Re: Comments of Leap of Faith Financial Services Inc. (October 18, 2011)

[George Kirikos <icann@xxxxxxxx>]

Just to followup, we did have on additional comment. If any
prospective registry operator wants a concesion on risk premiums, it
must give up something. The "trade" can't be one-sided.

In particular, if there is any concession, a registry operator MUST do
ALL of the following, BEFORE entering into a registry agreement

(1) implement permanent and mandatory price caps that can never exceed
the price of a dot-com registration, AND
(2) not auction off initial "premium" domain names, nor keep them for
the registry itself or its friends, AND
(2) give up their 'presumptive renewal', allowing the TLD to be
tendered every 5 years, so that consumer welfare is maximized via
competitive bidding.

Sincerely,

George Kirikos
President
Leap of Faith Financial Services Inc.
http://www.leap.com/

On Tue, Oct 18, 2011 at 2:55 PM, George Kirikos <icann@xxxxxxxx> wrote:
> Submission by: George Kirikos
> Company: Leap of Faith Financial Services Inc.
> Website: http://www.leap.com/
> Date: October 18, 2011
> Topic: RySG Alternative Proposal for Continuity Operations Instrument
>
> SUMMARY
>
> We strongly oppose the proposal by the RySG and question whether ICANN
> is giving equal access to alternative viewpoints in its new TLDs
> process.
>
> INTRODUCTION
>
> Leap of Faith Financial Services Inc. is a privately held company
> based in Toronto, Canada. We own and operate a network of some of the
> leading websites in the world, including Math.com and School.com.
> We've commented extensively during the new TLDs debate, and oppose
> their introduction.
>
> NEW TLDS PROCESS
>
> 1. In June of this year, the ICANN Board approved the new TLD program,
> despite the lack of proper economic studies, and over the vocal
> objections of the public. It placed the interests of a small number of
> industry insiders above those of the general public. In its
> resolution:
>
> http://www.icann.org/en/minutes/resolutions-20jun11-en.htm
>
> it cut-off debate from the public. However, now we see some of those
> industry insiders (members of the RySG) seeking to tilt the Guidebook
> in their favour one more time, even before new TLDs launch. This was
> not some proposal that came through the GNSO Council, but a privately
> developed proposal by one of its stakeholder groups. There is not even
> a public archive of the RySG mailing list, for example (nor an archive
> of many other individual constituencies, despite ICANN supposedly
> operating "transparently").
>
> 2. The RySG is not the only entity to have made "altervative
> proposals" during the new TLD process. Many others made careful and
> thoughtful submissions, yet ICANN has chosen at this late date to
> single out this proposal of the RySG, and place it on its homepage,
> opening up that sole proposal to the ICANN community for public
> comment? We question why ICANN is giving it preference over all other
> alternative proposals. ICANN should either:
>
> (a) remove the proposal completely (and compel it to go through the
> GNSO Council), or
> (b) provide equal access to all others with alternate proposals
>
> ALTERNATIVE ALLOCATION METHODS FOR NEW TLDS
>
> 1. ICANN has proposed a self-serving means of introducing new TLDs,
> without lack of proper economic studies. It completely ignored the
> input of Tim Berners-Lee, who opposed the introduction of new TLDs:
>
> http://forum.icann.org/lists/2gtld-guide/msg00020.html
> http://www.w3.org/DesignIssues/TLD
>
> 2. ICANN ignored the comments of the NTIA/DOC/DOJ
>
> http://www.ntia.doc.gov/files/ntia/publications/icann_081218.pdf
>
> who clearly stated that "(bottom of page 6) once it has decided to
> authorize a new gTLD, ICANN should implement a process by which
> prospective gTLD operators compete for the privilege of operating a
> particular gTLD by offering terms that benefit consumers. Effectively
> implementing such a process would require that ICANN evaluate bids
> from the perspective of the benefits they provide consumers, not
> merely the amount bidders are willing to pay to ICANN for the right to
> operate the gTLD." They went on to discuss price caps, and competitive
> bidding for renewals of the gTLD registry agreement.
>
> Instead, ICANN is auctioning new gTLDs to the highest bidders, not
> implementing price caps to protect consumers, and is granting
> presumptive perpetual renewal to registry operators.
>
