1. What an audit conclusion must establish
A compliance audit examines whether activity meets identified requirements and whether the arrangements supporting compliance are reliable. Its starting point is a clear criterion, not a general impression that a department looks well organised. The auditor compares evidence with that criterion, identifies exceptions, evaluates significance and communicates a supported conclusion. Audit therefore needs to connect documented intention with operating evidence. It also needs to state the limits of the work: the period, locations, products and population examined.
Consider an original audit of account-closure charges. The criterion includes the applicable rule and the bank's approved tariff. The population consists of relevant closures during a defined period. Evidence includes transaction records, customer categories, fee calculations and authorised reversals. A conclusion based only on the tariff document would establish policy design, not correct charging. Conversely, inspecting a few transactions without knowing the criterion can misclassify an authorised exception as a breach. The strongest reasoning links requirement, process, control, evidence and conclusion. This chapter develops that chain through risk-based planning, testing, reporting, financial disclosures and follow-up, with an emphasis on what an auditor can reasonably infer from the work actually performed.
2. Compliance testing and independent audit
Business teams perform controls as part of operations. Compliance may monitor and test whether obligations are being met, advise on interpretation and challenge management. Internal audit provides independent assurance over governance, risk management and controls, including the compliance function itself. These activities can use similar evidence but have different mandates. Coordination reduces duplication; it should not remove independent challenge. If internal audit designs and operates a control, its later assessment of that control creates a self-review concern. If compliance tests a process, management does not cease to own its execution.
For example, a branch checks customer documents before opening an account. A compliance review samples completed files to assess adherence to identified requirements. Internal audit examines whether onboarding governance, risk assessment, control operation and compliance oversight are effective. The same file may contribute to all three activities, but the questions and reporting routes differ. A clear assurance map records who examines which risk, with what frequency and evidence, and where gaps remain. An assertion that compliance has already checked this is not sufficient evidence. Coordination is strongest when responsibilities remain explicit and findings can still be escalated independently.
3. Risk-based internal audit
Risk-based internal audit allocates attention according to the significance and direction of risk rather than giving every location identical treatment. The audit universe includes activities, systems, functions and locations that could affect the bank's objectives. Risk assessment considers inherent exposure and the reliability of controls. High transaction volume is one indicator, but complexity, change, customer vulnerability, previous findings, staff turnover and data weaknesses can also increase risk. A small unit handling a sensitive process may deserve more attention than a large unit with stable, well-tested controls.
The cited RBI Commercial Banks, Internal Audit Function Directions of July 2026 provide the current-practice framework. Their illustrative risk matrix combines inherent and control risk, with greater attention to severe combinations and worsening trends. This is not permission to invent a numerical RBI score from arbitrary weights. In a teaching example, two branches may each have ten reported exceptions, but one processes a hundred transactions and the other ten thousand. Even the resulting rates are not enough: exception severity, selection method and control design matter. Risk-based planning asks where failure could be most harmful and where assurance is least reliable. It does not mean that low-risk areas are permanently excluded or that transaction testing disappears in favour of interviews alone.
4. Constructing the audit universe and plan
Begin by listing auditable activities in a way that matches how the institution actually operates. A digital lending journey may cross product, technology, operations, outsourced services and collections; limiting the universe to department names can miss the links between them. Record the objectives, obligations, owners, systems and significant dependencies of each activity. Use prior findings, incident data, regulatory observations, business changes and management information to assess risk. The plan should explain why some areas receive earlier or deeper work, and what coverage remains deferred.
An original annual plan might prioritise a newly migrated payment platform, a rapidly expanding distribution channel and branches with repeated unresolved findings. The current RBI framework requires at least annual risk assessment with updates for relevant changes. A major system migration can justify an interim reassessment even when the annual planning meeting is months away. Planning also considers skills: a complex automated control may require technology expertise rather than a larger number of generalist reviewers. Record changes to the approved plan and the reasons. Otherwise, resource pressure can silently remove difficult audits while the institution continues to report that its planned coverage is complete.
