Measurement under FIN 48: A tax professional's guide to oddsmaking

By John Kohler

Implementation of the Financial Accounting Standards Board's new Interpretation No. 48, "Accounting for Uncertainty in Income Taxes," presents many daunting tasks for tax professionals.

Previously a company recorded contingent liabilities for the tax exposure related to uncertain tax positions._¹ Now, before the tax benefit for any expense or other tax position₂ can be recorded in the financial statements, a company must complete an evaluation under the new FIN 48.

In short, the evaluation requires a company to first identify all tax positions. The company must determine that each tax position first meets the more-likely-than-not (MLTN) recognition threshold. For "highly certain" tax positions,₃ the company may need to document that the position is highly certain, however no measurement analysis may be required4 and the full tax benefit of the item may be recorded. For any other positions that meet the MLTN threshold the company must complete a probability analysis to measure the amount of the tax benefit that may be recognized for financial statement purposes. The creation of the probability analysis will challenge tax professionals to try their hands as oddsmakers.

Other articles address the many aspects of FIN 48. This article will focus only on one of the most challenging aspects — completion of the probability analysis required for FIN 48 measurements.

FIN 48 requirements for measurement

Under FIN 48, a tax position that meets the MLTN recognition threshold is then measured to the largest amount of the tax benefit that is greater than 50 percent likely of being realized upon ultimate settlement with the tax authority.₅ The company must consider the possible estimated outcomes using "the facts, circumstances, and information available at the reporting date."₆ Unlike the recognition test, where a company should assume resolution in the court of last resort₇ (for example, the U.S. Supreme Court), the measurement should consider what the company would ultimately accept on settlement with the tax authorities.₈

FIN 48 provides the following example₉ of an analysis of the possible outcomes for a $100 tax benefit:

FASB example:

Possible estimated outcome	Individual probability of occurring	Cumulative probability of occurring
$100	5%	5%
$80	25%	30%
$60	25%	55%
$50	20%	75%
$40	10%	85%
$20	10%	95%
$0	5%	100%

In the example, the benefit to be recorded would be $60, the largest amount with a cumulative probability greater than 50 percent. The FASB does not provide any detail on the possible estimated outcomes or the basis for the individual probabilities.

Must a company use probability tables? By including examples of probability tables, the FASB "does not intend to imply a documentation requirement."₁₀ FIN 48, however, states "[t]hat measurement is based on an analysis of the distribution of potential outcomes (that is, potential realized tax benefits) and their related probabilities."₁₁ The board rejected the use of a "best estimate," the single most likely amount in the range of possible estimated amounts.₁₂ The methodology illustrated by the probability tables therefore is required by FIN 48. A company may find it difficult to document that it applied the methodology for measurement under FIN 48 without completion of the probability tables or similar documentation.₁₃

Practical aspects of the probability analysis

So where do all the numbers come from? The FASB does not indicate where the probability data would come from. If this were professional sports, statistical data would be readily available or one could simply refer to the professional oddsmakers. In the tax arena, however, historical information on, for example, the median settlement of a particular issue at Internal Revenue Service Appeals is protected as confidential.

Even if a subject matter expert is consulted and the subject matter expert has prior experience with the same exact issue on examination by tax authorities, that subject matter expert is likely prohibited under professional ethical constraints from divulging the details of the outcomes obtained by other clients. The only outcomes published are court decisions, not settlements. Therefore the only support the company may be left with on expected settlements is the subject matter expert's opinion. This limited support may leave the company's analysis vulnerable to "second guessing."

Another practical problem with FIN 48 is that it applies to all tax positions. A company must complete the MLTN recognition standard and the probability measurement for all of its significant tax positions except perhaps the highly certain tax positions.₁₄ Accordingly, a company may be faced with completing the analysis for dozens or hundreds of such tax positions.

For a typical company, a number of these issues likely will be so significant that the company will want to, and need to, complete a detailed and thorough analysis. It will need to evaluate each possible outcome and prepare detailed calculations to determine the tax benefit under each possible outcome. The company would presumably use prior experience or consult with subject matter experts to determine the individual probability of each outcome.

