EXTERNAL comparability analysis, or benchmarking, is one of the most important yet often misunderstood aspects of transfer pricing. The essence of determining an arm’s length price lies in assessing the comparability of the conditions governing transactions between related parties against those between independent parties in the open marketplace. The term “comparability” is therefore synonymous with benchmarking.
A common belief is that benchmarking simply involves comparing financial ratios, such as gross margins, net margins and return on assets, among companies operating in the same industry and broadly carrying out similar activities. This perception is incorrect, as comparability is an essential consideration throughout the transfer pricing process in determining an appropriate arm’s length price.
Benchmarking the financial data
The process begins with identifying the appropriate tested party, selecting the appropriate Profit Level Indicator (PLI), determining the relevant period and establishing suitable search criteria.
Potential comparable companies are then identified from appropriate database, directories and subjected to qualitative and quantitative screening before their financial results are analysed.
The challenge, however, is not the mechanics. It is determining whether the companies selected are genuinely comparable.
Same industry does not mean comparable
One of the most common misconceptions is that companies operating in the same industry are automatically comparable. They are not.
A manufacturer may own valuable intangibles, carry significant inventory and market risks, while another manufacturer may operate as a routine contract manufacturer.
This is why a benchmarking exercise must begin with a proper FAR analysis i.e. functions performed, assets employed and risks assumed, rather than simply searching for companies using similar keywords.
The comparable set must be robust
The use of keywords is only a starting point. A company may perform similar functions but describe its business differently, while companies with similar descriptions may have fundamentally different business models.
The search may also produce too few potential comparables, creating a temptation to retain companies with significant differences merely to increase the sample size. The objective should not be to obtain a large number of comparables. A smaller set of reliable comparables is preferable to a larger set of weak ones.
Geographical differences, product or service differences, asset intensity and the availability of reliable financial information must also be considered. For example, a highly asset-intensive manufacturer or leasing company should not automatically be accepted as comparable to a routine service provider merely because both fall within a broadly similar industry classification.
Do not eliminate companies just because they are inconvenient
The treatment of loss-making companies is another area of concern. A loss-making company should not automatically be rejected. Losses may arise from an adverse industry cycle, restructuring or temporary market conditions. The relevant question is whether the circumstances causing the loss also affect comparability with the tested party.
Similarly, the elimination criteria must be applied consistently. A company should not be rejected for a particular characteristic while another company with the same characteristic is retained because its profitability produces a more favourable result. This raises an important principle: a benchmarking study should not be designed to produce a desired answer.
The statistics cannot fix a weak analysis
Under Malaysia’s current transfer pricing framework, the arm’s length range is generally between the 37.5th and 62.5th percentiles.
However, being within the range does not, by itself, make a benchmarking study reliable. If the comparable companies have significant functional differences, poor-quality data or inconsistent screening, the resulting range may give a false sense of comfort.
Adjustments to operating or non-operating items should have a clear and supportable basis. They should not be used simply to make an otherwise weak comparable fit the analysis.
End goal
The purpose of a comparability study is not to prove that the taxpayer’s result is correct. It is to demonstrate whether independent parties operating under sufficiently comparable circumstances would reasonably have achieved a similar outcome.
This article is contributed by Thannees Tax Consulting Services Sdn Bhd managing director SM Thanneermalai (www.thannees.com).