Margin

GS

Goldman goes into its July 14 Q2 print with the investment-banking recovery as its bull case — Advisory revenue nearly doubled YoY in Q1 on a rebound in completed M&A — but two things the reader should hold going in complicate the story: the IB fees backlog actually ticked *down* versus year-end 2025, and FICC was the Q1 soft spot (-10% YoY). The binding question the print resolves is whether the deal pipeline is converting into fees or stalling, and whether FICC normalizes now that Q2's equity backdrop recovered from Q1's -5% S&P quarter. Watch backlog commentary and the FICC line: a re-acceleration confirms the franchise is firing on all cylinders; a second soft FICC quarter plus a flat backlog would reframe the 14%+ ROE as advisory-cycle-dependent rather than structural.

high confidenceFreeScored

Highconfidence — what we can and can't see

Every Margin note states not just a rating but the caveats behind it — the source gaps, one-off distortions, and limits that shape how much weight to put on this read.

  • Q1 2026 'Other' revenue in GBM was $561M (+181% YoY), driven largely by net gains on direct investments — a lumpy, non-fee line; a Q2 beat leaning on similar marks would be lower-quality than a fee-led beat.
  • Q4 2025 comparisons are distorted by Apple Card markdowns (Platform Solutions posted a $(1.68)B revenue quarter) and a large Q4 net credit-loss benefit (~$2.12B reversal); sequential comps into Q2 should be read with these one-offs in mind.
  • FY2025 provision-for-credit-loss volatility (quarterly swings including benefits) makes provision a noisy line for YoY comparison.
  • No structured signal feed available for GS (get_signals returned zero); note is grounded in SEC filings plus recent news. Consumer-app / ad / app-store signals are not applicable to this ticker.
  • Consensus EPS/revenue estimates were referenced in preview news headlines but exact figures were not retrievable from source text, so no specific consensus number is cited; watch thresholds are anchored to disclosed run-rates instead.

Goldman Sachs — Q2 2026 preview (prints July 14)

The setup in one line: the market is treating Goldman as "the IB recovery trade," and Q1 handed that thesis real evidence — but the same filing quietly flagged a shrinking backlog. The print resolves whether the deal machine is building or just completing.

What we know going in (Q1 2026)

  • Global Banking & Markets $12.74B, +19% YoY. The engine.
  • Advisory $1.49B, +89% YoY; total IB fees $2.84B, +48% — completed M&A volumes jumped.
  • …but the IB fees backlog decreased slightly vs end-2025, on lower expected advisory transactions. This is the tell.
  • FICC $4.01B, -10% YoY — the drag, on weaker rates/mortgages/credit and "less favorable market-making conditions."
  • Equities $5.33B, +27% YoY — record prime financing; the quiet structural winner.
  • AWM $4.08B, +10% YoY on record average AUS.
  • ROE 14.2%, net to common $3.86B.

The three questions Q2 answers

  1. Is the backlog rebuilding? A pipeline that turns up again validates a durable up-cycle. A second flat/soft backlog reframes Q1 as a completion pull-forward.
  2. Does FICC normalize? Q1's S&P -5% tape hurt market-making. Q2's calmer equities and Goldman's own "best carry conditions since 2000" call set up a potential snap-back — or expose the weakness as structural.
  3. What's the quality of the beat? Watch the lumpy GBM "Other" line ($561M in Q1, +181% YoY on direct-investment gains). A fee-led beat is worth more than a marks-led one.

Bottom line

Near a 52-week high, GS is priced for the franchise firing on all cylinders. The headline will probably clear a low bar; the composition — pipeline vs. completions, FICC recovery vs. drag — is what should move the stock.

Signals tracked

Advisory / M&A rebound — the bull case going in

Positive

Q1 2026 Advisory revenue rose 89% YoY to $1.49B and total investment banking fees hit $2.84B (+48% YoY) on a jump in completed M&A volumes — the strongest evidence for the 'IB is back' thesis heading into Q2.

Per the Q1 2026 8-K and 10-Q, Advisory was $1,494M vs $792M a year earlier (+89%), Equity underwriting $535M (+45%, convertibles-led), Debt underwriting $811M (+8%). Management attributed the jump to a significant increase in completed M&A volumes. Q2 will show whether this is a durable up-cycle or a completion pull-forward.

IB fees backlog — the watch metric that resolves the setup

Watch

The Q1 filing disclosed the IB fees backlog DECREASED slightly vs end-2025, on lower expected advisory transactions. The Q2 print must show the backlog rebuilding to confirm the recovery is durable rather than a completion pull-forward.

CONFIRM: management commentary that backlog rose vs March 2026, driven by advisory — validates a durable pipeline. REFUTE: a second consecutive soft/flat backlog alongside softer sequential IB fees would reframe Q1's advisory surge as a pull-forward of completions, not a building pipeline. This is the single disclosure that separates a cyclical-durable thesis from a one-quarter beat. Grounded directly in the Q1 10-Q backlog language, not an invented threshold.

Sources

FICC — the swing factor into Q2

Watch

FICC was the Q1 drag at $4.01B (-10% YoY) on weaker rates, mortgages and credit amid less favorable market-making conditions. Q2's calmer equity tape and Goldman's own 'best carry conditions since 2000' call set up a potential snap-back.

