IDIOTS! Foreclosure value is an actual thing. In a zero sum game like a spectrum auction, where there is an absolute limit to the number of spectrum licenses, there is an economic value to denying the spectrum asset to your opponent. That value varies depending upon the incremental benefit of denying the spectrum asset, which is a function of how spectrum constrained your opponents are. If all opponents are equally constrained, then the value is a wash. Everyone accrues the same value of denying the license to their opponents and the foreclosure value does not show up as a distinct element in the bid price.
But when one or two bidders have a substantially greater set of spectrum assets going into the auction, and their opponents are spectrum-constrained, then foreclosure value becomes relevant. Why? Because foreclosure value raises the rational bid price to the spectrum privileged firms. But the benefit is no longer reciprocal, because the marginal cost of to the spectrum privileged firm of losing the asset is much lower.
In other words, if license "X" has a theoretic market value of "$X" based on the ability of the firm to extract value (and assuming that all firms have a relatively equal ability to extract revenue from the license -- a factor I have dealt with separately under the "incumbent discount" argument but lets assume it for now for simplification purpose), then the rational bid price for all firms is $X-$1.
However, to operate a successful wireless firm, a company must have a certain threshold of spectrum. Lack of spectrum prevents you from offering competitive services. Furthermore, the more customers you have, the more capacity you need to have to offer service, or service starts to degrade significantly. This inability of Spectrum Constrained Carrier (Carrier SC) to compete has value to the Spectrum Privileged Carrier (Carrier SP). Let us designate this value to Carrier SP as $Y. This value does not include the opportunity cost of not having the license, which is the same for both carriers. $Y is a distinct value to Carrier SP by foreclosing a needed input for which there is no replacement.
In theory, Carrier SP also suffers from from not having the spectrum. But because Carrier SP already has enough spectrum so that its service will not degrade relative to Carrier SC, the value of $Y to Carrier SC is effectively zero.
Thus, the rational bid for Spectrum Privileged Carrier is not $X-$1, but ($X+$Y)-1. Whereas the rational bid for the spectrum constrained carrier is merely $X-$1. True, you can argue that avoidance of loss for Spectrum Constrained Carrier likewise increases the rational bid (since avoidance of loss is the same as additional revenue), but this misses the point. Spectrum Privileged Carrier, by virtue of its superior spectrum position and auction function, is able to impose a wholly artificial loss on the Spectrum Constrained Carrier to its competitive disadvantage. While this may be efficient from the perspective of Coasian efficiency, it sucks from the perspective of competition policy.
This is true regardless of whether the Spectrum Privileged Carrier warehouses the spectrum or deploys the spectrum. The issue is not whether the Spectrum Privileged Carrier will deploy, but whether the Spectrum Privileged Carrier enjoys an anti-competitive advantage.
Foreclosure value is a thing!!! Deal!!! Stop treating this as a question of morality and bad intent and start treating this like an economic policy question! If you are mouthing platitudes like "auctions put spectrum in the hands of those that will put it to its highest best use" because the market "allocates spectrum to those best able to extract value," then you need to recognize that foreclosure value is part of the calculation of "value" and that the auction will -- absent a corrective mechanism -- allocate spectrum in the most anticompetitive manner.