Goldman Sachs Group Deep Value Investing Analysis

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it growing a bigger slice of an existing pie, or creating an entirely new market?5/10

  A

  Bottom line up front: Goldman Sachs's market ceiling is very high, but what it does is "capture a larger share within an already-mature existing pie," not "create an entirely new market." This is crucial to Baillie Gifford LTGG's "five-fold in ten years" paradigm: what Goldman lacks is the growth attribute that would convert that room into rapid compounding growth, rather than room itself.

  The pool it sits in is indeed vast. Global investment-banking advisory, equity and debt underwriting, market-making and financing, alternative assets, and ultra-high-net-worth wealth management together form a multi-trillion-dollar-revenue industry whose profits are highly concentrated at the top. In its 2025 annual report, Goldman itself positions its business as two "world-class interconnected platforms"—Global Banking & Markets and Asset & Wealth Management—and emphasizes that it has ranked #1 in M&A advisory for the 23rd consecutive year, advising on more than $1.6 trillion of announced M&A transactions in 2025 alone. Assets under supervision (AUS) reached a record $3.65 trillion. These numbers show the ceiling is not the problem.

  But what Baillie Gifford truly asks is whether this company can achieve exponential growth by enlarging the pool or opening a new pool, rather than how big the pool is. On this point Goldman is "capturing existing share" rather than "creating new increment": the report explicitly states, "Long-term demand is stable and persistent: corporations need financing, M&A, and risk management ... but profitability is highly influenced by capital-market cycles, interest rates, volatility, regulatory capital rules, and the credit environment," and characterizes it as "a mature industry, not a high-certainty growth industry." In other words, the total size of the M&A market, underwriting market, and market-making market is largely determined by the macro environment and capital-market conditions; what Goldman can do is maintain or slightly expand its wallet share within a given pie, rather than, like a platform-type growth stock, grow the pie itself from 1 to 10.

  What genuinely carries a "growing the existing pie" flavor is its shift of the revenue mix toward "more durable" fee bases and financing relationships: the report discloses that from 2021 to 2025, FICC and Equities financing revenue combined grew at a 17% CAGR and already accounted for 37% of the two segments' total revenue in 2025; within AWM, more durable revenue such as management fees and private banking/lending has grown at a 12% CAGR since 2021. The Q1 2026 earnings report further shows that Equities financing revenue reached a record $2.608 billion, up 59% year over year. This is "deepening the wallet share of the same set of top clients"—raising penetration within existing markets, rather than creating entirely new demand.

  The report's assessment of "industry attractiveness 3/5" and "this is not an easy, good industry" is honest and accurate. For a growth investor seeking a "creator of entirely new markets," Goldman does not fit the profile: its ceiling is high, but its growth mode is a cycle-driven battle for share, lacking the self-driven, exponentially scalable new-market attribute. On this dimension, Goldman has room but lacks explosive growth power.

  Jun 10, 2026
• QOver the next five years, can its revenue at least double? Is growth driven mainly by volume, price, or new business?3/10

  A

  Bottom line up front: Goldman's revenue is almost impossible to double over the next five years—this would require roughly a 15% annualized revenue CAGR, whereas Goldman is a mature, highly cyclical financial business whose revenue has historically swung sharply with capital-market conditions, and its structural growth drivers fall far short of the doubling threshold. Even if growth is achieved, it would come mainly from "cyclical recovery in volume × price + improving fee-base durability," rather than the exponential scaling of new business.

  First, size and cadence. Full-year 2025 net revenue was $58.28 billion; doubling in five years would mean reaching about $116.5 billion by 2030. But the historical trajectory in the report shows precisely that Goldman's revenue "fluctuates up and down" rather than "climbs one-directionally": a cyclical peak of $59.34 billion in 2021, a pullback to $47.37 billion in 2022, $46.25 billion in 2023, a recovery to $53.51 billion in 2024, and $58.28 billion in 2025. In other words, over the past five years (2021→2025) revenue essentially went from peak back to peak with limited net growth—nowhere near doubling. A business that has not even doubled in five years and is highly dependent on external market conditions has a very low probability of doubling in the next five years.

  Next, the growth structure. Goldman's growth comes from three types, but each has a ceiling:

  First is "volume"—M&A, underwriting, and trading volumes rise and fall with the capital-market cycle. The Q1 2026 earnings report shows investment-banking advisory fees of $1.494 billion, up a strong 89% year over year, and record equities revenue of $5.326 billion, up 27% year over year—a classic cyclical-recovery volume surge. But the report repeatedly warns that "the 19.8% annualized ROE in the first quarter of 2026 ... should not be mechanically extrapolated," because such volume surges are essentially a cyclical dividend that will reverse once trading heat and M&A windows pull back, and they are not a sustainable doubling engine.