> Despite these precise and reasonable recommendations which directly
> relate to the allocation method for new gTLDs, ICANN at no time put
> forth this proposal to the public on its homepage, or opened up a
> public comment period so that the public could express its support for
> the NTIA/DOC/DOJ recommendations. Yes, somehow ICANN *did* post for
> public comment the proposals of the RySG? Why is ICANN showing that
> favouritism?
>
> 3. ICANN's own "economic experts", in response to criticisms that they
> did not study efficient allocation methods of new TLDs, stated:
>
> http://www.icann.org/en/topics/new-gtlds/analysis-response-phase-ii-report-21feb11-en.pdf
>
> "The Phase II Report did not analyze all possible policy alternatives.
> The report is one component of a larger decision-making process and is
> intended to provide information on economic magnitudes of relevance to
> a wide range of policy analyses." (page 4)
>
> How can one possibly take ICANN seriously when it did not even
> instruct its own economists to analyze all possible policy
> alternatives, i.e. ones that are far superior to those that the
> self-serving ICANN has proposed?
>
> 4. ICANN's own experts stated "Contrary to Krueger’s assertion, the
> Phase II Report did not offer a conclusion as to whether the benefits
> of new gTLDs outweighed the costs." (page 5, same URL as above) This
> damning result was ignored by ICANN, despite it being repeated yet
> again on page 8 "The Phase II Report did not reach conclusions on the
> desirability or undesirability of moving forward with new gTLDs, nor
> was it intended to do so."
>
> 5. In light of the above, it's clear that more work needs to be done,
> including analysis of policy alternatives. In terms of new TLD
> allocation, there are multiple proposals that have been put forth that
> deserve equal access (and public comment periods), instead of ICANN
> giving preference to those proposals put forth by its insiders. Just a
> few of the proposals are:
>
> (i) no new TLDs (status quo)
>
> (ii) .com domains all ascend to the root: in this policy alternative,
> .com essentially becomes the root, so the owner of example.com would
> be entitled to .example, the owner of music.com would obtain .music,
> and so on. This has been suggested by many people at different times.
> This flattens the namespace for all but approximately 200 existing
> registrants.
>
> (iii) Ascended TLDs approach. We put forth this serious proposal in 2009, see:
>
> http://forum.icann.org/lists/irt-draft-report/msg00016.html
>
> which would allow ascension to the top-level if a new TLD applicant
> obtained "easements" from the owners of the .com, .net and .org
> domains. For example, Verizon could easily obtain (or block) .verizon,
> because it owns verizon.com, verizon.net and verizon.org. This would
> be a proposal far superior to ICANN's plan, with fewer objections from
> the public, because it takes externalities into account.
>
> (iv) Regular competitive bidding/tenders for lowest cost to
> registrants: This was the approach advocated by NTIA/DOC/DOJ, namely
> that if the public decided that .shop was desirable, there would be a
> tender process. If Registry A bid $5/yr per domain, Registry B bid
> $3/yr, and Registry C bid $4.50/yr, then Registry B would win the
> fixed term contract, because its bid was the lowest (and thus most
> favourable to consumers).
>
> (v) ICANN's own self-serving proposal, a free-for-all where consumers
> get gouged by profit-maximizing registry operators, who "own" the TLD
> for themselves and can set prices as high as they want, and can pay
> ICANN for that mini-monopoly. It is an allocation mechanism that is
> based on monetizing defensive registrations through FUD, as much of
> the latest "marketing" from new TLD parasite companies has
> demonstrated. ICANN would use the proceeds from auctioning desirable
> TLDs for itself, or "good causes" that its insiders favour. Most
> people do not consider "forced charitable donations" to be very
> different from "taxation."
>
> 6. ICANN should be compelled to give equal time to all these
> proposals, and not just proposals that are put forth by insiders like
> the RySG or by its own staff/Board, many of whom end up at new TLD
> companies as we've seen via the revolving door.
>
> CONTINUITY OF OPERATIONS
>
> 1. It is of critical importance that a registry be able to continue to
> serve its registrants. That is why there are safeguards in any
