Are you using substantive analytics in your audits? Many auditors rely solely on a test of detail when a better option is available. Substantive analytics, in some cases, provide better evidential matter. And they are often more efficient than a test of detail.
This article focuses on substantive analytics. But before we look at what substantive analytics are and how we can use them, let's see how analytics in general are used in an audit.
Analytics in Three Stages
Auditors use analytics in three stages of the audit:
Preliminary analytics are performed as a risk assessment procedure. We use them to locate potential misstatements. If we identify unexpected changes, we plan a response for that difference. For example, if we expect cost of goods sold to go down 5% but our planning analytics reveal an 8% increase, then we plan a response to determine why the change moved in an unexpected manner.
At the completion of the audit, we use final analytics to determine if we have addressed all risks of material misstatement. Here we put our numbers side-by-side and ask, "Have I dealt with all risks of material misstatement?" If yes, fine. If not, then we may need to perform additional substantive procedures.
So, how do we use substantive analytics? As a substantive procedure.
AU-C 330, Performing Audit Procedures in Response to Assessed risk and Evaluating the Audit Evidence Obtained, defines two substantive procedures:
- Tests of details
- Substantive analytical procedures
Substantive analytics can, in certain cases, be more effective and efficient than a test of details.
For example, if the profit margin has been in the range of 46% to 49% for the last five years, then you might decide to use that substantive analytic to prove accuracy and occurrence (assertions) of the cost of goods sold in the current year. (This will probably be more effective than vouching 50 invoices—a test of details--and will certainly take less time.) If you compute the ratio for the current year and it’s 47%, then you have sound evidence that cost of goods sold is accurate and that the transactions occurred.
Are there audit areas where substantive analytics should not be used alone? Yes. When the area is a significant risk. A test of details must be performed in relation to significant risks. A significant risk example is the allowance for loan losses in a bank. It is a highly complex estimate. Therefore, a test of details is required. The auditor could not, for example, just compare the allowance percent to prior years, though such a comparison could be added to the tests of details.
Now let's consider how auditors use tests of details and substantive analytics to respond to risk.
Responses to Risks of Material Misstatement
Many auditors use a test of details without performing substantive analytics. Why? For many, it's simply habit. We've always tested bank reconciliations, for example. But maybe we've never used analytics to prove revenues or expenses. I think this is the result of the old-school balance sheet audit approach.
Tests of details examples include:
- Testing a bank reconciliation
- A search for unrecorded liabilities in payables
- Confirming cash or debt or investments
- Vouching additions to plant, property and equipment
Tests of details are usually used in relation to balance sheet accounts such as cash or accounts payable.
Substantive analytics, on the other hand, are usually more fitting for income statement accounts such as revenue or expenses.
So, if you’re planning a response for accounts payable (a balance sheet account) and expenses (an income statement account), you might use a combined approach. A test of details for accounts payable (e.g., search for unrecorded liabilities) and substantive analytics for expense (e.g., departmental expenses divided by total expenses compared to the prior year).
One overarching principle to consider in your use of substantive analytics: use them in lower risk areas. AU-C 330 tells us that substantive analytics alone are more appropriate when assessed risk is lower. The higher your risk assessment, the more you should use tests of details.
Examples of Substantive Analytics
Here are examples of substantive analytics:
- Comparison of monthly sales for the current year with that of the preceding year (to test occurrence)
- Comparison of profit margins for the last few months with those subsequent to year-end (to test cutoff)
- Percent of expenses to sales compared with the prior year (to test occurrence)
- A comparison of balance sheet accounts with total assets compared to prior year (to test existence for assets and completeness for liabilities)
- Current ratio compared to prior year (to test for solvency and going concern)
- Comparing current year profit margins with prior periods (to test accuracy and occurrence)
- For pension or postemployment benefit plans: actuarial value of plan assets divided by actuarial accrued liability compared to prior year (to test completeness and accuracy)
- For debt: total debt divided by total assets compared to prior year (to test the financial strength of the entity and going concern)
- For inventory: cost of goods sold divided by average inventory compared to prior year (to test existence and occurrence)
Now let's see how to document your substantive analytics.
Documentation of Substantive Analytics
In performing substantive analytics, make sure you document your expectations and conclusions:
Expectation – Document what you expect the result of the computation or comparison to be (you can use a range).
A common peer review finding is the lack of a documented expectation. Prior to computing a ratio or comparing numbers to prior periods, document your expectation.
Conclusion – Document whether the computation or comparison falls within your expectation. If it does not, inquire of the client. You may need to perform a test of details if the substantive analytic result is not within an acceptable range. Regardless, make sure you respond to unexplained results (i.e., those that fall outside an acceptable range) and that you document your response.
Overall Substantive Analytical Considerations
Substantive analytics are not required. So, think of them as an efficient alternative to test of details.
But are there audits where substantive analytics don't work as well? Yes. If the company has weak internal controls or a history of significant errors, you may want to rely more on tests of details. Substantive analytics work better in stable environments.