How Safe Is Your Money in an ETF? The Real Risks — and the Imagined Ones

In short: For a broad, low-cost ETF held for decades, the catastrophic scenarios — provider bankruptcy, broker bankruptcy, fraud — are largely handled by the way the product is built. The real ways people lose money look different: panic selling in a crash, purchasing power lost to too much safety, costs. And the most dangerous risk of all isn't in the product — it's in you. There are no honest probabilities to hand you, but there is a clear ranking of what you should actually fear.

"But what if I just lose everything?" — that fear keeps a lot of people from ever starting. It's a fair question. But most of the fear is aimed at the wrong risks. Let's sort them into three buckets: the ones the structure absorbs, the real ones, and the rare extremes.

1. Risks the structure already absorbs

The fund provider goes bankrupt. A UCITS fund's assets are held separately from the management company's own balance sheet (a provider such as iShares or Xtrackers). If the provider goes under, your holding isn't part of the bankruptcy estate. Under the rules, the depositary takes over and the fund is either transferred or wound up and distributed to investors. Your units stay yours.

An important exception that often gets missed: this doesn't apply to ETCs and ETNs. Exchange-traded commodity and crypto products are legally debt securities, not segregated fund assets — so you carry issuer risk. If you hold a gold ETC, know that the ETF-style insolvency protection doesn't work the same way there. Reputable gold ETCs are instead backed by physical gold; check that before you buy.

The synthetic ETF case. A swap-based ETF doesn't buy the index constituents; it gets the index return through a contract with a bank, which creates counterparty risk. UCITS rules cap it: exposure to a single counterparty may not exceed 10% of the fund's net asset value (5% if the counterparty isn't a credit institution). In practice providers stay well below that, because swaps are typically reset daily and positions are collateralised, often above the required level. A residual risk remains — but it's capped and monitored.

Your broker goes bankrupt. The securities in your account belong to you, not the broker, and are held separately. In an insolvency they're returned to you or transferred to another institution — with no upper limit. Deposit insurance (€100,000 per customer per bank in the EU) separately covers the cash on your settlement account.

Alongside that sits investor compensation: 90% of the claim, capped at €20,000. That number worries people unnecessarily. It only applies in the rare case where securities can't be handed back at all — think embezzlement. In an ordinary insolvency your units remain your property and move to the next broker, however large the account.

Fraud. The classic total loss to fraud hits people who put money into unregulated "products" promising dream returns. A regulated UCITS ETF held through a regulated broker structurally rules that category out.

In short: "the provider runs off with my money" is one of the least likely outcomes for a regulated UCITS ETF.

2. The real risks — where people actually lose money

Panic selling. By far the biggest real driver of losses. As long as you hold, a crash is only a paper loss. Selling at the bottom is what makes it real — and then missing the recovery often costs more than the crash itself.

Morningstar's annual "Mind the Gap" study tries to measure the price of this: over the ten years to the end of 2024, the average invested dollar earned 7.0% a year while the funds themselves returned 8.2% — a gap of 1.2 percentage points annually, around 15% of the funds' total return.

Two honest caveats: the data covers the US fund market, and the size of the number is academically contested — a study using the same data finds substantially less, because part of the gap is a methodological artefact rather than bad timing. The exact magnitude is disputed; the direction isn't. Trade a lot, and you trail your own fund.

For a sense of scale: at €300 a month over 30 years, giving up 1.2 percentage points of return costs roughly €76,000 at the end.

Too much safety. Staying in a savings account out of fear isn't a nominal loss, but it's a real one: inflation eats your purchasing power. This isn't theoretical. In Germany, for example, the real return on instant-access savings was −2.4% in April 2026 and −2.2% on passbook savings — the 64th consecutive month in negative territory. More than five years without a break. Historically, negative real rates on savings deposits are closer to the rule than the exception. It feels safe and is a slow bleed. → Savings vs investing

Costs. Not a crash but a leak: ongoing fees add up noticeably over decades. A worked example — €300 a month, 30 years, an assumed 7% gross return:

• at 0.20% ongoing charge: about €352,000
• at 1.50% ongoing charge: about €274,000

That's roughly €78,000 apart, about 22% of the final amount — from 1.3 percentage points of fees. You paid in €108,000 either way. (Illustrative example at a constant return, not a forecast.)