This process may not be practical for all issues. In particular, less significant issues may not merit the rigor of a full analysis and consultation with a subject matter specialist.

Use of a probability template

As an alternative for the less significant tax positions, a company may want to consider adoption of a probability template as the basis to produce FIN 48 probability tables. The probability template, or templates, could be used as the starting point for default probability tables, subject to modification as necessary for facts and circumstances applicable to unique issues.₁₅

In considering the odds at the horse track, one must first consider the field and decide which horses are contenders and which are destined for the glue factory. Similarly, a company in the development of a probability template must determine a set of possible outcomes for settlement of tax positions.

The identification of the possible outcomes might start with the articulation of a corporate strategy on settlement of tax examinations. The strategy would provide guidelines as to the expectation for the company's tax executives to settle particular tax issues at different levels (for example, agent examination, appeals, and mediation) at prescribed percentages to avoid the time and expense of further contesting an issue to the next level.

The settlement percentages prescribed by a company's settlement strategy would likely vary depending on the size of the tax benefit at issue and the impact of the adjustment, whether temporary or permanent. The larger the amount, the more willing a company would be to take an issue to the next tax controversy level. Considering professional fees and corporate resources, a company may decide it is unwilling to take an issue below a certain threshold — for example, issues that represent less than $100,000 — to IRS Appeals. Similarly, a company may be unwilling to take an issue below a certain threshold — for example, $5 million — to mediation. Conversely, the smaller the amount at issue the more willing a company would be to settle for a lower amount to avoid further tax controversy.

Another factor a company would typically take into consideration in the settlement of an issue is the impact of the adjustment. For example, the adjustment proposed by a tax authority could result in a one-year deferral of a deduction, such as the deferral of the tax deduction of a one-time prepaid expense. The tax benefit of such an item would merely be delayed until the next year. Alternatively, the proposed adjustment could result in the capitalization and subsequent depreciation or amortization of an item over a longer period such as three, 15 or 39 years. An example of this type of adjustment would be the tax capitalization and depreciation of a repair. Yet another alternative is that the proposed adjustment would be permanent such that its disallowance would never reverse. An example of a permanent adjustment would be an adjustment to the disallowance of meals and entertainment expenses.

Considering the time value of money, a company would be much less willing to contest an adjustment of a one-year temporary item than, for example, an adjustment that would not reverse for 15 years. On the other hand, a company likely would be much more willing to contest an adjustment that is a permanent tax cost. Companies often calculate the net present value difference of recovering, for example, the tax benefit of depreciation over 15 years versus 39 years.

A company could use these two factors, the size and the impact, to determine appropriate guidelines for settlement. In order to satisfy the requirements of FIN 48, these guidelines should also reflect management's "best judgment" as to what would be accepted by the tax authorities.₁₆

Assume for illustration purposes that a company determines that the possible outcomes under examination in a tax jurisdiction₁₇ are:

acceptance in full,
settlement during examination at a reduced amount to avoid appeal,
settlement at a further reduced amount at IRS Appeals to avoid going to mediation, and
an expected "worse case" settlement at IRS mediation.₁₈

Based on these possible outcomes and the factors above, the company settlement strategy could provide for the settlement percentages spelled out in the Settlement Strategy Table.

Settlement Strategy Table

Alternative	Size of issue			Adjustment for impact of issue
	Less than $100,000	Mid-size	More than $1 million	1-year temporary item	Mid-term temporary item (3 to 8 years)	Temporary item (15 years or more)	Permanent item
Accepted	100%	100%	100%	100%	100%	100%	100%
Examination	70%	85%	90%	25%	50%	75%	100%
Appeals	0%	55%	60%	25%	50%	75%	100%
Mediation	0%	0%	50%	25%	50%	75%	100%

The settlement percentages in the table above are for illustration purposes only. In developing actual percentages, the factors a company might consider may include the company's cost of funds, the corporate emphasis of cash flow, the availability of tax attribute carryforwards to offset short-term adjustments of temporary items, and the availability and cost of internal, versus external, resources to contest examination issues.₁₉
Using the settlement strategy table, the acceptable settlement amount under each alternative would be the full tax benefit multiplied by both the size factor and the impact factor. Further illustration will be provided later in this article.

