ECON 302 - Microeconomics II IRYNA

DUDNYK

Auctions.

What is an auction? When and whhy do we need auctions?

Auction is a mechanism of allocating a particular object at a certain price.

‘Allocating’ part concerns who will get the object and the price part speaks for itself. As we
will see, there are several ways to design this mechanism, namely there is more than one way
to conduct an auction.
Why one needs an auction to allocate an object and to decide on the price? Recall that in
this part of the class we are in the realm of asymmetric information. Some players have private
information relevant for the decisions of other players, and as rationality requires them, they
use that private information to maximize their payoffs. Before we proceed to the auctions,
let’s take a look at the nature of the private information in the context of trade among people.
First of all we need more than one person interested in the object: there is more than
one potential buyer. The private information takes the form of a person’s valuation of the
object, which is not known to anybody except that person. Valuation is buyer’s maximum
willingness to pay, but actually it is simply the utility that a buyer will enjoy if he or she
obtains the object. When analyzing auctions we will usually assume that the seller has no
value of the object, meaning that he will be satisfied with any price that he can get (which,
of course, any rational seller will maximize). When valuations are not observable, if a seller
meets a buyer, seller has no clue about what that buyer is willing to pay, the only thing he can
be sure of is that buyers will always lie about their willingness to pay hoping to pay low price
(this means that the bargaining models are no good). If the seller tries to set a reservation
price and make take-it-or-leave-it offers, again he will have a problem: unless the price that
he asks is very low, some buyers will reject the offer and there will be no trade. What should
a seller do? Get a bunch of buyers in a room and make them compete for the price, which is
an auction.
A good auction is
• Efficient - of all buyers who are in a room, the one with the highest valuation gets the
object.

• Revenue Maximizing - extracts highest possible price from the buyer who gets the
object.
Modeling an auction as a game

Since we are in a game theory class we will analyze auctions using game theory language
and game theory tools: in this section we represent an auction as a game.
Like all games we studied so far, we have players who selfishly maximize their payoffs:
using all the information available to them, they choose their best responses to other player’s
strategies.

1 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

Any game starts with rules. Recall lecture 1. Rules of the game include: players, strategies
available for each player, payoffs for all possible combinations of strategies of all players.
Asymmetric information adds one more element: each player can be of more than one type
and each player’s type is his private information.
Players The players are the people who came to the auction. We call them buyers, or bidders.
We can count them and denote the total number of bidders by N . we can also label each
person as 1, 2, ..., i, ...N . For example, if there are only only two bidders, it is written
as N = 2, but, of course, we can analyze auction with any number of buyers. Note:
the auctioneer is not a player, he is third party or a “black box” who collects bids and
determines the outcome of the auction.
Player’s Types ‘Type’ is the private information that a player has. In our context it is buyer’s
valuation of the object. Buyer 1’s valuation is denoted by V1 , Buyer2’s valuation is
denoted V2 etc. up to VN What is valuation? It is the utility the buyer derives from
the object if he gets it, measured in dollars. At the same time valuation is a buyer’s
maximum willingness to pay for that object: nobody will pay more than an object is
worth to him. Why is it private information? Because by looking at a person we cannot
tell how much he or she is willing to pay for the object, but they privately know this
information.
Nature and Valuations. Formally economists say: “before buyers come to an auction, nature
assigns a valuation for each buyer”. How to imagine buyers’ valuations? Imagine a
huge slot machine. The number of wheels is equal to the number of seats (buyers). All
possible valuations are written on each wheel. Before the auction starts, Nature plays
the slot machine: presses Start and then Stop buttons. The numbers that appear on
the slot machine are the valuations of the buyers who came to the auction. With some
probability all valuations will be low, then all buyers who came to the auction are not
willing to pay too much for the object. With some probability all numbers will be high:
all buyers are willing to pay a lot. With some probability all intermediate combinations
will occur: there will be buyers with low and high valuations at the same time. Why am
I telling you this? Imagine yourself the auctioneer, for you, when you choose an auction
format, all you know is the numbers of wheels on the slot machine (the number of buyers)
and their possible valuations. Auctioneer has no clue whether actual valuations of all
buyers who came to his auction are high, low, one high and others low, two high and a
bunch of low, bunch of high and bunch of low, all valuations are exactly average etc. He
expects that any combination of valuations is possible with some probability. Also buyers
have no clue what are the other buyer’s valuations, they can only see what is written on
the wheel that corresponds to their seat. This description might seem counterintuitive
for you. But think again. Suppose the object is a Chinese antique box. When I come
to auction I know my valuation, it depends on: do I have appreciation for Asian art
(yes, because my grandmother had a nice collection that fascinated me as a child), do I
want a box right now (yes, because I have lots of jewelry and I want a nice storage for
it), do I find the shape of this box aesthetically appealing (not, because I think it’s too
square). So in a way my valuation of this object is determined by nature: my subjective
preferences and random experiences that shaped my perception. If my experiences were
different, my valuation would be different.
Knowledge Each buyer knows for sure his own valuation and the number of buyers who participate

