For small BVs, the dividend date matters less when cash, risk and tax relief part company.
A founder signs off a dividend after a careful year. The accounts are closed. Solvency has been checked. The holding BV is waiting for the cash. Then comes the familiar question to the bookkeeper: do we withhold 15 percent, or does an exemption apply?
The signal has to become readable
That is ordinary company administration. It should feel that way. Dividend tax in the Netherlands starts from a clear base: 15 percent, a return, and payment to the Belastingdienst within one month when withholding is due.
The pressure rises when the legal form and the economic story drift apart.
The Table Where The Question Starts
The shareholder register is only the first page. Dutch anti-dividendstripping rules look further when relief, refund, reduction or credit depends on who is ultimately entitled to the proceeds.
That is the key shift. The question is no longer only who held the shares on paper. It is also who carried the risk, who kept the dividend return and who gained the tax benefit.
For a normal holding structure, that distinction is often harmless. A holding BV, an operating BV, a foreign shareholder or an investment account does not create a problem by itself. The file becomes sensitive when the dividend, the shares, the hedge, the counter-payment and the tax claim stop telling the same story.
Since 1 January 2024, the Ministry of Finance has already moved the rules in that direction. The burden of proof changed. The record date was codified. Group situations were defined more sharply. A report to the Tweede Kamer on 27 June 2025 then set out further policy work.
The direction is clear. The law is moving from snapshot ownership toward economic ownership in the wider commercial story.
Cash Moves Before Certainty
For a small company, the dangerous moment is usually not a court hearing. It is the cash forecast.
Dividend tax relief can look like money before it has fully earned its place. A refund may be expected. A withholding exemption may support a distribution. A credit may reduce another payment. The money then moves into private tax, intercompany debt, a new investment or a loan repayment.
If the tax position later changes, the pressure lands in the ledger and the bank account. That is where the practical risk sits.
The Ministry’s 2025 report makes that plain. It described five court procedures involving €241 million in tax at stake. It also reported settlements covering €362 million in tax, including penalties, plus agreements that no further refund requests would be submitted for €142 million.
What the signal changes
Those numbers belong to larger enforcement files, not to the average family BV. Still, they show how quickly a dividend position can affect cash once the tax treatment is challenged.
Courts And Policy Are Moving The Same Way
The courts are reading the whole transaction set, not just the voucher.
In January 2024, the Hoge Raad dealt with corporate income tax, dividend-tax credit, opbrengstgerechtigde, uiteindelijk gerechtigde, dividendstripping and burden of proof. In March 2025, the Amsterdam Court of Appeal accepted that the inspector had made the anti-dividendstripping conditions plausible in a refund or credit case. In February 2026, the Advocate General addressed a file involving a foreign group company, cum-dividend shares and matching ex-dividend futures.
The detail is technical. The business lesson is not. If relief is claimed around a complex movement of shares or cash, the courts want the full chain.
The policy work points the same way. The Ministry has studied four further routes: a net-return approach, a pension-fund measure, clarification for group situations and a German-Austrian style economic-risk model. Those are policy options, not current Dutch rules.
The net-return idea is built around the retained dividend return after connected transactions. The pension-fund idea asks whether the dividend is tied to ordinary pension activity. The group option looks at related parties splitting one arrangement. The foreign models focus on economic risk around the record date.
For normal owner-managed BVs, the message stays calm. Ordinary dividends are not suddenly suspect. But the file has to be clean.
What The Small Company Should Have In The File
Return to the founder at the table.
Her dividend from the operating BV to the holding BV may be ordinary. If the exemption applies, the process can stay simple. But if shares moved close to the dividend date, if a foreign shareholder entered, if financing was arranged around the same time, or if an investment portfolio used lending, repos, futures or other hedges, the same dividend deserves a slower reading.
The practical record should line up. The board decision, solvency reasoning, shareholder status, dividend availability date, withholding analysis, return deadline, payment trail and ledger entries should tell one story.
When an adviser relies on an exemption or a treaty route, the business reason should still be readable in plain language. A tax file that only makes sense in statute references is a weak file.
What founders should check
For pension funds and asset managers, the issue has another layer. DNB’s prudent-person guidance puts participant interest, investment policy, risk control, procedures and outsourcing discipline at the centre of pension investing. If pension status creates tax value inside a transaction, the governance record should explain why the trade fits the mandate.
Tax status is not a business model by itself.
The Data Layer Is Getting Tighter
The future procedural layer is already visible.
The EU FASTER directive must be transposed by 31 December 2028 and apply from 1 January 2030. It is mainly a framework for faster and safer withholding-tax relief. It does not itself create a general Dutch anti-dividendstripping rule.
Still, faster relief needs cleaner data. Identity, entitlement, intermediary chains and claim routes will matter more in practice.
AFM’s SFTR context points in the same direction. Securities lending, borrowing, repos, buy-sell back transactions and margin lending all sit in a world where transaction details and changes can be reconstructed. SFTR does not decide dividend-tax entitlement. It does make timing, counterparties and collateral less invisible.
That matters for any founder who likes to leave dividends to the year-end tidy-up. The cleaner habit is earlier and simpler.
The Calm Work Before Distribution
A short pause before money leaves the company saves trouble later.
What exactly is being distributed? On which date was it made available? Is 15 percent withholding due, or is a clear exemption being used? If relief is claimed, who is the ultimately entitled party? Did that party also carry the economic risk and keep the real dividend return?
For the founder, the dividend conversation belongs before the transfer. For the adviser, the explanation should survive without theatrical tax language. For the bookkeeper, the ledger must agree with the contracts and bank statements.
The shareholder register still matters. It always will. But when cash, risk and tax benefit separate, it is only the first page.
The real question is simpler. Can the business tell, calmly and consistently, who carried the dividend and who kept the advantage?
Sources
- PwC Nederland
- Rijksoverheid
- Ministerie van Financiën via Open Overheid
- Belastingdienst
- Belastingdienst
- Wettenbank
- DNB
- AFM
Referenced in the article
Column | Governance
Dutch Tax Refunds Put Identity, Evidence and Cash in One Chain
A refund is not only tax arithmetic when access, data, evidence and payment meet at the same.
Column | Ledger & Tax
Why You Can’t Always Take What You Earned: The Dutch BV Distribution Test
Profit on your financial statements does not equal cash you're allowed to take out of your Dutch.
Column | Ledger & Tax
When a Criminal Conviction Becomes a Tax Liability: What I Learned from a Dutch Director’s Legal Nightmare
I've spent years watching directors navigate corporate liability in the Netherlands.
The Polder is written for readers who need the Dutch business environment translated into practical meaning. Corrections, source policy and editorial accountability are part of the publication record.
