1 ICBI Operational Excellence What do Banks Really Need to do to Address the Issues of Business and Operational Excellence? How will the Long Arm of Regulation.

1

ICBI Operational Excellence

What do Banks Really Need to do to Address the Issues of Business and Operational Excellence?

How will the Long Arm of Regulation Continue to Impact Banking Operations?

September 2010

BOB LYDDON – IBOS ASSOCIATION

2

The Payments business is not on a convergence trend

● Harmonisation within the EU / EEA has hit the buffers

Politicians’ stated aim Status

Cut 2% GDP off cost of payments

Direct Payments

Fewer Interbank events

Layered Market

Only 4% migration to SEPA instead of the critical mass that should, by end of 2010, have made migration irreversible (SEPA Roadmap 2004)

XML no longer integral to ISO20022 or SEPA

More room for national flavours

● Maxi-SEPA, universal usage of XML are not “just around the corner”….

3

Reg 924/2009 - partial abolition of CBR- reachability for SEPA Core Direct Debit- no MBPs after 2012

PSD - float income reduction- increased responsibilities and risk- new entrants with passports

SMED - mandatory adoption of SEPA schemes- but with many safety valves- national variations/”niche” schemes

More complexity

Lower price

Impact of latest regulations

4

SWIFT - standards co-existence- no retirements dates for MT- E&I and n92, 95, 96 co-exist beyond 2012

SEPA - SEPA does not mean XML any longer- “essential requirements” discussion- permits national variations/data formats- Credeuro (XCT) is SEPA-compliant- “niche” schemes formally excluded (RIBA)

Standards convergence no longer on the agenda

Proliferation and complexity

Very low migration to SEPA instruments so far (see Appendix with STEP2 statistics including for Credeuro)

Latest “Industry” developments

5EU Vision for future market structure: away from Vertical to Horizontal Integration

HORIZONTAL

Direct SWIFT Access

Bank-owned Channels

Mobile Social Networks

Activities requiring Banking Licence or Payment Institution Licence

Processors of non-licensed activities

VERTICAL = BANK

Customer Access Systems

Operations and data centres

Own backbone

ShareholdingsDirect memberships

Sub-memberships

Correspondents

SWIFT ISPs BT Radianz

PEACHes ACHs Net Settlement Systems Banks

Final Settlement activities:

● Direct RTGS membership

● Banks

Acquisition

Processing

Communications

Clearing &

Settlement

Only two areas of activity are the unalterable preserve of Banks:Activities requiring a full Banking License (as opposed to a PI or eMI)RTGS clearing membership

6

EPC Design Model for Card Schemes (Source EPC)

Cards market model is already layeredEPC design encapsulates this

7

EPC Design Model for Payments (Source EPC)

Layered – big change from status quoMore layers and more complex than for CardsIn principle a lot of space for non-bank “horizontal” players

8

Cards vs Payments

Cards Payments

Schemes Concentration of volume on Visa and Mastercard

Distribution over many CSMs to avoid duopoly

Standards Rigid definitions from the centre

Space for community-defined variations in the form of Additional Optional Services and Value-Added Services

Performance High speed, high volume, interoperable

Message size x line capacity x routing facilities >> quick response time

Bigger message size (XML) and usage of SWIFT (which runs on the lines of Equant etc) could compete with public internet usage

9

SWIFT Standards Co-Existence IR 535

IR 535 approved at SWIFT’s June Board Meeting is a major change of position:

Migration to XML is non-mandatory: no end dates for MT

ISO20022 does not automatically imply XML as the physical layer

The UML level (actors, data, process) is the fulcrum

By extension any Scheme that fulfils the EPC Rulebooks and the UML models can be SEPA-compliant – it does not have to use XML

10

Layered market should create many opportunities for industry utilities

For activities not needing a banking licence

Should banks lead this? What is the track record in collaborative, self-regulatory ventures?

What activities and tasks could be delegated to industry utilities?

What would be the hand-offs to them as Inputs to their Processes, and how would their Process Outputs be re-integrated into the banks’ processing flow?

Industry Utilities

11

Simple Outsourcing Model (Source PW)

Critical Non-critical

Core Do-It-Yourself Volume processer:Cheaper if output is commoditised

Non-Core Professional service expert: more expensive

Don’t do it at all

So which actions in the payment processing flow can be outsourced and who to?

Complicated in banking by the banking licence issue + VAT + no proof that the service itself is becoming harmonised.

12

Linear & Sequential Process - Amazon

Payment and Logistics are Non-Core but Critical.

What about Fedex?

Is there room here for a cheap volume processor?

13

Payment Process

Sporadic “Core and Critical” actions all the way along

PSP must retain decisions on credit, routing and liquidity.

Outsource does the legwork that does not require a banking licence.

Multiple hand-offs during the flow

Outsourced work is VAT-able towards the PSP: can the PSP reclaim it?