2 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

in the auction, but not the exact valuations of the other buyers. What else does he
know? Each buyer knows the distribution of valuations: range of all possible valuations
for the object and the probability with which each valuation happens. Why does he need
to know the distribution of the valuations? Because he has to form some expectations
about bids of other buyers to choose his own optimal bid. An example of a distribution
is: all possible valuations range from 1 to 100 and each valuation is equally likely; then
when each buyer looks at a person who sits next to him and only knows that: ‘my
neighbor’s valuation is 1 with probability .01, his valuation is 2 with probability 0.01,
etc, his valuation is 100 with probability .01 and the expected (on average) valuation
for this buyer is 50’ (recall that this is a uniform distribution: all values in the range are
equally likely, so we’d say that V ’s are uniformly distributed on the interval [1, 100]).

Strategies Each buyer i will place his bid bi hoping to get the object at price below Vi . so, in
this context the strategies are the bids. ‘Bid zero’ is one strategy, that we can write as
bi = 0 ‘bid 1 cent’, is another strategy, etc. Out of all possible bids the buyer will choose
the one that will maximize his expected payoff given the expectations about strategies
(bids) of the other buyers (remember in equilibrium all players play best responses to
each other strategies, same here). Keep in mind that although optimal bid depends on
buyer’s valuation, it does not have to be equal to the true willingness to pay. If a buyer
wants to bid more or less than his valuation, there is nothing that prevents him from
doing so.

Payoffs After the game is over the players receive their payoffs that are determined by the strate-
gies that all players chose to play. Same thing here: depending on all bids submitted
(strategies played by all players), the auctioneer will determine the winner and announce
some price, P . One buyer will win the auction and get the object. What is his utility?
His utility is his private valuation that he will get to enjoy with the object minus the
price he has to pay. If buyer indexed i wins the auction, then his utility, denoted Ui is

Ui = Vi − P

It is a very simple linear thing and it will be exactly the same equation for all auction
formats.
For all other buyers, the losers, the utility is equal to zero: they don’t get to enjoy having
the object but also do not pay anything (the only exception is all-pay auction in which
all buyers pay regardless of whether they win or lose).

Outcome What do we care about? Who got the object and how much he paid, the winner and
the price. The outcome depends on the bids placed by all buyers. In all auctions that we
will look at in this class the winner is the buyer who submitted the highest bid. The
price is determined differently in different auction formats.

Bayesian Nash Equilibrium. What happens when all players maximize their payoffs? We get
some kind of equilibrium, it is our familiar Nash Equilibrium (all player choose strategies
that maximize their payoffs given the strategies of other players). When information is
asymmetric, the equilibrium is called Bayesian Nash Equilibrium (BNE).

Not very formal definition of BNE. A Bayesian Nash Equilibrium is a combination of

players’ strategies such that

3 [idudnyk@sfu.ca]
 ECON 302 - Microeconomics II IRYNA
DUDNYK

• All players (all N buyers)

• All types for each player, (for any possible valuation that each buyer has)

• Maximize their expected payoffs (expected because they are not sure about what other
players will do and what payoff they will actually get in the end)

• When maximizing expected payoffs, each player forms beliefs about what are the strate-
gies of other players (each buyer does not know what other buyers will bid, and therefore
has to make some projections about that).

4 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

Classification of auctions.

There are several ways to auction an object off that are actually used. One can classify
auctions according to the bidding rules and by the nature of the object for sale.
Nature of object: there are 2 types of objects. Objects that have private value and
objects that have common value.

