[arin-ppml] Internet Fairness

[Mike Burns]

Hi Owen,

2014-14 calls for a /16 limit.
You expect some organization to advertise they will buy a /16, and then just 
change names and prosecute a series of transactions, each requiring a new 
ORG-ID. You know, there are a growing group of people who are striving every 
day to advertise the same thing. Brokers like me. We have to balance the 
advertising costs with the deal profits. There is no secret place where you 
can place an ad and magically find enough /16 sellers to change this market. 
And have sellers be serially faked into believing they are talking to 
separately spun-up organizations. And these organizations, though legally 
separate, are shells which are to collude under the control of some existing 
entity with billions of dollars to spend? An entity who is not afraid of 
driving us faster towards IPv6 and a zero value to IPv4? I deal with ARIN 
quite a bit, and I don't for one minute believe that they could start seeing 
ORG-IDS spun up, and then each receiving a needs-free transfer, repeated 
thousands of times, and fail to notice it. Nor do I believe it could 
conceivably happen in any sudden timeframe. In fact, the whole idea is just 
inconceivable.

Call me naive, but I would just speculate in RIPE. Maybe pick up some of 
those multiple /8s gathering dust? Save the creation of about 512 new dummy 
puppet corporations for /16s?

But wait, the evidence in RIPE, after dropping the needs-test, and the short 
experience in APNIC, when they dropped the needs-test, does not support your 
speculation speculation.

>From the ARIN staff review of 2014-14:
Exceptions to needs based review can be justified because the
smaller size of the blocs does not provide a significant vehicle
to 'game', 'hoard', or 'speculate' sizable IP resource blocks of size. Owen: 
Want to bet that there are not organizations out there that would be able to 
create enough entities to absorbe 4,000 /16s pretty quickly for the sake of 
attempting to flip them at a speculative profit?Mike: I would absolutely 
take your bet.Owen: Unfortunately, since this is a case where what would be 
bet is the best interests of the community, the stakes aren't mine to wager, 
nor are they yours.Mike: My bad?Regards,Mike
----- Original Message ----- 
From: "Owen DeLong" <owen at delong.com>
To: "Mike Burns" <mike at iptrading.com>
Cc: "Randy Carpenter" <rcarpen at network1.net>; "Steven Ryerse" 
<SRyerse at eclipse-networks.com>; <arin-ppml at arin.net>
Sent: Saturday, December 20, 2014 6:19 PM
Subject: Re: [arin-ppml] Internet Fairness



> On Dec 20, 2014, at 10:45 , Mike Burns <mike at iptrading.com> wrote:
>
> It's utterly simple to create artificial entities, if necessary.
>
> There are only about 48,000 /16s en toto in all of the IPv4 unicast space. 
> In reality, when you consider the potential transfer market, it's more 
> likely on the order of 4,000.
>
> Want to bet that there are not organizations out there that would be able 
> to create enough entities to absorbe 4,000 /16s pretty quickly for the 
> sake of attempting to flip them at a speculative profit?
>
> I think it's a bad bet.
>
> Owen
>
>
> Hi Owen,
>
> That old chestnut again. So little faith in ARIN staff members to detect 
> what would have to be monumental levels of fraud before any market 
> manipulation could be achieved. ARIN legal staff has reviewed 2014-14, 
> remember, and found your argument lacking.

No, legal stated that there were no issues of liability for ARIN with the 
policy. The potential for fraud was not stated one way or another by the 
legal team in the staff and legal review.

> First, where do you go to find 4,000 /16s? Not that I am accepting that 
> number, I argue the potential transfer market as more like 12,000 /16 
> equivalents (before more widespread CGN could enlarge that number).

I figure there are about 12 /8s or so that are well known to be 
underutilized and that a proportional number of other blocks would account 
for another 4 or so /8s worth of space. Admittedly, like any quantity 
described at this point, it's mostly a SWAG, but I consider it a fairly 
educated SWAG. YMMV.

Where do you get 12,000?

> How do you negotiate quietly with the many, many sellers whom you would 
> have to spend a long time finding?

Why would you have to negotiate quietly?

> Or would you put an ad in the NY Times and hope it escaped the notice of 
> the ARIN community, even though I assume it would be community members 
> engaging in at least one side of these transactions?

Why would an ad in the NY Times have to escape the notice of the ARIN 
community?

Advertise that your organization would like to buy a /n (whatever size gets 
passed with 2014-14 if it gets adopted).

Then, as the calls come in, change your org name after each successful call. 
Nobody looking at the ad knows you're shopping for more than one. People who 
compare notes think they're dealing with different organizations. I'm not 
sure how you expect ARIN to put it together as the transfer requests come 
in.

> Do you think ARIN staff, a group that has processed a few hundred 
> transfers, would be able to even process those transactions in a timeframe 
> and manner invisible to policymakers?

It doesn't have to be invisible, it just has to look like a lot of different 
unrelated transfers.

> And what is the entity willing to engage in arguable fraud, who also has 
> the desire and finances to invest billions of dollars in an asset whose 
> value will decline with every new incentive to transition to IPv6?

You don't have to corner the market to manipulate it. You just need to 
change its visible nature a little and then watch the lemmings do what you 
encouraged them to do. Especially in a small market.

> Or instead of one or two nefarious actors, you assume there would be many 
> smaller fraudsters engaging in enough rampant speculation to impact the 
> market.

I see no reason to believe that there would not be some combination of 
actors across the spectrum. Under the proposed policy, they wouldn't even 
necessarily be nefarious or operating fraudulent. A bit deceitful, perhaps, 
but through omission rather than falsification if they played it right.

> And how many flippers are active in a market with falling prices and an 
> asset value of zero?

You continue to assume falling prices, but there's nothing to indicate any 
likelihood that will be the case for some time.

> And if they flip them to a person who needs them, and makes money in the 
> process, what, exactly is the evil?

The evil is driving up the price for the person who had need and should have 
been able to obtain them at the lower cost. I have no interest in turning 
the IPv4 community resource pool into a source of monetization for parasites 
who offer no value to the community in the process.

> Where is the evidence of anyone, anywhere, buying Ipv4 addresses with an 
> intent to profit through their resale?
> Unless you can provide some, why do we hold policy in thrall to the 
> bogeyman?

Since you are the one arguing for a change in policy, I believe the burden 
of proof rests with you to prove that the policy will not damage the 
interests of the community rather than placing the burden on me to prove 
that it will.

> I guess that's a long way of saying I would absolutely take your bet.

Unfortunately, since this is a case where what would be bet is the best 
interests of the community, the stakes aren't mine to wager, nor are they 
yours.

Owen