To Site Map
Comments Regarding FCC Auction 66 - Advanced Wireless Services
On February 12, 2005, Colts Neck Solutions LLC responded to the FCC's request for comments regarding procedures and other matters pertaining to the FCC's upcoming Auction 66 - Advanced Wireless Services. Below is republished the Executive Summary of those comments. The full text plus other respondent comments are available on the FCC auctions web site.
The FCC's auction process helps to assure transparency in matching bidders with often very valuable spectrum. However, "transparent" is not necessarily synonymous with "readily understandable." Making the auction process more bidder-friendly and, especially, more new bidder friendly, is strategically important to build liquidity and reduce the risk of bidder overload.
A given FCC auction is in fact a set of diverse micro-auctions of hundreds of spectrum "properties." Even after normalizing prices, these exhibit substantial variability in bidder willingness to pay, with pricing outliers on both the high and low side. Also, some license properties do not attract minimum bids, even though the properties in question seem attractive. Either bidders ran out of money or ran out of knowledge and interest
Auctions 44 and 49 are particularly informative, because Auction 49 offered a "second time around" for properties unsold in Auction 44. A majority of the second-look offerings in Auction 49 sold for more than their minimum bids in Auction 44, often substantially more, which suggests some bidder fatigue in the first auction - too many license properties, too many "encumbrance" stories to research, too little bidder mindshare and, perhaps, too little liquidity. Potentially these factors could impact Auction 66.
Therefore expanding both liquidity and the base of bidder knowledge is important to Auction 66's success. In the interest of streamlining the auction process for bidders, especially for new recruits, as well as expanding the knowledge base, seven points are summarized below and are further described in the main body of this submission.
1.Replace the entire BU (Bidding Unit) scheme with a straight percentage upfront payment, if in fact it is worthwhile operating any upfront payment process.
Having people explain to other people what a "BU" is, how it is used, what the trade-offs are, etc. chews up valuable gray matter and time and is not prospective-bidder friendly. Also, the BU control process caps bidders with respect to a "mythical" dollar quantity (BUs) rather than bidder's actual dollar peak commitments. A straight percentage down payment is far more straightforward.
Note that the up front payments in effect function as a large three-four month non-interest bearing loan to the U.S. government, perhaps in the vicinity of $2 billion. Whether the government derives a net benefit, given administrative overheads plus the fact that the governments borrowing costs are lower than many bidders and bidder borrowing costs are tax deductible is an open question.
2.Replace "smoothed," dynamically generated minimum acceptable next bids per license with straight percentage increments for all licenses and rounds
The dynamic creation of bid increment percentages creates unneeded work for everyone apparently little or no gain. For example, for the 216 license properties sold in Auction 38, the smoothing algorithm generated terminal "next bid" percentage increments which had a median value and both quartile boundaries equal to 5% and with 207 of the 216 increments falling in a range of 4.8% to 5.2%. It is far simpler to publish a constant percentage that bidders can embed in their spreadsheets rather than dealing with multiple rates per property per round.
3.Permit bidders to bid in any rounded amounts above the minimum.
It is counterintuitive that the auctioneer dynamically generates "canned" bid amounts per license per round and that a bidder cannot simply bid any number higher than the minimum bid. Again, this complexity makes it more difficult for bidders to use their spreadsheets and other tools. It also encourages in-round bidding ties, which the auction process then "breaks" with another counter-intuitive scheme, a random pick that says, for example, that bidder A's $1 million bid beats bidder B's $1 million bid. Encourage independent choice in bid amounts.
4."Tilt" reserve price computation to align with market preference for higher population density locales.
As is well known, normalized prices for spectrum exhibit an upward trend with respect to increasing locale population. Therefore, a "flat line" reserve and minimum initial bid rule such as Auction 66's proposed $.05 x locale population x MHz inevitably generates prices that are comparatively too high at the low end and too low at the high end. A "tilted" rule that aligns with the "per pop" market trend is suggested. Note that, as indicated in the Auction 44/49 combinations, having properties go unsold is not necessarily a bad result, because the unsold inventory may go for higher prices later.
