Projecting misstatements and evaluating sample results

In this lesson you will

Investigate the nature and cause of misstatements under AU-C 530.12.
Project misstatements using ratio and difference methods.
Evaluate results against tolerable misstatement and decide on next steps.
Carry projected misstatement to the AU-C 450 evaluation.

Investigate before you compute

AU-C 530.12 requires you to investigate the nature and cause of any deviation or misstatement and evaluate its possible effect on the purpose of the procedure and other areas. AU-C 530.A22 adds that when many misstatements share a feature (a location, a product line, a period), you may identify all items with that feature and extend procedures to them. Misstatements may also be intentional and indicate fraud.

AU-C 530.A23 lists qualitative aspects to consider: whether a misstatement is a difference in principle or in application, an error or fraud, or a misunderstanding of instructions; and how it relates to other phases of the audit. A $500 cutoff error and a $500 forged invoice project the same arithmetic but lead to very different conclusions.

Projection is required

AU-C 530.13 requires you to project the results of audit sampling to the population. AU-C 530.A24: for tests of details, you project misstatements to obtain a likely misstatement, which may not be enough to determine an amount to record because of sampling risk.

Common methods for nonstatistical or CVS-style samples:

Ratio projection. Sample misstatement / sample book value x population book value. Fits when misstatements are roughly proportional to item size.
Difference projection. Average misstatement per sample item x number of items in the population. Fits when misstatements are similar in size regardless of item value.

For MUS, projection is tainting x interval per item, as covered in monetary unit sampling.

Project only the sampled stratum. Add actual misstatements from key items tested 100% without projecting them.

Evaluating with sampling risk

AU-C 530.14 requires you to evaluate the sample results, including sampling risk, and whether sampling provided a reasonable basis for conclusions about the population.

AU-C 530.A27: as projected misstatement approaches or exceeds tolerable misstatement, the likelihood that actual misstatement exceeds tolerable increases. If projected misstatement exceeds the expected misstatement used to size the sample, you may conclude there is unacceptable sampling risk.

Statistical samples give the allowance directly through an upper limit. Nonstatistical samples need judgment: a projection that is a small fraction of tolerable misstatement leaves room; one near tolerable does not.

When results are not acceptable

AU-C 530.A26 says an unexpectedly high misstatement amount may cause you to believe the class or balance is materially misstated, absent further evidence. AU-C 530.A28 lists options:

request management to investigate identified misstatements and the potential for more, and make necessary adjustments; or
tailor further procedures, for example by extending the sample or performing other procedures.

If management corrects the population after investigating, AU-C 450.08 requires additional procedures to determine whether misstatements remain.

Carrying results to AU-C 450

Projected misstatements are accumulated with factual and judgmental misstatements (AU-C 450.A6). Correcting the specific sample errors does not eliminate the projected misstatement for the untested part of the population. Record on the summary schedule:

the known misstatement in the sample, and whether corrected;
the projected misstatement for the untested remainder.

The summary of uncorrected misstatements keeps both. The evaluating sample results guide has more cases.

Common findings

Only known misstatements posted; projection omitted.
Projection applied to the full balance including key items.
No investigation of cause before concluding.
Results near tolerable accepted without further work.

Worked example

Projection that forces more work

Evergreen Home Health, Inc. tests billing accuracy on patient receivables. Remainder population after key items: $2,250,000 in 1,900 claims. Tolerable misstatement $65,000. Expected misstatement used to size the sample: $10,000. Nonstatistical sample: 60 claims totaling $90,000.

Findings: four overbillings totaling $2,160. All four involve the same therapy code billed at the old rate after a January payer contract change.

Ratio projection: $2,160 / $90,000 = 2.4%; 2.4% x $2,250,000 = $54,000. Difference projection: $2,160 / 60 = $36 per claim; $36 x 1,900 = $68,400.

Misstatements scale with claim value, so the team uses the ratio projection: $54,000. That is 83% of tolerable misstatement and more than five times the $10,000 expected, so under AU-C 530.A27 the team concludes sampling risk is unacceptable.

Because the misstatements share a feature (AU-C 530.A22), the team asks management to rerun all claims with that code after January. Management identifies 412 claims overbilled by $49,800 and records a credit. The team reperforms the query, tests 15 corrected claims, and records a remaining projected misstatement for other codes based on a follow-up sample.

Self-check

Answer each question before you open the answer.

What does AU-C 530.13 require after testing a sample for a test of details?
Show answer
Projecting the misstatements found to the population to obtain a likely misstatement.
When would you prefer ratio projection over difference projection?
Show answer
When misstatements are roughly proportional to item size. Difference projection fits when misstatements are similar in amount regardless of item size.
Projected misstatement is close to tolerable misstatement in a nonstatistical sample. Can you accept the population?
Show answer
Not without more work. AU-C 530.A27 says the closer projected misstatement is to tolerable, the more likely actual misstatement exceeds it. Extend testing, perform other procedures or ask management to investigate and adjust.
Management corrects the specific sample errors. Does the projected misstatement go away?
Show answer
No. The projected misstatement relates to the untested part of the population and stays on the summary schedule unless further procedures or a population-wide correction address it.

Sources

Back to the syllabus

Standards are paraphrased from the primary sources listed above; read the source before relying on a requirement. This lesson is training material, not professional advice. Report an error