  Second is "price"—Goldman has premium pricing ability in complex transactions, but the report explicitly notes that it is "closer to having premium pricing ability in complex transactions and high-value-added services than to being able to raise prices at will," so pricing power is not a lever that can independently drive a doubling.

  Third is "new business/fee-base durability"—this is the most substantive one. AWM management fees reached $3.077 billion in Q1 2026, up 14% year over year, with AUS at a record $3.65 trillion and a 33rd consecutive quarter of net inflows; the report also notes that durable revenue has grown at a 12% CAGR since 2021. But even if this segment grows steadily at double-digit rates, it is only a portion of total revenue and is not enough on its own to bring overall revenue to a doubling.

  Overall judgment: doubling Goldman's revenue over the next five years is unrealistic; the more likely shape is "fluctuating with the cycle within a range of $58 billion to over $70 billion, with the fee-base portion moving up steadily." Growth is driven mainly by cyclical volume-and-price recovery, supplemented by a rising durable fee base, and lacks the "new business exponentially scaling revenue" doubling engine that Baillie Gifford values most. On this dimension, Goldman is clearly on the weaker side.

  Jun 10, 2026
• QFive years from now, what will take over as the next growth engine? Does this "second curve" exist today?5/10

  A

  Bottom line up front: Goldman's "second curve" does already exist today, and it is the most substantive part of its strategic transformation—namely Asset & Wealth Management and the "financing-type revenue" within GBM. But to be honest, this curve takes over the baton of "lower cyclicality and more durable revenue," rather than the baton of "the next high-speed growth engine"; it improves quality, but it does not raise the growth rate in a step change.

  First, why AWM and the financing business count as an "already-existing second curve." The report's core narrative is precisely that "the company is actively broadening its moat from trading energy toward more stable fee bases and financing relationships": from 2021 to 2025, FICC and Equities financing revenue combined grew at a 17% CAGR and already accounted for 37% of the two segments' total revenue in 2025; within AWM, more durable revenue such as management fees and private banking/lending has grown at a 12% CAGR since 2021. These are already-realized numbers, not a PowerPoint vision. The Q1 2026 earnings report further confirms it: AWM management and other fees of $3.077 billion, up 14% year over year, with AUS at a record $3.65 trillion and a 33rd consecutive quarter of long-term fee-base net inflows; equities financing revenue reached a record $2.608 billion, up 59% year over year. Wealth-management client assets are about $1.8 trillion. Goldman also completed acquisitions of Industry Ventures and Innovator Capital Management in 1Q26/2Q26, continuing to strengthen its position in alternative assets and ETFs.

  But the key lies in the "slope" of this curve. When Baillie Gifford asks about the second curve, it is essentially asking "five years from now, is there a new engine that can recreate another Goldman, or pull the overall growth rate back up." AWM's role is different: its value lies in "reducing cyclical volatility and improving revenue repeatability," giving the company a stable fee base to fall back on during trading troughs. The report accurately positions it as improving "revenue durability" to "offset capital-market volatility," rather than generating explosive growth. In Q1 2026 AWM revenue actually fell 14% quarter over quarter (dragged down by private-banking deposit spreads and investment gains), which also shows that it still has volatility of its own and is not a one-directional, high-growth asset.

  More importantly, there is an implicit premise: "when the core business is disrupted, can the second curve become a lifeboat." Here Goldman's situation is mediocre: its three major segments (investment banking, market-making, AWM) are all exposed to the same macro and capital-market cycle, with high correlation. Once capital markets pull back systemically, M&A, underwriting, market-making, asset valuations, and net inflows will come under pressure simultaneously, and AWM can hardly fully hedge GBM's downside. The report puts its finger on this weak point in the bear case: "once trading heat, M&A activity, IPO windows, or private-asset business pulls back, ROE will quickly return to the low-to-mid double digits or even lower."

  Honest conclusion: Goldman's second curve already exists, points in the right direction, and is executed solidly—an important basis for the report's 4/5 moat score and its positive view of management's course correction. But it is a curve of "improving quality and dampening volatility," rather than one of "shifting gears and accelerating." For a Baillie Gifford investor seeking "the next tenfold growth pole," Goldman's second curve passes on soundness but fails on growth imagination.

