Compliance in Banks and Corporate Governance

Identifying Compliance Issues and Compliance Risks

Read at your own pace. Use chapter practice to check your understanding.

Chapters and chapter contents
100%
Practise this chapter ↓

Last updated · By the Clerk2Officer editorial team · Report a correction

1. Begin with the obligation and the activity

Identifying compliance risk starts with understanding what the bank does and which requirements govern that activity. A list of circular titles is not enough. The analyst must connect each relevant obligation to a process, customer journey, product, system or decision. Otherwise, a requirement may be acknowledged centrally but remain absent from the procedure used by staff. The process also works in reverse: a new activity should trigger a search for applicable obligations rather than be assumed permissible because the existing register contains no entry for it.

Consider an original example involving a new online request to close an account. The team should identify applicable customer-service, identity, recordkeeping and communication requirements, then trace how the request moves from submission to completion. Each step presents possible failures: the request may not be captured, a required verification may be missing, a charge may be wrong or the customer may receive no confirmation. Identifying risk means describing those potential failures and their consequences clearly enough to design controls. This chapter develops that method, distinguishes issues from risks, explains inherent and control risk, and connects testing, reporting, inspection and loan review into a coherent framework for finding and resolving weaknesses.

2. Risk, issue, incident and finding

A risk is a possibility of an adverse outcome; an issue is a condition requiring attention; an incident is an event that has occurred; and a finding is a supported conclusion from review or testing. Institutions may use slightly different taxonomies, so definitions should be explicit. The important distinction is between potential exposure and established facts. A weak control can create risk before any breach is detected. A complaint may signal an issue without proving an incident of noncompliance. A finding should state what evidence supports the conclusion.

Suppose a system permits an employee to approve their own refund entry. That is a control weakness and creates risk. If review finds that such approvals occurred, those are observed events. Whether they breached a specific requirement depends on the applicable criterion. If the approved transactions were otherwise accurate, the control failure may still matter because the required independent check did not operate. Conversely, a financial loss does not automatically establish a compliance breach; the cause and obligation must be examined. A useful register separates allegation, investigation status, confirmed condition, consequence and action. This prevents premature accusations and also prevents genuine weaknesses from being ignored merely because a measurable loss has not yet occurred.

3. Sources of identification

Compliance issues can emerge from regulatory change, business proposals, complaints, audit findings, incidents, management information, staff concerns, legal developments and supervisory observations. No single source provides complete coverage. Complaints reveal customer experience but may underrepresent people who do not know how to complain. Audit findings reflect the scope and timing of the audit. Self-assessments can provide broad coverage but may be optimistic. A strong identification process combines sources and investigates inconsistencies rather than treating each report as an isolated event.

An original monthly review notices rising account-closure complaints, a technology backlog affecting closure requests and a branch report of manual workarounds. Individually, each may appear manageable; together, they suggest a common process weakness. Link them through a shared process identifier and examine the full customer journey. External developments also matter. A new product, acquisition, vendor change or system migration can alter risk even when no internal incident has yet occurred. Staff should have a route to raise uncertainty early. The objective is not to maximise the number of issues recorded, but to make significant weaknesses visible in time for action. A low issue count may indicate strong controls, poor detection or a culture that discourages reporting; further evidence is needed to tell the difference.

4. The obligation and risk register

An obligation register records what must be done; a risk register records how failure might arise and why it matters. They should be linked but need not be identical. One obligation can affect several processes, and one control can support several obligations. Useful fields include source and version, applicability reasoning, effective date, process owner, risk description, control references, testing evidence, open issues and review date. Avoid vague entries such as regulatory risk across all branches. A useful risk statement identifies the event, cause and consequence in the actual activity.

For example: because the customer-category field is not validated, an exempt account may receive a charge, causing customer harm and breach of the applicable tariff requirement. This wording points to data validation, charging logic, exception review and customer correction. It is more actionable than fee risk. Registers should be maintained as living records, with changes linked to business and regulatory developments. An obsolete obligation should not simply disappear without a record of why it ceased to apply. Where interpretation is unresolved, record the uncertainty and responsible reviewer. A register is a management tool, not evidence that every requirement has been met. Its quality is demonstrated by traceable controls and reliable operation, not by the number of rows it contains.

