The strategies and technologies outlined in Banking Operations Strategies and Technologies are having a considerable impact on core banking processes. This report examines what strategies and technologies are affecting specific product processing functions within banks.
We divide these core banking processes into the following categories. Read a description of the major investment strategies and our opinions of them for each of the operational areas below:
- Lending Operations – Personal
- Lending Operations – Business
- Lending Operations – Security
- Customer Accounting
- Cheque Processing
- Other Payments Processing
- Account Maintenance
- Cash Handling
Lending Operations – Personal
For most banks, personal lending operations has been the beneficiary of substantial efficiency investment in recent years and so further break-through may not be easy. Activities that most banks have been undertaking to a greater or lesser extent include:
- Centralising the loan application processing/sanctioning activities. This facilitates:
- Call Centre/internet loan applications
- Outsourcing /offshore
- Efficiency from economies of scale
- Automating the sanctioning process via the use of on-line application forms, credit scoring, on-line links to Credit Agencies
- Automating production and posting of letters and documentation.
- Automating decision-making on excesses and automated action of consequences (e.g. automated charges taking).
In the opinion of the editors, this strategy has been/is successful in reducing costs without harming credit quality or customer service. We would welcome evidence to the contrary or in support.
Lending Operations – Business
We believe this area has lagged behind the personal sector in terms of automation and efficiency because of the emphasis on relationship banking to date. We believe this will change. Computers can beat Kasporov at chess and are reducing in price at 50% per annum; they can make good lending decisions, even for relatively complex things like businesses. Hence, the main operational design principles of personal lending will increasingly apply to business lending as well; i.e.
- Automate the credit sanctioning process using credit scoring techniques.
- Centralise and automate the monitoring process for excesses.
Probable investments that are more specific to the business market include:
- Greater use of electronic forms, filled in directly by the customer.
- Fund developments by the accounting package manufacturers (SAGE, SAP, etc) to automate the extract of statutory and management accounts into the credit monitoring software in the bank.
Lending Operations – Security
We think this really is an area that lends itself to cross industry consolidation in that it is usually only loosely connected to the main lending process and would benefit from centralisation and economies of scale. Firstly, for example, there are distinct processes involved in perfecting a security depending on the nature of the security (i.e. it is different for property or stocks and shares). Each of these processes is very similar in whatever bank it is carried out because much of the process is determined by law and external agencies. Finally, the systems that support it are not that integrated into the bank’s main systems, hence consolidation of the IT platforms supporting securities processing is relatively straightforward. For banks this cross industry consolidation essentially means the use of an outsourcer or multiple banks setting up a joint venture.
Even if the security operation is retained in house, we think considerable cost reduction progress will be made by the use of image technology. This is because managing security (perfection, reviewing, releasing) involves significant correspondence with external agencies such as the Land Registry, Share Registrars, Valuers etc. Image copies of all this documentation would –
- Allow for process re-engineering within the securities handling area (organise work by security type or by customer group)
- Speed up processes in other areas such as recoveries and sanctioning.
This latter item will require the ability to view the security documentation in those areas (for example, it would be necessary that these documents can be viewed via a Web browser in locations across the bank).
Customer Accounting Processes
There are many banking processes that are already very heavily automated (e.g. interest calculation) and have been so for a long time. Just because a process has been automated does not mean it does not need investment. A factory with a production line needs continual improvements and enhancements to that line. This is so for the bank’s computer systems. In particular we believe the core banking systems, mainframe based and mainly batch, need continued investment. This is because they form the heart of the bank’s operations and businessmen and IT professionals alike feel bitterly frustrated at how intractable they are to change.
In most banks, what we call “Customer Accounting” systems have the following functions packed into a very large, complex, knotty bundle
- Customer accounting
- Financial accounting
- Interest processing
- Charges processing
- Account opening /closing
- Group Accounts
- Statement processing
- Customer data
- Limits processing
- Loan repayments
- Product processing
- Auto transfers/sweep accounts
- Reports production
The result of this complexity is that a change to one part (e.g. a new limit structure) requires change and testing in all the other parts.
Everything indicates that these systems will be around for decades (although, as Banking Operations Strategies & Technologies concludes, there is a move toward investing in the ‘gradual break-up of legacy banking software’) and so there is a strategic business need to unpick the knot and break down the core systems into their constituent parts. The aim here will not be very ambitious in IT terms (previous attempts to leap several generations of technology have generally failed) but rather to stay with the mainframe and the reliable early nineties technologies of mainframe relational databases, COBOL mainframe transactions and combine them with a web browser presentation layer.