Determining probabilities

Once real contenders in a horse race are identified, a handicapper must still consider the odds for each horse. Similarly for FIN 48, once the possible outcomes are established, a company must determine the individual probability of actually settling an issue under each outcome. An experienced horse race handicapper likely has a selected number of factors for consideration — the history of the horses, the record of the jockey, the track conditions, etc. A company may also want to identify specific standard factors for use in developing the probability factors.

Assuming again that the possible outcomes for analysis are settlement at various levels of examination,₂₀ the individual probabilities would be affected by a number of factors. Many factors, such as the personality mix and attitudes of the company and government personnel may be unknown or impossible to assess. The company will likely want to focus on more specific factors.

Two factors that typically can be assessed, although perhaps not without some difficulty, are the opinion level and the complexity of the issue. Tax opinions are usually issued as either MLTN or "should" level opinions. Opinions at lower levels, such as "substantial authority," would not be relevant for this analysis since those positions would not meet the MLTN recognition threshold of FIN 48. Tax opinions may also be issued at the stronger "will" opinion level, however those positions might be considered "highly certain" under FIN 48 such that measurement of those tax positions may be unnecessary.

An issue with the stronger "should" level opinion presumably has a higher probability of being settled earlier in the exam process. This may also mean that an issue with a "should" level opinion has a higher probability of being settled for the full amount or a greater percentage of the tax benefit than an issue with only an MLTN opinion level.

Tax positions also vary in complexity. Some issues may be merely factual in nature — whether the company has adequate contemporaneous documentation for meals and entertainment expenses. Others may be relatively simple from a technical standpoint — whether a particular capital improvement qualifies for bonus depreciation. Others may be complex technical issues, such as whether a corporate reorganization qualifies for tax-free treatment.

Again, a company can presume that more factual-based issues and relatively simple technical issues are more likely to be settled at the agent examination level or at least earlier in the tax controversy process. On the other hand, complex issues may be difficult to resolve at the agent level and thus the company may expect a higher probability that it will need to resort to, for example, IRS Appeals for resolution. Similarly, issues relating to reportable transactions₂₁ are less likely to be resolved at the agent level and thus again a company may assume a higher probability that tax positions resulting in reportable transactions will need to go further through the tax controversy process before a settlement can be reached.

A company could then establish probability tables using these factors, such as in the Probability Source Table.

Probability Source Table

		Impact of opinion level	Impact of complexity issue
Alternative	Default probability₂₂	Should level	Factual	Simple technical	Complex technical reportable transaction
Accepted	20%	+10%	+25%		-15%-20%
Examination	30%	+10%	+15%	+40%	-10%-10%
Appeals	30%	-10%	-20%	-20%	+25%+20%
Mediation	20%	-10%	-20%	-20%	+10%
	100%	0%	0%	0%	0%0%

Note that the above probability tables are for illustration purposes only.₂₃ A company must determine such probabilities based, for example, on discussions with the tax controversy specialists in the appropriate jurisdictions. The actual percentages used by a company could be influenced by its prior history or current relationship with the authorities in the applicable jurisdiction and any examination program with which it may be involved, such as large-case examination.

A company could prepare a score card assessing the opinion level and complexity of each issue. The individual probability percentages can thus be generated for each issue by taking the default percentage for each outcome and increasing or decreasing it by the impact of the opinion level and complexity of the issue. For example, the individual probability of resolving by settlement at the examination level (30 percent) of an issue with a "should" level opinion (+10 percent) that is a complex technical issue (-10 percent) would be 30 percent.

The probability tables can thus be populated and a company can then determine the largest amount of the benefit that is greater than 50 percent likely of being realized upon settlement; the amount of the tax benefit reflected in the financial statements.