• An object has a private value if each buyer’s valuation is independent of how other
buyers value the object. Example: artwork by not so famous artists. When you look
at a painting of that sort, your willingness to pay for it is determined solely by how
much it aesthetically pleases you and to what extent you want to see it on your wall
every day. Your valuation is ‘private and independent’ because other buyers can value
it quite differently, they might not like it as much as you do, or they might really like
it a lot. Also from how much you value it yourself, you absolutely cannot predict how
much other people value it, because values are subjective.

• An object is common value if the value of the object is exactly the same for all buyers.
A perfect example of a common value object is an envelope with money, whichever
buyer gets the envelope, his value of the object will be exactly the same and equal to
the number of monetary units contained in the envelope. A more realistic example is an
oil field, the amount of oil in the field is fixed and all buyers’ value of the field amounts to
the profits they can extract from it. Why do we have an auction then? Because usually
in case of objects like oil field or a firm, buyers cannot know the exact value of the object,
they do some research about how much oil is in the field (how potentially profitable is the
firm) and form their expectation about the value of the object and place bids according
to their expectations. Since buyers’ estimates are not 100% accurate, they will make
mistakes. When object is common value, it can be the case that the winner pays more
than the object is worth P > V and gets negative utility U = V − P < 0, this is what
is called winner’s curse 1 .

• When you see a problem, how do you know it’s common or private value object? Very
easy. Look at buyers’ valuations: if all buyers have same (common) valuation, then it’s
a common value object. If all buyers have different valuations, then it’s private value
object.

As for the bidding procedure, there are four standard auction formats (Note: any pri-
vate/common value object can be auctioned off using any auction format). The system works
1
For more detail on how it works, see tutorial problem

5 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

as follows:

Can buyers observe each other’s bids?

Yes. Open outcry (oral) auction. In open auctions buyers bid by raising hand or sending
some other kind of observable signal that not only the auctioneer, but also all other
buyers can perfectly see.

No. Sealed-bid auctions: buyers submit their bids to auctioneer in sealed envelopes, so that
buyers cannot observe each other’s bids whatsoever.

In all actions the buyer who submits the highest bid gets the object. But auctions differ in
terms of the price that is charged for the object. What is the price that the winner has to pay?

Own bid. If the winner (the buyer who bid the highest) has to pay price equal to his bid,
then it is called first-price auction: P =winning bid.

Second highest bid (not winner’s own bid). If price is equal to second-highest bid, then it is
called second-price auction. P = 2nd highest bid, below the winning bid.

In their search for optimal auction format, people used all possible combinations of the above.
The four standard auction formats are:

English auction is second-price oral auction.

Second -price sealed-bid auction, speaks for itself.

First -price sealed-bid auction.

Dutch is oral first-price auction.

Let’s look at detailed rules, find optimal bidding strategies and outcome for each auction
format. Keep in mind that when we decide on the optimal bid we imagine ourselves as one of
the buyers who only knows his own valuation, number of buyers, and the distribution of V ’s.

6 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

English Auction
Is the most popular and most widely used auction format.

Bidding Rules. A standard textbook description goes as follows. The auctioneer starts at a
very low price and calls out successively higher prices. That is, he calls a low price
and several buyers raise hands to show that they are willing to pay that price. Then
the auctioneer raises price by a certain small amount (fixed increment), if several buyers
accept this new, higher price, auctioneer raises price again etc. The process goes on until
auctioneer increases a price and does not receive any more bids. Alternative procedure
applied in US used cars auctions is: buyers call out prices without any guidance from the
auctioneer, the auction ends when no higher prices are called. In any case, auctioneer
gives the object to the buyer who placed the last bid (highest bid wins), gives object
to that buyer and charges that last accepted price. Why is it called second-price? I
think because technically the buyer with the highest valuation in the room is waiting
for other buyers to drop-out and gets the object when the person with the second-
highest valuation (places second-highest bid) stops bidding, and that is how the price is
determined: the final price is the price at which the second-highest bidder stops bidding.