  Jun 10, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

  A

  Bottom line up front: Goldman's core competitive advantage is a composite moat of "brand + globally systemically important license + top-tier client relationships + scale platform + talent and risk-control culture," which is real and hard to replicate; the report's 4/5 score is fair. Over the next three to five years, this moat is likely to "widen slightly on a steady-state basis"—mainly by shifting the revenue mix toward more durable fee bases and financing relationships—but it is not the kind of network-effect moat that keeps self-widening and grows ever stronger, so the marginal widening is limited.

  First, what the moat is made of. The report breaks it down clearly: it "comes primarily from brand + regulatory licenses + global client relationships + scale + talent and culture + capital and risk-control capability, rather than from patents or classic network effects." The hardest evidence is the durability of its position—the 2025 annual report shows Goldman ranked #1 in M&A advisory for the 23rd consecutive year, advised on more than $1.6 trillion of announced M&A in 2025, and since 2020 has earned about $5 billion more in advisory net revenue than the #2 firm. Holding onto #1 for more than twenty years in a fully competitive, high-end service industry with sophisticated clients is itself the strongest evidence that a moat exists. As a global systemically important financial institution (G-SIB), its regulatory license and capital threshold naturally keep potential entrants out.

  Why judge it to "widen slightly" over the next three to five years? Because Goldman is actively shifting the center of gravity of its moat from "volatile trading energy" toward "more stable fee bases and financing relationships," and this direction is strengthening rather than weakening. The report notes that from 2021 to 2025, FICC and Equities financing revenue grew at a 17% CAGR and already accounted for 37% of the two segments in 2025; in the Q1 2026 earnings report, equities financing reached a record $2.608 billion, up 59% year over year, AWM management fees rose 14% year over year, AUS hit a record $3.65 trillion, and it was the 33rd consecutive quarter of net inflows. Financing-type business and management fees have higher client stickiness than one-off advisory transactions, which means the "switching cost" attribute of the moat is increasing. The report's sentence—"it is actively broadening its moat from trading energy toward more stable fee bases and financing relationships"—is precisely the basis for "widening."

  But the limits must be stated honestly, to avoid overstatement. First, the report explicitly notes that Goldman "does not have an extremely strong consumer-mindshare-type brand moat, nor strong network effects like a payment network or a search engine"—it lacks the compounding mechanism of "the more users, the more valuable, automatically snowballing," so the widening is "built up bit by bit through operations," rather than something the structure comes with. Second, pricing power is limited; the report says it is "closer to having premium pricing ability in complex transactions than to being able to raise prices at will." Third, the "technology and product substitution risk" the report lists is a real headwind: the rise of electronic trading, private credit, and alternative financing channels could redistribute the profit pool and erode the excess profits of traditional investment banking and market-making—a countervailing force in the "narrowing" direction for the moat. Fourth, the moat depends heavily on talent density and risk-control discipline; a single major risk-control or litigation event could locally weaken trust.

  Overall: Goldman's moat is real, wide, and durable; over three to five years it is more likely to "widen slightly" on a steady-state basis (fee-base durability) than to "narrow significantly"; but it also will not keep self-widening the way a top network-effect company does. This is a high-quality moat that "needs continuous operational maintenance"—worthy of 4/5, but not a perfect score.

  Jun 10, 2026
• QIf its core business is disrupted, does it have the gene to reinvent itself? How does it treat mistakes and bad news?5/10

  A

  Bottom line up front: Goldman has a genuine gene for "self-reinvention under disruption pressure"—over the past few years it went from an institution aggressively chasing consumer finance to actively contracting back to its two core platforms of investment banking and asset management, a successful strategic reinvention; it also treats mistakes relatively candidly, openly acknowledging the cost and quantifying the losses in its annual report. But to be honest: this reinvention was "correcting errors and returning to the core business," proving "the ability to admit mistakes" rather than "the ability to build something new from scratch when the core business is truly disrupted by technology"—the latter, deeper reinvention gene has not yet been tested.

  First, the positive evidence for self-reinvention. The report accurately summarizes the direction of Goldman's latest transformation: "shrinking the balance sheet of capital-intensive historical principal investments, exiting or scaling back consumer-finance expansion it is not good at, emphasizing the synergy of the core GBM and AWM platforms, and bringing capital intensity down." Quantitatively, the report cites the annual report: since 2020 the company's more durable revenue has doubled, historical principal investments have fallen from about $64 billion to about $6 billion (a reduction of over 90%), and the stress capital buffer has fallen by a cumulative 320 basis points. Being able, within a few years, to make a clean break from a business that consumed large amounts of capital yet delivered subpar returns, and to refocus resources on its strongest platform, is a genuine, top-down reinvention execution capability.