Examples

Based on the settlement strategy and the probabilities above, an example of a completed probability table using this approach could be:

Example 1.1

Issue example No. 1	Tax treatment of ABC issue
Total tax benefit:	$250,000
Factors impacting settlement amounts:
Size of issue	Mid-size
Tax impact of item	1-year temporary item
Factors influencing probability
Opinion level	MLTN
Complexity of issue	Simple technical

Example 1.2

Alternative	Amount	Individual probability	Cumulative probability	Benefit to record
Accepted—Accepted in full	$250,000	20%	20%	$--
Examination—Company agrees to a reduction to settle issue at the examining agent level	$53,125	70%	90%	$53,125
Appeals—Company agrees to a settlement at Appeals to avoid mediation	$34,375	10%	100%	$--
Settlement—Company expects settlement at IRS mediation	$--	0%	100%	$--
		100%

For example, the $53,125 settlement amount at the agent level would be based on the full tax benefit of $250,000 times the size factor of 85 percent (for mid-size items) times the impact factor of 25 percent (for a one-year temporary adjustment) per the Settlement Strategy Table. The individual probability percentage for settlement at the agent level, 70 percent, would be based, per the Probability Source Table above, on the default percentage of 30 percent, no adjustment for opinion level (the opinion is only MLTN), and increased by 40 percent for the impact of the issue being a simple technical issue.

In this example, the tax benefit for which the cumulative probability exceeds 50 percent is a settlement at the agent level, thus a tax benefit of $53,125 would be recorded. The balance of the benefit, $196,875, would be recorded as a FIN 48 liability or could be an adjustment to a deferred tax account.

In this example, although the opinion level on the tax position meets the MLTN threshold, the expected settlement amount at the agent level, $53,125, is a relatively small portion of the total $250,000 benefit because the issue in this illustration is a one-year temporary item. Since it is only a one-year temporary item, a company would likely be willing to settle the item for a very modest amount to avoid further tax controversy.₂₄ The FIN 48 credit established in year one of $196,875 would fully reverse in year two as the uncertainty of deduction is eliminated. In year two the company would presumably accrue interest, and if applicable, penalties on the one-year deferral of taxes.

Example 2.1

Issue example No. 2	Tax treatment of XYZ issue
Total tax benefit:	$5 million
Factors impacting settlement amounts:
Size of issue	Greater than $1 million
Tax impact of item	Temporary item (greater than 15 years)
Factors influencing probability
Opinion level	Should level
Complexity of issue	Complex technical

For the second example, the $2,250,000 settlement amount at the Appeals level would be based on the full tax benefit of $5 million times the size factor of 60 percent (for items greater than $1 million) times the impact factor of 75 percent (for temporary adjustments greater than 15 years). The individual probability percentage for settlement at this level, 45 percent, would be based on the default percentage of 30 percent, less 10 percent for the impact at this level of the "should" opinion and plus 25 percent for the impact of the issue being a complex technical issue.

In this example, the outcome for which the cumulative probability exceeds 50 percent is resolution through the IRS Appeals, thus a tax benefit of $2,250,000 would be recorded. The balance of the benefit, $2,750,000 would be recorded as a FIN 48 liability or could be an adjustment to a deferred tax account.

Example 2.2

Alternative	Amount	Individual probability	Cumulative probability	Benefit to record
Accepted—Accepted in full	$5 million	15%	15%	$--
Agent—Company agrees to a reduction to settle issue at the examining agent level	$3,375,000	30%	45%	$--
Appeals—Company agrees to a settlement at Appeals to avoid mediation	$2,250,000	45%	90%	$2,250,000
Settlement—Company expects settlement at IRS mediation	$1,875,000	10%	100%	$--
		100%

Closing comments

A company will want, and need, to complete a detailed thorough probability analysis for all significant issues. The use of probability templates as outlined above, however, could assist with the analysis of less significant issues. Even with this approach, the company should still review each tax position to determine if the results of using the settlement strategy and probability templates are appropriate for the particular issue. Past history of examination on an issue or unique facts and circumstances may suggest modifications. The template approach, however, at least may readily provide the creation of default tables or a starting point for the more challenging issues.

This template approach allows a company to provide its attest firm with a logic trail, as well as at least some support for the analysis in the probability tables.₂₅ The settlement amounts would be based on a stated company settlement strategy and the probabilities could presumably be based on discussions with tax controversy specialists. Another advantage that may be attractive to both the company and its attest firm is that this approach would assure some consistency in the analysis of large numbers of issues.

