How an ETF works internally: replication, TER, and liquidity

If you already have one or more ETFs in your portfolio, you probably know their benchmark index, their category, and perhaps even their historical performance. What is rarely explained in the same detail is the internal mechanism that makes all of that possible. How does the fund replicate the index, how much does it actually cost you to maintain it each year, and what happens the exact moment you click "buy" or "sell"?They are three distinct pieces, but very interconnected, and all three determine what you ultimately take home.

In this article we review these three elements at once: the replication method and the counterparty risk it entails, the TER and its cumulative impact on profitability, and the liquidity, spread and the market hours that determine the exact price of each transaction. By the end, you'll be able to look at any ETF's information sheet with a much more comprehensive perspective than simply "which index it tracks and how much it's risen this year."

Physical vs. synthetic replication: how an ETF tracks its index and what counterparty risk you assume

An ETF promises something simple to state but more complex to execute: to track the performance of an index like the S&P 500, the MSCI World, or the Ibex 35. The method the manager chooses to achieve this has real consequences for you, because it affects the risk you assume, the costs you pay, and how closely the fund actually tracks its benchmark index. There are two main types of replication: physical replication, in which the fund directly buys the securities that make up the index, and synthetic replication, in which the ETF uses a derivative called a swap to replicate that performance without owning the underlying assets.

Full physical replica: the ETF that owns exactly what it says

In full physical replication, the ETF buys each of the stocks that make up the index, in the same proportion as their weight within it. If you invest in an MSCI World ETF that uses this method, the fund owns actual shares of all 1.400+ companies that comprise the index, with no shortcuts.

This method offers maximum transparency because you know exactly which assets the fund holds at any given time, and there is no counterparty risk at the source, since the ETF directly owns the assets. It is also easy to audit, even without advanced financial training. The drawback arises with very broad indices or those with illiquid securities, such as emerging markets or small-cap companies: buying and rebalancing each component has a high operating cost, which can result in a slightly larger tracking error.

Sampling replication: when the ETF doesn't buy the entire index

When an index has several hundred or thousands of components, replicating it completely becomes impractical. Therefore, some physical ETFs opt for sampling: they buy a representative selection of the most heavily weighted and liquid securities, capable of statistically reproducing the index's performance without acquiring each of its components. It is still physical replication in the strictest sense, because the fund holds real shares, just not of all the securities.

The implication for you is that the tracking error may be somewhat higher than with full replication, especially during periods of high volatility or when the excluded securities behave differently from the rest of the index. To identify this method, review the Key Investment Document (KID) or the ETF prospectus: the investment policy section usually mentions "sampling" or "optimized sampling."

Synthetic replication: how a swap replicates the behavior of an index

Synthetic replication works completely differently. Instead of buying the index assets, the ETF enters into a total return swap agreement with a counterparty, usually an investment bank. The mechanism has four steps: first, the ETF signs the swap with the counterparty; second, the counterparty commits to delivering to the fund the exact return of the index, including dividends, according to the swap type; third, the ETF pays a periodic fee in return, or delivers the return of a basket of collateral assets that may be completely different from the replicated index; and fourth, the ETF's net asset value (NAV) moves as if the fund had bought the index, even though it technically does not own those assets.

An example clarifies this. Imagine a synthetic ETF tracking the S&P 500 that holds as collateral a basket of European bonds with no direct link to the US companies in the index: thanks to the swap, you still receive the S&P 500's return, while the collateral serves as a guarantee for the contract. This method can be particularly efficient for indices that are difficult to access directly, such as emerging markets with restrictions on foreign investment or commodities, and also when the taxation of dividends from the physical replication is unfavorable for the fund. It's not a worse approach: it's a different tool, with its own balance of advantages and risks.

Counterparty risk: what happens if the swap bank fails and what limits are imposed by UCITS regulations

This is probably the most relevant section from a risk perspective, and the one least explained in the product information sheets. If the entity that signs the swap defaults on its obligations, for example due to bankruptcy or non-payment, the ETF might not receive the promised return from the index, and in that scenario, the fund only has the collateral it holds.

European UCITS regulations directly limit this risk. ETFs domiciled in Europe under UCITS regulations are required to ensure that their net swap exposure does not exceed 10% of the fund's net asset value (NAV), which in practice means that collateral must cover at least 90% of the ETF's value at all times. This is a specific regulatory limit, established in the UCITS IV Directive, not a best practice recommendation. Additional mechanisms are also in place: overcollateralization, with many issuers requiring collateral exceeding 100% of the swap value; the use of high-quality collateral, such as European government bonds or shares of major indices; and daily collateral replenishment, which adjusts the portfolio daily even if the market moves sharply.

