07 · US and Global ETF Portfolio Construction
For ordinary people allocating globally, US ETFs are the lowest-cost, most disciplined vehicle: broad-base fees down to 0.03%, buyable from 1 share, a whole basket of global assets in one go. This article is a "copy-friendly" practical checklist: how to pick broad-base ETFs, how much style and sector to add, how to top up globally and with bonds and commodities, the US versions of the three-fund and permanent portfolios with ratios, and finally DCA practice, the math of DCA vs lump sum, and a portfolio health checklist. Core principle: hold long term at the lowest fees, the deepest liquidity, and the least hassle — do not churn.
1. The Broad-Base ETF List
S&P 500: SPY / VOO / IVV
| Ticker | Issuer | Fee (check latest data) | Size | Notes |
|---|---|---|---|---|
| SPY | State Street | ~0.09% | Largest | King of liquidity; the fullest options chain; for high-frequency trading and covered calls |
| VOO | Vanguard | ~0.03% | Large | Cheapest tier; one of the top picks for long-term DCA |
| IVV | iShares | ~0.03% | Large | Low fee, tight tracking |
All three track the same index (the S&P 500); return differences come down to a few hundredths of a percent in fees: over 30 years on USD 100,000, the gap between 0.09% and 0.03% is worth thousands to tens of thousands of dollars. Hold long term → VOO/IVV; trade options → SPY.
Nasdaq 100: QQQ and the cheaper alternates
| Ticker | Fee (check latest data) | Notes |
|---|---|---|
| QQQ | ~0.20% | The Nasdaq-100 benchmark; heavy tech weight, big swings, big elasticity |
| QQQM | ~0.15% | Same index at a lower fee; better value for long-term holding (cheaper, fractional-share friendly) |
The Nasdaq 100 is not "the whole Nasdaq" — it is the top 100 non-financial large caps, with Apple, Microsoft, Nvidia and other giants heavily weighted; single-sector tech exposure is roughly half or more. It is a "tech growth" style, not a "broad base".
Total market: VTI / ITOT
| Ticker | Fee | Notes |
|---|---|---|
| VTI | ~0.03% | The whole US market (~3,600 stocks); "broader" than the S&P 500 |
| ITOT | ~0.03% | The iShares version |
Russell and Dow
| Ticker | Fee | Notes |
|---|---|---|
| IWM | ~0.19% | Russell 2000 small caps; big elasticity, big volatility, uneven constituent quality |
| DIA | ~0.16% | Dow Jones Industrial Average; just 30 blue chips with dated methodology (price-weighted) — more reference value than investment value |
Practical advice: for the core, pick one of SPY/VOO/IVV vs VTI; treat IWM/DIA as style supplements, no need to double up.
2. Style and Sector ETFs
| Ticker | Category | Fee (check latest data) | Notes |
|---|---|---|---|
| VUG / SCHG | Growth style | ~0.04% | Dominated by high-growth big tech |
| VTV / SCHD | Value / high dividend | ~0.04% / 0.06% | Cheap financials, energy, consumer; SCHD is the dividend specialist |
| XLK | Technology sector | ~0.09% | Heavily overlapping the Nasdaq 100 |
| XLV | Healthcare sector | ~0.09% | Relatively defensive |
| XLF | Financial sector | ~0.09% | Benefits from rising rates |
| XLE | Energy sector | ~0.09% | An amplifier of the commodity cycle |
Style and sector ETFs are "satellite positions": keep the total under 20% of the portfolio. Their volatility and drawdowns far exceed broad-base funds (a single tech-sector drawdown can top 40%+); their purpose is extra elasticity, not replacing the core.
3. Global Allocation ETFs
| Ticker | Coverage | Fee (check latest data) | Notes |
|---|---|---|---|
| VT | Whole world | ~0.07% | One share buys global stocks (~9,000+); US weight ≈ 60% |
| VXUS | World ex-US | ~0.07% | VTI + VXUS equals VT |
| VGK | Europe | ~0.08% | Major eurozone countries |
| EWJ | Japan | ~0.50% | Pricey; replaceable with VEA (developed markets overall) |
| VWO | Emerging markets | ~0.08% | China/India/Korea/Taiwan and the rest; EEM (iShares) at ~0.68% is far dearer |
| FXI | China large cap (FTSE China 50) | ~0.74% | Mostly HK-listed SOEs |
| KWEB | China internet | ~0.68% | A basket of Chinese internet companies; extremely volatile (see Article 05, ADR and Cross-Border Listing) |
The point of global allocation: a single market (especially the tech-heavy US) cannot lead forever. Over the past two decades the US made up about 60% of global market cap; the simplest way to "own everything" is VT or VTI+VXUS.
