brain/
← all entities
entityGSstock-market

The Goldman Sachs Group

Notes

The Goldman Sachs Group

One-line summary: Capital-markets and alternatives franchise positioned as a direct beneficiary of the AI-capex financing supercycle — management frames AI infrastructure as a 3–5-year financing-demand cycle "outstripping the appropriate quantum," with private credit the flow and the peer-low SLR the governing constraint.

What it is

Global investment bank + asset & wealth manager. Revenue mix (advisory/underwriting, FICC & equities financing, alternatives fundraising, private-credit deployment) makes GS the cleanest listed read on the intermediary side of the AI-capex build-out — who captures the fees when the buildout is financed rather than self-funded.

Why it matters to stock-market

GS is the tradeable endpoint of the chain ai-financing-supercycle-to-gs-capital-markets-rerate: as hyperscaler/neocloud capex outgrows internal cash flow (the mega-issuance-peak-to-ai-capex-derate financing-shift also seen in michael-cembalest's hyperscaler read), the financing demand routes through capital-markets and private-credit intermediaries. GS also carries the built-in falsifier — Solomon's "recalibrations in 6–18 months" warning — and the constraint (SLR at 4.3%, peer-low). Its $31B one-quarter private-credit deployment is a notable counter-signal to the retail-BDC redemption stress in bdc-redemption-spiral-to-private-credit-repricing: primary lenders leaning in while retail vehicles gate.

Key facts

  • Q2 FY2026 (record): net revenues $20.3B; EPS $20.98; ROE 23.5% / ROTE 25.5%; efficiency ratio 58.8% (−320bps). From 2026-07-14-earnings-gs-q2-fy2026.
  • AI-as-financing-demand: david-solomon — AI cycle "expanding capital needs beyond core technology into infrastructure," a "three- to five-year" build-out; financing demand "outstripping what we think is the appropriate quantum."
  • Dated falsifier: Solomon — possible "recalibrations... in the next six months, the next 18 months."
  • Private credit / alternatives: denis-coleman — $59B fundraising in Q2, $31B private credit deployed in one quarter, alternatives target raised to >$125B; AUS record $4T (34th consecutive quarter of long-term net inflows).
  • Capital markets: large-cap M&A volumes +90% H1; first bank to cross $1T announced volumes; IB backlog "highest level in 5 years"; equity financing a record +91% YoY; financing revenues $4.5B (+62%).
  • Constraint: CET1 12.9% (150bps above requirement); SLR 4.3%, lowest among peers — "a governing constraint on growing the financing business."

Strengths (thesis-input perspective)

  • First-party, quantified financing-demand read from the intermediary's seat; dated durability falsifier.
  • Private-credit deployment scale is a hard datapoint on where AI-capex financing is flowing.

Weaknesses (thesis-input perspective)

  • "GS benefits from a deal/financing boom" is partly consensus; the non-obvious pieces are the SLR constraint and the durability recalibration window.
  • Management is talking its book; the AI-capex-durability assumption is the load-bearing (and self-warned) risk.

Open questions

Sources

Related

Referenced by