Does anyone know of a writeup with actual details of either the auction system or the loophole somewhere? For 4000 words, that transcript carried remarkably little actual information.
Edit: This looks like a good description of the process itself, and explains why the problem is difficult and requires the government to be a middleman. https://www.pnas.org/content/114/28/7202
I don't know if this specific auction is listed, but on the whole the auctions are generally the same. There are two types of Auctions: Simultaneous Multi Round (SMR) and Clock. The rules vary between those two Auctions significantly. SMR auctions are an older model, and clock auctions are more common in recent years, though SMR auctions are not completely dead.
The transcript of the podcast is exactly what I was talking about. As far as I can tell it is pure human interest fluff, it does not explain how the auction worked nor how this hack worked (understandable, since it's probably impossible to describe to people who don't understand the auction).
Here's the entirety of their explanation:
> SMITH: We've been using this metaphor of a superhighway in the sky. And you know how whenever someone wants to build a superhighway in the real world, there's always some run-down farmhouse in the way that the government has to pay way too much money for? The private equity firms had essentially bought all these key farmhouses blocking the new spectrum highway.
> ROMER: But the plan was not to stop the highway, to block the auction altogether. It was just to make it more expensive. Private equity and their consultants figured out which frequencies the FCC's computers valued the most, the frequencies you really needed to make the whole reorganization work. Once you had those, you could drive up the prices for all the TV stations, which private equity happened to own a lot of.
What reorganization? That's the first time it was mentioned. Why are some frequencies more valuable? (Or alternatively, why is it surprising and non-obvious that some are more valuable). Why would owning stations at these specific frequencies allow you to drive up the other prices?
Let me see if I can find the podcast I listened to (or I can read this transcript) because it went into detail about the auction (the science behind it), how they knew about the flaws before but had spent too much time and money to not use it, it talked about how the private equity ultimately gamed it and made a bucketful of money. And ultimately how the public who owns the spectrum at the end of the day did not make any money off of it. I'll update here if I can find it. It was very interesting.
Edit: you could probably also read the publications on two-sided simultaneous incentive auction and the associated nobel prize that was awarded for this.
I think I got most of the answers I needed from the publication I linked above. The goal of the auction was to convert the same frequencies across the country, rather than end up with a patchwork.
To make this possible, there were three basic possible outcomes for a TV station.
1. Get bought out at a price you are happy with.
2. Keep your frequency.
3. Have your frequency replaced with something equivalent (capable of teaching 99.5% of your current potential audience)
With that last point being the key. Even if you happened to own a bit of frequency that was needed for mobile networks, it should not allow you to extract any extra money from the system since you'd just get another frequency slice instead.
How do you game this? Obviously if you own all, or almost all, of the frequencies in one geography you can just set your own price, and the third option will not trigger since there is no equivalent frequency for them to swap you with. But that's just a run of the mill market cornering scenario, and the narrative was selling this as something a lot more interesting, so there is probably more to the story than that.
+ Reallocating the spectrum involved an auction that had a multi-year planning and implementation phase.
+ Existing licensees were incentivized to sell their spectrum into the government-run auction. All told, there was $10 billion in sales into the auction, and $20 billion in purchases from the auction.
+ The auction was well-designed in that existing licensees and would-be purchasers would simultaneously be bidding, and there were incentives for both buyers and sellers to price "honestly" i.e. based on how they actually valued their licenses.
+ The design assumed each license would be independent, so to speak, so there wouldn't be coordination/strategy problems. [My sense is this was a not entirely unreasonable assumption given the nature of the existing licenses.]
+ Private equity firms bought a number of small local stations to use as holdouts. The one the story centered on was bought from the licensee for $7 million, and sold into the auction for $20 million.
+ The holdouts raised net payouts by an estimated 7%, so private equity extracted an estimated $700 million (7% of the $10 billion spread).
I'm not sure why this is being downvoted, considering that Kelo v. City of New London[1] provides precedent that eminent domain can be use to forcibly transfer property for the purposes of economic development.
It's precedent at a federal level (which would apply here), but that doesn't mean people like it.
See [Art.] 12-a. [Power to Take Property Limited.] in https://www.nh.gov/glance/bill-of-rights.htm
The government gave the TV stations the spectrum in the 50's for free. The mistake in the auction design was the FCC not simply taking the spectrum space back from the TV stations and auctioning it off as fresh spectrum blocks. The frequencies weren't being used by TV stations since the digital switch-over in the mid 2000's.
There's a lot of mechanisms that can be defeated through collusion of the participants. The trick is to think like a cunning participant and sniff out every way to make money out of the system.
I recall the story of some similar spectrum auction (was it in Freakonomics?) where the participants were allowed to bid arbitrary $ amounts, and turned out they were encoding the geographic areas they wanted to buy in the last digits of their $ bids, and thus signaling to competitors to not compete and artificially hold down the price in those areas.
