Direct answer: beating SMH is not the same as beating QQQ
In this intraday snapshot, the equal-weight average return of Micron and SanDisk is +0.52%: above SMH at -0.06%, but below QQQ at +0.59%. The same prices support different relative-performance statements depending on the benchmark. The supported conclusion is that this two-stock sample outperformed the semiconductor ETF, not that memory stocks led the entire technology market. [1][2][3]
Generated and quotes retrieved: 2026-09-30 22:37:09 China Standard Time (UTC+8); equivalent to 2026-09-30 10:37:09 EDT. All US prices are regular-session observations for September 30, 2026, not closing results. The timestamp records completion of this request and calculation; it does not imply every security last traded at precisely the same second. [1]
One snapshot, two benchmarks
| Instrument | Intraday price, USD | Return from previous close | Versus SMH, pp | Versus QQQ, pp |
|---|---|---|---|---|
| MU | 1,070.62 | +0.52% | +0.58 | -0.07 |
| SNDK | 1,738.60 | +0.51% | +0.57 | -0.08 |
| SMH | 606.52 | -0.06% | +0.00 | -0.66 |
| QQQ | 742.31 | +0.59% | +0.66 | +0.00 |
| MU/SNDK equal-weight observation | — | +0.52% | +0.58 | -0.08 |
Source: Tiger live quotes; returns and spreads calculated for this article. Each previous-close denominator was checked against the September 29, 2026 close in Tiger historical daily bars. Calculations use unrounded inputs before displaying two decimals, so subtracting rounded table entries can produce a 0.01-percentage-point difference. The equal-weight observation is the arithmetic mean of MU and SNDK daily returns, not a traded product or a global memory-industry index. [1]
Why can the apparent lead over SMH be overstated?
Separate the observed spread into two components
For a common interval and return definition, an arithmetic identity holds:
Sample minus SMH = sample minus QQQ + QQQ minus SMH.
Using unrounded inputs and displaying three decimals:
+0.577 = -0.078 + +0.655 percentage points.
A positive spread over SMH can coexist with a negative spread against QQQ. Our calculation highlights that choosing the lower-performing benchmark produces a larger relative-return number. That entire number should not be interpreted as independent memory-stock strength. This is an arithmetic decomposition, not a capital-flow attribution or evidence that QQQ's performance caused memory stocks to rise. [1; our calculations]
The benchmarks answer different questions
- SMH: Its semiconductor production and equipment exposure offers an industry-oriented comparison. [2]
- QQQ: Its Nasdaq-100 exposure tests the result against a broader non-financial stock benchmark. It represents neither the entire US stock market nor all technology stocks. [3]
- Both together: They help prevent outperformance against a weaker benchmark from being described as broad leadership. This is descriptive comparison, not factor-adjusted or risk-adjusted alpha.
What does this snapshot not establish?
- It does not include Samsung Electronics or SK hynix, so it cannot establish synchronized strength across all four major memory companies. Korean closes are not subtracted from US intraday returns.
- It does not include a full constituent advance/decline count. Two stocks and two ETFs cannot establish industry-wide breadth.
- It does not include investor-category net transactions. Price spreads cannot identify foreign inflows, institutional rotation or the amount of new capital.
- Contracts, orders and earnings revisions were not independently verified for this article. Price performance is not proof of improving fundamentals.
- The historical percentile of these spreads was not calculated, so this observation is not labelled unusual or extreme.
What should readers verify next?
- Recalculate after the close. Keep the interval, benchmark and sample definitions unchanged, and check whether the signs persist. Do not describe this intraday article as a closing assessment.
- Track both spreads. If the sample later exceeds both SMH and QQQ, the descriptive conclusion can become “ahead of both benchmarks.” That alone would still not confirm a durable trend.
- Expand evidence before expanding the claim. An industry-wide conclusion requires more companies and aligned trading intervals. A capital-rotation claim requires relevant investor-category flow data.
Frequently asked questions
How can stocks rise without leading?
A positive return describes absolute performance. Leadership requires a named comparison. Here, positive absolute performance, a positive spread over SMH and a negative spread against QQQ coexist without contradiction. [1; our calculations]
Is the dual-benchmark table a trading signal?
No. It tests whether a headline matches the evidence. It does not estimate future returns, compensation for risk or trading costs, and contains no position sizing, trade instructions or promised returns.
Sources and methodology
[1] Tiger OpenAPI: Get Stock Briefs / Get Bars. Numerical observations come from actual API responses retrieved for this article. The public documentation explains fields; it does not host or independently reproduce this live snapshot. All returns and spreads are our calculations.
[2] VanEck's official SMH product page: used to verify the fund's semiconductor production and equipment index exposure, not to replace Tiger quotes with delayed website prices.
[3] Invesco's official QQQ product page: used to verify Nasdaq-100 tracking, not to equate QQQ with the entire market.
Scope: public-market observation and methodology. Intraday values change. This article does not attribute the day's moves to a news catalyst.