluded from this... So, it's never going to be a clean pattern of the last bid NEVER being the winning one. but rather a statistical tilt in that direction (in these "last two serious bidder" scenarios).

Things become much clearer if you are the winning bidder, thus privvy to your bid maximum, and keep seeing the pattern where in this "last two serious bidders" scenario it's not only that the underbidder comes in AFTER you, but also comes withing a bid increment or two of you. When you combine the mathematical probabilities of these factors all occurring at same time, and then see this being common (in my case in all 4 of 4 such "last two serious bidder" scenarios!) then one can only draw conclusions.

Edited December 18, 2023 by Heliodromus

[Prieure de Sion]

Shill biidding at Leu Auction? I have not experienced any shill bidding on my coins. I wish ghosts had carried my coins to the top, but ... 😂 😂 😂 

 

Edited December 18, 2023 by Prieure de Sion