  Next, how it treats mistakes and bad news—the core of this question. Goldman's approach here deserves credit: it openly acknowledges and quantifies mistakes rather than glossing over them. The report records two key self-disclosures: first, in 2023, "management itself acknowledged that several items related to strategic contraction and the FDIC special assessment reduced that year's net income by about $2.8 billion and EPS by $8.04"; second, in 2025, "Platform Solutions took an approximately $2.26 billion revenue-side writedown because Apple Card loans were moved to held-for-sale and because of contract-termination obligations," and it continued to disclose the Apple Card portfolio-related writedown impact in the Q1 2026 earnings report. Putting the failure of its consumer-finance strategy on the table and accounting for the cost with concrete figures, rather than papering over it, is transparency that counts as a plus for long-term shareholders. The report therefore offers the fair assessment that "management has already made costly mistakes and is currently correcting course, rather than being incapable of mistakes."

  But the implicit premise must be added: when the core business is truly disrupted, does it have the reinvention gene to "build another growth pole"? Here one should be conservative. Its recent reinvention was "contracting back and returning to the core business it is best at," which is entirely different in difficulty from "opening an entirely new battlefield after the core is disrupted." The "technology and product substitution risk" and "business-model disruption risk" the report lists—electronic trading, private credit, and alternative financing channels changing the distribution of profit—are precisely the scenarios that truly test the reinvention gene, and Goldman's response to such disruption (such as thickening financing-type and asset-management-type revenue) currently looks like "defensive reinforcement" rather than "offensive re-creation." In addition, the report still has reservations about capital allocation: in the first quarter of 2026, after a sharp run-up in the stock, it still aggressively repurchased $5 billion at an average price of $923.49, while standardized CET1 fell from 14.3% in 4Q25 to 12.5%, showing that its discipline after error-correction is not flawless either.

  Honest conclusion: on the dimension of "admitting mistakes—contracting—returning to the core," Goldman has shown a genuine and quantifiable gene for reinvention and candor, and its handling of bad news is transparent; but the higher-order reinvention ability of "building from zero after the core is disrupted" has not yet been through a true life-or-death test. The gene passes on error-correction and awaits verification on re-creation.

  Jun 10, 2026
• QDoes management (especially the founder) have a long-term vision, with interests deeply aligned with the company? Is it willing to sacrifice current profit for five to ten years down the road?4/10

  A

  Bottom line up front: Goldman has no "founder"-type soul figure; it is a century-old institution governed by professional managers with highly dispersed ownership; the alignment between management and shareholder interests is achieved mainly through heavily equity- and performance-based compensation, rather than through a large founder shareholding. On "long-term vision and willingness to sacrifice current profit for five to ten years down the road," its score is "moderate, with reservations"—the direction over the past two years has indeed been more long-term and more restrained, but historically it made the mistake of chasing short-term expansion, and its current capital-allocation discipline is still imperfect.

  First, let me spell out the alignment mechanism. Goldman has long since ceased to be a partner-owned private firm; after its 1999 IPO, ownership became extremely dispersed, with no founder or family controlling the company. Chairman and CEO David Solomon personally holds about 153,000 shares, a considerable absolute market value at the current price of about $1,001, but well below 1% of the roughly 300 million total shares outstanding—a classic "low absolute holding, strong compensation alignment" structure of a large systemically important bank, rather than deep founder alignment. The real handle for alignment is compensation design: most of Solomon's roughly $47 million annual total compensation is performance-based, tied to the company's long-term results and to shareholder returns through performance share units (PSUs) and carried-interest plans. The report also honestly acknowledges this information boundary: "the precise percentage of management's personal shareholding also needs a further careful reading of the original proxy statement to be quantified more precisely ... on the conclusion regarding the degree of alignment between management and shareholder interests, I give moderate confidence."

  Next, "whether it has a long-term vision and is willing to sacrifice the present for five to ten years down the road." The positive evidence is that the strategy over the past two years has indeed been more long-term and more willing to endure short-term pain: the report notes that management actively "shrank the balance sheet of capital-intensive historical principal investments and exited or scaled back consumer-finance expansion it is not good at," quantified as historical principal investments falling from about $64 billion to about $6 billion since 2020 and the stress capital buffer declining by a cumulative 320 basis points. Exiting consumer finance and cutting principal investments by over 90% means acknowledging losses and sacrificing book-value scale in the short term; the report records that related items in 2023 reduced that year's net income by about $2.8 billion and EPS by $8.04—for the sake of long-term capital efficiency, management did swallow a decline in short-term profit. This is genuine evidence of "sacrificing the present for the long term."