Optimal Bid . We have to find optimal bids for all N buyers, each of whom has his own private
valuation Vi . In this auction format optimal bidding behavior is quite straightforward.
Imagine Alfred is one of the N buyers and his valuation is VA = 50. Suppose price
starts at $1 and then is increased by one dollar as long as auctioneer receives bids.
Alfred can see the price and comppare it to his willingness to pay, 50 dollars. As long as
the price does not exceed his valuation, he wil place bids. As soon as price of 51 dollars
is announced, Alfred stops bidding and drops out. Why is it simple: because number of
other buyers does not matter for Alfred and also he does not have to know other buyer’s
valuations to find his optimal bid. In English auction each buyer places bids as
long as price (current bid which is observable) is below his valuation. As
soon as price exceeds buyer’s valuation, he stops bidding and drops out.
Notice that this strategy is dominant.

Outcome. Winner is the buyer with the highest valuation, price is determined by the second-highest
valuation and is equal to the second-highest valuation). 2 .

2
Some textbooks show that the price will be second-highest valuation plus the fixed increment. On the
exams you can choose whatever reasoning is more appealing to you as long as you are consistent, you will get
full marks

7 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

Second-price Sealed Bid Auction.

Bidding Rules. All buyers take a piece of paper and write down their willingness to pay for
the object and submit the sealed envelopes to the auctioneer. Notice: buyers can write
down numbers as they please, nobody can make them write their true willingness to pay
and any buyer can bid below or above his true valuation. After all buyers submit the
envelopes with the bids, auctioneer opens all envelopes. The winner is the buyer who
placed the highest bid. The price the winner has to pay is the second-highest bid.

Optimal Bid . Again imagine Alfred is one of the N buyers and his valuation is VA = 50. Can
we advise him how much to bid? In general Alfred has three strategies available to him:
bid something below 50, bid something above 50, and bid exactly 50. Let’s carefully
examine whether one of these strategies is better than the others.

Bid below 50. Is the first possible strategy. What can happen if Alfred bids something below his true
valuation? Suppose he bids 25 dollars. There are 3 possible scenarios that depend on
bids of other buyers.

• All other buyers submitted bids below 25. Then Alfred’s bid is the highest, he wins the
auction and pays some price below 25 (because all other bids, including second-highest
were below 25). Alfred gets positive utility and is very happy with bidding 25.

• One of the other buyers submitted a bid above 50, suppose 75. It means that there was
a bid not only higher that Alfred’s bid, but also higher than Alfred’s valuation. Does
Alfred want to outbid those buyers if he could? No. Because then he would have to
place a bid above 75, and pay 75 dollars, his utility would have been negative, so Alfred
is still happy with bid of 25 (does not wish he placed a higher bid).

• The highest bid submitted by other buyers was between 25 and 50. Suppose highest
bid was 30. Does Alfred get the object? No, because his bid is not the highest. Is
Alfred happy with his bid of 25? No. If he placed a bid above 30, he could win the
object and still pay 30 dollars and get positive utility. Therefore: If Alfred bids below
his true valuation, there is a possibility that he will not get the object at a price below
his valuation (forgone positive utility).

Bid above 50. is the second possible strategy. What can happen if Alfred bids something above his
true valuation, suppose 75 dollars? Again, there are 3 possible scenarios that depend on
bids of other buyers.

• All other buyers submitted bids below 50. Then Alfred’s bid is the highest, he wins the
auction and pays some price below 50 (because all other bids, including second-highest
were below 50). Alfred gets positive utility and is very happy with bidding 75.

• One of the other buyers submitted a bid above 75, suppose 95. It means that there
was a bid not only higher than Alfred’s bid, but also higher than Alfred’s valuation.
Does Alfred want to outbid 95 if he could? No. Because then he would have to place a
bid above 95, and pay 95 dollars, his utility would have been negative, so Alfred is still
happy with bidding 75 and losing the auction.

8 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

• The highest bid submitted by other buyers was between 50 and 75, suppose 65. Does
Alfred get the object? Yes, because his bid is the highest. Is Alfred happy with his bid
of 75? No. Because the price is equal to 65 and is above Alfred’s valuation, Alfred gets
negative utility. Therefore: If Alfred bids above his true valuation, there is a possibility
that he will get the object at a price above his valuation and receive negative utility.