4. Bond and Commodity ETFs
Bonds
| Ticker | Name | Fee (check latest data) | Notes |
|---|---|---|---|
| BND | Total US bond market | ~0.03% | The portfolio's "ballast"; weakly negatively correlated with stocks |
| TLT | 20+ year Treasuries | ~0.15% | Longest duration, biggest kick when rates fall, biggest swings too |
| IEF | 7-10 year Treasuries | ~0.15% | Moderate duration |
| SHY | 1-3 year Treasuries | ~0.15% | Near-cash substitute; tiny drawdowns |
Commodities
| Ticker | Name | Fee (check latest data) | Notes |
|---|---|---|---|
| GLD | Gold (spot fund) | ~0.40% | Physical gold trust; IAU at ~0.25% is cheaper |
| USO | Crude oil futures fund | ~0.59% | Holds crude futures; suffers roll decay (contango bleed); clearly inferior for long-term holding |
| XLE | Energy stocks | ~0.09% | Getting oil exposure via stocks is easier on the wallet, with no futures decay |
Bonds and commodities play hedging and stabilization in a portfolio, not return generation. Gold hedges stagflation and geopolitical risk (suggest 5-10%); oil-futures ETFs are for short-term tactical positioning only.
⚖ Bonds and commodities are for hedging and stability, not returns
Bonds and commodities play hedging and stabilization in a portfolio, not return generation. Gold hedges stagflation and geopolitical risk (suggest 5-10%); oil-futures ETFs suit short-term tactical positioning only — long-term holding bleeds roll decay.
5. The Three-Fund and Permanent Portfolios, US Edition
The Three-Fund Portfolio
Construction: stocks (VTI or VOO) + international (VXUS) + bonds (BND) — three funds cover global assets.
| Mix | Risk appetite | Notes |
|---|---|---|
| 70% VTI / 20% VXUS / 10% BND | Aggressive | 90% stocks; for the young who can stomach deep drawdowns |
| 60% VTI / 20% VXUS / 20% BND | Balanced | 80% stocks; the sensible starting point for most people |
| 40% VTI / 10% VXUS / 50% BND | Conservative | The defensive mix approaching retirement |
Rebalancing worked example (USD 100,000, balanced mix, once a year):
Start of year: VTI 60,000 / VXUS 20,000 / BND 20,000
End of year: VTI +20% → 72,000; VXUS +5% → 21,000; BND −3% → 19,400
Total = 112,400; targets should be: VTI 67,440 / VXUS 22,480 / BND 22,480
Action: sell 4,560 of VTI → buy 1,480 of VXUS and 3,080 of BND- Rebalancing is forced buy-low, sell-high: trim what ran, top up what lagged, and pull portfolio risk back to the preset level
- Frequency: once a year, or triggered when drift exceeds ±5 percentage points from target; over-rebalancing only adds taxes, fees, and friction
The Permanent Portfolio (US edition)
Construction: 25% stocks + 25% long Treasuries + 25% gold + 25% cash — four equal parts to face every macro regime:
| Asset | ETF implementation | Regime it handles |
|---|---|---|
| Stocks | VTI | Economic prosperity |
| Long Treasuries | TLT | Deflation / falling rates |
| Gold | GLD | Inflation / stagflation / geopolitical crises |
| Cash | SHY or a money-market fund | "Ammunition" for bear markets and recessions |
Allocation and rebalancing worked example (USD 100,000):
Initial: VTI 25,000 / TLT 25,000 / GLD 25,000 / SHY 25,000
One year later: stocks rally → VTI 34,000, the others 26,000/24,000/25,000
Action: sell 8,750 of VTI and top up the rest, restoring 27,250 each (=109,000/4)- Pros: some asset is always rising, low psychological load, essentially no market timing needed
- Cons: lower long-run return than a pure stock portfolio (gold and cash drag); fits investors seeking "smoothness", not "maximum return"
6. DCA into US ETFs in Practice
Choosing a Broker
| Broker | Characteristics | Fits |
|---|---|---|
| Interactive Brokers (IBKR) | Low fees, full product range, global account | Advanced and long-term investors |
| Charles Schwab | Commission-free US stocks, good service | US-focused long-term investors |
| Futu / Tiger / Longbridge | Chinese UI, user-friendly | Users comfortable with Chinese |
Funding and FX Costs
| Step | Notes |
|---|---|
| Getting funds out | Under FX control, the convenience quota may not fund offshore securities investment; gray channels (salami-sliced transfers, underground banks) risk freezing and penalties; legal offshore funds or status is the precondition (see Article 03, HK and US Stocks) |
| FX cost | Depositing (CNY→USD) and withdrawing (USD→CNY) each mean one conversion; FX moves count toward investment returns: CNY appreciation erodes the CNY-denominated return on USD assets |
| Small transfers | Cross-border wires cost cable fees and intermediary-bank fees; anything under USD 1,000 per transfer is poor value — batch up before wiring |
Dividend Tax (the 30% withholding basics)
- Individuals holding US ETFs via offshore brokers (filing W-8BEN): dividends are withheld at 30% (the standard US withholding rate on non-resident dividends)
- Capital gains (the sell price spread) are generally not taxed (non-residents)
- QDII funds investing in US stocks handle dividend tax differently at the fund level — possibly better, possibly more complex
