Origin

The origins of Taurus could be traced to the 1986 event known as the “Big Bang”, during which the London Stock Exchange (LSE) was turned into a self-regulating organization under the oversight of the Bank of England. Many regulations were lifted and ultimately the open-outcry market was replaced by electronic trading. The Big Bang led to huge trading volumes on the LSE. At that time, using the existing LSE batch trading system, settling a trade required three to six weeks even as trades were settled in the United States in three to five days and in Japan in only two days. One of the major problems for LSE was that its inefficient and inflexible system was primarily paper-based.

Concept

TAURUS (Transfer and Automated Registration of Uncertificated Stock) was a major IT project at the LSE designed to ease its back office functioning through paperless trading and computerized shareholdings. The purpose was to make the entire process cost efficient and time saving along with removing the need for face to face meetings.

TAURUS was conceptualized to be a computerized database of investors and their holdings maintained by the LSE in its role as a central operator. Stock ownership would be reduced to a simple entry in a computer database, which in turn would eliminate masses of paper work and hundreds of jobs, saving millions of pounds. It would lead to reduced costs for the share registrars (who maintain a list of holders of a company’s shares) and the institutions and brokers for future account controllers of share transactions. Total dematerialization would mean that share ownership would be stored electronically and investors would not receive share certificates. TAURUS would be linked to some 280 financial institutions, serving a range of stakeholders from registrars, brokers, market-makers, custodians to large investors. It would reduce the time taken to complete the legal and financial arrangements for the payment of shares being traded.

Though the first phase of TAURUS, known as INS (Institutional Net Settlement), was widely regarded then as a success, other parts of the project began to fall apart. The reasons for failure of a project of this size were complex.

Failures

The project selected Vista software from Vista Concepts, US, for database management. Although being very good for on-line real time processing, it could not handle distributed data processing or batch processing. Despite the online real-time processing forming the heart of the system at LSE, it decided that other processes should also be supported by the system. The LSE tried to augment the system with several high security communication packages. In the end, LSE tried to modify Vista by rewriting almost 60% of it. This resulted in the software becoming ridden by hidden bugs and long delays. The costs of adapting the Vista global custody system package, which should have cost only £1m, rocketed into the total unfinished bill of about £14m.

The most significant issue which led to failure was the fact that the system was designed to do “all things to all men”. Originally intended as a settlement only system, as time passed, it grew until it had become a full “share registration and transfer system”. Specifications were driven by a number of forces, including complex legislation and changing needs of investors. Implementing a standard system meant that standardized working was imposed on member firms, each of which ensured that the system was specifically tailored to their needs. Moreover, the government produced an extensive legal document detailing complex regulations about compliance for the system. The design team hence failed to get all the specifications together properly. The project was almost certainly doomed from the beginning.

Given that the system would have taken over the work of the share registrars, it is not surprising that the main pressure came from these same registrars. A market like a Stock Exchange is a fragile thing and exists only because of the large number of financial institutions that wish to trade within it. The smooth running of the market depends upon gaining a balance between the interests of the powerful and influential institutions that make up the market. It is impossible to propose a solution that ignored the interests of any significant part of that community, and LSE found it the hard way with the rejection of its original proposal of TAURUS.

Another reason for the failure was the outsourcing of this project. Two rival management consultancies became involved, with one running the Exchange’s day-to-day computer operations and the other’s staff trying to run TAURUS. Since total responsibility was shifted outside the LSE, no one from the Exchange’s technical staff was around to offer objective opinions or monitor the project otherwise.

A timetable produced in October 1991 by LSE compartmentalized the TAURUS project into a series of milestones, grouped by activity. These milestones were rarely hit. The same timetable document proved to come true wherever it had warned of project risks.

The development of Taurus became inflicted with many follies. Various committees formed failed to realize the full scope of the project. Communication between parties was poor, and there was no control linked to a project-wide process of evaluation. In addition, accountability was often unclear. The TAURUS project and its shortcomings became well documented, to the extent that Computer Weekly even published an extensive expose, adding to the onslaught which helped kill TAURUS off. The project was scrapped in 1993 at a cost of £800 million to the City of London.

The failure of the TAURUS project thus has elements of classic mistakes related to all of people, process, product and technology. Extensive analysis of the reasons for the failure have led to key insights into what went wrong and the best practices that would have led to avoidance of the failure. The lessons learnt have not entirely gone waste, as it helped in the successful implementation of CREST, which went “live” in 1996 to serve the need of a computerized system at LSE.

(I wrote this article with a few friends from my B-school)

If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!

Related posts:

1. Why Obama’s plan can fail …
2. IPOs – How good they are as investments?
3. Debt vs. Equity – A Tax Bias?
4. Windows 7 – Make or Break for Microsoft?
5. IPOs are no longer the same