You are here

Why aren't homes sold in second price auctions?

Newspapers reported a house sold in my town this week for almost $800,000 over asking, which is to say the buyers bid $2.5M for a house listed at $1.69M. Now the prices are already crazy but this takes the cake. The buyers had lost a few auctions before by not overbidding enough, and wanted to make sure they got this house, even though it is not that remarkable a place.

It was an auction, of course, but a simple sealed bid, highest price auction. Such auctions are "strategic" -- you want to bid the lowest amount that will still win. In this case the sellers did very well because the buyer probably bid more than they wanted to, unsure of what was needed to win.

Game theorists generally believe that second price auctions (the Vickrey Auction) are actually the best choice for both buyers and sellers. That when done properly, they bring in the best revenue for the seller but also make the buyer feel comfortable they bid the right number. Second price auctions are the most common in the world, since that's what eBay uses -- namely the winner bidder pays the amount of the second highest bid (in eBay's case, plus a small increment.) Google also uses them to price adwords.

To understand why they can be best for sellers, effectively the choice comes down to taking the #2 bid from a set of higher bids, or the #1 bid from a set of lower bids. The former option tends to be a good choice, and also minimizes buyer remorse at the same time.

So my question is, given that they are the best method, why don't these homes get sold in second price auctions?

The most likely reason is that the public, in spite of eBay, does not understand 2nd price auctions very well. In face, even on eBay, you see people who bid in advance or get annoyed at "snipers" who bid at the last minute. eBay combines what is actually a 2nd price auction with the appearance of a "going, going, gone" auction, and that confuses people.

For a 2nd price auction to work, the bidders must truly understand how to use it, which is to bid their "true heart." That is to say, to bid the maximum price they would really want to pay. The price at which, they are satisfied if they get it for that price, and they are satisfied if they don't get it at that price because they truly believe any more was too much. You must pick the price where you are (barely) happy if you win, and barely happy if you lose. It takes some practice to learn how to pick that price.

In practice of course, 2nd price auction bidders know they will almost always pay less than they bid, and thus be truly happy. That is what allows them to submit their truly highest bid. The fact that they submit their truly highest bid is what gives the seller the best outcome. You sacrifice the random swings of buyers who overbid for a reliable shot at the best market price.

Unfortunately, home buyers and sellers are usually amateurs. They do this once every decade, if that. They otherwise never spend or get this much money. They are likely to make mistakes or be irrational in their bidding.

To make it work, real estate agents would need to be trained in how to coach their clients. It's not that hard. You ask for their number, and when they say, "1 million" you say, "So if it sells for $1,050,000 you will be pleased that you didn't have to pay that much and will buy something else?" You need an increment that is large enough to make a difference. You can't just say, "You're happy if you lose with $1,000,001" because nobody would agree with that.

One reason 2nd price auctions confuse and anger unskilled bidders is that if you were the 2nd price, the winner gets the house for your price. You imagine, "If only I had bid a little more." Worse, on eBay, the winner gets the item for your bid plus $1, making many people believe, "Crap, if I had only bid a dollar more, of course I would have bid that."

In reality the winning bidder bid much more, so if you had bid $1 more, their price would also go up $1. You didn't lose by a dollar, you probably lost by a lot, though that is generally not revealed to anybody but the auctioneer.

As noted, a 2nd price auction has (almost) no strategy. You do not alter your bid based on your opinion on what others will bid. Even if the bidders collude or publish their bids, it doesn't change the result -- the person willing to pay the most gets it, and pays what the person willing to pay the 2nd most offered. In a 1st price auction, there can be strategy and collusion. You do well when you guess correctly what others will bid, and if you can find a motive, collusion can do very well for the bidders. (Usually that requires the bidders participate in multiple auctions and collude to cheat a group of sellers.) The one strategy in 2nd price auctions is that you should not publish your bid because it might make somebody realize they're actually willing to bid more, namely more than you. It's better for you if people underbid their true heart. When auctions repeat it can help people learn more about the market and, while it shouldn't, this adjusts their estimate of their true heart.

Differences among buyers

One problem in real estate auctions is that not all buyers are the same. Some have conditions on their offers. Some the seller just likes better than others. It is not unknown for sellers to take a 2nd highest bid if they liked the bidder more, or it came without conditions.

To deal with this, the auctioneer could allow the seller to pick the winner in a couple of ways:

  1. The seller might specify, "put a penalty of $10,000 on buyer A." This would be subtracted from their bid in deciding the winner and the winning price. So if A bids $1,050K and B bids $1M and C bids $900K, B wins and pays $950,000 -- it is treated as though A bid $950K.
  2. The seller could set a reserve price which is treated like a bid. Bids above the reserve price would be revealed to the seller, and the seller could pick any one of the bidders they like, not necessarily the highest. But the seller must pick a winning bidder, unless nobody qualifies, and that party pays the next highest bid from a qualifying buyer, or the reserve price if that is higher. The forced sale is needed because the seller has now learned the secret bids of the high bidders, which is unfair in another auction.