These mechanisms significantly reduce risk, but do not eliminate it entirely. To verify this yourself, look in the KID for the "Risk and Remuneration Profile" section and, in the full prospectus, the section describing the replication method, where you will find the swap exposure percentage and, in many cases, details of the collateral basket.

The hidden risk of physical ETFs: securities lending

It's tempting to think that "physical ETFs have no counterparty risk," but that statement needs clarification. Many physical ETFs engage in securities lending: they temporarily lend the shares they hold to third parties, usually to cover short positions, in exchange for a fee that reduces the fund's overall cost. This lending introduces its own counterparty risk, because if the borrower fails to return the securities, the ETF could suffer a loss.

Issuers mitigate this risk by requiring collateral, almost always exceeding 100% of the borrowed amount, and reinvesting it in instruments considered safe. Under normal market conditions, securities lending reduces the effective cost of the ETF and benefits investors, but the risk exists and it's important to be aware of it. The fund's annual report or the issuer's fact sheet—Vanguard, iShares, Xtrackers, among others—publishes securities lending income and the required collateral levels: a less visible figure than the TER, but just as important.

How to identify the type of replication before investing

With the context now clear, these are the steps to identify the replication method of any ETF before buying it:

1. On JustETF or Morningstar: search for the ETF and locate the "Replication Method" field in its information sheet.
2. In the KID, the key information document: the section “What is this product?” includes the description of the investment method.
3. In the full prospectus: check “Investment Policy” or “Risk Management” for details of the swap and collateral, if the ETF is synthetic.
4. Pay attention to the warning signs: if the fund's asset basket is not related to the index it advertises, you are dealing with a synthetic ETF, and it is advisable to check the percentage of exposure to the swap and the quality of the collateral.

In the portfolios available through Bit2Me Invest, the InbestMe team has already selected ETFs based on efficiency, cost, and risk management criteria, including the type of replication for each fund. This doesn't mean you should stop researching on your own; it simply means that when you do, you'll be starting with a selection already filtered according to those criteria.

TER in ETFs: what it is and how it impacts your profitability

With the replication method now clear, it's time to talk about the second factor that determines what you take home: the cost of holding the ETF. And this is where the TER comes in.

What is TER and what costs does it include (and which ones it doesn't)?

The TER, short for Total Expense Ratio, is the annual percentage of an ETF's assets that goes toward covering all its management and administrative costs: what it costs you to hold that ETF for a year. It's not a commission you pay when buying or selling, but rather a cost that runs in the background while the fund exists, and it's charged implicitly: the manager deducts it daily from the fund's net asset value, without you seeing any charge on your account or a line item on your statement.

Imagine an ETF with a TER of 0,20%. For every €100 you have invested in it, you're automatically paying €0,20 per year in costs. If you have €10.000 in that ETF, the annual cost is around €20, spread out in small amounts over the 365 days.

The TER (Total Expense Ratio) combines several components into a single percentage. It includes the management fee (the largest part), the deposit and custody fee, and the fund's audit and administration costs. It does not include the trading commissions charged by your broker when buying or selling the ETF, the bid-ask spread—which we'll discuss in the next section—or any taxes you may have to declare. This distinction is especially important if you make regular contributions using Dollar Cost Averaging (DCA): each recurring purchase may involve a broker fee that isn't included in the TER, so it's advisable to add this when calculating the true cost of your strategy.

The cumulative impact of the TER over 20 years

A TER of 0,07% or 1,50% might seem like a negligible difference on a product sheet, but over a twenty-year period, it becomes one of the most significant factors in your final return. Imagine investing €10.000 once and holding it for 20 years, assuming a hypothetical gross return of 7% per year, with no guarantee that this will be achieved. With an indexed ETF with a TER of 0,07%, the approximate net return would be 6,93% per year; with an actively managed fund with a TER of 1,50%, it would drop to 5,50% per year.

The difference between the two scenarios is around €9.000, solely due to the effect of the TER on the same initial investment and the same assumed gross return. This money doesn't disappear because of a bad investment decision; it's simply absorbed year after year by a slightly higher cost.

Liquidity, spread and timing: how your order is actually executed

You already know how your ETF tracks the index and how much it costs to maintain it each year. That leaves the third piece, the most practical of the three: what exactly happens the moment you decide to buy or sell.

