SPY Stalls Below Its Record As A 4.63% Earnings Yield Loses To A 4.74% Treasury

The SPDR S&P 500 ETF Trust trades near $768.75, up roughly $3.60 or 0.47%, tracking the underlying index to 7,702.54. The fund closed Wednesday at $765.16 after a $3.38 gain and printed $767.57 in Thursday's premarket. Total net assets stand at approximately $823.3 billion on roughly 1.07 billion shares outstanding, with daily volume around 36.3 million shares — close to $28 billion of notional turnover in a single session.

The record is close and it has not been cleared. SPY's all-time closing high is $777.88, set August 13, 2026, with a 52-week intraday peak of $779.37 and a 52-week low of $627.04. At $768.75 the fund sits 1.19% below the record close, 1.38% below the intraday high and 22.6% above the low.

Today's advance came from a single source. Federal Reserve Governor Christopher Waller said he is finally seeing signs of disinflation in recent data and would support holding rates steady in September if August inflation cooperates. CME FedWatch September hike odds fell to approximately 48% from 63%. The 10-year Treasury yield dropped for a second consecutive session to 4.74% after touching 4.818% Wednesday — its highest since October 2023 — and the 2-year fell six basis points to 4.33%.

The Dow gained 0.75% to 53,460.67 and the Nasdaq 0.67% to 26,392.37. The Russell 2000 added 0.04% to 2,954.36 after giving back an intraday gain of 1.13%. The VIX slipped 1.65% to 14.95.

That divergence between the Nasdaq and the Russell is the shape of the whole session, and it maps directly onto SPY's construction problem.

Information technology accounts for 36.32% of the fund. Apple and Nvidia together represented approximately 15.29% as of July 28, and the top six holdings exceeded 28%.

The thesis for this forecast is arithmetic rather than narrative. At a forward price-to-earnings ratio of 21.59, SPY offers a forward earnings yield of 4.63% against a 10-year Treasury at 4.74% and a 30-year above 5.3%. The equity risk premium is negative.

That is why the fund has been unable to clear $777.88 despite record corporate earnings, and it is why a 7-basis-point move in the 10-year was worth 0.47% today.

The Bond Market Is What Capped This ETF, And It Just Blinked

Every failed attempt at the record over the past three weeks traces to the same variable, and it is not equity fundamentals.

The 10-year Treasury reached 4.818% on Wednesday, its highest print since November 2023, having climbed from 4.672% in late August. That 15-basis-point move across a week was driven by three pressures stacking together: renewed U.S.-Iran strikes pushing Brent through $95 and lighting the inflation channel, estimates that AI companies have raised $1.5 trillion in debt this year squeezing dealer capacity for government paper, and Chair Kevin Warsh's Jackson Hole recommitment to fighting inflation forcing the front end to price a hike that had not existed a month earlier.

Yields in the United Kingdom, Germany and France rose alongside it. The 30-year Treasury has climbed above 5.3% — a level last seen before the global financial crisis.

The equity response was mechanical. SPY was described as capped by 30-year yields above 5.3% even as strong earnings landed, sellers took control as oil rose and yields hit new highs, and bond yields were explicitly identified as crushing the AI rally.

Thursday reversed part of it. The 10-year at 4.74% and the 2-year at 4.33% unwound roughly seven and eight basis points of the Wednesday spike, and the dollar index broke below 99 from Wednesday's 99.86.

The transmission is direct and it runs through the discount rate. SPY's 36.32% technology weighting is the highest-duration equity exposure in the index — cash flows sitting five and ten years out that get repriced hard by any change in the long-term rate. Every basis point removed from the 10-year raises the present value of that 36.32%.

That is why Snowflake ripped 21.94%, Palantir 7.35% and ServiceNow 6.42% today while Broadcom fell 6.31% and Ciena 11.07%. The rally was a duration-relief trade, not an earnings trade.

The fragility is obvious. August CPI lands around September 10 with Brent at $96.20 and the ISM Services prices index at 72.6. A hot print takes hike odds back through 68%, the 10-year retests 4.818%, and every basis point of relief is given back.

