outside money
danny
I'm most interested in your work. I've been reading through the dev
wiki and also some posts here and I have a few thoughts to offer,
mainly concerning the subject of debt clearance. Apologies if these
thoughts have already occurred but, just in case they have not...
It seems to me that some here are worried about how ripple boot-straps
given the fact people still need to go outside the system to settle. I
don't believe outside settlement is required : two parties payee A and
payer B only need the debt of a party C which is acceptable to the
payee (A) to settle. The point is A and B can't settle using the debt
of A or B (inside money). So the debt of C is _outside money_. The $US
is outside money to all but US citizens and the rest of the world uses
it to settle. The same principle applies I think in peer-peer land.
So my first observation is that most any debt in ripple ought to be
able to be cleared by the debt of a suitable third party node within
the ripple network which is acceptable to the payee.
My second observation is that entity C can provide the ripple
community with clearance assets by simply deficit spending. Let us
imagine that the ripple network is a village, and the ripple reserve
currency provider is the village shop. The shop pays part or all of
the wages of its staff and part or all the remittance to suppliers in
its own currency, $S, coined as a digital currency. Holders of $S may
redeem them for goods at the shop, or trade the $S amongst themselves
to clear peer-peer debts. There is no need for non-network money,
outside money - the goods of the shop serve as backing. Therefore in
theory all ripple nodes should have digital coin issuance built in.
My third observation is that the main technical issue of digital
currency -namely double-spending – is significantly mitigated in the
ripple trust network as long as double spending detected after the
fact is traced back to a specific node. The voluntary trust networks
established in ripple therefore allow for significant imperfections in
the security of the digital currency. This allows for double-spend
databases of imperfect accuracy to be maintained by nodes other than
the shop and the system will still remain stable. Thus the design of
suitable digital or cryptocurrencies to operate within ripple are
significantly simpler tasks than literature on these digital
currencies normally imply.
My fourth observation is that the set of nodes accepting a given
reserve currency maps nicely to your concept of cell structure
routing, apart from the fact that they need to overlap.
My fifth observation is that ad-infinitum deficit spending by a
reserve currency provider is not sustainable, either by a ripple node
or by the USA. Reserve currency providers must be able to set an
interest rate discount on their currency for as long as their debt is
demanded by other nodes. At the margin this means that the reserve
currency provider must be able to set a negative nominal rate, or
demurrage on their debt. This advantage will serve to sustain their
solvency and will be accepted by other nodes while the others still
have a liquidity preference for their debt. If the coined digital debt
of the reserve currency provider can be subject to demurrage (how are
the mechanics of interest applied to digital currency...) , then their
non coined debt (link credit) must also be capable of paying a
negative nominal rate. Thus when ripple establishes interest payment
features, these should support negative nominal rates of interest.
My final observation is that a ripple style credit network will likely
tend towards a hierarchy of reserve currencies associated with a
hierarchy of cells. The shop must trade with other shops and suppliers
using non-shop debt. At the top of this pyramid some accommodation
must be reached acceptable to all prime-reserve currency providers,
that is not based on inside money. Here the options are either gold,
outside fiat money, or some kind of IMF/SDR type arrangement.
I reckon that a P2P credit network having these features is always
stable without reference to any outside money at all.
Once again, apologies if this is old ground, in which case I hope to
hear the conclusion you guys have reached.
Best,
Dannyr.
Thomas Sherlock
My third observation is that the main technical issue of digital
currency -namely double-spending – is significantly mitigated in the
ripple trust network as long as double spending detected after the
fact is traced back to a specific node. The voluntary trust networks
established in ripple therefore allow for significant imperfections in
the security of the digital currency. This allows for double-spend
databases of imperfect accuracy to be maintained by nodes other than
the shop and the system will still remain stable. Thus the design of
suitable digital or cryptocurrencies to operate within ripple are
significantly simpler tasks than literature on these digital
currencies normally imply.
My fourth observation is that the set of nodes accepting a given
reserve currency maps nicely to your concept of cell structure
routing, apart from the fact that they need to overlap.
My fifth observation is that ad-infinitum deficit spending by a
reserve currency provider is not sustainable, either by a ripple node
or by the USA. Reserve currency providers must be able to set an
interest rate discount on their currency for as long as their debt is
demanded by other nodes. At the margin this means that the reserve
currency provider must be able to set a negative nominal rate, or
demurrage on their debt. This advantage will serve to sustain their
solvency and will be accepted by other nodes while the others still
have a liquidity preference for their debt. If the coined digital debt
of the reserve currency provider can be subject to demurrage (how are
the mechanics of interest applied to digital currency...) , then their
non coined debt (link credit) must also be capable of paying a
negative nominal rate. Thus when ripple establishes interest payment
features, these should support negative nominal rates of interest.
My final observation is that a ripple style credit network will likely
tend towards a hierarchy of reserve currencies associated with a
hierarchy of cells.
The shop must trade with other shops and suppliers
using non-shop debt. At the top of this pyramid some accommodation
must be reached acceptable to all prime-reserve currency providers,
that is not based on inside money.
Here the options are either gold,
outside fiat money, or some kind of IMF/SDR type arrangement.
I reckon that a P2P credit network having these features is always
stable without reference to any outside money at all.
danny
> As far as I can tell,Danny, _outside money_ would only come to play with nodes outside
> the system, e.g. utility, gov't, bank, etc.
My definition of outside money is as follows:
Two parties A and B cannot clear a debt between them using the debt of
A or B. They need to use either the debt of some external party C, or
some commodity. The third party debt, or commoditity is outside money
defined with respecy to A and B.
My observation regarding ripple is simply that there should be enough
nodes in the system that the vast majority of node pairs wishing to
clear debt between them will always be able to find a third node whos
debt is acceptable as a clearance asset. Therefore there should never
be a need to go outside ripple to clear debts, if the debt of the most
popular, liquid and trusted nodes can circulate in currency form.
>
> As a resident of the U.S. and citizen of NJ I would avoid using the term _coin_
> as the U.S. Constitution delegates to the Congress alone the power to coin money.
>
> I don't understand this issue of double-spending.
because it's just digits, spends it again. Later when the second
recipent of the coin tries to redeem his coin he finds it is invalid.
To avoid this the recipent of the copin always needs to check his coin
with the double-spend database, so that he can be sure it has not
already been spent. The problem is that this means a central authority
must be contacted during each transaction.
my observation here was that because each node implicitly trusts its
neighbours then the need to check the double spend database is
reduced. Also as long as the digital coin protocol can trace a double
spend to source after the fact, then the responsible node will be
punished by the rest of the network.
> I don't understand the connection between reserve currency map
> and "cell structure routing". Is this your term or a ripple term?
> Can you reference the source?
locals use this to clear ripple debt. But there is a shop in the next
village also coining its debt as currency. The two shops can't clear
debts between them using either of these monies. So they use say the
coined debt of the local wholsaler, which sells to all the shops. And
so on and so on. These hierarchical and supply chain relationships
exist in the real world so I am suggesting that left to themselves
ripple networks will take similar forms in which nodes cluster around
and connect with the main local provider of liquidity in their trust
network.
>
> Personally I don't like interest as it is form of wealth transfer.
>
> www.margritkennedy.de/pdf/BUE_ENG_Interest.pdf
>
> http://www.youtube.com/watch?v=QuBy3BzCXwg (7 minutes)
>
> And demurrage creates a false sense of urgency, serves
> as a discouragement to join the system and complicates matters.
>
and some won't. If the feature is not provided in ripple people will
just implement add ons that do, so the feature might as well be
offered.
But the point of negative interest on coined debt circulating as local
currency is very important because anyone coining their debt for
others to use as currency must by definition always be in a net
deficit position, otherwise there would be no currency circulating.
Over the long term this fact destroys the currency providers solvency,
just like what has happend to the US. So I imagine that many major
currency providers in ripple, would want to attach a small negative
interest rate to encourage people to redeem the debt of this currency
provider rather than hoarding it like the chinese have been hoarding
US dollars.
> As I understand P2P and ripple, they are artifacts or expressions of an open, decentralized
> loosely bound system. A hierarchy is centralized.
to think about them as clusters.
> Why not some other local or regional commodity to back the currency?
part using this coin, then it is the gods in the shop (priced in the
coin) that back the currency.
hope that make some sense.
danny
Ryan Fugger
On Fri, Jul 23, 2010 at 3:25 PM, danny <daniel.w...@btinternet.com> wrote:
> So my first observation is that most any debt in ripple ought to be
> able to be cleared by the debt of a suitable third party node within
> the ripple network which is acceptable to the payee.