When you buy a traditional mutual fund, there's only one snapshot of the day: a net asset value is calculated at the close of trading, and all subscriptions and redemptions are executed at that same price, whether you placed the order at nine in the morning or five in the afternoon. An ETF doesn't work that way: it trades on the stock exchange throughout the trading session, like a stock, and its price constantly changes according to supply and demand at that moment. This has an obvious advantage—being able to buy or sell at any time during the session without waiting for the close—but it also introduces a variable that a traditional fund doesn't have: the exact price at which your order is executed depends on the time of day, how many people are trading that ETF, and how liquid the market is at that precise second.

The two levels of liquidity in an ETF: secondary market and underlying assets

Many investors assume that the liquidity of an ETF is measured the same way as that of a stock: if few shares are traded daily, the ETF appears "illiquid." This is a common misconception, and understanding it correctly will save you from dismissing products that are actually perfectly viable.

There are two distinct levels of liquidity. The first is secondary market liquidity: the volume of ETF shares that change hands daily on the exchange where it is listed. This is the most visible level and has the greatest impact on the spread you pay when trading. The second level, less well-known but more crucial, is the liquidity of the underlying assets: the ease with which market makers can create or redeem shares by buying or selling the assets that make up the replicated index. This creation and redemption mechanism is precisely what connects the ETF's stock market price to the real value of the assets it claims to replicate, regardless of its replication method.

If the underlying index is composed of highly liquid stocks, such as those in the S&P 500, market makers can intervene efficiently and maintain a tight spread even if the ETF's daily trading volume is modest. An S&P 500 ETF with €100 million in assets and low daily volume might be perfectly liquid for an individual investor, whereas an ETF tracking a high-yield corporate bond index in emerging markets might present more challenges because its underlying assets are inherently less liquid. In short, the practical criterion for assessing an ETF's true liquidity is not just about daily trading volume: it's also important to consider assets under management (AUM) and the liquidity of the underlying index.

Bid, ask and spread: the cost that doesn't appear in the TER

Every ETF trades with two simultaneous prices, not one. The bid is the highest price at which someone is willing to buy the ETF at that moment—in other words, what you receive if you sell. The ask is the lowest price at which someone is willing to sell—what you pay if you buy. The difference between the two is the spread, and it's the implicit cost of each buy or sell transaction, the cost that isn't reflected in any prospectus or the fund's TER (Total Expense Ratio).

A numerical example makes it clearer. Imagine an ETF with a bid of €100,00 and an ask of €100,10: the spread is €0,10 per share, 0,10% of the price. If you buy 100 shares at €100,10 and immediately sell them at the bid of €100,00, you lose €10 just by crossing the spread, without the ETF price having moved a single cent in the market itself.

It's important to distinguish between the spread and the TER (Total Expense Ratio), as they are two different types of costs. The TER is deducted daily and progressively from the fund's assets, regardless of how often you trade. The spread, on the other hand, is paid every time you buy or sell, and its relative impact depends on your trading frequency: for a long-term investor who contributes once a month, a 0,10% spread is a negligible fraction of the annual TER; for someone who trades several times a day, that same spread becomes the largest transaction cost in their entire trading activity.

What determines the size of an ETF's spread

Not all ETFs have the same spread, and knowing what determines it helps you anticipate which products will cost you more to trade. The first factor is the liquidity of the underlying assets: ETFs tracking shares of large companies in developed markets typically trade with tight spreads, while ETFs tracking small caps, emerging markets, corporate bonds, or commodities tend to have much wider spreads.

The second factor is the size of the ETF, measured by its assets under management (AUM): larger funds tend to attract more market makers, and this competition among them tends to narrow the spread. The third factor is market volatility, since during macroeconomic data releases, crises, or at the opening of trading, market makers widen the spread to compensate for the increased risk of hedging their positions. The fourth factor, and probably the least intuitive, is the overlap of trading hours: when an ETF's benchmark market is closed, market makers have more difficulty hedging their positions, so they widen the spread to protect themselves, something we will discuss in the next section. Finally, the number of active market makers in a particular ETF also matters: in niche products, there may be only one, which reduces competition and widens the spread.

As a general guideline, ETFs tracking major developed-income equity indices tend to trade within the narrowest range in the market, while thematic or niche ETFs can have a much higher cost. For specific and up-to-date figures on a particular ETF, the issuer's fact sheet and comparison sites like JustETF are the most reliable sources when trading.