The more important change we expect to see will be in the application architecture by splitting out the functionality and making the component functions more online and less batch based. The actual timing of the changes will probably be opportunistic, coinciding with business driven developments that heavily impact on particular systems.
Many banks’ activities in this area have recently focused on outsourcing. It is our opinion that there remains considerable room for cost savings and service improvements in cheque processing across the banking industry through “image copy processing”. Photo imaging of vouchers would cut out the need for physical movement of vouchers within and between banks. The clearing cycle would no longer be dependent on this physical movement of the cheques; instead the clearing process would revolve around the transfer of the image copies. We think this will have a number of benefits –
- The exchange of images as well as IBDE files between banks could eliminate almost the entirety of the ‘in-clearing’ function from the banking industry;
- Image technology would free up centralisation opportunities (e.g. technical checks, exceptions and queries currently have to be handled where the voucher lands – a centralised unit could be set up to handle these instead). This could lead to cost savings (especially with opportunities to move such units overseas).
- Removing the time criticality around the physical transportation would have benefits as the transport legs of the clearing cycle are quite vulnerable (traffic jams; road traffic accidents; robbery; etc). It would therefore help improve security and service.
- Removing the time criticality around the physical transportation of cheques could reduce the clearing cycle by about a day as the cheque image could be received faster.
- Finally, those customers that receive their cheques back could be offered electronic image copies of their cheques, routed via email or electronic banking to wherever in their organisation they are wanted (and indeed with any number of copies).
As is expected in other areas image capture at the first point of entry would realise the most benefits. However, having to image scan large numbers of cheques brought to a branch by a commercial customer would cause long delays at counters. Therefore the relative benefits of centralised or decentralised image capture would depend on the customer profile of a bank (e.g. it may be more acceptable to image copy cheques in branches if the majority of customers are bringing in few cheques). Alternatively, it may be better to create data files from cheques when they land in the branch and have the image copying done at a central location.
As IPSL and EDS dominate the outsourced clearing market in the UK it is clear that moving image files between banks would in fact mean moving image files around within and between these companies.
Other Payments Processing
For the purposes of this framework we divide payments (other than cheques) into four types
- High value payments (CHAPS, CHAPS-euro, and Currency Payments)
- Low value automated payments (standing orders, direct debits and direct credits)
- Credit and Debit card payment processing (out of scope of this Framework)
- Cash Payments – see Cash Handling
We believe the high value payments arena is amenable to centralisation and specialisation (i.e. centralisation and specialisation of CHAPS keying, funds checks, money laundering checks, payments repair, payments enquiries). Most banks have moved some way to centralise these activities.
Probably the next big step in centralisation and specialisation would come with some cross industry sharing of the activities, presumably by outsourcing (for example payment repairs activities).
Other step changes in high value payments cost reduction would require change outside the payments centres by getting customers to key the details of the payments. The principal ways would be to
- Extend CHAPS payment entry to online personal banking (maybe using digital certificates to improve security)
- Using electronic forms in branches/online to be filled in by customers
In the area of low value automated payments the actual processing of the payments is achieved with a very high degree of STP and it is very hard to achieve significant extra savings. Where large numbers of staff are still involved is in the area of setup, amendment and deletion of mandates. This is covered in the section on Account Maintenance.
We include three large areas of activity under this heading:
- Create/delete/amend records – accounts, regular payment instructions, customer data (these normally require signed customer instructions);
- Bank driven processes – Unapplied items, suspense account processing, branch balancing etc;
- “Odds and Sods” – a large number of miscellaneous processes that are individually small in frequency but collectively add up to a lot (e.g. audit letters, cheque book orders, and copy statement requests).
The other way we look at these processes are the speed of interaction with the customer.
Postal speed – by which the customer signals he is prepared to accept some delay to an activity. Examples are:
- Letters in the post
- Dropping off a completed form at a branch
Real time – by which the customer signals he is expecting the action to be completed while he waits:
- Phoning a request
- Setting up a standing order mandate via the internet
- Opening an account with a Branch Officer on-line.
The bank can, to some extent, set the customer’s expectations on the speed of interaction. So if a customer wants to open an account over the phone, the operator can say that he will get a form sent to him hence re-setting the expectation that we are working at postal speed.
The strategy for handling the service/processing varies according to the type of activity and the time demands. See the suggested approach below.
