The Dynamics of FCC Auctions
Term
Paper for Electronic Commerce Course
Course Instructor: Prof. Y. Narahari
SUBMITTED BY:
SACHIN NAGPAL & SUGATO BANERJEE
ABSTRACT
The Federal Communications
Commission (USA) held its first auction of radio spectrum at the Nationwide
Narrowband PCS Auction in July 1994. The simultaneous multiple-round auction,
which lasted five days, was an ascending bid auction in which all licenses were
offered simultaneously. This paper describes the auction rules and how bidders
prepare for the auction and which strategies they follow during the course of
the auction. The minimum bid increment strategies of the auctioneer have also
been discussed. An assessment of the previous auctions has also been
undertaken. The paper also talks about the industry recommendation. The FCC
auctions were viewed by all as a huge success—an excellent example of bringing
economic theory to bear on practical problems of allocating scarce resources.
In the end the US government collected $617 million for ten licenses.
DEPARTMENT OF
MANAGEMENT STUDIES
INDIAN INSTITUTE OF
SCIENCE, BANGALORE
Index
Open Bidding is Better than a Single Sealed Bid
Simultaneous Open Bidding is Better than Sequential Auctions
Assessment of the Auction Design
Revelation of Extensive Information
Similar Items Sold for Similar Prices
Efficient Aggregations Formation
Minimum Bid Increments were Needed
Controlling the Pace of the Auction
Adopt contingencies for extending the auction
Increase the minimum bid increment in the early rounds
Extend the hours of the auction
Reduce the duration of a round
Close a round after all bids are in
Adjusting the Bid Increment in Response to Bidding Activity
Principles of Bid Increment Adjustment
Introduction
The
Federal Communications Commission (FCC) adopted a novel auction form to assign
licenses for the next generation of wireless communication services. Up for
auction were thousands of Personal Communication Services (PCS) licenses, each covering
exclusive rights for a particular slice of the radio spectrum over a geographic
area. These licenses, once developed by firms, promise to expand and improve
wireless services and increase competition throughout telecommunications. The
licenses were assigned using a simultaneous multiple-round auction. This
auction form was proposed by auction experts Paul Milgrom and Robert Wilson of
Stanford University and Preston McAfee at the University of Texas. A
simultaneous multiple-round auction is similar to a traditional ascending-bid
“English” auction, except that, rather than selling each license in sequence,
large sets of related licenses are auctioned simultaneously. In every round, a
bidder can bid on any of the licenses being offered. The auction does not close
until bidding has ceased on all licenses; that is, until a round goes by in
which there are no new bids on any of the licenses. There are three critical
features of this method. First, the ascending-bid design allows the bidders to
react to information revealed in prior rounds. This reduces the winner's curse,
enabling the bidders to bid more aggressively. Second, by auctioning a large
set of related licenses simultaneously, bidders are able to react to prices
across licenses. Since bidder valuations depend on the collection of licenses
held, providing this price information on related licenses is essential to the
formation of efficient aggregations of licenses. Some licenses are complements,
whereas others are substitutes. The simultaneous sale of related licenses in an
ascending bid auction, gives the bidders the flexibility they need to express
these value interdependencies.
Third,
keeping the bidding on all licenses open until there are no new bids gives the
bidders flexibility in switching among license aggregations as prices change.
In evaluating the auction design, it should be recognized that efficiency was
the primary goal of the FCC, not revenue maximization. The FCC sought to assign
the licenses in a timely manner to those firms that will put the licenses to
their best use. High prices are
consistent with an efficient auction, since only bidders with high values are
willing to pay high prices. This assumes that the high values are coming from a
firm's advantage at providing better services at lower prices, rather than an
advantage in restricting market competition. To reduce the chance that high
values are derived from more collusive industry structures, the FCC limits the
amount of spectrum a firm can hold in any market. The bidding process revealed
a great deal of information about likely assignments and relative prices.
Bidders were able to react to this information, shifting bids to alternative
licenses. The information allowed arbitrage across similar licenses, so prices
on similar licenses were close. Finally, the information revealed in the
bidding enabled firms to piece together complementary licenses into efficient
aggregations.
But there are two inherent limitations in any design
that seeks to assign and price the licenses individually. First, such designs
create strategic incentives for bidders interested in multiple licenses that
are substitutes to reduce their demands for some of the licenses in order to
reduce the final prices of the others; this is the demand reduction problem.
