By Christopher Combs
Chief Investment Officer, Silicon Valley Capital Partners
Key Takeaways
• Earnings consolidation appears complete. After years of AI‑driven multiple expansion, companies have largely re‑rated, allowing earnings to catch up to prices without a broad deterioration in profits.12
• Forward earnings revisions continue to trend higher where it matters most. Despite ongoing concerns around AI valuations, capital intensity, and economic impact, analyst expectations, led by index‑heavy U.S. mega‑cap companies—have remained resilient.21
• Falling forward rate expectations should allow a multiple‑expansion recovery in the second half of 2026. With the 10‑year Treasury yield peaking and drifting lower, discount‑rate pressure on long‑duration equities is easing, reopening the path for valuation expansion as earnings continue to compound.34
• Front‑loaded capex is increasingly giving way to project completions and positive forward guidance. As AI‑ and infrastructure‑heavy investment cycles move from peak spending toward deployment, capacity coming online and improving guidance can lift free‑cash‑flow visibility and sentiment into late 2026.15
Executive framing
Calendar year 2026 has been shaped by very scary and sometimes conflicting Wall Street narratives, while the fundamentals that typically shape markets have received far less attention. The loudest debate has been about AI, whether equity valuations ran ahead of reality, whether capital intensity will suppress free cash flow, and whether AI’s broader economic impact on labor markets and inflation will prove productive or destabilizing.67
Yet markets don’t ultimately turn on debate. They turn on arithmetic: earnings expectations and the discount rate applied to those earnings. On both fronts, the setup for the second half of 2026 is quietly improving, helped by a market that has already re‑rated, earnings expectations that remain resilient in the index’s heaviest weights, a long end that has started to ease, and an AI capex cycle that is beginning to shift from spending headlines toward tangible deployment and guidance.13
Force One: Earnings Consolidation Is Largely Complete and companies have re‑rated
The first force is not a heroic growth narrative, it is a digestive one. After a multi‑year run in which AI enthusiasm helped drive multiple expansion, the market has spent time consolidating. That process matters because it reduces fragility: prices stop doing all the work, and fundamentals are given room to catch up.12
FactSet’s own valuation snapshot shows this “re‑rating through time” dynamic. In mid‑February, FactSet pegged the S&P 500’s forward 12‑month P/E at 21.5, still above longer‑run averages, but below 22.0 at year‑end—a subtle yet consequential sign that valuation heat has been worked off without requiring an earnings collapse.1
At the same time, the earnings foundation has been firmer than the narrative implies. FactSet’s Earnings Insight shows the index tracking 13.2% blended year‑over‑year earnings growth for Q4 2025 at a late stage of reporting, alongside a record‑high revenue growth pace relative to recent years.1
Why this matters for 2H 2026: when an equity market consolidates while earnings remain intact, it reduces the odds that “valuation disappointment” becomes a self‑fulfilling downturn. It also means that if the discount‑rate headwind eases, multiple expansion can return without needing a new speculative wave.13
Force Two: Forward Earnings Revisions Remain Constructive—Where the Index Gets Its Earnings
The second force is the forward path of earnings—and here the story is less diffuse than headlines suggest. Even as AI skepticism dominates debate, revisions remain most resilient in the parts of the market that matter most for index‑level earnings.21
FactSet’s analysis of the “Magnificent Seven” is the cleanest summary of index math: for calendar year 2026, FactSet estimates ~22.8% earnings growth for the Magnificent Seven versus ~12.1% for the other 493 S&P 500 companies. That gap is large enough to support aggregate earnings expectations even if the broader market remains mixed.21
This matters because the AI debate has increasingly centered on whether the capex buildout is “too much, too soon.” But while investors argue about the return profile of AI spending, FactSet’s earnings season commentary shows positive surprises and revisions, especially in Information Technology and Communication Services—have been key contributors to the improvement in blended growth since year‑end.1
Why this matters for 2H 2026: markets don’t need uniform earnings strength to recover; they need earnings durability where index weight is concentrated. If those earnings engines keep compounding, they provide the “E” base for valuations—and create leverage to multiple expansion if rates cooperate.23
Force Three: Falling Forward Rate Expectations Should Reopen the Door to Multiple Expansion in 2H 2026
The third force is the hinge: the discount rate. For long‑duration equities, particularly innovation and AI‑adjacent growth, discount rates can dominate price action at the margin. Even modest changes in yields and real yields alter present value math meaningfully.38
Here, the facts have improved. The 10‑year Treasury yield peaked in late January and has drifted lower into late February. FRED shows the 10‑year yield around 4.02% on Feb 26, 2026, down from late‑January levels.34 Real‑yield measures show a similar easing trend through February, implying modest relief in the real discount rate applied to long‑duration cash flows.81
Crucially, lower long rates do not require a recession. Long yields reflect expected short‑rate paths and a term premium—the compensation investors demand for uncertainty. The New York Fed’s term premium framework makes explicit that term premium can compress as uncertainty and volatility fade, allowing yields to ease even in a still‑growing economy.910
One identifiable uncertainty window is Fed leadership. Chair Powell’s term as Chair ends May 15, 2026, and Reuters has emphasized that 2026 is likely to feature unusual scrutiny and uncertainty around policy leadership and independence. As clarity replaces uncertainty, term premium pressure can recede.1112
Why this matters for 2H 2026: if forward rate expectations keep easing at the margin, particularly on the long end, equities regain the capacity for multiple expansion. That is exactly the ingredient the market has lacked during consolidation: the ability for P/E to rise rather than merely rely on earnings growth.34
Force Four: Front‑Loaded Capex Is Shifting Toward Project Completions—and More Constructive Guidance
The fourth force resolves a key tension in the 2026 narrative: investors fear AI capex because it pressures free cash flow in the near term. That concern is real, and widely documented. In early February, CNBC reported that hyperscalers were projecting extremely large 2026 capex totals, raising investor concern about free‑cash‑flow compression and balance‑sheet implications.67
But the capex story has a second act: deployment.
