On 4 November 2026, at its eighth Tax Certainty Day, the OECD released the 2025 statistics on Mutual Agreement Procedures (MAP) and Advance Pricing Arrangements (APA) for the 137 jurisdictions of the Inclusive Framework on BEPS (available here). Globally, competent authorities closed a record 2,971 MAP cases (+17.6%) and the average time to close a case fell from 27.37 to 25.66 months.[1]
Belgium is one of the heaviest users of the procedure: roughly one in seven MAP cases pending worldwide involves the Belgian competent authority. In practice, the Belgian figures show a competent authority that resolves most cases by agreement and within the 24-month target, but also a sharp rise in transfer pricing disputes and in bilateral APA applications. This alert sets out the Belgian figures and what they mean for taxpayers with cross-border exposure.[2]
I. In More Detail: How the MAP and APA Statistics Work
The statistics are not a voluntary survey. Under the BEPS Action 14 minimum standard, members of the Inclusive Framework commit to seek to resolve MAP cases within an average of 24 months and to report their caseload every year under a common MAP Statistics Reporting Framework. The published figures are the measure against which that commitment is assessed.
A. Two generations of cases
The framework separates cases received before 1 January 2016 (or before the year in which a jurisdiction joined the Inclusive Framework) from cases received as from that date. Older cases are counted under each jurisdiction’s own method and are not reconciled, so the same case may appear in the figures of both treaty partners. Newer cases follow common definitions, are agreed between the two competent authorities and are reported per treaty partner. Partners with fewer than five cases are grouped in a single line.
B. Two categories of cases
Every case is classified either as an attribution/allocation case, commonly called a transfer pricing case, or as an “other” case. The first category covers the attribution of profits to a permanent establishment and the allocation of profits between associated enterprises. The second covers everything else: residence, employment income, pensions, withholding taxes, the existence of a permanent establishment, and so on.
C. What is reported
For each category, a jurisdiction reports the inventory at the start of the year, the cases started, the cases closed, the inventory at year-end and the average time taken to close cases. Closed cases are also allocated to one of ten outcomes, nine of which fall into three groups (the tenth is a residual “any other outcome”):
- cases that never reach the other competent authority on the merits (access denied, objection not justified, withdrawal by the taxpayer);
- cases resolved without a bilateral agreement (unilateral relief, domestic remedy);
- cases discussed between competent authorities (full agreement, partial agreement, agreement that the taxation was in accordance with the treaty, no agreement).
…
D. How time is measured
For cases under the common framework, the clock does not run from the date of the assessment. It starts shortly after the taxpayer files the MAP request and stops when the taxpayer is formally informed of the outcome. An intermediate milestone, the first position paper exchanged between competent authorities, splits that period in two. The more recent editions also show whether a case was closed at the unilateral or the bilateral stage and how old the pending cases are.
The OECD’s closing ratio compares the cases closed during the year with the start inventory plus half of the cases started. The higher the ratio, the larger the share of the available caseload that was cleared during the year.
E. APA statistics
Since the 2023 reporting year, jurisdictions with a bilateral or multilateral APA programme also report APA figures: start and end inventory, applications filed, APAs granted, applications rejected or closed for other reasons, and the average time to grant. Unlike the MAP figures, these are not reconciled between treaty partners and are not broken down by partner.
F. Three reading notes
- The 24-month target is an average per jurisdiction, not a deadline in an individual case.
- Average times are computed on cases closed during the year. A year in which old cases are cleared can therefore show a longer average, although the backlog improves.
- A jurisdiction’s figures reflect both competent authorities. A slow bilateral relationship does not say which side caused the delay.
…
II. Belgium in the Global Picture
Belgium closed more MAP cases than it opened in 2025, and its inventory fell from 914 to 891 cases.[3]
| Belgium, 2025 | Start inventory | Cases started | Cases closed | End inventory | Average time (months)* |
| Transfer pricing cases | 137 | 81 | 80 | 138 | 26.61 |
| Other cases | 777 | 437 | 461 | 753 | 20.30 |
| All cases | 914 | 518 | 541 | 891 | 21.22 |
* Cases started as from 1 January 2016.
Three points stand out against the global figures:
- Belgium started 518 new cases (432 in 2024) and closed 541 (423 in 2024). Closures rose by 28%, well above the global increase of 17.6%.
- Applying the OECD closing-ratio formula, Belgium closed about 46% of its available caseload, against about 39% globally.
- The Belgian average of 21.22 months remains below the 24-month target of the BEPS Action 14 minimum standard and below the global average of 25.66 months.
…
III. MAP: Four Findings for Belgium
A. The Belgian caseload is overwhelmingly non-transfer pricing
Of the 891 cases pending at year-end, 753 (85%) are “other” cases, i.e. cases that do not concern the attribution of profits to a permanent establishment or the allocation of profits between associated enterprises. Three neighbouring jurisdictions account for 64% of that inventory: the Netherlands (204), France (180) and Luxembourg (97). This concentration suggests that cross-border employment and pension income drive much of the Belgian caseload.
This category improved in 2025. Closures rose from 358 to 461, the inventory fell, and the average time decreased from 21.08 to 20.30 months, against 22.58 months globally.