SPY does not clear $777.88 while the 30-year is above 5.3%.

Information Technology Is 36.32% Of A Fund Marketed As Diversified

The concentration in this product has reached a level that changes what it actually is.

Information technology accounted for 36.32% of SPY as of the most recent disclosure, substantially more than any other sector. That is more than one dollar in three tracking a single sector's fortunes, in a fund that holds roughly 500 companies and is universally described as broad-market exposure.

Holding roughly 500 companies does not mean that every company contributes equally. A decline in a few leaders has a disproportionate effect.

The top-holdings disclosure fills in the picture. Apple sits at 7.04%, Amazon at 4.13%, Alphabet's class A shares at 3.24% with class C at 2.62%, Broadcom at 2.86%, Meta Platforms at 1.90%, JPMorgan Chase at 1.46% and Berkshire Hathaway at 1.46%. Nvidia and Microsoft carry weightings above Apple's.

Alphabet appearing twice, at a combined 5.86% across two share classes, is a structural quirk that further concentrates single-issuer risk.

The practical consequence showed up today. Broadcom at 2.86% of the fund fell 6.31% after guiding fourth-quarter revenue to $34.8 billion against $35.03 billion consensus and cutting consolidated gross margin guidance to 73% from 78% a year earlier — despite AI semiconductor revenue growing 221% year over year to $16.7 billion. That single position removed roughly 18 basis points from SPY on its own.

Offsetting it, Nvidia at a $5.449 trillion market capitalization gained 0.55% to $225.65 after agreeing to acquire Hugging Face for $12.93 billion, with approximately $1 billion in retention packages taking total deal value near $14 billion. Volume ran at a third of Nvidia's three-month average.

That is the concentration risk working in both directions on a single day.

The counterweight is that this weighting reflects where earnings growth actually is. Analysts expect roughly 8% revenue growth across SPY's holdings next fiscal year against 16% for the Nasdaq-100, with earnings growth of 16% against 25%.

SPY is buying less growth at a lower multiple. Whether that is prudence or a drag depends entirely on whether the AI capital cycle holds.

Apple And Nvidia Alone Are 15.29% Of The Fund

The two largest positions carry more weight than eight of the eleven sectors.

Apple and Nvidia represented approximately 15.29% of SPY combined as of July 28, 2026, with the top six holdings exceeding 28% of the fund. Weights change over time, and the direction of travel across this cycle has been upward.

Sizing that against the fund's $823.3 billion in assets, roughly $126 billion of SPY's capital sits in two companies. An investor buying a share of SPY is allocating 15 cents of every dollar to Apple and Nvidia before considering the other 498 constituents.

The dependency has consequences for how the fund behaves in a drawdown. In a broad selloff, diversification helps. In a concentrated selloff driven by AI capital expenditure disappointment, SPY behaves like a technology fund with a fixed-income sleeve attached.

The evidence for that scenario building is already visible in the tape. Four consecutive AI infrastructure names have reported excellent results in five sessions and been sold: Broadcom down 6.31% on 221% AI revenue growth, Ciena down 11.07% after beating adjusted EPS by 22% and adding $800 million to backlog, Hewlett Packard Enterprise down 6.81% after a beat, a raise and a $3.5 billion inference award, and Credo down 19.69% Wednesday after record revenue.

That is a crowded trade in the process of unwinding, and SPY carries it at index weight.

The bull framing is that these are the most profitable companies in the history of commerce and their index weight reflects genuine cash generation rather than speculative multiple expansion. Nvidia trades at 27.51 times trailing earnings — cheaper than the index's own trailing multiple.

The bear framing is simpler. Concentration works until it does not, and the last time the top six holdings exceeded 28% of the S&P 500, the subsequent decade's returns were driven by everything except those six.

For position sizing, the operative question is whether an investor already owns Apple and Nvidia elsewhere in the portfolio. If so, SPY is not the diversifier it appears to be.

The Broadening Nobody Priced: The Equal-Weight Index Is Beating SPY

Here is the datapoint that complicates the concentration narrative, and it deserves more attention than it gets.

