Re: the cost of untracability?

[Wei Dai <weidai@eskimo.com>]

On Sat, 6 Apr 1996, rick hoselton wrote:

> It wouldn't exactly have to be timestamped.  By convention, all interest
> bearing currency could be denominated as of some fixed date.  For instance,
> its future value as of Jan 1, 2200 A.D.  The issuer could then pay interest
> without knowing the date the currency was issued.  (Of course, some accounting
> rules are probably going to need changing, hehe) Neither the payee nor the
> issuer needs to know the actual issue date when settlement time comes.
>
> When you buy a t-bill, it is worth some amount on some date.  You don't
> know when the previous owner bought it or how much (s)he paid.
>
> The denominated date could even vary if it were "blinded".
> As long as the present value of the ecoin is the same, the issuing institution
> should not care how it is expressed.  A variable interest rate scheme could
> even prevent an announced fixed rate from conveying clues about the issue date.

I think you're right.  There is no need for the issuer to pay explicit
interest.  The easiest way to eliminate signorage would be to steadily
increase the value of each denomination of ecash.  It would be kind of
like a mutual fund that doesn't pay dividends.  In fact, if the ecash is
backed by a portfolio of investment securities and its value floats with
the value of the portfolio, then it would be almost exactly like a mutual
fund.

Of course, as Jonathan Wienke pointed out, the IRS would not be very happy
about this.  Then again, the IRS would not be happy with a lot of the
technology discussed on this list.

Wei Dai