Bottom line: A core ETF is not valuable because it "always goes up." Its value is that one trade can buy a diversified basket of companies with transparent rules, low costs, and automatic replacement of weaker constituents over time. Broad ETFs such as VOO and SPY are usually better suited as long-term core holdings than single stocks. Growth, technology, and semiconductor ETFs such as QQQ, VGT, XLK, SMH, and SOXX can offer more upside, but they also come with deeper drawdowns.
What These Eight ETFs Actually Own
An ETF is not just a ticker. It is a portfolio. The important questions are how broad the portfolio is, how concentrated the top holdings are, and whether the fee level is reasonable for long-term holding.
| Ticker | Role | Fee | Holdings | Top 10 Weight | Top Five Holdings |
|---|---|---|---|---|---|
| SPY | Tracks the S&P 500; best known for trading liquidity | 0.0945% | 504 | 37.26% | NVDA 7.84%; AAPL 6.83%; MSFT 4.56%; AMZN 3.71%; GOOGL 3.30% |
| VOO | Tracks the S&P 500; low-cost long-term core holding | 0.03% | 519 | 38.37% | NVDA 7.84%; AAPL 6.44%; MSFT 4.89%; AMZN 4.19%; GOOGL 3.62% |
| QQQ | Tracks the Nasdaq-100; large-cap growth and innovation tilt | 0.18% | 104 | 46.24% | NVDA 8.38%; AAPL 7.07%; MSFT 4.96%; MU 4.82%; AMZN 4.34% |
| QQQM | Nasdaq-100 exposure; lower-fee long-term version of QQQ | 0.15% | 106 | 45.79% | NVDA 8.22%; AAPL 7.23%; MSFT 4.98%; MU 4.68%; AMZN 4.31% |
| VGT | US information technology sector; high technology concentration | 0.09% | 322 | 59.76% | NVDA 18.60%; AAPL 14.82%; MSFT 10.02%; AVGO 4.60%; MU 2.62% |
| XLK | S&P 500 technology sector; narrower sector basket | 0.08% | 72 | 62.16% | NVDA 13.23%; AAPL 11.53%; MSFT 7.70%; MU 7.25%; AMD 5.15% |
| SMH | Semiconductor supply chain; very concentrated and high beta | 0.35% | 26 | 70.78% | NVDA 15.22%; TSM 9.72%; MU 7.60%; AMD 7.10%; INTC 6.88% |
| SOXX | Semiconductor ETF with a somewhat more diversified rule set | 0.34% | 30 | 62.00% | NVDA 10.84%; AVGO 9.70%; AMD 8.09%; QCOM 6.48%; TSM 6.21% |
Holdings change over time. The figures above are based on public disclosures around 2026-06-11 to 2026-06-12. SPY and XLK reference State Street disclosures; VOO and VGT reference Vanguard and Schwab; QQQ and QQQM reference Invesco and Schwab; SMH references VanEck and Schwab; SOXX references iShares and BlackRock.
Returns and Drawdowns: Do Not Only Look at the Winner
The table below uses Nasdaq historical closing prices to estimate price returns. It does not include dividend reinvestment, so it is best used to compare volatility and drawdown behavior rather than official total return.
| Ticker | Start | Through | Years | Annualized Price Return | Max Drawdown | Best Year | Worst Year |
|---|---|---|---|---|---|---|---|
| SPY | 2016-06-13 | 2026-06-12 | 10.0 | +13.54% | -34.10% | +28.79% | -19.95% |
| VOO | 2016-06-13 | 2026-06-12 | 10.0 | +13.57% | -34.30% | +28.77% | -20.01% |
| QQQ | 2016-06-13 | 2026-06-12 | 10.0 | +20.92% | -35.62% | +54.84% | -33.71% |
| QQQM | 2020-10-13 | 2026-06-12 | 5.7 | +17.18% | -35.57% | +54.98% | -33.71% |
| VGT | 2016-06-13 | 2026-06-12 | 10.0 | +24.07% | -35.46% | +52.99% | -30.86% |
| XLK | 2016-06-13 | 2026-06-12 | 10.0 | +23.88% | -34.02% | +56.16% | -29.10% |
| SMH | 2016-06-13 | 2026-06-12 | 10.0 | +36.21% | -45.30% | +73.67% | -35.70% |
| SOXX | 2016-06-13 | 2026-06-12 | 10.0 | +34.21% | -46.24% | +90.08% | -37.09% |
SMH, SOXX, VGT, and XLK look stronger over long periods because they take more concentrated sector risk. SOXX had a roughly -46.24% max drawdown over the sample, and SMH had a roughly -45.30% max drawdown. These are satellite holdings, not substitutes for broad-market core ETFs.