As of a June measurement, SPY had returned 7.8% year to date against 10.3% for the equal-weight S&P 500 vehicle, 8.8% for the total-market fund and 15.8% for the Nasdaq-100. Over the trailing twelve months SPY returned 21.9% against 18.2% for equal weight, 22.8% for total market and 32.1% for the Nasdaq-100.

Equal weight beating cap weight year to date means the average S&P 500 constituent has outperformed the index. That is the definition of broadening, and it is the opposite of what a concentration-risk framing predicts.

The sector data confirms it. Energy has led the S&P 500 in 2026 with a 43% gain and led the third quarter with 22%, while the energy sector fund is up 47.3% year to date against 27.7% for technology — a 19.6-percentage-point gap in a year defined by artificial intelligence. Eight of eleven sectors are higher in 2026, with consumer discretionary lagging at negative 2.3% and industrials down 7% quarter to date.

The market has been broadening, and the case for looking beyond megacap technology has been made repeatedly through August.

What that means for SPY specifically is uncomfortable. A cap-weighted fund underperforms in a broadening market by construction, because its largest weights are the names losing relative share. SPY's 36.32% technology allocation was an asset through 2023 and 2024 and has been a headwind through 2026.

The counter is that broadening reverses. Today's session saw the Nasdaq gain 0.67% against the Russell 2000's 0.04%, which is narrowing rather than broadening, and it happened on a duration-relief catalyst that specifically favours the highest-multiple names.

The honest read: SPY is a bet that megacap technology reasserts leadership. If the market keeps broadening, equal weight is the better vehicle and has been for eight months.

Today's Internals: Nasdaq Up 0.67%, Russell Up 0.04%

The gap between the indices is the most informative number on the board and it undercuts the headline gains.

The Nasdaq Composite rose 0.67% to 26,392.37 while the Russell 2000 added 0.04% to 2,954.36. Small caps opened up 1.13% and gave essentially all of it back. That 63-basis-point spread on a day when the entire market rallied on the same catalyst says the money went to a specific set of large, liquid, high-multiple names.

The composition of the movers confirms it. The largest advances were dominated by crypto proxies, brokerages and enterprise software: Snowflake up 21.94%, Robinhood 14.88%, Circle 12.35%, Strategy 11.96%, Palantir 7.35%, ServiceNow 6.42%, Coinbase 7.37%, Tesla 6.47%. Nine of the twenty-five largest gainers were direct Bitcoin derivatives.

The largest declines were earnings casualties and consumer staples: Ciena down 11.07%, Campbell's 10.66% on a 36% dividend cut and fiscal 2027 EPS guidance of $1.65 to $1.80 against $2.17 delivered, Hewlett Packard Enterprise 6.81%, Broadcom 6.31%, Tyson Foods 6.47%, General Mills 4.41%.

That is a rotation, not a broadening. It is what a market looks like when one macro variable flips and everyone re-books positions against it inside four hours.

For SPY the implication is that today's 0.47% understates what happened inside the fund. Roughly a third of the portfolio moved sharply on duration relief while another meaningful slice sold off on company-specific margin problems, and the net was under half a percent.

The VIX at 14.95, down 1.65% and holding a 14-handle for the first time this week, says nobody is paying for protection into Friday's payrolls report. That is either complacency or conviction depending on the print.

Data released Thursday cut both ways. Initial jobless claims rose to 206,000 against a 205,000 forecast, showing no labour market stress. The August ISM Services PMI rose to 55.4 from 54.1, beating 54.3, with Prices Paid climbing to 72.6 and Employment at 47.8 in contraction.

Activity accelerating, prices accelerating faster, hiring shrinking. That is not a dovish datapoint.

Valuation: 21.6 Times Forward Earnings And A Negative Risk Premium

The valuation arithmetic is where this analysis has to be blunt.

SPY trades at a forward price-to-earnings ratio of 21.59, measured against a portfolio-level forward earnings estimate, and at roughly 20.3 times on an alternative aggregation — a 14% discount to the Nasdaq-100's 23.6 times. Trailing metrics put earnings per share at $24.83 with a headline multiple near 27.6.