CONFIRM the normalization thesis: FICC back toward the ~$4.4B Q1-2025 run-rate or up sequentially, driven by rates/credit intermediation recovery. REFUTE: a second sub-$4.1B quarter would signal the FICC weakness is structural/positioning-driven, not transient. Commodities and currencies were the Q1 bright spots within FICC; watch whether rates and mortgages recover.

Equities — the quiet structural winner

Positive

Equities revenue was $5.33B in Q1 (+27% YoY), driven by record prime financing and stronger intermediation — the franchise is compounding independent of the deal cycle and is the largest single revenue line.

Equities intermediation $2.72B (+7% YoY) and Equities financing $2.61B (+59% YoY, prime-financing-led) per the Q1 8-K. This is the durable, less-cyclical earnings anchor; continued growth here cushions any FICC or advisory softness in Q2.

Sources

Asset & Wealth Management — fee engine and mandate wins

Positive

AWM revenue was $4.08B in Q1 (+10% YoY) on higher management fees from record average AUS; recent large mandate wins ($70B from Verizon and Lockheed Martin) support the fee-based growth story into Q2.

Management and other fees $3.08B (+14% YoY) on higher average assets under supervision, partially offset by lower Marcus deposit spreads in private banking and lending (-12% YoY). The Verizon/Lockheed OCIO-style wins reported in early July feed the recurring-fee base that GS is steering toward. Watch the management-fee trajectory as the durable offset to markets volatility.

Platform Solutions — Apple Card overhang largely behind

Positive

Platform Solutions returned to a small positive Q1 (revenue $411M, near-breakeven pre-tax) after the Apple Card portfolio was moved to held-for-sale and marked down through 2025 — removing a multi-quarter earnings drag.

Q1 2026 PSol revenue $411M vs $610M a year earlier and vs a $(1.68)B Q4-2025 that carried Apple Card markdowns. FY2025 PSol pre-tax swung across large negatives (e.g. $(2.0)B, $(2.06)B quarters). With the consumer-card wind-down substantially recognized, Q2 removes a source of headline-number noise — cleaner comps ahead.

Returns run-rate — the bar the print is measured against

Flat

GS printed a 14.2% ROE in Q1 2026 (net earnings to common $3.86B, diluted EPS drivers), against a stock only ~4.6% off its 12-month peak — a demanding valuation that leaves little room for a low-quality beat.

Q1 2026 total pre-tax $5.39B, net earnings to common $3.86B, ROE 14.2% (vs 10.4% Q1-2025). With shares near highs, the market is already discounting a strong franchise; the print's job is to defend a mid-teens ROE. A beat driven by one-off DIP gains ('Other' was $561M in Q1, +181% YoY on direct-investment gains) rather than core fees would be lower quality.

Margin's read

The consensus frame is "IB is back, buy the cyclical" — and Q1 gave that thesis real ammunition: Global Banking & Markets revenue rose 19% YoY to $12.74B, with Advisory up 89% on a jump in completed M&A. But the same filing that showed the advisory surge also disclosed that the fees backlog decreased slightly versus end-2025, driven by lower expected advisory transactions. That is the tension the print must resolve: is Q1's advisory print the front edge of a durable pipeline, or a pull-forward of completions against a backlog that is no longer building? Meanwhile FICC (-10% YoY in Q1 on weaker rates and mortgages) is the swing factor — Q2's calmer equity tape and improved carry conditions Goldman itself has been flagging could see it snap back. The quiet structural story remains Equities (+27% YoY, record prime financing), which is doing the heavy lifting regardless of the deal cycle. Net: the number will likely beat the low bar, but the quality of the beat — pipeline-led vs. completion-led — matters more than the headline EPS.

How this note scored

Margin's notes are scored by an LLM judge after the earnings print, against what actually happened. Wins and misses appear with equal prominence.

Relevance5 / 5
Accuracy4 / 5
Non-obviousness4 / 5
Calibration4 / 5
Utility5 / 5
Overall4.4 / 5

Wins

  • Correctly identified the IB backlog shrinkage as the central tension — a non-obvious, filing-grounded insight that resolved in the print (IB fees +55% YoY, backlog presumably rebuilt)
  • Named FICC normalization as the swing factor; FICC did snap back strongly (+32% YoY to $4.59B), validating the thesis
  • Flagged the quality-of-beat question (fee-led vs. marks-led) — the actual beat was overwhelmingly fee-led, making this framing highly relevant
  • Equities identified as 'quiet structural winner' — confirmed with +72% YoY in Q2
  • AWM mandate wins ($70B Verizon/Lockheed) correctly linked to fee-base growth story
  • Quality flags on lumpy GBM 'Other' line and Apple Card comp distortions were specific and diagnostic
  • Watch-list items (backlog commentary, FICC line, beat composition) were all directly resolvable from the actual print

Misses

  • Confidence set to 'high' was slightly aggressive given the acknowledged backlog uncertainty and FICC structural-vs-transient ambiguity
  • No explicit disambiguation of constant-currency vs. reported figures, missing the level-5 accuracy bar
  • The note did not anticipate the magnitude of the beat (ROE 23.5% vs. 14.2% in Q1; revenue +39% YoY vs. +19% in Q1) — the 'low bar' framing underestimated the upside, though directional stance was appropriately null
  • Equity underwriting growth (+130% YoY in Q2) and debt underwriting growth (+75% YoY) were not specifically flagged as potential upside drivers beyond the general IB recovery thesis

Synthesised 10 Jul 2026 · v3.3.1 · 9 tool calls · Scored 15 Jul 2026 · v2-judge