From: Robert Hettinga <rah@shipwright.com>
Date: Mon, 26 Oct 1998 07:28:07 +0800
To: cryptography@c2.net
Subject: e-cash, banks, systemic risk, and financial safety in the Metro
Message-ID: <v04020a1bb25943283e11@[139.167.130.246]>
MIME-Version: 1.0
Content-Type: text/plain




--- begin forwarded text


Date: Sun, 25 Oct 1998 16:07:04 -0400 (AST)
From: Ian Grigg <iang@systemics.com>
To: dbs@philodox.com
Subject: e-cash, banks, systemic risk, and financial safety in the Metro
Cc: daveb@hyperion.com, djackson@jackson-trading.com
Reply-To: iang@systemics.com
Sender: <dbs@philodox.com>
Precedence: Bulk
List-Subscribe: <mailto:requests@philodox.com?subject=subscribe%20dbs>
X-Web-Archive: http://www.philodox.com/dbs-archive/

I've been a faithful believer in the past that banks are not
the right organisations to get involved with electronic cash.
To some extent, this has been merely a learning exercise in
order to determine the shape and composition of the ideal
organisation that should be involved;  starting off by
rejecting the status quo is always educational.  When building
electronic cash systems, it is an obvious question to ask, and
if not, your customers soon do so for you.

Over time, another face has arisen, that of defending society
from threat of the new world being monopolised by the
dinosaurs of yesteryear.  In short, the need to keep banks
out of electronic cash.

At the end of this rant, I have attached a post from Dave
Birch of Hyperion, in which he presents some evidence of how
banks are failing in the smart card area.  This I find
personally amusing, as I argued in my paper on the EMI's 1994
Opinion that smart cards were something that the banks could
get into and sensibly monopolise.  I guess that argument now
looks somewhat weak.

Whether banks should be involved in electronic cash is part
of a wider question as to whether they should be involved
in payment systems.  Such a question is not really discussed;
It is fairly to clear to me after several years of central
bank watching that much of the regulators' thinking in these
areas is muddied due to their lack of appreciation of the
key distinction between banking and payment systems.  In
private and sometimes in public, the more enlightened of the
insiders tend to agree that such a distinction doesn't really
make sense to most central bankers.

In such a void of discussion, it is insufficient to simply
state that the business of payment systems is not the business
of banking, as I have been doing for the last few years.

Making such a statement does gives rise to the conclusion
that as different businesses, payment systems and banking
should at least be separately regulated.  From any sensible
business principles, unless there is a compelling reason to
combine them, one separates them (and it should said that
the marketing benefits of securing additional customer
ownership for banks are not really compelling enough to
argue for public intervention).

Such an argument is consistent and valuable, but has not
really made the grade as far as convincing the world to
change.  We need more.  We need to collect more information
on why payment systems need to be decoupled from banking.

One argument that has arisen, which I have been recently made
aware of through discussions with Douglas Jackson, is that
the combination of payment systems and banking raises the
risk of systemic failure in the economy.

The logic runs like this.  Banks, being the places where the
money is kept, are subject to attachments by goverments;
one of the bugbears of Free Banking is that it is not stable
in the long run, being unable to survive a long-term attack
by a regulator that seeks to attach the funds managed.  Now,
leaving aside from our natural distaste of theft with
coercion, we need to identify how this makes the system
weaker.

When a regulator coerces a loan out of a bank, such as occurs
regularly in periods of systemic turmoil (LTCM, the Asian
crisis, S&L, need we go on?), then the bank that extends these
funds will be made weaker.  That is, it has somehow been
lumbered with a non-performing asset, which weakens its
balance sheet.

To see how this must be the case, consider what would happen
if the loan made was in fact a performing one.  Trivially,
any bank would take on a performing loan without the advice
of any regulator, as this can only be good business.
Therefore, any business that the regulator gets involved in
must be non-performing, and therefore must involve coercion
(or perhaps some other subsidy, but that is not considered
here).

Now, this pushes the bank in the direction of failure.  If
the bank fails, we now have two problems, being the original
failure that caused the coerced loan and the second failure
of the coercee.  It takes no learned education to see that
this process is starting to look like an epidemic.

If the bank does not fail, then all is apparently well and
good.  Regulators might argue that their coercion and the
consequent dead loss to the bank's balance sheet did in
fact pay off, in societal terms, and they simply need to be
careful to only load up a bank as far as it can support the
load.

The problem with this is that the regulators can make no
such argument.  Was it Mises who said that governments have
no special monopoly on good decisions, and therefore no
better ability to regulate than the market itself?  Sooner
or later, the regulators will make a mistake, forcing their
new found compliant bank into bankrupcy in order to bail out
the previous disaster.

Can we accept that?  We might be able to, if it was some
statistical thing - every now and then, our leaders make
a mistake, and something breaks.  Every now and then a
bank fails that did not need to, but most people shed only
crocodile tears over bank failures.

