Executive answer
A transfer pricing method is selected after delineating the transaction, testing functions, assets and risks, reviewing terms and determining which comparable data exist. It is not selected by habit, prior-year practice or database convenience. The file should explain why the chosen method is the most appropriate and why reasonable alternatives are less reliable.
Article 180 of the Mexican Income Tax Law recognizes six methods: comparable uncontrolled price, resale price, cost plus, profit split, residual profit split and transactional operating profit margin. The statutory text gives relevant sequencing to CUP. Documentation should show that this priority was evaluated, not merely state that “no CUP was found.”
The method answers an economic question. Is there a comparable price for the same dealing? Can a resale gross margin be observed? Does a cost base represent a manufacturing or service function? Do both parties make unique and integrated contributions? Can a less complex party be tested through a net margin? A tree organizes the inquiry, but evidence decides the answer.
Before selection: delineate the dealing
Define the parties, property or service, rights, volume, currency, period, market, agreement, conduct, functions, assets, risks and strategy. Separate components with different treatment. A “service” agreement may contain a license, financing and reimbursement; a distribution arrangement may include marketing or warranty.
Determine which party controls relevant decisions and contributes unique assets. Identify internal comparables and segmented data. Without clear delineation, any method produces precise calculations for the wrong question.
Document data constraints before calculation. Missing information does not change the facts, but it may alter relative method reliability. First pursue reasonably available information rather than treating an avoidable limitation as justification.
Selection tree
Is a reliable internal or external CUP available?
Compare characteristics, market, volume, timing, terms, credit, security and functions. When material differences can be adjusted reasonably, CUP generally provides direct evidence. Otherwise identify the candidate, difference and why its effect cannot be measured.
Is this distribution without unique value and are gross margins comparable?
Consider resale price when the third-party resale price is observable and a comparable gross margin can be determined after distribution functions. Accounting consistency is critical: discounts, freight, marketing and warranty classification may distort gross margin.
Does the transaction remunerate manufacturing or services over a reliable cost base?
Cost plus can be suitable when functions, costs and independent markups are comparable. Before pricing the markup, define included costs, pass-throughs and differences caused by assets or risks.
Do both parties make unique or highly integrated contributions?
Profit split may be suitable when no simple party can be assessed in isolation or both contribute unique intangibles. It requires combined profit, perimeter, accounts and allocation factors linked to value. It is not a residual response to poor data.
Can a less complex party be tested using a net margin?
TNMM compares a net profit indicator over costs, sales or assets. Select the party with reliable information and no difficult-to-compare unique contribution. The indicator should fit the function and be segmented by transaction.
Are other recognized approaches reasonably applicable?
The law also recognizes residual profit split. Valuation and other techniques may support particular cases, but connect the conclusion with statutory methods and facts. Document alternatives, not only the final formula.
Method matrix
| Method | Core question | Needed data | Common risk |
|---|---|---|---|
| CUP | What price did independent parties agree? | Comparable dealing and terms | Unreliable adjustments |
| Resale price | What gross margin rewards distribution? | Resale price, gross margin and accounting | Incompatible classifications |
| Cost plus | What markup belongs on costs? | Controlled cost base and comparables | Arbitrary cost inclusion |
| Profit split | How should combined profit be divided? | Combined accounts and contributions | Arbitrary factors |
| Residual split | Which routine and residual returns apply? | Basic returns and unique value | Double counting contributions |
| TNMM | What net margin does a comparable party earn? | Segment, indicator and companies | Convenient tested-party choice |
Add availability, adjustments, limitation, result and conclusion. A method may be conceptually strong but impracticable because of data; another may be applicable with lower comparability. Selection compares overall reliability.
If the file only names the method used, document the alternative tree and CUP rejection before relying on the result.
Justifying CUP rejection
Begin with internal dealings: sales, purchases, loans, services or licenses with independent parties. List candidates and compare every factor. Do not use a generic information statement when the company holds agreements and invoices.
Potentially material differences include product, unusual volume, geography, supply-chain level, exclusivity, term, warranty, currency and period. Determine whether data exist for an adjustment. Perfection is unnecessary, but an adjustment should improve reliability and avoid speculation.
If rejecting CUP, explain why another method better tolerates the differences or uses more reliable data. If accepting it, document adjustments and sensitivity. Apply the same standard to favorable and unfavorable references.
Tested party and indicator
Under unilateral methods, the tested party is generally the party to which the method can be applied most reliably and that lacks material unique contributions. It is not automatically Mexico or the entity with less profit. Compare complexity, intangibles, risks and data on both sides.
The indicator should fit the function. Return on sales may fit distribution; markup on costs may fit services or manufacturing; return on assets may fit asset-intensive functions, with care. Analyze the denominator, classification and assets used. A stable indicator over the wrong base is not defensible.