Bid exactly 50. is the third possible strategy. What can happen if Alfred bids 50 dollars? Again, there
are 3 possible scenarios that depend on bids of other buyers.

• All other buyers submitted bids below 50. Then Alfred’s bid is the highest, he wins the
auction and pays some price below 50 (because all other bids, including second-highest
were below 50). Alfred gets positive utility and is very happy with bidding 50.

• One of the other buyers submitted a bid above 50, suppose 75. It means that there
was a bid higher than Alfred’s valuation. Does Alfred want to outbid 75 if he could?
No. Because then he would have to place a bid above 75, and would pay 75 dollars, his
utility would have been negative, so Alfred is still happy with bid of 50.

• If the highest bid of all other buyers was exactly fifty - then there are two highest bids
of 50. One of the buyers will get the object and pay price of 50, get zero utility and be
indifferent between winning and not winning. Is Alfred happy with bidding 50 dollars
in this case and does he want to bid 51 dollar and win? Even if Alfred placed bid of 51
dollar, he would get the object and still pay 50, and get zero utility, wining in this case
does not make him any happier than losing. But also Alfred does not know that the
second highest bid is fifty, could be 50.5 and then Alfred would win and overpay by 50
cents. So in this case, when the second highest bid is also 50, there is no other bid that
makes Alfred any happier than bidding his true valuation.

Finally, compare the how well all of the three strategies do compared to each other:
bidding below true valuation some times is not a good idea; bidding above the true
valuation is also sometimes not a good idea. Bidding exactly the true valuation is
always a good idea because whatever are all other bids, Alfred never pays too much and
always gets the object when the price (second-highest bid) is below his true valuation.
Therefore, in the second-price sealed-bid auction each buyer has a weakly
dominant strategy to bid exactly their true valuation, whatever it is. To be
totally precise, it is a weakly dominant strategy: bidding truthfully is never worse and
sometimes is better than bidding above/below the true valuation.

Outcome. Winner is the buyer with the highest valuation, price is equal to the second-highest val-
uation. How do we know? Because buyers have a dominant strategy, we know that
if buyers are smart, then the bids that auctioneer actually collected are the same as
the true actual valuations of the buyers. Since the highest bid is equal to the highest
valuation, then the object goes to the person with the highest valuation. Since the price
is equal to the second highest bid and second highest bid is equal to the second highest
valuation, by simple logic price is equal to the actual true second-highest valuation.

So far we analyzed two auctions. Let’s compare them. If you look at the outcomes, the
outcomes are the same (winner and price in terms of buyers’ valuations). What does it mean
for you? It means that if you take a room with buyers with private valuations, you will get

9 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

exactly the same winner and exactly the same price regardless whether you conduct English
or second-price auction. The only difference is that in English auction you will never know
what was the highest valuation, but in the second-price sealed bid auction you can see what
was the highest valuation among all the buyer who were participating in the auction. A few
more comments. Both these auctions are easy on buyers: notice that when we derived optimal
bids we did not use the number of buyers, buyers optimal bids would be exactly the same
regardless there are 2 or 10 000 potential buyers. Also when we derived each buyer’s optimal
bid we did not use any information about possible valuations of other buyers. The buyers did
not have to form beliefs about what would be the bids of other buyers.

10 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

First-price Sealed Bid Auction.

Bidding Rules. All buyers take a piece of paper and write down their bids, or how much they are
ready to pay for the object and submit the sealed envelopes to the auctioneer. Notice:
buyers can write down whatever they want any buyer can bid below or above his true
valuation. After all buyers submit the envelopes with the bids, auctioneer opens all
envelopes. The winner is the buyer who placed the highest bid. The price the winner
has to pay is equal to his bid.

Optimal Bid . Now buyers have no clue what to do. It is clear what they should not do: they
should not bid their true valuations. Why not? Because if a buyer wins, the price is
equal to his own bid, if he bids his true valuation his is sure to get zero utility. So
buyers should bid something lower than their true valuation, but how much lower? The
problem is that there is a trade-off that buyers face when choosing their bid in a first-
price auction. The lower is the bid, the higher is the difference between buyers valuation
and the price he will pay if he wins. But at the same time decreasing the bid gives a
buyer smaller chance of winning the auction at all. The optimal bid balances these two
effects. In terms of math, the optimal bid maximizes buyer’s expected utility:

EU = (V aluation − bid)(probability − to − win)

How does a buyer know his probability to win? He has to form beliefs of what kind of
bids and with what probabilities will be placed in the auction. It also depends on the
number of buyers, because to win you have to place a bid higher than all other N −1 bids.

