Box 3's Unrealized Gains Tax: How the Netherlands Systematically Punishes Deeptech Innovation
Executive Summary
On February 12, 2026, the Dutch House of Representatives passed the Wet werkelijk rendement box 3 (Actual Return in Box 3 Act), introducing a 36% tax on unrealized capital gains effective January 1, 2028. While startup exemptions exist, this legislation represents the latest chapter in a decades-long pattern of Dutch tax policy simultaneously incentivizing and penalizing innovation: creating structural disadvantages particularly harmful to capital-intensive deeptech ventures.
This analysis examines:
The technical mechanics of Box 2 and Box 3 taxation systems
How the unrealized gains tax specifically impacts deeptech funding dynamics
Historical patterns of Dutch tax policy undermining startup ecosystems
Quantitative evidence of the Netherlands' deeptech scaling failure
Structural reforms necessary for competitive deeptech development
Key Finding: While the Netherlands offers world-class R&D tax credits (WBSO at 36-50%) and Innovation Box treatment (9% tax rate), these benefits are systematically undermined by capital formation barriers, particularly affecting deeptech companies requiring patient, long-term investment.
The Dutch Tax System: A Three-Box Framework
Understanding the unrealized gains tax requires understanding how the Netherlands categorizes and taxes different income streams.
Box 1: Work and Home
Employment income and freelance earnings
Primary residence imputed rental value
Progressive rates: 36.97% to 49.50% (2026)
Not directly relevant to investment capital
Box 2: Substantial Interest (Aanmerkelijk Belang)
Applies when individual holds ≥5% ownership in a company
Taxation: Annual mark-to-market on December 31st value change
Rate: 36% flat
Key point: Tax triggered whether or not you sell
Track 2: Capital Gains Tax (Vermogenswinstbelasting)
Assets: Real estate, qualifying startup shares
Taxation: Only upon realization (sale, gift, emigration, death)
Rate: 36% on realized gains
Annual income (dividends, rent) taxed yearly
Tax-free allowance: €1,800 annual return (replaces old €57,684 capital threshold)
This structure creates a critical bifurcation: liquid securities face annual taxation on paper gains, while illiquid assets (including startup shares meeting specific criteria) receive realized-gains treatment.
The Startup Exemption: Technical Specifications and Limitations
The legislation includes a carve-out designed to protect early-stage startup investment. Understanding its precise boundaries is critical.
Qualifying Startup Criteria
A company qualifies for capital gains treatment (Track 2, tax only on realization) if all of the following conditions are met:
Age: Incorporated ≤5 years ago
Revenue: Annual turnover <€30 million
Independence: Not >25% owned by a non-startup business entity
Structure: Must be a private limited company (BV) or equivalent
Shares in qualifying startups receive preferential treatment: investors pay 36% tax only upon actual sale, not on annual paper appreciation.
Who Benefits from the Exemption
Founders with >5% ownership:
Automatically taxed under Box 2 (substantial interest regime)
Never subject to Box 3 unrealized gains tax
Pay 26.9% only on dividends and realized capital gains
If startup reaches unicorn valuation while founder holds >5%: zero tax on paper appreciation
Angel investors and small shareholders:
If shares qualify under startup criteria: Tax only on realization
36% rate applied when shares are actually sold
Protection lasts as long as company meets all four criteria
The Exemption Cliff: Deemed Sale Trigger
The exemption creates a dangerous threshold effect. When a company ceases to qualify,by exceeding the 5-year age limit, crossing €30M revenue, or violating independence requirements, a deemed sale is triggered for tax purposes.
At that moment:
Small shareholders (<5%) transition from Track 2 to Track 1
A fictional "sale" occurs for tax calculation
Going forward: Annual taxation on unrealized gains at 36%
Example scenario:
Angel investor owns 2% of DeepTech AI (incorporated 2023)
2028: Company qualifies, shares under Track 2 (safe)
2029: Company reaches €35M revenue → exemption expires
Tax authority treats shares as "sold" at 2029 fair market value
2030+: Investor owes 36% annually on paper appreciation, even if shares remain illiquid
This creates perverse incentives to exit before qualification cliffs rather than build for long-term scale.
