Types of ETFs — Index, Sector, Thematic, Bond, Commodity, Leveraged

Index ETFs — the most important financial innovation of the 20th century

John Bogle launched the first index mutual fund in 1976, and the concept was initially ridiculed as "Bogle's folly." The idea was radical: instead of paying professional fund managers to pick stocks, simply buy every stock in the index in proportion to its market cap, charge next to nothing, and let the market do the work. Today, Vanguard — the company Bogle built around this idea — manages over $8 trillion in assets, and index ETFs represent the most evidence-backed approach to long-term wealth accumulation.

An index ETF tracks a pre-defined, rules-based list of securities. The S&P 500 index selects the 500 largest US companies by market capitalisation, subject to profitability screens and a committee review. An S&P 500 ETF buys all 500 in exact proportion. When Apple is 7% of the index by market cap, Apple is 7% of the ETF. When Apple grows and becomes 8% of the index, the ETF automatically rebalances to match. There is no human judgement involved — no portfolio manager deciding to buy more Apple or reduce Exxon. The rules decide, automatically, every time the index reconstitutes.

The most important consequence of this structure is cost. Without a research team, traders, or analysts to pay, index ETF expense ratios are extraordinarily low. Vanguard's S&P 500 ETF (VOO) charges 0.03% per year — $3 per year on a $10,000 investment. The comparable actively managed large-cap fund averages 0.8–1.2% per year — $80–$120 on the same investment. That difference, compounded over 30 years at 8% annual returns, is the difference between $1,006,000 and $761,000 on a $100,000 starting investment. The fee gap alone accounts for $245,000 of missing wealth.

The main index categories every investor should know

Index ETFs exist for virtually every market and asset class, but a handful cover what most investors actually need. US Total Market ETFs (VTI, ITOT) hold approximately 4,000 companies — every publicly listed US business, from mega-cap to micro-cap. S&P 500 ETFs (SPY, VOO, IVV) hold the largest 500. The practical difference in long-run performance between VTI and VOO is historically tiny — roughly 0.1–0.2% per year in favour of VTI in most periods — because the S&P 500 dominates the total market's returns anyway. International developed market ETFs (EFA, VEA) and emerging market ETFs (EEM, VWO) extend coverage globally. A single all-world ETF (VT, ACWI) combines all of this in one fund, holding approximately 9,000 companies across 50+ countries.

Bond ETFs — the portfolio stabiliser most beginners underestimate

Bond ETFs hold fixed-income securities — loans made to governments or corporations that pay regular interest (the coupon) and return the principal at maturity. Bond prices move inversely to interest rates: when rates rise, existing bonds become less attractive compared to new bonds paying higher coupons, so their price falls. When rates fall, existing bonds paying higher coupons become more valuable. This is why 2022 was historically unusual — both stocks and bonds fell simultaneously as rates rose faster than at any point since 1980.

The critical portfolio construction benefit of bond ETFs is their low correlation to equities in normal market conditions. During the 2008 financial crisis, VTI (US equities) fell approximately 50% peak to trough. BND (US total bond market) rose roughly 5%. During the 2020 COVID crash, VTI fell 30% in five weeks; BND fell only 5% and then recovered to new highs within months. This ballast effect means a portfolio containing even 20–30% bonds experiences dramatically lower drawdowns than a 100% equity portfolio — which matters enormously for investors who might otherwise panic-sell during crashes.

Bond ETFs subdivide by maturity and credit quality. Government bond ETFs (TLT for long-term US Treasuries, SHY for short-term, IEF for intermediate) carry essentially zero default risk — the US government has never defaulted on domestic obligations. Corporate investment-grade bond ETFs (LQD, VCIT) hold bonds from creditworthy companies — slightly more yield, slightly more default risk. High-yield bond ETFs (HYG, JNK) hold bonds from lower-rated companies, sometimes called "junk bonds" — meaningfully more yield, meaningfully more default risk, and significantly higher correlation to equities during stress periods. High-yield ETFs should not be treated as a portfolio stabiliser the way government bond ETFs are.

