Physical vs Synthetic ETFs Explained

If you are comparing ETFs and you have come across the phrase physical vs synthetic ETFs, the good news is that the idea is less scary than it sounds.

An ETF is usually designed to track something: an index such as the S&P 500, MSCI World or FTSE All-World, a bond index, a sector index or another basket of assets. The question is: how does the ETF actually copy that index?

That copying process is called replication.

Some ETFs copy an index by buying the underlying investments directly. These are known as physical ETFs.

Other ETFs copy an index using a financial contract called a swap. These are known as synthetic ETFs.

Neither label automatically makes an ETF good or bad. They are different structures with different trade-offs. My aim here is not to tell you which one to buy, because that would be financial advice. It is to explain the moving parts in plain English so you know what to look for before you make your own decisions.

As always, your capital is at risk. ETF prices can fall as well as rise, and past performance is not a reliable guide to future returns.

Quick answer: physical vs synthetic ETFs

A physical ETF aims to track an index by holding the actual shares, bonds or other assets in that index.

A synthetic ETF aims to track an index using a swap agreement with another financial institution. Instead of directly holding all the index assets, the ETF receives the return of the index through that contract.

The simple version:

That is the overview. The details matter.

What ETF replication means

ETF replication is simply the method an ETF uses to follow its benchmark index. Vanguard’s ETF replication guide gives the same broad split: physical ETFs hold underlying securities, while synthetic ETFs use derivatives, mainly swaps, to follow an index (Vanguard).

If an ETF tracks an index of 500 companies, there are a few ways it might try to mirror that index:

1. Buy all 500 companies in the correct proportions.
2. Buy a representative selection of those companies.
3. Use a swap contract that pays the ETF the return of the index.

Those three ideas broadly map to:

• Physical full replication
• Physical sampling, sometimes called optimised sampling
• Synthetic replication

When I first started looking at ETFs, I cared mainly about the big obvious things: what index it tracked, whether it was accumulating or distributing and whether it fitted into a simple long-term approach. The replication method was one of those details I saw in factsheets but did not properly understand at first.

Now I see it as one of the checks worth making before investing in any ETF, alongside the index, fees, currency, fund size, domicile, dividend policy and platform availability.

If you are still getting comfortable with the ETF basics, I would start with the wider ETF section here: ETFs. If you are completely new to investing, the gentler starting point is Start Here.

What is a physical ETF?

A physical ETF buys the actual investments it is trying to track.

If it tracks an index of shares, it holds shares. If it tracks an index of bonds, it holds bonds. In beginner terms, the fund is trying to own the real ingredients rather than recreate the recipe through a contract.

There are two main types of physical replication:

• Full replication
• Sampling

Full replication explained

With full replication, the ETF aims to hold every security in the index in roughly the same weight as the index.

For example, if a company makes up 4% of the index, a fully replicated ETF would aim to hold around 4% of the fund in that company.

This is the easiest version to understand. If the index contains 100 shares, the ETF tries to hold those 100 shares.

Full replication can make sense when the index is relatively straightforward, liquid and not too expensive to copy. A large developed-market share index is often easier to replicate physically than a niche index with lots of smaller or harder-to-trade holdings.

The appeal is obvious:

• It is simple to understand.
• The holdings usually look familiar.
• It can feel more transparent because the ETF owns the underlying assets.

But full replication is not automatically perfect. The ETF still has costs, trading spreads, tax drag, rebalancing and operational details. It can still perform slightly differently from the index.

Sampling explained

With sampling, the ETF does not hold every single security in the index. Instead, it holds a representative selection designed to behave similarly to the full index.

You may also see this described as optimised sampling.

This can happen when an index has a large number of holdings, some holdings are expensive to trade or some assets are difficult to access efficiently. Rather than buying everything, the ETF provider uses a model to choose a smaller basket that aims to track the index closely.

A simple example:

• An index contains 1,500 shares.
• The ETF holds 700 of them.
• The provider chooses those 700 to reflect the wider index by sector, country, company size and other characteristics.

The ETF is still physical because it holds real securities. It just does not hold every security in the index.

Sampling can be efficient, but it adds an extra layer to understand. The ETF is now relying on the provider’s selection process to behave like the index. If the sample does not match the index closely, the ETF may drift away from it.

That does not mean sampling is bad. Many large, mainstream ETFs use sampling. It simply means you should know whether the ETF is fully replicating or sampling, rather than assuming all physical ETFs work in the same way.