Prepare a reconciled segmented statement. Do not test an entire entity combining manufacturing, distribution, services and finance unless aggregation criteria are satisfied and no distortion results.
Aggregating transactions
Dealings may be analyzed together when closely linked or continuous and separate evaluation would be unreliable. Aggregation should not hide transactions with different methods or risks.
Document the economic relationship, agreements, flow and why the combined indicator represents the dealings. Compare aggregation with forms that may require separate disclosure. Preserve amounts by type even when the method uses one segment.
Distribution and marketing may be integrated; a loan should not be folded into an operating margin. Different services may share a base when provider, beneficiary and allocation are coherent, while a license needs its own analysis.
More than one method
One appropriate method may be enough, but a second analysis can corroborate uncertain or complex cases. Do not average incompatible outcomes. Explain the question each method answers and which carries more weight.
A valuation may support a royalty; an imperfect CUP may corroborate TNMM; a rate analysis may coexist with debt-capacity analysis. A second method cannot repair poor delineation.
Method changes between years
There is no indefinite entitlement to a historical method. Changes in facts, comparables, systems, functions or interpretation may support a new choice. Keeping the method can also be appropriate when the dealing remains stable.
Prepare a bridge showing earlier and current facts, new limitations, quantitative effect and implementation date. Do not change merely because another method produces a convenient result. Align policy, agreements, accounting and returns.
Practical scenarios
Distributor with an internal CUP. It sells the same product to third parties, but in different volumes and countries. Before TNMM, test market, volume and term adjustments.
Service center. It performs routine processes with identifiable costs. Cost plus may be direct when gross markups exist; TNMM may be more reliable when gross classifications differ.
Integrated manufacturer with local technology. If both parties contribute unique intangibles and decisions, a unilateral method may miss value; evaluate profit split without assuming it.
Loan. A CUP based on financial instruments may work after defining debt, rating, currency, term and security. An operating margin does not answer the financing question.
Common errors
- Selecting the method before FAR.
- Repeating the prior-year method without testing.
- Rejecting CUP with boilerplate.
- Choosing the tested party by residence.
- Applying TNMM to unsegmented statements.
- Confusing missing data with no comparables.
- Using profit split as a last resort without unique value.
- Mixing financial and operating dealings.
- Changing method after seeing results.
- Failing to align method with agreement and policy.
Quality control and approval
Independent review should rebuild the tree and check FAR, internal comparables, data availability, tested party, indicator and segmentation. Recalculate results and test sensitivity. Leadership should approve material changes and understand operating consequences.
The final memorandum includes the delineated dealing, alternatives, evidence, rejection analysis, selected method, application, limitations and refresh triggers. It should show not only what was done but why it was most reliable.
Method decision record
Maintain a compact decision record for every material transaction. It should state the economic question, candidate methods, data requested, data obtained, internal comparable review, reason for each rejection, selected method and approving reviewer. Link each assertion to the FAR, agreement, segmented accounts or benchmark rather than repeating conclusions.
The record should distinguish “not available” from “not reliable.” Data may be unavailable because the group does not collect it; that is a control gap. Data may exist but be unreliable because terms or accounting cannot be compared; that is a technical limitation. The remediation differs.
Add a trigger table. A new product, change in decision-making, acquisition of an intangible, new third-party transaction, ERP improvement or persistent loss may require reconsideration. This prevents historical method selection from becoming a permanent assumption.
When the selected method affects operating policy, translate it into actionable pricing instructions: rate, base, frequency, monitoring metric, exception owner and true-up process. A technically correct annual method that cannot be executed during the year will repeatedly create closing adjustments.
Related topics
- PT-017 and PT-018: FAR and comparability.
- PT-020: range and median.
- PT-021 through PT-025: method-specific guides.
Method selection-and-rejection memorandum
The memorandum starts with the delineated transaction: parties, rights, conduct, functions, assets, risks and conditions. It then evaluates each recognized method through required data, strengths, limitations and possible adjustments. Selection cannot rest on a method being “most common”; it should explain why that method gives the most reliable result for these facts.
Address CUP explicitly. Identify internal and external prices, conditions and differences; if rejected, show why adjustments would be unreliable. For profit methods, justify the tested party, indicator, segmentation and comparables. Include sensitivity where a reasonable choice changes the outcome.
Tax, operations and the economic reviewer approve the version, while disagreements remain documented. Refresh the analysis when agreements, conduct, data or markets change. This file defends both the chosen method and the disciplined process that rejected alternatives.
Sources and verification date
- Mexican Income Tax Law, current text, Articles 179 and 180.
- OECD Mexico Transfer Pricing Country Profile.
- OECD Transfer Pricing Guidelines, a technical reference.
Sources checked on August 2, 2026. Method selection depends on facts and data; no label guarantees application.
Request a Method Selection Workshop to document the dealing, alternatives, CUP, tested party, indicator and most reliable method.