This is essentially an escalation clause (or am I missing something). It has been nearly 15 years since I bought a house (going to one of your points), and at the time, the market was still a bit hot. Buyers would commonly add a clause to their bid, something like, we will pay X percent over your top bid, up to value Y. I thought this was still a common thing today. I recall doing this on a few highly desirable properties. We did not do this on the home we ultimately bought. We put forward our best, fair offer, but we also knew we were the only bidders.

To answer your question, it is done this way because it benefits the seller. Why take 1.7M when some crazy person might offer 2.4M?

Other articles have pointed out that in this particular case, the final purchase price was in line with other properties in the area, the asking price was set artificially low to generate interest in the property.

2nd price auctions will never be accepted voluntarily because both agents and sellers want the maximum price possible, and are willing to accept the downsides of highest-bidder-wins method.

There is a perception that auctions are for bargain or distressed sales, which is why there is a disconnect between people's value of an item and the price they are willing to pay. As long as that auction=bargain association exists, auctions won't be successful in finding "true value" for any transaction.

In reply to both comments above -- the reason to go with 2nd price is that auction theory states it is the best for the seller and the buyer. While it is true that sometimes people bid something "crazy" and thus you capture it, the use of 2nd price auctions encourages bidders to bid higher than they would in fixed price auctions. Yes, you get the #2 bid, but because everybody bids higher -- their true maximum bid -- it's a win for the seller.

This does require that you have a decent number of bidders (which you do in a hot market.) In a working market, #1 and #2 should not really differ by that much. If they are that different, we have bidders with very different motivations, which of course sometimes we do have. Indeed, those different motivations may answer my question.

In a standard highest bid auction, most bidders should not be bidding their true maximum. You instead just want to make sure you bid what you can get away with. You would never bid over your true maximum. 2nd price auctions encourage you to bid your true maximum by rewarding you by an automatic algorithmic calculation based on the true maximum of your next competitor.

So as a seller you are choosing between two strategic bids of say $1M and $1.2M -- and getting $1.2M or two true bids of $1.3M and $1.5M and getting $1.3M. Of course, anything that gets more money for the seller is against the interests of the buyers, but here the buyer interest that is satisfied is that they paid less than their maximum, and they also have no buyer's remorse because they don't worry that they could have gotten it for less. They could not have.

...the result of an emotional bidding war. In this case, each side is driven by a (possibly unconscious) fear of the regret they will have if they loose. The sting from a lost transaction is a powerful punishment, and the desire to avoid reliving this was the motivator in your example.

In my experience, interactive auction rounds lead to the highest bids. The more iterations in the auction, the higher the anchor price goes making it easier for the bidders to go beyond what would have been their limit price in a single pass auction ("If they are willing to pay X, then we are willing to pay X+Y"). The repeated iterations also increase the bidders perceived sunk costs ("We've come this far, don't want to loose it now"). Good real-estate agents feed these emotional responses.

Interestingly, this does not happen in commercial real estate sales so maybe this would be a better place to start promoting 2nd price auctions.

I see this often on eBay as well where people base their internal highest price on what others are willing to pay. It even happens in the stock market - why else would people care so much about non-actionable last trade information when the only information that can matter in the future is the bid and ask?

I think the more interesting question is how to best deal with a single seller, single buyer situation. The trivial case is where the buyer's highest price is lower than then seller's lowest price, where no transaction is possible. But what about when the prices overlap? How do you efficiently find the correct transaction price in between the two? Typically people use bargaining, which is inefficient and path dependent and can even lead to outcomes where no transaction happens even though there was a range of prices where both parties would have been better off if it had. Any ideas on how to solve this problem?

I do not see many reports in real estate auctions of actually iterative bidding wars. Most of what I have seen is sealed bids. It should not be possible for a bidding war to drive you up above your true value. Instead, it can discover that your true value was higher than you first estimated. And if there is no other way to discover that, you could argue that it changed the true value.

When you speak of single seller, do you mean something different from an ordinary sale where there is always one seller, or do you mean a very slow market where only one house is on the market? In a general sale with only one bidder, which is the case in slow markets, real estate agents figure out a sense, from the asking price, of prices the seller will likely accept, and guide their clients to offer that, and then negotiation goes back and forth between there and the asking price. Something works out.

(I bought this house in a buyer's market but for some reason, after 30 days without an offer, mine and another came in the same day, so I did have to do a small bidding war, which I won by accepting their counter -- asking price and minimal contingencies -- unchanged.)

>I do not see many reports in real estate auctions of actually iterative bidding wars. Most of what I have seen is sealed bids.

Iterative bidding wars are common here in NYC, and I can not remember the last time I saw a true sealed bid auction for residential real estate here. I can not imagine a case where the seller would announce a sealed bid auction, receive all the bids, and then *not* then turn around to the 2nd place bidder and ask if they wanted to outbid the highest bidder. As you you noted, people typically only do one residential real estate transaction per decade, so the potential reputational costs of making that phone call to the 2nd runner up are almost zero (what are the chances that person will be a bidder on the next home you sell?) and the potential upside is large (they do out bid the highest bidder, at least getting you a better price... and potentially starting bidding war with all the price anchoring and regret-mitigation noted above).

