1994-05-23 - Kick-starting the eeconomy - my view of ecash…

Header Data

From: “gtoal@gtoal.com” <gtoal@pizzabox.demon.co.uk>
To: cypherpunks@toad.com
Message Hash: 7c84b159719576b2f3d9ee334c27bb34867f2d4327773c2383111297280fbb7d
Message ID: <9405230629.AA09270@pizzabox.demon.co.uk>
Reply To: N/A
UTC Datetime: 1994-05-23 06:35:36 UTC
Raw Date: Sun, 22 May 94 23:35:36 PDT

Raw message

From: "gtoal@gtoal.com" <gtoal@pizzabox.demon.co.uk>
Date: Sun, 22 May 94 23:35:36 PDT
To: cypherpunks@toad.com
Subject: Kick-starting the eeconomy - my view of ecash...
Message-ID: <9405230629.AA09270@pizzabox.demon.co.uk>
MIME-Version: 1.0
Content-Type: text/plain


Well, I once dived into the anon remailer thread when knowing nothing about
them, and something useful came of it, so I'll risk diving into the
electronic money argument today, and I warn you in advance I know even
less about that - it's something I haven't been following closely because
I couldn't see how anonymous cash could work without some non-anonymous
means of stopping double spending, like a central bank.

So, my conclusion, based on no research whatsover :-) is that anonymous
e-cash is probably not going to work but that a reputation-based system
might, and I'm now going to describe what I think is how a reputation-
based system would work: it's based on the exchange of personal IOUs and
it seems to get round the other problem I've seen here of how to kick-start
the eeconomy...

Firstly, there's no central money-issuer.  Let's say I am running a
service for sale - say a newsclipping service where I will scan my mailspool
and send you articles matching a set of patterns you've registered with
me (I just pick this example because it's a program I'm currently working
on), then if you want to use my service, you write me a personal eIOU for
the fee, demonimated in the currency of this scheme.  Let's call the
currency the Ob.  (SF readers might recognise it.  It stands for an
'Obligation' - a personal IOU of one unit)

So, you mint however many Obs I decide I want to charge for the service,
and we enter a contract where you transfer the Obs to me.  I can either
accept them directly, or via a third-party exchange at a central bank
which *doesn't* hold any money, but is simply a reputation server - this
is necessary when accepting new Obs from someone you've never heard of,
on a transaction for an amount greater than you are willing to risk on
an unverified transaction.

In the simple case, if, at any time in the future, I want something from
you, I give you your Obs back in exchange for the service.  However, you
*have* to be willing to exchange your Obs for cash *or* the Obs of someone
with a good reputation at the bank.  If you can't, your reputation value
at the bank goes down, but that doesn't mean people can't still do business
with you - individual people who trust you personally might still take
your minted money if they believe that you will be able to repay someday -
but total strangers most likely won't accept your personal Obs any more...
but they will still deal with you, as long as you pay in Obs from other
trustworthy people.  You might however be better off enhancing your
reputation back up by using these trustworthy Obs you've earned to buy
back your own IOUs. (Later you'll see that simply acquiring these Obs is
equivalent since it enhances your reputation and revalues your personal
IOUs, but I'm talking about the early days of this system where personal
IOUs still are treated as such and haven't morphed into a more notional
currency)

(Note that if I cash in your Obs for servives, that's straight barter, albeit
rather complex, if they've been passed round a lot, but provably barter and
therefore exempt from tax.  If I demand cash from your obs then it's income
and expenditure and one would have to pay tax on the money exchanged.)

This scheme allows a lot of money to get into circulation quickly, probably
starting with small sums exchanging between friends, until the reputation
server gets enough data to be meaningful.  I haven't worked out the details
of how a good reputation server would structure its algorithms - that's
probably a full-time PhD-level project, but it looks to me intuitively
obvious that something could be made to work.  (Note that you *could*
simply make the reputation server a strict account-balancer, weighing up
what a person has issued against what they've received, and the reputation
value being the bottom line, but I think for the scheme I describe here
to work, it has to be much more than that, and in fact I'd like to explore
the possibility here of working on reputation alone and ignoring the strict
bottom-line...)

The significance of a good reputation server is to block frauds such as
person A mints a large amount of money and exchanges it with a similar
number of Obs from person B a few times (or in practice a larger clique
of conspirators) in order to enhance the reputations of A and B as
trustworthy individuals that people do business with.

Anyway, the end result of this is that the economy takes off straight away,
possibly even with large sums involved, but restricted at first to cliquish
groups of friends who all trust each other, but slowly the cliques merge
as people accept Obs from others outside their clique - initially they
would make personal real-world reputation checks to verify that someone
could back up an Ob, but after enough transactions had happened at this
level, a pgp-like web-of-trust would build up in the reputation server
which would allow you to accept people's Obs whom you'd never met.

Of course, once the system was in place, when you did a transaction with
someone you could have the choice of insisting not on their personal Obs
as payment, but for an Ob from someone with a sound reputation.  This
*isn't* like a central bank note - there wouldn't be millions out there
under one person's name - but there *might* be millions of people with
very sound reputations.  Eventually it would get to the point as with
current currency that you no longer feel you have to trade it in for
gold (as we used to do) to be sure that the currency is still valued -
because the reputation system would give more confidence than the gold
reserves did, especially if you had a good spread of originators for
all the Obs you held.  And it would make you appreciate the riskiness
of big business at a personal level - would you want to hold a million
bucks in money backed only by Bill Gate's personal IOU?

