3 slides on Decentralization and Events

R. A. Hettinga rah@shipwright.com
Wed, 19 Jun 2002 23:06:12 -0400


--- begin forwarded text


Status:  U
Date: Wed, 19 Jun 2002 10:47:47 -0700
From: Paul Harrison <pth-02@pacbell.net>
Subject: Re: 3 slides on Decentralization and Events
To: "R. A. Hettinga" <rah@shipwright.com>
User-Agent: Microsoft-Outlook-Express-Macintosh-Edition/5.02.2022

on 6/18/02, Rhohit may have expostulated:

> A Tale of Three Markets
> How do multiple traders set prices?
> NYSE: one human, a 'specialist', aggregates all data
> NASDAQ: many 'dealers' quote their own prices
> FOREX: every trade is settled pairwise (unregulated)
>
> There is a spectrum here:
> Centralized: single immediate  arbiter in space-time
> Distributed: multiple arbiters with timeouts
> Decentralized: no arbiters at all

Three expansions:

The NYSE and NASDAQ have been augmented by the un-coordinated third market
(Instinet, Island, european market, etc.)

The NYSE and the NASDAQ have augmented their main market mechanisms
with SuperDOT (Designated Order Turnaround) and SOES (Small Order
Execution System) which are centralized alternative methods for
smaller transactions, coordinated with the specialist/dealer pricing.

The FOREX market has been augmented with multiple, independently
centralized market mechanisms (Currenex, FXAll, State Street's market).

Two observations:

The huge, highly liquid capital market is well served by multiple,
competing price discover mechanisms, some of which use the information
efficiency of centralization to advantage,  but all of which are kept more
price efficient by having competitive market makers a few electronic clicks
away.

Independent of the price discovery mechanism, these markets live on
ubiquitous price information distribution.  The Consolidated Ticker System
pushes transaction prices from nearly all markets out to everyone concerned,
and the Reuters, Bloomberg, etc. services provide redundancy and commercial
distribution.  This "information exhaust" is a critical part of any high-
functioning market.

One speculation:

Very large decentralised systems will not evolve in a pure sense, because
their stability depends on the coexistence multiple models as found in the
capital markets.  The constant will be that the component systems will
feature useful information exhaust.

A parting random shot:

A stable future for P2P file sharing networks would likely involve gnutella
(limewire), Kaaza, Morpheus, the music labels and others -- each at least
somewhat specializing in their content or service and possibly somewhat
interoperable.  The big impediment to this future is that the legal issues
effectively prevent the informal  P2P networks from providing good
information exhaust.  If it were possible to determine where fast, reliable
service could be obtained, then legal harassment would ensue.  Similarly, if
a gnutella servant began charging for downloads it would become a target.
Too bad, because the human need for music is so broad and deep that this
application could provide an interesting experiment in very large scale
decentralization.

--- end forwarded text


-- 
-----------------
R. A. Hettinga <mailto: rah@ibuc.com>
The Internet Bearer Underwriting Corporation <http://www.ibuc.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'