QQQM ETF $280.99 as Amazon's 13.6% Cancels Apple's 9.8% Drop

The Invesco Nasdaq 100 ETF closed Thursday at $280.99 against a Wednesday settle of $272.51 — a 3.11% single-session gain — after trading a $277.22 to $281.53 band. The fund quoted $277.72 in Friday's premarket and spent the session close to flat as the Nasdaq Composite held 25,115, down 6.78 points or 0.03%.

That two-day sequence contains the entire month. On Wednesday the Nasdaq Composite fell 1.74% to 24,442.94, closing more than 10% below its early June all-time high and entering correction territory. On Thursday it rallied 2.8% to 25,122.18, ending a six-day losing streak on the back of Microsoft's 16% surge. QQQM captured both moves nearly one-for-one.

The fund sits at $280.99 against a 52-week range of $227.00 to $308.21, leaving it 8.8% below the high and 23.8% above the low. Year-to-date return runs 10.2% with a trailing one-year total return of 19.7% to 21.08%, a three-year average of 21.6% and a five-year average of 13.8%.

Net assets stand near $101 billion with an expense ratio of 0.15% and a dividend yield of 0.51% on a $1.41 annualized distribution. The fund holds 46.7% of assets in its top ten positions.

The flow data captures the sentiment shift precisely. QQQM saw net AUM decline $1.5 billion over five days and $1.75 billion over one month, against a three-month gain of $15.77 billion, a six-month gain of $23.43 billion and a twelve-month gain of $40.65 billion. Investors added aggressively into the spring rally and have been trimming through July.

The technical read is negative. Daily buy/sell signals across moving averages register Strong Sell.

What makes this month analytically interesting is that the index went into correction while its largest constituents delivered the strongest earnings quarter of the cycle. Amazon beat by 216% on EPS. Microsoft added roughly $450 billion of market value in a single session. Four hyperscalers guided to $720 billion to $745 billion of combined 2026 capital projects. And the Nasdaq-100 still fell more than 10% from its June peak before recovering.

The reason is the discount rate. The 10-year Treasury hit 4.731% and the 30-year 5.263%, a 19-year high, with the Fed holding on a 9-3 vote and September hike odds at 63%.

Wednesday Confirmed the Correction and Thursday Erased It

The speed of the round trip is what defines this tape.

The Nasdaq-100 traded 9.7% below its record on Tuesday July 28 — just short of a correction — after the semiconductor index fell as much as 6% on a fourth consecutive losing session. Wednesday the Federal Open Market Committee held at 3.50% to 3.75% on a 9-3 vote with three regional presidents dissenting for a hike, the bond market sold off hard, and the Composite fell 1.74% to 24,442.94, crossing the 10% threshold.

The Dow lost 1,153.18 points or 2.19% that session to 51,594.14, its worst decline since April 2025. The S&P 500 slid 1.52% to 7,316.15. The 30-year Treasury jumped 10.5 basis points to 5.201% and touched 5.244%, its highest since July 2007.

Thursday reversed it entirely. Microsoft reported and rallied 16%, adding roughly $450 billion of market value. The Nasdaq Composite finished 2.8% higher at 25,122.18. The Dow surged 613.92 points to 52,208.06 and the S&P 500 climbed 1.7% to 7,437.64. The Philadelphia semiconductor index rose more than 7% by mid-afternoon.

QQQM's 3.11% gain from $272.51 to $280.99 tracked the Nasdaq-100's move, which ran slightly ahead of the Composite because the index is more concentrated in the megacaps that drove it.

Friday delivered the third act. Nasdaq futures were up 342.50 points or 1.21% at 28,580 before the open, with Nasdaq-100 futures climbing 0.9% and a chip-tracking ETF gaining 3.3% in early trading. The gap evaporated at the bell as Treasury yields ripped to 4.731% on the 10-year.

Three sessions producing a correction, a 2.8% rally and a fully retraced 1.2% gap is a market with no directional conviction and enormous single-name dispersion underneath.

Breadth data from earlier in the week showed the internals were healthier than the index level implied. The percentage of S&P 500 components trading above their 200-day moving average rose to 73%, the highest reading since December 2024, which indicates the July decline concentrated in a single sector rather than spreading across the market.