Sector ETFs — concentration is a double-edged sword

Sector ETFs slice the equity market into its 11 industry classifications defined by GICS (Global Industry Classification Standard): Technology, Healthcare, Financials, Consumer Discretionary, Consumer Staples, Energy, Utilities, Real Estate, Industrials, Materials, and Communication Services. SPDR's Select Sector ETFs (the XL-series) are the most popular implementation: XLK for Technology, XLF for Financials, XLE for Energy, XLV for Healthcare.

The appeal is clear: if you have conviction that healthcare will outperform over the next decade due to aging demographics, XLV lets you express that view across 60+ healthcare companies without picking individual stocks. If you believe rising commodity prices will boost energy firms, XLE gives you diversified exposure without the single-company risk of owning Exxon alone.

The risk is equally clear: sector ETFs remove cross-sector diversification. In 2022, XLK (Technology) fell 33% while XLE (Energy) rose 65% — a 98 percentage point divergence within the same calendar year. An investor holding only XLK believed they were diversified across 60+ companies, but they were concentrated in a single economic sensitivity: high-multiple growth stocks that are disproportionately hurt by rising interest rates. Sector ETFs amplify a specific economic bet; they do not reduce risk relative to individual stock picking in the traditional sense.

Thematic ETFs — ideas that cross industry lines

Thematic ETFs invest around a concept rather than an industry classification. An artificial intelligence ETF might hold semiconductor companies (Nvidia — traditional tech), cloud infrastructure providers (Amazon Web Services), enterprise software firms (Microsoft Azure), and even healthcare diagnostics companies applying AI to imaging. These companies sit in four different GICS sectors but share exposure to a single technological trend.

Thematic ETFs became enormously popular after 2015 as ETF providers recognised that retail investors want to express views on trends rather than sectors. Clean energy, genomics, cannabis, the metaverse, cybersecurity, water scarcity — every major trend has spawned multiple competing ETFs. ARKK (ARK Innovation) became the poster child of thematic ETF investing: from 2017 to 2020, it returned over 300% total. Its $50 billion peak AUM in early 2021 made it briefly the world's largest actively managed ETF. Then rates rose. By late 2022, ARKK had fallen 75% from its peak — a larger decline than the Nasdaq during the dot-com bust. Investors who bought at peak AUM in 2021 had lost most of their investment within two years.

The structural problem with thematic ETFs is timing. Fund companies launch thematic products when retail interest peaks — which coincides exactly with peak valuations. The theme itself may be legitimate and important: AI is genuinely transforming industries, clean energy genuinely is expanding. But being invested in the right theme at a 100× earnings multiple produces poor returns even if the theme plays out perfectly over 20 years. The entry price determines the outcome, and thematic ETFs are typically launched after the easy gains have already been made.

Commodity ETFs — real assets and their hidden costs

Commodity ETFs provide exposure to raw materials — gold, silver, crude oil, natural gas, agricultural products, or broad commodity baskets. They serve as inflation hedges (hard assets tend to hold purchasing power when paper money depreciates) and portfolio diversifiers (commodity returns have historically had low correlation to equity returns).

The structure of the ETF determines its tracking quality. Gold ETFs like GLD (SPDR Gold Shares) hold actual gold bars in allocated storage at HSBC London. Each share represents approximately 0.093 ounces of gold. The fee is 0.40% per year — primarily storage and insurance costs. GLD has tracked the gold spot price with less than 0.5% annual error since its 2004 launch. Physical commodity ETFs are straightforward to understand and own.