What is a synthetic ETF?

A synthetic ETF tracks an index using a swap contract.

A swap is an agreement between the ETF and another financial institution, often an investment bank. In simple terms, the swap provider agrees to deliver the return of the index to the ETF.

Xtrackers describes synthetic replication in similar terms: the ETF uses derivatives, usually swaps, to track the reference index rather than directly holding the securities in that index (Xtrackers glossary).

Instead of saying, “We will buy every company in this index”, a synthetic ETF says something closer to, “We will use a contract so the fund receives the performance of this index.”

That can sound more complicated, and it is. But the core concept is still manageable.

Swaps explained in plain English

Imagine an ETF wants to track Index A.

Rather than buying all the holdings in Index A, the ETF enters into a swap agreement with a counterparty. The counterparty agrees to pay the ETF the return of Index A.

In exchange, the ETF may pay the counterparty the return from another basket of assets, a fee or both, depending on the structure.

The exact mechanics vary by fund, so this is where the official documents matter. You do not need to become a derivatives expert, but you do need to know whether the ETF is physical or synthetic and what risks the provider discloses.

Why would an ETF use synthetic replication?

Synthetic replication can be used for several reasons:

• The index may be hard or expensive to replicate physically.
• Some markets may be difficult to access directly.
• The provider may be trying to reduce tracking difference.
• The structure may be useful for certain commodities, sectors or specialised exposures.
• The provider may believe a swap is more efficient than holding all the underlying assets.

That does not make synthetic ETFs better. It just explains why they exist.

For a beginner, the main point is this: a synthetic ETF is not fake exposure, but it is different exposure. It aims to give you the return of the index through a contract rather than direct ownership of every underlying asset.

Counterparty risk in synthetic ETFs

The biggest beginner concept with synthetic ETFs is counterparty risk.

A counterparty is the institution on the other side of the swap contract. If that institution cannot meet its obligations, the ETF may be affected.

That is counterparty risk.

Synthetic ETFs usually have safeguards such as collateral arrangements and counterparty exposure controls described in the fund documents. In the UK rules for UCITS schemes, the FCA’s COLL sourcebook says exposure to one OTC derivative counterparty must not exceed 5% of scheme property, rising to 10% where the counterparty is an approved bank (FCA Handbook COLL 5.2.11R). That regulatory limit is useful context, but it does not mean risk disappears. The details still depend on the exact ETF structure, provider and legal documents.

In beginner language, the question is not “Will the swap provider definitely fail?” The better question is:

If something goes wrong with the swap counterparty, what protection does the ETF have and where is that explained?

Look in the official documents for:

• Swap counterparty information
• Collateral policy
• Collateral quality
• Counterparty exposure
• Whether the swap is funded or unfunded
• Risk disclosures relating to derivatives

You do not need to memorise every legal term, but you should be able to find where the provider explains them.

Physical ETFs have risks too

It is easy to hear “physical” and assume it means simple, safe and risk-free. That is not right.

Physical ETFs can still have risks, including:

• Market risk: if the index falls, the ETF can fall.
• Tracking difference: the ETF may not match the index perfectly.
• Sampling risk: if it uses sampling, the chosen basket may behave differently from the full index.
• Liquidity risk: some underlying holdings may be harder to trade.
• Securities lending risk: some physical ETFs lend out holdings to earn extra revenue.
• Currency risk: if the holdings are in foreign currencies, exchange rates can affect returns.
• Concentration risk: some indices are heavily weighted to certain countries, sectors or companies.

The words “physical ETF” do not remove the need to understand what you own.

This matters to me because my own investing approach has become more focused on keeping things simple, especially with broad ETFs. But simple does not mean ignoring the details. Even a plain-looking global ETF has a structure, a benchmark, a currency, costs and risks.

Securities lending in physical ETFs

Some physical ETFs lend out some of their holdings to other market participants. This is called securities lending.

The ETF may receive a fee for doing this, which can help offset costs and potentially improve tracking. In return, the borrower provides collateral.

This is not unique to ETFs and it is not automatically a problem. But it is another risk area to understand.

If a physical ETF uses securities lending, the provider should explain its policy in the official documents. Vanguard’s overview lists securities lending as one of the due-diligence questions for physical ETFs, including borrower selection, collateral, revenue and maximum lending levels (Vanguard). Look for:

• Whether securities lending is used
• What proportion of the fund can be lent
• Who receives the lending revenue
• What collateral is accepted
• What happens if a borrower fails to return securities

Again, the aim is not to panic. The aim is to know where the moving parts are.