Amazon Up 13.6% Against Apple Down 9.8% Cancelled the Index Out

Friday's session is the clearest demonstration of why concentration cuts both ways.

Amazon surged 13.6% intraday after reporting second-quarter revenue of $200.61 billion against a $196.47 billion estimate, adjusted EPS of $5.75 against $1.82 expected, and AWS revenue growing 36.7% to $42.2 billion — its fastest in 18 quarters and well above the $40.54 billion forecast. Operating income rose 43% to $27.5 billion, AWS backlog reached $496 billion, and management lifted 2026 capital expenditure guidance to $220 billion from $200 billion.

Apple fell 9.17% to 9.8%, its sharpest drop in 16 months, after guiding fiscal fourth-quarter revenue growth to 9% to 11% against Street expectations above 12%, citing component shortages and severe DRAM and NAND price inflation. Services revenue of $30.74 billion missed a $31.22 billion estimate and Greater China came in at $18.8 billion, also short. Market capitalization fell to $4.4 trillion from $5 trillion days earlier.

Those two positions are the largest and second-largest weights in the Nasdaq-100 alongside Microsoft and Nvidia. A 13.6% gain in one and a 9.8% loss in the other, on the same day, produce an index change close to zero.

That is what happened. The Composite finished down 0.03% and QQQM tracked it.

The Roundhill Magnificent Seven ETF captured the same dynamic, gaining just 0.45% to $64.21 on a session where its two largest members moved a combined 23 percentage points in opposite directions.

The dispersion underneath was extreme across the rest of the index. Monolithic Power Systems ran 15%, taking its year-to-date gain to 67% after adjusted EPS of $6.50 beat a $5.87 estimate on revenue of $980.6 million against $902.5 million expected, with a $500 million buyback added. Roblox tumbled 22.5% after withdrawing full-year guidance and guiding third-quarter revenue to $1.41 billion to $1.49 billion against a $1.86 billion consensus.

For a fund holding 46.7% of assets in ten names, single-stock outcomes at that magnitude determine the daily return more than any macro variable.

46.7% in Ten Names Is the Structural Feature

Concentration is not a flaw in this product; it is the product. Understanding the magnitude is what determines whether it fits a portfolio.

QQQM holds 46.7% of net assets in its top ten positions, tracking a modified market-cap weighted index of the 100 largest non-financial companies listed on the Nasdaq. Nearly half the fund sits in ten names, and the largest four — Apple, Microsoft, Amazon and Nvidia — carry the majority of that.

Morningstar's analyst work published in April framed the risk directly, arguing the Nasdaq-100's tremendous returns obscure key flaws in its construction and cast doubt on whether the performance can continue.

Those construction flaws are specific. The index requires a Nasdaq listing, which is an exchange-selection criterion rather than an investment one — it excludes NYSE-listed companies of identical size and quality for reasons unrelated to fundamentals. It excludes financials entirely, which removed the sector from participation during the 2023 to 2025 rate cycle when banks and insurers performed well. And its modified market-cap weighting concentrates the largest names beyond what a standard cap-weighted approach would produce.

The result is a fund with the sector profile of a technology ETF and the label of a broad index product. Communication services, consumer discretionary and information technology dominate, with token allocations to consumer staples, energy, health care, industrials, materials, real estate and utilities.

The index implemented its June quarterly rebalance, which periodically resets weights and adds or removes constituents based on market capitalization ranking.

For an allocator the practical question is overlap. An investor holding a broad U.S. equity fund alongside QQQM owns the top ten names twice, at a combined weight that can exceed 30% of a supposedly diversified portfolio.

The July experience demonstrated the consequence. A 10%-plus correction driven by one sector — semiconductors posting their worst month since 2002 — took the entire index down while 73% of S&P 500 components sat above their 200-day averages.

The fund did exactly what it is designed to do. Whether that is the intended exposure is a separate question.

A 0.15% Fee and a Live Price War

The economic case for QQQM over its larger sibling is straightforward and it has been getting stronger.

QQQM charges 0.15% in total expenses with no fee waiver, no other expenses, and management fee comprising the entire figure. The category comparison is stark: large-cap ETFs average 0.65% total expense and 0.62% net, equity ETFs average 0.72% and 0.63%, and Nasdaq-focused ETFs average 0.68% and 0.62%.

