Fiscal Reform Proposal · Romania 2026 · Version 5.0
Not a reform. A pure replacement. One tax, differentiated by elasticity, zero cascade through UFT-90. The only fiscal architecture that survives AI displacement.
Pilot 10% · Reference 13% · Maturity 15%
Executive Summary
TTF SHIFT v5.0 replaces all taxes with a single transaction tax. Not additive — replacement. Rates are differentiated by elasticity, not uniform. UFT-90 blocks cascades by design. Cost of AI displacement is explicitly modeled. Every number is anchored in 2025–2026 budget execution.
CPMF failed because it was added to the system, not replacing it. TTF SHIFT is the inverse: it eliminates VAT, income tax, profit tax, social contributions. One tax remains. No hybrid, no supplement.
Six rate tiers, not one. HFT at 0.01–0.1%. Financial non-HFT at 0.1–0.4%. B2B essential at 3–5%. B2C at 5–7%. ATM at 15–20%. UFT-flagged money at 0%. Every tier is calibrated against its escape velocity.
30% job loss scenario quantified: 1.68M displaced, €1,000/mo National Card top-up = 90.5 bn lei/year cost = 1.5% additional TTF rate. The only system where automation increases the tax base.
The CPMF experiment proved that transaction taxation works — it failed for political reasons, not technical ones. TTF SHIFT v5.0 solves the technical flaw (cascade) through UFT-90 architecture and the perception flaw (uniform rate) through elasticity-calibrated tiers. It is the only fiscal architecture that is mathematically solvent under AI displacement, legally defensible in the EU, and empirically anchored in real data.
Chapter 1 — The Crisis
The 2026 Romanian budget crystallises two simultaneous failures: the fiscal model (35–40% evasion, chronic deficit) and the labor model (automation eroding the tax base itself).
| Indicator (Law 43/2026) | mld lei |
|---|---|
| Total revenues | 390,6 |
| — Gross VAT planned | 156,3 |
| — Excise duties | 52,5 |
| — Profit tax | 42,4 |
| — Payroll + income tax | 59,7 |
| Total expenditures | 526,3 |
| Deficit | −135,7 |
EU excessive deficit procedure active. VAT gap 35–40% (EC 2024). Compliance cost 2–3% of turnover.
WEF, McKinsey, MIT: 40–60% of current jobs structurally eliminated by 2040. Payroll tax + CAS collapse proportionally. In Romania alone, 77.6 bn lei in budget payroll taxes + 158.7 bn lei in CAS contributions become unstable.
A tax system based on human labor cannot survive the automation of human labor. This is not a political opinion. It is arithmetic. The classical system has no mechanism to tax a robot that replaces a worker.
Chapter 2 — Why Pure, Not Mixed
The most common objection — "TTF failed in Brazil" — misidentifies the cause of failure. CPMF did not fail because transaction taxation is unworkable. It failed because it was layered on top of an existing system, creating double burdens and resentments.
Critics say "TTF SHIFT has a high rate — 10–15% — how can that be better than VAT at 21%?" The answer: they are not comparable. The 21% VAT is in addition to 16% profit tax, 37% payroll contributions, and 10% income tax. The cumulative burden on value added is 60–80%. TTF SHIFT replaces all of them with a single 10% on new flows — a 6–8× reduction in total fiscal burden.
Chapter 3 — Rate Matrix v5.0
v4.0 treated rates as a single lever. v5.0 treats them as a matrix. Each segment has a different elasticity — a different speed at which it flees to cash, offshore, or non-bank channels. The rate must be inversely proportional to that escape velocity.
Academic studies on CPMF (Brazil) identify the Laffer inflection point at approximately 0.62% — above this, revenue falls as the base flees. Sweden's 1% transaction tax lost 30% of equity volume and 80% of bond volume. France's 0.4% non-HFT rate remains stable; a unilateral extension to derivatives would trigger "important flow movements outside France" (French Treasury). For HFT, the welfare-maximizing rate is 0.37%, and the revenue-maximizing rate is 0.78%. Any rate above these thresholds is revenue-negative. v5.0's tiers are calibrated to stay below the Laffer ceiling of each segment while maximizing total revenue from the segments that cannot flee.
| Segment | Success signal | Failure signal (reduce rate) |
|---|---|---|
| UFT (0%) | Zero UFT recognition errors at all banks | Any cascade detected → UFT protocol fails |
| HFT (0.01–0.1%) | Volume stable; yield rises with rate up to 0.62% | Volume falls >15% → migrated to non-RO venues |
| Fin. non-HFT (0.1–0.4%) | Spread stable; French benchmark holds | Spread widens >20bps → rate too high |
| B2B (3–5%) | SENT/ReGIS volume drops <10% vs baseline | Netting outside banks >20% → rate too high |
| B2C (5–7%) | Card transactions stable; cash usage <+15% | Cash withdrawal surge >20% → reduce |
| ATM (15–20%) | Share of electronic payments rises | Withdrawals flat → rate not discouraging, just penalizing |
Chapter 4 — The Critical Engineering Parameter
v4.0 introduced the Unique Flow Token (UFT) as a permanent marker of already-taxed money. That design is incomplete — a permanent UFT collapses the tax base to a single year's PIB. v5.0 corrects this with a calibrated 90-day lifetime, which is the single parameter that determines whether the reform is solvent or not.