  But the reservations are equally grounded. First, this long-term vision is "restraint after error-correction," and management previously plunged headlong into low-return consumer finance precisely because it lacked long-term discipline; the report characterizes it as "having already made costly mistakes and currently correcting course," rather than consistent long-termism. Second, current capital allocation still carries a short-term flavor: in Q1 2026, after a sharp run-up in the stock, it still aggressively repurchased $5 billion at an average price of about $923.49, while standardized CET1 fell from 14.3% in 4Q25 to 12.5%—on which the report explicitly states "this is where I have the most reservations about capital allocation," arguing that making large buybacks at a high valuation while the buffer narrows is "not perfect from a long-term owner's perspective." A capital allocator who truly operates on a ten-year scale should have repurchased more restrainedly when the stock re-rated significantly, prioritizing thickening the capital buffer.

  Honest conclusion: Goldman is governed by professional managers, has no founder, and has dispersed ownership, aligning interests through performance-based compensation rather than a large shareholding; management has shown genuine "sacrifice the short term for the long term" actions over the past two years, but this is error-correction-style long-termism, and its current buyback discipline still has flaws. This dimension merits the report's "pass but with reservations," and falls short of the "founder-level long-term deep alignment" that Baillie Gifford values most.

  Jun 10, 2026
• QIf it disappeared tomorrow, how much would its clients miss it? Is its growth model sustainable and not reliant on harming society or regulatory arbitrage?6/10

  A

  Bottom line up front: if Goldman disappeared tomorrow, its top-tier institutional clients would "miss it quite a lot"—in high-value-added scenarios like large complex M&A, equity and debt underwriting, market-making and financing, and ultra-high-net-worth wealth management, it is one of the few indispensable players with global execution ability; but its "indispensability" is of the "one member of a leading oligopoly" kind, rather than "an irreplaceable sole infrastructure." On the composite sub-item of "whether growth relies on harming society and regulatory arbitrage," Goldman is in a neutral-to-cautious zone: as a heavily regulated entity its compliance costs are high, and it has had regulatory fines and litigation historically; growth itself does not rely on harming society, but regulatory friction is a real constraint on its long-term returns.

  First, "how much clients would miss it"—that is, indispensability. The evidence is its durable lead in high-value segments: the 2025 annual report shows Goldman ranked #1 in M&A advisory for the 23rd consecutive year, advised on more than $1.6 trillion of announced M&A in 2025, and since 2020 has cumulatively earned about $5 billion more in advisory net revenue than the #2 firm. The report summarizes that its brand "has global recognition in large complex M&A, equity and debt underwriting, market-making and financing, alternative assets, and ultra-high-net-worth wealth management." For the issuer in a ten-billion-dollar-scale cross-border M&A, or a large IPO, the institutions that can provide top-tier execution and an underwriting network are few, and Goldman is a key one among them—without it, the client's choice set would clearly narrow and its bargaining power and execution quality would decline. This is the source of "would miss it a lot."

  But the "degree" of indispensability must be honestly defined. The report pinpoints the limit: Goldman "does not have an extremely strong consumer-mindshare-type brand moat, nor strong network effects like a payment network or a search engine"; in investment-banking advisory, market trading, and underwriting, client switching costs "depend on client relationships and execution quality and cannot simply be treated as rigid switching costs." In other words, its indispensability is built on "relationships + execution + scale," while leading rivals such as JPMorgan, Morgan Stanley, and Bank of America can provide highly similar substitutes—the report explicitly lists these as main rivals and notes that "the profit pool is clearly concentrated among a handful of leading platforms." So clients would miss Goldman, but would not be "left with no way out"; it is the standout among the oligopolists, not a monopolistic sole node. By comparison, wealth management and private banking have higher stickiness (larger switching costs), and those clients would miss it more deeply.

  Next, the implicit second premise—"whether the growth model is sustainable and does not rely on harming society and regulatory arbitrage." Goldman's growth logic (financing, M&A, and risk management for corporations; liquidity for institutions; asset allocation for the wealthy) is itself legitimate financial-intermediation service, not built on harming consumers or regulatory arbitrage. But it is a global systemically important financial institution, strongly constrained by capital requirements, stress tests, and G-SIB rules—the report lists "regulatory risk—capital requirements, stress tests, G-SIB rules, and litigation penalties could all eat into shareholder returns" as one of the most important risks, and lists "a major risk-control, compliance, or litigation event forcing additional provisions or significantly higher capital requirements" as a signal that would overturn the investment judgment. Historically Goldman has indeed carried regulatory and litigation baggage (one reason its compliance costs are high). From a "social sustainability" angle, it does not grow by systematically harming society, but its earnings and expansion are always "braked" by the regulatory-capital ceiling; in Q1 2026 standardized CET1 fell to 12.5% and risk-weighted assets rose, showing precisely that growth must continuously consume regulatory capital—a real constraint on its growth sustainability.