Second, even if bidders behave non-strategically, there is a fundamental
problem with the basic concept of individual license pricing when licenses are
complementary. In simultaneous ascending-price auctions, from a bidder’s
perspective this is the exposure problem. A bidder who is unsuccessful
in bidding for a large package of licenses may be left with a partial package
whose total price cannot be justified in the absence of those complementary
licenses it failed to win. This problem is present in any auction mechanism
that sells licenses individually, with no opportunity to bid on packages. In
principle, bidding for packages can resolve both problems, but it brings along
new difficulties that any successful design must overcome. Thus the paper will
mainly concentrate on simultaneous multiple round auction which has also been
implemented in the programming assignment. The discussions in the paper will
also cover some of the aspects of the package bidding which are common to SMR.
Why SMR Auction
Open Bidding is Better than a Single Sealed Bid
An
essential advantage of open bidding is that the bidding process reveals
information about valuations. This information promotes the efficient
assignment of licenses, since bidders can condition their bids on more
information. Moreover, to the extent that bidder values are affiliated, it
raises auction revenues since the winner's curse is reduced. Bidders are able
to bid more aggressively in an open auction, since they have better information
about the item's value. The advantage of a sealed-bid design is that it is less
susceptible to collusion. Open bidding allows bidders to signal through their
bids and establish tacit agreements. With open bidding, these tacit agreements
can be enforced, since a bidder can immediately punish another that has
deviated from the collusive agreement. Signaling and punishments are not
possible with a single sealed bid. A second advantage of sealed bidding is that
it may yield higher revenues when there are large ex ante differences among the
bidders. This is especially the case if the bidders are risk averse. In a
sealed bid auction, a strong bidder can guarantee victory only by placing a
very high bid. In an open auction, the strong bidder never needs to bid higher
than the second-highest value. There was a consensus among experts in favor of
open bidding. The advantage of revealing more information in the bidding
process was thought to outweigh any increased risk of collusion. Collusion was
viewed as unlikely and revenue maximization was a secondary goal.
Simultaneous Open Bidding is Better than Sequential Auctions
A
disadvantage of sequential auctions is that they limit information available to
bidders and limit how the bidders can respond to information. With sequential
auctions, bidders must guess what prices will be in future auctions when
determining bids in the current auction. Incorrect guesses may result in an
inefficient assignment when license values are interdependent. A sequential
auction also eliminates many strategies. A bidder cannot switch back to an
earlier license if prices go too high in a later auction. Bidders are likely to
regret having purchased early at high prices, or not having purchased early at
low prices. The guesswork about future auction outcomes makes strategies in
sequential auctions complex. In a simultaneous auction, a large collection of
related licenses is up for auction at the same time. Hence, the bidders get
information about prices on all the licenses as the auction proceeds. Bidders
can switch among licenses based on this information. Hence, there is less of a
need to anticipate where prices are likely to go. Moreover, the auction
generates market prices. Similar items sell for similar prices. Bidders do not
regret having bought too early or too late. Proponents of sequential auctions
argue that the relevant information for the bidders is the final prices and
assignments. They argue that simultaneous auctions do not reveal final outcomes
until the auction is over. In contrast, the sequential auction gives final
information about prices and assignments for all prior auctions. This final
information may be more useful to bidders than the preliminary information
revealed in a simultaneous auction. Since nothing is assigned until the end in
a simultaneous auction, bidders can punish aggressive bidding by raising the
bids on those licenses desired by the aggressive bidder. In a sequential
auction, collusion is more difficult. A bidder that is supposed to win a later
license at a low price is vulnerable to competition from another that won an
earlier license at a low price. The early winner no longer has an incentive to
hold back in the later auctions. A
final advantage of a sequential auction is that it has been used extensively in
practice and is easier to implement than a simultaneous auction.
Package Bids are Too Complex
A
bidder's value of a license may depend on what other licenses it wins. With a package bid, the bidder either gets
the entire combination or nothing. There is no possibility that the bidder will
end up winning just some of what it needs. With individual bids, bidding for a
synergistic combination is risky. The bidder may fail to acquire key pieces of
the desired combination, but pay prices based on the synergistic gain. Alternatively,
the bidder may be forced to bid beyond its valuation in order to secure the
synergies and reduce its loss from being stuck with the dogs. This is the
exposure problem. A second problem with allowing package bids is complexity. If
all combinations are allowed, even identifying the revenue maximizing
assignment is an intractable integer-programming problem when there are many
bidders and licenses. The problem can be made tractable by restricting the set
of allowable combinations. Increased
complexity is a legitimate concern when considering package bids. Furthermore,
allowing package bids would weaken a central advantage of auctions:
transparency. A bidder who offered a higher bid for part of a combination might
be unable to see why it lost.