Concentration. A single stock, one sector, or one country — here you can lose for real and for good. A broad world ETF defuses this: the MSCI World holds around 1,280 companies across 23 developed markets; the MSCI ACWI, which adds emerging markets, around 2,460. Worth knowing: that spread isn't even — the US currently accounts for roughly 70% of the MSCI World. Broad means widely spread, not equally weighted. → What is UCITS?

Entry timing. A large lump sum invested right at the peak hurts. A savings plan smooths the entry over time. → Dollar-cost averaging

3. The extreme risks nobody can fully hedge away

Let's stay honest: there are scenarios that even the best structure can't fully protect against.

Total market or country wipeout. War, expropriation, a permanently closed exchange. This isn't a thought experiment — it happened in 2022: MSCI reclassified its Russia indexes on 9 March at a price of effectively zero, and FTSE Russell had already removed Russian constituents at zero value on 7 March. The trigger wasn't the price collapse alone but unsellability: a closed exchange, suspended trading in depositary receipts, and a ban on foreigners selling locally.

What matters is who this actually hit — and that's precisely where diversification shows its worth:

• Russia country ETFs: effectively a total loss.
• Broad emerging-market ETFs: Russia made up around 3% of the index before removal. Funds wrote the position down to zero — a loss of roughly 3%. Painful, not catastrophic.
• MSCI World: unaffected. Russia was never in it, because it isn't a developed market.

So the extreme scenario is real. But it wipes out a portfolio only if you concentrated on a single country beforehand. The one effective answer is exactly the spread a broad ETF already gives you.

Currency risk. As a euro-based investor, a world ETF holds most of its value in foreign currency — with the US at around 70%, a large part rides on the dollar. That risk doesn't disappear, and the common line that it "evens out over time" is too convenient.

The more accurate version: exchange rates have no systematically positive or negative expected return. Over short periods they matter a lot; over long ones they recede behind equity returns, and whether hedging would have helped or hurt moves in cycles that can't be forecast reliably. Hedging, by contrast, costs money for certain — hedged share classes are typically more expensive than their unhedged siblings. That's why most long-term equity investors skip it. A common compromise: leave equities unhedged, hedge bonds — because there, currency swings would otherwise swamp the entire return.

And the "probabilities"?

You asked about probabilities — so here's the honest answer: there is no precise percentage for "the market never recovers," and anyone who gives you one is inventing it. What can be said seriously is how things have gone so far.

For the US market, which has the longest data series, roughly: the average bear market since 1928 ran a little over 400 days from peak to trough, with a decline of about a third. The severe cases: the dot-com crash of 2000–2002 at around −49%, the 2008 financial crisis at around −48%, both taking several years to fully recover. In 12 of the 15 bear markets since 1945, investors were back to break-even in under three years; the outliers were 1973–74, 2000 and 2008, each taking more than four. At the other end: the 2020 covid drop was recovered in about four months.

Two caveats so these numbers don't promise more than they can. First, they mostly describe the US market, not a world index. Second, they're nominal — adjust for inflation and the recoveries took considerably longer. After the 1973 oil crisis, more than nine years in real terms.

The practical core stays the same: whether you end up losing money depends less on the product than on your behavior.

Bottom line

Spread broadly, keep costs low, hold for the long run, and don't sell in the storm — that alone absorbs most of the catastrophic scenarios. The extreme risks are real but rare, and the only answer to them is diversification. The biggest avoidable risk is in the driver's seat: you.

---

This article is general information, not investment advice. Worked examples are illustrations at assumed returns, not forecasts. Past performance does not guarantee future results.