The Invesco QQQ Trust — the older, larger, more heavily traded vehicle tracking the identical index — charges 0.20%. That five basis point gap is the fund's entire reason for existing. Invesco launched QQQM on October 13, 2020 designed explicitly for buy-and-hold investors, while traders and institutional buyers stay in QQQ for its deeper options market and tighter spreads.

On $101 billion of net assets, five basis points is $50.5 million annually. For an individual holder, the difference compounds meaningfully over a multi-decade horizon on identical exposure.

The competitive environment has intensified. A fee war has broken out in the Nasdaq-100 category, with investors now holding considerably more optionality when seeking that exposure. Multiple asset managers offer competing wrappers on the same index, and the pricing floor has been moving down.

That competition is a benefit to holders and a margin problem for issuers. It also raises a strategic question about the product itself: if the index can be accessed at progressively lower cost from multiple providers, the differentiation shifts entirely to liquidity, tracking quality and tax efficiency.

QQQM's tracking has been sound. Total return over the trailing year ran 21.08% including dividends, against an average annual return of 16.38% since inception in October 2020.

The distribution is minimal by design. A 0.51% to 0.52% yield on a $1.41 annualized payout reflects an index of growth companies that reinvest rather than distribute. Against a 10-year Treasury at 4.731%, this fund pays 422 basis points less in income, which places the entire return burden on price appreciation.

That is the correct configuration for a growth vehicle and the wrong one for an income allocation, and the distinction matters more when the risk-free rate sits above 4.7%.

$101 Billion in Assets and a Flow Reversal Underway

The asset trajectory shows exactly when sentiment turned, and the answer is this month.

QQQM's net AUM declined $1.5 billion over the trailing five days and $1.75 billion over one month. That contrasts against gains of $15.77 billion over three months, $23.43 billion over six months, $40.65 billion over one year, $81.98 billion over three years and $94.89 billion over five years.

A fund that added $40.65 billion across twelve months and shed $1.75 billion in the last four weeks is experiencing its first sustained outflow of the cycle.

Some of that AUM change reflects price rather than flow — the index fell more than 10% from its June peak before recovering, and mark-to-market losses reduce assets without any redemption. Separating the two matters, and the five-day figure of negative $1.5 billion against a period when the index rose 3.11% on Thursday points to genuine redemption rather than pure price effect.

The broader ETF flow picture provides context. The first half of 2026 capped a historic stretch for the industry, with assets pouring into U.S.-listed ETFs despite global returns struggling and equities falling 0.9% in June. More recently flows have shifted toward international valuation gaps and domestic large-cap equities, which is a rotation away from concentrated growth exposure rather than out of equities entirely.

The comparison to the crypto ETF complex is instructive on scale. QQQM's $101 billion in net assets exceeds the entire U.S. spot Bitcoin ETF category at roughly $74 billion to $79 billion, and dwarfs the Ether, Solana and XRP wrappers combined.

That scale creates its own dynamic. A fund holding 46.7% in ten megacap names with $101 billion of assets is a meaningful marginal buyer of those specific stocks, and redemptions convert into programmatic selling of the same names — the identical mechanism that drives Bitcoin ETF flows into spot pressure.

The three-month gain of $15.77 billion establishes that the buyer base is present and responsive to price. The one-month decline of $1.75 billion establishes that it is not unconditional.

The Index Is a Levered Bet on AI Capex Converting

The fundamental question determining this fund's next twelve months is whether hyperscaler spending produces revenue, and July delivered the first genuine evidence.

Four hyperscalers — Amazon, Microsoft, Meta and Alphabet — guided to combined 2026 capital projects of $720 billion to $745 billion. Those four plus Apple and Nvidia constitute the majority of QQQM's top ten weight. The fund is, functionally, a concentrated position in that spending programme and its returns.

Through most of July the market treated the spending as a liability. Alphabet fell 6% on a capex increase. Semiconductors entered their worst month since 2002. A $24 billion AI-focused hedge fund launched by a former OpenAI researcher slumped 67% in July after dumping positions in SK Hynix, SanDisk and Nebius into margin calls. The Nasdaq-100 traded 9.7% below its record.

