Bank IT costs set for material increase

http://www.finextra.com/News/FullStory.aspx?newsitemid=25196

Banks operating ageing IT platforms can expect to experience a material increase in running costs over the coming years, only partially offset by the introduction of smaller cheaper-to-run branch outlets.

Analysts at Deutsche Bank Research say most big bank core systems rely on too many applications patched too many times to cope with rising transaction volumes, regulatory change and digital channel changes in particular.

“This will drive up to a 10% increase in overall operating costs and will be encouraged by an exponential increase in regulatory attention given to operational IT risk,” says the DB Research report.

The rise in IT spend will be offset by a five per cent decline in branch costs, as banks move to introduce smaller banks with fewer teller staff.

An interesting post, I’d love to read the report. What I think we have seen across many enterprises and market places is continued short term remediation, fixing things point in time but never going back to do the analysis both in terms of technology and process to consolidate processes, applications and the underlying infrastructure.

Application consolidation empowers you to then build upon that as a grounds to reduce business processes, thereby reducing cost and lead times to the customer. If we reduce the components involved in the transactions we can offer the client a unified presence regardless of the medium of distribution, mobile web or terminal based.

System complexity combined with ‘enterprise’ grade approaches not to mention legal and regulatory requirements will impact costs, what companies need to do is embrace the platform that they have chosen and go with it.

For commodity x86:

• Replace the hardware every 18 months
• Strip out the over engineering – RAID 1+0 for local storage, no redundant or backup disks. Use on-board fibre/NIC connections standardise the specifications, its cheaper to ‘over order’ and swap out legacy than it is to marginal purchase.
• Switch to warranty support and have your own own gold stock – which might just be a few spare servers
• Introduce redundancy at the app or network layer

For midrange or mainframe:

• Consolidate work loads and applications – less is more
• Choose platforms optimized for that workload either in terms of cost, licensing cost or ability to consolidate on less more powerful servers
• Reduce the number of single instance databases, middleware and messaging systems running at 12% utilisation

This site uses Akismet to reduce spam. Learn how your comment data is processed.