technology as driver of change: automation and cognitive

Technology as driver of change: Automation and Cognitive Solutions in Business Compliance

CeTIF OPEN SUMMITUNIVERSITA’ CATTOLICA DEL SACRO CUORE5 LUGLIO 2018 - MILANO

Raffaele Cosimo

Amministratore Delegato

Promontory Italy S.r.l., an IBM Company

The Financial Services Industry is at a Crossroads

The financial services industry is operating on inadequate, legacy systems and

being held back by a manual, labor-intensive approach to regulatory compliance.

Managing the abundance of data – banks could be handling data volumes in the

tens of exabytes before 2020 – is making the industry’s job even harder.1

And compliance responsibilities and expectations for banks and their executives

have never been higher.

These limitations are amplified by many firms struggling to achieve meaningful and

consistent gains in return on equity

1. Nasar, Mohammed and Christopher J. Sifter, “Putting Big Data to Work in Your Bank.” Crowe,

Howarth, LLP. June 2016. http://www.aba.com/Products/Endorsed/Documents/Crowe-Putting-big-data-to-

work-in-your-bank.pdf1

Cognitive RegTech is the Answer

By leveraging RegTech across the entirety of the banking enterprise, firms can

combine data, technology innovations and new methodologies for regulatory

compliance and augment the ability of compliance teams to do their jobs more

efficiently and effectively.

RegTech is the digitization of manual reporting and compliance processes. It

promises substantial cost savings for the financial services industry.2 Deploying the

right combination of expertise and technology can help firms manage new and

changing regulations. Cognitively enabled banks will have the tools to help

prevent fines, penalties and reputational risks, and they will be able to scale

these solutions without continually adding to compliance teams.

2. Arner, Douglas W., Janos N. Barberis and Ross P. Buckley. “FinTech, RegTech and the Reconceptualization of Financial Regulation.”

Northwestern Journal of International Law & Business, Forthcoming; University of Hong Kong Faculty of Law Research Paper No.

2016/035. October 1, 2016. Available at SSRN: https://ssrn.com/abstract=2847806

An Integrated Aproach

An integrated approach calls for:

Industry expertise and deep domain knowledge: Cognitive systems are literally

taught by experts – whether doctors, researchers, wealth advisors, customer

service reps or veterinarians.

When considering RegTech systems, inquire about who did the training. Was it

done by programming experts or regulatory experts? Was the solution trained to

understand the intent of regulations, or just to recognize keywords? Is there

consideration for the front, middle and back office and the implications of the

same regulatory mandates as applied across the firm?

Simply identifying new regulations is not a path to success. Understanding the

intent of the regulation is critical in determining how best an organization should

respond. Again, this requires deep domain expertise.

An Integrated Approach

Data insights: Data is a key factor in gaining competitive advantage.

Organizations that are able to access, understand and generate actionable

insights from their data put themselves in a position to foster and strengthen

trust, which can lead the transformation of the firm as well as professions and

industries.

Similarly, it is not just about crunching on large volumes of data; it’s about finding

the hidden insights in the data, across both structured and unstructured data

(such as video and social media). This requires truly cognitive technologies that

can understand, reason and learn.

Most important, it is no longer sufficient to rely solely on domain expertise or

advanced technology; it is the integration of these capabilities – with many

abundant data sources – that differentiates a RegTech solution. And, it is exactly

this combination that is required to tackle the increasingly complex risk and

compliance challenges the industry faces today.

IT’S NOT HUMAN VS MACHINE

DANGER IS

making automated

decisions based

on untrustworthy

learnings that are not curated.

IT’S HUMAN AND MACHINE

RegTech: Regulatory PerspectiveRegtech could address a wide array of requirements related to regulatory reporting, financial crime, operational risk

(including cyber-security and fraud detection), consumer protection and data protection regulation. The technologies

used include IT (software, cloud computing, API, automation and AI), data technologies (big data, machine learning, risk

scoring, real-time monitoring), identity technologies (biometrics, vocal recognition) or new technologies such as the DLT

combining cryptography and IT solutions (BCBS - Sound Practices Implications of fintech developments for banks and

bank supervisors, 02/2018, based on IOSCO, Research Report on Financial Technologies, 02/2017)

The European Banking Authority (EBA) published the first products of its FinTech Roadmap, namely a thematic report on

the impact of FinTech on incumbent credit institutions' business models and a thematic report on the prudential risks and

opportunities arising for institutions from FinTech (EBA - Report on the impact of FinTech on incumbent credit

institutions' business models, 07/2018; Report on the prudential risks and opportunities arising for institutions from

FinTech, 07/2018)

• The first report sets out five factors that might affect incumbents' business models from a sustainability perspective:

1. Digitalisation/innovation strategies pursued to keep up with the fast-changing environment

2. Challenges arising from legacy ICT systems

3. Operational capacity to implement the necessary changes

4. Concerns over retaining and attracting staff

5. Increasing risk of competition from peers and other entities

• The second report assesses seven use cases, where new technologies are applied or considered to be applied to

existing financial processes, procedures and services:

1. Biometric authentication using fingerprint recognition

2. Use of robo-advisors for investment advice

3. Use of big data and machine learning for credit scoring

4. Use of distributed ledger technology and smart contracts for trade finance

5. Use of distributed ledger technology to streamline customer due diligence processes

6. Mobile wallet with the use of near-field communication

7. Outsourcing core banking/payment system to the public cloud

RegTech: EBA - Level of involvement in the

selected FinTech use cases (37 EU banks)

Source: EBA - Report on the prudential risks and opportunities arising for institutions from FinTech, 07/2018

10Time

The changing body of unstructured

regulatory content and manual,

labor-intensive processes require

nuance and judgment.