The order book: how to read market depth before trading

The order book is a list of all pending buy and sell orders for an ETF at any given time, sorted by price and quantity. You'll find it on any broker platform as a table with two columns—bid and ask—each with various price levels and the available volume at each.

For an ETF investor, the order book matters for two reasons. The first is depth: if there are only 500 shares available at the nearest price level and you want to buy 2.000, the excess will be executed at progressively worse prices, a phenomenon known as slippage. The second is that it acts as a real-time liquidity signal: an order book with many price levels and high volumes indicates good liquidity at that moment, while one with few levels and low volumes warns that a large order could move the price on its own.

The practical recommendation is simple. For large orders, above €5.000-€10.000, it's advisable to consult the order book before placing a trade or use a limit order to control the final price. For most individual investors trading between €500 and €2.000 in major index ETFs, the order book rarely poses a real problem, although it's always worth checking it before placing a substantial order.

Market hours: Europe, America and overlap with Wall Street

Market hours largely determine what you can expect from the spread at any given time of day. In Spain, ETFs are primarily traded on BME, Xetra (Frankfurt), and Euronext (Paris, Amsterdam, Brussels), and all three exchanges share trading hours from 09:00 to 17:30 CET. The London Stock Exchange (LSE) operates from 08:00 to 16:30 GMT, which during summer time is equivalent to 09:00 to 17:30 CET.

The most interesting case is that of ETFs that track American indices such as the S&P 500, the Nasdaq 100, or the Russell 2000. These products are traded on European stock exchanges throughout the session, but their benchmark markets—the NYSE and the Nasdaq—don't open until 15:30 p.m. CEST in the summer (16:30 p.m. in the winter). This time difference explains why the spread of these ETFs can vary so much throughout the day: while Wall Street remains closed, European market makers have more difficulty hedging their positions against the underlying assets, so they widen the spread as a protective measure.

So, what's the best time to buy an ETF if yours tracks a US index? It depends primarily on the amount invested and the time horizon of your trade. If you're investing long-term with small orders, under €1.000, the spread's impact is marginal regardless of the time of day, and it's more important to prioritize consistency—for example, a fixed day of the month—than chasing the exact minute. If you're going to execute orders of a significant amount, above €5.000, on US index ETFs, it might make sense to wait until Wall Street opens to trade under more liquid conditions. In any case, it's advisable to avoid the first and last 15-30 minutes of the trading session for niche ETFs, because the opening and closing volatility tends to temporarily widen the spread.

Market orders vs. limit orders: which to use in ETFs

Besides liquidity and timing, the type of order you choose also affects the final price you pay. A market order is executed immediately at the best available price in the order book: its advantage is speed, but its drawback is that you don't control the exact price, and in an ETF with a wide spread or a shallow market at that moment, you could end up paying significantly more—or receiving significantly less—than you expected when placing the order.

A limit order works differently: you specify the maximum price you're willing to pay if you buy, or the minimum price you're willing to accept if you sell, and the order is only executed if the market reaches that level. Its advantage is complete control over the execution price; its disadvantage is that it may not be executed if the market doesn't reach your limit during the validity period you've set. As a general rule, for highly liquid index ETFs—such as the S&P 500, the MSCI World, or the Ibex 35—with very tight spreads, a market order during the regular trading session is perfectly reasonable for an individual investor. For niche ETFs, emerging market ETFs, or any trade outside of Wall Street's trading hours, it's always wiser to use a limit order, precisely to avoid price surprises during periods of wider spreads. For illustrative purposes, this comparison does not constitute investment advice on any specific ETF: each decision should be evaluated based on your individual circumstances and objectives.

The three pieces together: what really determines what you get out of an ETF

Physical replication offers greater transparency because the fund directly owns what it claims to own; synthetic replication can be more efficient in certain markets, but introduces counterparty risk that UCITS regulations limit, without eliminating it entirely. The TER (Total Expense Ratio), meanwhile, is the silent cost that weighs most heavily in the long run, while liquidity, the spread, and trading hours determine how much you pay more—or less—at the specific moment of each transaction. No ETF is completely risk-free, regardless of its replication method, cost, or liquidity level, and understanding these three elements is probably what gives you the most confidence before adding an ETF to your portfolio or reviewing the ones you already hold.

If you prefer that these criteria—replication method, TER, and execution—are already factory-checked, in Bit2Me Invest lets you explore ETF portfolios selected by InbestMe and see which one fits your investment horizon.

Frequently asked questions about the internal mechanics of an ETF