Case Study - Taurus

Brian Varela
4/20/2010
Management Information Systems
What was Taurus?
Taurus was a project with the goal of:

Developing an automated transaction
settlement system for the London Stock
exchange

Successfully accomplish de-materialization of
all transactions
Background of Securities Business
Securities involves 2 processes:
Dealing
A trade is dealt when a broker is to buy or sell
shares for a client at a specific price

Settlement
Settlement involves the transfer of money and
shares, amending the companys register of
shareholders and either issuing and/or cancelling a
share certificate

Inefficiencies of Securities
Business (at the time)
It is no surprise that a system that can be traced
back to the 18
th
century could be inefficient

due to the large number of transactions occurring
within a day, the paper trail was becoming
unmanageable

In 1987 the antiquated paper driven system almost
collapsed under the sheer volume of trades resulting
from a rising market
Project: Taurus
The Taurus project began in 1998.

The aim was to create a simple system for
the large investment houses, which
account for over 70% of the value of
transactions on the London Stock
Exchange.
Collapse of Taurus
From the beginning there were signs of failure
which included:

An ambitious 18 month time frame

It used new and untested technology (untested due to
the time frame)

Too many different interests from organizations which
would later cause complexity in the project
Collapse of Taurus
1
st
Warning:

the original project design called Taurus 1 was abandoned due to the
broker constituency, who believed that it placed them at a commercial
disadvantage.

This was a project that could have been completed in 6 months with
tested technology

Conflicts of interest:

There were over 30 committees linked to the Taurus project, each with
its own special interest

Not on schedule:

The design stage, which was scheduled for 2 months, lasted 2 years



Final Design
Finalized by John Watson, hired by a
consultancy firm to direct the project.

The final design was a hybrid structure
comprising of over 17 alternative versions of
Taurus welded together to reflect existing
business practice.

The CEO at the time was opposed to this
solution but he was not in a position to stop it
Faults in the Final Design
The design phase was constantly being changed by
securities

It made everything more complex

Too much control was given to too many bodies and this lead to
constant changes in the software design

There was a lack of leadership and ownership was too
spread out

The best firms/people in the industry headed the Taurus project

Power and control were the variables that steered the Taurus project
into failure


Faults in the Final Design
Use of off the shelf software

Their intention was to speed up development

They used this state of the art technology and
software on 20 year old hardware which made no
sense

The different builds were constantly not matching

It was noted that the re-engineering of the software
package at the time of the failure of Taurus cost 14
million

Faults in the Final Design
The system design

There was no logic behind it

Its goal was to speed up project implementation

It was believed that by building the outward part of the
system first, and then the central architecture would
speed things up and make the interested parties
happy.

Conclusion
In mid-1991 the board granted a time extension
and extended budget, but by 1993, with no
visible sight of the finished Taurus project, it was
cancelled.

The original budget of 6 million and with
estimate cost at time of abandonment was 800
million; this was 13,200 % over budget, five
years over time and without a viable solution in
sight