Tracking: the part beginners often miss

The whole point of most ETFs is to track an index. So the next question is obvious:

How closely does the ETF actually track it?

There are two useful terms here:

• Tracking difference
• Tracking error

They sound similar, but they are not the same.

Tracking difference

Tracking difference is the gap between the ETF’s return and the index return over a period of time.

For example, if an index returns 8% over a year and the ETF returns 7.7%, the tracking difference is minus 0.3 percentage points.

This gap can happen because of:

• Ongoing fund charges
• Trading costs
• Tax treatment of dividends
• Rebalancing
• Sampling
• Swap costs or benefits
• Securities lending revenue
• Cash drag
• Timing differences

The ongoing charge is important, but it is not the whole story. A lower-fee ETF can still track worse than expected, and a higher-fee ETF can sometimes track surprisingly well because of other factors.

That does not mean you should ignore fees. It means tracking is broader than the headline fee.

Tracking error

Tracking error is about how consistent the tracking gap is.

An ETF with low tracking error follows the index closely and steadily. An ETF with high tracking error moves around the index more unpredictably.

For beginners, I think of it like this:

• Tracking difference asks, “How far away was the ETF from the index?”
• Tracking error asks, “How bumpy was that difference?”

Both can matter, especially if you are comparing funds that track the same index.

Past tracking is not a promise

When you look at past tracking data, remember the caveat: past performance is not a reliable guide to future returns.

An ETF that tracked well in the past may not always track well in the future. Markets change, index rules change, fund size changes, costs change and replication methods can change.

That is one reason I prefer treating ETF research as a checklist rather than a one-line answer.

Physical vs synthetic ETFs: neutral comparison

Here is the balanced version.

Physical ETFs may appeal because

• They are easier to understand.
• They hold the underlying securities directly.
• Holdings are often easier to inspect.
• They can feel more intuitive for beginners.
• Full replication is especially straightforward when practical.

Physical ETFs may be less ideal when

• The index has thousands of holdings.
• Some holdings are hard or expensive to trade.
• Sampling leads to tracking differences.
• Securities lending adds another layer of risk.
• The fund cannot access the market efficiently.

Synthetic ETFs may appeal because

• They can track some indices efficiently.
• They can provide access to markets that are harder to replicate physically.
• They may reduce certain types of tracking difference in some cases.
• They can be useful for specialised exposures.

Synthetic ETFs may be less ideal when

• You are uncomfortable with swap structures.
• Counterparty risk is not clearly understood.
• The collateral policy is hard to assess.
• The ETF documents are too complex for your comfort level.
• You prefer direct holdings where possible.

The key point is that replication method is one factor, not the only factor.

I would not personally look at an ETF and stop at “physical” or “synthetic”. I would want to know what index it tracks, what it costs, how it has tracked historically, how big and liquid it is, whether it accumulates or distributes, what currency it trades in, what risks the provider lists and whether I actually understand it.

That still does not tell anyone else what to buy. It is just the kind of process I find useful.

Where to check whether an ETF is physical or synthetic

Do not rely only on a platform summary, a social media post or a random table online. ETF details can change, and third-party summaries can be incomplete.

The best place to verify the replication method is the provider’s official documentation.

One current UK wrinkle is disclosure wording. The FCA says the old PRIIPs regime has been replaced by the Consumer Composite Investments regime, with transitional provisions running from 6 April 2026 to 7 June 2027. During that period, manufacturers may be using a product summary, Key Information Document, Key Investor Information Document or EEA Key Investor Information Document depending on the product and regime (FCA PRIIPs page, FCA DISC TP 2). In practice, the beginner habit is the same: use the provider’s current official product page and current disclosure documents, not an old screenshot or platform summary.

Look at:

1. The ETF provider’s official fund page
2. The factsheet
3. The Key Information Document
4. The prospectus
5. The annual or semi-annual report, if you want more detail

You are looking for phrases such as:

• Physical replication
• Direct replication
• Full replication
• Optimised sampling
• Representative sampling
• Synthetic replication
• Swap-based replication
• Indirect replication
• Derivatives
• Counterparty risk
• Collateral

If the document says full replication, the ETF aims to hold all or substantially all of the index securities.