Other Issues
Having
settled on a simultaneous ascending-bid auction without package bids, several
issues of implementation remained. How much information should the bidders be
given? The insights suggest that typically more public information is better.
Hence, with the exception of bidder identities in the nationwide auction, the
FCC decided to reveal all information: the identities of the bidders, all the
bids, and the bidders' current eligibility. So long as collusion and predatory
bidding are not problems, revealing more information should improve efficiency
and increase revenues. It also makes for a more open process. Should the rounds
be discrete or continuous? The FCC decided on discrete rounds, which would give
the bidders a specific amount of time to respond to bids. Continuous bidding
has the advantage that it makes endogenous the time between bids. Bidders can
respond quickly when the strategic situation is simple, and take more time when
it is complex. Discrete bidding is easier to implement and it gives the bidders
a specific schedule to follow. Bidders know exactly when new information will
be available and when they have to respond. How will the pace of the auction
controlled? There are three key instruments: the frequency of rounds, the
minimum bid increments, and an activity rule, which sets minimum levels of
bidding activity. These are discussed later.
Auction Rules
Payment
Payments
are received at three times: (1) An upfront payment before the bidding begins
assures that the bidder is serious. Any withdrawal penalties are taken from the
bidder's upfront payment.2
(2) A down payment of
20 percent is paid within five business days after the close of the auction.
(3) A final payment of the remaining 80 percent is paid within five business
days of the award of the license. Licenses are awarded one to three months
after the auction. The upfront payment, due two weeks before the auction
begins, defines the bidder's maximum eligibility in any round of bidding.
Acceptable Bids
In order for a bid to be acceptable in any
round, it must be greater, by a pre-specified increment, than the standing bid
for that license.
Minimum Bid Increments
To
assure that the auction concludes in a reasonable amount of time, the FCC
specifies minimum bid increments between rounds. Initially, bid increments are
set at 5%. Before a license receives a bid, the minimum bid is 0. Bid
increments are adjusted in response to bidder behavior. In the early rounds,
when bid activity is high, the FCC sets larger bid increments; in the later
rounds, when bid activity is low, the FCC sets smaller bid increments.
Activity Rule
The
activity rule is a further device for controlling the pace of the auction. It
forces a bidder to maintain a minimum level of activity to preserve its current
eligibility. As the auction progresses, the required activity increases in
stages. There are three stages in the activity rule. In the initial stage each
bidder must be active on at least one-third of its current eligibility.
Activity is measured as the sum of the MHz-pops on which the bidder submitted a
valid bid or was the high bidder. If activity falls below the one-third level,
then the bidder's current eligibility is reduced to three times its current
activity. In stage 2, a bidder must be active on at least two-thirds of its
current eligibility. If its activity falls below the two-third level, the
bidder's current eligibility is reduced to 1.5 times its current activity. In stage 3 of the auction, a bidder
must be active on 100 percent of its current eligibility. If its activity falls
below 100 percent, the bidder's current eligibility is reduced to its current
activity. A waiver prevents a reduction in eligibility in the event of bidder
error or some other problem. Bidders are given five waivers. Waivers are applied automatically. An
automatic waiver is used whenever a bidder's eligibility would otherwise fall
as a result of its reduced bid activity. A bidder that does not wish to
maintain its eligibility from the prior round may override the automatic
waiver.
Number of Rounds per Day
A
final means of controlling the pace of the auction is the number of rounds per
day. Many rounds are conducted each day. This makes sense given the relatively
small number of licenses and more modest stakes. The time between rounds is
longer early in the auction (about 2 hours), but gets shortened toward the end
when bidding activity is low and strategies are simple. At the end of the
nationwide auction, rounds are occurring every 20 minutes.
Stopping Rule
A simultaneous
stopping rule is used to give the bidders maximum flexibility in pursuing
backup strategies. All markets close if a single round passes in which no new
bids are submitted on any license. But the FCC retains the right to keep the
auction open if there are no new bids in a round. This prevents a premature
close of the auction at the end of stages 1 and 2 if bidders simply are bidding
to maintain eligibility. It also allows the FCC to use a larger bid increment.
If the increment chokes off activity, then the FCC can drop the increment
and/or move to the next stage in order to restore bid activity.
Bid Information
Each bidder is fully informed about the identities of the
bidders, the size of the upfront payments, and which bidders qualify as designated
entities. High bids and bidder identities are posted after each round. In
addition, all valid bids and bidder identities are displayed at the conclusion
of each round, together with each bidder's eligibility and waivers.
