A borrower in Chakan questioned the stock auditor: "You only counted 180 items out of 2,400 SKUs in our warehouse. How can you say the stock is Rs 4.2 crore when you did not count everything?" The auditor explained that the 180 items selected through value-weighted stratified sampling represented Rs 3.78 crore -- 90% of the total stock value. The remaining 2,220 SKUs collectively represented only Rs 42 lakh. By verifying 90% of the value through 7.5% of the SKUs, the auditor obtained sufficient evidence to conclude on the entire population.
Sampling is not a shortcut -- it is a scientifically designed approach that provides reasonable assurance about the entire inventory population without the impracticality (and cost) of counting every item. Understanding how auditors sample is important for businesses because it explains why certain items are always counted, why others are sometimes counted, and why sampling-based conclusions are reliable enough for banks, statutory auditors, and tax authorities.
This guide explains the sampling methods used in stock audits, the SA 530 framework, how ABC analysis drives sample selection, practical worked examples, what businesses should prepare, and why sampling-based stock audits are accepted by banks and regulators.
Why Auditors Use Sampling in Stock Audits
Sampling is the application of audit procedures to less than 100% of items in a population, such that all sampling units have a chance of selection, providing the auditor with a reasonable basis to draw conclusions about the entire population (SA 530, Para 5(a)).
In stock audits, 100% physical verification of every SKU at every location is often impractical. A manufacturer with 5,000 SKUs across 3 warehouses and 2 job workers would require 15-20 auditor-days for complete enumeration -- a cost and time investment that exceeds the benefit for most audits. Sampling allows the auditor to verify the items that matter most (by value, risk, or materiality) and draw reliable conclusions about the rest.
Businesses engaging professional stock audit services should understand that a well-designed sample covering 70-90% of inventory value provides assurance equivalent to -- and sometimes better than -- a rushed 100% count that sacrifices accuracy for completeness.
Key Terms
- SA 530 (Audit Sampling): The Standard on Auditing issued by ICAI governing the design, selection, and evaluation of audit samples. Applies to both statutory and stock audits.
- Population: The entire set of inventory items from which the sample is selected. In a stock audit, the population is all SKUs in the stock register or ERP.
- Sampling Unit: An individual item in the population -- e.g., a single raw material, a finished goods batch, or a line item in the stock register.
- Sampling Risk: The risk that the auditor draws a wrong conclusion from the sample -- either missing a discrepancy that exists, or flagging a discrepancy that does not exist at the population level.
- Tolerable Misstatement: The maximum misstatement in the population that the auditor is willing to accept and still conclude that the audit objective is achieved. For bank stock audits, this is typically 1-5% of total stock value.
- ABC Analysis: Classification of inventory into three categories: A items (top 10-20% of SKUs representing 70-80% of value), B items (next 20-30% representing 15-20% of value), and C items (remaining 50-70% representing 5-10% of value). The Pareto principle applied to inventory.
Who Needs to Understand Sampling in Stock Audits?
- Bank borrowers whose stock is audited by a bank-appointed CA -- understanding sampling prevents unnecessary alarm when not every item is counted
- CFOs and finance controllers responsible for coordinating stock audits -- they must prepare data that enables effective sampling
- Warehouse and operations managers who facilitate the physical count -- they need to know which items will be verified and how to organise access
- Statutory auditors relying on stock audit reports for CARO 2020 compliance -- they must evaluate whether the sampling methodology was adequate
- Internal auditors designing their own inventory verification programmes -- sampling methods are directly applicable
Manufacturers in Pune requiring CA-led stock audits with rigorous sampling methodology can explore stock audit in Pune -- our audit approach combines 100% verification of Category A items with stratified sampling of B and C categories.