A forward multiple of 21.59 is above many long-term historical averages for U.S. large caps. High valuations do not automatically predict an immediate decline, but they reduce the margin for disappointment.

The comparison that matters is not historical, though. It is against bonds.

A 21.59 forward multiple implies a forward earnings yield of 4.63%. The 10-year Treasury yields 4.74%. The 30-year sits above 5.3%.

The equity risk premium — the compensation an investor receives for accepting equity risk over a government guarantee — is negative 11 basis points against the 10-year and negative 67 basis points against the 30-year. Investors are paying for the privilege of owning equity risk rather than being compensated for it.

The dividend picture is worse in isolation. SPY pays $7.61 per share annually for a yield of 0.99%, with a 35.18% payout ratio and 34 consecutive years of distributions. Against a 4.74% risk-free rate, the income gap is 375 basis points in the bond's favour.

That configuration is not unprecedented — it characterised the late 1990s and much of 2021 — but in both cases it resolved through multiple compression rather than through earnings catching up.

The bull rebuttal has substance. Earnings growth of 16% expected across SPY's holdings next fiscal year on 8% revenue growth implies meaningful margin expansion, and a 4.63% earnings yield growing 16% annually beats a fixed 4.74% coupon inside two years.

The bear rebuttal is that 16% earnings growth on 8% revenue growth requires margins to keep expanding from record levels, and Broadcom's gross margin guidance cut to 73% from 78% is what that assumption failing looks like.

At a negative risk premium, the burden of proof sits entirely with the bulls.

Earnings: 97% Reported And The Guidance-Raise Signal

The completed earnings season provides the fundamental case, and it was genuinely strong.

Approximately 97% of S&P 500 companies had reported as of late August, with many delivering earnings and revenue well ahead of expectations and a notable cohort raising guidance. Raising guidance during uncertain economic times, when forward visibility is poor, has been identified as a catalyst for potential market-beating returns in subsequent quarters.

The most recent examples from this week illustrate both sides. Snowflake raised full-year product revenue guidance to $6.07 billion from $5.84 billion — lifting implied growth to 36% from 31% — while also raising its operating margin guide to 14.5% from 13.5%, and the stock ripped 21.94%. Broadcom reported fiscal third-quarter revenue of $29.591 billion, up 86%, with adjusted EPS of $3.32 beating $3.24, and fell 6.31% because fourth-quarter revenue guidance of $34.8 billion came in below $35.03 billion consensus.

The market is rewarding raises and punishing anything short of one, which is the behaviour of an index priced for perfection.

Sector composition of the beats matters for SPY's forward path. Energy, at 43% year-to-date leadership, is producing earnings on $96.20 Brent and record refining margins with the ICE gasoil crack at $79 per barrel. Technology is producing earnings on AI capital expenditure that Broadcom quantified at $350 billion of shipments to six hyperscalers across two years — $115 billion in fiscal 2027 and $230 billion in fiscal 2028.

Both of those are cyclical peaks by any reasonable reading. Neither is a base that supports the 16% earnings growth embedded in the forward multiple indefinitely.

The financial sector at 1.46% for JPMorgan and 1.46% for Berkshire provides the offset, and financials benefit from a steep curve with the 30-year above 5.3%.

The near-term earnings catalyst calendar is thin until October. Lululemon, DocuSign and Campbell's peers report after Thursday's close, and the next full reporting cycle begins with the banks in mid-October.

That leaves macro as the only driver for six weeks.

$823 Billion Of Assets And The Liquidity Franchise

SPY's competitive position rests on something no competitor can replicate quickly.

The fund holds approximately $823.3 billion in net assets and trades roughly 36.3 million shares daily — around $28 billion of notional turnover. It launched January 22, 1993, making it the oldest exchange-traded fund in the United States, and it has delivered an average annual return of 10.86% since inception with a 21.59% total return over the trailing twelve months.

That turnover is the product. SPY has become extremely popular with active traders as a way to toggle between risky and safe assets, and its daily turnover reveals a short average holding period. It is a trading instrument first and a buy-and-hold vehicle second.

The options ecosystem built on top of it is the deepest in the equity market. August produced a record-breaking week for options that powered an S&P 500 surge, and Cboe recorded a single session with four million calls traded on August 5.