Bid Withdrawal
After
the close of each round, there is a brief withdrawal period in which the high
bidders can withdraw their bids subject to a bid withdrawal penalty. If a
bidder withdraws its high bid it is then penalized or its upfront money is
forfeited. A withdrawn high bid counts as bidding activity for the high bidder
in the round the bid is withdrawn. This enables the bidder to switch licenses
without losing eligibility. To discourage insincere bidding, there are
penalties for withdrawing a high bid.
Quantity Restrictions
To
promote competition, a firm is limited in the quantity of spectrum it can hold
in any market. For narrowband spectrum, a firm can hold no more than three
licenses in any market
Strategic Bidding
The
simultaneous multiple-round auction, by revealing information and giving
bidders enormous flexibility in responding to information tends to minimize the
need for elaborate bidding strategies. Nonetheless, a simple strategy of
bidding up licenses until price exceeds value is probably far from optimal.
Firms bidding in multiple markets have an incentive to under bid; that is, to
bid in fewer markets than they desire at current prices. The bidders face a
complex matching problem: who should get which licenses? They have a strong
interest in resolving this question before prices get too high. The auction can
be thought of as a negotiation process in which bidders begin by making
conflicting demands. The auction ends when enough bidders reduce demands, so
that excess demand is zero. Signaling is a device to facilitate the sorting.
Signaling can take the form of public announcements or it can be through
bidding behavior. Code bidding, double bidding, jump bidding, raising one's own
bid, and strategic drops or shifts in bids are all examples of strategic
bidding.
Code Bidding
Code
bidding involves using the last few digits of a bid to signal information to a
rival. The code bid can threaten the bumped bidder to drop out of the market or
face retaliation in another market.
Double Bidding
Bidding
on both licenses in a market — became a common strategy in last rounds of the
auction. If the market is contested (there are three or more bidders active in
the market) or prices are low, the double bid is a cheap way to maintain
eligibility. It also facilitates tacit collusion by maintaining eligibility
without moving into a rival's territory. Double bidding exposes the bidder to
withdrawal penalties and it may increase prices in the markets with double
bids. Hence, the double bidder often pushed price above what it would have been
without the double bid.
Strategic Shifts or Drops
Strategic
shifts or drops can be used to facilitate collusion. In a strategic shift, a
bidder shifts to another license to keep prices in other markets from escalating.
If firms X and Y are competing in market 1 and firm X is in market 2, then Y
switches out of market 1 and into market 2, implicitly telling X to drop 2 to
prevent further competition in market 1. In a strategic drop, a bidder drops a
license, prompting a reciprocal drop from a competitor. If X and Y are
competing in markets 1 and 2, then Y drops market 1, implicitly telling X to
drop market 2. Strategic shifts and drops have two difficulties, which limit
their use. First, the implicit message is much less clear than with a gift
withdrawal or code bidding. Second, strategic shifts and drops are only
effective once the competition is down to two bidders. Prices at this point may
already be high. There is little evidence that strategic shifts or drops were
used successfully to limit competition. In special circumstances, raising one's
own bid may be a good strategy. If the high bidder believes that the remaining
competitor would be willing to bid up one bid increment, but not two, then the
high bidder may benefit from raising its own bid.
Bid Withdrawals
Bid
withdrawals are another example of strategic bidding. The purpose of allowing
withdrawals is to let bidders back out of failed aggregations. However, none of
the withdrawals seems to be motivated by an exit from a failed aggregation.
Rather the withdrawals appeared to be for some other strategic purpose. There
are several reasons for withdrawing a bid:
·
To
back out of a failed aggregation. The withdrawal follows being bumped on
complementary licenses. The bidder
either drops eligibility or shifts to another set of complementary licenses.
·
To
increase flexibility in the next round of bidding. A bidder with little free
eligibility might want to shift among licenses in the next round.
·
To
maintain eligibility or raise rivals' costs. A bidder might engage in a fight
for a license it is not truly interested in. It then withdraws when the
competitor drops out.
·
To
maintain eligibility without raising prices. A bidder withdraws from a license
and then places a minimum bid. When repeated, this maintains eligibility, but
prices do not rise, so long as a competitor places the minimum bid. The
withdrawal signals to others that the bidder is not truly interested in the
license.
·
To
make room for another bidder to drop down. In a fight with another bidder, a
bidder might withdraw to suggest that the competitor move to the withdrawn
license rather than continue the fight. This facilitates tacit collusion by
offering a gift and then lowering the cost of punishment. It is easier to
punish bad behavior by the bidder that takes over a withdrawn license. A raise
by the bidder that withdrew is essentially costless, since the withdrawn bid
amount is already committed. A possible implication of the withdrawals in last
stages some licenses might go unsold. Later in the last rounds, bidders might
not have the eligibility to pick up withdrawn licenses. Fortunately this did
not happen. Most of the withdrawals near the end of the auction were to
increase flexibility in the next round. If the licenses were not picked up by a
competitor, then the withdrawing bidder picked up its own withdrawals.
