Statistical sampling and resulting allocations under fixed-asset studies

By Joseph Tawfik and Eric Wilhelm, Washington

Editor: Annette B. Smith, CPA

Determining the proper tax treatment of tangible property can be a challenge for taxpayers. For example, issues can arise because contractor invoices for the construction of real property do not identify the cost of property by tax recovery period. As a result, taxpayers may treat the entire construction cost as real property for tax purposes and depreciate it over a 39-year period. In recent years, taxpayers have focused on whether certain remodeling or repair expenditures from their book accounts should be considered repairs for tax purposes.

Tax practitioners have been performing fixed-asset studies for businesses by reviewing capital accounts to determine whether expenditures should be expensed as repairs under Sec. 162(a) or treated as capital under Sec. 263(a). Taxpayers with many tangible property expenditures frequently find the only practical way to review and document the appropriate treatment of these expenses is through statistical sampling. Rev. Proc. 2011-42 provides guidance on how to develop an acceptable statistical sample, but it does not provide a methodology for applying the results of the sample.

This discussion focuses on a methodology to sample expenditures by reviewing amounts capitalized for book purposes and to determine the extent to which they should be classified as tangible personal property with a shorter recovery period or expensed for tax purposes. A similar exercise could be applied in reviewing accounts expensed for book purposes to determine whether any should be capitalized based on the original asset's placed-in-service date.

Upon completing the sample review, documentation for some sampled expenditures may indicate that the item should be reclassified, for example, from a 39-year asset to either a shorter recovery period or a repair expense. These sample results then are statistically extrapolated to the population. The final estimate represents the total population amount that should be reclassified to reflect the correct tax treatment.

Many fixed-asset studies cover multiple years, and the taxpayer's records can include different types of property and recovery periods. While the results for the sampled assets are known, the reclassification for the nonsampled assets must be determined to properly calculate the depreciation for all assets in the population.

Example 1: Allocation of extrapolated reclassification amount by year

The following example demonstrates one approach to allocating the reclassification amount to nonsampled assets. As illustrated in the table "Example of Allocation by Year," the cost of property for the population is $600 in column (2); the sample amount is $80 in column (3); and the nonsampled population is $520 in column (4). The projected reclassified amount, in this case the repair expense, is $440 in column (5); and the sample reclassification result is $60 in column (6). After backing out the sample results, the nonsampled reclassification estimate is $380, as seen in column (7). To recalculate depreciation, those total values must be separated by placed-in-service years and the appropriate recovery period or repair category. This example assumes all the expenditures are reclassified to repairs.

Example of allocation by year

The allocation methodology described below demonstrates the determination of repair expense amounts attributable to the nonsampled population. First, separate extrapolations are performed to estimate the reclassification amount for each placed-in-service year as illustrated in column (5). The sample results then are removed from the statistical estimate to obtain the reclassified, nonsampled allocation percentages by year in column (8). These percentages are used to allocate the nonsampled reclassified amount by year.

Example 2: Allocation of extrapolated reclassification amount by detailed allocation

As illustrated in the table below, "Example of Detailed Allocation (Nonsampled Assets)," the nonsampled allocation percentages in column (4) are used to allocate each nonsampled expenditure into two buckets: expense or depreciated over 39 years. The categories can differ for each project. This table illustrates how the nonsampled cost of property of $520 is allocated into the reclassified tax treatment for each nonsampled asset. As calculated in the first table, the non-sampled expense amount is $380, and the nonsampled capital amount is $140. The percentages derived in the first table can be applied to each nonsampled asset to determine the expense amount and the 39-year amount. This example assumes all of these assets originally were classified as 39-year property.

Example of detailed allocation (nonsampled assets)

Taxpayers can use their sample results to calculate the proper depreciation amount for the fixed assets in the sample population.

This allocation methodology is not limited to two categories. If the sample results include reclassifications of five-year property, seven-year property, 15-year property, and expenses, this allocation methodology could be used to appropriately allocate the nonsampled expenditures into the various categories.

Statistical sampling and allocation often can be used to determine the proper tax treatment of tangible property. To be useful, the statistical result will be an aggregate amount of reclassified expense that must be allocated by recovery period or repair by placed-in-service date. While this discussion explains two methods for allocating results, other reasonable methods also exist.


Annette Smith is a partner with PricewaterhouseCoopers LLP, Washington National Tax Services, in Washington.

For additional information about these items, contact Ms. Smith at 202-414-1048 or

Unless otherwise noted, contributors are members of or associated with PricewaterhouseCoopers LLP.

Newsletter Articles


CPEOs provide peace of mind around payroll services

The creation of these new IRS-certified service providers for small businesses clarifies some issues around traditional professional employer organizations.


2016 Best Article Award

The winners of The Tax Adviser’s 2016 Best Article Award are Edward Schnee, CPA, Ph.D., and W. Eugene Seago, J.D., Ph.D., for their article, “Taxation of Worthless and Abandoned Partnership Interests.”