The Historical Pattern: Death by a Thousand Tax Cuts
The unrealized gains tax is the latest manifestation of a systemic contradiction in Dutch innovation policy. The Netherlands simultaneously offers some of Europe's best R&D incentives and worst capital formation dynamics for high-risk ventures.
The Incentive Side: What Works
1. WBSO (R&D Tax Credit)
The Wet bevordering speur- en ontwikkelingswerk (Research and Development Promotion Act) is genuinely world-class:
Structure:
Reduces wage tax liability for R&D employee hours
Also covers certain material costs, tooling, and external testing
2025 Rates:
First €380,000 in R&D labor costs: 36% tax credit
Remaining R&D costs: 16% tax credit
Technostarters (<5 years old): 50% on first €380,000
The WBSO effectively makes R&D labor 36-50% cheaper. For a deeptech startup burning €1M annually on engineering talent, that's €360,000-€500,000 in direct cost reduction, significant for early-stage capital efficiency.
2. Innovation Box
Companies with qualifying intellectual property can pay 9% tax on profits derived from innovation, compared to standard 19-25.8% corporate tax rates.
Requirements:
Valid WBSO certificate for the R&D project
Demonstrable R&D activities
Profits attributable to self-developed IP
For a deeptech company generating €5M in annual profit from patented technology, Innovation Box saves approximately €550,000-€840,000 annually compared to standard corporate tax rates.
The Contradiction: What Destroys Capital Formation
Despite these R&D incentives, the Netherlands systematically undermines the capital side of the innovation equation:
Historical Tax Policy Failures:
1. Box 2 Debt Limitation (2023)
Founders previously could borrow from their companies tax-free
Common practice: Rather than take taxed dividends, founders would lend from company
January 2023: New rule treats loans exceeding €700,000 as "fictional dividends" taxed in Box 2
Impact: Reduced founder flexibility to manage personal finances without triggering tax events
2. Stock Option Taxation (Pre-2027 Reform)
Historically: Stock options taxed at grant/vest as wage income in Box 1 (up to 49.50%)
Problem: Employees owed tax on illiquid paper value
Competitive disadvantage: UK, France, Germany offered capital gains treatment
2027 partial fix: Taxable base reduced to 65% for qualifying innovative startups
Remaining issue: Still wage income treatment, not capital gains; 65% reduction helps but doesn't eliminate disadvantage
3. Wealth Tax (Box 3) on Entrepreneurial Capital
Even before 2028 reform, Box 3 taxed investment portfolios at fictitious returns
High-net-worth individuals faced annual wealth erosion regardless of actual performance
Constitutional challenge in 2021: Supreme Court ruled system unconstitutional
Government response: Not to eliminate wealth tax, but to "fix" it with mark-to-market system
4. Now: 36% Unrealized Gains Tax
Most aggressive capital growth tax in developed world
No comparable system in UK, Germany, France, Italy, Sweden, Denmark, Finland, Belgium
Creates unique anti-competitive environment for patient capital
The Pattern: Each "fix" maintains or increases tax burden on entrepreneurial capital while international competitors move in opposite direction.
Deeptech's Unique Vulnerability: Why This Matters More Than SaaS
Deeptech companies (those building fundamental breakthroughs in materials, biotech, quantum, advanced hardware, energy, semiconductors) face capital requirements and timelines fundamentally different from software startups.
Key implication: Deeptech requires patient capital willing to wait 10-15 years for exits, accepting illiquidity, and funding through long development valleys.