>
Actually, the point of Ripple is precisely to convert obligations of
or held by the payer into those acceptable to the payee. I suppose
"acceptable" is relative though, so there might always be further
clearing required in the future, either through Ripple, or outside of
it.
> My second observation is that entity C can provide the ripple
> community with clearance assets by simply deficit spending. Let us
> imagine that the ripple network is a village, and the ripple reserve
> currency provider is the village shop.
This is one possible scenario. Another is that Ripple users are happy
holding obligations between each other for longer periods of time and
don't require clearing to a relatively central entity C very often.
But local shops could and should be users in the network.
In your scenario, the local reserve currency provider would probably
end up specializing solely in that business and would call itself a
bank, and the system would be little different than what we have
today. The goal of Ripple is to use more than just the trust
relationships to banks as the basis for a monetary system.
> The shop pays part or all of
> the wages of its staff and part or all the remittance to suppliers in
> its own currency, $S, coined as a digital currency. Holders of $S may
> redeem them for goods at the shop, or trade the $S amongst themselves
> to clear peer-peer debts. There is no need for non-network money,
> outside money - the goods of the shop serve as backing. Therefore in
> theory all ripple nodes should have digital coin issuance built in.
>
Why not just use C's Ripple issued currency instead of digital coins?
What's the benefit of digital coins?
> My third observation is that the main technical issue of digital
> currency -namely double-spending – is significantly mitigated in the
> ripple trust network as long as double spending detected after the
> fact is traced back to a specific node. The voluntary trust networks
> established in ripple therefore allow for significant imperfections in
> the security of the digital currency. This allows for double-spend
> databases of imperfect accuracy to be maintained by nodes other than
> the shop and the system will still remain stable. Thus the design of
> suitable digital or cryptocurrencies to operate within ripple are
> significantly simpler tasks than literature on these digital
> currencies normally imply.
>
I believe you're referring to double-spending of digital coins in the
previous paragraph, since Ripple shouldn't have any issues with
double-spending itself.
> My fourth observation is that the set of nodes accepting a given
> reserve currency maps nicely to your concept of cell structure
> routing, apart from the fact that they need to overlap.
>
He's referring to
http://ripple-project.org/wiki/Main/CellStructureRouting, something I
cooked up as a possible way to route payments in a distributed Ripple
network.
Sure. But in the world of a strict hierarchy of reserve currencies,
there would be no need for something as complicated as cell-structure
routing. Routing in a hierarchy is easy -- there's a single path
between any two nodes.
> My final observation is that a ripple style credit network will likely
> tend towards a hierarchy of reserve currencies associated with a
> hierarchy of cells. The shop must trade with other shops and suppliers
> using non-shop debt. At the top of this pyramid some accommodation
> must be reached acceptable to all prime-reserve currency providers,
> that is not based on inside money. Here the options are either gold,
> outside fiat money, or some kind of IMF/SDR type arrangement.
>
You are basically saying that Ripple would likely degrade back to the
banking system we have now, which is possible, for sure. I would hope
not though. It all depends whether users are happy holding and
managing obligations from many sources, or whether they would rather
have the simplicity of one bank account. Part of this depends on how
easy the software makes it for them.
> I reckon that a P2P credit network having these features is always
> stable without reference to any outside money at all.
>
I think so too.
Ryan
> Once again, apologies if this is old ground, in which case I hope to
> hear the conclusion you guys have reached.
>
> Best,
>
> Dannyr.
>
> --
> You received this message because you are subscribed to the Google Groups "Ripple users" group.
> To post to this group, send email to rippl...@googlegroups.com.
> To unsubscribe from this group, send email to rippleusers...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/rippleusers?hl=en.
>
>
danny jpw
> This is one possible scenario. Another is that Ripple users are happy
> holding obligations between each other for longer periods of time and
> don't require clearing to a relatively central entity C very often.
> But local shops could and should be users in the network.
some of their debts for quite a while but the problem with this is it
quickly gums up the network.
Examining the limit scenario, settling outside ripple could become
difficult, e.g. in a scenario where the mainstream monetary system is
suffering from a severe deflation or inflation. So at the limit, some
from of instant clearance asset would help. Whether this matters in
the non limit scenario when there is plenty of reliable outside money
is another question.
I'm also wondering how when there is plenty of outside money, whether
ripple can effectively compete against the mainstream infrastructure
on traditional metrics such as transaction fees, interest rates, small
business credit availability etc. I assume it ought to be able to.
What are your views on that?
> In your scenario, the local reserve currency provider would probably
> end up specializing solely in that business and would call itself a
> bank, and the system would be little different than what we have
> today.
deposits. In my example the shop is more like a central bank than a
commercial bank. In my examlpe, the shop is sovereign in ots own
currency. Perhaps thats what you meant...
> Why not just use C's Ripple issued currency instead of digital coins?
> What's the benefit of digital coins?
allows nodes not connected to C to own C's debt. I guess these nodes
could connect for a relationship in which these nodes only grant C
credit but don't receive credit in return. I'm not sure aht advantage
this confers on the nodes connectnig to C so I'm not sure why they
would do that.
The benefit of a digital coin is that it (an encrypted zero maturity
bearer bond claim on C) can circulate freely peer to peer without them
needing to be connected to C. However once again I'm not sure what you
mean by ripple issued currency so maybe I'm missing something.
> I believe you're referring to double-spending of digital coins in the
> previous paragraph, since Ripple shouldn't have any issues with
> double-spending itself.
incentive to double spend in almost all transactions is tiny, as long
as such double spending is always trace-able after the fact.
The only case in which double spending remains a serious threat is
when two unconnected nodes make cash transactions. In which case the
payee still needs to contact a double spend database to make sure he
is not being defrauded. However assuming these cash only transactions
are a much rarer than debt-floowed-by-clearance transactions (which is
the point of ripple after all) then the load on the provider of the
double spend database is much reduced compared to typical cash-only
digital coin architectures which always assume zero trust exists
between parties.
> Sure. But in the world of a strict hierarchy of reserve currencies,
> there would be no need for something as complicated as cell-structure
> routing. Routing in a hierarchy is easy -- there's a single path
> between any two nodes.
a rigid hierarchy. Actually a better way of describing the way I would
imagine ripple would evolve is a series of overlapping soverign-yet-
private currencies. Leaf node zones represent small communities like a
village and its shop, and larger zones presumably would take shape
around wider area supply chains and fiscal zones. And ripple provides
a way for this self-organised network to reconfigure itself and route
round problems ni a way todays central intrastructure can't possibly
do.
This vision of course is materially changed from what prevails today.
Its more analagous to the situation that holds at the international
forex level.
Just for the avoidance of doubt, I think ripple and similar systems
are the future of money for many people and the work you are doing is
hugely important.
Ryan Fugger
>
>> This is one possible scenario. Another is that Ripple users are happy
>> holding obligations between each other for longer periods of time and
>> don't require clearing to a relatively central entity C very often.
>> But local shops could and should be users in the network.
>
> For sure many nodes are going to be perfectly comfotable with holding
> some of their debts for quite a while but the problem with this is it
> quickly gums up the network.
>
Actually, this paper, posted here recently, says that this is not the case:
http://arxiv.org/abs/1007.0515
> I'm also wondering how when there is plenty of outside money, whether
> ripple can effectively compete against the mainstream infrastructure
> on traditional metrics such as transaction fees, interest rates, small
> business credit availability etc. I assume it ought to be able to.
> What are your views on that?
>
My view is that a lot of the cost of mainstream financial
infrastructure comes from building and maintaining trust relationships
between entities that cannot naturally trust each other, and that by
building on more organic trust relationships, costs should go down.
>
>> In your scenario, the local reserve currency provider would probably
>> end up specializing solely in that business and would call itself a
>> bank, and the system would be little different than what we have
>> today.
>
> commercial banks don't provide currency - they provide bank debt/
> deposits.
As I'm sure you know, nearly all currency in circulation is issued by
banks. I think we're just tripping over the word "currency". I don't
mean to say that banks define a unit of value -- that's the central
bank. Rather, I mean to say that banks issue obligations we use as
money.
> In my example the shop is more like a central bank than a
> commercial bank. In my examlpe, the shop is sovereign in ots own
> currency. Perhaps thats what you meant...
>
But you also raised the idea that somehow different shops in different
communities would need a higher level of reserve currency to provide
liquidity between them. *That* would be the central bank in this
scenario, and the shops' local currencies would become defacto
placeholders for the higher-level currency, just as today, even if
they weren't denominated as such.
