[ppml] Markets, pricing, transparency, 2008-2 / 8.3.9

[Tom Vest]

Looking over the various threads on 2008-2 over the last couple of  
weeks, it seems that many people are counting on price signals to  
accomplish a variety of different things, e.g., incentivize IPv4  
surplus address space holders to offer their surplus for sale,  
incentivize other IPv4 holders to create surplus (e.g., by migrating  
some requirements to NAT or IPv6), encourage at least some nonzero  
supply of IPv4 prefixes of various sizes, encourage aggregation, etc.

As far as I can tell, nothing currently in 2008-02 supports any  
persuasive argument about how any of these price signals will achieve  
any the stated goals. It may be that Section 8.3.9 permits such  
details to be worked out in the course of the policy's  
"implementation", but I think they're far too important to set aside  
until then...

A brief and incomplete survey of pricing beliefs/doubts follows:

On Mar 6, 2008, at 4:43 PM, Geoff Huston wrote:

> We all share a common view that post unallocated pool exhaustion  
> there will be aa period where demand outstrips supply and in a  
> market situation the price of IPv4 escalates, while the price of  
> IPv6 remains relative constant and relatively insignificant.  
> Ultimately consumers are exposed to the price differential as is  
> conventional when prices of supply shift, and it is reasonable to  
> anticipate that most consumders would express a preference for a  
> lower priced service.
>
> Perhaps the objective here is NOT "to try and construct a perfect  
> Ipv4 market". Perhaps it is more along the lines of "to try and  
> expose some basic economic signals about the relative scarcity of  
> Ipv4 and Ipv6 addresses to provide more direct and focussed impetus  
> for industry-wide IPv6 adoption"

Hi Geoff,

That's a very reasonable argument, but sometimes even the best  
argument can be undermined by its context. Today many of us operate  
in a competitive, "relatively low margin commodity" market (a  
paraphrase of your own description from another message). However,  
it's fair to say that not every operator agrees with that  
description, much less embraces it as inevitable or appropriate. In  
the places where the industry actually does fit that description, it  
only arrived at that state fairly recently, and (for many operators)  
quite reluctantly, and (for all but a handful of the world's largest  
networks) very gradually, as price signals revealing the actual state  
of supply in the marketplace slowly trickled out beyond the veil of  
NDAs and strategic vendor relationships. Even today, there are many  
passionate market advocates that reject this outcome as illegitimate,  
precisely because it was precipitated by regulatory interventions  
that either restricted or compelled  some commercial behavior by  
incumbents. However, you never hear them criticize this development  
because it undercut monopoly rent seeking -- instead the critics  
suggest that by forcing the incumbents to make network inputs  
available for less than monopoly-maximum-$, such rules obscured  
"basic economic signals" and undermined incentives for non-incumbents  
to invest in constructing their own vertically integrated end-to-end  
platforms. By implication, they suggested that incumbents that  
withheld critical bottleneck inputs, or sold them only at absurd  
prices, were actually doing a noble and highly principled thing; they  
were promoting investment, and a transition to a more sustainable  
market!

Now, no doubt there are some staunch libertarians that would make  
that argument "on principle" regardless of any relevant personal  
stakes that might be advanced or harmed as a result. But when the  
incumbents themselves and/or their agents are making the argument,  
non-incumbents and the general public tend to be a little more  
skeptical -- and rightly so.

I am concerned that the market-based IPv4 transfer proposals are  
heading down a very similar path, and that the results are also  
likely to be similar -- and not much to anyone's liking. We do have a  
real supply challenge to overcome in the near term, and I concede  
that market-based mechanisms may be a good way to meet that  
challenge. But given the lessons of the past, I think that any market  
mechanisms adopted this time should require full transparency from  
all parties involved, e.g., public disclosure of non-anonymized  
identities of buyers and sellers, actual prefix exchanged, price  
paid, etc. Without transparency, there can be no "market price", at  
least not for the first 4-5 years -- e.g.,  until enough address  
traders have whispered enough stories to each other at some *OG,  
perhaps enabling someone to write a white paper about it ;-). Without  
transparency, it will never be possible to know or demonstrate how  
the costs and the benefits of this major policy turn have been  
distributed. With transparency, however, these "facts" will remain  
accessible to the whole community, thereby permitting the address  
transfer policies to be sensibly adjusted later as necessary, and  
preserving the "lessons" that come out of this experience for future  
policy deliberations.


On Mar 10, 2008, at 5:34 AM, bmanning at vacation.karoshi.com wrote:

> 	it could be, leo, that for the right price, there might be a
> 	"steady flow" of prefixes.  at least this is what i think
> 	Geoff and David have argued.