Aggressive References: DRAM and EWY
Positioning: DRAM and EWY are better treated as aggressive references or small satellite positions. They should not be confused with core holdings such as VOO or SPY. DRAM is a newly launched memory ETF with a very short live history. EWY is a single-country South Korea ETF with a much longer history, but it is heavily influenced by Samsung Electronics, SK Hynix, the Korean won, and Korea-specific market risk.
Performance and Volatility Profile
| Ticker | Start | Through | Price Performance | Max Drawdown | How to Read It |
|---|---|---|---|---|---|
| DRAM | 2026-04-02 | 2026-06-17 | About +152% since launch | About -20% | The live history is extremely short. The surge mainly reflects AI memory and HBM enthusiasm and should not be extrapolated into a long-term return assumption. |
| EWY | 2000-05-12 | 2026-06-17 | About +10.8% annualized price return | About -74% | EWY has shown strong long-term upside, but it has also suffered very deep drawdowns. Its 2025 and 2026 performance was helped by the repricing of Korean technology stocks. |
Methodology: Yahoo Finance daily adjusted close, approximate price return only. This is not the same as official fund total return. DRAM is too new for annualized figures to be meaningful.
Holdings: What They Are Really Betting On
| Ticker | Theme | Fee | Concentration | Main Holdings / Exposure |
|---|---|---|---|---|
| DRAM | Global memory chips, HBM, NAND, SSD, HDD | 0.65% | Highly concentrated | Roundhill disclosures list names such as Samsung, SK Hynix, Micron, Kioxia, and SanDisk among major exposures. Weight calculations may combine equity holdings and total return swaps, so the official Roundhill CSV or PDF should be the primary source. |
| EWY | South Korean equities, with technology and industrial tilt | 0.59% | Top-heavy | iShares discloses roughly 78 holdings, with information technology around 52.5%. Large weights usually include Samsung Electronics, SK Hynix, Hyundai Motor, KB Financial, and SK Square. Samsung and SK Hynix can together approach or exceed 40%. |
Why They Became Popular
DRAM: AI Memory Bottleneck
- HBM demand: AI GPU clusters need high-bandwidth memory, turning memory from a background component into a key AI infrastructure bottleneck.
- Higher purity: Compared with SMH or SOXX, DRAM is more focused on the memory supply chain rather than the whole semiconductor industry.
- Supply cycle: When DRAM and NAND pricing turns upward, memory producer earnings can expand sharply.
EWY: Korea Technology Repricing
- SK Hynix and Samsung: Korea is central to global memory and advanced manufacturing, so the AI memory cycle directly affects EWY.
- Single-country leverage: EWY packages Korean technology, industrials, financials, and currency exposure into one trade.
- Policy and valuation: Market reform, governance improvement, valuation rerating, and foreign inflows can amplify returns.
Main Risks
| Ticker | Main Risk | Meaning | Best Use |
|---|---|---|---|
| DRAM | New fund, narrow theme, derivatives exposure | DRAM only launched on 2026-04-02, so the sample is extremely short. Memory is cyclical, and reversals can be fast. Some exposure may be expressed through total return swaps, so third-party weight data can differ from the official source. | Useful for watching the AI memory theme or as a small satellite position. It should not replace SMH, SOXX, or a broad-market ETF. |
| EWY | Single-country, currency, and top-holding concentration | Korea is sensitive to exports, semiconductor prices, geopolitics, and the Korean won. When Samsung and SK Hynix dominate the fund, EWY behaves partly like a large memory-stock amplifier. | Useful for expressing a Korea, memory, or Asia technology cycle view, with more conservative sizing than broad-market ETFs. |
Where to check latest holdings: For DRAM, use Roundhill's official Holdings / Download CSV / Fact Sheet. For EWY, use the iShares official Holdings / Data Download page. For a fund like DRAM that may use swaps, official fund disclosures should take priority over brokerage or third-party websites.
