The Spectrum of Auctions

Submitted by admin on Tue, 08/14/2007 - 13:09.

eBay Auction

The Spectrum of Auctions

The Spectrum of Auctions
By: Dr. Sam Vaknin

Months of procrastination and righteous protestations to the contrary led to the inevitable: the European Commission assented last week to a joint venture between Germany's T-mobile and Britain's mmO2 to share the mammoth costs of erecting third generation - 3G in the parlance - mobile phone networks in both countries. The two companies were among the accursed winners of a series of spectrum auctions in the late 1990's. Altogether telecom firms shelled well over $100 billion to secure 3G licences in markets as diverse as Germany, Italy, the UK, and the Netherlands.

There is little doubt that governments - and, through them, the public - have made a killing in these auctions. But paying the fees left the winners' coffers depleted. They are now unable to comply with the licence terms and provide the service that is supposed to revolutionize wireless communications and data retrieval.

Judged narrowly, from the sellers' point of view, these auctions have been an astounding success. But the outcomes of the best auctions encompass the widest possible utility - including the buyers' and the public's. From this wider angle, go the critics, spectrum auctions have been an abysmal failure.

This is surprising. Auctions are nothing new. The notorious slave fairs of the 18th and 19th century were auction markets. Similar bazaars existed in ancient Greece. Many commodities, such as US loose leaf tobacco, are exclusively sold in such tenders as are government bonds, second hand goods, used machinery, artworks, antiques, stamps, old coins, rare books, jewelry, and property foreclosed by financial institutions or expropriated by the government. Several stock and commodity exchanges the world over are auction-based. A branch of game theory - auction theory - deals with the intricacies of auctions and how they can be frustrated by collusion implicit or explicit.

All auctions are managed by an auctioneer who rewards the desired article to the highest bidder and charges the seller - and sometimes the bidder a fee, a percentage of the realized price. In almost all auctions, the seller sets a - published or undisclosed - "reserve" price - the lowest bid it is willing to accept and below which the item is "reserved", i.e., goes unsold.

In an English "open outcry" auction, bids are made public, allowing other bidders to up the ante. In a first-price - or discriminatory - sealed bid auction, bids remain secret until the auctioneer opens the sealed envelopes at a pre-determined time. In the Vickrey - or uniform second price - auction the winner pays an amount equal to the second highest bid. In a Dutch auction, the auctioneer announces a series of decreasing prices and awards the article to the first bidder. These epithets are used in financial markets to designate other types of auctions.

Auctions are no longer considered the most efficient method in markets with imperfect competition - as most markets are.

Steve Kaplan and Mohanbir Sawhney noted in an article published by the Harvard Business Review two years ago that the advent of the Internet removed two handicaps. It allows an unlimited number of potential bidders and sellers to congregate virtually on Web sites such as eBay. It also eliminated the substantial costs of traditional, physical, auctions. The process of matching buyers with sellers - i.e., finding equilibrium prices which clear supply and demand efficiently - was also simplified in e-hubs.

Yet, as Paul Milgrom of Stanford University pointed out to "The Economist":

"Arguments that online exchanges will produce big increases in efficiency ... implicitly assume that the Internet will make markets perfectly competitive - with homogeneous products and competition on price alone ... (ignore the fact that) markets for most goods and services in fact have 'imperfect competition' - similar but slightly differentiated products competing on many things besides price."

Moreover, as Paul Klemperer of Oxford University observes, bidders sometimes collude - explicitly, in "rings", or implicitly, by signaling each other - to rig the process or deter "outsider" entrants. New participants often underbid, expecting incumbents to overbid.

An FCC auction of wireless data transmission frequencies in April 1997 raised only $14 million - rather than the $1.8 billion expected. This was apparently achieved by signals to warn off competitors embedded in the bids themselves. Salomon Brothers admitted, in August 1991, to manipulating US treasury auctions - by submitting fake bids - and paid a fine of $290 million.

Another problem is the "winner's curse" - the tendency to bid too high to ensure winning. Wary of this propensity, bidders often bid too low - especially in sealed bid auctions or in auctions with many bidders, says Jeremy Bulow of Stanford University in a paper he co-authored with Klemperer. And, as opposed to fixed prices, preparing for an auction consumes resources while the risk of losing is high.

So, are the critics right? Have the 3G auctions - due to their inherent imperfections or erroneous design - brought the winners to their pecuniary knees? will the sunk costs of the licence fees be passed on to reluctant consumers? Should the European Commission and governments in Europe allow winners to co-invest, co-own, co-operate, and co-maintain their networks?

This, at best, is debatable.

Frequencies are a commodity in perfect competition - though their price (their "common value") is unknown. Theoretically, auctioning the spectrum is the most efficient way to make bidders pay for their "monopoly rent" - i.e., their excess profits. Bidders know best where their interests lie and how much they can pay and the auction process extracts this information from them in the form of a bid. They may misread the market and go bust - but this is a risk every business takes.

Economic theory decouples the size of the bids from the marginal return on investment. But, in the real world, the higher the "commitment fees" in the shape of costs sunk into obtaining the licenes - the more motivated the winners are to recoup them by investing in infrastructure, providing innovative services competitively, and aggressively marketing their offerings. The licences are fully tradable assets whose value depends on added investment in networks and customers.

Too late, telcoms are realizing the magnitude of their mistake. Consumers are ill-prepared for the wireless Internet. Clashing standards, incompatible devices, reluctant hardware manufacturers, the spread of broadband, the recession - all conspire to undermine the sanguine business plans of yesteryear. Yet, getting it wrong does not justify a bail-out. On the very contrary, the losers should be purged by that famous invisible hand. Inexorable and merciless as it may be, the market - unencumbered by state intervention - always ends up delivering commercial, non-public, goods cheaply and efficiently.