14

ACH models – direct + indirect submission

Vocalink acts as an effective outsource for UK banks

It prepares files of accounting entries

There is no “choice of routing” for UK Credit Transfers and Direct Debits as Vocalink is the sole system

Banking controls are done upfront by the posting of credit limits by system members for corporates + other FIs (who are not direct members)

Those other FIs in turn post limits for their customers

Three rings: not equal access for all, as required by PSD Title 2

15

ACH Payments – Credits and Debits: Submission direct to clearing

Account-holding

bankClearing System

File upload into EB

Client Accounting

system

DR/CR ACCOUNT

BENEFICIARIES/DEBIT PARTIES’ ACCOUNTS

OTHER BANKS

OTHER BANKS

OTHER BANKS

16

ACH Payments – Credits and Debits: Submission model through the bank

Account-holding

bank Clearing System

File upload into EB

Client Accounting

system

OTHER BANKS

OTHER BANKS

OTHER BANKS

DR/CR ACCOUNT

BENEFICIARIES/DEBIT PARTIES’ ACCOUNTS

17

Is Cloud Computing the answer?

Is this the way the hand-off to Outsourcers is effected?

What is the certainty that the computing resources at the Outsource remain superior to in-house ones?

How nimble is this set-up when requirements change and what are the costs of change?

Cloud computing is internet-based computing, whereby shared resources, software and information are provided to computers and other devices

18

What about Blade Computing?

Massive increase in power and flexibility.

Enables datacentre consolidation and shrinkage.

Game-changer – allows processes to be run in much smaller scale without normal penalty of poor scale economics.

A blade server: houses multiple server modules ("blades") in a single chassis (rectangular, usually steel frame).

Used in datacenters to save space and improve system management.

19

What’s the environment we are going into?

Proliferation of processes defying achievement of scale economics

Variations in data and service level

Learning from other industries:

oBanking processes are not linear and can be hard to fully automate

oActivity has to be strictly standardised with rigorous data and banking checks applied upfront

oThe Cards business and a utility model like Vocalink for UK BACS follow this line

oThat is not the SEPA environment as it appears to be unfolding

20

SEPA environment unlikely to be 6σ compliant

Bank B

Bank A

STEP 2

SEPA Compliant

ACH

SEPA Compliant

ACH

PEACH

Too many players Too many routing choices Expandable data requirements (VAS, AOS..) May work in small volumes to begin with As more players come on and more volumes,

standard deviation moves away from the mean over time

Many instances then fall outside Upper/Lower Specification Limits which are only 2σ away from the mean

Not 3.4 Defects Per Million Opportunities (6σ) but 000s of DPMO

non-STP rate in 15-20% range

Account Blocked

SCT

SCT

SCT

SCT

R-message

R-message

R-message

R-message

21

Curve is same shape but μ moves by 1.5σ

Hardly any instances below LSL

LSL USL

+1σ +2σ +3σ

μ

μ -3σ -2σ -1σ

But huge numbers – 45% - fall outside USL

Non-compliance / not STP

Proliferation of official Schemes (Core, B2B..)

Annual Scheme upgrades

AOS

What happens when more volume is brought onto SCT and the normal assumed movement of 1.5 occurs in or over time?

μ moves by 1.5σ because μ itself moves or σ widens

22

Conclusions

The emerging environment is one of divergence, not convergence

Detailed management of inputs and outputs becomes a Core and Critical function of a Transaction Bank

That means “do it in-house”

Whether the IT configuration can include Cloud will depend on line capacity and contention with public internet usage

Blade seems to be a better bet: higher power in one place and more flexibility >> lower costs and a change of scale economies

23

Appendix

EBA STEP2 Volumes as a proxy for SEPA migration:

STEP2 frequently used a bridge between other CSMs

Credeuro (XCT) statistics are steady

SEPA Credit Transfer (SCT) statistics show steady increase:oHigher average value than XCToHigher average value than national payments

Volumes are therefore:oCross-border paymentsoCorporate payments

Virtually no Direct Debit volumes migrated to the SDD yet – key date is 30th November when all Euro-in banks must be reachable

24

APPENDIX: STEP2 SCT

  Average daily volume Average daily value

July 2010 663,259 3.84 bio. euro

June 2010 653,925 3.77 bio. euro

May 2010 604,561 3.55 bio. euro

April 2010 594,092 3.53 bio. euro

March 2010 553,841 3.13 bio. euro

February 2010 466,026 2.76 bio. euro

25

APPENDIX: STEP2 XCT

  Average daily volume Average daily value

July 2010 189,292 841 mio. euro

June 2010 192,439 822 mio. euro

May 2010 189,131 805 mio. euro

April 2010 202,067 852 mio. euro

March 2010 193,253 789 mio. euro

February 2010 195,586 778 mio. euro

26

APPENDIX: COMBINED STEP2 SCT & XCT

 Average daily volume:

XCT + SCTAverage daily value:

XCT + SCT

July 2010 852,551 4.681 bio. euro

June 2010 846,364 4.592 bio. euro

May 2010 793,692 4.355 bio. euro

April 2010 796,159 4.382 bio. euro

March 2010 747,094 3.919 bio. euro

February 2010 661,612 3.538 bio. euro