Cognitive computing can enable

rapid augmentation of human

capability; address nuanced,

ambiguous problems; and scale the

application of industry knowledge.

Cognitive

Capabili

ty Digital and automated

Highly manual,best efforts

Digital computing and process

automation can enable people

to multitask efficiently and

connect globally.

Many banks

are here.

RegTech is a Journey

Applications: Cognitive Computing Solution

RegTech: Vision, Objectives and Applications

Know-Your-Customer (KYC)

New regulations (e.g.,

ownership & control),

heightened expectations,

and enforcement actions

Manual, repetitive,

time-intensive,

never complete,

non-standardised,

and risk laden

Silos across business

lines and banks;

basic information

not shared

RISING

EXPECTATIONS

NATURE OF

THE WORK

LIMITED

VISIBILITY

IMPACT

High – and rising – costs

Insufficient information – many high risks

False PEP or negative news hits

Negative customer experience

Stressful jobs

Growing EDD roles

KYC Industry Challenges


Challenges & Needs

High operating costs and

complexity from disparate

legacy systems

Robotic Process Automation coupled

with various analytics minimize

workloads and deliver greater

standardisation, consistency, and

efficiency

Manual processes for KYC records

are costly and lead to negative

customer experiences

Automating input from external data

aggregation avoids unnecessary

customer requests, reducing friction

while speeding up processes

False positives related to

negative news bogs

down analysis

Cognitive analytics, which quickly

identifies more relevant articles, and

context annotation accelerate analyst

decision making

How Cognitive Can Help

Technology Can Help Address the Inefficiencies of Today’s KYC

Operations


Speed EDD investigations up to 2x

Standardise highly manual processes

Reduce false positives in negative news

by applied cognitive analytics

Shorten investigative time by proactivity

ingesting unstructured and structured

data

Elevate the productivity of your high-cost

resources

Intelligent robotic automation, cognitive content analytics, and identity

resolution streamline processes can:

Enhanced Due Diligence


16

KYC is a particular area for improvement

Mission: To elevate the financial services industry to a common best practice for authenticating and

validating KYC documentation to sustain a sharing economy for KYC.

KYC is shared, trusted, and secure

reducing duplication and providing easier

access to new technologies and

innovation.

F u t u r eKYC is a siloed effort by each bank,

resulting in duplicative processes making

it difficult to take advantage of

technology developments.

C u r r e n t

17

Shared KYC – Key Benefits

Financial

Institutions

Costs

• Time. Reduced time spent on

on-boarding frees up RMs and

Operations personnel.

• Costs. Reduced document

storage and security costs

due to digitization.

Good for Business Reduced Risk

$

• Better Client Experience.

Eliminates duplication and

frustration for the clients. Less

need for RMs to chase clients

for documents.

• Speed to Onboard. More

efficient on-boarding leads to

quicker and greater access to

new client relationships.

• Standards. Harmonized standards ensure

consistency and ability to anticipate and

keep up with changing regulatory

requirements.

• Technology. Systems risk lowered through

enhanced document retention and recording.

Immutable audit trail creates golden source.

• Regulatory Relationships. Potentially

smoother inspections and shorter time on-

site.

Convenience

• Convenience. Documents

are only provided once and

updated documents are

seamlessly provided to all

banks.

• Consistency. Harmonized set

of documents and verification

standards among banks.

Control Greater Choice

• Visibility. Clients maintain

visibility over which

documents have been

provided to each bank.

• Control. Clients retain control

over who can access each of

their documents, eliminating

unauthorized sharing.

• Speed to On-board. Quicker

on-boarding means greater

flexibility to establish new

banking relationships.

Clients

18

Envisioning Approaches to Shared KYC

Third-Party Utility Service Provider Blockchain-based Shared KYC

Summary:

With a third-party utility service

provider, the service provider owns and

controls the data, acting as a central

repository.

Summary:

A Blockchain-based Shared KYC Platform

enables secure, permission based

sharing, leaving control of the data with

the owner.

Third-

Party

Utility

FI

FI

FI

FI

FI

FI

FI FI

FI

FI

FI

FI

FI

FI

FI FI

Blockchain

Platform

Summary:

Each Financial Institution conducts

collects customer data and conducts due

diligence activities with their own data

repositories.

