A standard for STP
02 June 2006 | 8238 views | 0
Source: Ian Benn, Misys Banking Systems
Does Unifi hold the answer to the standards muddle asks Ian Benn, chief marketing officer of Misys Banking Systems
Since the 1970s, we have been working on ways to free our working world from paper, automate the mundane and fast track our evolutionary way from office worker ant to domestic couch-potato.
In the banking world, information free-flow is defined by three little letters – S-T-P. Straight-through processing will mean that banks and customers, banks and banks and - like it or not - customers and other customers, can transact seamlessly, safely and free of human intervention.
Standards bodies have been alive for decades and new industry groups still peep up through the soil like new shoots in springtime, yet the smooth flow of financial information seems to get harder, not easier.
Do you ever sit in a meeting and wonder what someone from outside the industry would make of the conversation? Last week I was in just such a session and in occurred to me that as terms like Twist, Omgeo, Target2, Sepa, Chips, CLS, Fed, Swift, IFX, XBRL, MDDL, FIX, RIXML, FpML and Unifi glided by, our outsider might be wondering what language was being spoken let alone what the meeting was about.
And then it struck me: the biggest obstacle to standardisation is....standards. Really. Think about it: each of these different initiatives, offerings and regulations has been set up with a perfectly sensible mandate. Of course we should have a single European payments approach – it is good for trade flows and good for fairness and competition. But by presenting difficult systems design challenges often with ambitious deadlines, a bank has to constantly re-prioritise resources from one initiative to another. Each of these standards is sensible, each viable and each widely supported. The problem lies in trying to meet the mix of demands, timescales and expectations.
One bank I know used to see every new release of Swift standards as a six-month project. Given that these come along at least once a year, half the year was gone in keeping systems up to date before they could even think about the next demand. And it is rare for a standards group to voluntarily disband once it has published its particular approach. Normally there is a set of revisions, versions two, three and four and a whole lot more. This means that whilst new generations are entering the ecosystem, the older generation stubbornly refuses to die off. Population explosion.
What’s the answer? Two phases. The first is to structure IT systems such that messaging and transaction flow is harmonised and separate from core banking applications, insulating the main processing engines from todays standards problems. There are many message management systems on the market – including ours, Misys Message Manager. Once implemented, they should mean that every new standard can be implemented once and flow out to every system in the bank (The bank that was spending six months implementing changes before is now a Misys Message Manager user and handles them in a few days).
The second is to create a standard for standards. Imagine if every new organisation who wanted to improve information flows had first to submit their approach, requirements and timescales to a single standards group which could work out whether these conflicted with anyone else’s group. In an attempt to solve this problem, ISO proposes ONE single standardisation approach, a common ‘recipe’ - which includes a common development methodology, a common process, a common registration body and a common repository - to be used by ALL financial standards initiatives.
This recipe is called ISO 20022 or Unifi – ‘UNIversal Financial Industry message scheme’.
If all financial standards initiatives used this recipe, we could avoid duplication of effort and conflicting demands and our evolutionary path to free-flowing information would have fewer forks in it. A life of paperless leisure awaits.