  Honest conclusion: Goldman is "quite indispensable" to top-tier clients and would clearly be missed if it disappeared, but it is one member of the oligopoly tier rather than the sole substitute; its growth does not rely on harming society, yet is subject over the long term to the hard constraints of heavy regulation and compliance costs. On this dimension it passes, but does not dazzle.

  Jun 10, 2026
• QWhat are the unit economics (gross margin, incremental returns) of this business? Do they improve or worsen as it scales? Where does the money it earns go?6/10

  A

  Bottom line up front: Goldman's unit economics are excellent in "good years"—ROE can reach 15% or even nearly 20% annualized, with the efficiency ratio pressed to the low 60s%; but its incremental returns are a hybrid that "swings dramatically with the capital-market cycle and requires regulatory capital to grow," rather than the structural increase of "asset-light, the bigger the better." The money it earns is returned to shareholders mainly through dividends and large buybacks, a reflection of its strong cash generation, but its current aggressive buybacks at a high stock price are also where the report has the most reservations.

  First, the "unit economics" level of profitability. The numbers from the report and earnings releases are quite impressive: full-year 2025 ROE of 15.0%, net margin of about 29.5%, and an efficiency ratio of about 64.4%; Q1 2026 annualized ROE of 19.8%, ROTE as high as 21.3%, and the efficiency ratio pressed to 60.5%. For financial institutions, the report explicitly notes that "gross margin/operating margin has limited industry meaning, and the efficiency ratio and ROE should be examined instead"—the correct lens. Through this lens, Goldman's unit economics in prosperous periods are first-tier on Wall Street, as the scale platform lets it spread fixed talent, technology, and compliance costs over a huge revenue base, giving the efficiency ratio flexibility.

  But "do things get better or worse as it scales" must be answered honestly from two sides.

  The positive: scale does bring dilution effects and synergy. The report notes that the platform effects of GBM and AWM are evident—"top clients bring advisory revenue as well as financing, hedging, asset-management, and private-wealth business"—monetizing one client in multiple places, which is economies of scale. The efficiency ratio can be pressed to the low 60s% in prosperous periods, showing that the marginal cost of incremental revenue is relatively controllable.

  The negative (and the crux): Goldman's incremental returns have two structural drags. The first is cyclicality—the report's historical table clearly shows unit economics swing violently: ROE of 23.0% in 2021, dropping to 10.2% in 2022, 7.5% in 2023, back to 12.7% in 2024, and 15.0% in 2025. The same company's ROE can be halved within two years, showing that its incremental returns depend heavily on external market conditions and are not steadily increasing. The second is that "growth eats capital"—the report points out this is the core difference from asset-light companies: "even if book profits are high, if business expansion, market risk exposure, and loan growth push up RWA, part of the cash distributable to shareholders will be consumed by regulatory capital needs." In Q1 2026, risk-weighted assets rose and standardized CET1 fell from 14.3% in 4Q25 to 12.5%—a portrait of incremental growth consuming capital. The report's conclusion is precise: "GS is not a laggard that needs more cash the more it grows, but it is certainly not a top student whose growth barely consumes capital."

  In addition, the high unit economics in prosperous periods also contain "non-recurring water," so extrapolation would overstate them. The report discloses that in 2025 tax benefits related to the settlement of employee equity awards were about $620 million, adding about $1.95 to full-year EPS; in Q1 2026 such tax benefits were about $895 million, adding about $2.91 to quarterly EPS and lifting annualized ROE by about 3.1 percentage points. So the 19.8% ROE is "clearly above a normalized return level." The report therefore sets normalized Owner Earnings at about $33.5 / $40.0 / $48.0 per share for the conservative/neutral/optimistic cases, far below the headline EPS of prosperous periods.

  Where does the money go? Mainly returned to shareholders. The report records that in 2025 it returned $16.78 billion of capital to common shareholders, of which $12.36 billion was buybacks (average price about $654.45); in Q1 2026 it repurchased another $5 billion, about 5.4 million shares, with the average price rising to $923.49. Strong capital-return ability reflects good unit economics, but the report gives a clear reservation on "still aggressively buying back after a sharp run-up while the capital buffer narrows"—good money not spent prudently enough.