- ⚠️ Dividend tax and the application of tax treaties defer to professional tax advice; if reducing withholding matters, low-dividend broad-base ETFs are friendlier than high-dividend ETFs
7. DCA vs Lump Sum: The Math
Lump Sum (one-shot investing) and DCA (scheduled buying) are two completely different ways in:
| Dimension | Lump Sum | DCA |
|---|---|---|
| Expected return | Higher: all capital enters the market earlier and rides the trend longer | Lower |
| Downside risk | High: a crash right after entry starts you deep underwater | Lower: averaging in dilutes cost |
| Psychological load | Heavy: fully invested, watching the tape daily | Light: mechanical execution, no agonizing |
| Fits | A stock of money (a lump that arrived at once) | Monthly cash flow (paycheck investing) |
Worked example (USD 100,000, market up a steady 8% for the year):
Lump Sum: buy everything in January → 108,000 at year end (+8,000)
DCA: buy 10,000 monthly over 10 months → capital averages ~half a year in the market → ~104,000 at year end (+4,000)
(Rising market: Lump Sum earns roughly half again as much)Conclusions:
- Statistically (US stocks trend up over the long run), Lump Sum's expected return always beats DCA
- But DCA solves the "psychology problem": those who dare not go all-in at once can stick with scheduled buying
- The two combine: deploy a stock of money in 3-6 planned batches (shortening the DCA cycle) while continuing paycheck DCA — secure executability first, then talk returns
8. Portfolio Construction Checklist
| Check | What to watch |
|---|---|
| Fees | The cost of long-term holding: prefer the 0.03%-0.15% tier; above 0.5%, ask yourself whether it is worth it |
| Liquidity | Average daily volume (thin ETFs have wide spreads); whether a market maker backs the benchmark |
| Tracking error | How far the long-run NAV drifts from the index (bigger for small caps, EM, commodities) |
| Premium / discount | The gap between the traded price and NAV — EM ETFs and market manias can push premiums to 5-10%; buying at a fat premium is donating money |
| Taxes | High-dividend ETFs suffer the unfriendly 30% non-resident withholding; tax-advantaged accounts (e.g. an IRA, if available) are better |
| Correlation | Are the holdings "truly diversified" — five tech ETFs equal one fully invested sector bet |
9. US ETF Portfolios vs Domestic Fund Portfolios
| Dimension | US ETF portfolio | Domestic fund portfolio (QDII / on-exchange ETF) |
|---|---|---|
| Fees | 0.03%-0.2%, extremely low | Off-exchange subscription fees + management fees generally higher (QDII management mostly 0.5%-1.5%+) |
| Product breadth | Direct access to global assets (US stocks, global bonds, commodities, gold) | Constrained by QDII quotas; frequent purchase caps and fat premiums |
| Trading efficiency | Intraday real-time trading, flexible T+0/T+1 | Off-exchange T+1 confirmation, redemption takes about T+7 to arrive |
| Threshold | Requires an offshore account and compliant outbound funds | CNY, a few thousand yuan to start, fully compliant |
| FX | You carry the USD exposure yourself | Handled at the fund level, denominated in CNY |
| Taxes | 30% dividend withholding (W-8BEN basis) | Handled at the fund level; individuals never face US tax directly |
| Biggest obstacle | Outbound funding compliance | Fees, quotas, and premiums |
The realistic conclusion: the compliant domestic QDII channel (including on-exchange China-internet, S&P, and Nasdaq ETFs) is already enough for most people; the US ETF edge lies in fees and product range, but only on the precondition of fully compliant outbound funds — "saving 0.3% in fees while risking a frozen account" gets the priorities exactly backwards.
⚠️ Saving 0.3% in fees while risking a frozen account gets priorities backwards
The compliant domestic QDII channel (including on-exchange China-internet, S&P, and Nasdaq ETFs) is already enough for most people. The US ETF edge lies in fees and product range, but only on the precondition of fully compliant outbound funds — "saving 0.3% in fees while risking a frozen account" gets the priorities exactly backwards.
⚠️ Risk Warning
⚠️ Risk Warning
① Portfolios still lose money: no stock-bond mix avoids drawdowns; the balanced 70/20/10 still drew down 15%+ in the 2022 double slaughter of stocks and bonds. Diversification reduces "single-name blow-up risk", not "market-wide risk"; ② FX risk: USD assets price in USD; CNY-denominated return = USD return × FX change, and CNY appreciation erodes it; ③ Premium risk: QDII and EM ETFs can run 5-10% premiums in euphoric phases; buying at a fat premium plus a NAV drawdown = a double loss; ④ Commodity futures ETF decay: USO-style futures ETFs roll at a loss and will most likely trail spot over the long run; ⑤ Compliance red line: outbound funds must use compliant channels; gray funding risks frozen accounts and penalties; ⑥ All ETF tickers, fees, and allocation data in this article defer to the latest filings and latest data; tax arrangements defer to professional tax advice.
This article is education and worked examples; example prices and ratios are not investment advice.