Then the earnings arrived and the framing inverted. Microsoft reported Azure growth and rallied 16%, adding $450 billion of market value. Amazon reported AWS accelerating to 36.7% with a $496 billion backlog, raised capex to $220 billion, and gained 13.6% — the market rewarding an incremental $20 billion of spending because it came with contracted revenue attached.

One market participant summarized the shift precisely: the market is no longer focused on spending levels but on whether that spending works and produces a payoff.

That distinction is the single most important variable for this fund. An index carrying 46.7% in ten names, most of which are either spending on AI infrastructure or selling it, re-rates entirely on whether the return on that capital shows up.

Amazon answered for its own business. AWS added over $4.6 billion of sequential revenue — roughly 80% more than its largest increase ever — and disclosed AI and custom chip businesses each above $25 billion annualized run rates.

Apple demonstrated the other side. A company that never went on the capex binge got billed for everyone else's spending through DRAM and NAND price inflation it does not control, guided below consensus, and fell 9.8%.

The index holds both.

Semiconductors Inside the Index Had Their Worst Month Since 2002

The sector that drove the correction sits at meaningful weight inside this fund, and its month was historic.

The semiconductor complex closed July with its worst monthly performance since 2002. Month-end fund flows drove the Friday technical recovery — a chip-tracking ETF gained 3.3% in early U.S. trading — rather than any change in conviction.

Individual damage was severe among Nasdaq-listed constituents. Micron trades at $868 against a $1,255 high, down 41%, despite jumping 18% Thursday after Samsung confirmed structural memory shortages. Intel gave back 21% across seven trading sessions after entering July up roughly 270% year to date, on reports its 18A foundry process would not reach profitable yields until late 2026 or 2027 and after losing data-center revenue leadership to AMD for the first time.

The internal rotation is the more useful signal. The correlation between semiconductors and hyperscalers fell to negative 31% this month, an all-time low, with semis trading in line with capital goods rather than software. That is the market reclassifying chips as cyclical equipment suppliers rather than secular AI beneficiaries — and QQQM holds both categories at index weight.

The Korean rebound Friday showed how violently the trade reverses. The Kospi closed up 17.91% at 6,595.45, its largest single-day gain on record, with SK Hynix rising 29.95% to hit its daily limit and Samsung gaining 26.81%. That followed a 22.19% July decline and back-to-back circuit breakers on July 28 and 29 — the first consecutive halts in the exchange's history.

Korean retail investors had financed the first half of that rally with margin debt and single-stock leveraged ETFs launched only in May 2026. When prices turned, the positions liquidated themselves.

The read-through for QQQM holders is about volatility rather than direction. A fund concentrated in the AI supply chain inherits the deleveraging dynamics of that supply chain, and those dynamics now include retail leverage products in a foreign market triggering circuit breakers.

Hedge funds and other large investors have likely completed their tech and memory deleveraging, and it happened faster than the sell side expected.

A 4.731% Ten-Year Is the Valuation Constraint

The mechanism that produced the correction is not earnings. It is the rate at which those earnings get discounted.

The 10-year Treasury jumped almost 7 basis points Friday to 4.731% with an intraday print at 4.737%, the highest since January 2025. The 30-year surged to 5.263%, a 19-year high. The 2-year rose 6.6 basis points to 4.295%.

For an index of long-duration growth companies where the majority of present value sits in cash flows a decade out, the discount rate is the dominant input. A 100 basis point move in the 10-year changes the fair value of a 25-times-forward-earnings growth stock considerably more than a 5% earnings revision does.

The Fed held at 3.50% to 3.75% on July 29 in a 9-3 vote, with Hammack, Kashkari and Logan dissenting for an immediate hike — the most hawkish dissent since September 2016. September hike odds sit near 63%. Chair Kevin Warsh declined to offer forward guidance, stating there is no soft inflation target and only 2%, and described the stance as watchful thinking rather than watchful waiting.

The inflation backdrop supports the hawks. June PCE fell 0.1% month over month with the annual rate easing to 3.7% from 4.1%, but core rose 0.1% and held 3.3% year over year. That decline came almost entirely from a 9.2% drop in energy goods prices during a ceasefire window that has since collapsed, with Brent closing July at $90.36 after a 22% gain and gasoline back above $4 a gallon.

Final July consumer sentiment came in at 55.2 with one-year inflation expectations at 4.2%.

