The Unconstitutionality of the Portuguese Stamp Tax: A Case Against the Taxation of Guarantees
Introduction
The Portuguese Stamp Tax (Imposto do Selo, “IS”) is an anachronistic levy that evolved from a documentary tax into a convenient revenue-raising device, unmoored from genuine economic ability. While its application to financing already raises constitutional doubts, the problem is sharper with guarantees: the tax bites where no contributive capacity is manifested.
Contributive Capacity as a Constitutional Benchmark
The Constitution requires taxes to align with real economic ability. Contributive capacity is not a slogan; it legitimises taxation only where there is wealth, income, or consumption. A tax that targets acts revealing liability rather than wealth is constitutionally suspect.
Guarantees Reflect Liabilities, Not Wealth
Some contend that providing a guarantee evidences economic strength. This is incorrect. A guarantee may be furnished either by the debtor or by a third party. If furnished by the debtor, it merely secures an existing liability and evidences exposure, not wealth; if furnished by a third party, any “capacity” belongs to the guarantor, while the fiscal burden typically falls on the debtor. Guarantees therefore do not, as such, manifest the taxpayer’s contributive capacity; they allocate risk and secure performance.
This shows why guarantees cannot serve as reliable indicators of economic ability. The act of guaranteeing does not enrich the taxpayer or generate additional wealth. It only reflects obligations already in existence or risk transferred between parties.
Misattribution of Capacity (Debtor vs. Third Party)
The constitutional anomaly is acute when the guarantor is a third party. A parent company pledging shares to secure a subsidiary’s loan, or an individual standing surety for a friend, reveals the guarantor’s assets; yet the tax burden may attach to the debtor in the main transaction. This misattribution violates the principle that taxation must rest on the taxpayer’s own capacity, not someone else’s.
Equality and Double (or Misplaced) Taxation
Even if one assumes a guarantee signals patrimony, taxing it collides with equality. The assets mobilised as collateral—homes, land, shares—already bear specific taxes (e.g., IMI, IMT, income taxes, and VAT where applicable). Encumbering an asset does not create new wealth; it merely reconfigures legal risk. A homeowner who mortgages the property is not richer than an identical owner who does not.
The same arbitrariness appears in leases. Under the TGIS, lease contracts are specially taxed at 10% of one monthly rent (the landlord is the taxpayer). If the tenant delivers a cash deposit (caução), that is a guarantee potentially within Verba 10; however, the law exempts guarantees that are materially accessory to a contract specially taxed in the TGIS and constituted simultaneously with the principal obligation—three cumulative conditions repeatedly underlined in tax arbitration. Thus, a deposit given simultaneously with the lease may be exempt, whereas later reinforcements and/or substitutions may trigger IS as a guarantee. By contrast, prepaying rent is not a guarantee at all; beyond the base 10% rule on one month’s rent, it does not generate an additional guarantee tax. The economic reality—the landlord’s security—may be similar, yet the fiscal outcome diverges based on formalities rather than capacity.
The Contradiction of Exemptions
The exemption for guarantees that are accessory and simultaneous to contracts already taxed by the TGIS tacitly recognises that no capacity arises from the guarantee itself. If capacity existed, there would be no principled reason to exempt it. The statute thus taxes some guarantees and exempts others not by reference to ability to pay, but by formal timing and linkage. That is the hallmark of arbitrary design.
The Anatomy of Verba 10 (and Why It Fails)
Verba 10 taxes “guarantees of obligations” (including caução, fiança, hipoteca, penhor, seguro-caução) on an ad valorem basis that scales by term (0.04% per month under one year; 0.5% one year or more; 0.6% five years or more or open-ended). The provision expressly carves out the accessory-and-simultaneous exemption where the principal contract is “specially taxed” in the TGIS (e.g., leases). The CAAD case law crystallises the three cumulative requirements: (i) material accessoriness, (ii) principal obligation specially taxed in TGIS, and (iii) simultaneity with the birth of that obligation (even if in a separate instrument).
Constitutionally, this architecture is untenable. A formal nexus (accessory + simultaneous) is treated as the determinant factor of taxability, while economic ability—the constitutional north star—is marginalised. Two tenants with identical means are treated differently merely because one replaced a deteriorated guarantee later in time; two borrowers with identical solvency are treated differently because one could constitute the mortgage on day 0, while the other only on day 1. Such distinctions are not rationally connected to capacity.
Territoriality
The territorial scope of the Stamp Tax on guarantees, set out in Article 4(b) of the CIS, covers guarantees granted in Portugal and those granted to resident entities, including Portuguese branches of foreign institutions. The burden, however, is always imposed on the debtor as the holder of the guaranteed obligation. This design leads to discrimination: two debtors with identical economic ability may face different tax outcomes depending solely on where the guarantee is provided, where the secured assets are located, or whether the creditor is resident in Portugal. Such factors do not reveal contributive capacity; they merely reflect jurisdictional connections. The territorial rule therefore deepens the constitutional fragility of the tax by tying liability to formal criteria divorced from real economic ability.
Comparative Glimpse
Comparative practice has generally retreated from taxing guarantees as such. Several common law systems abolished stamp duties on securities and mortgages; many civil law countries narrowed their scope. Portugal’s persistence with Verba 10 reveals an older, form-driven paradigm detached from modern capacity-based taxation.
Conclusion
Guarantees do not manifest the taxpayer’s contributive capacity. They are liability-facing devices that allocate risk; where a third party guarantees, any “capacity” lies with the guarantor, not the debtor who often bears the fiscal burden. Verba 10’s ad valorem charge, offset by a narrow accessory-and-simultaneous carve-out, turns formal timing into the key to taxability, producing unequal outcomes among economically identical cases.
In leases, the contrast between simultaneous deposits (potentially exempt) and later guarantee events (taxed)—while prepaid rent attracts no additional guarantee tax—lays bare the disconnect between the tax and true economic ability. The territorial rule compounds the problem by discriminating between debtors based on criteria unrelated to capacity.
A constitution based on ability-to-pay requires taxes to track real capacity, not formalities. The Stamp Tax on guarantees fails that test. It is outdated, arbitrary, and unconstitutional—and it should be retired in favour of a coherent framework that taxes actual manifestations of wealth, income, or consumption.