Record Issuance Meets a Buyback Wall: Can $1.4 Trillion Absorb the AI Capex Flood?

US equity markets are absorbing a wall of new supply and doing so with remarkable calm. Goldman Sachs estimates total corporate equity issuance will reach roughly $700 billion in 2026 — a dollar record driven by IPOs, follow-on offerings, convertible bonds, and SPACs combined^[1]. Yet that headline number, when measured against the Russell 3000’s market capitalization, represents only about 1% — squarely in line with the 2015–2019 annual average^[2]. The characterization from Goldman’s strategist Ben Snider captures it: this is “more of a return to normal than a boom”^[2].

What is unusual is the concentration. AI-related companies account for roughly 40% of US follow-on offering volume, and the technology, media, and telecom sector represents nearly 30% of year-to-date follow-on volume — more than double its share over the preceding five years^[1]. The top three deals by size account for nearly half of total year-to-date issuance^[2]. This is not broad-based exuberance; it is a targeted capital-raising cycle led by hyperscalers and AI infrastructure plays.

The AI Capex Engine Behind the Issuance Wave

AI-focused capital expenditure across US businesses will approach $600 billion in 2026, representing approximately 2% of GDP and more than 10% of business fixed investment in recent quarters^[1]. Goldman identifies Amazon, Alphabet, Meta Platforms, Microsoft, and Oracle as the principal hyperscale operators driving this wave, with consensus forecasts suggesting their collective capex will surpass $1 trillion annually in coming years^[1].

These companies’ capital expenditure is anticipated to outpace their operational cash generation by approximately $150 billion in 2027; if spending reaches $1.4 trillion as some market participants expect, the financing gap could exceed $300 billion^[1]. Debt will shoulder the majority — Goldman’s credit strategists expect hyperscalers to finance 35% of 2027 infrastructure spending through debt, translating to roughly $400 billion in global bond issuance^[1]. Equity issuance is the supporting channel, not the primary one.

Secondary offerings have priced at an average markdown of approximately 7% relative to pre-announcement trading prices, and post-offering stock performance has remained consistent with historical patterns — a signal that markets are digesting the supply without indigestion^[1].

The Buyback Counterweight

The more underappreciated story is on the demand side. US corporate buyback authorizations have reached $989 billion year-to-date, a record high^[2]. Goldman projects total open-market share repurchases of $1.4 trillion for full-year 2026 — double the expected primary issuance^[2].

S&P 500 buyback activity grew 11% year-over-year in Q2 2026^[1]. Even as hyperscalers redirect cash flow from buybacks toward capex, other sectors — particularly banks and semiconductors — are expanding their repurchase programs^[2]. Goldman’s John Flood frames the supply-demand balance bluntly: even assuming all post-IPO lock-up shares are sold immediately, corporate demand still overwhelms supply^[2].

The tension worth watching: Goldman’s research notes that buybacks’ biggest benefit to investors — making each remaining share more valuable by shrinking the share count — is fading as companies redirect cash toward AI infrastructure. A Capital Group global equity study highlighted the same dynamic, with buyback yields potentially compressing in the quarters ahead^[3]. If capex continues to crowd out repurchases at the margin, the buyback ballast that has offset issuance may weaken.

Today’s IPO Pipeline: A Concrete Snapshot

The near-term calendar is thin on US-listed names but active at the margins. The notable pricing this week is Londian Wason New Energy Tech (FOIL), a China-based producer of electrolytic copper foil for lithium-ion EV batteries, which priced its upsized IPO at $22 per ADS — the top of its $20–$22 range — raising approximately $95 million and targeting a valuation of up to $1.7 billion^[4]. The company trades on the NYSE starting August 12, 2026^[4].

Renaissance Capital’s calendar for the week of August 10–15 shows a handful of small-cap and ETF pricings alongside FOIL, with no large marquee IPOs scheduled^[5]. Goldman notes that IPOs account for slightly more than $225 billion of the $700 billion full-year issuance estimate, with follow-ons, convertibles, and SPACs making up the remaining ~$450 billion^[2]. The pipeline is concentrated rather than broad.

The Reg NMS Rescission: Rewiring Market Plumbing

While the issuance story dominates the supply-demand conversation, a potentially more consequential structural shift is working its way through the SEC’s rulemaking pipeline. On June 11, 2026, the Commission voted to propose the rescission of Exchange Act Rule 611 — the Order Protection Rule, or “trade-through rule” — and Rule 610(e), the locked-and-crossed markets prohibition^[6]. These two provisions have shaped US equity market structure since their adoption in 2005 as part of Regulation NMS^[6].

