Accumulating vs distributing ETFs:
measuring the real tax drag
Every time a distributing ETF pays a dividend, you owe tax on it — even if you reinvest immediately. That recurring leakage is tax drag. This study uses real monthly NAV and dividend data from IUSA/CSPX (S&P 500) and VWRL/VWCE (FTSE All-World) to measure exactly how much compounding you lose.
Some of the links on this site are affiliate links, meaning we may earn a commission at no extra cost to you if you sign up through them. This does not affect our reviews or recommendations — we only feature products we genuinely believe are useful for investors. This site provides educational content only, not personalized investment advice. Investments can lose value and past performance does not guarantee future results. You are responsible for your own financial decisions and for confirming the tax and legal rules that apply in your country.
What the real data shows
- S&P 500 (IUSA, 15.3 yr): at 26% dividend tax, distributing investors ended with 7.4% less — about €6,934 lost on a €10,000 start.
- FTSE All-World (VWRL, 13.7 yr): at 26%, the gap was 6.9% — roughly €3,503 on €10,000.
- Accumulating ETFs (CSPX, VWCE) avoid this drag by reinvesting dividends inside the fund — deferring the taxable event.
- In some countries accumulating ETFs are still taxed annually — German Vorabpauschale, Dutch Box 3, deemed distributions.
- In those cases the advantage shrinks or disappears entirely.
- FX drag and wide spreads can easily outweigh the share-class difference for many investors.
- Always check your own country’s rules first.
What “tax drag” means in ETFs
Tax drag is any recurring tax leakage that reduces the capital available to compound. With ETFs it comes from several layers.
- Distribution taxation: taxes paid when dividends are paid out from distributing share class.
- Reinvestment gap: even immediate reinvestment is net-of-tax, so less capital compounds from day one.
- Fund-level withholding: taxes withheld before dividends reach the ETF (see UCITS vs US study).
- Annual deemed taxation: Vorabpauschale, Box 3, or similar systems that neutralise accumulating’s advantage.
How the simulation works
Instead of illustrative models, this study uses real monthly NAV and dividend data from two distributing UCITS ETFs. For each fund we simulate an investor who starts with 10,000, receives every actual quarterly distribution, pays tax at the assumed rate, and immediately reinvests the net amount at the closing price. The “0% tax” line represents what an accumulating share class achieves — no distributions leave the fund, so no taxable event occurs.
| Input | S&P 500 pair | All-World pair |
|---|---|---|
| Distributing ETF | IUSA (iShares S&P 500 Dist) | VWRL (Vanguard FTSE All-World Dist) |
| Accumulating equivalent | CSPX (iShares S&P 500 Acc) | VWCE (Vanguard FTSE All-World Acc) |
| Period | Oct 2010 – Mar 2026 (15.3 yr) | Jun 2012 – Mar 2026 (13.7 yr) |
| Starting capital | 10,000 (fund currency) | |
| Tax scenarios | 0% (= accumulating), 15%, 26%, 30% | |
| Reinvestment | Net-of-tax dividends reinvested at month-end close | |
Data source: Alpha Vantage (monthly adjusted). Dividends as reported by the exchange. The simulation isolates the tax-on-distributions effect only — it does not model capital gains tax at sale, broker fees, or FX conversion costs.
S&P 500: growth of 10,000 — IUSA at different dividend tax rates
The green line shows full reinvestment (0% tax = accumulating equivalent). Each additional line applies a higher tax rate to each quarterly payout before reinvesting the remainder. Over 15.3 years, the gap grows to 7.4% at 26% and 8.5% at 30%.
| Scenario | End value | CAGR | Gap vs acc |
|---|---|---|---|
| 0% tax (acc equivalent) | 93,374 | 15.67% | — |
| 15% dividend tax | 89,310 | 15.33% | −4.35% |
| 26% dividend tax | 86,440 | 15.09% | −7.43% |
| 30% dividend tax | 85,420 | 15.00% | −8.52% |
IUSA (iShares S&P 500 UCITS Dist), Oct 2010 – Mar 2026. Growth of 10,000 in fund currency with dividends reinvested net of tax at month-end close.
FTSE All-World: growth of 10,000 — VWRL at different dividend tax rates
Same methodology, different index. VWRL’s quarterly distributions are taxed and reinvested. Over 13.7 years the gap reaches 6.9% at 26% and 7.9% at 30%. The accumulating equivalent is VWCE.
| Scenario | End value | CAGR | Gap vs acc |
|---|---|---|---|
| 0% tax (acc equivalent) | 50,676 | 12.59% | — |
| 15% dividend tax | 48,625 | 12.25% | −4.05% |
| 26% dividend tax | 47,173 | 12.01% | −6.91% |
| 30% dividend tax | 46,656 | 11.92% | −7.93% |
VWRL (Vanguard FTSE All-World UCITS Dist), Jun 2012 – Mar 2026. Growth of 10,000 in fund currency with dividends reinvested net of tax at month-end close.
