How to build an ETF portfolio

The short version

Start with a core that already covers a large, well-defined market. Add satellites only when they express a risk you actually want and that the core does not already provide. Count overlap by looking through holdings, not by counting tickers. Write a rebalancing rule before markets move the weights. Treat expense ratios as a durable cost, and treat past fund returns as history, not a forecast.

There is not a single mix that is right for every investor. A buy-and-hold investor in a tax-advantaged account and a frequent trader in a taxable account can reasonably prefer different wrappers that track similar indexes. This article is a construction process, not a recommendation to buy any fund.

What an ETF portfolio is — and what it is not

An exchange-traded fund is a pooled vehicle that trades on an exchange. The U.S. Securities and Exchange Commission describes ETFs as funds that issue shares investors can buy and sell throughout the trading day at market prices, which may differ from net asset value. That structure is why people use ETFs as building blocks: one ticker can stand in for hundreds of securities, with a published objective, a published index or strategy, and a published fee.

A portfolio of ETFs is still a portfolio. It has a stock-versus-bond-versus-cash split, sector weights, geographic weights, and a handful of issuers that often dominate cap-weighted indexes. Naming the funds does not finish the job. You still need a target mix you can defend, a way to see what you already own across accounts, and a rule for what you will do when the mix drifts.

It is also not a collection of last year's winners. Funds that track the same index are substitutes for one another on the exposure that matters. Funds that track different indexes can still share the same largest holdings. Building from performance tables is how people accidentally stack the same bet three times and call it diversification.

Core versus satellite

A practical ETF mix has a core and, optionally, satellites. The core is the holding you would keep if you could own only one or two funds. It should be broad, cheap to hold, and aligned with the job you wrote down. Satellites are smaller sleeves that express a specific view: a different geography, a different company-size segment, a sector, or a style. They are optional. Many durable portfolios are a core plus cash and a bond fund, with no satellites at all.

The core should do most of the work. If a satellite is half the portfolio, it is not a satellite; it is a second core, and you should treat the overlap and the risk as such. Size satellites so that a complete miss in that sleeve cannot wreck the plan. There is no universal percentage, but a sleeve you would not notice if it halved is usually too small to bother with, and a sleeve that would force you to sell the core if it failed is too large.

Overlap is the construction problem people skip

Ticker count is a poor proxy for diversification. Two ETFs with different names can share the same largest issuers because cap-weighted indexes concentrate in the largest companies. A technology satellite on top of an S&P 500 core often increases weight in names that already dominate the core. Adding the same company as an individual stock on top of both funds triples an exposure that looked like three lines on a statement.

Look through holdings before you add a fund. Compare the top ten names and the sector weights with what you already own, including funds in other accounts. Published holdings arrive on a lag, so the picture is directional rather than live, especially around index reconstitutions. Directional is still better than assuming different tickers mean different risks.

A simple overlap test: if you removed the new fund, would the portfolio's largest issuers and largest sector meaningfully change? If not, you are adding weight, not breadth. Adding weight can be intentional — some investors want more of a theme they already have — but it should be a choice you can explain.

• Compare top holdings of each fund, not just the strategy name on the fact sheet
• Check whether a "growth" or "innovation" fund leads with the same companies as the core
• Watch for the same issuer appearing through several sector or style funds
• Remember that cap-weighted indexes are concentrated by construction
• Look across every account; overlap is a household problem, not a single-broker problem

How many ETFs you actually need

You need enough funds to cover the risks in the job you wrote, and no more. One broad equity fund can be a complete equity sleeve. A second fund is justified when it adds a market the first fund does not cover — for example, a dedicated international fund next to a U.S. core, or a bond fund next to an equity core. A fifth U.S. large-cap fund is rarely a new market; it is usually overlap with extra paperwork.

Complexity has a cost even when expense ratios are low. More funds mean more distributions to track, more rebalancing decisions, and more chances to tinker. If you cannot explain why a fund is in the mix without looking it up, it is a candidate to merge into the core.

There is not a magic number. Investors who use a target-date fund already own a packaged mix; stacking several broad index ETFs on top of that package is a common way to duplicate U.S. large-cap exposure. Investors who prefer to assemble the mix themselves still do not need a fund for every headline.

Cost, structure, and who might care about which wrapper

Expense ratio is the fee the fund states for running the portfolio, expressed as an annual percentage of assets. It is not the only cost — bid-ask spreads, premium or discount to net asset value, and tracking difference also affect what you earn — but it is the cost you can compare from issuer documents without treating a yield table as a recommendation. Prefer structure and fees over long return tables. Any past performance you see on a product page is past, and it is not predictive.

Funds that track the same index can still differ in legal structure and in how expensive they are to hold. VOO and IVV are examples of S&P 500 exchange-traded funds with a 0.03% stated annual cost in the issuer materials cited below. SPY is an S&P 500 tracker organized as a unit investment trust, with a 0.0945% gross expense ratio on State Street's product page. QQQ's stated total expense ratio is 0.18% after Invesco's UIT-to-open-end structure change, which reduced the stated ratio from 0.20%. QQQM's stated total expense ratio is 0.15% for the same Nasdaq-100 index.

For a buy-and-hold core, the holding cost compounds for years, so the difference between 0.03% and 0.0945% is a real, if unspectacular, drag. For someone who trades frequently and cares about the tightness of the market in a particular ticker, liquidity and spreads can dominate a few basis points of expense ratio on a short holding period. That is a description of trade-offs, not a verdict that one fund is better.