As projects complete and capacity comes online, what has been “spending” begins to look like “throughput”, revenue capacity, delivery schedules, and monetization pathways. Industry reporting suggests that major hyperscaler AI campuses and expansions are moving from construction to operation across 2025–2026. For example, Data Center Knowledge notes Microsoft’s AI campus buildouts becoming operational and additional facilities expected to come online in early 2026, while the broader hyperscaler pipeline remains substantial.57 Construction‑focused coverage similarly highlights large AI data center projects slated to come fully online during 2026, reinforcing the notion that portions of the build cycle are reaching usable deployment.1314
The second piece is guidance—and here we can ground the narrative in FactSet. In its February Earnings Insight, FactSet reports that for Q1 2026, 38 S&P 500 companies issued positive EPS guidance versus 31 issuing negative guidance. That is not universal optimism, but it is a meaningful data point: guidance is not uniformly deteriorating beneath the capex headlines.1
Why this matters for 2H 2026: markets often struggle most at “peak spending” moments, when uncertainty is highest and cash flow optics are worst. As projects are completed and guidance stabilizes, the narrative can shift from “capex shock” to “execution and monetization.” That transition, capacity online plus constructive guidance, can tighten the link between AI investment and earnings visibility, helping both multiples and revisions into late 2026.71
How the Four Forces Combine Into a 2H 2026 Recovery Template
Put the pieces together and the late‑2026 recovery mechanism becomes unusually clear:
1. Re‑rating and consolidation have reduced fragility: forward P/E eased from year‑end while earnings growth remained solid.1
2. Revisions remain constructive where the index gets its earnings: FactSet still sees materially faster 2026 earnings growth for the Magnificent Seven than the rest of the index.21
3. Long‑end yields have begun to ease, improving the probability of multiple expansion rather than a market that can only advance through earnings alone.34
4. Capex is beginning to turn into capacity, and guidance is not collapsing: project completions and constructive guidance data help convert fear into execution.51
This is also why the market can feel stalled even as the setup improves: the narratives (AI valuation fear, macro uncertainty) tend to stay loud after the underlying math begins to turn. Markets typically bottom in ambiguity, not clarity.97
Europe: Improving at the Margin, but a Different Earnings Engine
Europe’s earnings outlook has improved modestly as the reporting season progressed, according to Reuters reporting based on LSEG I/B/E/S forecasts, though results are still expected to be weak relative to the U.S. on average.1516
Structurally, Europe also lacks the same degree of mega‑cap technology earnings concentration that drives U.S. index math. That difference matters when the catalyst is falling discount rates and resilient revisions in mega‑cap platforms. Europe can improve, but the transmission mechanism is different.152
What Could Derail the Setup
A second‑half recovery built on revisions and multiple expansion is not guaranteed. The most credible derailers are those that reverse Force Two, Three, or Four:
• Inflation or energy shocks that push term premium higher and lift long yields again.910
• A disorderly Fed transition that extends uncertainty rather than resolving it.1211
• AI monetization disappoints—capex stays high while revenue conversion and margins lag, turning revisions negative in the index heavyweights.72
• Project delays or power constraints that postpone the “capex to completions” transition and keep cash‑flow optics under pressure.57
Conclusion
Wall Street’s narratives may remain scary and conflicting. But the structural forces that typically drive markets are strengthening: consolidation has reduced valuation fragility; revisions remain resilient where index earnings are concentrated; long‑end yields have started to ease; and the capex supercycle is beginning to produce tangible completions and steadier guidance.13