>
>> Why not just use C's Ripple issued currency instead of digital coins?
>> What's the benefit of digital coins?
>
> You mean C's ripple issued debt? If so then my response is because it
> allows nodes not connected to C to own C's debt.
If you hold C's debt, then you might as well be connected to C -- that
is the meaning of being connected to C, that you agree to hold C's
debt.
>> Sure. But in the world of a strict hierarchy of reserve currencies,
>> there would be no need for something as complicated as cell-structure
>> routing. Routing in a hierarchy is easy -- there's a single path
>> between any two nodes.
>
> I didn't mean to give the impression that ripple would degenerate into
> a rigid hierarchy. Actually a better way of describing the way I would
> imagine ripple would evolve is a series of overlapping soverign-yet-
> private currencies. Leaf node zones represent small communities like a
> village and its shop, and larger zones presumably would take shape
> around wider area supply chains and fiscal zones. And ripple provides
> a way for this self-organised network to reconfigure itself and route
> round problems ni a way todays central intrastructure can't possibly
> do.
>
I fundamentally agree with this vision -- the Ripple concept is
exactly as you say, a self-organized network of private currencies.
> This vision of course is materially changed from what prevails today.
> Its more analagous to the situation that holds at the international
> forex level.
>
Agreed.
> Just for the avoidance of doubt, I think ripple and similar systems
> are the future of money for many people and the work you are doing is
> hugely important.
>
Thanks.
Ryan
danny jpw
>
> Actually, this paper, posted here recently, says that this is not the case:
>
> http://arxiv.org/abs/1007.0515
again, more slowly :-)
IIRC the paper assumes that connections between nodes sustain
'repeated transactions'.
This is the assumption I feel one needs to be wary of. I'm not sure
what justification the paper presents for that assumption, however I
admit my reading of it so far is just a first pass.
> My view is that a lot of the cost of mainstream financial
> infrastructure comes from building and maintaining trust relationships
> between entities that cannot naturally trust each other, and that by
> building on more organic trust relationships, costs should go down.
future. Unless one posits imminent total collapse (which could also
invalidate the communication infrastructure ripple relies on) , we
should assume such tricks must be competed with and bested in the
short to medium term.
The question therefore for me is how and where ripple can beat the
existing paradigm on its own ground.
> As I'm sure you know, nearly all currency in circulation is issued by
> banks. I think we're just tripping over the word "currency". I don't
> mean to say that banks define a unit of value -- that's the central
> bank. Rather, I mean to say that banks issue obligations we use as
> money.
extended by ripple nodes are currency. The position of commercial bank
and ripple node are competely analagous. Bank debt is not currency.
Bank debt says, I, the bank *owe* you 100 units of *currency*. It is
only deposit insurance that leads people to equate bank debt with
currency (i.e. banks are infallible). Banks don't provide curreency,
and neither, as currently conceived, do ripple nodes.
currency does more than define a unit of value, it promises to redeem
that unit for something actual. A unit of currency issued by the
government is good to redeem your tax obligations. A unit of shop
dollars is promised, by the shop, to be good for a pint of milk,
always and under all circumstances.
I am suggesting that ripple needs currency to make it work, and also
suggesting how individual nodes can issue *currency*, in the same way
that your existing work to date has shown how individual nodes can
issue debt. the world "currency" implies in its language that the debt
is immediate, i.e. of zero maturity. So all I am suggesting is that
ripple incorprate the concept of zero maturity debt.
> If you hold C's debt, then you might as well be connected to C -- that
> is the meaning of being connected to C, that you agree to hold C's
> debt.
>
DESIRE to hold its debt, if it fulfills these desires C must
eventually become insolvent, like the USA. That was my point about
ensuring that debt issuance in ripple can accomodate negative nominal
rates of interest.
However the key point is this, if entity B trusts entity C to be the
local issuer of liquidity, and agrees to hold Cs debt for no premium,
what payback does B get? The answer is nothing, unless entity C mints
zero maturity claims on Cs own ouput and lets B hold them for
liquidity purposes. That is, B will enjoy no increases in B's own
credit capacity by extending credit to C, so what is the point?
The liquidity issue is a quid pro quo thing, and I can't see how that
is currently incorporated in the ripple framework.
Sepp
On Jul 26, 12:04 am, danny jpw <danny...@gmail.com> wrote:
> However the key point is this, if entity B trusts entity C to be the
> local issuer of liquidity, and agrees to hold Cs debt for no premium,
> what payback does B get? The answer is nothing, unless entity C mints
> zero maturity claims on Cs own ouput and lets B hold them for
> liquidity purposes. That is, B will enjoy no increases in B's own
> credit capacity by extending credit to C, so what is the point?
>
> The liquidity issue is a quid pro quo thing, and I can't see how that
> is currently incorporated in the ripple framework.
exchange-network) as being based on mutual trust.
If B and C are friends and would like to set up a currency between
themselves, or rather as part of a circle of mutual and trusted
acquaintances, B would hold some of C's debt and in turn C would hold
some of B's debt. This need not be limited to two people.
The group of friends that creates a currency by mutually expressing
(and formalizing) their trust in each other, could be quite large, and
the borders of it could be somewhat "hazy", in the sense that not all
in the group trust all others, but there is enough of a mutual trust
to form a network, of which each one is a part.
In my imagination, B for instance, could trust C, H, M, and T with
varying amounts of money. In turn, C, M and T as well as K, P and L,
trust B with varying amounts. Each participant in the network thus
adds a certain amount of currency to the network, in tune with the
amount of trust the participant enjoys with others in the network.
It's a new way of looking at currency as a monetization of trust
relationships. It's mutual, and Ripple in this scenario has two
distinct functions:
1) It allows the expression and formalization of permanent or semi-
permanent trust relations and thus the creation of potential liquidity
among a group of friends/acquaintances and
2) it allows exchanges to be mediated by using that trust as a
placeholder, to move as payments from one member of the group to
another and eventually, between the members of different groups that
don't know each other. This is possible because Ripple would form a
network of trust networks where exchange between strangers is
possible, each one of the parties being embedded in their own trust
network, and the different trust networks being joined by Ripple
connections.
Sepp
danny jpw
maturity attached to a promise to redeem the debt for a specific
reference good on demand. The debts in ripple are not that - me and
you would settle our debt whenever we choose. This is entirely
reasonable in a mutual, friendly relationship. The current definition
of ripple is I think, still relying on the existence of outside
currencies, whether domestic/national, or private, to allow nodes to
clear debt from ripple and therefore make room for new/repeated
transactions along connections.
If I owe you a ripple debt of 100 dollars, it is impossible for us to
clear that until you happen to need something (say, your lawn mowing),
which I can provide. If that's all I can do, then we shall have to
wait until summer when your grass is growing and your lawn needs
mowing before we can clear. In the meanwhile, our connection is
blocked (which blocks the network from routing the transactions of
other users via me and you), unless I can come up with outside
currency to clear my debt. And in order to clear that debt in that
currency we'll need to visit the clearing house that manages that
currency, which sort of defeats the point of ripple, to my mind.
I think an interesting way to frame the debate is to assume that no
domestic/national currency is available, or if it is, no-one wants it.
Lets say we have a hyperinflation underway and there is no outside
money. How then can a stock of infinite maturity debts (perpetuities)
accumulated within a P2P credit network be periodically cleared so the
system can retain liquidity? If the only means for clearance of our
debt is for me to mow your lawn or something, then the clearance
currency is barter.
Another way of asking the same question is to start from a no-
currencies-exist position, and then ask what forms of currencies might
arise, and how they would affect a ripple network that begins from a
barter-currency only clearing process. And from there ask what if any
features the credit network should have so it can nurture these
currencies and in turn be nourished by them.
I know ryan believes that ripple is effectively a network of private
currencies (a point which I fully agree with), but what I am
struggling with is the nature of such private currencies and how they
would actually operate along the trust network connections and the
effect they would have by their presence (or absence) on the network
and how any arrangements they make outside ripple (such as issuing
paper notes or running a clearing house) affect network performance. I
believe this issue is very important in understanding credit network
behaviour.
I guess my position is simply that it is tempting to regard the issue
of the nature of private currencies as orthogonal to ripple design
when in fact they are central to it.
Ryan Fugger
>
>
>>
>> Actually, this paper, posted here recently, says that this is not the case:
>>
>> http://arxiv.org/abs/1007.0515
>
> yes, I read that, and I'm currently in the process of reading it
> again, more slowly :-)
>
> IIRC the paper assumes that connections between nodes sustain
> 'repeated transactions'.
