Taurus & Crest case - why did Taurus fail and Crest succeed?

If a stock exchange is to function well, then having the right technology in place essential. For the London Stock Exchange (LSE), the largest stock exchange in Europe, implementing new technology has proven to be problematic over the years.

The LSE has had somewhat contrasting fortunes in selecting a computer system to handle the settlement of trading transactions. TAURUS, a well publicised technical failure set the LSE back years and cost millions. On the other hand, CREST was a fantastic success and for a time was widely regarded as the best securities processing system in the world.
(http://commentisfree.guardian.co.uk/daniel_davies/2006/09/dont_just_do_something_stand_t.html)

Why though did CREST succeed and TAURUS fail?

TAURUS was designed to tackle the back office function of the LSE. Its purpose was to handle the settlement of trading transactions in a more efficient cost effective way.

The new settlement process provided by TAURUS once live would dematerialise stock certificates, reducing stock ownership to a simple entry in a computer database, which in turn would eliminate masses of paper work and hundreds of jobs, saving the LSE millions. Unfortunately though, the project failed miserably for the reasons outlined below.

The most significant issue which led to failure was the fact that the system was designed to do too much. TAURUS had to be “all things to all men” (Head, C., (2001), “TAURUS and CREST, Failure and Success in Technology Project Management”, Henley Management College), and as a result, was overly complex. Implementing a standard system meant that standardised working was imposed on member firms, who each ensured that the system was specifically tailored to their needs. On top of this, the government produced an extensive legal document detailing masses of complex regulations with which the system had to comply. Not surprisingly, the design team were unable to get to grips with the sheer complexity of all the requirements. The project was almost certainly doomed from the beginning.

Secondly, the package solution selected required significant modification to meet the various demands referred to above. This led to further delays and increased costs, as well as an increased level of risk stemming from the use of unproven software.

Even the computer trade press publicly reported problems. 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.

When it became clear that the LSE was in serious danger of losing its position in Europe, it was decided that the TAURUS project was to be terminated. The collapse in 1993 was a catastrophic disaster for the LSE, resulting in a public inquiry which concluded that TAURUS had been fundamentally misdesigned. (http://commentisfree.guardian.co.uk/daniel_davies/2006/09/dont_just_do_something_stand_t.html).

It was crucial for the LSE to rectify the situation and move forward once again. The Bank of England decided to step in to resolve the problem. A new design team of talented civil service individuals was formed and a new system named CREST was designed. CREST was thought of as the simple solution which provided the London market with the settlement facilities required to remain competitive into the next century (http://www.practicallaw.com/7-100-4206).

In contrast to TAURUS, the CREST team adopted a minimalist approach, and used tried and tested technology. Because this system was designed by the Bank of England and deemed as optional (in terms of usage purposes), there was no requirement to consider LSE member needs – something which was detrimental to the outcome of the TAURUS project.

CREST was designed to perform just two vital business processes. Comparing this to the twenty-one specified by TAURUS highlights the clear difference in complexity between the two systems. The CREST system was similar to other systems which the design team had worked on, meaning they kept to what they knew, rather than experimenting with the unknown. This was a safer option, which paid off.

On the whole, CREST was far better managed than TAURUS was. A senior level team was put in place to overview all aspects of the project, and the main team consisted of a small number of experienced individuals, in contrast to the over crowded, less experienced and less efficient team who worked on TAURUS. Any requests to add more business functions to the CREST system were only recognised if 80% of customer representatives within a business area demanded it, as opposed to the constant adhering to member needs implicit with TAURUS.

CREST, unlike its predecessor, was a fantastic success which proved that less is definitely more. Delivered on time and on budget, this system was at one point classed as one of the best securities processing systems in the world. Fundamentally, the minimalist approach employed by CREST was the critical factor which led to its success. TAURUS, as previously stated, was too ambitious. Constant modifications combined with a high level of complexity and poor management ultimately led to its downfall.