FIFI

FI

FI

FI

FI

FI

FI

FI

FI

Independent KYC

19

KYC – Governance Solution Options

Consortium (Banker’s Association) Promontory

• Expert subject matter knowledge

• Global regulatory connections

• Global banking connections

• Difficulty reaching consensus and

decision-making

• Focused on one jurisdiction only

(i.e., no international view)

• Lacking independence

Banker’s Associatio

n

FI

FI

FI

FIFI

FI

FI

20

Transaction Monitoring

RISING

EXPECTATIONS

Consistent processes,

enterprise-wide imputed

knowledge, risk

appetites, customer-

specific

NATURE OF

THE WORK

Static rules,

subjective analysis,

manual monitoring,

labor intensive,

expensive

LIMITED

VISIBILITY

KYC & TM silos,

aged/disconnected

systems, ownership

& control, entity

resolution

IMPACT

High costs – labor & systems

90%+ “false positives”

Enforcement actions

Human errors

Diminished customer relations

Focus is on the signal, not the noise

Transaction Monitoring Industry Challenges


Challenges & Needs

Effective transaction monitoring with

cognitive accelerators that augment

current transaction monitoring

investment

Faster identification of suspicious

activity and with accurate insights,

including contextual information and

analysis

Leverage automation, machine

learning, cognitive capabilities, and

Promontory domain expertise for an

enhanced view of risk

How Cognitive Can Help

Legacy, manual AML transaction

systems hinder the ability to

meet evolving regulatory

requirements

High volume of false positives

impedes efficiency

Lack of a holistic view of data,

systems, and domain expertise

Derive Insights and Increase Effectiveness and Efficiency


Optimise KYC and customer transaction monitoring for AML programs by

improving quality of insights, reducing false positives, and accelerating

investigations can:

Increase speed and accuracy of insight to

identify true positives and false positives

Contextualise transaction risk with related risks

about the customer and counterparties

Understand related transactions and entities to

get better sense of true suspicious activity

Evaluate data from historical alerts, cases, and

SARs

Enhance Alert Triage and Investigations for Transaction Monitoring


Cognitive technology and machine

learning enhance and simplify the

onboarding and customer due diligence

process

Technology integrates the existing

customer risk rating, periodic customer

review, and transaction monitoring

processes into a single, dynamic,

customer-focused process for identifying

money laundering risk

Information sharing across financial

institution consortia facilitated through

shared business process services to

increase efficiency and effectiveness

Transforming AML Compliance

Anti-Money Laundering

Trade Surveillance Solutions

Trade Surveillance

Fuse insights to detect suspicious activities

View linkages with instant drill-down & playbacks

Build employee profiles with personality & behavior traits

Order

Trade

Execution

Reference

News

Social

Email

Chat

Voice

Transactions

Communications

External

Su

rveilla

nce

See alerts & investigationsfrom continuously updated risk models

Insights for Financial Services

27

Regulatory Compliance Analytics (RCA)

Overwhelming and complex feeds: Regulatory information is received by

individuals and clients in not standard forms and different channels;

Manual process: Several information are not automatized, and managed manually

for updates;

High volume of alerts: Number of false positives alerts arises from non

standardization in information; and

Difficulty to search across regulations: Limited ability to provide ad-hoc Q&A

functionalities on clients side.

Risk Assessments transparency: Clients find challenging to guarantee completely

transparent solid assessment due to the difficulty in collecting and controlling

information that are not highly standardized;

High intensity, high-cost manual process: Existing tools do not have a mechanism

to link control requirements to regulatory requirements, and to the level of inherent

risk without manual effort and constant updates; and

Lack of deep insights: Information characteristics at the current status and their

process reduce the ability to produce reports and perform analysis of obligations,

controls and risk characteristics dynamically across the risk assessment process.

RCA

Challenges

Risk

Assessment

Challenges

Challenges in the current regulatory environment


Single delivery channel:

Production and maintenance of

a standard library on regulatory

content (containing rules,

guidance, speeches, based on

the clients profiles and needs)

Change management: follow

regulatory releases and update

the library automatically with

regulatory changes, including

version controls

Customization: Provide

company-specific content and

taxonomies to the regulatory

library in order to support

alerts generation and Q&A

search and processes

Training: provide the client a

series of workshop on how to

detect and classify relevant

new information from

regulatory releases

Obligation identification:

Support in identifying and

categorize new and existing

obligations based on

applicable regulatory

library areas for the

compliance needs of the

clients

Obligations library: Manage

applicable obligations

tagged with company

Target alerting: Support in

the prioritization and

confirmation of alerts

based on the detected

obligations

Control identification:

Support the compliance

officers in the

identification and

categorization of new and

existing controls for the

defined obligations

Control library: Manage

applicable controls tagged

company identifiers; and

Integrate with GRC

RCA is designed to provide a simple, low-footprint service allowing

banks to manage obligations and controls

Regulatory Content

Acquisition

Obligation

Identification &

Management

Control

Identification and

Linkages

Dynamic risk

assessment: Recognize

gaps in the company’s

controls by porting

obligation and control

updates to OpenPages

Visualization Analytics:

Provide insights through

data analysis &

customized dashboards

Process improvement:

Manage the

implementation of new

controls tied directly to

regulatory requirements

for end-to-end coverage

Risk

Assessment


Thank you