B. Transfer pricing disputes are rising and taking longer
New transfer pricing cases increased by 50%, from 54 to 81. Closures kept pace (80, against 65 in 2024), so the inventory is stable at 138 cases. The average time, however, rose from 24.05 to 26.61 months. That is above the 24-month target, although still below the global transfer pricing average of 30.29 months.
C. Most cases end in full relief
Of the 541 cases closed, 371 (69%) ended with an agreement fully eliminating double taxation or fully resolving the taxation not in accordance with the treaty. For transfer pricing cases the share is 88% (70 of 80). Only three cases ended without agreement.
Two figures deserve attention. The Belgian competent authority considered the objection not justified in 42 cases, twice the 2024 number, and access to MAP was denied in 33 cases. Both outcomes are concentrated in non-transfer pricing cases.
D. The treaty partner matters
Time to resolution differs widely from one bilateral relationship to another.
| Treaty partner | Type of case | Pending at 31 December 2025 | Started in 2025 | Average time (months) |
| Netherlands | Other | 204 | 149 | 14.16 |
| France | Other | 180 | 92 | 20.24 |
| Luxembourg | Other | 97 | 46 | 19.56 |
| Spain | Other | 43 | 22 | 38.50 |
| Germany | Other | 42 | 25 | 29.23 |
| Germany | Transfer pricing | 24 | 15 | 14.80 |
| Italy | Transfer pricing | 24 | 11 | 33.97 |
| Luxembourg | Transfer pricing | 17 | 16 | 11.81 |
| France | Transfer pricing | 10 | 10 | 26.54 |
| Spain | Transfer pricing | 6 | 4 | 31.81 |
Transfer pricing cases with Germany, Luxembourg and the Netherlands (8.70 months) were closed quickly. Cases with Italy and Spain took more than two and a half years on average. Spain nevertheless cleared most of its backlog with Belgium in 2025: 24 transfer pricing cases were closed by full agreement, against none in 2024.
The Belgian inventory is also comparatively young: 64% of the cases started as from 2016 are less than two years old (57% globally) and only 4% are older than six years (8% globally).
IV. APAs: Demand Outpaces Grants
Applications for bilateral and multilateral APAs in Belgium rose by 68% in 2025, from 22 to 37, while the number of APAs granted barely moved.[4]
| Belgium | Start inventory | Applications filed | Granted | Rejected | Closed for other reasons | End inventory | Average time to grant (months) |
| 2024 | 51 | 22 | 13 | 0 | 1 | 59 | 35.70 |
| 2025 | 43 | 37 | 15 | 0 | 3 | 62 | 31.49 |
Belgium corrected its 2025 start inventory following an error in its 2024 submission, which explains the difference with the 2024 end inventory.
The Belgian programme compares well on timing: 31.49 months to grant an APA, against a global average of 40.19 months. No application was rejected in either year.
The pressure point is capacity. With 37 applications and 18 closures, the inventory grew from 43 to 62 cases in twelve months. Unless the number of APAs granted increases, taxpayers filing today should expect longer lead times than the 2025 average suggests.
The programme also remains small. Belgium holds about 1.4% of the 4,334 APA cases pending worldwide, a fraction of its share of the global MAP inventory. The rise in transfer pricing MAP cases and in APA applications in the same year indicates that groups with Belgian operations increasingly seek certainty on transfer pricing, both after and before the fact.
V. Takeaways for Clients
The 2025 figures confirm that MAP is a credible route to relief in Belgium, provided the request is well prepared and expectations on timing are set by reference to the treaty partner concerned.
- Treat MAP as a real option, not a last resort. Nearly nine in ten Belgian transfer pricing cases closed in 2025 ended in full relief, and cases ending without agreement are rare.
- Set the timeline by treaty partner. A transfer pricing case with Germany or the Netherlands and one with Italy or Spain are different propositions. Build the expected duration into provisioning and into the decision on interim measures.
- Prepare the request carefully. The number of objections found not justified and of cases denied access shows that admissibility is not a formality. Verify the time limit under the applicable treaty, document the taxation not in accordance with the treaty, and safeguard domestic remedies in parallel.
- Consider a bilateral APA early. Belgian APAs are granted faster than the global average, but the inventory is growing quickly. File well ahead of the first year to be covered and raise roll-back at the outset: the OECD peer review notes that the Belgian programme does not provide for roll-back in all appropriate cases.[5]
- Use the statistics. The jurisdiction-by-jurisdiction data allow a realistic assessment of duration and likely outcome before choosing between MAP, an APA and domestic litigation.
For any question on mutual agreement procedures, advance pricing arrangements or the management of cross-border tax disputes, please contact Simont Braun’s Tax team.
Jacques Malherbe | Of Counsel
Rik Strauven | Associate
[1] OECD, press release of 4 November 2026.
[2] This alert is for general information only and does not constitute legal advice or legal opinion. Percentages, year-on-year changes and closing ratios are our own calculations based on the OECD data. Please consult with a legal counsel of your choice before taking any action based on the information provided.
[3] OECD, 2025 Mutual Agreement Procedure Statistics, including the Belgium country file (2025 MAP and APA statistics).
[4] OECD, 2025 Advance Pricing Arrangement Statistics.
[5] OECD, Making Dispute Resolution More Effective: peer review of Belgium (2025).
