## Techniques for Substantive Analytical Procedures

Four main techniques are used when AP is applied as a substantive procedure:

---

### i) Trend Analysis

Compare current period data with one or more prior periods (or with trends across multiple prior periods).
Identifies unexplained changes in account balances over time.
Best for: Revenue, expense, and balance sheet line items with stable historical patterns.

> Example: Comparing salary expense of the current year with the previous year. An increase disproportionate to headcount growth signals a potential misstatement.

---

### ii) Ratio Analysis

Examine mathematical relationships between account balances (e.g., asset-to-liability ratios, expense-to-revenue ratios).
Best for: Asset/liability accounts and revenue/expense accounts.
Key insight: An individual account balance is hard to predict in isolation, but its ratio to a related account is far more stable and predictable.

> Example: Gross profit margin should be stable year-on-year for a business with consistent pricing. A sudden drop from 40% to 25% signals possible overstatement of COGS or understatement of revenue.

---

### iii) Structural Modelling (Regression / Statistical Modelling)

A statistical tool that builds a predictive model using financial or non-financial data from prior accounting periods.
The model is then used to predict current period account balances; recorded amounts are compared to predictions.
More sophisticated than trend or ratio analysis; useful when multiple variables drive an account balance.

> Example: Building a regression model where electricity expense = f(production volume, season, energy prices). If actual expense deviates significantly from the model's prediction, further investigation is warranted.

---

### iv) Reasonableness Tests

Unlike trend analysis, does not rely primarily on prior-period data.
Instead, relies predominantly on non-financial data (independent, current-period information) to form an expectation.
Best for: Accounts that have a strong, direct relationship with a measurable non-financial variable.

Example	Non-Financial Driver
Sales discounts / commissions	Sales volume
Rental revenue	Occupancy rate
Interest expense	Interest-bearing obligations × interest rate

Worked example

### Example 1

Trend Analysis: Audit of a manufacturing company. Depreciation expense was ₹15L in Year 1, ₹15.5L in Year 2, and ₹16L in Year 3. In the current year, the client reports ₹24L. No new major assets were acquired. The spike is inconsistent with the trend → potential error in depreciation rates or capitalisation.

### Example 2

Ratio Analysis: Current ratio was consistently between 1.8–2.0 over three years. This year it is 0.9. The auditor drills into current liabilities and discovers a large creditor balance that was off-balance-sheet in prior years — a completeness issue.

### Example 3

Structural Modelling: A hotel chain's room revenue is modelled as: Average room rate × Occupancy % × Number of rooms × 365 days. Predicted revenue = ₹8.2 Cr; recorded revenue = ₹6.1 Cr. The significant shortfall leads to investigation of revenue recognition or unrecorded receipts.

### Example 4

Reasonableness Test: Commission expense should be 3% of net sales (as per the agency agreement). Net sales = ₹50 Cr → Expected commission = ₹1.5 Cr. Recorded commission = ₹2.2 Cr. The difference of ₹0.7 Cr triggers a request for the commission ledger and agent agreements.

⚠️ Common exam mistakes

Treating reasonableness tests as just another form of trend analysis — the key distinction is that reasonableness tests use current non-financial data, not prior-period financial data.
Applying ratio analysis to individual accounts in isolation instead of to relationships between accounts — ratios only work when there is a logical, stable relationship between the numerator and denominator.
Assuming structural modelling is always superior — for simpler businesses, a straightforward reasonableness test or trend analysis may provide equal or better evidence at lower cost.
Forgetting that all four techniques produce only an 'expectation'; they must be compared against recorded amounts and unexplained differences must be investigated.

Reference: SA 520 — Application Material A7–A14 — SA 520 — Analytical Procedures (ICAI)

Now that you've read this — what's next?

Move from understanding → mastery in 3 clicks. Each option below picks up from this lesson's topic.

🔥 Drill 5 Qs on this → ✍️ Adaptive practice →