  Honest conclusion: Goldman's unit economics are excellent in prosperous periods, scale brings genuine dilution and synergy, and money is generously returned via dividends and buybacks; but incremental returns swing violently with the cycle and growth continuously consumes regulatory capital, so it lacks the growth-stock trait of "the bigger the more profitable, asset-light compounding." Its unit economics are strong on profitability, weak on stability and extrapolability.

  Jun 10, 2026
• QFor it to rise fivefold in ten years, what conditions must hold simultaneously? Are these conditions realistic? What expectations does today's stock price imply?2/10

  A

  Bottom line up front: Goldman rising fivefold in ten years (Baillie Gifford's standard threshold, about 17% annualized) is almost unrealistic—based on the report's three Owner Earnings scenarios, buying at the current price of about $1,001 implies a long-term annualized return of only about 6% (conservative) / 8% (neutral) / 10% (optimistic), far short of 17%. Achieving fivefold requires a string of highly optimistic conditions to hold simultaneously, whereas the current stock price has already largely priced in the "optimistic scenario," leaving almost no room for error.

  First, quantify "what fivefold in ten years requires." Fivefold means the market value (or per-share value) must reach about $5,000/share in ten years. Absent a large expansion of the valuation multiple, this requires intrinsic value per share to grow at about 17% annualized. But Goldman is a mature cyclical financial stock: the report records its historical ROE swinging violently between 7.5% (2023) and 23.0% (2021), 15.0% in 2025, with book value per share growing from $336.77 at the end of 2024 to $361.19 in Q1 2026—the endogenous growth of book value is on the order of mid-to-high single digits to the teens, not 17%. For fivefold, the following conditions must hold "simultaneously":

  First, through-the-cycle ROE stably staying in the high teens or even higher over the long term, without being knocked back to the low-to-mid double digits by cyclical troughs; second, AWM and financing-type durable revenue continuing to grow rapidly and significantly reducing earnings volatility, permanently raising the valuation multiple the market is willing to grant; third, regulatory capital requirements not deteriorating, or even loosening, allowing more profit to be returned or reinvested; fourth, capital markets maintaining high prosperity (active M&A, underwriting, and market-making) for many consecutive years without a deep pullback; fifth, management no longer drifting strategically and conducting buybacks when undervalued rather than chasing at highs. The report lists the first four almost verbatim as "key assumptions," and frankly states that "the current stock price already reflects the combined optimistic expectation of strong GBM + improving AWM + continued capital lightening + a friendlier regulatory environment, with almost no room for error."

  Are these conditions realistic? Each one alone is not absurd, but the probability of them "holding simultaneously for ten consecutive years" is very low—especially the first and fourth, which directly conflict with Goldman's cyclical nature. The report's bear case cuts to the point: "once trading heat, M&A activity, IPO windows, or private-asset business pulls back, ROE will quickly return to the low-to-mid double digits or even lower." Moreover, the current high ROE still contains tax-item water—the report notes that of the Q1 2026 annualized ROE of 19.8%, about 3.1 percentage points came from equity-award tax benefits, "clearly above a normalized return level," and cannot be extrapolated. A business whose headline figure does not even reach "normalized return" and which depends heavily on the external cycle achieving 17% compound value growth for ten consecutive years is unrealistic.

  What expectations does today's stock price imply? The report gives a clear answer by cross-checking three methods—the current price is essentially priced against the upper edge of the "optimistic scenario":

  • Owner Earnings discount model: conservative/neutral/optimistic intrinsic values of about $475 / $682 / $1,081, with the current price of about $1,001 "essentially close to the optimistic scenario and clearly above the neutral and conservative."
  • Relative valuation: based on Q1 2026 BVPS of $361.19, the current level is about 2.77x book and about 18.3x earnings (current price about $1,001); the report notes this is more expensive than JPM, BAC, and C, and only slightly below MS. Based on Q1 2026 tangible book value per share of $336.28, it is as high as about 2.98x tangible book.
  • Asset value method: the report argues that only if one believes it can steadily deliver ROE/ROTE near the high teens and keep optimizing capital intensity does 2.3–2.7x book become roughly reasonable (corresponding to about $830–975), while the current 2.77x is "already slightly above the upper edge of the optimistic asset-value range I am willing to give."

  On this basis, the report sets the required margin of safety at 20%–30%, the ideal buy price at $550–650, and above $950 as "clearly overvalued." The current level of about $1,001 falls squarely in the "clearly overvalued" range.