Short enough to prevent the base from collapsing (at 90 days, a company with quarterly cash cycles still gets UFT protection for its normal production chain — raw material → production → sale → collection, all within 90 days). Long enough that inter-company supply chains are not re-taxed (the median B2B payment cycle is 45–60 days in Romania). Calibrated so that hoarded money — savings, unused reserves, dormant accounts — re-enters the tax base when mobilized. The parameter is not arbitrary: it matches the NBR's quarterly reporting cycle and the standard ISO 20022 settlement window.
| Day | Stage | Flow | UFT Status | TTF applied |
|---|---|---|---|---|
| 0 | Customer → Firm A (revenue) | 10,000 lei | UFT generated (exp: day 90) | 5% = 500 lei |
| 15 | Firm A → Supplier B | 6,000 lei | UFT active, propagated | 0% |
| 45 | Supplier B → Producer C | 4,000 lei | UFT active, propagated | 0% |
| 88 | Producer C → Firm A (final goods) | 3,000 lei | UFT about to expire | 0% |
| 91 | Firm A's next transfer (post-expiry) | 2,000 lei | UFT expired → new event | 5% = 100 lei |
Zero cascade within the 90-day window. Post-expiry events create new taxable base, ensuring the reform is solvent.
UFT-90 is not a new protocol — it is an additional metadata field in the existing ISO 20022 payment messages already used by NBR's ReGIS and SENT systems. The field encodes: (1) UFT issuance date, (2) originating transaction ID, (3) chain lineage. Banks verify and propagate automatically. No additional infrastructure is needed.
Chapter 5 — The Brazil Lesson
The empirical test that TTF SHIFT is built on: CPMF (Brazil, 1993–2007) collected 30% of PIS/Cofins revenue at 1/10 the rate. Evasion below 5%. Zero filing. Zero compliance cost. It was abolished by corporate lobbying, not economic failure.
1. Unresolved cascade: CPMF was applied indiscriminately to every transfer, at 0.38%. Manageable at that rate, but structurally hostile to long production chains — hence a powerful political argument for abolition. 2. Corporate lobbying: large players orchestrated abolition using the cascade argument, real but technically solvable. 3. No citizen protection: the abolition was not accompanied by any alternative social safety net, so the political coalition for keeping it collapsed.
1. Cascade: solved through UFT-90 metadata — technically impossible, not just illegal. 2. Corporate lobbying: defused through the "economic superiority" model — capital stays because it earns more inside the system than outside (Chapter 7). 3. No safety net: the National Card is the constitutional guarantee that abolition of old taxes does not mean abolition of social protection.
Chapter 6 — The AI Displacement Model
The classical system collapses. TTF SHIFT is the only fiscal architecture that gains revenue as automation accelerates. Here is the full arithmetic — the model that no other reform proposal in Romania has produced.
| Step | Calculation | Result |
|---|---|---|
| 1. Displaced workers | 5.586.187 × 30% | 1.675.856 |
| 2. Annual income loss | 1.675.856 × 4.500 lei × 12 | 90,5 bn lei |
| 3. TTF rate to cover only this | 90,5 / 6.000 (taxable base) | 1,51% |
| 4. Combined with full budget replacement | 699 + 90,5 = 789,5 bn lei | 789,5 bn lei |
| 5. Effective TTF rate needed | 789,5 / 6.000 | 13,16% |
The 13% reference rate from the fiscal verification is validated by the AI displacement model. Same number, two independent derivations.
When a robot replaces a worker in the classical system, the payroll tax base shrinks. When a robot replaces a worker in TTF SHIFT, the robot's own transactions (energy, components, maintenance, cloud, logistics) generate new TTF base. The larger the automation wave, the larger the TTF base. This is the only fiscal architecture in which AI is a revenue amplifier, not a revenue destroyer.
The National Card (€1,000/mo for adults 20+) is not a replacement for all salaries. It is a floor — a constitutional guarantee that no citizen falls below a dignity threshold regardless of employment status. Displaced workers who earn more than €1,000 through gig work, freelance, or partial employment receive the difference only up to the ceiling. The National Card does not discourage work; it eliminates the poverty trap that makes re-entry impossible.
Chapter 7 — Independent Fiscal Verification
| Revenue Category — BGC 2025 actual | Million lei | Bn lei |
|---|---|---|
| Total fiscal revenues | 323.131 | 323,1 |
| — Net VAT | 133.901 | 133,9 |
| — Excise duties | 48.320 | 48,3 |
| — Corporate + income tax | 104.955 | 105,0 |
| Total social contributions | 208.034 | 208,0 |
| — Pension (CAS) | ~118.700 | ~118,7 |
| — Health (CASS) | ~69.600 | ~69,6 |
| TOTAL to replace | 531.165 | 531,2 |
| 2026 estimate (+9% nominal) | ~579.670 | ~580,0 |
Source: Ministry of Finance Romania, BGC 2025 annual execution.