Jump Bidding
This
behavior seems to fly in the face of common bidding wisdom and perhaps even
common sense. However, there are good reasons for
jump bidding, especially in an auction where the aggregation of licenses plays
a role. The basic idea is that the jump bid conveys information about a
bidder's valuations. It is a message of strength, conveying that the bidder has
a high value for the particular license. Moreover, it conveys this message in a
credible way. Jump bidding has a cost — it exposes the bidder to the
possibility of leaving money on the table. It is precisely this cost that makes
the communication credible. A bidder with a low value would not find it in its
interest to make a large jump. The gain, increasing the chance of winning the
license, would not exceed the cost, the risk of overbidding. All bidders, high
value and low value alike, have an interest in making such a statement. To make
the statement credible, the words must be backed up by an action that a low
value bidder would find too costly.
Figure
5 displays the bid increments by round, stated as a percent of the previous
high bid. The figure also shows bid activity by round, defined as the number of
new bids in the round. Interestingly, the size of the largest jump bids follows
bid activity closely. Intuitively, this is consistent with the tradeoffs
involved in jump bidding. The benefit from signaling a high value is weighed
against the cost of overbidding. When bidding activity is high, the cost of
overbidding is negligible unless a substantial jump is made. Hence,
assuming that the benefit from signaling a high value remains roughly constant,
the size of the jumps should decline with bidder activity.
The
minimum increment introduces a discontinuity in the bidding functions, which
prompts strategic jump bidding. Jump bidding also arises when there is a cost
of bid submission. A desirable collection of licenses is worth more than the
sum of the values of the individual licenses. This provides a strong incentive
for jump bidding. The value of a license depends on the probability that it and
related licenses can be acquired. In bidding with aggregation gains, the
bidders are effectively engaged in a war of attrition. One way to think of jump
biding is that a jump bid effectively “selects” the asymmetric equilibrium
favoring the jump bidder in this war of attrition. However, jump bidding
actually may be a more efficient means of signaling a high value than the
screening in the symmetric equilibrium.
Joint Bidding and Collusion
A
political issue that has been troubling the FCC throughout the auction design
process is how to encourage broad participation in the spectrum auctions. To
encourage small firms to participate, the FCC allows the formation of bidding
alliances for joint bidding. By banding together small firms may be able to
reap economies of scale that would not be available to them if they bid alone.
In future auctions, the FCC allows bidders to form an alliance during the
bidding, so long as the firms have applied for disjoint sets of licenses. Joint
bidding in a common value auction does not necessarily lead to lower revenues.
Although the number of bidders is reduced, so too is the winner's curse, which
allows the bidders to bid more aggressively. Joint bidding could be a problem
if it became so extensive that there would be only a few bidders for each
license. However, given the large number of firms involved and the enormous
gains from deviating from such a collusive posture, it is almost inconceivable
that collusion could persist in such a public process. Collusion has been a
problem in some auctions in the past;28 however,
it should not be difficult for the FCC to discourage collusion among large
public firms. One of the rules in the initial auction intended in part to limit
collusion was keeping bidder identities secret. Based on the nationwide auction
experience, the FCC has since decided to reveal bidder identities in subsequent
auctions. It was felt that the bidders were able to figure out who was who,
despite the confidential bidder numbers, so that little was gained by the
attempt at secrecy. This change will enable firms to focus more on bidding
strategy and less on public surveillance. It should have no effect on
collusion, since under the old rules colluding firms could exchange bidder
numbers. Really the only advantage of concealing identifies is that it might
reduce predatory bidding. However, predatory bidding is most likely in
situations where there are large ex ante asymmetries among the firms, but this
is the case where firms are most able to uncover identifies, regardless of the
FCC's efforts to conceal them. On balance, the benefits of revealing identities
would seem to outweigh the costs.
Assessment of the Auction Design
Since we do not observe the values firms place on licenses, it is impossible to directly assess the efficiency of FCC auctions. Nonetheless, we can indirectly evaluate the auction design from the observed behavior.
Revelation of Extensive Information
Two
essential features of the design are (1) the use of multiple rounds, rather
than a single sealed bid, and (2) simultaneous, rather than sequential sales.
The goal of both of these features is to reveal information and then give the
bidders the flexibility to respond to the information. This should reduce the
winner's curse and more importantly facilitate efficient aggregations.
