9.2 Monetary Unit Sampling (MUS / PPS) Mechanics & Sizing

9.2 Monetary Unit Sampling (MUS / PPS) Mechanics & Sizing

Core Principle: Monetary Unit Sampling (MUS)—alternatively designated as Probability-Proportional-to-Size (PPS) sampling or Dollar-Unit Sampling (DUS)—is a statistical sampling technique that uses monetary units (individual dollars) rather than physical items as the individual sampling units. In MUS, every individual dollar in the population has an equal probability of selection, meaning that an invoice or account containing 10,000 dollars has 10 times the probability of being selected as an invoice containing 1,000 dollars. MUS is the most widely applied statistical sampling method for substantive testing of asset balances where overstatement is the primary audit risk.

---

1. Conceptual Architecture: Individual Dollar vs. Logical Unit

Understanding MUS requires a clear conceptual distinction between the individual sampling unit and the logical sampling unit:

                                 MONETARY UNIT SAMPLING: THE SELECTION CONCEPT
                                                       |
               +---------------------------------------+---------------------------------------+
               |                                                                               |
      [ INDIVIDUAL SAMPLING UNIT ]                                                    [ LOGICAL SAMPLING UNIT ]
               |                                                                               |
      Each Individual Dollar ($1)                                                     The Entire Recorded Balance
      - Every dollar has an equal                                                     (Invoice, Customer Account, Voucher)
        probability of selection.                                                     - The entire transaction containing the
      - Total sampling units = Recorded Book Value.                                     selected dollar is examined.

How Selection Works in Practice

Consider an auditor testing an accounts receivable ledger with a recorded book value of $1,000,000. The population consists of 1,000,000 individual sampling units (each $1 bill). When the statistical plan selects dollar number 412,850, the auditor identifies which customer account or invoice contains dollar number 412,850. That entire customer account or invoice is the logical unit, and the auditor tests that entire invoice for compliance and valuation.

---

2. Advantages of Monetary Unit Sampling

Monetary Unit Sampling has become the dominant statistical tool for substantive testing in modern financial statement audits due to several powerful design features:

1. Automatic Stratification

Because each dollar has an equal chance of being chosen, items with larger dollar balances automatically have a higher probability of selection:

• An item with a book value of $50,000 is 50 times more likely to be selected than a $1,000 item.
• Guaranteed 100% Selection: Any item with a recorded book value equal to or greater than the Sampling Interval is guaranteed to be selected at least once. Items greater than twice the sampling interval will be selected multiple times. This eliminates the need for the auditor to manually create complex stratification boundaries.

2. No Estimate of Standard Deviation Required

In classical variables sampling, the auditor must calculate or estimate the population standard deviation (σ) from prior-year workpapers or a pilot sample. In MUS, no standard deviation estimate is needed; sample size comes from reliability factors based on the Poisson distribution, so the auditor does not estimate population variance before sampling.

3. Highly Efficient for Low-Error Populations

When the auditor expects few or no misstatements (which is the typical circumstance for well-controlled accounts like cash, trade receivables, or marketable securities), MUS produces a significantly smaller sample size than unstratified classical variables sampling.

4. Objective, Statistically Defensible Results

MUS yields a clear statistical conclusion expressed in monetary terms (e.g., "There is a 95% confidence level that the misstatement in accounts receivable does not exceed $72,400").

---

3. Disadvantages & Structural Limitations of MUS

Despite its popularity, MUS possesses structural constraints that make it completely unsuitable for certain audit objectives:

+-------------------------------------------------------------------------------------------------------+
|                                    STRUCTURAL LIMITATIONS OF MUS / PPS                                |
|                                                                                                       |
|   1. OVERSTATEMENT BIAS                 2. ZERO & NEGATIVE BALANCES         3. HIGH ERROR INEFFICIENCY|
|   Cannot test Understatement            Cannot select $0 or credits         Sample size balloons if   |
|   or Completeness assertions;           directly; must be removed and       frequent misstatements    |
|   favors large recorded amounts.        tested via separate procedures.     are anticipated.          |
+-------------------------------------------------------------------------------------------------------+

1. Inherent Overstatement Bias (Ineffective for Completeness)

MUS assigns selection probabilities based on recorded book values. Consequently:

• An unrecorded liability or unrecorded sale has a recorded book value of $0.00. Its probability of selection in an MUS sample is exactly zero.
• Understated accounts have an artificially depressed probability of selection.
• Conclusion: MUS is excellent for testing Existence, Occurrence, and Rights (overstatement risks), but is completely ineffective for testing Completeness or searching for unrecorded liabilities.

2. Zero Balances and Negative (Credit) Balances

• Zero Balances: Accounts with a zero recorded balance contain no monetary units and cannot be selected by the cumulative dollar run.
• Negative Balances: Negative (credit) balances in an asset population disrupt the cumulative summation mechanics and distort the sampling interval.
• Auditor Protocol: The auditor must filter out all zero and negative balances before running the MUS selection algorithm. Negative balances are evaluated separately (either through 100% audit testing or an independent classical variables sampling design).