5.Rather than take on the complexities of packaged bidding, the FCC should perhaps provide somewhat easier bid withdrawal rules.
The complexities of the FCC attempting to formalize "package" bidding are likely to outweigh the gains. Bidders today can package bid, with great flexibility. A modest easing of the bid withdrawal penalties should be as far as the FCC goes to further that cause.
Note that having a license property go "unsold" because of a withdrawn bid is not necessarily, or even usually, a net loss from the seller's perspective. Additionally, the FCC also should encourage development of secondary markets to enable all bidders to dispose of no longer wanted license properties.
6.It is suggested that the FCC not constrain its level of information-distribution and indeed perhaps even increase it.
The proposed withholding of bidder-identified round results to battle perhaps non-existent bidder collusion is may be "cure" probably worse than the purported disease. For the FCC to attain secrecy regarding who bids for what over the multi-day term of the auction, the FCC must plug every leak and loophole. A nominally secure, but still leaky environment creates a new form of bidder advantage, so every relevant communication, every help desk call, every report program etc., must be sanitized before and throughout the multi-day auction process.
Further, even if the FCC is successful in sanitizing results, most bidding contests consists of two-way license-specific duels between entities with fairly predictable interests, so in the end many bidder identities will be eminently guessable.
With respect to bidder's Form 175 pre-auction expression of interest in given licenses, it is suggested that bidders be free to bid on any line item in Auction 66. In that case (which some bidders approximate by filing very wide-ranging interest lists), the FCC would no longer need to collect such data on Form 175, saving bidder effort on a not very useful process. It appears that the Form 175 data collection originated to double-check bidders' advance payments and BU's, another candidate for simplification.
The FCC's auction system's ability to "filter" feedback to bidders is a user convenience that should be made a self-service function, with the user having ability to expand or contract "field of view" during the auction. The fact that bidders' overlook opportunities may be an unintended result of the Form 175 preference process plus undue filtering.
Besides considering publishing less information, the FCC should also consider the advantages of publishing more - for example, publishing bids in real-time, so that unlike today's practices there could be a succession of competitive bids per round. Moreover, it also should consider permitting "late joining" - e.g., if Day One results indicate to a non-enrolled potential bidder that some properties are undervalued, perhaps prospective bidders should be permitted to join to add their knowledge and bidding power to Day 2 and Day 3, etc. of the auction. Exploiting the FCC's investment in automation in these ways can help expand liquidity.
7. A question not asked by the FCC, but an important one, is whether the prospective bidders for Auction 66 have been provided with enough information regarding "encumbrances" to make the auction effective and efficient. The answer seems to be no.
A critical success factor for Auction 66 is the sufficiency of the information provided to support bidder "due diligence." The 1,100+ license line items in the Auction 66 catalog each potentially intersect with the 1,600+ line items in the NTIA "catalog" of current users of the spectrum being auctioned. Many of those NTIA items are non-geographically aligned or are expressed as group items or are classified, or all three. What bidders need is not great detail regarding incumbents, but Auction 66 line item level "available to promise" dates and geographic diagrams of areas of likely interference.
Although it is certainly appropriate for the FCC and all participants to be vigilant regarding improper bidder behavior and its impact on auction selling prices, the far greater risk to the proceeds of Auction 66 is the difficulty of assessing how the encumbrances impact each of the 1,100 license line items.
From a prospective bidder's viewpoint, each spectrum property in Auction 66 presents a unique "story" in terms of radio-frequency relevant geography, socio-economic conditions, competitive conditions, tower and other infrastructure availability, spectrum "encumbrances," time value of money and others. As described above, the FCC can help bidders by simplifying the auction process and increasing the amount of pre-auction information on each license property's "encumbrances."