The Dutch Deeptech Performance Gap
Quantitative evidence demonstrates the Netherlands' structural failure to scale deeptech ventures:
TechLeap State of Dutch Tech 2023 Findings:
Spinoff stagnation:
80% of Dutch spinoffs created remain active
But: After 10 years, only 50% employ >10 people
Translation: Companies survive but don't scale
Underfunded relative to European peers:
Dutch deeptech raised only 1/3 of capital compared to Germany, France, Sweden
Even lower compared to UK deeptech funding
Not a quality problem, but a capital availability problem
IO+ State of Dutch Tech 2026 Analysis:
AI scaleup gap:
Only 21.2% of Dutch AI startups scale up (raise >€10M after initial funding)
European average: 31.1%
US average: 80.9%
89% of Dutch AI startups build vertical solutions on US foundation models (lack of fundamental research commercialization)
Funding concentration:
€2.64 billion deployed across 265 deals in 2025
Number of deals down 14.5% vs 2024
Capital flowing to fewer, later-stage companies
Early-stage funding gap widening
Deeptech represents:
12% of all Dutch tech companies
But 40% of scaleups (companies that successfully scale)
Attracts 41% of all VC money in Netherlands
Scaleup ratio: 39% (vs 17% for non-deeptech)
Paradox: Deeptech has higher scaleup success rates when funded, but faces lower funding availability, particularly at growth stages.
Why the Unrealized Gains Tax Exacerbates This
The 36% unrealized gains tax compounds existing deeptech capital constraints:
1. Discourages Patient High-Net-Worth Capital
Family offices and high-net-worth individuals represent critical growth-stage deeptech investors. They can:
Write €1M-€10M checks
Accept 10-15 year horizons
Bridge between VC early-stage and institutional late-stage
Under the unrealized gains tax:
If they invest in a deeptech company that doesn't qualify for startup exemption (e.g., >5 years old, university spinout with complex ownership)
Or if the company "graduates" from exemption before exit
They face annual 36% tax bills on paper appreciation of illiquid shares
Example:
HNW investor deploys €5M into QuantumTech BV (year 6, post-exemption)
Year 1: Company raises Series B at 2x valuation → investor's stake now "worth" €10M on paper
Tax due: (€5M gain) × 36% = €1.8M cash tax on unrealized gain
Problem: Shares are completely illiquid, no secondary market
Investor must pay €1.8M from other sources or sell unrelated liquid assets
This is structurally guaranteed to reduce willingness to deploy capital into long-duration deeptech.
2. LP Taxation Kills VC Fund Economics
Dutch venture capital funds structured as commanditaire vennootschappen (CVs, limited partnerships) are "looked through" for tax purposes. The tax authority treats LPs as directly owning underlying portfolio assets.
Implications:
If portfolio companies don't qualify for startup exemptions
Or if they age out of exemption periods
LPs face annual unrealized gains taxation on their fund positions
VC Fund Example:
DeepTech Fund I raises €100M from Dutch family offices and institutions
Fund invests in 10 deeptech companies, average hold period 10 years
Years 1-5: Some companies qualify for exemption
Years 6-10: Companies mature beyond 5-year threshold → deemed sale triggers
LPs suddenly face annual tax bills on paper gains in illiquid fund positions
Traditional VC economics assume no distributions until exits. LPs expect to pay tax when fund realizes gains and distributes proceeds, not annually on paper valuations.
Annual unrealized gains taxation forces:
Fund distributions to cover LP tax bills (reducing capital available for follow-on investments)
Or LP exit from Dutch funds in favor of non-Dutch structures
Or reduced LP commitments to Dutch deeptech funds altogether
3. The Exemption Cliff Accelerates Premature Exits
The 5-year age limit and €30M revenue threshold create artificial urgency completely disconnected from business fundamentals.
Real deeptech timeline:
Year 1-3: R&D, prototype development, early validation
Year 4-6: Pilot customers, regulatory approvals, manufacturing setup
Year 7-10: Revenue growth, scaling production, market expansion
Year 10-15: Profitability, market leadership, strategic exit or IPO
The 5-year startup exemption expires precisely when deeptech companies hit critical scaling inflection: when they need maximum capital, longest patience, and least pressure for premature liquidity.