For any institution needing to move meaningful equity beta in a single trade — a pension rebalancing, a hedge fund toggling gross exposure, a market maker hedging a structured product — SPY is the only venue with sufficient depth. That franchise persists regardless of fee competition.

Flow data reflects the trading character rather than allocation. A single session in a prior period saw SPY pull in $3.1 billion, and a monthly reading showed net inflows of $15.24 billion — roughly 2.0% of assets — with capital rotating into U.S. large cap core.

Those are large absolute numbers and modest relative ones. A fund that turns over $28 billion daily and takes $15 billion of net creations across a month is being used predominantly for tactical positioning.

The distinction matters for interpreting flows. Heavy SPY creation is not the same signal as heavy creation in a buy-and-hold vehicle, because a substantial share of it reverses within weeks.

Comparison across the tracking complex frames the competitive picture: Vanguard's S&P 500 fund carried $668.7 billion at a prior measurement and has been gaining ground on assets while losing badly on turnover.

The Fee Gap And Why Long-Term Money Left

SPY's one structural weakness is cost, and it has cost the fund its title.

The trust charges an expense ratio of 0.0945%. The two largest competitors tracking the identical index charge roughly a third of that. For an investor holding across a decade, that difference compounds to a material sum on a large position.

The structural reason is the fund's legal form. SPY is organised as a unit investment trust rather than as an open-end fund, which prevents it from lending securities or reinvesting dividends between distribution dates. Those constraints cost basis points that newer structures do not incur.

The consequence is a clean split in the market. Long-term allocators — retirement plans, advisory model portfolios, index-tracking mandates — have migrated to the cheaper vehicles. Traders, hedgers and options users have stayed with SPY because liquidity beats fee for any holding period measured in days.

That is why the competitor now leads on assets while SPY leads on volume by an enormous margin.

For anyone reading this as a forecast rather than a fund comparison, the practical implication is straightforward. If the intent is multi-year exposure to the S&P 500, the fee difference is real money and the cheaper vehicle is the better instrument. If the intent is to express a view over days or weeks, or to write options against the position, SPY's spread and depth outweigh the expense ratio several times over.

The dividend structure adds a wrinkle for income investors. SPY pays $7.61 per share across four quarterly distributions with a 35.18% payout ratio and 34 years of consecutive payments, yielding 0.99%. That yield has ranged from 0.97% to 1.87% across the past five years, and the current reading sits at the very bottom of that band.

A 0.99% yield at the low end of a five-year range, against a 4.74% ten-year, is another expression of the same valuation problem.

None of this changes SPY's utility. It changes who should own it.

September Seasonality Against A Bull-Strengthening Signal

The calendar is working against this fund and the market knows it.

The U.S. equity market entered September with strong momentum and faces a familiar seasonal test, heading into a month that has historically delivered the weakest average returns of the year. That framing produced a wave of cautionary commentary through late August, including warnings of a bearish September setup with seasonality aligning against the tape.

The counter-argument has been equally loud, with several assessments arguing this year should defy the September effect.

The structural case for defiance rests on positioning rather than on seasonality. The market has been described as continuing to defy gravity, steamrolling macro challenges and driving indices into record territory, and one systematic framework set a year-end 2026 S&P 500 close range of 7,892 to 8,542 while rating the market Bull Strengthening.

Translated into SPY terms at the current tracking ratio, that range maps to roughly $787 to $852 — 2.5% to 10.9% above the current price.

The internal market structure carries a warning that seasonality does not. Stocks in the S&P 500 are increasingly trading in opposite directions on a given day, which mechanically produces higher index-level volatility than the headline performance suggests. Rising dispersion inside a cap-weighted index means the average constituent's move is larger than the index's move, and that condition historically precedes volatility expansion.

Today's tape is a live example. Snowflake up 21.94% and Ciena down 11.07% inside the same index on the same session produced a 0.47% index move.

The seasonal risk is real but it is second-order. The first-order variable remains the 30-year yield above 5.3% and the 10-year at 4.74%.