>
> This is the assumption I feel one needs to be wary of. I'm not sure
> what justification the paper presents for that assumption, however I
> admit my reading of it so far is just a first pass.
>
If there are no repeated transactions, then there's nothing to gum up, is there?
>> My view is that a lot of the cost of mainstream financial
>> infrastructure comes from building and maintaining trust relationships
>> between entities that cannot naturally trust each other, and that by
>> building on more organic trust relationships, costs should go down.
>
> Of course much of those costs are externalised or pushed into the
> future. Unless one posits imminent total collapse (which could also
> invalidate the communication infrastructure ripple relies on) , we
> should assume such tricks must be competed with and bested in the
> short to medium term.
>
> The question therefore for me is how and where ripple can beat the
> existing paradigm on its own ground.
>
I'm thinking about the cost of operating bank branches and paying bank
employees, lawyers, running associated bureaucracies, etc. Ripple
might be able to cut down on that significantly.
>> As I'm sure you know, nearly all currency in circulation is issued by
>> banks. I think we're just tripping over the word "currency". I don't
>> mean to say that banks define a unit of value -- that's the central
>> bank. Rather, I mean to say that banks issue obligations we use as
>> money.
>
> Money issued by banks is not currency, any more than credit lines
> extended by ripple nodes are currency. The position of commercial bank
> and ripple node are competely analagous. Bank debt is not currency.
> Bank debt says, I, the bank *owe* you 100 units of *currency*. It is
> only deposit insurance that leads people to equate bank debt with
> currency (i.e. banks are infallible). Banks don't provide curreency,
> and neither, as currently conceived, do ripple nodes.
>
> currency does more than define a unit of value, it promises to redeem
> that unit for something actual. A unit of currency issued by the
> government is good to redeem your tax obligations. A unit of shop
> dollars is promised, by the shop, to be good for a pint of milk,
> always and under all circumstances.
>
> I am suggesting that ripple needs currency to make it work, and also
> suggesting how individual nodes can issue *currency*, in the same way
> that your existing work to date has shown how individual nodes can
> issue debt. the world "currency" implies in its language that the debt
> is immediate, i.e. of zero maturity. So all I am suggesting is that
> ripple incorprate the concept of zero maturity debt.
>
Ripple debt can certainly be redeemable for something actual, as in
your case with the village store, or something as simple as getting
your neighbour to mow your lawn.
>> If you hold C's debt, then you might as well be connected to C -- that
>> is the meaning of being connected to C, that you agree to hold C's
>> debt.
>>
>
> if entity C genuinely enjoys such community trust, and all local nodes
> DESIRE to hold its debt, if it fulfills these desires C must
> eventually become insolvent, like the USA. That was my point about
> ensuring that debt issuance in ripple can accomodate negative nominal
> rates of interest.
>
> However the key point is this, if entity B trusts entity C to be the
> local issuer of liquidity, and agrees to hold Cs debt for no premium,
> what payback does B get? The answer is nothing, unless entity C mints
> zero maturity claims on Cs own ouput and lets B hold them for
> liquidity purposes. That is, B will enjoy no increases in B's own
> credit capacity by extending credit to C, so what is the point?
>
> The liquidity issue is a quid pro quo thing, and I can't see how that
> is currently incorporated in the ripple framework.
>
Replace "C" with "bank" and you've just described the whole banking
system. People hold bank debt for essentially no premium precisely
because it is easy to use to buy things and get paid with, same reason
they would hold C's debt for no premium in Ripple. No quid pro quo
required.
Ryan
Ryan Fugger
> Hi sepp. A reasonable definition of currency is a debt of zero
> maturity attached to a promise to redeem the debt for a specific
> reference good on demand. The debts in ripple are not that - me and
> you would settle our debt whenever we choose.
Debts in Ripple can be *anything*, and certainly can be based on an
agreement to settle regularly.
> This is entirely
> reasonable in a mutual, friendly relationship. The current definition
> of ripple is I think, still relying on the existence of outside
> currencies, whether domestic/national, or private, to allow nodes to
> clear debt from ripple and therefore make room for new/repeated
> transactions along connections.
>
It's certainly possible in the current Ripple concept to work without
relying on any outside currencies. Just earn what you spend inside
the network, exactly as we do in the existing banking network. It's
even theoretically possible for the Ripple network as it is conceived
to subsume the existing national currency and banking networks (not
likely obviously).
> If I owe you a ripple debt of 100 dollars, it is impossible for us to
> clear that until you happen to need something (say, your lawn mowing),
> which I can provide.
Not at all. It will clear as payments are routed in the opposite
direction of the payment that initially created the debt.
> If that's all I can do, then we shall have to
> wait until summer when your grass is growing and your lawn needs
> mowing before we can clear. In the meanwhile, our connection is
> blocked (which blocks the network from routing the transactions of
> other users via me and you),
Our account is only blocked in one direction if it is at its limit.
Payments can pass in the other direction, clearing the debt. When one
account is blocked, payments that would use it are routed around. In
simulations I've run, very few payments get completely blocked, and
those are usually because one of the endpoints has used up all his or
her incoming or outgoing credit. This intuition is confirmed by the
paper I mentioned earlier.
> unless I can come up with outside
> currency to clear my debt. And in order to clear that debt in that
> currency we'll need to visit the clearing house that manages that
> currency, which sort of defeats the point of ripple, to my mind.
>
> I think an interesting way to frame the debate is to assume that no
> domestic/national currency is available, or if it is, no-one wants it.
> Lets say we have a hyperinflation underway and there is no outside
> money. How then can a stock of infinite maturity debts (perpetuities)
> accumulated within a P2P credit network be periodically cleared so the
> system can retain liquidity? If the only means for clearance of our
> debt is for me to mow your lawn or something, then the clearance
> currency is barter.
>
Yes, all money is delayed barter. There is no value in debt except
for the work people do for each other because of it.
> I know ryan believes that ripple is effectively a network of private
> currencies (a point which I fully agree with), but what I am
> struggling with is the nature of such private currencies and how they
> would actually operate along the trust network connections and the
> effect they would have by their presence (or absence) on the network
> and how any arrangements they make outside ripple (such as issuing
> paper notes or running a clearing house) affect network performance. I
> believe this issue is very important in understanding credit network
> behaviour.
>
> I guess my position is simply that it is tempting to regard the issue
> of the nature of private currencies as orthogonal to ripple design
> when in fact they are central to it.
>
Ripple would work with any type of obligation that can be exchanged
electronically, so pretty much any private currency design would fit
fine. Someone would just have to write a software client for it, and
hook it in to a Ripple server via an API (which hasn't been
implemented yet either :)
Ryan
danny jpw
> Debts in Ripple can be *anything*, and certainly can be based on an
> agreement to settle regularly.
would show the various ripple debts under various contractual
arrangements on its books, for accounting compliance purposes. It
would be necessary to account for which debts are countable as cash,
which as receivables and so on. The business has to demonstrate its
status as a going concern using these methods to the relevant
authorities.
> It's certainly possible in the current Ripple concept to work without
> relying on any outside currencies. Just earn what you spend inside
> the network, exactly as we do in the existing banking network.
system stable. Some of my points have been about liquidity preference
- that people will prefer to hold on to their shop-debt and spend
their danny-debt, since they are more likely to obtain what they want
on any given day using shop-debt. People preferentially hanging onto
their shop debt will become a problem for the shop won't it?
Have you modelled liquidity preference in your simulations?
>
> Our account is only blocked in one direction if it is at its limit.
> Payments can pass in the other direction, clearing the debt. When one
> account is blocked, payments that would use it are routed around. In
> simulations I've run, very few payments get completely blocked, and
> those are usually because one of the endpoints has used up all his or
> her incoming or outgoing credit. This intuition is confirmed by the
> paper I mentioned earlier.
In this case, what bearing might liquidity preference within the
network have on the activity seen on all my connections?
So the shop, to use your terminology is going to tend towards using up
its incoming credit.
> Yes, all money is delayed barter. There is no value in debt except
> for the work people do for each other because of it.
a promise to redeem it *upon demand*.
Then we have various forms of credit, e.g. agreement with a fixed end
date, an agreement which is a line of credit (a ripple connection)
and agreements with various forms of interest or fees and agreements
which are entirely informal. These distinctions matter in the real
world - there are accounting rules to be followed. If ripple
connections are specifying end dates, settlement frequency and so on
then the connections IMO can no longer be modelled as simple lines of
credit.
