From: hughes@ah.com (Eric Hughes)
Date: Sun, 28 Aug 94 23:31:55 PDT
To: cypherpunks@toad.com
Subject: In Search of Genuine DigiCash
In-Reply-To: <199408280516.BAA15355@zork.tiac.net>
Message-ID: <9408290556.AA28298@ah.com>
MIME-Version: 1.0
Content-Type: text/plain


To review, I said the following:
   >-- The financial structure matters when things go right.
   >-- The legal structure matters when things go wrong.

The reply: 
   The law and the enforcebility of agreements is what makes financial
   instruments exist.  Their behavior is a direct result of their legal
   underpinnings. 

This is absolutely false.  Both a promissory note and a bond can have
identical financial structure, but the legalities are completely
different.

   The
   financial behavior of a security can thus be predicted just by assuming the
   efficacy of the legal system they're written in.

Certainly the probability of transaction failure can be factored into
the face value and behavior of the instrument, but the actions in case
of transaction failure are not determined by how the financial
transactions around the instrument are governeed.

   If you break the law or agreements creating a market, say if people didn't
   make their margin calls and got away with it, there wouldn't be a market on
   margin for very long. 

Sure, the legal system creates the stability that allows the financial
structure to become significant.  But neither side determines the other.

   Thus, by collateralizing what you would call a
   digital banknote, you are agreeing with the person you issued it to that at
   the very least, that dollar-for-dollar, there's money to back the note up.

Well, no.  At the _very_ least, you promise that there will be money
for them when they redeem the note.  There's no necessity to make any
promise about what happens to the money in the meantime.  

Here, then, is most of the answer to the earlier pop quiz.  Promissory
notes need not be secured, whereas bonds by definition are securities.
Money paid for a promissory note might, for example, be immediately
lent out.  As long as there's money for redemption when it becomes
due, everything is OK.

In bankruptcy, secured debt is paid off entirely before unsecured debt.

   By the way, I figured out just now why this can't be called a digital bank
   note [...]   The issuing underwriter isn't anymore a bank
   than an institution offering any other piece of collateralized paper [...]

Even though the issuer need not be a bank, the phrase digital banknote
still captures most all of the intent of what these instruments are
meant to be used for.

   >Merely saying that the money sits
   >somewhere while it's in transit (which it clearly does) does not make
   >the instruments secured.

   But it does, Eric. Especially if the underwriter says at the outset that
   the money's secured (collateralized).  

You are merely _assuming_ that the digital notes are secured; you do
not seemed to have considered the possibility that they are not.

   If money isn't secured dollar for
   dollar, especially in the early stages, you get a whole mess of legal, not
   to mention financial problems.  

If I say that the notes I issuer are not secured, and yet for
convenience keep the money in 100% liquid reserves, is there a
contradiction?  No, because security is a legal issue, namely promises
to the holders of notes, and reserve structure is a financial
property, namely where the money sits for the duration of the
issuance.

   It should be possible to keep an issue of
   digital cash fully collateralized (secured) and still make money.

You are confusing here, very clearly, the promise to keep a fund in a
particular way, and actually keeping that fund in that way.  If you
undertake a legal responsibility, that will affect you financial
structure, but merely naming some financial structure does not
determine the legalities around it.

   Again, Eric, if one digital cash underwriter has to unwind a fully
   collateralized bunch of digital cash, what's the problem?  

Go do some reading.  In the case of bankruptcy, for example, the
issuer is not around anymore to do any unwinding.

   If the
   underwriter isn't fully collateralized, he's in violation of his issuance
   covenants and is likely to be sued by the trustee for the instruments, at
   the very least, long before a run on the cash started.  

Finally the hidden assumption of full collateral is revealed.  Why on
earth are you assuming that this has to be the case?  Reasoning from a
particular model about a set of properties is a good way to ensure
that you don't see all the possibilities.

Eric