5. Inherent risk, control risk and audit response
Inherent risk describes exposure before considering the effectiveness of controls. Control risk concerns the possibility that controls are inadequate or fail to prevent, detect or correct relevant problems. Strong controls can reduce residual exposure, but they do not make the underlying activity inherently harmless. A high-value payment process remains intrinsically sensitive even when dual authorisation works well. Conversely, a modest process with poor access controls and repeated overrides can become an important audit priority. The distinction helps explain why both business characteristics and control evidence matter.
Suppose a central team processes high-value settlements with automated limits and independently reviewed exceptions. Another team handles low-value refunds but permits one person to create and approve entries. The first has high inherent sensitivity with potentially strong controls; the second has a significant control weakness. The audit response should examine the relevant risks in each, not simply rank them by transaction value. Test design should address whether controls are appropriately designed and whether they operated consistently. A control described in a manual is not necessarily implemented. A control implemented in software may still be bypassed through an administrator route. An auditor should examine normal operation, authorised exceptions and the possibility of management override before relying on the control.
6. Walkthroughs and evidence quality
A walkthrough follows a transaction or event from initiation to completion to understand the process and identify key controls. It is useful for discovering handoffs, manual workarounds and differences between the written procedure and actual practice. A walkthrough is not, by itself, a test that establishes consistent operation across a period. It helps design that test. Ask the operator to demonstrate the process with suitable records, then reconcile the explanation with system behaviour and retained evidence. Avoid collecting unnecessary personal information when a masked record can establish the same point.
Evidence should be relevant, reliable and sufficient for the conclusion. A system-generated report may be more useful than a recollection, but its completeness and extraction logic still need validation. A screenshot can show a screen at one moment without proving what happened earlier. A signed declaration may establish that someone attested to a condition, not that the condition was true. Independent corroboration strengthens difficult findings. In an original maker-checker test, compare user identities, timestamps, authority limits and exception logs rather than merely counting two signature fields. If a report omits cancelled or failed transactions, the auditor should understand whether those exclusions are appropriate. Evidence quality often depends on the population and data lineage before any sample is selected.
7. Sampling and the exception-rate laboratory
Define the population and sampling unit before selecting records. A unit might be an account, a transaction, a customer complaint or a daily control performance. Selection can be random, stratified or deliberately focused on high-risk cases. Each method supports different conclusions. A judgemental sample of unusual overrides is useful for investigating those overrides, but its exception rate should not be presented as an unbiased estimate for all transactions. Document the selection method, period, exclusions and reasons. Ensure that the population actually contains the activity the audit intends to examine.
The calculator reports the observed exception rate as exceptions divided by tested items multiplied by one hundred. With eight exceptions in two hundred tested records, the observed rate is 4%. It also shows the untested count when the population is supplied. If the population is five thousand, 4,800 records remain outside that sample. The tool deliberately does not extrapolate a guaranteed number of failures or certify compliance. Zero exceptions in a sample means none were found in those tested items; it does not establish zero failures in the population. A critical single exception can justify immediate escalation and wider testing even when the percentage is small. Statistical inference requires an appropriate sampling design and assumptions beyond this descriptive calculation.
8. Testing design and operating effectiveness
Design effectiveness asks whether a control, if performed as intended, addresses the identified risk. Operating effectiveness asks whether it actually performed that function during the period examined. A control can fail either test. An exception report that omits the most important transaction category is weak by design. A well-designed report that nobody reviews is weak in operation. Audit should distinguish these findings because the remedies differ: redesign the report in the first case, and correct ownership, capacity, escalation or execution in the second.
An original test of access removal illustrates the distinction. The stated control requires prompt removal of access after an employee leaves. Design testing examines whether human-resources notifications reach all relevant systems, including vendor tools and privileged accounts. Operating testing selects departures and compares notification dates, actual disablement evidence and subsequent activity. A central application may show successful removal while a separate reporting platform remains accessible. The conclusion should identify that gap rather than calling the entire process effective because most systems worked. Include negative and exceptional paths: temporary staff, transfers, extended leave and emergency access may need different treatment. The auditor should not invent an institution-wide deadline; the test criterion must come from the applicable requirement and approved process.
