The Central Attack Trap in Madrid Protocol Portfolios

The Madrid Protocol is genuinely attractive on paper. A single international application, filed through the applicant's home trademark office, can designate protection in more than 130 member countries. One language. One WIPO fee schedule. One renewal cycle. For a company expanding to five or ten markets simultaneously, the efficiency case is real.

But there is a structural risk baked into the Madrid system that the efficiency pitch doesn't mention. It's called central attack. It can unravel a multinational trademark portfolio faster than most trademark owners expect, and it is cheap enough to execute that competitors have learned to use it strategically.

The Five-Year Dependency Rule

Every international registration under the Madrid Protocol is dependent on the basic mark — the application or registration in the applicant's home country that underpins the international filing. Under Article 6 of the Protocol Relating to the Madrid Agreement Concerning the International Registration of Marks, this dependency is absolute for five years from the date of the international registration.

If the basic mark is cancelled, invalidated, or withdrawn during that five-year window — for any reason — the international registration falls with it. Not in one designated country. In every country the applicant designated.

The five-year clock runs from the WIPO international registration date, not from the home-country filing date. Those dates can be weeks or months apart. And actions that are merely commenced before the five-year mark but not resolved until afterward can still drag down the international registration — the dependency period extends until the action is finally determined.

WIPO's Madrid Working Group has documented that central attacks grew from roughly 200 total or partial cancellations in 1996 to more than 7,000 in 2017, and the trend has continued upward. The International Trademark Association flagged the issue formally in 2024, urging policy reform to "limit the adverse effects of the central attack and encourage increased use of the Madrid system." As of the twenty-third session of the WIPO Madrid Working Group, scheduled for September 2025, the reform proposals remain under discussion. Nothing has changed in the Protocol itself.

The Goods Specification Trap

For U.S.-designated international registrations — filed under Section 66(a) of the Lanham Act, 15 U.S.C. § 1141f — there is a second structural problem that is less catastrophic but encountered far more often in practice.

The identification of goods and services in a Section 66(a) application is frozen at whatever WIPO has on record. Under USPTO examining practice, the applicant cannot broaden or materially amend the identification beyond what is in the international registration. If the home-country office accepted a goods description that is too vague or too broad for U.S. standards, the USPTO will issue a provisional refusal — but the applicant can only narrow, never fix, the underlying problem.

Compare this with a direct U.S. 1(a) or 1(b) filing, where the applicant has room to work with an examining attorney on the identification language before publication. The Madrid route eliminates that flexibility. Description problems discovered after a Section 66(a) filing are consistently more expensive to resolve than they would have been in a direct application, and some cannot be resolved at all.

Provisional Refusals and the Hidden Local Counsel Cost

The Madrid Protocol's efficiency depends on designated jurisdictions accepting the application without objection. When a designated country issues a provisional refusal — because the mark is too descriptive, conflicts with a prior registration, or the goods specification doesn't meet local requirements — the applicant must engage local counsel in that country to respond. That cost is not included in the Madrid system fees.

For an applicant designating fifteen countries, the probability that at least one jurisdiction issues a provisional refusal is not low. The cost of local counsel responses in even two or three problematic jurisdictions can approach or exceed what direct national filings in those same countries would have cost, erasing the efficiency advantage entirely.

Provisional refusals have time limits — typically twelve to eighteen months from the notification date, depending on jurisdiction — and failure to respond results in final refusal in that country with no recourse. Monitoring foreign office actions across a dozen jurisdictions simultaneously requires the same diligence infrastructure that direct national filings would require. The Madrid route doesn't eliminate that need; it consolidates the filing step while leaving the post-filing monitoring obligation fully in place.

The Transformation Safety Net — and Its Limits

The Lanham Act provides a partial remedy for central attack casualties. Under 15 U.S.C. § 1141j (Section 70 of the Lanham Act), if an international registration is cancelled because the home-country basic mark was cancelled, the trademark owner can file a transformation application — a new national U.S. filing under Section 1(a) or 1(b) — within three months of the cancellation date, and carry forward the original WIPO international registration date as the U.S. priority date.

This looks like insurance. In a limited sense, it is. But transformation is only available when the cancellation is a direct result of the basic mark's cancellation at the home office. It is not available if the international registration expired because the applicant failed to renew, if the applicant voluntarily surrendered the international registration, or if the cancellation arose from causes other than the basic mark's cancellation. When transformation is available, the applicant still pays full U.S. national filing fees, still has to demonstrate intent to use or actual use, and still goes through examination from scratch. The retained priority date is the only preserved benefit.

For non-U.S. designated countries, equivalent transformation procedures exist in some jurisdictions under national law, but they are not uniform. The practitioner cannot assume that every designated country will offer a path to preserve the priority date after a central attack.

When Madrid Is — and Isn't — the Right Tool

The Madrid Protocol works well in specific circumstances: the basic mark is stable and preferably already registered (not pending); the goods specification in the home country is specific enough for every designated jurisdiction; the applicant is filing in five or more countries, where the consolidated-filing efficiency actually materializes; and the applicant has a monitoring process for provisional refusals and a budget for local counsel responses.

It works less well when the basic mark is contentious or freshly filed, when the mark is close to descriptive or borderline inherently distinctive in one or more markets (making it an attractive target for central attack), when the applicant is filing in only two or three countries (where direct national filings are often cheaper and more flexible), or when speed-to-registration in a specific market is critical and that market's examination timeline is uncertain under Madrid.

The calculus shifts further toward direct filing for U.S. designations when the applicant's home-country goods specification is complicated or broadly worded, given the Section 66(a) goods-specification limitation.

The Portfolio Risk That Accumulates Quietly

Companies that have built international trademark portfolios through Madrid over a decade may have dozens of registrations all tethered to a small number of basic marks — typically one in the home jurisdiction and, for U.S.-based applicants, their U.S. registration. A competitor who identifies those basic marks can commence a cancellation proceeding in the home jurisdiction and place the entire international portfolio on the clock simultaneously.

The attack is economically asymmetric: inexpensive to initiate, expensive to defend, and highly effective if it succeeds within the five-year window. INTA has noted that the current structure of the Madrid dependency rule makes it "disproportionately burdensome" for trademark owners who file in good faith.

Portfolio hygiene that tracks which international registrations remain within their five-year dependency period — and which basic marks underlie each one — is the kind of systematic monitoring that gets deferred in a busy practice. For firms counseling clients on international trademark strategy, that tracking is as important as the filing itself. Trademark watch services built for this kind of ongoing portfolio surveillance can flag dependency windows before they become emergencies, rather than after.

The conversation about Madrid Protocol should include the central attack risk, the goods specification limitation, and a clear-eyed assessment of whether the filing efficiency justifies the structural exposure. For many portfolios, it does. For some, the same budget spent on direct national filings produces a more defensible result.