Direct communication would have been easily provable collusion. This way they could use the excuse "Our math guys came up with that offer price based on complex financial calculations. Total coincidence that the last part of the number happens to look like a zipcode."
"Your honor, as part of the settlement we admit to speaking with our auction competitor to agree not to outbid each other by a geographically arranged scheme, in violation of the rules of the auction"
versus:
"Huh, our team loves to do funny things with numbers and has this habit of putting our area code in the last several digits of any bid for our record keeping purposes. Every time we did it, we won the auction, so we kept on doing it as a good luck charm!"
21 comments
[ 1.9 ms ] story [ 52.8 ms ] threadEdit: This looks like a good description of the process itself, and explains why the problem is difficult and requires the government to be a middleman. https://www.pnas.org/content/114/28/7202
I don't know if this specific auction is listed, but on the whole the auctions are generally the same. There are two types of Auctions: Simultaneous Multi Round (SMR) and Clock. The rules vary between those two Auctions significantly. SMR auctions are an older model, and clock auctions are more common in recent years, though SMR auctions are not completely dead.
Here's the entirety of their explanation:
> SMITH: We've been using this metaphor of a superhighway in the sky. And you know how whenever someone wants to build a superhighway in the real world, there's always some run-down farmhouse in the way that the government has to pay way too much money for? The private equity firms had essentially bought all these key farmhouses blocking the new spectrum highway.
> ROMER: But the plan was not to stop the highway, to block the auction altogether. It was just to make it more expensive. Private equity and their consultants figured out which frequencies the FCC's computers valued the most, the frequencies you really needed to make the whole reorganization work. Once you had those, you could drive up the prices for all the TV stations, which private equity happened to own a lot of.
What reorganization? That's the first time it was mentioned. Why are some frequencies more valuable? (Or alternatively, why is it surprising and non-obvious that some are more valuable). Why would owning stations at these specific frequencies allow you to drive up the other prices?
Edit: you could probably also read the publications on two-sided simultaneous incentive auction and the associated nobel prize that was awarded for this.
To make this possible, there were three basic possible outcomes for a TV station.
1. Get bought out at a price you are happy with.
2. Keep your frequency.
3. Have your frequency replaced with something equivalent (capable of teaching 99.5% of your current potential audience)
With that last point being the key. Even if you happened to own a bit of frequency that was needed for mobile networks, it should not allow you to extract any extra money from the system since you'd just get another frequency slice instead.
How do you game this? Obviously if you own all, or almost all, of the frequencies in one geography you can just set your own price, and the third option will not trigger since there is no equivalent frequency for them to swap you with. But that's just a run of the mill market cornering scenario, and the narrative was selling this as something a lot more interesting, so there is probably more to the story than that.
Promarket has more articles as well.
+ Reallocating the spectrum involved an auction that had a multi-year planning and implementation phase.
+ Existing licensees were incentivized to sell their spectrum into the government-run auction. All told, there was $10 billion in sales into the auction, and $20 billion in purchases from the auction.
+ The auction was well-designed in that existing licensees and would-be purchasers would simultaneously be bidding, and there were incentives for both buyers and sellers to price "honestly" i.e. based on how they actually valued their licenses.
+ The design assumed each license would be independent, so to speak, so there wouldn't be coordination/strategy problems. [My sense is this was a not entirely unreasonable assumption given the nature of the existing licenses.]
+ Private equity firms bought a number of small local stations to use as holdouts. The one the story centered on was bought from the licensee for $7 million, and sold into the auction for $20 million.
+ The holdouts raised net payouts by an estimated 7%, so private equity extracted an estimated $700 million (7% of the $10 billion spread).
But laws are only for little people.
[1] https://en.wikipedia.org/wiki/Kelo_v._City_of_New_London
I recall the story of some similar spectrum auction (was it in Freakonomics?) where the participants were allowed to bid arbitrary $ amounts, and turned out they were encoding the geographic areas they wanted to buy in the last digits of their $ bids, and thus signaling to competitors to not compete and artificially hold down the price in those areas.
"Your honor, as part of the settlement we admit to speaking with our auction competitor to agree not to outbid each other by a geographically arranged scheme, in violation of the rules of the auction"
versus:
"Huh, our team loves to do funny things with numbers and has this habit of putting our area code in the last several digits of any bid for our record keeping purposes. Every time we did it, we won the auction, so we kept on doing it as a good luck charm!"
Here's an FCC write-up of one of the more egregious cases, where the participant was fined $650k https://wireless.fcc.gov/auctions/anticollusion/releases/FCC...
To answer your question, I believe they weren't allowed to communicate.
... invent a conspiratorial narrative ... fabrications and accusations of corruption .. In the fashion of a flat earther...
I'm sorry but if you have to resort to such ad hominem attacks your defense is significantly devalued and someone must have hit a nerve.