DRAM holdings and official CSV guide · QQQM holdings and sector allocation
DRAM holdings and official CSV guide · QQQM holdings and sector allocation
How to Choose Similar ETFs
| Pair | Key Difference | Who It Fits | Simple Rule |
|---|---|---|---|
| SPY vs VOO | Both track the S&P 500. SPY has deeper trading liquidity and options markets; VOO has a lower fee and is usually better for long-term holding. | SPY fits traders and options users. VOO fits buy-and-hold investors. | Long-term: VOO. Trading tool: SPY. |
| QQQ vs QQQM | Both track the Nasdaq-100. QQQ is older, larger, and more liquid; QQQM is the lower-fee holding version. | QQQ fits trading and options. QQQM fits long-term allocation. | Long-term: QQQM. Trading: QQQ. |
| VGT vs XLK | Both are technology-heavy. VGT is broader; XLK is the S&P 500 technology sector and is usually more concentrated in mega-cap tech. | VGT fits broader technology exposure. XLK fits a narrower S&P 500 tech leader bet. | More diversified tech: VGT. More concentrated leaders: XLK. |
| SMH vs SOXX | Both own semiconductors. SMH is more concentrated; SOXX holds around 30 names and is somewhat more diversified by rule. | SMH fits investors seeking higher concentration. SOXX fits investors who want semiconductor exposure with slightly more diversification. | More upside beta: SMH. More balanced semiconductor basket: SOXX. |
Why Core ETFs Fit Most Investors Better Than Single Stocks
ETF Advantages
- Diversification: Buying SPY or VOO is not a bet on one company's earnings report. It is exposure to a broad set of large US companies.
- Transparent rules: Index rules add, remove, and rebalance companies over time.
- Low cost: Fees such as VOO at 0.03%, VGT at 0.09%, and XLK at 0.08% are low compared with many active funds.
- Less stock-picking pressure: You do not need to decide every day whether to own NVDA, AAPL, MSFT, AMZN, or another winner.
- Better suited for dollar-cost averaging: Volatility remains, but the portfolio rules are stable.
ETF Trade-Offs
- You cannot only own winners: ETFs include both companies you like and companies you would not choose individually.
- Market-cap weighting can chase price: As large companies rise, their weights rise too.
- No automatic market timing: Index funds usually stay invested through bear markets.
- Theme ETFs are volatile: Technology and semiconductor funds can perform well over time while still suffering painful drawdowns.
- Lower jackpot potential: ETFs sacrifice single-stock upside for a higher probability of survival.
Core and Satellite Portfolio Roles
| Role | ETF Examples | Reason | Risk Reminder |
|---|---|---|---|
| Core holding | VOO / SPY | Broad exposure to roughly 500 large US companies, diversified sectors, and low fees. | Still exposed to market-wide bear markets; not a replacement for cash or bonds. |
| Growth enhancer | QQQ / QQQM | Large non-financial Nasdaq companies with stronger technology and innovation exposure. | More sensitive to valuation and interest rates; drawdowns can be deeper than the S&P 500. |
| Technology sector tilt | VGT / XLK | Direct exposure to US technology companies and AI, software, and semiconductor cycles. | Top holdings can approach or exceed 60%, so this is not a broad-market substitute. |
| High-beta theme | SMH / SOXX | Semiconductor supply-chain exposure with strong sensitivity to AI capital spending. | When the theme reverses, drawdowns can be severe. |
Portfolio framing: Many long-term investors can start with VOO or SPY as a 60%-80% core, then use QQQM or QQQ for 10%-25% growth exposure. VGT, XLK, SMH, and SOXX are better treated as 0%-15% satellite positions. Semiconductor ETFs should not replace broad-market ETFs as the core because their industry cycle and valuation volatility are much more intense.
Data Sources and Verification Links
- Price history: Nasdaq historical closing price API, through 2026-06-12. Returns are price returns and exclude dividend reinvestment.
- SPY and XLK: State Street ETF pages for fees, holdings, and sector exposure.
- VOO and VGT: Vanguard and Schwab disclosures.
- QQQ and QQQM: Invesco and Schwab disclosures.
- SMH: VanEck website, Fact Sheet, and Schwab holdings data.
- SOXX: iShares / BlackRock website and holdings data.
- DRAM: Roundhill Memory ETF official page, Holdings / Download CSV / Fact Sheet.
- EWY: iShares MSCI South Korea ETF official page, Holdings / Data Download.