Proponents of sequential auctions have argued that the information revealed in
a simultaneous auction is of little help to the bidders, because it is only
preliminary information. The final outcome may be far from the current state,
even near the auction's end. There are two dimensions to the information: the
assignment of licenses and the prices of the licenses. Each is considered in
turn. As observed earlier, the upfront payment is an excellent indicator of the
quantity of spectrum won. It, however, tells us nothing about which licenses a
firm will win. For this bidders must look at the bids during the auction. In
each of the auctions, much about the final assignment was determined well
before the auction's end. The clarity of the assignments stems from the fact
that most bidders have focused interests. They bid on a relatively small set of
licenses throughout the auction, although they were typically eligible to bid
on much more. As a result, the number of active bidders in each market are
small. The current bids provide good information about final assignments, but
what about prices. The remaining question is whether bidders have the
flexibility to act on the information. By the time firms have a good sense
about prices and the assignment, they may not have sufficient eligibility to
respond.
Similar Items Sold for Similar Prices
An
advantage of the simultaneous ascending-bid design is that it tends to generate
market prices. Similar items should sell for similar prices. The price
differences among similar licenses were at most a few percent and often zero.
The importance of forming nationwide aggregations within the same band was
probably the source of the larger differences in prices.
The
generation of market prices is important from an efficiency viewpoint. In
addition, it contributes to a sense among the bidders (and observers) that the
auction is fair. The simultaneous stopping rule is an important factor in
achieving market prices and efficiency. Market-by-market closing would not give
the bidders sufficient flexibility. With market-by-market closing, the auction
is essentially a sequential auction with endogenous order. A license may close
by the time a bidder wants to shift to it.
Efficient Aggregations Formation
Valuations
depend on the set of licenses won. Hence, it is important to use an auction
form that allows bidders to express these value interdependencies. Such a
design would encourage the formation of efficient aggregations. Supporters of
the simultaneous ascending-bid design argued that bidders would have sufficient
flexibility to express valuations for combinations of licenses, even without
package bids. The absence of package bids did not
seem to prevent firms from forming efficient aggregations. However, it is certainly
possible that efficiency will reduce, because of under bidding. High-value
bidders may drop out of markets too soon to keep prices on other markets from
escalating. Further evidence of efficient aggregations comes from the absence
of bid withdrawals. There are no withdrawals in the nationwide auction.
Reasonable Auction Durations
An
important advantage of auctions is their ability to quickly assign licenses to
high value uses. The sooner licenses are assigned, the sooner companies can
provide services demanded by consumers. Hence, in judging the auction design,
we must consider how long it takes to conduct the auction. Companies needed
time to think through their options. The short auction durations were possible
in those auctions, where small number of licenses are up for auction. This
meant that many rounds could be conducted in a day. Toward the end of the
auctions, when bidding activity is low and few decisions were being made, more
than one round occurred each hour.
Minimum Bid Increments were Needed
Minimum
bid increments play an important role in controlling the pace of the auction.
If set too high, the increments choke off bidding, even when the high bidder
does not have the highest value. If set too low, the auction may last too many
rounds if bidders bid at the minimum level. Large increments are especially
useful early in the auction when activity is high and prices are low. There is
little cost to large increments early in the auction. Large increments are
inefficient only when they prevent the highest valuer from placing a bid. But
if prices are low, the highest valuer can easily top the high bid by the
minimum increment. Inefficiencies only appear when a license is about to close
and the size of the inefficiency is at most one bid increment. Thus, the
auctioneer can start with a large increment and then reduce the increment as
the probability of closure increases. In the nationwide auction, where all of
the licenses were good substitutes, overall bid activity was an excellent
measure of when licenses were about to close, so a sensible rule tied the bid
increment to bid activity.
Controlling the Pace of the Auction
A
main concern of bidders in the nationwide auction was whether the auction would
come to a natural end, according to the simultaneous stopping rule. Many
bidders feared that the FCC would be forced to declare a final round of bidding
if the FCC did not take steps to control the pace of the auction. Bidders
believed that the FCC had grossly underestimated the number of rounds needed to
reach closure in a simultaneous multiple-round auction. Bidders and the FCC
wished to avoid a declared final round because of the inefficiencies it would
introduce. Much of the benefits of the simultaneous multiple-round auction is
lost if a final round of bidding is declared. There are two main problems.
First, equivalent licenses may go for substantially different prices. If
several bidders with high values happen to bid on the same license, that
license will sell for a premium, while other equivalent licenses may go for
substantially less if only bidders with low values happened to bid on them.