Perverse incentive structure:
Angels invested in years 1-2 face deemed sale tax event in years 6-7
Those angels may demand earlier exits (years 5-6) to realize gains before cliff
Or demand liquidation provisions, secondary sales, dividend policies, anything to access cash for tax bills
Companies face pressure to stay under €30M revenue (avoid growth!) or exit quickly
This is anti-scaling by design.
International Context: How Competitors Are Pulling Ahead
While the Netherlands introduces one of the world's most aggressive unrealized gains taxes, competing European ecosystems are enhancing startup capital formation advantages.
UK: Capital Gains Treatment and EIS/SEIS
Enterprise Investment Scheme (EIS) and Seed EIS:
Income tax relief up to 30-50% of investment amount
Capital gains tax exemption on qualifying startup investments
Loss relief if startup fails
Capital Gains Tax rates:
10% on first £1M lifetime gains (Business Asset Disposal Relief)
20% on additional gains
vs. Dutch 36% on unrealized gains
Recent enhancements:
Extended EIS qualifying criteria
Increased investment limits
Expanded sector eligibility for advanced tech
France: French Tech Visa and Tax Advantages
JEI (Jeune Entreprise Innovante) status:
Corporate tax exemption for first profitable year
50% reduction for following year
Social security contribution reductions
R&D Tax Credit (CIR):
30% credit on first €100M R&D spending
5% on additional spending
Cash refundable for unprofitable startups
BSPCE (Stock options for growth companies):
Favorable capital gains treatment (flat 30% tax, not wage income)
No social contributions on stock option gains
vs. Dutch 65% of wage income treatment post-2027 reform
Germany: Zukunftsfonds and Growth Financing
€10 billion Zukunftsfonds (Future Fund):
Co-investment with private VCs in growth-stage companies
Focus on deeptech, climate, digital infrastructure
Explicit goal: Close Series B-C "valley of death"
Tax advantages:
No wealth tax (eliminated 1997)
Capital gains 60% tax-exempt for corporate investors in startups
Simplified stock option taxation
The Competitive Disadvantage
A 2023 survey by the Dutch Association of Tax Advisors found zero of 12 comparable countries use a capital growth tax similar to the Netherlands' unrealized gains system.
The Netherlands is creating a unique anti-competitive environment precisely when competitors are enhancing advantages.
Capital is mobile. Deeptech companies can incorporate anywhere in the EU. Investors can deploy capital across borders.
The unrealized gains tax creates a structural incentive for:
Dutch deeptech founders to incorporate in UK, Germany, or France
Dutch angel investors to invest in non-Dutch startups
International VCs to avoid Dutch fund structures
Dutch LPs to commit capital to non-Dutch funds
Network effects in innovation ecosystems are self-reinforcing: Success attracts capital attracts talent attracts more startups attracts more success. The Netherlands risks triggering a negative network effect spiral.
The Liquidity Paradox: When Paper Gains Become Real Tax Bills
The government's own explanatory memorandum acknowledged "liquidity risk" as the reason for exempting real estate and qualifying startup shares from unrealized gains taxation. This admission is telling: it reveals awareness that taxing unrealized gains creates genuine liquidity crises.
But the exemptions are too narrow. They protect some players while leaving critical ecosystem participants exposed.
Who Faces the Liquidity Nightmare
1. Employees with Stock Options
Post-2027 reform improves stock option treatment (65% taxable base vs 100%), but employees still face Box 1 wage income taxation at grant/vest.
Scenario:
Deeptech engineer receives options in year 2 (company qualifies for exemptions)
Options vest in year 5
Company raises Series C in year 6 at high valuation → employee exercises options, becomes small shareholder
Year 7: Company crosses 5-year age threshold → deemed sale triggered
Employee's shares valued at €200K on paper
Shift from Track 2 to Track 1 taxation
Year 8: Company raises Series D at 50% higher valuation
Employee's stake now "worth" €300K
Tax bill: €36K on €100K unrealized gain
Shares remain completely illiquid (no secondary market, no buyer)
That employee must pay €36K in cash from their salary to cover tax on shares they cannot sell.