Chairman Paul Atkins, who dissented from the original adoption of Rule 611 as a commissioner in 2005, described the rule as having “hindered — rather than enhanced — the long-term growth of our markets”^[6]. The SEC’s rationale rests on several pillars:

• Exchange proliferation. In 2005, roughly eight exchanges traded NMS stocks. Today, 17 operate, with three more approved. Rule 611 effectively guarantees connectivity revenue for any new exchange displaying a protected quote, driving up costs and fragmenting liquidity^[6].
• Connectivity costs. The SEC estimates a broker-dealer connecting to all exchanges spends approximately $5.7 million per year on market data and connectivity fees, with onboarding a new exchange costing an estimated $1.5 million^[6].
• Institutional friction. The rule can force brokers to interact with small-sized protected quotes across many venues before executing the balance of a large parent order, potentially signaling trading intentions and increasing slippage^[6].
• Complex order types. Compliance with Rules 611 and 610(e) spawned dozens of specialized order types — intermarket sweep orders, price-to-comply, post-only — adding layers of complexity^[6].
• Best execution as sufficient. The SEC argues that broker-dealers’ existing duty of best execution under FINRA Rule 5310 will continue to protect investors independent of Rule 611^[6].

The proposal drew a spectrum of responses at two public roundtables. Some institutional investors argued the rule forces interaction with small quotes on distant exchanges rather than allowing judgment about best execution of large orders. Others cautioned that best execution as currently constituted is “neither enforced nor enforceable” — “likened to trying to nail Jello to a wall”^[6]. Commissioner Peirce described Rule 611 as having “completed its work, and may be causing more mischief than good,” while former Commissioner Crenshaw urged, “Let’s not assume that doing away with the order protection rule is a magic bullet”^[6].

Comments are due 60 days following Federal Register publication, referencing File No. S7-2026-20^[6].

What Would Have to Be True

For the bullish case to hold — Goldman’s S&P 500 target of 8,000, against a recent record close of 7,757^[2] — three conditions need to remain intact. First, buybacks must continue at or near the projected $1.4 trillion pace; if capex displacement accelerates and repurchase growth stalls, the supply-demand calculus shifts. Second, secondary-offering discounts and post-deal performance must remain benign; a widening of the 7% average markdown or deteriorating post-announcement returns would signal market indigestion. Third, earnings growth — which has averaged 25% year-over-year across the first two quarters excluding private companies^[2] — must continue at a pace that justifies current multiples despite the supply absorption.

For the Reg NMS rescission, the key question is whether best execution obligations can substitute for the mechanical protections of Rule 611 without a deterioration in displayed liquidity and price discovery. The SEC itself is soliciting comment on whether volume-based thresholds for protected quote status might serve as an alternative to full rescission — a middle ground that could preserve some order-flow protection while reducing connectivity burdens^[6].

What to Watch Next

• Nvidia earnings on August 26 — the same week as the Fed’s Jackson Hole symposium. Goldman’s Tony Pasquariello flags this convergence as a liquidity test; NVDA has rebounded 18% from its prior-week low and announced a partnership with SK Group exceeding $500 billion^[2].
• Post-Labor Day seasonal pressure. Pasquariello cautions that increased supply and unfavorable seasonal factors may pressure markets in September, though he expects fundamentals and technicals to sustain the broader bull trend^[2].
• Reg NMS comment period. The 60-day comment window from Federal Register publication will determine whether the SEC proceeds with full rescission, adopts a compromise (e.g., volume thresholds), or stalls. Watch for 605/606 reporting reform proposals as a potential companion^[6].
• Hyperscaler financing mix. If capex exceeds consensus — as most equity investors Goldman surveyed expect^[2] — the balance between debt, equity, and buyback trade-offs will tighten. The $400 billion debt-issuance estimate for 2027 could rise, crowding credit markets even as non-AI spreads remain near historically tight levels^[1].
• India IPO pipeline. August–October windows are targeting listings for OYO, PhonePe, and potentially Reliance Jio in what Indian market sources describe as one of the busiest IPO stretches in recent memory^[5]. While these are not US-listed, they represent a parallel global issuance wave that competes for the same institutional capital pools.