RESEARCH MEMO · Core Beta
Leveraged Index ETFs: Long-Term Holding and Stress Testing
SPY / QQQ / SSO / QLD / UPRO / TQQQ | Path, Drawdown and Capital Efficiency
SPY · QQQ · SSO · QLD · UPRO · TQQQ
Leveraged ETFs can be held for more than one day, but long-term results are not the index return multiplied by two or three. Daily reset makes volatility, path, financing cost and drawdown decisive. SPY and QQQ fit permanent Core Beta; SSO and QLD are conditional capital-efficiency tools; UPRO and TQQQ are closer to tactical high-beta instruments.
长期回报高度依赖波动率与收益路径。
2000–2002理想化3x Nasdaq路径的最大回撤。
S&P 500长期资本效率增强的重点研究区，而非保证最优。
Nasdaq长期杠杆更保守的重点研究区。
Method confirmed | Daily reset is not a fixed long-term multiple | 2x/3x are not unconditional permanent core
01 · RESEARCH BRIEF
The one-minute brief
ProShares states that SSO, QLD, UPRO and TQQQ target two or three times the daily—not multi-year—return of their benchmarks. Returns beyond one day can differ significantly from the daily target, with higher volatility generally worsening outcomes. ACIS stress-tested idealized daily-reset paths from 1.0x to 3.0x: a 3x Nasdaq path lost about 99.9% in the 2000 crash, while 1.2x lost about 89%; a 3x S&P 500 path lost about 95.2% in 2008. Long-term capital efficiency starts with surviving the full cycle, not maximizing backtested CAGR.
Audio transcript
Leveraged ETFs can be held for more than one day, but daily two-times or three-times targets are not fixed long-term multiples. Volatility, market path, financing cost and maximum drawdown determine the outcome. In the modeled 2000 Nasdaq crash, a three-times path lost nearly one hundred percent and even a one-point-two-times path lost about eighty-nine percent. In 2008, the modeled three-times S and P five hundred path lost about ninety-five percent. ACIS therefore places SPY and QQQ in permanent Core Beta, SSO and QLD in conditional capital-efficiency research, and UPRO and TQQQ in tactical high beta. Long-term research must evaluate recovery time, dry powder and the ability to survive the next extreme cycle—not only the highest annualized return.
Known facts and open questions
- Primary section
- Investment Intelligence
- Secondary category
- Core Beta
- Report type
- Core Beta Research Note
- Scope
- Structure, path and stress capacity—not personalized leverage or trade instructions
Beta foundation
Identify the underlying index, concentration, valuation and earnings base.
Daily leverage
2x/3x is a daily target, not a multi-year promise.
Path stress
Volatility, sideways markets, slow bears and recovery speed change compounding.
Survival
Maximum drawdown and recovery time determine whether capital can stay invested.
Capital efficiency
Discuss efficiency only after tail risk is survivable.
| Vehicle | Benchmark | Daily target | ACIS role |
|---|---|---|---|
| SPY | S&P 500 | 1x | Permanent Core Beta |
| SSO | S&P 500 | 2x | Leveraged Beta / Regime dependent |
| UPRO | S&P 500 | 3x | High-Beta Tactical |
| QQQ | Nasdaq-100 | 1x | Growth Core Beta |
| QLD | Nasdaq-100 | 2x | Leveraged Growth Beta |
| TQQQ | Nasdaq-100 | 3x | High-Beta Tactical / Cycle Position |
| Asset | 5-year | 10-year | Basis |
|---|---|---|---|
| SPY | 13.65% | 15.18% | State Street NAV · Sep. 30, 2026 |
| SSO | 18.96% | 23.46% | ProShares NAV · Sep. 30, 2026 |
| UPRO | 21.86% | 28.85% | ProShares NAV · Sep. 30, 2026 |
| Nasdaq-100 benchmark | 18.16% | 22.04% | Invesco/Bloomberg · May 31, 2026 |
| QLD | 21.56% | 33.44% | ProShares NAV · Sep. 30, 2026 |