2. Technical Co-Founders Below 5% Threshold
Early employees who join pre-incorporation or as first hires may receive 2-4% equity stakes, just below the 5% Box 2 protection threshold.
They are the most risk-taking, earliest believers, often working below-market salaries for equity. Yet they face the harshest tax treatment:
No Box 2 protection (hold <5%)
Subject to Box 3 unrealized gains after startup exemption expires
Annual tax bills on illiquid shares
Cannot afford to maintain positions through growth phase
Forced to sell early (if secondary market exists) or face annual tax bleeding
3. Academic Spinouts with University Ownership
Dutch universities often retain 10-30% ownership in spinout companies, creating complex ownership structures that may violate the "not >25% owned by non-startup business entity" independence requirement.
Even if the spinout qualifies initially, university ownership can create ambiguity about when exemption applies, adding administrative burden and risk.
4. Follow-On Investors in Maturing Companies
Growth-stage investors entering in Series B-C (years 5-7) may invest in companies that have already exceeded or are about to exceed the 5-year age limit.
These investors provide critical scale-up capital but receive no exemption protection:
Invest €10M at year 6 valuation
Company continues growing over years 7-10
Investor faces annual 36% tax on paper appreciation throughout
Even though the company remains private and illiquid
This explicitly discourages growth-stage investment in Dutch deeptech, the exact capital type TechLeap, TNO, and policy analysts identify as the primary constraint on scaling.
Constitutional and Political Realities: A Law Nobody Wants
Perhaps the most absurd aspect of this legislation is that even its supporters oppose it.
The Parliamentary Paradox
The bill passed the House of Representatives on February 12, 2026 with 93 votes, well above the 75 required majority.
Yet simultaneously:
A parliamentary majority (including parties that voted yes)passed motions demanding the government present a realized-gains-only alternative by Budget Day 2028
State Secretary Eugène Heijnen called the bill "an intermediate step"
The January 2026 coalition agreement explicitly states intention to convert Box 3 to pure capital gains tax
Why pass a law everyone plans to repeal?
The Constitutional Constraint
December 2021: Dutch Supreme Court ruling
Found the Box 3 "fictitious returns" system unconstitutional
Government required to implement actual-return-based system by 2028
Each year of delay costs treasury €2.3-2.4 billion
The government's dilemma:
Constitutional mandate: Must tax actual returns
Time constraint: Must implement by January 1, 2028
Administrative capacity: Tax authority can only implement mark-to-market system within timeline
Political reality: No consensus on better alternative
So they passed a law that:
Tax advisors warn creates capital flight risk
International experts condemn as internationally divergent
The government itself plans to replace
Creates perverse incentives for startups
All because the only constitutional solution the Belastingdienst (tax authority) could implement by deadline was a mark-to-market system.
The 2028 Replacement Window
There is a realistic possibility this law never fully takes effect:
Implementation: January 1, 2028
First tax year: 2028 (assessed on December 31, 2028 valuations)
Tax filing: 2029 (for 2028 tax year)
Budget Day 2028: Government must present realized-gains alternative
Timeline suggests:
Law takes effect January 2028
Throughout 2028, political pressure builds
Government presents alternative in September 2028 Budget Day
Legislative process in late 2028/early 2029
Potential replacement before first major tax bills come due in 2029
This is policy via constitutional crisis,driven by court deadlines rather than economic rationale.
Structural Solutions: What Actually Needs to Change
Assuming the 2028 replacement occurs, what should it look like? And what broader reforms does Dutch deeptech need?