Taxable versus tax-advantaged accounts, at a high level

Where you hold a fund can matter as much as which fund you hold. Tax-advantaged accounts (workplace plans and IRAs, in broad terms) shelter ordinary income and capital gains until withdrawal rules apply. Taxable accounts do not. That is not a StockLift product claim, and it is not a reason to pick a fund because a marketing page mentioned "tax efficiency."

At a high level, funds that distribute more taxable income, or that you expect to sell and replace often, tend to create more tax paperwork in a taxable account than a low-turnover broad index fund you rarely touch. Frequent trading in a taxable account can also turn a low expense ratio into a rounding error next to realized gains. The opposite is also true: a slightly higher expense ratio inside a tax-advantaged account may be a smaller issue than a short-term gain in a taxable account. Those are location questions for a tax professional when the dollars are large, not a ranking of tickers.

Do not treat this article as a location map. Account type, withholding, and distribution character depend on your facts. The construction point is simpler: decide the mix first, then decide which account should hold which sleeve so you are not rebalancing in the most expensive place by accident.

Rebalancing: pick the rule before you need it

Markets will move the mix. A core that starts at 70% of the portfolio will not stay at 70%. Rebalancing is the decision to restore the target you wrote down, which usually means trimming what has recently done well. FINRA describes rebalancing as bringing a portfolio back to a target allocation; the SEC's investor education on asset allocation treats it as a periodic discipline, not a market call.

Two common rules: a calendar (review quarterly or annually) and a band (act when a sleeve drifts by a set number of percentage points). A hybrid — check on a schedule, trade only if a band is breached — reduces both neglect and fidgeting. Choose the rule while the mix is still close to target. A rule invented after a large move is usually a story about the move.

Measure drift across every account you own, not inside one brokerage window. Directing new contributions toward the underweight sleeve is often a lower-friction way to rebalance than selling in a taxable account. StockLift does not execute transactions; any change happens at your own brokerage after you have decided what should change.

A worked example that is not a recommendation

Suppose the job is long-horizon U.S. equity growth with a smaller international sleeve and a bond sleeve for ballast, held mostly in a workplace plan and an IRA. One construction: a U.S. large-cap or total-market fund as the equity core, an international fund as a distinct market, and a broad bond fund. That is three jobs and three funds. A fourth fund is justified only if it covers a gap those three leave open.

Now suppose the same investor adds a Nasdaq-100 fund because it is familiar. The U.S. core already holds many of the same large companies. The new fund raises concentration in those names and raises the blended expense ratio if the satellite is QQQ at 0.18% or QQQM at 0.15% while the S&P 500 core examples sit at 0.03%. That can still be a deliberate tilt. It is not automatic diversification.

A different investor who trades the S&P 500 frequently might care more about the market in a particular ticker than about a few basis points of expense ratio, and might look at SPY's structure and liquidity rather than at VOO or IVV's lower stated cost. That preference does not make SPY a better core for a buy-and-hold household. It makes it a different tool.

Replace the tickers in this example with whatever you already own and run the same questions: what is the core, what does each extra fund add, and what would you sell first if you had to simplify tomorrow?

Common mistakes when assembling ETF portfolios

Collecting funds that tell the same story. A U.S. large-cap index, a "quality" large-cap fund, and a technology fund can be three labels on one concentrated bet.

Using a sector or thematic fund as a core. Cores should survive a full cycle in that theme. Satellites can fail without ending the plan.

Ignoring the funds inside a target-date or allocation product you already hold, then adding a "simple" S&P 500 ETF on top.

Comparing funds with long return tables and calling the winner the core. Past returns are not a forecast, and they often reflect the same index twice.

Rebalancing on headlines instead of on a written rule, or never rebalancing and calling the drift a strategy.

Treating expense ratio as the only cost while trading often enough that spreads and taxes dominate.

Building the mix in one account while a second account already holds the same index under a different ticker.

A practical checklist

Use this as a pass/fail list, not as a score. If you cannot complete a line, the mix is not finished.

• Write the job of the money: goal, horizon, and the decline you can tolerate
• Name the core fund and the market it is supposed to represent
• List every fund and stock you already own across accounts, including workplace plans
• Look through top holdings and sectors; drop or resize anything that only duplicates the core
• Assign each remaining fund a job: core, ballast, or satellite
• Record the stated expense ratio from the issuer, not from memory; fees change and should be re-checked
• Choose a rebalancing rule (calendar, band, or hybrid) and where you will measure drift
• Note which sleeves sit in taxable versus tax-advantaged accounts, and ask a tax professional before large taxable sales
• Write what would cause you to remove a satellite, so the next interesting fund has a hurdle

Where this leaves you

A durable ETF portfolio is mostly a core you understand, satellites you can justify, overlap you have measured, and a rebalancing rule you will actually follow. The funds in the fee table above are examples of how similar indexes can still differ in cost and structure. They are not a menu to complete.

If you already hold several funds, the next useful step is look-through: see what you own inside the wrappers before you add another one. StockLift's Learn guide on ETF and fund holdings walks through that overlap problem. You can also review linked holdings in StockLift's iOS app; StockLift analyzes portfolios and does not execute transactions.

StockLift's iOS app listing is https://apps.apple.com/us/app/stocklift/id1553250387. Use it to inspect the mix you already have, then make any change at your own brokerage.