1994-08-22 - Re: In Search of Genuine DigiCash

Header Data

From: Jason W Solinsky <solman@MIT.EDU>
To: rah@shipwright.com (Robert Hettinga)
Message Hash: 77f8c5bc7314985c22c23ac3d0fc76670853ca53392aff66f79b83c21a19c37a
Message ID: <9408220336.AA09238@ua.MIT.EDU>
Reply To: <199408220047.UAA24562@zork.tiac.net>
UTC Datetime: 1994-08-22 03:36:57 UTC
Raw Date: Sun, 21 Aug 94 20:36:57 PDT

Raw message

From: Jason W Solinsky <solman@MIT.EDU>
Date: Sun, 21 Aug 94 20:36:57 PDT
To: rah@shipwright.com (Robert Hettinga)
Subject: Re: In Search of Genuine DigiCash
In-Reply-To: <199408220047.UAA24562@zork.tiac.net>
Message-ID: <9408220336.AA09238@ua.MIT.EDU>
MIME-Version: 1.0
Content-Type: text/plain

> >Let's see, I'm selling spindles for $2.59 and you come up with a piece of
> >ecash you bought ten years ago for $1.00, which is now worth $2.59, and I
> >sell my spindle to you for it.  I deposit the cash in the bank and it's 
> >$2.59.  Now who isn't this fair to?  How is it different from you putting
> >$1.00 into your interest-bearing checking account ten years ago and writing
> >me a check for $2.59 today, the amount your $1.00 grew to?
> The problem is, you have to price the cash before you use it to buy
> something, and then you and the seller has to agree that that's the value
> of it. To do that, you or the person you're offering the cash to need to
> somehow communicate with the underwriter, thus destroying the anonymity of
> the cash transaction, and also increasing it by the communication costs,
> and creating an on-line cash system when we wanted an off-line one.

WHAT?!? If I want to sell some stock and I want to figure out how much it is
worth, I go to the Bloomberg in the Sloan basement and get a 15 minute
delayed quote. If I want to buy something in Mexico with dollars, I look at
the exchange rate in the bank or in my hotel. If I want to buy something
in digicash, I check the exchange rate, and then I conduct the transaction.
Where is the problem here?  

> Of
> course, the issuer could publish the prices based on the compounded
> interest accrued *for each certificate*, for the time period it's
> outstanding, possibly complete with the compounding factors for each
> compounding period used. (a day, a month, a year, or even continuous over
> the life of the instrument)  Lot of overhead there, but mutual funds do it
> all the time. You'd want to just take their word for it, I suppose, and
> trust their price, right?

OK, I see the problem. You are assuming that certificates will be issued
at a consistent set of notional values. (like ten bucks, five bucks ect.)
The correct way to do things [:-] is to set the notional value of new
certificates based on the trading value of old certificates. Suppose the
first certificate had a principal of $10 and is now worth $11, then the
new certificates that I issue will have their principal adjusted so that
including the effect of interest rates, my new certificate is worth as much
as your old certificate. Thus, there is only ONE value that needs to be
published at any given time.

> There's nothing awful about keeping the interest, folks. (Unless you're a
> moslem, of course :-) )  It's really just a type of liquidity premium paid
> to the underwriter to offset whatever risk (business risk, and legal risk
> at this point) taken to issue e$ for use in internet commerce.  As more
> people get into internet commerce and underwriting it, then the premium
> goes down because the risk goes down.

Seting prices based on convenience instead of value derived? *BLECH*. That
sort of thing is anathema to free markets.