e$, Liquidity, and Economic Granularity

[rah@shipwright.com (Robert Hettinga)]

Flame-bait warning.

The last week or so, I've been trying to cobble together some business
models for electronic commerce, particularly how the banking system fits
in. I'm posting some of the more far-out stuff for discussion here.  To get
here, I've had to thrash some stuff that's probably obvious to the
cypher-gerontocracy. I beg their indulgence in advance, because some of the
stuff I'm going to yak about probably has been been said here numerous
times in prehistory.  I'd like pointers to those discussions, FAQs(?), etc.
as I couldn't find specific mention of them in the file heirarchy at soda.
As it is, I've bumped into some interesting stuff to me, and thought I'd
share... Hokay... Here goes.

One of the things I like about e$ is that strong crypto provides the
linkages to  disperse it into quite small units and still reunite quite
usefully into big batches for the purposes of financial markets. This is
necessary, because like all technology, successful new stuff usually
creates a superset of the old stuff. It has to offer the same functionality
and add a significant improvement. Here's how.

Let's start with where people keep their money. It can be said that given
the size of a bank's average retail demand deposit account, and the
interest and fees on those accounts, it might at some time behoove people
to keep their disposable money *in cash* on a personal hard drive
somewhere, probably at home, and probably backed up offsite for security.

The principal reasons that people have checking/savings accounts these days
are pretty much as follows: 1. to protect that money from theft, 2. to be
able to conveniently safely spend that money in small increments with
checks or with ATM cards, 3. (recently) to be paid through direct-deposit
mechanisms, 4. to get interest on the money while it sits there.

I think that reasons 1,2, and 3 can be taken care of with e$ protocols, and
that for most demand deposits, 4 is not meaningful because fees outweigh
interest most of the time. You might as well keep your money at home.

Like a lot of other things, retail demand deposits are largely an
industrial phenomenon.  With e$, information technology does to banking
what it did with the industrial telephone network.  A heirarchical network
is replaced with a geodesic one, and demand deposits, except as
concentrator points for large institutional cash distributions, cease to be
meaningful in an economic sense for individuals.

When people accumulate surplus money (:-)) and want to sell that money to
an entity in the financial community, the transaction can be taken care of
with automated secure transmissions of e$.

Organizational concentrations caused by efficiencies of scale would tend to
dissapate as well.  Imagine if Peter Lynch's replacement(s?) could run
Magellan as he saw fit and had all his fund concentration and distribution
activity taken care of automatically without the cost of the Fidelity
administrative armature. He'd still drive a multi-billion dollar fund. His
customers would still hold shares of Magellan. However, those customers
would be doing business with an automated digital cash transaction server,
which would take their money or redeem their digital shares of Magellan for
it's current market value following whatever redemption criteria exist in
the funds prospectus and deposit agreement.

The disbursement/concentration code's already there in Sybase and on the
Heavy Iron (yes, it's still there...).  The user interface just needs
changing. With a digital cash transaction server, there is no need to train
a cast of thousands of clean-scrubbed young econ majors to answer the
phone. (Fidelity Joke: "Camp Fido. It's a great place to work if you're
parents can afford to send you there.")  Grove's Law tells us the iron
keeps getting smaller.  So, our Lynch-analog could (in theory) have a small
cash transaction server handling his client relations while spent his time
looking out the window at Marblehead, at his Quotron-replacement, or at his
collection of Ren-n-Stimpy cartoons.

With e$, the capital markets could still operate the way they always do,
but with more functionality. The growth of communications technology
originally allowed financial information to move more quickly.  Then,
centralized information technology allowed transactions to be processed
more quickly. Finally, distributed information technology allowed decisions
to be made more quickly.  The increased functionality contributed by strong
crypto enables decision-making ability to be pushed out of investment firms
and onto the network, the same way that automated switching technology
created more more nodes in the telephone network.

Here's how that could happen. Most serious individual equity investors know
what p/e ratios and book values are, and what they mean.  Software can
allow them to understand and manipulate fixed-income concepts like
duration, convexity and total return. This means that people can do more
and more sophisticated things with their money and get better returns.

The first limitation for sophisticated individual transactions is small
transaction size. However, it's possible to see how if they're trading on
their own, investors could take "physical" delivery of e$-based investment
instruments. A person's "portfolio" could consist of various "securities"
physically resident on a storage medium that they physically control.
Because of the automation of transactions allowed by this kind of "physical
delivery", the minimum certificate sizes could come down for the most
common securities.

For uncommon securities or market strategies, it is possible to envision
the ability to anonymously concentrate large purchases of various
positions, much in the way odd-lot trades are consolidated in the equity
markets today.

The second limitation is the ability to securely communicate these
transctions with the markets. I suppose that's a straw man to those on this
list, but as you've probably guessed, this is the most important part.
Strong crypto allows you to send money and money equivalents over the
network with the confidence that it doesn't get waylayed.  Thats *real*
important for the efficient function of capital markets. ;-).

I bet that the roles of the really important players in the capital markets
won't really change much. Portfolio managers still function like editors.
They add value by synthesizing information.  The people on the sell side,
the investment bankers, securitizers, and pool-builders, etc. all still
create securities so that markets can cope with technological change in
information technology. (A charitable way of looking at *those* guys,
anyway...)

The thing that holds this all together is strong encryption and it's
various offspring, including digital cash and other forms of e$.

This crypto-stuff has a lot of really spiffy applications in finance and
financial operations. I *love* this place....




-----------------
Robert Hettinga  (rah@shipwright.com) "There is no difference between someone
Shipwright Development Corporation     who eats too little and sees Heaven and
44 Farquhar Street                       someone who drinks too much and sees
Boston, MA 02331 USA                       snakes." -- Bertrand Russell
(617) 323-7923