Sensitivity: end-value gap vs dividend tax rate
How much of the final portfolio value do you lose at each dividend tax rate? The S&P 500 line is steeper because IUSA had a longer compounding period (15.3 vs 13.7 years) — more time for the drag to compound.
Gap = (taxed end value ÷ full-reinvestment end value − 1). Based on real quarterly dividends from each fund’s full history.
What actually changes the result
Before changing share class, check these. Country rules and execution costs often matter more.
- Whether accumulating funds are taxed annually — Vorabpauschale (Germany), Box 3 (Netherlands), deemed distributions.
- Dividend tax rate vs capital gains tax rate, and whether gains can be deferred until sale.
- Tax-advantaged accounts (if available) that can neutralise the share-class difference entirely.
- Dividend yield level — higher yield means a larger distribution-tax surface area each year.
- Withholding taxes before cash reaches the fund (often invisible — see UCITS vs US study).
- Tracking difference and spreads — execution drag can dominate in smaller accounts.
Practical takeaways for UCITS investors
At 26% tax, real data shows a 7.4% end-value gap for S&P 500 and 6.9% for All-World over 13–15 years. It matters — but FX spreads or a careless broker can cost more.
Unless you want the income cashflow or your tax system makes it neutral, there’s no reason to choose distributing as the default. The path of least resistance favours accumulating.
Where your country taxes accumulating funds annually (Vorabpauschale, Box 3), the gap narrows significantly. Check before assuming the acc share class is always better.
One listing, consistent buys, minimal currency conversions, simple rebalancing. Never optimise one drag layer while ignoring larger ones like FX friction or wide spreads.
Minimise drag in your own portfolio
The biggest practical win is pairing the right share class with low FX friction and repeatable execution. IBKR gives you institutional FX rates and full UCITS ETF access. TradingView helps you research and screen ETFs before you buy.
Or use the calculator below to model your own inputs.
Some of the links on this site are affiliate links, meaning we may earn a commission at no extra cost to you if you sign up through them. This does not affect our reviews or recommendations — we only feature products we genuinely believe are useful for investors. This site provides educational content only, not personalized investment advice. Investments can lose value and past performance does not guarantee future results. You are responsible for your own financial decisions and for confirming the tax and legal rules that apply in your country.
Go deeper on drag
Share class is one layer of the total drag stack. Use these together: wrapper → withholding → FX → execution.
Frequently asked questions
Are accumulating ETFs always better than distributing ETFs?
No. It depends on whether your tax system treats accumulating growth as genuinely deferred (the advantage) or taxes it annually anyway — like the German Vorabpauschale or Dutch Box 3 wealth tax. In those cases the share-class advantage shrinks or disappears. It also depends on whether you actually need the income cashflow.
If I reinvest distributions immediately, is there still tax drag?
Yes. If distributions are taxed at payout, you reinvest less than 100% net of tax. Even immediate reinvestment doesn’t recover the portion paid in tax. In our S&P 500 data, even at 15% tax the investor ended with 4.3% less than the accumulating path over 15.3 years — despite reinvesting every distribution immediately at month-end close.
Does ETF domicile matter for tax drag?
Yes. Withholding can occur at different layers depending on fund domicile and underlying holdings. This study focuses on the share-class timing effect only — it doesn’t model fund-level withholding. Use the UCITS vs US ETF total drag study for the full stack including withholding, TER, and execution drag.
What matters more than the accumulating vs distributing choice?
Often: FX spreads, repeated currency conversions, wide ETF spreads, and inconsistent contribution behaviour. Fix execution drag first if it is large — it can easily dwarf the share-class difference, especially for smaller portfolios. See the FX drag study for the numbers.
Why use IUSA and VWRL instead of CSPX and VWCE directly?
Accumulating ETFs have no distribution events — their NAV already reflects full reinvestment. To measure tax drag we need actual dividend amounts and dates, which only the distributing share class provides. The 0% tax line in our simulation is mathematically equivalent to what the accumulating fund achieves.
How large is the tax drag in annual CAGR terms?
For the S&P 500 at 26% dividend tax: 15.67% CAGR (accumulating equivalent) vs 15.09% (distributing) — a difference of 0.58 percentage points per year. Small annually, but it compounds to a 7.4% end-value gap over 15.3 years. The gap is non-linear: it grows faster at higher dividend yields and longer periods.
QuantRoutine provides educational content only. Nothing on this page is an offer, solicitation, or recommendation to buy or sell any security or to open an account with any specific broker. Investments can lose value, and past performance does not guarantee future results. You are responsible for your own investment, tax, and legal decisions. Always confirm the tax rules that apply in your country before acting on any information here.