If it says sampling or optimised sampling, it is physical but does not necessarily hold every index security.

If it says synthetic, swap-based or indirect replication, it is using a swap or derivative structure to track the index.

A beginner checklist before comparing two ETFs

If I were comparing two ETFs that track the same or similar index, I would not make the replication method the only question. I would look at it as part of a wider checklist.

Here are the questions I would ask:

1. What index does it track?

Two ETFs can sound similar but track different indices. For example, one might track developed markets only while another includes emerging markets. One might track a broad market while another tracks a sector.

Before worrying about physical vs synthetic, I would first make sure I understand the actual benchmark.

2. Is it physical or synthetic?

Then I would check the replication method in the official documents.

If it is physical, I would check whether it uses full replication or sampling.

If it is synthetic, I would check the swap structure, counterparty risk explanation and collateral policy.

3. How closely has it tracked its index?

I would look for tracking difference and tracking error where available.

Past data is useful, but not a guarantee. The point is to understand whether the ETF has historically behaved as expected, while remembering that future results can differ.

4. What are the costs?

The ongoing charge is usually the easiest cost to find, but it is not the only cost that matters. Trading spreads, platform fees, currency conversion and tracking difference can also affect investor outcomes.

Do not assume the cheapest headline fee always means the best overall result.

5. Is the ETF understandable?

This is underrated.

If I cannot explain what an ETF does in a few sentences, I probably need to slow down. That does not mean the ETF is bad. It means I have not done enough work to understand it.

For me, keeping investing simple has been one of the biggest lessons. Automation and broad ETFs helped remove a lot of emotion from the process, but simplicity only works if I actually understand what is sitting underneath the automation.

Common beginner misunderstandings

”Physical means safe”

Physical does not mean safe. It means the ETF holds underlying assets directly.

A physical ETF tracking a volatile sector can still move sharply. A physical ETF tracking a concentrated index can still be risky. A physical ETF using sampling can still track imperfectly.

Market risk remains.

”Synthetic means dangerous”

Synthetic does not automatically mean dangerous. It means the ETF uses a swap-based structure.

That structure adds counterparty risk and can be harder to understand, but synthetic ETFs are a normal part of the ETF market. The question is whether the structure, risks and protections are clear enough for you to be comfortable.

”The ETF return should exactly match the index”

In reality, ETFs rarely match their index perfectly forever.

Charges, taxes, cash, timing, trading costs, lending revenue and swap terms can all create differences. The aim is close tracking, not magic.

”The platform tells me everything I need”

Platforms are useful, but I would always verify important details in the official fund documents. Platforms can summarise, but the provider documents are where the fund’s structure and risks should be properly explained.

If you use Trading 212, I have a separate topic area here: Trading 212. Just remember that the ETF provider’s documents matter regardless of the platform you use.

My own view as a beginner-focused investor

I am not trying to be clever with ETF structures. My own investing style is built around simple, long-term ETF investing, with automation doing a lot of the behavioural heavy lifting.

That does not mean every ETF I have ever used is perfect or that my approach is right for everyone. It means I have learned that complexity has a cost, even when that cost is just mental energy.

When I look at physical vs synthetic ETFs, I see it as a clarity question:

• Do I know how this ETF tracks its index?
• Do I understand the risks?
• Can I find the method in the official documents?
• Am I comfortable with the structure?
• Does it fit the role I want it to play?

If the answer is no, I would rather pause and read more than rush because an ETF looks popular.

Nothing here is financial advice or a recommendation. It is just the framework I would use to understand the difference.

For the wider investing basics, see Foundations and my general disclaimer.

Final thoughts

The physical vs synthetic ETF debate is not about one being universally good and the other being universally bad.

A physical ETF buys the underlying assets directly, either through full replication or sampling.

A synthetic ETF uses a swap contract to receive the return of an index.

Physical ETFs can be easier to understand, but they still carry market risk, tracking risk and sometimes securities lending risk. Synthetic ETFs can be efficient for certain exposures, but they add counterparty risk and can be more complex.

For beginners, the most useful habit is simple: check the replication method in the official fund documents before investing.

Look at the factsheet, Key Information Document, prospectus and provider page. Find the words physical, sampling, synthetic, swap, collateral and counterparty risk. If you cannot find them or do not understand them, slow down.

Investing does not need to be complicated, but it does need to be understood.