I guess the rest of my questions aside from the liquidity preference
issue have to do with maturity transformation, which seems to me to be
an important feature of the existing monetary/banking system. Of
course maturity transformation and liquidity preference are linked
(perhaps the spontaneous emergence of liquidity preference leads to
the need for maturity transformation?).
So I guess I can restate my entire line of argument as two simple
questions:
* How if at all does ripple provide maturity transformation? If it
doesn't then is there a justification for not doing so?
* Is liquidity preference an issue for P2P credit networks?
Thanks for bearing with me on this - I'm really interested to hear the
answers (and I'm still trying to get into the habit of ripple-
thinking - e.g. that payments can pass in two directions).
Ryan Fugger
> Sure. Actually this left me wondering how a business using ripple
> would show the various ripple debts under various contractual
> arrangements on its books, for accounting compliance purposes. It
> would be necessary to account for which debts are countable as cash,
> which as receivables and so on. The business has to demonstrate its
> status as a going concern using these methods to the relevant
> authorities.
>
The Ripple client software could be made to produce a report at any
point, I suppose.
> There is also the requirement to spend what you earn to keep the
> system stable. Some of my points have been about liquidity preference
> - that people will prefer to hold on to their shop-debt and spend
> their danny-debt, since they are more likely to obtain what they want
> on any given day using shop-debt. People preferentially hanging onto
> their shop debt will become a problem for the shop won't it?
>
I don't know. The USA has done quite well by having other people hang
on to their debt. The main problem seems to be if it gets out of
balance in the long run, which is the fault of all parties involved...
> Have you modelled liquidity preference in your simulations?
>
I've mostly thought about liquidity preference in terms of:
a) letting users specify that preference in the routing software so it
prefers certain routes,
b) letting users specify transaction fees so they make more money when
they make certain exchanges as intermediaries, and
c) letting users specify interest on accounts.
I haven't modeled these into simulations, and Ripplepay only
implements (c), and doesn't take it into account for routing as of
yet.
This line of thinking usually makes me wonder if Ripple wouldn't
degrade into a giant forex casino. That's why the direction I'm
personally going with Ripple is quite different, using it to track
favours between friends and friends-of-friends in the community, and
then prioritize favour requests within the community so that the
favours balance out. These kind of obligations are "soft", in that
there is no hard requirement for repayment, only a friendly desire to
keep up when you can. I call it a "favour economy", somewhere between
a pure gift economy and a coercive debt-based economy.
I should probably talk about this more, because my heart is in that
direction far more than the original hard-obligation-style Ripple, and
it makes implementation so much simpler, by relying even more on
established organic trust.
>
>>
>> Our account is only blocked in one direction if it is at its limit.
>> Payments can pass in the other direction, clearing the debt. When one
>> account is blocked, payments that would use it are routed around. In
>> simulations I've run, very few payments get completely blocked, and
>> those are usually because one of the endpoints has used up all his or
>> her incoming or outgoing credit. This intuition is confirmed by the
>> paper I mentioned earlier.
>
> OK, that makes sense. I have to keep reminding myself of this feature.
> In this case, what bearing might liquidity preference within the
> network have on the activity seen on all my connections?
>
My intuition is that the nodes whose debt others will want to hold
will also be granted proportionally larger credit limits, so it may
just balance itself out.
I can imagine situations where two nodes that may not even be
connected end up fighting back and forth to clear debt they would each
rather have in a different form, but that clearing one's undesirable
holdings ends up reinstating the other's. My thought is that if they
assign higher transaction fees to transactions they find undesirable,
and the routing algorithm is made smart enough, then this situation
should resolve in the right way, either by saddling the person who
minds the least having their undesirable debt, or by routing
differently to clear so both parties are satisfied.
> Sure. I guess I would define money as debt with a contract specifying
> a promise to redeem it *upon demand*.
> Then we have various forms of credit, e.g. agreement with a fixed end
> date, an agreement which is a line of credit (a ripple connection)
> and agreements with various forms of interest or fees and agreements
> which are entirely informal. These distinctions matter in the real
> world - there are accounting rules to be followed. If ripple
> connections are specifying end dates, settlement frequency and so on
> then the connections IMO can no longer be modelled as simple lines of
> credit.
>
> I guess the rest of my questions aside from the liquidity preference
> issue have to do with maturity transformation, which seems to me to be
> an important feature of the existing monetary/banking system. Of
> course maturity transformation and liquidity preference are linked
> (perhaps the spontaneous emergence of liquidity preference leads to
> the need for maturity transformation?).
>
> So I guess I can restate my entire line of argument as two simple
> questions:
>
> * How if at all does ripple provide maturity transformation? If it
> doesn't then is there a justification for not doing so?
> * Is liquidity preference an issue for P2P credit networks?
>
Yes, these are good questions. The logical conclusion is that
hard-obligation Ripple must become more than a payment routing system,
it must be a generic debt-clearing marketplace, with buy and sell
orders, options, and all the rest of it. If we could get enough
people participating, this kind of grand-unified debt exchange might
very well be useful, and probably also quite dangerous. I'm
personally not interested in pursuing this direction, but I'd be
interested to watch if someone was...
Ryan
Daniel
I think what makes a forex casino is currency issuers arbitrarily acting
in ways that affect the value of currencies, resulting in gamblers
jumping around after them. Ripple enables users to route around
arbitrary actors, settling on a more stable equilibrium with reliable
currencies. This would make forex more of an exchange and less of a casino.
Not to put too fine a point on it, but I think it makes sense to treat
contractual obligations as "semi-hard", and that this would be far more
valuable than tracking "soft" obligations. Personally I think that
rigorous tracking of soft obligations can even be a bad thing.
Daniel
danny jpw
> rigorous tracking of soft obligations can even be a bad thing.
>
1) monetise the favour economy
2) favourise the monetary economy
3) decentralise the monetary economy
the dev wiki, project pages and ripple.com made me assume you were
after (3) ryan. I'm not sure its possible to monetise the favour
economy, and even if you can whether it is a good idea - it might just
create more demand for bank debt. No doubt bernanke would thank you
for that!
Perhaps a good target is to go for decentralisation and softening of
the monetary economy. Decentralisation is what you have already in
ripple, and softening can be achieved perhaps by dispensing with:
* fixed rate lending
* 0% bound on interest rates
* short-selling and speculation
Also it occurs to me that if a favour biased ripple were successful,
people would likely bolt their own bits onto it that would harden
obligations anyway. Even favours get called in eventually and
presumably the calling in of a favour at one node could start
hardening other obligations along the chain. Certainly if I had
favours denominated in dollars or pounds via ripple I would want to
perform some analysis on the maturity (whether implicit or explicit)
of my liabilities and assets to make sure I wasn't going to dissapoint
my connections.
This was my point about money in ripple really - there's nothing like
having a few zero maturity claims on others on your balance sheet to
make you feel a bit more secure when it comes to your liabilities.
danny jpw
> The Ripple client software could be made to produce a report at any
> point, I suppose.
be valid under FASB.
> I don't know. The USA has done quite well by having other people hang
> on to their debt. The main problem seems to be if it gets out of
> balance in the long run, which is the fault of all parties involved...
currency. Its fine until people stop deciding to roll over your debt.
Translated to shop-example terms, the shop can provide currency/
liquidity to its community only until a hypermarket gets built up the
road, after which the shop is left with large debts it can't repay in
full.
> This line of thinking usually makes me wonder if Ripple wouldn't
> degrade into a giant forex casino.
Transacting with nodes far removed from yours would presumably become
prohibitively expensive to do on a regular basis so I think the
networked nature of ripple would localise speculation.
> I should probably talk about this more, because my heart is in that
> direction far more than the original hard-obligation-style Ripple, and
> it makes implementation so much simpler, by relying even more on
> established organic trust.
lending is unsecured there is a limit to how actually hard obligations
can be.
One question that comes to mind is whether the ripple routing algo can
objectively assess the credit worthyness of each node somehow and
route on that basis.
If it could and there are fees associated with each node in a payment
chain that creates a great incentive to keep yourself honest.
> I can imagine situations where two nodes that may not even be
> connected end up fighting back and forth to clear debt they would each
> rather have in a different form, but that clearing one's undesirable
> holdings ends up reinstating the other's.
Ultimately that cycle can be broken by having the nodes being the
target of liquidity preference simply lower their rates or impose
higher fees.
