Share article

Dutch tax policy has a talent for making "simple" investing feel like an endurance sport. The latest example: a proposed 36% levy on "actual returns" in Box 3 (the Dutch tax bucket for savings and investments) that would have treated crypto and listed stocks as if everyone has spare cash lying around to pay tax on gains they have not even realized yet. Now the government says it will amend the proposal ahead of the broader Box 3 overhaul. [1] Because of course it will.

Here are the core numbers and the core problem:

  • Rate discussed: 36% (the Box 3 tax rate policymakers have been working with). [2]
  • Tax base under the draft plan: "Actual return," including price appreciation in assets such as crypto and publicly traded shares, even if unsold.
  • Timeline: the new system is tied to the long-running Box 3 reform that has been targeted for later this decade (often referenced around 2027 to 2028, depending on legislative and implementation constraints). [3]
The government's decision to revise the draft is not a sudden ideological awakening. It is what happens when a policy runs head-first into liquidity reality, volatility math, and years of court pressure to fix Box 3 properly. [4]

Enjoy articles without ads?

Register for free and get unlimited access to all articles.

What the Netherlands is trying to do with Box 3

Box 3 taxes wealth held as savings and investments. For years, it relied on a deemed (assumed) return model, where taxpayers were taxed as if their assets produced a certain yield, regardless of what actually happened in markets.

That approach has been under sustained legal and political strain. Dutch courts have repeatedly scrutinized whether taxing fictional returns is fair, especially when real returns were lower or negative. The reform push aims to move to something closer to an "actual return" framework, meaning tax is calculated based on what you truly earned.

The controversy is in the details: if "actual return" includes unrealized gains, then taxpayers can owe money on assets that increased in price on paper but were never sold. That is manageable for some portfolios. It is a headache for others.

Why the 36% unrealized gains idea blew up (fast)

Taxing annual price changes sounds tidy on a whiteboard. In practice, it creates three obvious friction points.

1) Liquidity: tax bills do not pay themselves

If your Bitcoin$62,588.20 position rallies over the year, an accrual tax model says you have "income" even if you never sold. The tax authority still wants cash. That can force selling to cover the bill, particularly for investors who hold volatile assets or who are otherwise cash-light.

This is not a niche scenario. Crypto is known for sharp drawdowns and sudden spikes. A system that taxes the spike immediately, but only recognizes losses under specific rules, can quickly feel one-sided to taxpayers.

2) Volatility: you can get taxed at the top, then watch the bottom arrive

Accrual taxation interacts badly with the way crypto and equities behave across calendar boundaries.

A simplified example shows the problem:

  • Asset rises from EUR 50,000 to EUR 80,000 by year-end.
  • Under an unrealized gains approach, the EUR 30,000 increase may be treated as taxable "return."
  • At a 36% rate, that implies a tax claim tied to the year-end mark-to-market valuation.
  • If the asset then drops back to EUR 55,000 early the next year, the taxpayer may have paid meaningful tax on gains that evaporated.

Policymakers can mitigate this with robust loss offset rules and carryforwards, but the public debate suggests many observers were not convinced the initial design solved the timing mismatch.

3) Complexity: "actual return" is easy to say, hard to administer

Crypto introduces additional complications that do not show up in a savings account:
  • Multiple exchanges and wallets (including self-custody)
  • Forks, airdrops, staking rewards (what counts as income and when?)
  • Stablecoins, wrapped assets, and token migrations
  • Valuation standards at year-end (which pricing source, which timestamp?)

The more "actual" the system becomes, the more it needs crisp definitions. Without them, enforcement becomes inconsistent and taxpayers end up guessing.

What "amending the proposal" likely means in practice

The government's signal that it will revise the unrealized gains element is best read as a design pivot, not an abandonment of reform. Officials still need a Box 3 model that meets legal expectations and can be implemented at national scale.

While exact amendments depend on the final legislative text, the policy toolset is fairly standard. The revisions may include a combination of the following:

A realization-based approach for some assets

A common compromise is to tax realized gains (when you sell) for assets where annual mark-to-market taxation is seen as too disruptive, while keeping accrual treatment for more liquid, easily valued instruments. Another approach is to use realization for illiquid assets and accrual for liquid assets, but crypto's "liquid yet volatile" profile makes it a political hot potato either way.

A deferral option or liquidity relief

If accrual taxation remains, lawmakers can add mechanisms that reduce forced selling, such as deferring the tax on paper gains until sale, or allowing payment plans under certain conditions. These features add complexity, but they directly address the "tax without cash" complaint.

Stronger loss offset and carryforward rules

Loss relief is the difference between "tough but workable" and "tax roulette." If the state taxes annual gains, taxpayers will expect symmetrical treatment for annual losses, including the ability to offset gains in future years.

How this hits crypto and stock investors specifically

For Dutch residents with crypto or equity exposure in Box 3, the direction of travel matters as much as the final rate.
  • If the system stays accrual-based: expect more frequent valuation reporting, clearer sourcing requirements for prices, and potentially more year-to-year tax volatility.
  • If the system shifts toward realization: expect fewer forced sales but potentially more detailed cost-basis tracking and transaction reporting, especially for crypto where trades can be frequent and fragmented.
Either way, the reform is not just a "crypto tax." It is a re-architecture of how the Netherlands taxes household investment wealth, and crypto simply happens to be the asset class most likely to stress-test the design.

The political reality: Box 3 reform has deadlines, and court pressure

The Netherlands is not redesigning Box 3 for fun. The existing system has faced sustained legal criticism, and the longer the state relies on assumed returns, the more pressure builds to deliver something that better matches real-world outcomes.

That creates an uncomfortable constraint: policymakers need a model that is:

  • Legally durable
  • Administratively feasible
  • Perceived as fair enough to survive politically
  • Ready on a timeline that tax authorities can actually implement (a non-trivial constraint, especially for data-heavy frameworks)

Amending the unrealized gains feature is a tactical retreat to keep the broader reform alive.

Takeaways (clearly labeled, because everyone is tired)

  • The 36% number is not the whole story. The real fight is about the tax base: unrealized versus realized returns, and how losses are treated.
  • Crypto's volatility makes it the worst possible test case for annual mark-to-market taxation, especially without ironclad loss symmetry.
  • The government is recalibrating, not abandoning Box 3 reform. The policy objective remains "actual return," but the method is being reconsidered.

What to watch next

  • Draft amendment details: Look for whether crypto and listed shares remain under mark-to-market treatment, shift to realization, or get a hybrid rule.
  • Loss treatment language: Watch for explicit loss carryforward provisions, offset limitations, and whether losses are fully symmetrical with gains.
  • Valuation and reporting standards for crypto: Any final bill needs practical rules on pricing sources, custody disclosures, and treatment of staking and airdrops.
  • Implementation timeline realism: If the reform target sits around 2027 to 2028, delays are still possible if the administrative burden is judged too high.
  • Market behavior around year-end: If accrual taxation survives in any form, year-end portfolio management and tax-loss harvesting will become more prominent, because incentives always win.

Box 3 reform is supposed to make taxes reflect reality. Taxing "profits" you did not realize was always going to invite questions about which reality the government had in mind. The upcoming amendments should clarify whether policymakers are aiming for practicality, principle, or the usual compromise that makes everyone equally annoyed.