The curve shape carries additional information. The front end priced a Fed on hold while the long end priced inflation the Fed is not containing — a bear steepener following a hawkish-dissent hold. Long-end yields rising on credibility rather than growth expectations compress equity multiples without the offsetting earnings benefit that a growth-driven yield rise would provide.

Small caps showed where the pressure lands first. The Russell 2000 fell 1.45% to 2,903 on Friday while the Nasdaq held flat, because unprofitable companies with distant cash flows suffer most from a higher discount rate.

QQQM sits between the two — profitable megacaps with long-duration growth profiles.

The Technical Map: $272.51 Floors and $308.21 Caps

The chart is a correction that found support and produced a violent snapback, with the trend intact but the momentum broken.

QQQM closed Thursday at $280.99 after trading a $277.22 to $281.53 band, up 3.11% from the $272.51 Wednesday settle. The fund quoted $277.72 in Friday's premarket and traded near flat through the session.

The 52-week range runs $227.00 to $308.21. That places the fund 8.8% below the high and 23.8% above the low — the upper third of its annual band despite the correction.

Support is defined by the correction low. The $272.51 Wednesday close corresponds to the Nasdaq Composite's 24,442.94 print that confirmed the 10% drawdown. A break below it re-establishes the downtrend and opens the $265 area. Below that, the $227.00 52-week low is the structural floor, roughly 19% lower.

Resistance runs at $281.53, Thursday's high, then the $285 to $290 zone where the fund traded before the July breakdown. The $308.21 high is the reclaim target and would require a 9.7% advance.

Momentum signals are negative despite the price recovery. Daily buy/sell indicators across moving averages register Strong Sell, which reflects the fund trading below shorter-term averages that have rolled over even after Thursday's 3.11% gain.

The July 28 close at $278.14, down 0.97%, marks the level the fund has now oscillated around for four consecutive sessions — $278.14, $272.51, $280.99, and roughly $278 to $280 on Friday. That compression around a single price after a 10% drawdown and a 3% rally describes a market waiting for information.

The information arrives quickly. U.S. manufacturing PMI and the ISM employment index land August 3, JOLTS August 4, ADP and services PMI August 5, jobless claims August 6, and July nonfarm payrolls with the unemployment rate August 7. That sequence determines September Fed hike pricing, and September Fed hike pricing determines the discount rate applied to every holding in this fund.

Warsh speaks at Jackson Hole in August, which is the next scheduled opportunity to move the curve before the September 15-16 meeting.

How to Frame This Exposure Against the Alternatives

The allocation question for QQQM is not whether the Nasdaq-100 is a good index. It is what role a 46.7%-concentrated growth vehicle plays alongside everything else.

Against a broad U.S. market fund, QQQM adds concentration rather than diversification. The top ten holdings appear in both, and combining them produces a portfolio where four to six companies drive the majority of the return. The July experience quantified that: a semiconductor sector having its worst month since 2002 dragged the entire index into correction while 73% of S&P 500 components held above their 200-day averages.

Against a dividend or income vehicle, the comparison is not close. QQQM yields 0.51% against Verizon at 6.17%, AT&T at 4.78% and a 10-year Treasury at 4.731%. This fund is a total-return instrument where 100% of the expected return comes from price appreciation.

Against a sector technology ETF, QQQM is broader in name and similar in effect, with token allocations to consumer staples, energy, health care, industrials, materials, real estate and utilities that do not change the risk profile.

Against QQQ specifically, the case is entirely fee-based. Identical index, identical holdings, 0.15% versus 0.20%. Buy-and-hold investors take the cheaper wrapper. Traders take the deeper options market and tighter spreads in QQQ. That distinction has not changed since launch and the fee war in the category is narrowing it further.

The return record supports the strategy over long horizons. Since October 2020 inception the average annual return has been 16.38%, with three-year and five-year averages of 21.6% and 13.8%. Those figures include a 2022 drawdown and the current correction.

What that record does not include is a sustained period of 5% risk-free rates with a hawkish Fed. The index's outperformance was built during an era of zero and near-zero policy rates that compressed discount rates and inflated long-duration valuations.

A 30-year Treasury at a 19-year high is a different environment for this construction than the one that produced the track record.