> Yes, these are good questions. The logical conclusion is that
> hard-obligation Ripple must become more than a payment routing system,
> it must be a generic debt-clearing marketplace, with buy and sell
> orders, options, and all the rest of it. If we could get enough
> people participating, this kind of grand-unified debt exchange might
> very well be useful, and probably also quite dangerous. I'm
> personally not interested in pursuing this direction, but I'd be
> interested to watch if someone was...
dangerous?
Ryan Fugger
>> Personally I think that
>> rigorous tracking of soft obligations can even be a bad thing.
>>
>
> Yes that is what I was thinking. Is it the aim to:
>
> 1) monetise the favour economy
> 2) favourise the monetary economy
> 3) decentralise the monetary economy
>
> the dev wiki, project pages and ripple.com made me assume you were
> after (3) ryan.
I was, a few years ago. That goal, though, came out of an original
goal of eliminating the conflict between the things we do because we
love to do them, and the things we do because we must make money.
This is the true goal for me. Decentralizing the monetary system is
only useful for me to the extent that it serves that greater goal.
> I'm not sure its possible to monetise the favour
> economy, and even if you can whether it is a good idea - it might just
> create more demand for bank debt. No doubt bernanke would thank you
> for that!
>
One reason it's commonplace to do favours for friends, is that it's
easy to remember an informal "favour balance" with those around you.
My idea is to extend the circle where it is easy to do favours for
people by having a system that makes it easy to acknowledge these
favours, account for them via a common friend, and use the goodwill
with that friend to encourage repayment, all without having users
actually think about numerical balances (unless they really want to).
This system of soft obligations really just supports a community
network of people doing things they love for other people, on their
own terms.
> Perhaps a good target is to go for decentralisation and softening of
> the monetary economy. Decentralisation is what you have already in
> ripple, and softening can be achieved perhaps by dispensing with:
>
> * fixed rate lending
> * 0% bound on interest rates
> * short-selling and speculation
>
> Also it occurs to me that if a favour biased ripple were successful,
> people would likely bolt their own bits onto it that would harden
> obligations anyway. Even favours get called in eventually and
> presumably the calling in of a favour at one node could start
> hardening other obligations along the chain. Certainly if I had
> favours denominated in dollars or pounds via ripple I would want to
> perform some analysis on the maturity (whether implicit or explicit)
> of my liabilities and assets to make sure I wasn't going to dissapoint
> my connections.
>
I intend to design the system so that users don't even need to know
they're doing Ripple underneath. They are just presented with
community requests that they might fulfill, and hopefully feel that
they are getting enough of their own requests fulfilled to make it
worthwhile. All Ripple does is inform the ordering of requests, and
there would probably be some kind of colour-coding -- green for
requests that settle obligations, yellow for requests that don't
change the value of outstanding obligations relative to credit limits
too much, and red for requests that create new obligations.
It's really just a bit of structure to limit free-riding in what is
essentially a gift economy, making it easier to participate when there
are more people involved than our monkey brains can keep track of
individually.
Most of the tricky bits of hard-obligation Ripple just don't come into
play. For example, in soft-obligation Ripple, credit limits are soft
as well, meaning that they aren't hard limits on obligations, but just
indicators of levels of trust that tell the system how important it is
to prioritize request to clear that balance. If the balance goes
higher than the "limit", then it just becomes that much more important
to reduce that balance.
I originally thought that this system would get people using Ripple,
and then they could transition to a hard-obligation Ripple once they
got used to the idea. Now I think the soft obligation Ripple might be
nice on its own as a complement to the regular monetary economy, and
hopefully help people earn more regular money doing things they love
by creating connections for people.
Sure, it would be nice if a hard-obligation Ripple could replace the
regular monetary system, but I burnt out a few years ago pursuing
that, and I'm still not ready to start pushing that stone again...
I'm happy to help anyone who wants to though :)
Ryan
danny jpw
> One reason it's commonplace to do favours for friends, is that it's
> easy to remember an informal "favour balance" with those around you.
> My idea is to extend the circle where it is easy to do favours for
> people by having a system that makes it easy to acknowledge these
> favours, account for them via a common friend, and use the goodwill
> with that friend to encourage repayment, all without having users
> actually think about numerical balances (unless they really want to).
>
> This system of soft obligations really just supports a community
> network of people doing things they love for other people, on their
> own terms.
> I intend to design the system so that users don't even need to know
> they're doing Ripple underneath. They are just presented with
> community requests that they might fulfill, and hopefully feel that
> they are getting enough of their own requests fulfilled to make it
> worthwhile. All Ripple does is inform the ordering of requests, and
> there would probably be some kind of colour-coding -- green for
> requests that settle obligations, yellow for requests that don't
> change the value of outstanding obligations relative to credit limits
> too much, and red for requests that create new obligations.
fluffy ... and not like money.
Would make a nice extension to ebay for small ticket items.
> Sure, it would be nice if a hard-obligation Ripple could replace the
> regular monetary system, but I burnt out a few years ago pursuing
> that, and I'm still not ready to start pushing that stone again...
> I'm happy to help anyone who wants to though :)
able to quickly post say the top 10 big problems that seemed
insurmountable (or at least, took ripple too far from the goals you
initially had for it)? Just 10 bullet points would be great...
danny jpw
that touches on many aspects of my initial thesis in this thread, but
also offers the possibility that in some cases no outside money might
be required. I've pasted in the conclusions from this paper here, and
edited in a few notes that relate it back to my original points, and
too ripple. Its worth a read in full.
Inside and Outside Money
Ricardo Lagos
Federal Reserve Bank of Minneapolis
and New York University
Inside money is private debt that also circulates as a tangible medium
of exchange. Thus, an
economy with inside money must perform a delicate balancing act: On
the one hand, it must have
enough commitment or enforcement for credit to be feasible, but at the
same time, credit must not
function too well, for otherwise a tangible medium of exchange would
be inessential. For
example, Kocherlakota (1998) shows that a tangible medium of exchange
is not essential if
agents can commit to future actions or if their trading histories are
public. Starting from this
observation, Cavalcanti and Wallace (1999a) consider an environment
where trading histories are
public for a subset of agents but private for the rest, and show that
a social optimum requires note
issue by those agents with public trading histories(*1). In addition,
those notes are in turn used in
trade among the agents whose trading histories are private. Thus, in
their environment, an
optimum requires inside money.
Kiyotaki and Moore (2002) instead consider an environment where
everyone is anonymous, and
emphasize the importance of the agents’ ability to make bilateral and
multilateral commitments.
The degree of (bilateral) commitment a borrower can make to an initial
lender when selling a
paper claim places a bound on the entire stock of private debt(*2).
The degree of (multilateral) commitment a borrower can make to repay
any bearer determines the extent to which the
borrower’s debt can circulate in equilibrium(*3). Kiyotaki and Moore
find that only outside money
circulates in economies with very low degrees of bilateral
commitment(*4). For higher but still low
degrees of bilateral commitment, outside and inside money circulate
alongside in equilibrium(*5). For
yet higher degrees, only inside money circulates, and when the agents’
ability to make bilateral
commitments is large enough, the economy can manage without any money,
inside or outside(*6).
(*1) this was my point about some socially trusted entity with open
history (e.g. the village shop) needing to issue a medium of exchange
to make a credit network function optimally.
(*2) this is equivalent to the total credit capacity of a ripple
network
(*3) this was my point about digital coins - the ability for debt
issued by a 'reserve currency issuer' to circulate anonymously
(*4) e.g. middle ages metal currency, or for example only $US now in
zimbabwe
(*5) This is where we are now, the anglo-saxon banking model with fiat
money
(*6) This is what ripple could be (i.e. a pure trust/credit network),
as long as links have the ability to create suitable commitments to
one another, and (according to this paper) publish their entire
trading history to their trusted connections.
If the above didn't make sense you may need to read the paper from
scratch (its quite short).
it's here:
rl561a.googlepages.com/SR374.pdf
Ryan Fugger
> do you have notes on what caused you to give up? If not, would you be
> able to quickly post say the top 10 big problems that seemed
> insurmountable (or at least, took ripple too far from the goals you
> initially had for it)? Just 10 bullet points would be great...
>
It probably only needs one bullet point: lack of customers. I'm
pretty much done with putting something out there to see who will use
it -- Ripplepay was enough. I would need to find a group of private
currency issuers who have active users of their currencies, and have
cobbled together some kind of ad-hoc exchange mechanism that is
failing to scale for them. Then we could work together to build a
Ripple system that works for their needs, and go from there. Until
then, it's a lot of pie in the sky I think.
A few years ago Mats, Thomas, Jevgenij, and I did some outreach to
potential customers, and didn't find any promising leads, so we
stopped actively searching.