5. Inherent compliance risk

Inherent risk is exposure before considering the effectiveness of controls. Relevant factors include the complexity of requirements, transaction volume, customer vulnerability, discretion, product novelty, geographic reach and reliance on third parties. The assessment should consider the nature and possible severity of failure, not just historical loss. A new process can have high inherent risk despite having no past incidents, because it has not operated long enough to generate a history. Similarly, a low-value transaction category may carry significant conduct risk if it affects many vulnerable customers.

An original comparison involves a standard internal administrative report and a complex retail product sold through several channels. The second activity may have more opportunities for misunderstanding, inconsistent disclosure and unsuitable execution. That does not mean it is necessarily noncompliant; it means the control framework needs to address a more demanding exposure. Use clear rating definitions and document reasoning. If a qualitative scale is used, explain what low, medium and high mean in the institution's context. Avoid presenting an arbitrary numerical score as a regulatory formula. The value of inherent-risk assessment is to identify where controls and assurance deserve attention, while remaining open to revision when the activity, customer base or applicable obligations change.

6. Control risk and residual exposure

Control risk concerns inadequate design, implementation or operation of controls. A process may contain many controls yet still be weak if none addresses the main failure mode. Evaluate whether the control is appropriate, assigned to a capable owner, supported by reliable data and evidenced in practice. Residual exposure is what remains after considering controls, but it should not be calculated mechanically without understanding their effectiveness. Several controls may depend on the same data feed or reviewer, so their apparent benefits are not necessarily independent.

Suppose both a daily exception report and a monthly dashboard use the same incomplete transaction extract. The presence of two reports does not create two independent layers of protection. A data omission can defeat both simultaneously. The review should test data completeness and identify correlated weaknesses. Likewise, a preventive system restriction may be strong in normal use but bypassable through emergency access. Evaluate that route and its monitoring. A control assessed as effective last year may weaken after staff turnover or a system change. Residual-risk assessment should therefore use current evidence and record uncertainty. It helps prioritise action, but it does not grant permission to breach a binding obligation simply because the remaining numerical score appears acceptable.

From obligation to residual exposureClockwise: identify the requirement, describe inherent exposure, evaluate actual controls, and assess remaining exposure. A favourable score cannot waive a duty.From obligation to residual exposureObligationEntity, activity and dateInherent riskBefore control effectsResidual riskWhat remains and why?ControlsDesign and operating evidence→↓←
Clockwise: identify the requirement, describe inherent exposure, evaluate actual controls, and assess remaining exposure. A favourable score cannot waive a duty.

7. A transparent illustrative risk calculation

The chapter calculator uses a deliberately simple scenario: annual failure probability multiplied by the financial impact of one event gives an illustrative expected annual loss. A supplied control-reduction percentage then reduces that modelled amount. With probability 10%, impact of 20 lakh and an assumed 60% reduction, inherent modelled loss is 2 lakh and residual modelled loss is 0.8 lakh. The arithmetic is useful for understanding assumptions and sensitivity. It is not an RBI compliance score, a penalty estimate or a statement that the risk can lawfully be accepted.

The model assumes a single relevant event outcome and a proportional control effect. Real events may recur, interact or cause nonfinancial harm that is not captured by the amount entered. A control-reduction estimate should not be invented merely to produce a favourable result. If evidence is weak, explore several assumptions and state the uncertainty. Increasing impact from 20 to 50 lakh while keeping other inputs unchanged increases residual modelled loss from 0.8 to 2 lakh. The same percentage can therefore conceal very different exposures. Use the laboratory to challenge how a conclusion changes with assumptions, then supplement it with severity, customer impact, legal requirements and control evidence before making any practical decision.

A transparent single-event model10% × 20 lakh = 2 lakh inherent modelled loss. Assumed control reduction 60% removes 1.2, leaving 0.8. This is not an RBI compliance score.A transparent single-event modelLakh; illustrative assumptions2Inherent1.2Reduction0.8Residual
10% × 20 lakh = 2 lakh inherent modelled loss. Assumed control reduction 60% removes 1.2, leaving 0.8. This is not an RBI compliance score.