- Create/Delete/Amend – Postal Speed
- Create/Delete/Amend – Real Time
- Bank Driven Processes (Postal Speed)
- Odds & Sods – Real Time
- Odds & Sods – Postal Speed
Create/Delete/Amend – Postal Speed
The principal characteristic of these processes is a requirement to have a high degree of certainty that the customer has authorised the task and that the customer is who he says he is. This boils down to
- Signatures on paper in most cases,
- Additional proof of identity (passport etc) in some cases,
- And a high degree of authentication of the customer if the change is communicated over the phone or internet (some cases).
Because of the signature and proof of identity, many of the processes have to be carried out at postal speed.
We believe postal speed activities will be (if they are not already) centralised (e.g. not carried out in the branch or relationship manager’s office). We expect the branch will act as an intelligent post box, converting (via branch scanners) any proof of identity documentation and signed forms into electronic images and then E-mailing these to a central processing area.
Create/Delete/Amend – Real Time
We think Banks want customers to key in data as much as possible themselves and that banks still have work to do to let them. Most banks offer a capability to set up standing orders and open a second account via the internet with the customer effectively keying the data directly into the product engine. We think this will be supplemented with a number of additional approaches.
- Electronic versions (e.g. HTML or Word documents) of all forms will be available over the internet.
- Workstations for filling in and/or downloading forms by customers will be available in branches (the Easy Jet ticket purchase model at airports).
- Branch staff will encourage people to key in data in the branch, much as they encouraged people to use ATMs in the early days of that technology.
- Customers may well be able (and encouraged) to bring in diskettes with forms or e-Mail them to a central ID.
In central processing areas, software will be needed to “scrape” the data out of the forms and automatically transfer it into the legacy systems. Although some forms will be filled in incorrectly, the amount of Straight-Through-Processing will be hugely increased.
In Call Centres the expectation is that the work should be keyed wherever possible by the Call Operator directly (i.e. in real time). Although Call Centre staff will be more expensive than processing staff, the cost incurred in the Call Centre staff writing the work down and faxing or emailing the work to a processing centre will be too high.
Bank Driven Processes – Postal Speed
These are essentially driven by the 24-hour batch processing cycle that nearly all banks run. The overnight batch processing of the legacy banking systems produce large numbers of reports that require action by the account maintenance function; examples are –
- Branch balancing
- Unapplied items
- Returned items
- Failed Data Create/Delete/Amend
- Diarised actions
Some of the processes are control processes and others are about errors processing. In many respects these are only indirectly in support of customers and, because of the 24-hour cycle, they inherently run at “Postal speed”.
It is our view that in the spirit of “only electrons move in the bank”, (see Banking Operations Strategies and Technologies) the current implementation of these processes based around printed reports will substantially reduce. It appears that electronic versions of the reports online or e-mailed versions of the reports are now feasible.
In a longer term sense, many of the error handling issues require better data validation at the point of data entry and a move away from batch-based processing. This typically will require the “gradual break-up of the core processing systems” (see Banking Operations Strategies and Technologies). For example, separating out statement data and its maintenance update from the customer accounting system.
“Odds and Sods” – Real Time
Many of these processes are similar to create/update/amend (see above) but do not require forms and signatures. They are often service requests and as such should be handled as quickly as possible, with minimum bank staff intervention in the delivery channel. In principle Banks would make more self service investment in the channels; particularly
- Electronic banking (more self service functions, such as a “How do I?” query capability)
- Branches (more self-service technology)
- Call Centres
However, the problem appears to be that each of the processes is relatively low volume and so it is hard to cost justify IT developments. It is not clear to us what Banks will do about this workload/cost base.
“Odds and Sods” – Postal Speed
Here the assumption is that the service request is signalled as being not time critical because it has come in by mail or e-mail and Banks will pass the work to a processing centre by emailing an image of the customer letter or forwarding the customer e-mail.
The principle costs associated with cash handling are those of
- Secure transportation
- Stock holding costs charged by the Bank of England
The latter is not the object of this framework, it requires very specific strategies. Notwithstanding some false starts in the recent past, the former is amenable to centralisation and cross industry sharing of resources, presumably by outsourcing or a joint venture arrangement. The centralisation of activities allows for investment in high volume cash counting and coin weighing machines as well as other packaging technology to allow re-use of notes and coins. There could also be room for sharing secure transport runs.
A crucial element in this strategy will be in persuading the customers, particularly business customers, to allow the cash to be counted in the centres and not in the branches.