BRIIDGE Analytics

A positive index can conceal radically different investment outcomes. Through 10 November, the S&P 500 gained 9.9%, while its equal-weight counterpart rose 15.2%. The gap points to broader market leadership: reducing the influence of the largest companies has helped performance in 2026.

The Magnificent Seven illustrate that shift. Apple, Amazon and Microsoft beat the standard index on price return, but Alphabet, Meta, Nvidia and Tesla fell. An equally allocated, buy-and-hold basket of the seven companies gained approximately 2.8%, using Alphabet Class A once. That is a simple company-level comparison, not the return of a regularly rebalanced Mag Seven index.

One index, hundreds of very different outcomes

We retrieved all 503 current S&P 500 share classes and calculated YTD returns for the 500 with a comparable 31 December 2025 closing price. Of those, 350 rose, 149 fell and one was unchanged. A total of 259 beat the standard index; the median share-price return was 10.9%. These are current-constituent statistics, not a reconstruction of historical index membership.

Moderna’s 750.6% rise and FIS’s 57.4% decline span approximately 808 percentage points. Neither result resembles the index’s 9.9% gain. Nor does a strong share price necessarily mean strong current cash conversion: Moderna’s trailing FCF margin remains negative despite its extraordinary rally.

The leaders and laggards

The first two rows are the actual price indices, followed by the ten highest and ten lowest YTD share-price returns. FCF/Sales uses the latest available trailing statements, subject to the disclosed Aptiv comparability exception. Market capitalization is approximate.

Ranks refer to the 500 securities with usable year-end baselines. Sandisk and Reddit have no comparable five-year standalone trading history. Index financial columns are blank because an index is not a corporate issuer; no synthetic issuer FCF margin or company market capitalization has been assigned.

Aptiv exception: its parent-only price return excludes the value of Versigent shares distributed on 1 April 2026, one for every three Aptiv shares. The displayed decline is therefore not the wealth loss of a shareholder who retained both holdings. A comparable continuing-operations TTM FCF margin could not be verified after the separation, so it is shown as N/A rather than combining the original full-year accounts with recast interim accounts. Latest nine-month continuing CFO was $402m against $404m of capex and $9,505m of sales.

The Magnificent Seven: no longer one trade

Apple’s 28.2% gain contrasts with Tesla’s 15.6% decline and Nvidia’s 9.4% fall. Amazon’s 17.4% advance also matters: heavy infrastructure investment has not prevented every hyperscaler from outperforming. Microsoft is only narrowly ahead of the index. The evidence argues for analysing each business, rather than treating seven different earnings models as a single position.

Alphabet uses GOOGL in the seven-company basket; GOOG returned −9.2% and is included separately in the full constituent calculations. The two share classes are not counted as two Magnificent Seven companies. Relative returns are percentage-point differences, not percentage changes in relative wealth.

Size determines influence, not the winner

The seven companies accounted for approximately 33.2% of iShares Core S&P 500 ETF holdings on 9 November, counting both Alphabet classes. Moderna represented just 0.15%. These fund weights are a transparent proxy for concentration, not official index weights or historical return contributions.

A spectacular gain in a smaller index holding can consequently have less influence than a modest move in a giant. As a static illustration, a simultaneous 10% decline in the seven companies, with all other holdings unchanged, would move that fund by roughly 3.3%. At a company-level equal-weight rebalance, seven of 500 companies represent about 1.4%, making the equivalent move approximately 0.14%. Actual index outcomes also depend on changing weights, rebalancing and corporate actions.

Equal weight’s 5.3-point advantage supports a change in 2026 leadership. It does not establish a permanent regime change. Over the same five-year window, the capitalization-weighted price index still gained 61.9%, compared with 37.4% for equal weight. Investors should distinguish this year’s broadening from the longer period of large-company leadership.

AI capex: the cash-flow test behind the rotation

Microsoft, Alphabet, Amazon and Meta are the hyperscalers within the Magnificent Seven. Apple, Nvidia and Tesla have different exposure and investment structures. The latest trailing cash-flow statements show why investors are scrutinising infrastructure returns: cash purchases of productive assets consumed most of operating cash flow at three of the four hyperscalers and exceeded it at Amazon.

TTM ended 30 September 2026 · USD billions · Standardised CFO less gross cash PPE/productive-asset purchases; finance-lease principal and asset-sale proceeds excluded.

Alphabet’s cash capex absorbed 89.0% of trailing operating cash flow. Its latest quarter generated $42.3bn of operating cash flow against $51.4bn of capex, producing negative $9.1bn of company-reported FCF. Management also identified inventory investment for TPU system sales and a $5.2bn European Commission fine payment as cash-flow headwinds; AI construction is not the entire explanation.

Amazon’s trailing gross-cash-capex FCF is negative $46.9bn on the common basis above. Its own reported FCF is negative $42.4bn because it nets $4.5bn of asset-sale proceeds and incentives against purchases. Management links the increase in investment primarily to AI and continues to expect approximately $220bn of 2026 cash capex. Revenue and operating-cash-flow growth therefore do not automatically translate into cash left after infrastructure investment.

Microsoft generated $58.0bn of operating cash flow and spent $39.4bn of cash on PPE in its September quarter, leaving $18.6bn of FCF. Its broader reported capex measure was $45.2bn, including finance-lease commencements. Meta reported $32.79bn of quarterly operating cash flow and $34.91bn of capital expenditure including finance-lease principal, leaving negative $2.12bn of reported FCF. Those lease-inclusive measures differ from the common cash-PPE measure in our trailing table.

The financial pressure is real, but causality requires care. Interest rates, valuation, earnings expectations, competitive developments and regulation can all affect prices. Nvidia’s infrastructure customers are making the spending decisions; its own FCF margin remains 41.9%. Its share-price decline cannot simply be described as the same capex burden carried by a data-centre operator.