8. Indicators and thresholds

An indicator is a measure that helps detect change or weakness. Examples include overdue regulatory actions, repeated documentation exceptions, delayed complaint resolution, unexplained overrides and failed reconciliations. A useful indicator has a clear definition, reliable source, meaningful denominator, owner and response threshold. A threshold should prompt a specified action, not merely change the colour of a dashboard. Some events require immediate attention regardless of percentage. An aggregate indicator should therefore be accompanied by information on critical individual cases.

Consider two months with ten exceptions each. If activity fell from one thousand to two hundred cases, the observed rate rose from 1% to 5%. If the selection method changed, even that comparison needs qualification. Counts, rates and severity should be interpreted together. Leading indicators, such as a growing control backlog, can warn of future failures; lagging indicators, such as confirmed breaches, describe events already discovered. Both are useful. Avoid target designs that encourage staff to suppress issues or relabel them to remain below a threshold. Review unusual improvements as well as deterioration. A sudden fall to zero complaints after a reporting change may reflect broken capture rather than a genuine improvement in customer experience.

9. Independent testing and assurance

Testing examines whether specified requirements and controls operate as intended. Independence helps ensure that an unfavourable result can be reported without the tester protecting their own work. The degree and form of independence depend on the function and mandate, but the underlying concern is self-review and undue influence. Compliance testing and internal audit should coordinate while preserving their distinct roles. Testing plans need risk-based coverage, clear criteria, reliable populations and documented methods. Repeating last year's sample without considering changes can miss newly important exposure.

An original test of customer communications might include different products, channels, languages, exception paths and dates around a system change. The reviewer checks what the customer actually received, not just what the template contains. If evidence is missing, investigate whether the communication failed or the retention process failed. Both may matter, but they are different findings. Report the limits of the work and avoid universal conclusions from a narrow sample. Independent testing also applies to the compliance function itself: its risk assessment, change management, reporting and closure processes should be subject to appropriate review. A specialist control function is not exempt from assurance simply because its purpose is to oversee others.

10. Inspection and audit as information sources

Inspection and audit can reveal transaction-level errors, systemic weaknesses and discrepancies between reported and actual practice. Their value depends on the quality of the response after the report is issued. A bank should map observations to underlying processes and obligations, identify common causes and allocate action owners. Treating each observation as an isolated administrative task can produce repetitive closure letters while the same defect appears in later inspections. The response should distinguish correcting the reviewed case from fixing the wider control weakness.

Suppose an inspection identifies incomplete customer records in one branch. The bank corrects those files, but the cause is an onboarding form used throughout the network. A complete response examines wider exposure and changes the form and validation process. Compliance monitors relevant observations and supports sustainable closure; operational owners implement remedies; internal audit provides independent follow-up within its mandate. Reports from different assurance providers should be reconciled where they describe the same weakness. Duplicate records can otherwise create conflicting owners and closure dates. Conversely, apparent duplication should not be used to delete a distinct finding with a different scope. Maintain traceability from each observation to the common root cause and the evidence supporting each required response.

11. Loan review mechanism and credit audit

A loan review mechanism provides a structured assessment of credit quality and the effectiveness of lending processes. Credit audit examines relevant aspects of appraisal, sanction, documentation, disbursement, monitoring and adherence to policy or requirements. These activities connect credit risk with compliance risk, but they are not identical. A loan can deteriorate because a borrower's business fails even when procedures were followed. A loan can remain fully performing while documentation, authority or end-use controls are defective. Therefore, repayment status alone cannot establish compliance quality.

An original review finds that a performing borrower's limit was renewed without an updated analysis required by the bank's process. The absence of default does not erase that weakness. The reviewer should examine the applicable criterion, actual evidence, authority and possible wider pattern. Conversely, an overdue loan should not automatically be labelled a compliance failure or fraud without investigating facts. Credit review can use portfolio indicators, exception reports, rating changes and account-level examination to identify areas needing attention. Findings should feed back into credit policy, training, approval practices and monitoring. The compliance perspective asks which obligations were met or missed; the credit perspective asks what drives repayment risk. Combining them improves understanding without collapsing their different purposes into one label.