If those retreat toward 4.50% on a soft CPI, September seasonality will be forgotten within a week. If they push toward 5.0%, no amount of earnings strength holds $777.88.

Technical Structure: $777.88 Above, $765.16 And $627.04 Below

The chart is a narrow consolidation beneath a record, and the levels are tight.

SPY near $768.75 sits 1.19% below its August 13 record close of $777.88 and 1.38% below the 52-week intraday high at $779.37. The 52-week low is $627.04, 18.4% beneath current levels.

Recent session closes map the range precisely. The fund closed Wednesday at $765.16 after gaining 0.44%, having tracked the index through 7,631.47 on September 1, 7,666.60 on September 2 and 7,702.54 today. The index closed at 7,730.99 on August 27 and 7,686.14 on August 31, which places today's level inside a band that has held for two weeks.

The immediate structure: $765.16 is first support at 0.47% below, and the August 27 index equivalent near $771.50 is first resistance. Above that, $777.88 is the level that defines the entire chart.

The distances are small. Clearing the record requires 1.19%. A meaningful correction to the $735 area — where the index would sit near 7,365 — requires 4.39%.

For a fund with a 22.6% range between its 52-week high and low, those are tiny moves, which is characteristic of a compressed market approaching a decision point.

The index has been described as remaining within striking distance of a record even as a key measure of U.S. borrowing costs climbed to levels last seen before the global financial crisis. That tension — a record equity market and pre-2008 borrowing costs — is the defining feature of this chart.

The confirmation signal is a daily close above $779.37 with the 10-year holding below 4.75%. That combination would confirm the market has decided the rate ceiling is in and would open the $787 to $800 zone.

The invalidation is a close below $765.16 with the 10-year back above 4.80%, which puts the August 31 lows in play and then the $735 area.

Given the VIX at 14.95 and options positioning at record volumes, whichever level breaks first will produce an outsized follow-through.

Verdict And Forecast: $800 Above $779.37, $735 On A September Hike

SPY near $768.75, sitting 1.19% below its August 13 record close of $777.88, deserves a neutral-to-constructive stance with the bond market as the sole determining variable.

The constructive case is genuine. The 10-year has retreated two consecutive sessions to 4.74% from 4.818% and the 2-year to 4.33% as September hike odds fell to roughly 48% from 63%. Earnings season closed with approximately 97% of constituents reported, many well ahead of expectations and a meaningful cohort raising guidance. Eight of eleven sectors are higher in 2026, with energy up 43% and the equal-weight index outpacing SPY year to date — evidence of genuine broadening rather than narrow megacap dependence. The fund has returned 21.59% over twelve months and 10.86% annually since its January 1993 inception, holds $823.3 billion of assets, and trades roughly $28 billion daily with the deepest options market in equities. One systematic framework carries a year-end index range of 7,892 to 8,542, mapping to roughly $787 to $852.

The cautionary case is the price. At a 21.59 forward multiple the earnings yield is 4.63% against a 4.74% ten-year and a 30-year above 5.3% — a negative equity risk premium of 11 to 67 basis points. The dividend yield of 0.99% on $7.61 per share sits at the bottom of its five-year range. Information technology is 36.32% of the fund, Apple and Nvidia are 15.29% combined, and the top six holdings exceed 28%. Four AI infrastructure names have reported excellent quarters in five sessions and been sold, with Broadcom guiding gross margin to 73% from 78%. The Russell 2000 gained 0.04% against the Nasdaq's 0.67% today, and rising intra-index dispersion means realized volatility exceeds what the index level suggests.

The forecast: rangebound between $765.16 and $779.37 until the data breaks it. A daily close above $779.37 with the 10-year holding under 4.75% opens $787 and then $800, worth 4.07%. A close below $765.16 with the 10-year back above 4.80% targets $735, 4.39% lower, and September seasonality gives that path historical support. Three events decide it: August payrolls Friday, expected at a 58,000 rebound; August CPI around September 10; and the FOMC on September 16 with the range at 3.50% to 3.75% after a 9-3 July vote. Verdict: hold, do not chase into the record, and treat the 10-year as the position's real underlying.

That's TradingNEWS