For example, if we have N = 2 named Alfred and Ben with V ’s uniformly distributed
on [0,1]. Suppose for each of them his bid is a linear function of the his valuation of the
object bi = αVi , or the buyer bids α percent of his reservation price. Also each buyer
believes that all buyers will choose exactly the same α. Let’s find Alfred’s optimal bid.
Alfred wins if ba > bB , notice that bB = αVB . Alfred wins if bA > αVB , the probability
of which is bA /α. Then Alfred’ expected utility is UA = (VA − bA )(bA /α). To maximize,
take the first-order conditions and find that b∗A = V2A . Similarly, b∗B = V2B . which is the
BNE. Recall the definition of BNE, what we just found is:

• if there are only two buyers, Alfred and Ben (for all players)

• whatever are Alfred’s and Ben’s true valuations, VA and VB (for all types of each player)

• given their beliefs about each other’s strategies (given their beliefs of what the other
guy will bid depending on his valuation) It is an equilibrium (situation when all players
maximize their expected payoffs given the strategies of other players) for them to bid
exactly 1/2 of their valuations. Of course, when Alfred and Ben come to an auction
they have a certain valuation and place a certain bid. If there are N bidders Ui =
(Vi −bi )(bi /α)(N −1) , where (bi /α)(N −1) is the probability that all other buyers’ valuations
are smaller than Vi and therefore the submitted bids are lower. This solves for optimal
bid bi = NN−1 Vi . Notice, that the more buyers participate in the auction, the closer is
the optimal bid to the true valuation.

11 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

In first-price sealed bid auction each buyer’s optimal bid is equal to some
fraction of the buyer’s true valuation (means that optimal bid is always
below the valuation). As the number of buyers goes up, chances of winning
for each buyer go down, therefore, buyers should bid a higher proportion of
their valuation. Buyers’ bids depend on buyers beliefs about other buyers’
bids and also on the number of buyers.

Outcome. Winner is the buyer with the highest valuation, price is equal to some fraction of the
highest valuation. How do we know? Because bids are proportional to buyers’ valuations
and the higher is the valuation, the higher is the bid. Therefore, buyer with the highest
valuation will submit the highest bid and will win. Price is equal to the highest bid and
it is some proportion of the highest valuation.

12 [idudnyk@sfu.ca]
  ECON 302 - Microeconomics II IRYNA
DUDNYK

Dutch Auction.
Bidding Rules. Price starts really high, hopefully above the highest possible valuation. All
buyers observe the price going down. The auction ends when one of the buyers accepts
the price and wins the object at that price. Why is it a first-price auction - because you
can think of the price as your own bid that you will have to pay once you accept it.

Optimal Bid . Again each buyer knows his own valuation, but not the valuations of the other
buyers. The thing not to do is not to take the object as soon as price reaches your
valuation, because then you will pay the price equal to your valuation and get zero
utility. You should wait. How long to wait? Same as in the first-price auction it
depends on your beliefs about when will somebody else decide to take the object. The
optimal price at which you should place your bid balances the difference between your
valuation and the price and the risk of loosing the object. Technically, you can think
about this price same way you think about the bid in the first-price sealed bid auction.
Math is exactly the same and logic is exactly the same.

Outcome. Winner is the buyer with the highest valuation, price is equal to some fraction of the
highest valuation. How do we know? Because the buyer with the highest valuation
will be the most impatient since he has the most to lose in the case he does not get the
object. The price at which he will take the object depends on his (the highest) valuation
and the number of buyers. Why does the number of buyers matter? Because the more
buyers who are interested in the object the higher are the chances of losing the object
if you wait too long.

Compare first-price sealed-bid auction and Dutch auction. They are strategically identical
in the sense that the buyer with the highest valuation will win, his bid will be equal to some
proportion of his valuation. The optimal bids of buyers depend on their beliefs about the
behavior of other buyers and also the number of buyers participating in the auction.

13 [idudnyk@sfu.ca]