The use of probability templates should be discussed in advance with a company's attest firm. This approach could help tax professionals beat the long odds of completing the implementation of FIN 48 on time.

Footnotes

The previous accounting arose under Financial Accounting Standards 5, Accounting for Contingencies.
Tax positions include, for example, claiming of tax credits, exclusion of income from taxation, allocation of income between jurisdictions, and the decision not to file returns in a jurisdiction.
FIN 48 Paragraph A-19.
A "highly certain" deduction might be uncertain as to timing such that measurement may be required.
FIN 48 Paragraph 8.
Id.
FIN 48 Paragraph A-2.
FIN 48 Paragraph A-3.
FIN 48 Paragraph A-21.
FIN 48 Paragraph A-1.
FIN 48 Paragraph B-43.
FIN 48 Paragraph B-42.
The FASB also adopted a probability approach in FIN 46R, Consolidation of Variable Interest Entities. In FIN 46R enterprises use probability tables, when appropriate, to calculate expected losses and expected residual returns. In FIN 46R the FASB states, "Enterprises will not necessarily be able to estimate the probabilities to use a precise computation of the type illustrated, but they should use their best efforts to achieve the objective result." See FIN 46R Paragraph A-1.
Again, a highly certain tax deduction may be uncertain as to the timing of the deduction such that measurement may be required. See FIN Paragraphs A-28 and A-29.
Note the probability-template approach outlined below is not endorsed by Deloitte Tax LLP or any of its affiliates as a surrogate to the application of professional judgment or as a mechanical solution to the FIN 48 measurement requirement. Rather the author is merely suggesting this as possible approach for the myriad of smaller tax positions for which it may not be practical, for a number of reasons, for a company to develop a more robust probability analysis. Whether a probability template approach is appropriate for the analysis of any FIN 48 tax positions should be thoroughly considered by the company and discussed with its attest firm.
FIN 48 Paragraph A-3.
The settlement levels and alternatives would vary depending on the applicable tax jurisdiction. Also, the company should consider the applicability of other alternatives, such as the IRS fast-track settlement program. For information on the IRS fast-track settlement program see Revenue Procedure 2003-40, 2003-25 I.R.B. 1044.
Litigation is not considered in this illustration. Litigation may be an option, however, it likely would only be a viable alternative for the most significant tax positions for which the suggested template approach would probably not be appropriate. For information on IRS mediation see Rev. Proc. 2003-41, 2003-25 I.R.B. 1047.
The above illustrative settlement table is built with the assumption that the settlement percentages decrease as a company goes through the tax controversy process (i.e., the company expects a smaller settlement in appeals than exam). In some cases, however, a company may envision a better settlement in appeals or mediation because of the ability of those forums to consider hazards of litigation. This expectation of a greater settlement, however, requires additional time and cost of further controversy.
Note that under FIN 48 the possible outcomes for consideration need not be based on settlement at various levels of the tax controversy process. The alternatives could be based, for example, on different alternatives for structuring a compromise settlement at the same IRS examination level.
A reportable transaction in this context refers to transactions requiring special disclosure under Internal Revenue Code Section 6707A(c)(1).
This illustration assumes that the default probabilities are based on an MLTN opinion level.
Also note that using certain combinations of the percentages in this table could result in cumulative probabilities exceeding 100 percent, however only the 50 percent cumulative probability is critical for FIN 48. Again, these percentages are for illustration only.
On the other hand, another company faced with a great number of uncertain one-year temporary differences may be much less willing to settle such issues for such a low settlement percentage.
A company may also want to consider whether having a settlement strategy as is being suggested would create a potential "road map" for IRS or other tax authorities on how a company would be willing to settle an issue.

John Kohler is based in St. Louis and is a partner in Deloitte Tax LLP's FAS 109 Competency Group. He can be reached by e-mail at jkohler@deloitte.com. This article does not constitute tax, legal, or other advice from Deloitte Tax LLP, which assumes no responsibility with respect to assessing or advising the reader as to tax, legal, or other consequences arising from the reader's particular situation.

This content has not yet been Rated.

To Rate content, please Login.