  Honest conclusion: Goldman rising fivefold in ten years is unrealistic, requiring multiple optimistic conditions to hold simultaneously and consecutively while its cyclical nature makes them hard to deliver; the current stock price already implies the optimistic expectation of "optimal operations + the friendliest market + continued capital lightening," and the buyer is betting that the company will stay near-optimal over the long term—far more than merely that "the company is decent." This is exactly the fundamental basis for the report's "Watch, insufficient margin of safety" rating.

  Jun 10, 2026
• QWhy hasn't the market realized all this yet? Does it not understand, look down on it, or fail to see far enough? What would become the "narrative inflection point"?3/10

  A

  Bottom line up front: this question must be answered in reverse for Goldman—the market has already fully, even excessively, recognized it, rather than "not yet realizing its value." Goldman is not a "dusty stock" that is misunderstood, looked down on, or neglected by the market; on the contrary, the market has given it a fairly generous premium (about 2.77x book, about 18.3x earnings, close to the upper edge of the report's optimistic scenario). So the real "narrative inflection point" is more likely the optimistic narrative being falsified and the valuation returning downward, rather than good news being discovered and the valuation repairing upward.

  First, let me clarify the current pricing to avoid applying the "the market doesn't get it" growth-stock template. Based on Q1 2026 book value per share of $361.19 and the current price of about $1,001, Goldman trades at about 2.77x book and about 18.3x earnings, and based on tangible book value per share of $336.28, at about 2.98x tangible book. After a cross-sectional comparison, the report notes that it is "more expensive than JPM, BAC, and C, and only slightly cheaper than MS." A large-cap financial stock that is thoroughly researched by the market, densely covered by the sell side, and highly liquid being given such a multiple shows that the market has not only understood its quality improvement (more durable revenue, principal-investment balance-sheet contraction, lower SCB, ROE back to 15%) but has largely priced in this good news. The report's judgment is direct: "you cannot assume GS is cheap just because its peers are not cheap," and "the premium the market currently gives it is already quite generous."

  So why can the stock stay high, and even keep being chased? This reflects "the optimistic scenario being extrapolated and a strong quarter being realized," rather than "undervaluation from short-sightedness." The report's bear case pinpoints the source of market sentiment: the high earnings of 2025 and Q1 2026 "benefited both from the recovery in capital markets and from tax and structural-adjustment factors"—of the Q1 2026 annualized ROE of 19.8%, about 3.1 percentage points came from equity-award tax benefits and were inherently non-extrapolable; investment-banking advisory fees jumped 89% year over year and the equities business hit a record, a cyclical-prosperity volume surge. The market linearly extrapolating these prosperous-period figures and assigning optimistic pricing is "paying for the excitement of a strong quarter," which is why the report marks "am I only wanting to buy because of the rising price or market sentiment" as "fail—it is currently easy to be lured by the strong rally" in its checklist. So the direction of the cognitive bias here is "excessive optimism," not "excessive pessimism."

  So what would the "narrative inflection point" be? Since an optimistic narrative currently supports the high valuation, the inflection point is the moment this narrative begins to be falsified—the report has in fact already laid out an inflection-point checklist under "what facts would overturn the investment judgment" and "signals that trigger reassessment," and reading them as downward narrative inflection points fits best:

  First, a downturn in capital-market prosperity—M&A activity, IPO windows, and market-making heat pulling back, with ROE returning from near 20% to the low-to-mid double digits or even lower, falsifying the "high returns are sustainable" narrative; second, AWM durable revenue, wealth-client assets, and net inflows stagnating for two or more years, falsifying the core bull logic of "revenue durability"; third, a clear decline in the wallet share of the investment-banking and markets businesses and in M&A/Equities/FICC rankings, falsifying the moat; fourth, standardized CET1 approaching the regulatory requirement for a long time while still buying back at high prices, or a major risk-control/litigation event raising the capital threshold, falsifying capital-allocation discipline; fifth, a renewed low-return, capital-intensive strategic drift, repeating the consumer-finance detour. If any signal materializes, the market could re-rate the valuation downward from about 2.77x book—the report even gives an extreme downside scenario: the valuation compressing to 1.2–1.5x book and the stock falling to $450–550, corresponding to roughly 45%–55% downside from the current level.

  Honest conclusion: Goldman's problem is that it is "a good company the market has already fully priced, even over-optimistically," rather than "a good company the market does not understand." It has no upward inflection point of "value buried, awaiting discovery"; the real inflection point is a downward re-rating after the cycle turns down and the durability or moat narrative is falsified. This is precisely the internal logic behind the report's final rating of "Watch, insufficient margin of safety at the current price, not worth buying at any price."

  Jun 10, 2026