The biggest flaw in this scheme you might have spotted already is what
to do when someone with a good reputation turns bad - or dies.  Well,
dying is relatively easy - when someone dies, notionally all the Obs they've
issued are returned to their estate to be replaced with all the Obs they've
received (ie their current bank balance held in Obs).  Any shortfall is
taken out their estate in real terms by the executors; and surplus is
given to their inheritors.  If they're insolvent and can't exchange
their Obs for those of people with good reputations, well, unfortunately
the holders of the Obs have to take the loss and their reputation is
wiped out at the bank.

*However*... what would really happen is that in general people's Obs
would be continued to be traded after their death.  The executor would
do the calculation above, to see if the person was in real debt when
they died, and if they were their reputation at the bank would be lowered
by a percentage, but not wiped out.  This would devalue the utility of
their Obs for use in large transactions but they could continue to be
used in small transactions where people didn't care too much about
the risk (eg trivial net services, like my newsclip example, which I'd
be using to generate low-level money)  Of course, people would have to
trake care never to build up a portfolio of money from too many people
of poor reputation - it would be effectively like small change - no-one
ever keeps millions in nickles and dimes, but everyone needs them for
day-to-day small transactions.

So in all likelihood after someone's death the money would simply remain
in the economy, because if their books balanced it wouldn't be necessary
to call them in.  And their books balancing would at a first approximation
be equivalent to them having a good reputation at the time of death.

The other problem I mention is the harder one; someone genuinely builds
up a good reputation, then makes some transactions and ducks out of the
system with real good paid for with their personally-minted cash which
is now valueless.  Well, we have similar problems at the moment with 
real cash - it's called counterfeiting.  Counterfeiters are criminals
and once identified have to go on the run and avoid being detected;
anyone who copped out of this system would be instantly detected:
If reputation servers would be based on real IDs, they would never be
able to get back into society, so there would be a great disincentive
to doing this in the first place, if reputation servers were based
on nyms, it would take them a *long* time to build up a new reputation
that would allow people to do business with them.  (I guess they could
sell off all the stolen goods for 'clean' Obs, but they could never
trade on their own reputation for a long time, and anyone doing business
with them would insist on reputable Obs, so as soon as the stolen money
ran out, they'd be in trouble - broke)

But that's just waffle saying the event is unlikely; in reality, it
probably would happen.  So how is it handled? I guess by trashing the
reputation of that person, which would mean that the loss from the
theft would be spread out over all the personal Obs they've ever issued -
anyone who held a lot of them would be wiped out - they wouldn't be
valueless, but they could only be used as small change and would take
a long time to offload.  Which is fitting I guess, because it's people
who took a lot of money from this person who built up the undeserved
strong reputation in the first place.  Though by the time this scheme
turned into a global large-scale economy, I guess personal integrity
checks for large transactions would be rarer - on the other hand, once
the economy gets to that size, the risk of leaving bad money in the
system lessens, as it's slowly decoupled from the original strong
concept of a personal IOU.

Well, hey, it's Science Fiction, I don't expect it all to make sense.

So what do folks think?  Is this like any of the existing schemes?  (I
don't think so - the guy who did the Ghostmark scheme seemed to have
limited the currency supply to a fixed amount and only money that
he minted could be used, although I may have misunderstood entirely)
- I'm afraid I've never been interested in this field until I had
the thoughts above last night, so I haven't read the Chaum papers
yet.  Is this the sort of scheme he proposes?  I know you guys are
gung-ho on anonymity, and a reputation-based system seems to preclude
that.  But I'm not so sure - either it could be made to work on
nyms (the reputation server software would have to be pretty slick,
but it needs to be anyway, and the input from a real-life credit
check part of the initial reputation value would become less and
less significant as time went on, when this scheme worked effectively
from birth...) or anyone who wanted anonymity *in a particular
transaction* would merely have to deal in other people's well-reputed
Obs.  In fact, once there was a good money supply going, people
would tend towards doing that anyway - issuing of a personal Ob
would have to be done *very* carefully, because it would automatically
decrease your reputation - it would be like taking out a bank loan
that *could* be called in at any moment, so you'd be sure to do it
only when you could afford to, and you'd only be *able* to when you
had a good strong reputation or could make a transaction with
someone with a strong reputation, who trusted you to pay *them*
off - in fact, this mechanism would cover both a real live electronic
bank manager, *or* a personal backed loan, like a parent's guarantee
to help you through college, for instance.

Hell, I was almost finished, but I ran off at the mouth again.  I'll stop
now until I can get some feedback...

G
PS If its an idea worth trying on a small scale, what do you say we
try it ourselves? - now we find out what our reputations are *really*
worth :-)  Would *you* sell a used car for tmp@netcom notes? :-)

Oh hell, I feel another tract coming on... here goes (sorry)...

I mentioned at the start that the scheme didn't rely on a central
bank but on a shared unit of currency.  There's no reason why there
only needs to be *one* shared unit of currency.  Long-term, globally,
yes, it could be desirable.  But in kick-starting the system there
could be several independant schemes running - for instance, the
Cypherpunk Ob, started by trading among ourselves; the Extropian Ob
for those people next door who I've never really understood what 
they're up to; the RKBA Ob shared by everyone that hasn't yet
left rec.politics.guns; the Queer Ob shared by everyone on soc.motss;
the Worthless Ob, used by the clique on alt.religion.kibology;
the Boston Ob, used geographically among people in Boston because
of physical proximity being used to kick-start the reputation server,
etc etc.  So it's reasonable to trial it here, and if the scheme
takes off, the various currencies might eventually set up an exchange
rate like national currencies, or they might merge into one - who knows...
(this is how the banking system in Scotland developed - we originally
had lots of competing banks issuing their own money)

G (*really* signing off this time...)





Thread