> reasonable registry agreement, in order to ensure that consumers are
> protected.
>
> 2. The alternate proposal by the RySG of a $50,000 flat fee represents
> a clear one-sided attempt to "game" new TLDs, again, before they even
> launch. One need only look at the table on page 2 of:
>
> http://www.icann.org/en/topics/new-gtlds/rysg-proposal-cof-addendum-01sep11-en.pdf
>
> to see that registry operators would save up to 99% or more of what
> they are supposed to set aside under the current guidebook. While this
> may save money for registry operators, it would put registrants at
> risk. This is pure cost-shifting.
>
> 3. There is an important concept in economics called "signalling"
>
> http://en.wikipedia.org/wiki/Signalling_(economics)
> http://en.wikipedia.org/wiki/Information_economics#Signaling
>
> There is asymmetric information, and various choices can be used to
> "signal" between "good" and "bad" properties. Asymmetric information
> is critical in insurance. For example, if everyone had to pay
> $3,000/yr for car insurance, bad drivers would love it. Those who
> drive a lot or own expensive cars would also love it. Those "higher
> risks" aren't reflected in the policy. In the real world, good drivers
> might pay $800/yr, and bad drivers might instead pay $10,000/yr. Some
> drivers might be uninsurable. That's left up to the market.
>
> 4. By implementing a flat rate group insurance plan, ICANN would be
> eliminating the "signals" provided by the marketplace that would
> eliminate or punish bad actors, and instead would be saying that
> everyone is equally risky. This is obviously not the case, and would
> benefit the "bad drivers" and "riskiest drivers" the most.
>
> 5. People have bemoaned the lack of economic studies, and ICANN has
> not heard those cries. Instead, it listens to the whispers behind
> closed doors of its insiders. This proposal, like a "tell" in poker,
> actually *is* an economic study of its own. Going back to the driver
> example, who on earth would be proposing flat rate insurance plans? It
> would be the *worst* drivers, the riskiest prospective registry
> operators. They would be the ones who would benefit the most. The fact
> that they have the gall to make this proposal before the new TLDs have
> even launched, at this late date, tells us everything. It tells us not
> to trust the past observations of the RySG, the past stories of "great
> demand" for new TLDs. These new proposals seek to subsidize the
> registry operators, instead of allowing the market itself (via third
> party insurance) to determine who is risky and who is not risky. It
> tells us that instead of long-run business models, ICANN wants to
> promote high risk speculation at low cost. ICANN intends to "privatize
> the gains", but "socialize the losses".
>
> 6. This proposal totally ignores the concept of "correlation." By
> pooling the insurance, the RySG is attempting to assert that registry
> failures are independent of one another, and thus one doesn't need as
> large a contingency fund. Nice try, but that's not how things work in
> the real world. Bad things tend to happen at the same time, as those
> with any knowledge of the financial markets or who has observed the
> economy in the past few years can easily recognize. There are
> linkages, or correlations, between events. These bad events can even
> become perfectly correlated at the worst of times --- and that's when
> you need the insurance coverage the most.
>
> For example, suppose you had an instrument that paid $1 million if
> your house was hit by a hurricane, and that insurance costs
> $10,000/yr. Your house is on 123 Main Street, Miami, Florida. If you
> tried to buy house insurance for 125 Main Street, Miami, Florida (i.e.
> right next door), it would still cost $10,000/yr (identical houses).
> If you tried to "pool" the risk together, to pay *less* than
> $20,000/yr for *both* houses, the insurance company would laugh at
> you, because the event risk is nearly *perfectly* correlated (i.e. a
> hurricane hits Miami, it's likely both houses will be damaged at the
> same time, and the risks are purely additive).
>
> If you wanted to get insurance on a 2nd house in Nice, France at the
> same time, instead of the house next door in Miami, there might be
> some limited support to the arguments to be made for what the RySG is
> proposing (global climate change might correlate those risks).
> However, it would *not* support such a drastic reduction in