If these four forces hold, the probability rises that the second half of 2026 is defined less by debate and more by the quiet return of multiple expansion—built on earnings delivery rather than hype.42
Footnotes references
1. FactSet Research Systems (2026) FactSet Earnings Insight, 13 February. Available at: https://www.factset.com/earningsinsight (Accessed: 27 February 2026). 1
2. Butters, J. (2026) ‘Are “Magnificent 7” Companies Still Top Contributors to Earnings Growth for the S&P 500 for Q4?’, FactSet Insight, 26 January. Available at: https://insight.factset.com/are-magnificent-7-companies-still-top-contributors-to-earnings-growth-for-the-sp-500-for-q4 (Accessed: 27 February 2026). 2
3. Federal Reserve Bank of St. Louis (FRED) (2026) Market Yield on U.S. Treasury Securities at 10‑Year Constant Maturity (DGS10). Available at: https://fred.stlouisfed.org/series/DGS10 (Accessed: 27 February 2026). 3
4. YCharts (2026) 10 Year Treasury Rate (I:10YTCMR). Available at: https://ycharts.com/indicators/10_year_treasury_rate (Accessed: 27 February 2026). 4
5. YCharts (2026) 10 Year Real Treasury Rate (I:10YRTR). Available at: https://ycharts.com/indicators/10_year_real_treasury_rate (Accessed: 27 February 2026). 8
6. Federal Reserve Bank of New York (2026) Treasury Term Premia (ACM model). Available at: https://www.newyorkfed.org/research/data_indicators/term-premia-tabs (Accessed: 27 February 2026). 9
7. Board of Governors of the Federal Reserve System (2022) ‘Jerome H. Powell sworn in for second term as Chair of the Board of Governors of the Federal Reserve System’, Press Release, 23 May. Available at: https://www.federalreserve.gov/newsevents/pressreleases/other20220523e.htm (Accessed: 27 February 2026). 11
8. Reuters (2025) ‘Factbox—Fed faces turbulent 2026 as Powell’s term ends, independence tested’, published via U.S. News, 9 December. Available at: https://money.usnews.com/investing/news/articles/2025-12-09/factbox-fed-faces-turbulent-2026-as-powells-term-ends-independence-tested (Accessed: 27 February 2026). 12
9. Reuters (2026) ‘European corporate outlook brightens as results defy worst fears’, published via U.S. News, 19 February. Available at: https://money.usnews.com/investing/news/articles/2026-02-19/european-corporate-outlook-brightens-as-results-defy-worst-fears (Accessed: 27 February 2026). 15
10. Reuters (2026) ‘European corporate outlook improves, but earnings overall expected to fall’, published via MarketScreener, 5 February. Available at: https://www.marketscreener.com/news/european-corporate-outlook-improves-but-earnings-overall-expected-to-fall-ce7e5ad8da8ff524 (Accessed: 27 February 2026). 16
11. Elias, J. (2026) ‘Tech AI spending approaches $700 billion in 2026, cash taking big hit’, CNBC, 6 February. Available at: https://www.cnbc.com/2026/02/06/google-microsoft-meta-amazon-ai-cash.html (Accessed: 27 February 2026). 6
12. Nicol‑Schwarz, K. (2026) ‘The Tech Download: Can hyperscalers justify their huge AI capex?’, CNBC, 13 February. Available at: https://www.cnbc.com/2026/02/13/tech-download-newsletter-ai-capex-hyperscalers.html (Accessed: 27 February 2026). 7
13. Kerner, S.M. (2026) ‘Hyperscalers in 2026: What’s Next for the World’s Biggest Data Center Operators?’, Data Center Knowledge, 6 January. Available at: https://www.datacenterknowledge.com/hyperscalers/hyperscalers-in-2026-what-s-next-for-the-world-s-largest-data-center-operators- (Accessed: 27 February 2026). 5
14. Thorpe, B. (2026) ‘Data Center Construction Boom Continues: 8 Projects to Watch in 2026’, Equipment World, 6 February (updated 23 February). Available at: https://www.equipmentworld.com/market-pulse/article/15816534/data-center-construction-boom-to-grow-in-2026 (Accessed: 27 February 2026). 13
15. Thibault, M. (2026) ‘Breaking down the data center opportunity for builders in 2026’, Construction Dive, 22 January. Available at: https://www.constructiondive.com/news/data-centers-construction-2026-trends/810016/ (Accessed: 27 February 2026). 14
February 2026