>Instead, it can discover that your true value was higher than you first estimated.

Maybe you could argue that I am assuming that the other people bidding had more information than me, and their higher bid reflects this knowledge. I then can rationally increase my bid to also reflect this new information that I assume they have.

Of course then they will see my increased bid and make the same assumption... and an irrational bidding war has begun! Prices artificially inflate based solely on self-referential false information.

If you have any doubts that this is how people actually operate, just look at any active ebay auction. The bidding pattern is almost always the same... two or more bidders take turns ratcheting up each other's maximum bid by the minimum increment with shorter and shorter turn times up until the moment the auction closes. This is especially irrational since the ebay system has surprisingly large increments and favors the first entered bid in the case of a price tie. The only model I can imagine that explains this pattern is that both bidders are each adjusting their internal limit price based on the false information of the previous bid- driven by a regret-mitigation strategy rather than utility.

>I bought this house in a buyer’s market but for some reason, after 30 days without an offer, mine and another came in the same day

This happens so often that I doubt it is coincidence. I have seen this exact scenario myself twice in the past few months - once as a buyer and once as a seller. Savvy real estate agents precipitate this by delaying a first comer and accelerating subsequent ones. They also use well-paced sales pushes (as opposed to continuous marketing) to bunch interest.

When I was the initial bidder, the agent delayed steps of the transaction until a 2nd potential buyer appeared and then cleverly rushed the process with them so our two offers would be concurrent to the day. I intentionally withdrew from that negotiation when I discovered the manipulation (thanks to chance ex-parte communications with the other buyers)... and now I have crushing regret! :)

Do you know who your bidding counterparts were? Did you sense that you were being delayed up until the moment that they showed up, and then you were rushed?

Create a service to connect potential residential Real-estate buyers and allow them create a synthetic 2nd price pre-auction.

Buyers would register their interest in a property along with their best prices. The result would be a winner and a price (which would be the winning 2nd price).

The winner would then be able to offer the 2nd price to the seller and avoid competing with the other buyer in a 1st price auction.

I don't think there is any benefit to buyers to misrepresenting their best price since the loosing buyer can always compete in the real auction if the winning buyer submits a price lower than the winning 2nd price. The winning buyer is always better off since he gets the property for the 2nd price, and losing buyer are never worse off.

Maybe sellers would attempt to subvert the system by entering false bids to artificially inflate the results of the pre-auction. You could discourage this by having the true identity of all parties disclosed at the end of the auction, but sellers could still use proxy bidders to hide their identity. Maybe the system automatically follows up with a public record search to ensure that the winning buyer actually closed - and if not investigates. In the case where the winning buyer turned out to be shill, the system could either file a civil suit or try to get fraud charges filed against them. Just the (well publicized) threat of these actions could deter false bidding.

Sadly, the DOJ would likely try to come after you for anti-competitive practices. It would be misguided though since, as you point out in the article, encouraging 2nd price auctions would ultimately lead to more efficient outcomes.

BTW, the above inflationary false bidding actually happens on ebay. I was once outbid, only to get a 2nd chance offer right after the close of the auction. I looked deeper and saw that the account that won had previously bid on other auctions form the same seller in the same pattern. I wonder how many 2nd chance offers are really unintentionally wining shills? Yet another reason to bid your true limit price once early in the auction and never look again until after the close!

Business model:

Charge each bidder a fee that is a percentage of their bid to enter it. Rebate the loser the full fee paid. Rebate that winner's fee such that the net of the pre-aution 2nd price and the fee is lower than the winners' best price. In no case is anyone worse off for participating in the pre-auction and the winner is always better off (assuming he is a true bidder).

Besides being a revenue source, the fee discourages sellers from entering high false prices with the intention of discovering the true buyers' best price.

See any flaws?

No, the winner is worse off if any fee is paid above the 2nd price. The whole point is that people bid higher in 2nd price auctions (their true price) than they do in first price auctions (their strategic estimate of just enough to be likely to win, but no more.)

Actually, on eBay you should bid you true price in the last minute. The truth is that many ebay bidders do not know how to calculate their true price, and so if they see that they are outbid, they change their minds and bid higher. It is not in your interest to give them the chance to do that.

The same could happen in the pre-auction described. Sadly, the seller has no interest in punishing those who entered false bids.

Escalation clauses do exactly this, and they're in common use in markets where sellers are getting a lot of offers. When I bought in Washington DC in 2015, I had 6 failed bids before my seventh offer was accepted (we used an escalation clause in several bids, including the successful one).

But in a typical market, the dynamic is different. Buyers expect the house to go for a bit less than the asking price, and the market is often thin enough that offers come in days apart. Without simultaneous bids, the second-price auction mechanism doesn't work.

It's only in a hot market that you can have an auction like this.

Add new comment

Subscribe to Comments for "Why aren't homes sold in second price auctions?"