| TQQQ | 21.42% | 40.81% | ProShares NAV · Sep. 30, 2026 |
| Daily leverage | Maximum drawdown | $10,000 at trough | Recovery |
|---|---|---|---|
| 1.0x | -83.0% | $1,706 | 14.9 years |
| 1.2x | -89.0% | $1,103 | 16.8 years |
| 1.3x | -91.2% | $877 | 16.9 years |
| 1.5x | -94.6% | $544 | 17.5 years |
| 2.0x | -98.5% | $146 | 19.7 years |
| 3.0x | -99.9% | ~$6 | Not fully recovered in sample |
| Daily leverage | Maximum drawdown | $10,000 at trough | Recovery |
|---|---|---|---|
| 1.0x | -55.2% | $4,481 | 4.9 years |
| 1.2x | -62.8% | $3,723 | 4.9 years |
| 1.3x | -66.2% | $3,384 | 5.3 years |
| 1.5x | -72.2% | $2,777 | 5.4 years |
| 2.0x | -83.7% | $1,633 | 5.6 years |
| 3.0x | -95.2% | $481 | 6.7 years |
| Daily leverage | 2020 S&P | 2020 Nasdaq | 2022 S&P | 2022 Nasdaq |
|---|---|---|---|---|
| 1.0x | -33.7% | -28.6% | -24.5% | ~-32% |
| 1.2x | -39.4% | -33.6% | -29.0% | ~-38% |
| 1.3x | -42.1% | -36.1% | -31.2% | ~-41% |
| 1.5x | -47.3% | -40.9% | -35.5% | ~-46% |
| 2.0x | -58.8% | -51.8% | -45.6% | ~-57% |
| 3.0x | -76.1% | -69.6% | -62.6% | ~-74% |
| Exposure | Definition | Typical setting | Core risk |
|---|---|---|---|
| 1.0x | Permanent Core Beta | Long-term market participation | Full equity drawdown |
| 1.1x–1.3x | Capital-Efficient Core | Growth with tolerance for deeper losses | Extreme bears still amplify loss |
| 1.3x–1.5x | Aggressive / Regime-dependent | Trend, valuation and liquidity align | Tail risk rises materially |
| 2.0x | Tactical Leveraged Beta | Post-drawdown risk expansion | Not unconditional permanent core |
| 3.0x | High-Beta Cycle Position | High-conviction cyclical trend | Structural bears can destroy capital |
02 · FACTS → IMPACT → VIEW
Why does this change matter?
From return multiplier to path amplifier
- What is confirmed
- A long bull-market sample can make 2x and 3x funds look like stable multiples of index returns.
- Why it matters
- The products target daily multiples; 2000, 2008 and 2022 show drawdowns and recovery time worsening nonlinearly with leverage.
- ACIS view
- Keep SPY and QQQ as permanent Core Beta, study SSO and QLD as conditional enhancements, and classify UPRO and TQQQ as tactical high beta rather than unconditional long-term core.
03 · EVIDENCE & ANALYSIS
Evidence and analysis
01｜What leveraged ETFs actually target
SSO and QLD seek 2x daily benchmark performance; UPRO and TQQQ seek 3x. ProShares warns that returns beyond one day can differ significantly from the daily target, while the SEC highlights amplified divergence in volatile markets.
02｜Why the index can finish flat while the leveraged fund loses
If an index rises 10% and then falls 9.09%, it returns to its starting level. A 3x path rises 30% and then falls about 27.27%, ending near 94.55. Daily compounding makes volatility and sequence matter.
03｜Five- and ten-year windows can tell different stories
Through September 30, TQQQ's five-year NAV return was about 21.42%, slightly below QLD at 21.56% and UPRO at 21.86%; its ten-year return was 40.81%. Changing only the window can reverse the ranking.
04｜The 2000 Nasdaq crash is the limiting case
The modeled 1x path lost 83%, 1.2x lost 89%, and 3x lost 99.9%. The lesson is not that Nasdaq cannot be held long term; it is that permanent high leverage assumes away structural bubbles.
05｜The S&P 500 can also create capital-destruction paths
The modeled 3x S&P path lost about 95.2% in 2008, taking $10,000 near $481. UPRO is not merely a somewhat more aggressive SPY; systemic crises create a different order of loss.
