In Search of Genuine DigiCash

Robert Hettinga rah at shipwright.com
Sat Aug 20 19:21:33 PDT 1994


At 12:20 PM 8/19/94 -0700, Eric Hughes wrote:

>A raw, non-modal "is"??  Digital cash doesn't exist yet, so saying
>that it "is" something, is, well, premature. The real question is
>"What happens if we set up a digital cash system as a callable bond?"
>
>And my answer to that is, "You really _want_ the SEC involved?"
>

I meant "is". Like a triangle, or a limit, or an asymptote, "is".  It's
okay to be non-modal here. Digital cash has to be issued by someone, who
*really should* back it up with real money, and should thus receive real
money as collateral for the digicash on the net.  Thus, there's a float.
Thus it's really a loan with a security (ecash) to prove it, with the
collateral in the bank of the issuer earning the issuer interest.  Thus
it's a bond. And since it has no maturity date, and it's not a perpetuity,
then it has an implicit call provision. Thus, it's a callable bond.
Example: A CMO is a callable bond, whether it's called one or not.  When a
tranche's principal comes in, the tranche is "called", and the investors in
that tranche are paid off. By the way, most people refer to a callable bond
as a series of options, and that's how modern portfolio analysis is done on
them... Wittgenstein would laugh.

The SEC has nothing to do with the mechanics of a security's behavior.
There are some bonds which are illegal here, but not illegal outside the
country. They're still bonds.  The obligation is held by the issuer, and
the issuer keeps the interest, which discounts their price.

Also, so what if the SEC is involved, or not?  I expect that there has to
be a test of the technology, forced by the possibility of competition from
overseas (regulatory arbitrage).  If the market test is successful, then
the SEC will not willfully restrain trade if the market's big enough (the
revolving door), and perceived to be benign enough. Frankly, I don't see
what the fuss is about, do you?  It's just a low-cost settlement mechanism
for retail transactions on the internet. ;-).

>   The issuer gets to
>   keep the interest accrued on that money while the ecash is in circulation.
>
>Perhaps in some systems this is so, but not all.  The unit of account
>must be fixed, but the unit of account may not be constant currency,
>but rather currency at a fixed interest rate.

Is "unit of account" a formal term here? Could you define it?

The problem about not keeping the interest on the float is, who do you pay
it to otherwise? If you have a truly anonymous digital cash system, you
couldn't find the original purchaser if you tried.  If you want to treat
this like a settlement problem in securities operations then you have to
track each owner's interest share for the time they held the instrument and
pay them back. Again impossible. If you pay back the accrued interest on
that specific ecash certificate to the person who "walks in the door" with
it, is it fair?

The solution is, keep the interest, use the money to fund the issuer's
operations. If that's not enough, charge exchange fees. A competitive
market will sort out who's got the most efficient operations, and thus
ecash users get ecash at its most efficient price. It's just like
insurance.  An insurance underwriter collects premiums, some portion are
direct fees for handling the transaction.  The remaining premiums are put
into a fund which accrues interest (for want of a better term).  Some or
all of that interest ends up in the insurers pocket, and the rest is held
for loss reserve (which may be itself reinsured) so the insured are paid
when calamity strikes.  It's a living.

>Why do you assume that the only source of income for the "underwriter"
>is the return on investment from the float?  Sure, that's one business
>model.  Transaction and participation fees can also be levied.

It's not really like you're quoting me out of context here, but I really
did say further on in the post you're talking about here that exchange fees
were how an issuer made up the difference between his cost of operations
and the actual return he got on the float...

>The issuer has a debt mediated by an instrument, yes.  There are,
>however, more instruments than bonds available for use.

Yes. But probably short term bonds (money markets, t-bills) are safe places
to earn higher returns than a demand deposit account.  It's all cash
management technique, which is pretty straightforward, boring stuff.

>Is the debt
>secured or unsecured?

It's secured by the cash which bought the ecash in the first place, which
can be put into secure money instruments of some sort. See the post you're
quoting from about durations, total return, etc.  If you want the issuer to
put it into a demand deposit at, say, Shawmut National here in Boston, and
let *them* invest the money in the money market, you can do that.  They'll
gladly take your money. (This is a good reason for a bank to get into the
market, in my opinion, because of this synergy.) But it doesn't take much
to manage your own portfolio of cash instruments by yourself.

>What happens during bankruptcy of the issuer?

This probably won't happen except in cases of fraud.  I expect this
business to be pretty boring. After all, you're the one with a portfolio of
(real) cash to manage. Unwinding a position in the money markets is not
really a scary proposition at all.

When an ecash bank "fails" if ever, it'll be just like the old days
(actually, not so old, really; Continental in Chicago was the last famous
big one).  The ecash banking community will circle the wagons and honor the
unfortunate's ecash. More probably the bank will be quietly merged, and no
one will know the difference.

>These and similar issues determine the nature of the instrument.

The instrument is e-cash.  It's backed up by dollars, probably money market
instruments, or maybe government bills.  There may be "brands" of ecash
which may have to charge higher and lower fees, depending on their risk. A
rating system could evolve. I bet that the differences between issuers
could be pretty marginal after a while. It's as simple and as boring as
running IBM's corporate treasury accounts.  It's just not that complex.

>   If you thought that
>   the ecash duration was 3 days and it stayed out there 3 months,
>
>It's unlikely that these sorts of figures are not going to be known
>shortly after rollout, during which phase the cash management function
>for income is much smaller.

Agreed. Pardon my hyperbolic example. I just put them out there for
illustration. Fees will be higher at first.  They might be too high to
sustain a market in the long run. There's no way to find out except to try,
which was the ultimate point of the post.

>
>   In
>   theory, if the fees are high, the money may never come back, and stay in
>   circulation forever.
>
>I think you may be getting confused here between "on-us" transactions
>and a first class currency, which does circulate.  Digital cash cannot
>"circulate forever".

That's why I said "in theory". I thought I qualified that further in the
same paragraph. Again my hyperbolic rhetorical style does me in. In the
first few pages of finance text books (I read Brealy & Meyers in 1985),
they like to talk about British securities called "perpetuities".  They are
literally perpetual bonds with no expiration date, and a few have no call
provisions at all. Whoever holds them keeps getting interest until he sells
them.  This type of asymptotic behavior was what I meant by "forever". I
forgot at the time that digicash grows every time it's exchanged, for
instance, and was simply making a point about the behavior of a system at
it's extremes. My apologies.

By the way, what does "on-us" mean?

>
>I should note, however, that I agree with the basic point, that the
>portfolio management problem for digital cash is not unusual.
>
>Eric

I'll leave this here. I feel better now.  Between Eric and Tim, I feel a
little like the gopher in the game at Chuck E. Cheese's.

Cheers,
Bob Hettinga

-----------------
Robert Hettinga  (rah at 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








More information about the cypherpunks-legacy mailing list