Stock Audit Sampling Methods: Complete Comparison
The following table compares the principal sampling methods used in stock audits, with practical application examples.
| Method | How It Works | When Used in Stock Audit | Advantage | Limitation |
|---|---|---|---|---|
| ABC Analysis + 100% / Sampling Hybrid | 100% count of A items (high value); sampling of B and C items | Most common approach for bank and statutory stock audits | Covers 70-90% of value by counting only 20-30% of SKUs | Requires reliable ERP data to classify A/B/C accurately |
| Value-Weighted Selection | Items with higher monetary value have proportionately higher chance of selection | Selecting additional samples beyond Category A; testing for overstatement | Focuses effort on items where misstatement has the greatest financial impact | May miss errors in large volumes of low-value items |
| Stratified Sampling | Population divided into sub-groups (strata) by category, location, or value; sample drawn from each stratum | Multi-location audits; audits with diverse inventory types (raw, WIP, FG, spares) | Ensures representation across all categories and locations | Requires clear classification criteria; more complex to design |
| Random Selection | Every item in the population has an equal chance of being selected (computer-generated random numbers) | Testing completeness of stock register; verifying that low-value items are accurately recorded | Eliminates selection bias; statistically valid for projection | May select trivial items while missing important ones if not combined with stratification |
| Systematic Selection | Select every nth item from the stock register after a random starting point | Quick selection from large ERP-generated stock lists; cycle counting | Easy to implement; provides even coverage across the population | If the population has a pattern (e.g., every 10th item is from the same category), systematic bias can occur |
| Haphazard Selection | Items selected without a formal method -- auditor picks items across the warehouse | Supplementing primary sampling; quick spot-checks during walk-through | Flexible; can target areas of concern | Not statistically valid; cannot be used as the sole method for drawing conclusions |
| Block Sampling | All items in a specific block are selected (e.g., all items in Rack 5, Row 3) | Testing a specific area flagged as high-risk; verifying a particular production batch | Complete coverage of a targeted area | Results cannot be projected to the entire population |
For details on what bank auditors specifically check during stock verification, see our guide on bank stock audit RBI guidelines.
ABC Analysis: The Foundation of Stock Audit Sampling
ABC analysis is the most widely used framework for determining which items to count during a stock audit. It applies the Pareto principle (80/20 rule) to inventory -- a small number of SKUs typically represent a large proportion of total inventory value.
| Category | % of SKUs (Typical) | % of Value (Typical) | Stock Audit Approach | Example (Manufacturer with 2,400 SKUs, Rs 4.2 Cr stock) |
|---|---|---|---|---|
| A -- High Value | 10-20% | 70-80% | 100% physical verification -- every item counted | 240 SKUs = Rs 3.36 Cr (80%); all 240 counted |
| B -- Medium Value | 20-30% | 15-20% | Sampling -- 30-50% of items verified | 600 SKUs = Rs 63 L (15%); 200 items sampled |
| C -- Low Value | 50-70% | 5-10% | Sampling -- 10-20% of items verified; focus on anomalies | 1,560 SKUs = Rs 21 L (5%); 150 items sampled |
| TOTAL | 100% | 100% | 590 items counted out of 2,400 = 25% of SKUs covering 90%+ of value |
Key Insight: By counting 590 items (25% of SKUs), the auditor physically verifies Rs 3.78 crore+ (90%+ of total stock value). This is more effective than counting all 2,400 SKUs superficially -- the deep verification of high-value items catches material discrepancies that a rushed full count might miss.
What Determines Sample Size?
| Factor | Effect on Sample Size | Explanation |
|---|---|---|
| Population size | Moderate increase for larger populations | A warehouse with 10,000 SKUs needs a larger sample than one with 500 -- but not proportionately larger |
| Materiality threshold | Lower materiality = larger sample | If the auditor wants to detect discrepancies above Rs 50,000 (vs Rs 2 lakh), more items must be tested |
| Assessed risk of misstatement | Higher risk = larger sample | If the business has poor internal controls, history of discrepancies, or first-time audit, the sample is larger |
| Internal control strength | Stronger controls = smaller sample | If the business uses ERP with real-time stock tracking and regular cycle counting, the auditor can rely more on controls and sample less |
| Number of locations | More locations = larger total sample | Each location needs its own sample to verify location-specific accuracy; total sample is the sum across locations |
| Tolerable deviation rate | Lower tolerance = larger sample | For bank audits where even 1-2% discrepancy affects drawing power, the sample must be larger to detect small errors |
| Expected error rate | Higher expected errors = larger sample | If previous audits found significant discrepancies, the auditor designs a larger sample anticipating similar issues |
How a Stock Auditor Designs the Sampling Plan: Step-by-Step
1. Define the population and audit objective. The population is all inventory at all locations as of the audit date. The objective is to obtain sufficient evidence that the stock register balance is materially correct (for statutory audit) or that the drawing power computation is based on accurate stock values (for bank audit).