Second, a bidder's values will typically depend on the combination of licenses
it is able to obtain. In a final round, a bidder is unable to express its valuations
for license combinations in an effective way. Such a strategy is risky — the
bidder may lose on both if it happens that another bidder happened to place its
high and low bids on the same licenses. The alternative of bidding high on both
(in the hopes of winning one) is equally risky. Exposing the bidders to these
strategic risks creates inefficiencies, discourages bidding, and reduces
auction revenues. Faced with these potential inefficiencies, it is not
surprising that there was an industry consensus against declaring a final round
of bidding. In assessing how long the auction may last, it is important to make
conservative assumptions about bidding behavior and bidder values. Only in this
way can the FCC be confident that the auction will conclude in the available
time. A conservative bidding strategy is for a bidder to look at the current
high bids and to bid the minimum bid on the licenses that represent the best
value, given the bidder's preferences. Bidding at or near the minimum bid
avoids the risk of leaving money on the table. Given the great uncertainty
about when bidders are likely to drop out, this is a prudent strategy. This
assumption is insufficient to predict the number of rounds needed to reach
closure. There remain two large uncertainties. First, no one knows what the
market-clearing price of the licenses will be. The higher these prices, the
longer it will take to reach closure. Second, it is difficult to predict how
quickly prices will rise, since it is likely that bidding activity will shift
to different groups of licenses from round to round. It is likely that a
license will increase in price in only a minority of the rounds. Toward the end
of the auction it is especially likely that the pace will slow as bidders drop
out.
Industry Recommendations
In
order to avoid a declared final round of bidding, the industry recommended that
the FCC modify the auction procedures as follows:
Adopt contingencies for extending the auction
Given
the large uncertainties about bidder values, bidder behavior, and auction
implementation, the FCC should have viable contingency plans for extending the
auction. Even if it is unlikely an extension will be necessary, it would be
unfortunate if a final round were forced in this first auction, because of a
lack of viable options.
Increase the minimum bid increment in the early
rounds
Increasing
the minimum bid increment in the early rounds can make a substantial difference
in the time to reach closure. Moreover, this is the most desirable way to
increase the pace of the auction. It increases the pace precisely when it is
most valuable
Extend the hours of the auction
The
FCC's proposal allowed for only about 30 rounds. By extending the hours, about
60 rounds could be conducted over the five days.
Reduce the duration of a round
The
FCC proposed one hour between rounds. This could be shortened to 45 minutes or
less. This change is self-correcting. If logistical problems prevent the
auctioneer from processing all bids within the shorter time frame, the round
would simply be extended.
Close a round after all bids are in
Since
the auctioneer knows who the eligible bidders are in any round, the auctioneer
can close the round and immediately post the high bids as soon as every
eligible bidder has bid.
Adjusting the Bid Increment in Response to Bidding Activity
A
critical element of the simultaneous multiple-round auction is the minimum bid
increment.
With
proper adjustment of the bid increment, the auction will come to a timely
closure and yet allow bidders to express small differences in valuations.
Without proper adjustment, the auction will not close in a reasonable amount of
time or it will share the undesirable properties of a single sealed bid
auction. The simultaneous multiple-round auction is most similar to an English
auction. In an English auction, the auctioneer plays a key role in adjusting
the bid increment throughout the auction. With a simultaneous multiple-round
auction, adjustment of the bid increment is even more important for two
reasons. First, there is a significant amount of time between rounds (say, an
hour) rather than a few seconds as in an English auction. Adopting a pace that
is too slow would add days to the auction, rather than minutes. Second, because
there are many items for sale in the simultaneous multiple-round auction, many
more rounds are needed to reach closure than in an English auction for a single
item. As bid activity falls, it becomes more likely that it will take several
rounds for the price of a particular type of license to increase by a single
bid increment. The importance of the bid increment stems from the fact that
bidders may bid near the minimum bids in any round. As in an English auction,
aside from signaling issues, there is little reason to raise a bid by more than
the bid increment. Doing so exposes the bidder to the risk of leaving money on
the table. A bidder knows that if its bid is not the highest in this round, it
will have an opportunity to bid again in subsequent rounds. Of course, a bidder
could bid well above the minimum bids in an effort to hasten the auction. But
this individual action is unlikely to be effective. Other bidders will simply
shift to the other licenses and continue with the minimum increments. The
bidder could place a high bid on all licenses, thereby forcing a rapid pace.
But this strategy involves enormous risk (the possibility of substantial
withdrawal penalties) for little gain (a more timely closure and a reduced risk
of a declared final round). Moreover, there is a free-rider problem. Every
bidder has the incentive to let the other bidders take the risk of overbidding.