06｜Recovery speed separates 2020 from 2022
The 2020 decline was fast, followed by rapid policy support and a V-shaped recovery. The 2022 regime combined higher rates, multiple compression, repeated volatility and slower repair—conditions leveraged ETFs dislike.
07｜Why 2010–2020 looked like a compounding machine
Low rates, technology earnings growth and persistent trends rewarded daily leverage. A modeled 3x Nasdaq path could turn $10,000 into about $732,000. TQQQ launched in 2010 and therefore omitted the 2000 and 2008 crashes from its live history.
08｜The capital-efficiency research zone
ACIS currently studies roughly 1.2x–1.3x for the S&P 500 and a more conservative 1.1x–1.2x for Nasdaq. These are research zones, not fixed optimal answers or recommendations.
09｜Putting leverage into a capital-allocation system
Separate permanent core from tactical leverage; vary leverage with risk premium and liquidity rather than holding it permanently; require dry powder and rebalancing capacity; and judge maximum drawdown, recovery time and behavioral execution alongside CAGR.
04 · INVESTMENT IMPLICATIONS
Industry and asset implications
Permanent Core
SPY and QQQ carry long-term market and growth beta.
Capital Efficiency
SSO and QLD belong in conditional low-leverage enhancement research.
Tactical Beta
UPRO and TQQQ fit episodic, tightly risk-budgeted use.
Risk Management
Maximum drawdown, recovery time and execution matter as much as CAGR.
These are research roles and stress-test categories, not personalized leverage targets.
05 · VALIDATION & RISKS
What to watch next
Volatility and path
Persistent trend and lower volatility support compounding.
What would weaken the view: High-volatility sideways markets or slow bears create repeated drag.
Financing and tracking
Financing cost, fees and tracking remain controlled.
What would weaken the view: High rates and derivative costs erode returns.
Tail capacity
Liquidity remains sufficient through the stress case without forced selling.
What would weaken the view: Drawdown breaks cash flow, risk limits or behavior.
Underlying fundamentals
Earnings, cash flow and valuation support long-run returns.
What would weaken the view: A structural bubble breaks or earnings reset lower for years.
What would change our view?
Historical simulations cannot guarantee future outcomes. The idealized paths do not fully include financing cost, management fees, derivative spreads, tracking error, taxes, distributions or implementation friction. Actual results may be worse, and future drawdowns can exceed the historical sample.
06 · FAQ
Key questions
Can TQQQ, UPRO or QLD be held longer than one day?
Yes, but their objectives are daily. The longer the holding period, the more volatility and path can drive a significant deviation.
Why is 3x not always best when the index rises long term?
Daily reset, volatility drag, fees and financing costs alter the compounded path. A rising benchmark can coexist with a leveraged-fund loss.
What environment is more favorable?
A lower-volatility, persistent uptrend supported by earnings and accommodative financing conditions.
What is most dangerous?
A structural bubble collapse, high-volatility range, tightening slow bear or prolonged credit and liquidity contraction.
Why not judge only by CAGR?
A drawdown near 90% can require more than a decade to recover and may force an exit before the recovery arrives.
Does ACIS recommend permanently holding 2x or 3x?
No. The framework separates 1x core, low-leverage enhancement and tactical 2x/3x tools; fit depends on the entire portfolio and risk capacity.
07 · TERMS & SOURCES
Terms, sources and related research
Key terms
- Daily Reset
- The fund re-establishes target leverage each day, so long-term return compounds from the daily path.
- Volatility Drag
- Volatility and compounding can reduce geometric return versus a simple average or fixed-multiple intuition.
- Leveraged Beta
- Market beta magnified through derivatives or financing.
- Maximum Drawdown
- The largest peak-to-trough decline over a period.
- Recovery Time
- Time required to regain the prior peak after the maximum drawdown begins.
[1] ProShares | TQQQ official fund page ↗
[2] ProShares | QLD official fund page ↗
[3] ProShares | UPRO official fund page ↗
[4] ProShares | SSO official fund page ↗
[5] SEC Investor.gov | Leveraged and inverse ETFs ↗
Public data are updated through September 30, 2026 or the date shown by each source. Historical stress tests are idealized daily-reset simulations, not tradable fund history before inception. For research and education only; not personalized investment, tax or trading advice.