2. Obtain the complete stock register and classify using ABC analysis. Request the stock register (ERP or Tally) as of the audit date. Sort items by value (quantity x unit cost). Classify into A (top 70-80% of value), B (next 15-20%), and C (remaining 5-10%). This classification drives the sampling approach.
3. Determine materiality and tolerable misstatement. For statutory audits, materiality is typically 1-5% of total inventory. For bank audits, the tolerable misstatement is the amount that would cause drawing power to change materially -- typically Rs 1-5 lakh for smaller borrowers, Rs 10-50 lakh for larger ones.
4. Design the sample for each stratum. Category A: 100% verification. Category B: 30-50% sampling (random or value-weighted). Category C: 10-20% sampling (random or systematic). Additional items may be selected based on risk factors -- new products, items with recent price changes, items with prior discrepancies, items at remote locations.
5. Select the specific items using the chosen method. Use random number generation (for random selection), interval calculation (for systematic), or value-weighted ranking (for MUS). The selection must be documented -- the auditor records the method, the population, and the selected items before visiting the warehouse.
6. Execute the physical count and compare. Count the selected items at each location. Compare physical quantity against book quantity. Record discrepancies. Investigate the cause of each discrepancy -- is it a counting error, recording lag, pilferage, or measurement difference?
7. Project findings to the population and conclude. If discrepancies found in the sample are material when projected to the full population, the auditor extends the sample or requests explanation. If projected misstatement is within tolerable limits, the auditor concludes that the population is materially correct.
Common Mistakes Businesses Make During Sampled Stock Audits
Mistake 1: Assuming sampling means the auditor is cutting corners. Sampling is a professional methodology governed by SA 530 -- not a shortcut. A well-designed sample covering 90% of value is more reliable than a hurried 100% count. Businesses should focus on preparing accurate stock data, not questioning the sampling methodology.
Mistake 2: Not having a complete stock register available on audit day. If the stock register is incomplete or not updated to the audit date, the auditor cannot classify items or design a representative sample. This delays the audit and may result in a larger sample (higher cost and time) or an adverse observation.
Mistake 3: Hiding items or restricting warehouse access during audit. Auditors specifically note if access to certain areas was restricted. Unverified stock is excluded from eligible stock for bank drawing power. If the business has nothing to hide, full access to all locations -- including job workers -- should be provided.
Mistake 4: Expecting the auditor to count every screw and washer. For Category C items (high volume, low value), counting every single item is impractical and unnecessary. The auditor may use weighment-based estimation (weighing a box of 10,000 screws vs counting them) -- this is an accepted technique that saves time while providing reliable results.
Mistake 5: Not reconciling ERP data before the audit.If the stock register has unposted GRNs, pending dispatches, or unrecorded transfers, every sampled item will show a discrepancy -- not because of real inventory problems but because of recording lag. Reconcile ERP data up to the audit date before the auditor arrives. For common audit findings, see our guide on common stock audit deficiencies.
Bank Stock Audit vs Statutory Audit: How Sampling Differs
| Feature | Bank Stock Audit Sampling | Statutory Audit (CARO 2020) Sampling |
|---|---|---|
| Primary objective | Verify collateral value for drawing power | Verify closing stock for financial statements |
| Who designs the sample | Bank-appointed CA (independent of borrower) | Statutory auditor (or relies on management count under SA 501) |
| Category A treatment | 100% verification -- bank relies on every high-value item | 100% or high coverage -- statutory auditor tests material items |
| Sample size driver | Drawing power sensitivity -- how much DP change can a discrepancy cause? | Materiality -- would a discrepancy change the financial statements? |
| Tolerable misstatement | Tight -- even 2-3% affects DP and may trigger excess drawing | Broader -- typically 1-5% of total inventory |
| Location coverage | Must cover all locations in the hypothecation agreement | Must cover all material storage locations |
| Scrap and rejection | Specifically counted and excluded from DP | May be verified for NRV write-down compliance |
| Result if sample shows errors | DP reduced; excess drawing notice if material | Audit report may be qualified if material |
How Sampling Connects with Other Audit Standards
SA 530 (Sampling) works in conjunction with SA 501 (Audit Evidence -- Inventory), SA 500 (Audit Evidence), and SA 315 (Risk Assessment). The statutory audit combines these standards: SA 315 identifies inventory as a risk area, SA 530 designs the sample, SA 501 requires the auditor to attend the physical count, and SA 500 evaluates whether the evidence obtained is sufficient.