3. Rapid Loss of Efficiency When Errors Are Prevalent

If the auditor anticipates a substantial number of misstatements across the population, the required MUS sample size expands dramatically, often becoming significantly larger than a stratified classical variables sample.

---

4. Mathematical Determination of MUS Sample Size

The required sample size in Monetary Unit Sampling is calculated using the following authoritative formula:

$n = \frac{\text{Recorded Book Value} \times \text{Reliability Factor}}{\text{Tolerable Misstatement} - (\text{Expected Misstatement} \times \text{Expansion Factor})}$

Where:

• Recorded Book Value (BV): The total recorded dollar balance of the population being sampled.
• Reliability Factor (RF): A factor determined by the auditor's acceptable Risk of Incorrect Acceptance (Beta risk) assuming zero misstatements (derived from Poisson distribution tables):
  • 5% Risk of Incorrect Acceptance (95% Confidence): RF = 3.00
  • 10% Risk of Incorrect Acceptance (90% Confidence): RF = 2.31
  • 15% Risk of Incorrect Acceptance (85% Confidence): RF = 1.90
• Tolerable Misstatement (TM): The maximum monetary misstatement the auditor will accept in the population without concluding that the financial statements are materially misstated.
• Expected Misstatement (EM): The auditor's realistic estimate of dollar misstatement in the population based on prior audits, interim testing, or risk assessment.
• Expansion Factor (EF): An adjustment factor based on the acceptable risk of incorrect acceptance that accounts for anticipated errors:
  • 5% Risk: EF = 1.60
  • 10% Risk: EF = 1.50
  • 15% Risk: EF = 1.40

Calculating the Sampling Interval (SI)

The Sampling Interval (SI) is the dollar distance between selected sample items:

$\text{Sampling Interval (SI)} = \frac{\text{Recorded Book Value}}{n}$

Alternatively, substituting the sample size formula into the interval formula yields:

$\text{Sampling Interval (SI)} = \frac{\text{Tolerable Misstatement} - (\text{Expected Misstatement} \times \text{Expansion Factor})}{\text{Reliability Factor}}$

Comprehensive Numerical Sizing Walkthrough

An audit team plans to test an inventory population with the following parameters:

• Recorded Book Value: $3,600,000
• Acceptable Risk of Incorrect Acceptance: 5% (Reliability Factor = 3.00, Expansion Factor = 1.60)
• Tolerable Misstatement: $180,000
• Expected Misstatement: $30,000

Step 1: Calculate the denominator (Net Precision Allowance):
        Denominator = Tolerable Misstatement - (Expected Misstatement × Expansion Factor)
                    = $180,000 - ($30,000 × 1.60)
                    = $180,000 - $48,000 = $132,000

Step 2: Calculate the required sample size (n):
        n = ($3,600,000 × 3.00) / $132,000
          = $10,800,000 / $132,000
          = 81.82 --> Round UP to 82 items

Step 3: Calculate the Sampling Interval (SI):
        SI = $3,600,000 / 81.82 = $44,000
        (Or directly: $132,000 / 3.00 = $44,000)

Important Exam Rule: When calculating sample size, standard audit practice mandates always rounding UP to the next whole integer (81.82 becomes 82) to preserve the required statistical confidence level.

---

5. Systematic Selection Mechanics in Practice

Once the sampling interval (SI = $44,000) has been determined, the auditor executes systematic selection using a cumulative monetary schedule:

Step 1: Select a Random Start

The auditor uses a random number generator to select a random start between $1 and the Sampling Interval ($1 to $44,000). Assume the random start is $12,500.

Step 2: Establish Cumulative Selection Targets

The auditor adds the sampling interval sequentially to identify all dollar checkpoints:

• Target 1: $12,500
• Target 2: $12,500 + $44,000 = $56,500
• Target 3: $56,500 + $44,000 = $100,500
• Target 4: $100,500 + $44,000 = $144,500
• ... and so forth until the entire $3,600,000 population is traversed.

Step 3: Match Targets to Logical Units (Cumulative Schedule)

The auditor generates a cumulative run of recorded account balances:

Key Observations from the Selection Schedule

1. Beta LLC (Item 102): The cumulative range spans $8,001 to $23,000. Because Target #1 ($12,500) falls within this range, Item 102 is selected as the logical unit.
2. Omega Global (Item 107): Recorded at $95,000, which is greater than twice the sampling interval ($44,000 × 2 = $88,000). It receives two selection hits ($144,500 and $188,500). Omega Global is audited once, but in evaluation, it accounts for two sampling units.

---

6. MUS / PPS vs. Classical Variables Sampling: Master Comparison