Ryan
danny jpw
> It probably only needs one bullet point: lack of customers.
new, colour coded favour friendly ripple you described earlier?
> I'm pretty much done with putting something out there to see who will use
> it -- Ripplepay was enough.
the wrong market. Perhaps the alternative currency market is just too
small, for now.
matabele
and public.
Private trust (friendship) supports soft exchange. If any attempt is
made to harden the relationship (via legal or social obligation), the
relationship will itself become hard (no longer friendly.)
Public trust supports hard exchange - but is easily gamed in the way
of the current banking cartel. Members of a cartel may employ a
private trust network with one another, whilst abusing public trust by
non disclosure of information. Of course, if this deceit is unveiled,
public trust is lost and a financial crisis results.
Two entirely separate currency networks are therefore required. A soft
currency (based upon private trust), the purpose of which is to
discourage free riders; and a hard currency (based upon public trust),
the purpose of which is to enable commercial transactions between
strangers.
Ripple can easily support a soft type network, but may require a more
public disclosure model to support a hard type network.
danny jp
They are bilateral and multilateral. An example of a bilateral trust relationship is a ripple connenction or the relationship between a buyer of a corporate bond and the seller of it.
An example of a multilateral relationship is that between private individuals and a bank. Many people trust the bank to pay.
The trust relationship upon which our whole economy is based is as follows:
1) individuals trust banks but not one another, hence they clear transactions between them using a bank (banks enjoy limited multilateral commitment from individuals)
2) banks trust the government but not one another, hence they clear transactions between them using the central bank (aka government)
3) everyone trusts the government (enjoys full multilateral commitment)
belief in deposit insurance comes from (3).
the reason we use public debt as outside money is that it is easier to trust taxpayers to meet future commitments (since they can be compelled to pay) than it is a given private bank.
Dan Miller
> I would need to find a group of private currency issuers who have
> active users of their currencies, and have cobbled together some
> kind of ad-hoc exchange mechanism that is failing to scale for them.
MMO Games always come to mind when I think along these lines. I tested
a a personal ripple server with custom currencies for a little while,
when I had more free time in general and also more patience for
Django.
I don't know if this is appropriate, but here
<http://firstmetaexchange.com/> is an example of a concern that is
small enough to be agile and experimental, yet organized and
established enough to take action. I see these types of organizations
as potential adopters of the Ripple concept.
Kevin
> The trust relationship upon which our whole economy is based is as
> follows:
>
> 1) individuals trust banks but not one another, hence they clear
> transactions between them using a bank (banks enjoy limited
> multilateral commitment from individuals)
> 2) banks trust the government but not one another, hence they clear
> transactions between them using the central bank (aka government)
Quick clarification here: In the USA, the "central bank" is an odd mix
of private companies and public entities, so it is not "the government".
The Federal Reserve keeps (many) secrets from the government, and
arguably works against the government at times.
> 3) everyone trusts the government (enjoys full multilateral
> commitment)
I'll let that generalization slide :)
>
> belief in deposit insurance comes from (3).
Right, which really means in point (1) that individuals don't trust
banks (I sure don't). They do trust that the government (not the central
bank) will step up and cover reasonable losses if the banks go broke.
Kevin
Ryan Fugger
> I don't know if this is appropriate, but here
> <http://firstmetaexchange.com/> is an example of a concern that is
> small enough to be agile and experimental, yet organized and
> established enough to take action. I see these types of organizations
> as potential adopters of the Ripple concept.
>
I'm not sure there's a role for Ripple in this exchange. The exchange
is the sole intermediary here, so there's no multi-hop routing
necessary. (One way to tell if Ripple might be helpful is to ask if
multi-hop routing is involved.) If there were a bunch of these
exchanges trying to interoperate, *then* Ripple might be helpful...
Ryan
danny jp
In fact I would say that this is an ideal deployment scenario for ripple, which will allow people already having buddy relationships in the virtual worlds to establish their own financial community. The somewhat non-real nature of MMO gaming currencies should also mitigate against the lack of hard obligation functionality in ripple.
A trust network of virtual avatars exchanging virtual and real CREDIT is a wonderful showcase for ripple - showing how credit creation can be applied to these virtual worlds. If successful then there is no reason why ripple couldn't propagate throughout these various gaming communities.
I can see the headline now: Second Life and warcraft invaded by virtual bankers!
danny
Ryan Fugger
scenario? How would people use it? I don't think I'm clear on what
you're proposing. Thanks.
Ryan
danny jp
Danny, could you explain more about how Ripple would work in this
scenario? How would people use it? I don't think I'm clear on what
you're proposing. Thanks.
FMX is a virtual currency exchange. People can bid to buy and sell various virtual currencies.
They can upload funds of whatever denomination (real dollars or virtual currency) using their email or their avatar from secondlife, frenzoo or whatever.
The about page describes what firstMeta do:
"First Meta provides financial and payment services based in virtual currencies used in virtual worlds, social networks and MMOs.
We believe that people living in virtual worlds need, and deserve, real financial services too. This is why at First Meta, we have real people, with real financial knowledge and expertise, to fulfill that very real need.
First Meta works closely with publishers and developers of virtual worlds and MMOs to provide users with the best experience possible, at the same time, helping them more effectively monetize their in-world economies."
However all FirstMeta offer is a clearing house for foreign currency transactions. They do not offer credit or loans in second life currency. This is where ripple comes in. Now I have one ripple node per avatar I have in these various worlds. My secondlife node is connected only to my secondlife buddies. my frenzoo node is connected only to my frenzoo buddies. All my own nodes are connected to each other with infinite credit limits.I am now in a position to be a virtual banker using FMX as the clearing house for settling debts outside ripple. Later as more MMOs get added, I could borrow the 15,000 golden orcs I need to buy the magic sword I'm after, if I'm connected to a node that transacts in golden orcs which has extended me a credit limit of 15K.
Of course this lets one 'leverage up' in virtual worlds to progress more quickly. Presumably if my application of the magic sword is successful and I slay many orcs, I'lll easily earn back the 15K golden orcs I need to repay the loan.
An additional innovation would be bi-lateral repos between non trusted nodes - I ought to be able to post some real estate in second life as collateral for the 15k golden orcs, on the basis that I will agree to repurchase the land in second life when my quest is complete.
danny jp
"The financial services world continues to expand in Second Life.
Singapore-based financial firm First Meta has just launched Second
Life’s first
credit card.
So, how will the new Second Life credit card work? Well, pretty much like a regular credit card. You can use the card to buy goods in Linden dollars at participating merchants. Just like in real life, you also have a credit limit.
To pay using the virtual card, Second Life residents just need to right click on the item being purchased and then click on 'MetaCard'.
Cardholders can also earn reward points by using the virtual card for purchases. Customers can also use the card to get a "cash advance" from First Meta ATMs in Second Life.
Users will receive statements at the end of each billing cycle
showing the minimum payment required. The monthly fee for a card is 300
Linden dollars. Card balances and fees can be paid at First Meta's
Second Life ATMs.
First Meta deals exclusively with virtual worlds, but a number of
real-world financial services have already opened branches in Second
Life.
Germany’s Wirecard Bank moved into Second Life this past May. Earlier in the year, Denmark’s online investment bank Saxo, and Dutch banking groups ING also entered the virtual Second Life world, joining existing resident ABN Amro.
There is also a stock market in Second Life, called World Stock Exchange, which enables virtual companies to raise capital and allows investors to buy stock using the fictional Linden Dollar and World Internet Currency that can then be sold for real US Dollars. It is operated by Australian entrepreneur Luke Connell."
Now, you can even leave aside FMX and posit a ripple network existing entirely in secondlife that competes with FirstMeta. FirstMeta charges 300 lindendollars for their secondlife creditcard. Presumably they charge interest too.
Is there any reason why you couldn't setup advertising for ripple in secondlife, for example?
An API to ripplepay.com and secondlife might be possible and would create an actual real network of users to obtain critical mass which can then spill into the real world.
danny jp
Now he advertises loans and deposit taking services to denizens of all these virtual worlds and social networks affiliated with FMX. To deposit you connect to the vBanker in ripple and extend him a small credit limit representing your desired deposit. To apply for a loan, you connect to the vBanker in ripple and he initially assigns you a 0 credit limit.