Gross top-ups for all eligible adults: ~304 bn lei/year. Already budgeted in current system (pensions, social assistance, unemployment): ~185 bn lei/year. Net additional cost: 119 bn lei/year. Total requirement: 580 (fiscal) + 119 (National Card net) = 699 bn lei/year.
| Rate | Yield (base 6.000 bn, 2% evasion) | vs. 699 bn requirement | Verdict |
|---|---|---|---|
| 8% | 470 | −229 | ❌ Excluded |
| 10% | 588 | −111 | ⚠ Pilot only (legacy covers gap) |
| 12% | 706 | +7 | ✓ Minimum viable |
| 13% | 764 | +65 | ★ Recommended reference |
| 15% | 882 | +183 | ✓✓✓ Maturity (Sovereign Fund) |
v4.0 presented 10% as a full replacement rate. v5.0 explicitly corrects this. 10% is viable only as a pilot alongside the legacy system (the 111 bn lei gap is covered by remaining VAT). 12–13% is the minimum viable full-replacement rate. 15% enables a sovereign investment fund. The reform is honest about its own thresholds.
Chapter 8 — The National Card
Every adult Romanian citizen with residence in Romania receives an automatic monthly top-up — no application, no paperwork, no bureaucracy. Funded entirely from TTF revenue. The mechanism that makes AI displacement survivable.
| Category | Transfer Rule | State Top-up |
|---|---|---|
| Public sector salaries | First to National Card (up to ceiling), surplus to personal card | Difference to €1,000 |
| Private sector salaries | To personal card; state auto-tops-up separately | Difference to €1,000 |
| Pension < €1,000 | Entire amount to National Card | Difference to €1,000 |
| Pension ≥ €1,000 | Entire amount to National Card — no surplus to personal card | Zero (pension covers) |
| Special pensions (any amount) | 100% to National Card — no euro to personal card | Zero |
A special pension of €5,000/month enters entirely on the National Card. The pensioner's purchasing power is not limited — it is structurally directed toward the Romanian economy via the RO 90% platform. Aggregate effect: 154.8 bn lei/year of pension spending circulates 100% through Romanian merchants, services, and producers. Captive demand. Jobs. Endogenous growth. Zero additional cost to the state.
Chapter 9 — Capital & Platform
The 10–15% rate will generate a perception of friction. The solution is not to reduce the rate — it is to make compliance more profitable than avoidance, through a complete economic platform.
Chapter 10 — Legal Path
The EU legal framework does not forbid the reform. It requires a specific procedure. Article 395 of Directive 2006/112/EC allows the Council, acting unanimously on a Commission proposal, to authorise a Member State to introduce special measures derogating from the Directive for the purpose of simplifying VAT collection or preventing certain forms of tax evasion or avoidance.
| Legal Step | Mechanism | Status |
|---|---|---|
| VAT Directive Art. 395 derogation | Council unanimity, Commission proposal | Precedent exists (multiple Member States) |
| Import/export treatment | Zero-rated TTF, VAT per existing rules | No derogation needed |
| Free movement of capital (Art. 63 TFEU) | TTF as internal tax — same for residents & non-residents | Compatible if non-discriminatory |
| State aid rules | RO 90% platform = neutral incentive scheme | Needs careful formulation |
| EMU fiscal framework | Deficit reduction path | TTF surplus eliminates excessive deficit |
Article 395 requires unanimity in Council — 27 Member States. This is a high bar. But it is not an insurmountable one: Romania has secured such derogations before, and the fiscal solvency argument (surplus vs. deficit) is strong. The strategy is to negotiate the derogation as part of a broader fiscal reform package, framed as "Romania as EU pilot for flow-based taxation," rather than as an isolated request. The pilot phase (Phase 0–1) does not require derogation — TTF coexists with reduced VAT during this phase. Derogation is only needed for Phase 2 (full replacement).
Chapter 11 — Implementation Roadmap
Chapter 12 — Conclusion
Traditional tax systems were designed for an economy that no longer exists.
TTF SHIFT v5.0 is not a reform of the existing system — it is its pure replacement, calibrated by elasticity, engineered against cascade, and designed to survive the automation of human labour.
One tax. Six tiers. Zero filings. Zero ANAF. Zero cascade. With a National Card that makes AI displacement survivable. With a Platform that makes capital compliance more profitable than avoidance. With a Constitutional guarantee that no future government can reintroduce classical taxes.
Technically feasible with existing NBR/TRANSFOND infrastructure. Anchored in the 2025 BGC execution data. Empirically validated by CPMF Brazil 1993–2007. Legally navigable through Article 395 derogation.
This is not a utopia. It is an engineering problem with a solution.
References