At this point in the inventory appraisal process it is presumed that the appraiser has already determined the scope of the assignment, the value concept to be performed, the basic classifications that are applicable, and what approach to value will be employed. Given this set of facts, the appraiser would now consider the individual items contained in the subject inventory from which global results could be developed. Key at this stage in the appraisal process is the accuracy of the inventory regarding both costs and counts.
As opposed to a full audit, which is in the domain of accounting, and is usually outside the scope of an appraisal assignment, the appraiser will typically conduct some sort of sampling or testing (or both), the purpose of which is twofold. First, a sample count provides a means of determining to what extent, if any, the actual shelf quantities depart from those shown in the supplied report. Second, a sample test proves or disproves the efficacy of internal cost accounting procedures.
1. Sample Counting: Sample counts are based on a random method for any of the inventory upon which the appraiser later performs a sample test. The objective is for the appraiser to sample count a reasonable percentage of the inventory to determine if the counts indicated, at the date of the supplied report, are reasonable to that which is observed on-site at the time of inspection.
Naturally, there can be some differences, as inventory could be moved in or out on the day of inspection. This movement can be traced to bills of lading or shipping documents. If there is an unreasonable difference between the count and the supplied report over a short length of time, the appraiser must determine the reason for the difference and confirm, through documentation, the response. Sample counting is necessary in the event the appraiser is asked, some time in the future, if the counts were reasonable based upon their investigation rather than that of an audit team or company personnel. Remember, the appraiser’s credibility can be in question if, at the very least, the inventory counts were not found to be reasonably accurate.
2. Sample Testing: This is far more extensive than sample counting, and to a different end. In the sample count, the appraiser samples enough items to determine that the supplied inventory report is either reasonable, or that any differences are justified. However, counting every item that is to be tested is not necessary for the appraiser, as the purpose of this exercise is to determine whether the indicated cost is proper. In addition, the sample test considers the quantity on-hand and all causes and effects that could be applicable to the applied value concept.
Sample testing is a more in-depth analysis of the costing method being used, and is required when the sample counts, compared with the perpetual inventory, contain costs that appear out of line. At this point, general notes and indicated amounts can be applied to each SKU identified.
There are two different identifiable reasons why sample testing and sample counting differ in the amount of sampling. With sample testing, the appraiser is attempting to sample from each category he has identified in the initial review. Once there has been enough sampling for the appraiser to feel comfortable with the supplied inventory report, a final review with the client may be necessary. This test refers specifically to verification of the inventory’s individual line item cost as indicated in the purchase invoices.
Generally, both procedures should be undertaken to some extent, though in practice sample testing is recognized in the domain of an appraisal assignment. Sample counting can be construed as more of an audit function and, if undertaken, should be highly disclaimed in the appraisal as limited in scope.
3. Statistical Sampling: As is generally known, all sampling is governed by the field of statistics, and the inventory appraiser can be expected to demonstrate some knowledge in this area. Where definable standards do not exist due to purchasing and/or accounting practices, there are methods available for adjusting indicated cost to that which is standard for the industry. One method for making this adjustment is through statistical sampling. To achieve a statistically correct sample, the appraiser must first define the population to be sampled, and then determine the tolerance level. The tolerance level is assigned according to the confidence interval the appraiser determines is appropriate, which is influenced by the characteristics of the respective inventory. For instance, a large inventory with wide variances in price would warrant a low tolerance, whereas, a small inventory displaying a narrow variance in price would not require such a strict tolerance. An inventory such as this would be best valued on an individual basis, and not through statistical sampling.
When statistically sampling an inventory, the priority is to select a random sample of the population to capture a reasonable cross section. This is accomplished when the rolling standard deviation begins to stabilize and decline, which typically occurs in a sample of fifty items. An example of tolerance levels and confidence intervals would be if a 10% tolerance level were required and/or desired, that would dictate a tolerance level of no more or less than 10% of the rolling mean. Subsequently, the rolling standard deviation (range) could be no more or less than 5% of the rolling mean. The rolling mean is the mean ratio of the adjusted book value to price, which is calculated on a line item basis, and includes stated quantities. With the assigned tolerance level of 10%, the appraiser can discontinue sampling the population when the rolling standard deviation is equal to or less than 5% of the rolling mean, which achieves a 95% confidence interval or a 10% tolerance. Therefore, it can be expected that 95% of the time, the average or rolling mean of the population will fall within the calculated confidence interval. This method is best applied when there is no definable standard cost.
Statistical sampling may be required of auditors, but is not necessarily required of appraisers. If the method and extent of the sampling are fully and clearly described, and if the intent is specified in the appraisal, the level of sampling is another function of the appraiser’s judgment. This level of testing may require input as to both methodology and content of a certified statistician.