9. From observation to a defensible finding
A finding becomes useful when it explains the difference between expected and observed conditions, supported by evidence and an assessment of consequence. Identify the requirement, the affected process, the period and the extent established by testing. Explain the likely cause without pretending that a hypothesis is proven. A phrase such as staff negligence may be premature if the system permitted an error and training was unclear. Distinguish root cause, contributing conditions and the immediate mistake. This supports fair accountability and a remedy that prevents recurrence.
For example, twelve of sixty sampled refund transactions lacked evidence of independent approval. The finding should state how the sixty were selected, which approval requirement applied and whether missing evidence means the approval was absent or merely unrecorded. Further work may be needed to resolve that distinction. Management's response should identify corrective actions, responsible officials, target dates and dependencies. Disagreement should be recorded with the evidence and final resolution. Audit reports should avoid unnecessary disclosure of sensitive personal data while retaining enough detail for action. Both evidence discipline and clear communication are necessary.
10. Reporting and monitoring framework
Serious deficiencies require timely escalation to the appropriate authority. Routine reports then provide a structured view of findings, management responses and follow-up. The reporting framework should distinguish immediate alerts, final audit reports, periodic summaries and overdue-action escalation. It should identify who receives each report and who can challenge closure. A delayed final report must not become an excuse to withhold an urgent finding that threatens customers or the bank. Equally, preliminary concerns should be described accurately as preliminary until the supporting work is complete.
Monitoring should preserve the history of an issue. If an action is reopened after failed verification, the dashboard should not erase its original age or present it as a fresh minor item. Report recurring findings across locations, not only location-by-location closure percentages. An original example involves ten branches correcting incomplete files after inspection while the central form remains defective. A report focused solely on branch closure could miss the common cause and allow the problem to continue elsewhere. Link related issues, assign a central owner where appropriate and test the broader remedy. Management reporting should combine count, severity, age, recurrence and exposure. Averages alone can conceal a small number of important unresolved matters.
11. Disclosure requirements and applicability
Disclosure is the controlled communication of information required by an applicable framework. Banking disclosures, company-law reporting, securities-market disclosures and customer communications serve different audiences and may operate on different schedules. The compliance auditor first establishes which obligations apply to the entity and instrument. A listed bank may face securities-listing requirements in addition to banking requirements; an unlisted entity should not be assigned every listed-entity obligation by analogy. Identify the legal entity, listing status, security type, reporting period and any relevant exemption before preparing a checklist.
A useful disclosure control connects the requirement to source data, preparation, review, approval, filing and evidence of submission. It also covers corrections and subsequent developments. An accurate draft filed late may fail a timing requirement; a timely filing with incomplete information may fail a content requirement. Reconcile information across financial statements, regulatory returns, investor communications and internal records, while recognising that legitimate differences in definitions may exist. The auditor should investigate unexplained differences rather than force all numbers to match mechanically. This chapter provides the control logic, not a universal filing calendar. Exact forms, thresholds and deadlines must be taken from the applicable current instrument and the examination cutoff relevant to the learner's sitting.
12. Accounting standards in a compliance audit
Accounting standards influence recognition, measurement, presentation and disclosure. Recognition asks whether an item enters the statements; measurement determines its amount; presentation determines where it appears; disclosure explains material information and judgements. Compliance audit does not replace the statutory auditor's opinion, but it may examine whether the bank has processes to identify applicable standards, maintain accounting policies, support estimates and produce required disclosures. The applicable accounting framework must be established for the entity; one should not assume that every bank and every NBFC uses an identical framework.
Consider a hypothetical contingent obligation. A reviewer should understand the underlying facts, the applicable recognition and disclosure criteria, management's assessment and supporting evidence. It is not enough to see that a note exists somewhere in the accounts. The note may omit a material uncertainty or use an amount inconsistent with underlying records. Accounting estimates also need governance: assumptions, data sources, approvals, sensitivity and review of later outcomes. A change in estimate differs from a correction of an error, and the reporting consequences depend on the applicable framework. For study purposes, focus on the control questions and conceptual distinctions. Do not use this chapter's illustrative cases as a substitute for the precise accounting standard or entity-specific prudential direction governing a real reporting decision.
