Dear Legal Files: I read your article on shill bidding (“Legal Files,” November 2020) this week. It was funny because I had just bid at auction for a piece of art that I didn’t win. The bidding went to my bid limit, which was me, and then $20k higher on the next bid. I was on the phone and they tried to get me to outbid by another $20k. I said no, then they said $10k. I said no, I had hit my limit. Then they came back and said they’re not sure what happened, but would I match that bid. I said no. The painting didn’t sell. The auction house wanted to know later if I’d pay my high bid price. I said no.
The reason is that there is no true price discovery. If you know there are three real bidders and someone real bid below your highest bid, you can feel comfortable that at least one if not more people felt the object was worth something close to what you paid. Then you know that your evaluation of value was reasonable. But if the only bidder was really the auction house, the best real bid might be 25% below your price or lower, and that makes me uncomfortable having confidence buying. It really isn’t right.
This ties in to auction price ranges published in catalogs. People who are only marginally knowledgeable use those as the basis for bidding. This is somewhat scary. For cars, I feel competent enough to bid what I think is right and not worry. Art is very different. Not only am I not knowledgeable, but the variance in one artist’s prices for similar-size paintings can be huge. Plus, the buyer premium is 25% under $600k! So a big spread if you don’t end up wanting to keep what you bought long term. — Irv Kessler, a longtime SCMer
In the news
In the interest of full disclosure, Kessler happens to be one of my best friends. I know that he is a smart guy and absolutely the most disciplined bidder I know. He sets a maximum bid for himself and sticks to it almost invariably — quite the lesson for our own Publisher Martin.
Just after I read his letter, it was announced that the Nobel Prize for economics was awarded to Stanford professors Paul Milgrom and Robert Wilson for their decades of pioneering work on auction science. Milgrom and Wilson made their first mark working with the FCC in the ‘90s on its auctions of cell-phone licenses across the country. They have continued their scholarly and real-world work through the present day, with numerous books and articles to their credit. They have not only researched and analyzed the theory of auctions, but their work has actually changed some of the ways that auctions are conducted today.
Reading some of the literature raised the question whether the academic science of auction theory might offer enhancements to our personal anecdotal and practical theories about how to buy or sell at a collector-car auction.
Making it personal
The behavior of bidders is determined by their values, which fall into two categories: personal values and common values.
Personal values relate to the bidder’s personal connection to the auction lot. Let’s say the car being auctioned is a nicely restored 1957 Porsche Speedster. You’ve always wanted one, mainly because your father taught you how to drive a manual-transmission car in his. You had always wished that your mother had not made him sell it because you and your sister couldn’t both ride in it with him at the same time. This Speedster on offer is the same color combination as your father’s, and after doing your homework, amazingly, you discover that this actually is your father’s Speedster. It’s safe to say this car is worth more to you than any other bidder. This is personal value.
Common values are the norm. Bidders see the car for what is and how it fits into the overall market at the time of the auction. When all bidders share the same values, the bidding should fit into a fairly tight range and the result should be fairly predictable.
All else being equal, the bidder with the high personal value should win the lot at a slight or modest premium over its true market value. But watch out — if there are multiple bidders who have high personal values, the hammer price can end up being rather fantastic.
The winner’s curse
High personal value is not the only auction wild card. Even when all bidders share the same common values, the result is greatly affected by the amount of information they have about the car.
Information about the car can be public and private. Public information is what is written in the auction listing, information disclosed by the seller while the car is displayed, etc. All bidders have access to all the public information, at least if they look for it.
Private information is created when one bidder learns more about the car through individual efforts, such as a bidder arranging a pre-auction expert inspection and learning important information, good or bad, that is not known by the other bidders.
If bidders worry that they lack information, it can make them hold back to avoid the winner’s curse: Learning afterward that you paid too much because you didn’t know what everyone else did. This fear causes bidders to bid less than they are willing to pay, as in the case of Mr. Kessler above.
Smart auction companies and sellers take this into account and provide as much information as possible to dispel these worries.
The FCC licenses
When the FCC was directed to sell broadband licenses to create a national cell-phone network, it worried that its accustomed practice of sealed-bid auctions wasn’t the best way to go. So they brought Milgrom and Wilson in as consultants.
The main problem was that the licenses all covered limited geographic areas while the larger bidders were trying to construct regional and national networks. That meant they needed to buy many contiguous licenses. If these were sold one at a time, bidders ran the risk of getting outbid on one or a few, leaving them with a disconnected patchwork.
Also, there were alternative ways to build the network. For example, they could use licenses that went around the north side of a mountain, or licenses that went around the south side. If bidders knew they would get outbid on the north, they could still bid on the south. But they didn’t know. This meant that the FCC ran the risk that the auction would result in a large number of small local networks rather than the national network the country needed.
Milgrom and Wilson solved this by creating a new auction format for the FCC: the simultaneous ascending auction. Each license was auctioned individually in an ascending auction, where bidders outbid each other until a high bid is achieved. The trick was that all licenses were auctioned at the same time, and none of the auction lots closed until there were no further bids on any auction lot.
It worked splendidly. Each license sold for its highest attainable price. Bidders with high personal values were able to outbid the other bidders. When the bid got too high on one license, the bidder looking to create a network could switch to an alternative location. When the network lacked only one critical license, that bidder then had high personal value and would outbid all others.
What about with cars?
While the simultaneous ascending auction worked well for FCC licenses, could it be adapted to collector car auctions?
Part of it — the extended closing — is already here, thanks to Bring a Trailer. Just as on eBay, its auctions produce a flurry of activity at the end. But eBay closes the auction at the stated time, allowing late bidders to “snipe” the auction and steal the lot away, even though other bidders were willing to bid more. BaT extends the auction closing until the bidding ends. That produces higher sales prices for the sellers and higher commissions for the auction company.
Could a simultaneous multiple-lot format work for cars? Let’s say you want to buy a Mercedes-Benz 300SL and you have $1.1 million to spend. You register to bid at an auction that has two of them available. You do your homework and decide that both are fine cars, both are worth about the same, and either would be fine for you, but you prefer the one that will be auctioned second.
The first 300SL sells for $1 million. You confidently let it go and bid on the second 300SL. Unfortunately, it has some personal value for someone, and it sells for $1.2 million.
But what if the auction company had sold both cars in a simultaneous auction? The second car would have still sold for $1.2 million, but you would have bid to $1.1 million for the first car. You might have still been outbid, but that would have meant at least an extra $100,000 for the seller of the first 300SL, along with a correspondingly higher commission for the auction company.
Bidding strategy
Bidders choose different styles. Some bid early and aggressively to deter competitors. Some wait to swoop in at the end. The research shows that the natural tendency is to hang back and let things settle out before bidding seriously. But it probably doesn’t matter, as Milgrom writes, “all naïve bidding paths lead to the same competitive equilibrium outcome.”
But what you really do need to watch for are bidders with high personal value or a lack of public information about the car. Both can result in artificially high bids, which you don’t want to outbid. As my friend Kessler counsels, set a limit beforehand and stick to it. ♦
John Draneas is an attorney in Oregon and has been SCM’s “Legal Files” columnist since 2003. He can be reached through www.draneaslaw.com. His comments are general in nature and are not intended to substitute for consultation with an attorney.