Meet in the middle: HMRC’s transfer pricing settlement policy

Unfortunately, in the real world, establishing the arm’s length price or range of prices can be challenging because of data limitations and involve the exercise of considerable professional judgement. In a Transactional Net Margin Method (TNMM) benchmarking analysis for example, a decision is usually required as to whether differences in economic circumstances between each potential comparable party and the taxpayer, or the absence of information to make detailed comparisons, are sufficiently minor to permit a reliable conclusion on pricing. In some cases, all potential comparables are likely to have comparability defects but are the closest that can be identified. Sometimes these differences can be reliably adjusted for and sometimes this is not possible due to data limitations or other factors.

This need for judgement means that reasonable people can sometimes disagree over the conclusions. For example, a tax authority such as HMRC may take the view that a taxpayer’s filed position falls outside the arm’s length range. If the taxpayer is unable to convince HMRC that no adjustment is required, then HMRC ultimately need to determine both the arm’s length range and the point within that range to which they will adjust in order to close the audit (see paragraph 3.61 of the OECD Transfer Pricing Guidelines (TPG)). It is this point in the range determination that is the subject of HMRC’s policy update.

Where HMRC considers that the TP outcome reflected in the tax return is outside the arm’s length range, they have now explained that they expect a central point in the arm’s length range to be the adjustment point in the great majority of cases. In practice we expect that the median will be the expectation of HMRC’s governance processes in all cases unless there is compelling evidence that a different point is more appropriate. Previously HMRC’s position had been more variable.

Often in practice, limitations on available information mean that a transactional profit method, such as the TNMM is selected. The available information may not be sufficient to establish very precise comparability but the TNMM tests Net Profit Indicators (ratios), which the TPG explain are less affected by transactional differences and may be more tolerant to some functional differences than direct price testing. Even after a careful analysis in this way, there are two possible situations: either the identified comparable companies are of relatively equal and high reliability, or some unidentified and/or unquantified comparability defects remain. The former is unusual in practice in our experience. By way of example, it is relatively common to see comparable sets of distributors exhibiting interquartile ranges with spreads of approximately five percentage points or more.

The TPG explain that limited information means that “determining a reliable estimate of an arm’s length outcome requires flexibility and the exercise of good judgement.”. Further, “a substantial deviation among points in that range may indicate that the data used in establishing some of the points may not be as reliable [as others]”. Where the points are of relatively equal and high reliability, “it could be argued that any point in the range satisfies the arm’s length principle.”

Where some comparability defects remain, paragraph 3.62 of the TPG suggests it may be appropriate for tax administrations to determine the adjustment point using measures of central tendency, such as the median (or perhaps the arithmetic mean or others), to minimise the risk of error due to these defects. Put simply, in cases where there are non-quantifiable comparability defects, adjusting to a central point is likely to minimise the likelihood of error.

HMRC’s policy change is now generally to expect adjustment to the central point and to be much clearer about this policy expectation. HMRC’s approach is not novel. Several of the UK’s major partners, including France, Germany, India and the US, will adjust to the median or mean as a matter of law or policy if the taxpayer’s results fall outside the arm’s length range.

HMRC’s new approach to TP settlement standards is explained by reference to their Litigation and Settlement Strategy (LSS) which “requires HMRC to adopt an approach that secures the best practicable return to the exchequer. It also states that where there is a range of possible figures for tax due, HMRC will not settle by agreement for less than it would reasonably expect to obtain from litigation”

However, the negotiation process involved in MAP means that a settlement under the LSS does not provide final certainty over the tax ultimately payable in the UK: the Competent Authorities of the two treaty partners may agree a different resolution. It is also possible for taxpayers to suspend payment of tax pending the outcome of MAP which can delay when HMRC is able to establish and collect the final tax liability. HMRC have explained that they are taking particular care to ensure consistent audit settlement positions in cases with potential to go to MAP. Such consistency may facilitate efficient resolution of double taxation and provide value to the UK Exchequer.

As noted above, many countries, including the US, already take the position that an adjustment to the median or another central point in the range will be appropriate as a logical and principled approach where the taxpayer has filed outside the arm’s length range.

We understand that HMRC does not want to be in the position of settling cases based on adjustments to (say) the lower quartile, and then find themselves agreeing MAP relief for adjustments by other countries to the median or mean. They are therefore effectively instructing their audit teams to now generally adjust to the median of the arm’s length range where the tested party has fallen outside unless there is very strong support for another point in the range.

Such clarity from HMRC is useful in our view, although the arm’s length position in each specific case should be determined on its specific facts (as HMRC continue to recognise). The practical point is that HMRC view the median as generally most appropriate to use when making a TP audit adjustment and they are now making their position clear. A secondary motive from HMRC for this clarity may be to provide an extra incentive for taxpayers to self-assess transfer pricing adjustments where they have fallen outside the arm’s length range in a period and consider an adjustment to a different point in the range would be more appropriate.

Although HMRC’s approach to settlements where MAP is not available has not been updated, in practice we would expect to see HMRC’s governance boards adopt similar principles in those cases too.

Although this is new guidance on settling TP enquiries, much of the underlying reasoning is not new and recommended practice for taxpayers hasn’t changed.

A best practice approach here is to:

• Invest in establishing a robust arm’s length range. Undertake a careful review of potential comparables upfront, with careful consideration as to whether some are more reliable than others and whether any comparability adjustments would improve reliability;
• Unless there are strong technical arguments for an alternative approach, establish a target point that is well within that arm’s length range reflecting sensitivity analysis on potential Tax Authority adjustments to the comparables set;
• Ensure robust operational processes for regular ongoing monitoring of transfer prices and profit outcomes for group companies, making timely adjustments as necessary; and
• Document judgements made regarding outcomes falling outside of expected benchmarking ranges including whether to adjust or not, and document how any adjustment was quantified.

The good news here is that developments in Artificial Intelligence (AI) are opening up exciting new opportunities to make these best practices more achievable. For example, we are already seeing AI unlocking significant improvements in the efficiency of undertaking benchmarking searches and the reliability of the outputs.

For further information please contact:

• Phil Roper
• Nick Stevart