12. Reporting and prioritisation

Issue reporting should make decisions easier. A useful record identifies the obligation, observed condition, evidence, extent, severity, cause, action, owner and target date. It distinguishes confirmed facts from estimates and unresolved questions. Prioritisation considers potential harm, ongoing exposure, legal urgency, recurrence and the reliability of controls. A large number of minor items should not obscure a small number of critical ones. Equally, a dramatic description without evidence can distort resource allocation. Use consistent definitions and allow justified escalation when the standard rating does not capture the significance.

An original report on fee errors should identify whether charging continues, how many customers may be affected, which period is covered and what containment is in place. The financial amount may be modest while the systemic nature is serious. Management needs enough information to decide on resources, customer correction and any required external communication. A dashboard should show ageing and dependencies, including actions waiting on technology or another department. If a target date changes, preserve the original date and the reason for revision. Otherwise, repeatedly extending dates can make persistent failure disappear from overdue statistics. Reporting should create accountability for resolution rather than a competition to maintain favourable presentation.

13. Remediation and evidence of good compliance

Good compliance is demonstrated through reliable operation and honest response to weaknesses. It includes clear requirements, capable owners, proportionate controls, accessible advice, timely escalation and verified remediation. It does not mean an institution never discovers a problem. In fact, effective detection may initially increase reported issues. The important questions are whether material matters are identified promptly, affected people are treated appropriately and the underlying causes are addressed. A culture that hides issues can look successful until a larger failure becomes unavoidable.

Closure should require evidence matched to the finding. If the issue concerns a system rule, evidence includes tested implementation and relevant transaction results. If it concerns missing historical records, evidence includes the completed population and unresolved exceptions. If it concerns staff understanding, training attendance alone may be insufficient; observed performance or targeted testing can provide stronger support. Independent verification is particularly valuable for significant or repeated failures. After closure, follow-up monitoring tests sustainability. A recurring defect may indicate that the original cause was misunderstood or that the control deteriorated. Record that lesson and update the risk assessment. Good compliance is therefore a continuing feedback process, not a one-time certificate issued when an action spreadsheet becomes empty.

14. Integrated identification case

Meadow Bank sees a rise in complaints about delayed release of loan documents. The branch team attributes the complaints to customer impatience. A wider review combines complaint dates, closure records, document-custody logs and technology tickets. It finds that a system change stopped sending notifications to a central custody team. The first issue is the failed handoff; further work is needed to identify all affected customers and determine the applicable obligations and consequences. The complaint sample alone is not the full population.

The bank assigns an operational owner, restores the notification route through authorised changes and reviews historical closures. Compliance interprets the relevant requirements, challenges the population and monitors remediation. Credit review considers whether the weakness affects related security-release processes. Internal audit can assess the system-change controls and the adequacy of the response. Reporting includes ongoing exposure, customer impact, unresolved cases and verification evidence. An illustrative loss calculation might help compare scenarios, but it cannot determine whether the bank may leave a binding obligation unmet. This case shows why identification requires several information sources, a process view and careful distinctions between symptoms, causes, risks and confirmed failures. The best response fixes the journey rather than merely replying to the customers who complained.

15. Revision and examination reasoning

For revision, practise writing one precise risk statement for a familiar banking process. Identify the obligation, possible failure, cause, consequence, preventive control, detective control and evidence of operation. Then ask how the risk could change after a new product, system migration or staff turnover. This exercise makes the concepts concrete and exposes weak statements that merely rename the process. It also helps distinguish inherent risk from control weakness and residual exposure. Keep qualitative judgement visible even when numerical tools are used.

In examination questions, avoid assuming that no loss means no risk, that a performing loan proves compliant lending or that a high closure percentage proves effective remediation. An alert is not a finding, a finding is not necessarily fraud, and a management response is not verified closure. When given a formula, state its assumptions and units. Probability percentages must be converted to decimals, and the control-reduction input must be interpreted consistently. In descriptive cases, choose the response that investigates evidence, assigns ownership, addresses wider impact and verifies the remedy. The overall framework is a loop: understand obligations, identify exposure, assess controls, test, report, act and learn. Each completed loop should improve both the risk register and the way the underlying banking process actually operates.

Open the interactive chapter exercise and offline practice

LEARN THROUGH EXPLANATIONS

Chapter practice

This practice is separate from timed mock tests. Answers and explanations appear after you choose an option.

Your reading progress

Sign in to save your progress across devices.

Browse published timed tests for this paper