Timely closure is unlikely to be assured by the aggressive behavior of an
individual bidder.
Principles of Bid Increment Adjustment
There
are five basic principles for an effective method of adjusting the bid
increment:
1.
Start large
– It is best to
have a large bid increment early, when there are many bidders. A large
increment quickly brings the price level to a point where bidders start to drop
out. In the early rounds, bidding activity is great and there is no cost to a
large increment. The large increment simply saves valuable rounds—rounds that
are better used at the end of the auction.
2.
End small – Toward the end of the auction, the
bid increment should be small, allowing the bidders to express small
differences in valuations. A small increment increases efficiency and
government revenues. It reduces the importance of gaming and strategy.
3.
Reduce the increment
as bidding activity falls – As bidding activity drops the bid increment should fall. The
best way to measure bidding activity is the number of new bids in the prior
round across all licenses. This is a better measure than the number of new high
bids. Since the licenses are substitutes, it is important to use the same
bid increment (adjusting for the size of the license) across all licenses.
Waivers should not be counted as new bids, since a waiver does not indicate a
willingness to raise prices. Indeed, waivers may be used by bidders who are
unwilling to bid under the current bid increment, but who are hopeful that the
increment will fall in subsequent rounds.
4.
Avoid large
drops in the bid increment – The bid increment should decline gradually with bidding
activity. Large drops (say, dropping from 10% to 5% when activity falls below a
particular level) introduce more complex strategies. For example, bidders may
use waivers strategically, anticipating a large drop in the increment. Faced
with a large drop in the increment, bidders are more apt to regret a prior bid.
5.
Adopt and
announce a plan for bid increment adjustment – The FCC should adopt and announce a
plan for adjusting the bid increment before the auction begins. By adopting a
sensible plan up front, the FCC can focus on other things during the auction.
Some modifications may be needed as the auction proceeds, but these changes
would be minor relative to the changes needed under the current plan. With the
current plan, the FCC's behavior is more apt to be viewed as arbitrary. The
prior announcement of the plan allows the bidders to develop bidding strategies
in light of specific rules. Bidding is more apt to be orderly and rational. As
a result, efficiency improves.
Conclusion
Any
auction would look good relative to the FCC's past experience with comparative
hearings and lotteries. Hence, it is remarkable that the FCC chose an
innovative and untested design to auction the spectrum. Fortunately, there is
now substantial evidence that the simultaneous ascending auction worked well.
It raised large revenues. It revealed critical information in the process of
bidding and gave bidders the flexibility to adjust strategies in response to
new information. As a result, similar licenses sold for similar prices, and
bidders were able to piece together sensible sets of licenses. The setting of
the spectrum auctions is too complex to guarantee full efficiency. Bidders with
the highest private values may not have the highest social values. To keep
prices low, large bidders may reduce demand, inefficiently making room for
smaller rivals. Preferences for designated bidders may distort assignments. And
bidders may hesitate to bid for synergistic combinations for fear of not
obtaining the synergies. Nonetheless, an examination of the bidding suggests
that these problems, although present, probably did not lead to large
inefficiencies. Moreover, the cures to these problems have side-effects that
may be worse than the disease. The spectrum auctions are a major step toward
creating a market for spectrum. The greatest room for improvement lies, not in
the assignment of licenses, but in the allocation process.
Bibliography
Ausubel, Lawrence M., Peter
Cramton, R. Preston McAfee & John McMillan, Synergies in Wireless Telephony:
Evidence from the Broadband PCS Auctions, 6 J. Econ. & Mgmt. Strategy
The FCC Auction Design Problem, (Working paper, California Inst. Technology 1995).
Cramton, Peter, Money Out of Thin Air: The Nationwide
Narrowband PCS Auction, 4 J. Econ. & Mgmt. Strategy
Cramton, Peter, The FCC Spectrum Auctions: An Early
Assessment, 6 J. Econ. & Mgmt. Strategy
McAfee, R. Preston, Auction Design of Personal Communication
Services, attached to Comments of PacTel Corporation
Milgrom, Paul & Robert J. Weber, A Theory of Auctions
and Competitive Bidding, 50 Econometrica
Theory, Experiment and the FCC Spectrum Auctions Submitted
to the Federal Communications Commission By Cybernomics, Inc.
McMillan, John (1994), “Selling Spectrum Rights,” Journal
of Economics Perspectives
An Experimental Comparison of the Simultaneous Multi-Round
Auction and the CRA Combinatorial Auction Submitted to the Federal
Communications Commission By Cybernomics, Inc.
WEBSITES REFERRED: citeseer.nj.nec.com/cs
, www.fcc.gov