When the vBanker has sufficient loan customers he takes up the credit extended to him by deposit customers by using FMX to transfer their currency to him. Now, a deposit has been made, and the vBanker initiates a payment of these funds to his borrowers. Ripple records the debts he owes to his deposit customers and the debts his borrowers owe him. To his deposit customers he promises to clear his debt with them on demand via FMX. He can insure himself against bad borrowers and borrowers of bad faith by requiring that the borrower obtain a reference from another ripple user acting as an insurer who deposits funds with the vBanker to cover a certain fraction of his contracts as per TrustDavis.
While ripple doesn't currently set maturity limits on connections the vBanker can run his own bookkeeping software for doing this.
Having set himself up like this the vBanker can now be a critical ripple clearing node such that all parties using FMX traded currencies can transact without actually ever using FMX except when they have reached their credit limits with the vBanker.
FMX is like the central bank (or perhaps more analogous to the bank of international settlements) and the vBanker is like a commercial bank.
Dan Miller
By accepting ripple obligations, the exchange gains more customers,
both from ripple enthusiasts and from virtual currency users who have
ripple credit to spend but not currency. These new customers are
trades and commissions the exchange would otherwise not have. This
benefits the ripple project by expanding adoption of the idea. Game
players may not be willing to extend "money" credit to a connection,
but they may be happy to loan "game money". With Ripplepay or a
similar system, players can participate in the ripple concept with a
form of value they are comfortable with.
danny jp
the banking model I outlined simply extends that basic peer-peer model to provide for more intermediation, liquidity and, god forbid, maturity transformation and interest payments in the virtual currencies.
Ryan Fugger
virtual gaming currencies. Seems like a good idea. I think to really
be useful for this, Ripplepay would need the following additional
features:
* ability to define custom currency units
* ability to define custom exchange rates between currency units
My questions:
* Are the trust connections in-game strong enough to bear these kinds
of obligations?
* Are there enough of these strong trust connections in-game to create
a useful network? (You don't need Ripple to lend virtual currency to
your buddy -- just do it and write it down somewhere. The usefulness
of Ripple comes when the network gets several layers deep, allowing
you to deal with strangers as you would with friends.)
* How would we get enough gamers signed up to Ripple[pay] to create a
useful network?
* Are in-game marketplace/communication functions flexible enough for
would-be borrowers to find lenders who might be friends-of-friends
in-game?
Ryan
On Fri, Aug 6, 2010 at 2:49 AM, danny jp <dann...@gmail.com> wrote:
> However all FirstMeta offer is a clearing house for foreign currency
> transactions. They do not offer credit or loans in second life currency.
> This is where ripple comes in. Now I have one ripple node per avatar I have
> in these various worlds. My secondlife node is connected only to my
> secondlife buddies. my frenzoo node is connected only to my frenzoo buddies.
> All my own nodes are connected to each other with infinite credit limits.
>
Actually, you would really only need one Ripple node for this, but
that's a technicality.
> I am now in a position to be a virtual banker using FMX as the clearing
> house for settling debts outside ripple. Later as more MMOs get added, I
> could borrow the 15,000 golden orcs I need to buy the magic sword I'm after,
> if I'm connected to a node that transacts in golden orcs which has extended
> me a credit limit of 15K.
>
> Of course this lets one 'leverage up' in virtual worlds to progress more
> quickly. Presumably if my application of the magic sword is successful and I
> slay many orcs, I'lll easily earn back the 15K golden orcs I need to repay
> the loan.
>
> An additional innovation would be bi-lateral repos between non trusted nodes
> - I ought to be able to post some real estate in second life as collateral
> for the 15k golden orcs, on the basis that I will agree to repurchase the
> land in second life when my quest is complete.
>
danny jpw
> virtual gaming currencies.
outside currencies, who knows where it could lead...
> * ability to define custom currency units
by reference to the clearing house (FMX) or issuer.
perhaps you mean hyprid currencies (e.g. a basket of liquid virtual
currencies), like SDR for gamnig.
> * ability to define custom exchange rates between currency units
would refer to this.
I think it would be useful for ripple users to be able to specify a
discount or premium for a particular virtual currency on a given
connection, but this would be referenced to the prevailing spot rate
on FMX.
>
> * Are the trust connections in-game strong enough to bear these kinds
> of obligations?
burden is lower. after all, its a game!
this is why I feel its an ideal target for 'soft obligation' ripple.
It would provide invaluable feedback for progress toward hard
obligation ripple, which I still think is worth persuing.
there is no reason I can see why gaming relationships that develop
trust can't be slowly carried into the world of real money.
of course the main issue is that of virtual identities - but this is
not really a ripple problem.
> * Are there enough of these strong trust connections in-game to create
> a useful network? (You don't need Ripple to lend virtual currency to
> your buddy -- just do it and write it down somewhere. The usefulness
> of Ripple comes when the network gets several layers deep, allowing
> you to deal with strangers as you would with friends.)
run the gaming communities. Gaming is a solitary persuit. But the
introduction of something like ripple has the potential to be the
cement that bridges the divide between the virtual and the real, with
the value being carried between them the interpersonal relationships
developed in the virtual world and carried to a real world ripple
connection denominated in dollars.
having said that, I do feel ripple is in serious need of some outbound
marketing collateral for such a pitch to be successful. The wiki and
FAQ seem to be wildly out of date since they position ripple as a hard
obligation system, which is what prompted my original post.
on the upside, many of these virtual communities are flounding now
after the past hype, and casting around for a new business model and
new sense or purpose. SecondLife is a case in point.
> * How would we get enough gamers signed up to Ripple[pay] to create a
> useful network?
managment and they see the advantages for them then this may well be
free. Same goes for social network currencies.
* if you wanted to persue this you need to do some market research on
existing gaming communities:
- which communities are trying to develop currencies
- of those that are, what is their market penetration? How many users
do they have?
- of those that are, which have the highest turnover of virtual
currency. ripple will be most attractive to those communities
suffering from low turnover or excess demand for a scarce amount of
currency.
> * Are in-game marketplace/communication functions flexible enough for
> would-be borrowers to find lenders who might be friends-of-friends
> in-game?
Jeffrey Cliff
(Personally, I use negative interest in some accounts -- ie a loan that is slowly forgiven with time, but I also could use higher if it made sense to do so...it really depends on how closely you follow existing currencies with your debt's value imho)
jeff
danny jp
Its odd that people designing a system that allows people to establish their own trust relationships would want to put limits on how consenting adults manage their credit relationships with each other!
danny
Dan Miller
> FirstMeta charges 300 lindendollars for their secondlife creditcard.
> Presumably they charge interest too.
I just want to clarify that First Meta has stopped credit card
operations several months ago to focus solely on the exchange. Of
course that doesn't affect any of the ideas discussed, but I still
wanted to clarify the point.
danny jp
--
Vitaly
money as commodity has killed millions and developing a new system
incorporating this feature is ... gruesome. Demurrage is a very
effective method of stimulating money flow. I can't make enough stress
on how important this is. Please, heed...
Thomas Hartman
social values in technology.
You could have a non-interest version of ripple. You could have a
"child safe" web browser that filters violent and sexual content.
But the ultimate say on how to use these tools falls to the (all too)
human user.
> --
> You received this message because you are subscribed to the Google Groups "Ripple users" group.
> To post to this group, send email to rippl...@googlegroups.com.
> To unsubscribe from this group, send email to rippleusers...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/rippleusers?hl=en.
>
>
--
Thomas Hartman
Chief Technical Officer
MarketPsy Capital, LLC
2400 Broadway, Suite 220
Santa Monica, CA, USA
______________________
This e-mail message and any attachments may contain information that
is confidential, proprietary or privileged, and is for the named
person’s use only. No confidentiality or privilege is waived or lost
by any mistransmission. If you are not the intended recipient or
agent responsible for delivering it to the intended recipient, you are
hereby notified that any dissemination, distribution, copying or other
use of this message or its attachments is strictly prohibited. If you
have received this message in error, please notify us immediately and
destroy all copies of this message and attachments without disclosing
the contents to anyone. Neither this message nor any information
included herein constitute an offer to sell or the solicitation of an
offer to buy any securities or investment product, or legal, tax,
accounting, or investment advice.
The fees and expenses charged in connection with the MarketPsy Long
Short Fund LP ("the Investment") may be higher than the fees and
expenses of other investment alternatives and may offset profits. No
assurance can be given that the investment objective will be achieved
or that an investor will receive a return of all of her/his
investment. Investment results may vary substantially over any given
time period. An investment in this Fund is speculative and involves a
high degree of risk. The performance data represents the performance
of the Fund as calculated by the management company. The performance
figures include the reinvestment of any dividends and other earnings
as appropriate. Past performance is not necessarily indicative of
future results and there is always the possibility of loss.