Short-Term: Box 3 Reform Principles
1. Pure Realized Gains Taxation
Tax capital gains only upon actual sale
Maintain 36% rate if necessary for revenue, but only on realized transactions
Eliminates liquidity crisis entirely
2. Broaden Startup Exemptions
Extend age limit from 5 years to 10 years (matches realistic deeptech timelines)
Increase revenue threshold from €30M to €100M (allows genuine scaling)
7 of 10 largest tech companies globally are American
US venture capital: ~$238B (2023) vs Europe: ~€50B
China leads in manufacturing, hardware, batteries, solar:
BYD, CATL, Huawei, Xiaomi
60-80% of global solar panel production
70%+ of global battery production
Advanced manufacturing at scale
Europe's competitive advantage must be in deeptech:
Advanced materials (composites, semiconductors)
Precision manufacturing (ASML, Carl Zeiss)
Cleantech and energy transition
Biotech and life sciences
Quantum and advanced computing
The Netherlands specifically has world-class positions in:
Semiconductor equipment (ASML photolithography)
Agricultural technology (Wageningen ecosystem)
Quantum computing (QuTech Delft)
Offshore wind and maritime technology
Photonics and integrated optics
These advantages require patient, long-term capital to maintain.
The Geopolitical Stakes
European Commission President Ursula von der Leyen's 2024-2029 agenda explicitly prioritizes:
Technological sovereignty
Deep tech innovation
Reducing dependence on US digital platforms and Chinese manufacturing
Yet member states like the Netherlands actively undermine these goals with capital-hostile tax policies.
If Europe cannot fund its own deeptech champions:
US investors will own European IP
Chinese manufacturers will out-compete European production
Strategic dependencies on foreign technology will increase
Economic and security vulnerabilities will grow
The unrealized gains tax is a small policy in a large context but it moves in precisely the wrong direction.
Conclusion: A Test of National Priorities
The Netherlands stands at a crossroads.
On one side: A future as a leading European innovation hub, where deeptech companies commercialize world-class research, where patient capital funds breakthroughs, where technical talent builds the next generation of advanced technologies.
On the other: A future as a cautionary tale, where talented founders incorporate in London or Berlin, where Dutch capital deploys in foreign startups, where universities produce excellent research that gets commercialized elsewhere.
The unrealized gains tax tips the balance toward the latter.
This legislation reveals a fundamental misalignment in Dutch innovation policy. The government simultaneously:
✅ Subsidizes R&D labor at 36-50% (WBSO)
✅ Taxes innovation profits at 9% (Innovation Box)
❌ Taxes unrealized investment gains at 36% annually
❌ Creates artificial exemption cliffs that discourage scaling
❌ Adopts internationally divergent policies that drive capital elsewhere
You cannot build a thriving deeptech ecosystem while systematically punishing the capital that funds it.
What Needs to Happen
For policymakers:
Deliver on the 2028 replacement commitment with pure realized-gains taxation
Extend startup exemptions to 10 years and €100M revenue
Create growth-stage capital programs (Dutch Zukunftsfonds equivalent)
Reform stock option taxation to capital gains treatment
Simplify university technology transfer
For founders:
Understand your exemption status and tax implications
Plan strategically around exemption cliffs
Communicate proactively with investors
Participate in advocacy efforts
Build for long-term value despite short-term policy headwinds
For investors:
Demand clarity on tax treatment before deploying capital
Support policy advocacy for startup-friendly reforms
Consider fund structures that protect LPs from unrealized gains exposure
Maintain patient capital commitments despite policy uncertainty
For the ecosystem:
Amplify the message: This law threatens Dutch competitiveness
Share real company stories demonstrating impact
Benchmark internationally: Show what competitors are doing
Build coalition across founders, investors, universities, corporates
The Window Is Narrow
The 2028 Budget Day replacement represents a genuine opportunity to fix this policy before it causes lasting damage. But that only happens with sustained pressure.
Deeptech companies are fundamentally different from software startups. They require:
Longer timelines (10-15 years vs 5-7 years)
More capital (€50M-€200M vs €10M-€30M)
Patient investors accepting illiquidity
Risk tolerance for technical failure
A tax system that punishes paper gains on illiquid, long-duration investments is structurally incompatible with deeptech development.
The Netherlands has the research base, the talent, the infrastructure, and the R&D incentives to be a European deeptech leader. Don't let a misguided wealth tax implementation destroy what decades of smart policy built.
The choice is clear. The time is now. The stakes couldn't be higher.
Build the future. Fix the tax code. Choose innovation.