For bank stock audits, the ICAI Technical Guide on Stock and Receivable Audit provides additional sampling guidance specific to borrower audits -- including ABC analysis application, drawing power-based materiality, and location-specific sampling requirements. The bank credit policy may also prescribe minimum coverage percentages (e.g., "verify at least 80% of stock value") that the auditor must meet regardless of the SA 530 sample design.
Statistical vs Non-Statistical Sampling: Which Is Used?
| Feature | Statistical Sampling | Non-Statistical (Judgmental) Sampling |
|---|---|---|
| Method | Mathematical formulas determine sample size; random selection | Professional judgement determines sample size and selection |
| Bias | Minimal -- every item has a calculable probability of selection | Possible -- auditor may unconsciously select items that "look right" |
| Projection | Misstatements can be statistically projected to the population with measured confidence | Misstatements cannot be mathematically projected; auditor uses judgement to conclude |
| Documentation | Requires formal documentation of confidence level, tolerable misstatement, expected error rate | Less formal but must document rationale for selections |
| When used in stock audits | Large inventories (5,000+ SKUs); bank stock audits requiring defensible conclusions | Smaller inventories; internal audits; supplementary testing within a statistical framework |
| Recommended for | Bank borrower audits; listed company audits; any audit where the conclusion may be challenged | SME internal audits; supplementary spot-checks |
In practice, most stock audits use a hybrid approach: statistical methods for Category A (100%) and Category B (stratified random sampling), combined with non-statistical judgmental selection for Category C (auditor picks items based on risk factors like location, age, or anomalies in the stock register).
Key Takeaways
Stock audit sampling is governed by SA 530 and the ICAI Technical Guide on Stock and Receivable Audit. It is not a shortcut -- it is a scientifically designed methodology that provides reasonable assurance about the entire inventory population by verifying a representative subset.
The most common sampling approach is ABC analysis-based: 100% verification of Category A items (high value), 30-50% sampling of Category B, and 10-20% sampling of Category C. This typically covers 90%+ of inventory value by counting only 20-30% of SKUs -- applying the Pareto principle to audit efficiency.
Sample size depends on: population size, materiality, assessed risk, internal control strength, number of locations, and tolerable misstatement. For bank stock audits, the tolerable misstatement is driven by drawing power sensitivity -- even small discrepancies can affect available credit.
Businesses can facilitate effective sampling by: maintaining an accurate and up-to-date stock register (ERP/Tally), providing the complete register to auditors in advance for ABC classification, ensuring all locations are accessible on audit day, and reconciling ERP data (unposted GRNs, pending dispatches) before the audit.
Statistical sampling (formal, mathematically defensible) is preferred for bank stock audits and large inventories. Non-statistical sampling (judgmental) supplements the statistical approach for targeted testing of high-risk areas. Most stock audits use a hybrid of both methods.
Need a Professional Stock Audit with Rigorous Sampling?
A stock audit is only as reliable as the sampling methodology behind it. Our CA-led teams use ABC analysis-based stratified sampling, 100% verification of high-value items, and SA 530-compliant documentation -- providing audit reports that banks, statutory auditors, and management can rely on with confidence.
Explore our stock audit services -- on-site physical verification with documented sampling methodology, value-weighted selection, drawing power computation, and bank-format reporting. Available across India.
For queries, reach out at +91 945 945 6700 or WhatsApp us directly.