Well, no.  Unfortunately, the mere action of succeeding in
repairing the situation by coercing one good bank to bail out
another bad bank causes a reward scheme that self-perpetuates
the activity.  More simply, bad behaviour gets rewarded, and
good gets punished.

Where does this incentive lead to?  Evidently, repeated
bail-outs and bad loans across the entire banking industry,
which net out to a general weakening of the industry balance
sheet.

This industry-wide weakening to the system of banks is not
a good sign, but when it is coupled, as it causally must be
in this case, with a regulator's propensity to shift bad
onto good, and risk an occasional mistake in the act, then
the whole system is at risk from the very people sworn to
protect it.

People need banking, not banks.  Who said that?  He might
have been thinking of the above, and perhaps predicting that
we would not be able to escape from a situation of regulated
weakness in the banking sector.

Assuming we cannot avoid the above problem, and it is
historically evident that many best efforts have failed, then
we must look at containment.  Banks may be unsound;  the
economy must go on.  And herein lies the nub of Jackson's
argument.  What better way to spread failures than through
payment systems?

Payment systems by their nature link banks into a system of
banking.  Better than that, or worse as the case may be,
payment systems are generally offered to and used by many
other sectors of the economy.  Such a powerful invention as
the payment system, or money as the economists like to call
it, will be made universally available.

Their power is a double-edged sword.  Failure of a payment
system is capable of crippling economies, toppling governments
and causing widespread destruction of wealth.  Russia today,
Albania yesterday, and the Weimar Republic in yesteryear are
powerful reminders of what happens when the payment system
fails.

Clearly, to then reserve payment systems to banks, which are
fundamentally regulated as to be unsound, is to raise systemic
risk.  The hitherto unchallenged argument of the central banks,
that their role is to protect the system against systemic risk,
should now be presented back to them as a reason for them not
reserving payment systems for banks.

Independant payment systems, separated from the regulated,
risky and weakened world in which banks operate, must
therefore decrease systemic risk in the financial system,
ceteris paribus.



This argument is one that supports separation.  Is there
another?  And to be rigourous, is there an argument that
supports the combination of banking and payment systems?

iang



PS: follows is Dave Birch's post that catalysed this rant,
although the thoughts have been permeating for many a year.
Also note that this inclusion does not in any way evidence
Dave's endorsement of my arguments.

=======8<==========8<==========8<==========8<==========8<===
From: Dave Birch <daveb-lists@mail.hyperion.co.uk>

Frank Sudia said

>Several large organizations have undertaken consumer purse trials in the
>US, which have been miserable failures.

The only consumer purse trial I can think of is the Mondex/Visa trial in
New York. I saw a report on the web a couple of weeks ago (Business Week?
Internesia sets in). The top three places where consumers in New York
wanted to use their electronic purses were for subway tickets, taxis and
payphones: not one of these was included in the trial. I bet that you can
buy a fur coat in Bloomingdales with your Mondex card but not pay a
parking meter.

Worldwide, the situation is much therefore more interesting than your
comment suggests. I might characterise it as follows...

Several large organisations (banks) have undertaken consumer digital
money trials which haven't been runaway successes. At retail point of
sale, which is where electronic purses are being trialled, smartcards
have no competitive advantage over cash whatsoever. Why do they persist?
It's because banks treat electronic purses as cards (instead of
computers, which is what they actually are). As a consequence, they are
handled by the "card services" (or whatever) department of the bank. What
is "card services" main business? Merchant acquiring. Who do "card
services" talk to about electronic purses? The merchants that they
already acquire credit card transactions for, rather than they guys who
run parking meters, vending machines, payphones, taxi meters etc etc etc.

Other large organisations (not banks) have undertaken consumer trials
which have been great successes. Examples are

* mass transit operators (e.g. Hong Kong mass transit does more smartcard
transactions every day than Mondex and VisaCash have done in their
history) where people find smartcards far more convenient than tickets or
coins. London Transport are soon going to issue more than 7 million
contactless smartcards for people to pay for subway and bus fares: how
long before the newspaper vendors at the stations accept these cards as
well?

* campuses (in Europe and the US), where people need change for a million
reasons everyday (photocopiers, payphones etc etc) and it saves the
campus operators a lot of money to not have collect cash from machines.

I'm just as frustrated as everyone else that electronic purses aren't
catching on as quickly as I'd like. All I'd say is that purses will come
(even in America), so it's best to just take it easy and wait.

Regards,
Dave Birch.

=== mailto:daveb@hyperion.co.uk ===== http://www.hyperion.co.uk/ ===

--- end forwarded text


-----------------
Robert A. Hettinga <mailto: rah@philodox.com>
Philodox Financial Technology Evangelism <http://www.philodox.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'