How will the community know if there is a steady flow, or if the  
price is "right"? If prices *and* involved parties are not completely  
transparent (and the transaction costs are not a large share of the  
sales prices), it will be trivially easy for anyone with IPv4 to  
pollute any partial/anonymized market information, making it less  
than useless.

On Mar 10, 2008, at 6:28 PM, Scott Leibrand wrote:

> My own amateur-economist take on it is that, initially, supply (of
> currently unrouted space) will exceed demand, so price will stay low,
> and the quantity demanded (your "churn rate") will be only slightly
> lower than it is today (with price ~= 0).  Price will then start to  
> rise
> as the easy supply is put into production, bringing new, more  
> expensive
> supply online, and reducing demand.


If there is transparency in the marketplace, then all of these things  
may happen -- in any case you'll be able to know whether they did or  
not. Without transparency no one will ever know.


On Mar 6, 2008, at 1:43 AM, Geoff Huston wrote:

> ...Others may be of the view that one of the major elements in an IPv6
> transition is the incentive to stop deploying IPv4 and that may  
> well be
> based in an escalating value of IPv4 addresses, which would  
> increasingly
> provide economic incentives for entities to deploy IPv6 as their
> mainstream technology base with minimal IPv4 translation services.


That may be true, but as noted above, historical (and current)  
precedent doesn't provide much confidence. At one time many people  
made the same argument about international telecom settlement flows;  
just send us more, some countries said, and we'll use it to bankroll  
our own network rollout, and thereby gradually offset the settlement  
imbalances over time. Needless to say, no one ever found any evidence  
that this actually worked -- usually it worked the other way round.  
Some incumbent access facilities owners make the exact same argument  
today (and every day since 1984): just permit us to charge end uses  
more, or charge other indirect beneficiaries lots more, or change our  
pricing model with the same overall effect, and we'll deliver the  
latest/fastest/most advanced platform ever. Needless to say, in  
places where they had their way no one has ever found any evidence  
that this actually works, much less works better than current/viable  
alternatives.

Bottom line: revenue is revenue; just because an incumbent operator  
makes additional revenue from IPv4 sales does not mean that that  
marginal addition to the bottom line will be used to fund IPv6  
transition. In fact, if incumbent IPv4 address holders stand to earn  
lots of revenue from IPv4 holdings in an IPv4 & IPv6 scarce  
environment, it's not clear to me that (or when) the incentives to  
capitalize on that arrangement are surpassed by the incentives to do  
their part to make IPv6 ubiquitous.

Out of curiosity, would you support making that linkage between IPv4  
revenues and IPv6 investment an explicit requirement in APNIC's  
proposal?

On Mar 6, 2008, at 1:43 AM, Geoff Huston wrote:

> ...The diversity of views abot scenarios and timings will drive  
> different
> behaviours - some entities may act on the view that once the shock of
> exhaustion wears off and the related market price has factored in
> scarcity premiums as well as the initial shock, the drive to IPv6 will
> assume a new level of momentum, and the IPv4 demand will drop, and  
> with
> declining demand so will market prices - i.e. some folk would see no
> value in hoarding and believe that maximal value will be seen in the
> initial operation of short term market that will be overtaken by IPv6
> relatively quickly. Other entities may be of the view that the inertia
> of this industry is sufficiently large, and the capability of dense  
> NAT
> deployments sufficiently tenable that an IPv4 market may be used for a
> long period. If this is accompanied by a view that demand will  
> continue
> to outpace supply then an escalating price is an obvious outcome.

These are reasonable and useful speculations -- no doubt we could  
hypothesize a wide variety of stakeholder views, and maybe even model  
the weighted results of their behavior on average prices, collateral  
growth of NAT, IPv6 substitution, etc. Maybe we could then bootstrap  
that model, and tweak the parameters, and see how different kinds of  
market arrangements lead to different kinds of outcomes. If we  
actually had a model like that, then maybe some community members  
would be more persuaded that markets would, under some set of  
arrangements, be quite likely to lead to a particular outcome (e.g.,  
IPv6/industry growth and diversity) and not to others (e.g., perma- 
NAT/industry consolidation).

Unfortunately we don't have a model like that, and we're only going  
to have one opportunity to try this out in reality once. If we  
engineer transparency into the transfers mechanisms, however, maybe  
the resulting real-time window onto how things are really working  
will be good enough...

On Mar 6, 2008, at 1:43 AM, Geoff Huston wrote:

> And its not as if such forms of market outcomes spell
> doom and disaster of an out of control stampede. This is not quite the
> wild days of 2000 and the GSM auction debacle appears to have hosed  
> down
> some of the more insane players in this industry. This is a relatively
> low margin commodity business down at the plumbing side and it may  
> well
> be that market pricing may well be influenced by the limited  
> amounts of
> capital that are within the industry these days.


Ahh, right back where we started ;-)

TV