13. SEBI listing disclosures and control ownership
SEBI's Listing Obligations and Disclosure Requirements framework is a relevant source for listed entities. Its provisions cover multiple subjects, so an auditor should not treat one generic checklist as sufficient for every security and every event. Material-event assessment and periodic financial reporting illustrate different workflows. An event may require prompt evaluation and communication outside the normal reporting cycle, while financial results require controlled preparation, review and approval for a defined period. The reference register links the retrieved consolidated version and records its date rather than asserting that an old checklist remains permanently current.
An original listed-bank example involves a significant operational incident discovered after the routine disclosure committee meeting. Waiting for the next scheduled meeting without checking the applicable obligation would be poor control design. The incident owner should notify the designated disclosure function through the established escalation route, with facts distinguished from estimates. Compliance and legal specialists help determine applicability; authorised management and governance bodies perform their prescribed roles. Maintain an evidence trail showing when the bank became aware, what was assessed, who decided and what was communicated. Confidentiality and market sensitivity require controlled access, but they do not justify suppressing a required disclosure. Avoid memorising a single deadline for all events: the applicable provision, event category and current amendments govern the actual timing.
14. Worked audit case and closure test
Harbour Bank's audit team tests two hundred fee reversals from a population of five thousand. Eight lack evidence of the required independent approval. Six involve a recently introduced workflow and two involve manual exceptions. The observed sample exception rate is 4%, but the team deliberately selected more new-workflow cases than their population share. Therefore, it cannot present 4% as an unbiased population estimate. It should explain the selection, investigate the common pattern and consider additional testing targeted at the new workflow and manual-exception route.
Management proposes to circulate a reminder and close all eight findings. That response is incomplete if the workflow does not reliably require an approver. A stronger plan combines correction of affected cases, system changes, revised exception handling and later verification. Audit should assess whether missing approval evidence indicates a documentation gap, an actual approval failure or both. If the same user could create and approve an entry, the control weakness is broader than eight missing records. Closure requires evidence that the replacement control works and that relevant historical impact has been addressed. This case combines the chapter's main lessons: define the criterion, understand the population, respect sampling limits, investigate causes, report clearly and verify a sustainable remedy.
15. Revision framework and common traps
An effective revision sequence is scope, risk, criterion, evidence, conclusion, action and follow-up. At each stage ask what could invalidate the next step. An incomplete population undermines sampling; an incorrect criterion undermines exception classification; unreliable evidence undermines the finding; an unverified management response undermines closure. This chain is more useful than memorising audit terminology in isolation. It also explains why independence, competence and clear documentation matter throughout the engagement rather than only when the final report is signed.
Common examination traps include treating low risk as no audit need, treating zero sample exceptions as proof of universal compliance, and treating management acceptance as proof that an issue is resolved. Other traps equate compliance monitoring with independent internal audit, assume every listed-entity requirement applies to every bank, or confuse financial-statement disclosure with customer disclosure. For calculations, identify tested items and exceptions carefully; population size is not the denominator of a sample exception rate. In descriptive answers, distinguish design failure from operating failure and explain what evidence would support closure. The chapter's calculator intentionally provides descriptive arithmetic only. Its restraint is part of the lesson: an honest audit conclusion says what the evidence supports, what remains uncertain and what additional work is needed.
Audit sample exception rate
Descriptive sample arithmetic only. No population inference or compliance certification.
Practice and review
One answer per question. No negative marking. You can read every explanation even without JavaScript.
Optional browser-local answer saving, with no account synchronisation.
Ready: 25 questions.
References and conventions
- IIBF CAIIB Rules and Syllabus 2026: Module D seven headings, PDF page 7; examination cutoff page 4. Observed version: 2026 syllabus
- RBI Commercial Banks: Internal Audit Function Directions, 2026: Risk-based audit, inherent/control risk, testing and reporting. Observed version: Issued July 31, 2026; retrieved November 4
- SEBI LODR Regulations: consolidated July 14, 2026: Dated listing-disclosure framework; no universal filing calendar asserted. Later-version completeness not claimed. Observed version: Consolidated through July 14, 2026
Scope and cited reference pages checked 5 November 2026. Original exposition, branch cases, diagrams and question wording. Apply the cited source scope, subsequent changes and IIBF examination cutoff. Observed source versions are recorded in the reference register. Calculators are educational, not live regulatory or legal decision engines.