Futures-based commodity ETFs are fundamentally different. USO (United States Oil Fund) does not own barrels of crude oil — it holds crude oil futures contracts. Every month, as contracts approach expiry, the fund must sell the expiring contracts and buy the next month's contracts. When the futures market is in contango — the normal condition where far-dated futures trade at a premium to near-term futures — this monthly roll incurs a systematic cost. From 2008 to 2018, crude oil fell approximately 26%. USO fell approximately 77% over the same period. The 51-percentage-point gap is almost entirely attributable to contango drag across a decade of monthly rolls. This is why commodity futures ETFs, unlike physical gold ETFs, are generally unsuitable for long-term buy-and-hold investors.

Leveraged ETFs — why the math destroys buy-and-hold investors

Leveraged ETFs use financial derivatives — primarily swap agreements and futures — to deliver a multiple of the underlying index's daily return. TQQQ (ProShares UltraPro QQQ) targets 3× the daily return of the Nasdaq 100. If the Nasdaq rises 1% on a given day, TQQQ should rise approximately 3%. If the Nasdaq falls 2%, TQQQ should fall approximately 6%. For a single trading day, this behaves exactly as advertised.

The critical word is daily. Leveraged ETFs reset their exposure at the end of every trading day. This daily reset is what causes volatility decay — also called beta slippage — a mathematical phenomenon that erodes returns in any period where the underlying index is not trending strongly in one direction. The arithmetic is unintuitive: after a 10% gain and a 9.09% loss, the index returns to exactly its starting value. But the 3× leveraged version gains 30%, then loses 27.27%: $100 × 1.30 × 0.7273 = $94.55 — a 5.45% loss while the index went nowhere. The faster the index whipsaws, the faster this decay accumulates.

The real-world consequences are severe. TQQQ rose approximately 290% from 2019 to the peak of late 2021 as the Nasdaq trended strongly upward. Then in 2022, as the Nasdaq fell 33%, TQQQ fell 79% — more than twice the index decline — because volatility decay amplified the drawdown beyond 3× the linear calculation. An investor who put $100,000 into TQQQ at the peak and held through 2022 watched it fall to roughly $21,000. Recovering from a 79% loss requires a 376% gain just to break even. This is not a cautionary tale about bad timing — it is the mathematical property of daily-reset leverage applied to volatile assets over multi-month periods.

Inverse ETFs — the same decay problem in reverse

Inverse ETFs deliver the opposite of the index's daily return. SH targets −1× the S&P 500 daily return; SQQQ targets −3× the Nasdaq 100 daily return. They suffer identical volatility decay. An investor who holds SH for six months in a market that rises 20% does not simply lose 20% — they lose more, because every day the underlying index gains, their loss that day is locked in, and the next day starts from a lower base. In addition, inverse ETFs incur swap costs and borrowing fees that erode returns further. They are designed for professional intraday hedging, not multi-day speculation.

Synthetic ETFs — counterparty risk hiding in plain sight

A synthetic ETF does not hold the actual underlying securities. Instead, it enters a swap agreement with a counterparty bank — typically a major investment bank like Goldman Sachs, BNP Paribas, or Societe Generale. The bank promises to pay the ETF the return of the index; the ETF pays the bank a small fee. Internally, the ETF might hold entirely different assets as collateral — whatever the swap counterparty agrees to post.

Synthetic ETFs were developed primarily in Europe, where regulatory structures and tax treaties sometimes make it inefficient to physically hold all the underlying securities. They can track illiquid or exotic indices more precisely than physical ETFs and are sometimes slightly more tax-efficient in specific jurisdictions. However, they introduce a risk that physical ETFs do not have: counterparty risk. If the swap counterparty bank fails to deliver on its contractual obligation — as nearly occurred with Lehman Brothers in 2008 — the ETF may not receive the promised return. EU UCITS regulations cap swap counterparty exposure at 10% of fund assets, which limits but does not eliminate this risk.

For the vast majority of broad market indices — S&P 500, MSCI World, global bonds — physical replication is perfectly feasible and transparent. There is no reason for a retail investor to accept counterparty risk in exchange for marginal benefits they will not notice. When choosing between a physical and synthetic ETF tracking the same index, prefer physical unless the synthetic offers a compelling, quantifiable advantage.