> fees/coverage. If one TLD fails, it is highly likely that many other
> TLDs will also fail, as their market circumstances are *highly*
> correlated. Many of the applicants will be in the same corporate
> group, for example. Or they face the same exposure to the internet
> economy (all target similar entitities who are buying domain names).
> In other words, if business is bad for one soup company, it's probably
> very likely that business is bad for all the other soup companies too.
> While there is firm-specific risk, there is also industry-wide risk
> that is highly correlated.
>
> In other words, high correlation, an economic and financial reality,
> completely destroys the arguments of the RySG. ICANN should not be put
> in the position of even further engaging in "mission creep" to become
> a private insurer, rather than let the rational marketplace determine
> the true risks. Indeed, ICANN's own operations are correlated to the
> entities it is purporting to insure! When "disaster" strikes at
> registry operators, ICANN would not have the financial cushion itself,
> because its highly likely "disaster" would also befall ICANN at the
> same time.
>
> If I'm buying life insurance, I would not buy it from my parents or my
> children. I would obtain it from an entity which is totally
> uncorrelated with my life events (i.e. "an insurance company"). ICANN
> knows nothing about insurance. Yet it now wants to insure its
> "children"? Preposterous. It'd end up that its "other children" (i.e.
> dot-com registrants) end up paying for all of its mistakes.
>
> 7. Given the RySG proposal is preposterous, it begs the question of
> who would benefit from the proposal. Certainly "fly by night"
> operators, who auction off the highest paying keywords in a TLD,
> immediately pay off its own shareholders with the proceeds (rather
> than saving for a "rainy day"), and then abandon the TLD would benefit
> from a fixed-rate continuing operations fee. If we go back to the
> "signalling" discussion, it's like all the bad drivers being attracted
> to the cheapest insurance plan, one that misprices the risk. These are
> the "get rich quick" schemes that ICANN pretends that it is not
> targeting via the new TLD program.
>
> In the real world, insurance companies are smart, and will assess the
> risk. They don't take input from the "bad drivers", asking them how
> their premiums should be set. The premiums are set by a scientific
> actuarial process. ICANN instead is allowing itself to be lobbied, to
> set the premiums to token minimums, replacing the judgment of the
> marketplace itself.
>
> When all the fly-by-night operators who have engaged in "pump and
> dump" schemes have abandoned their TLDs, it will be left to dot-com
> registrants to foot the bill.
>
> ICANN reveals its true intentions by supporting the "pump and dump"
> model, by having the gall to highlight this proposal on its home page,
> and not highlight the serious alternate proposals advocated by others,
> including my company. ICANN demonstrates that the public and
> regulators like the GAC/NTIA/DOC/DOJ have been bamboozled, that
> registry operators aren't willing to put their own many where their
> mouths have been for the past few years. Instead, these registry
> operators want to shift even more risk to the public, thereby
> threatening the stability and security of the DNS. This
> "cost-shifting" shall not stand.
>
> CONCLUSIONS
>
> 1. ICANN should reject this preposterous proposal, which would
> position itself as an "insurer" for bad actors.
>
> 2. Since ICANN has no ability to even assess the risks (having not
> done the proper economic studies in the first place), not only should
> it not operate as an insurer, it needs to halt the new TLD process and
> complete the required studies, instead of jumping to a predetermined
> outcome that favours itself and its insiders at the expense of the
> public.
>
> 3. Since ICANN has created this precedent of opening up comment
> periods developed by independent parties, and highlighting them on the
> ICANN homepage and through their announcements mailing list, we demand
> equal time. For starters, we resubmit our Ascended TLDs allocation
> method:
>
> http://forum.icann.org/lists/irt-draft-report/msg00016.html
>
> for a formal ICANN comment period and invite others to do the same.
>
> Sincerely,
>
> George Kirikos
> President
> Leap of Faith Financial Services Inc.
> http:///www.leap.com/
>