The Number on the Check
On July 15, 2010, Goldman Sachs agreed to pay $550 million to the Securities and Exchange Commission — the largest penalty ever assessed against a financial services firm in the agency's history — and in doing so, acknowledged that its marketing materials for a synthetic collateralized debt obligation called ABACUS 2007-AC1 "contained incomplete information." The firm did not admit guilt. It did not deny it. It simply wrote the check, reformed certain business practices related to mortgage securities offerings, and moved on. The amount — split between $250 million returned to harmed investors and $300 million paid to the U.S. Treasury — represented roughly four days of Goldman's net revenue at the time. The markets barely flinched. Goldman's stock actually rose on the news of the settlement.
That detail — the stock going up on the day the firm paid the largest regulatory penalty in Wall Street history — tells you almost everything you need to know about Goldman Sachs. Not that it is invulnerable, though it sometimes appears so. Not that it is corrupt, though it has been accused of corruption in language that would make a medieval pamphleteer blush. What the detail reveals is something more precise: Goldman Sachs occupies a position in the architecture of global capital that is so deeply embedded, so structurally load-bearing, that the market's collective judgment of a half-billion-dollar fine was essentially relief. The penalty was finite. Goldman's position was not.
For more than 155 years — across financial panics, world wars, the dissolution of entire industries, the invention of entirely new ones, and at least three moments when the firm came within weeks of extinction — Goldman Sachs has operated as something closer to a utility of capitalism than a mere bank. It is the intermediary that sits at the center of the global capital system: between companies and markets, between governments and investors, between risk and the price of risk. It has been called "the most powerful investment bank on Wall Street," "Government Sachs," and, in Matt Taibbi's immortal Rolling Stone formulation, "a great vampire squid wrapped around the face of humanity, relentlessly jamming its blood funnel into anything that smells like money." The firm's own preferred self-description is more anodyne — "long-term greedy" — a phrase attributed to senior partner Gus Levy in the 1960s that has become Goldman's Rosetta Stone, the two-word operating principle that explains both the discipline and the appetite.
By the Numbers
Goldman Sachs at a Glance
$53.5BNet revenues, FY2024
$14.3BNet income, FY2024
~$200BMarket capitalization (early 2025)
~46,500Employees worldwide
$3.1TTotal assets under supervision
155+Years in continuous operation
#1Global M&A advisory (2024)
1M+Annual lateral job applicants
The Janitor's Assistant and the German Immigrant
The origin myth of Goldman Sachs contains two improbable figures whose biographies — separated by decades — define the firm's twin obsessions with capital innovation and human capital.
Marcus Goldman arrived in the United States from Bavaria in 1869, settled in New York City, and started a one-man commercial paper business operating out of a basement office at 30 Pine Street. Commercial paper — short-term promissory notes issued by businesses — was Wall Street's most pedestrian product, the financial equivalent of running a laundromat. Goldman would buy the notes from small merchants and jewelers in lower Manhattan, tuck them into the band of his top hat, and sell them to commercial banks at a slight markup. The margins were minuscule. The volume was everything. By 1882, he was handling $5 million in commercial paper annually, and he brought his son-in-law Samuel Sachs into the business. The name became Goldman, Sachs & Co.
What Marcus Goldman intuited — and what his son
Henry Goldman would later formalize into a revolution — was that the unglamorous work of connecting capital-seeking businesses with capital-providing institutions was itself the valuable franchise. You didn't need to be the capital. You needed to be the
connection. Henry Goldman, who joined the firm in the 1890s, took this insight and applied it to a product that would transform American finance: the initial public offering. In 1906, he underwrote the IPO of Sears, Roebuck and Company, using a then-novel approach of valuing the retailer based on its earnings rather than its physical assets — a methodology that allowed investors to price growth companies for the first time. It was, in the context of early-twentieth-century Wall Street, a genuinely radical act. Henry Goldman essentially invented the framework that would later value every technology company from IBM to Google.
The other figure is Sidney Weinberg, who joined Goldman Sachs in 1907 as a janitor's assistant at the age of 16. An eighth-grade dropout from Brooklyn, Weinberg's initial duties consisted of cleaning the office and brushing off the partners' boots. For years, he was known around the firm simply as "boy." But Weinberg possessed two qualities that Goldman would come to prize above almost all others: an extraordinary ability to listen, and an almost feral instinct for relationships. After serving in the Coast Guard during World War I, he returned to Goldman Sachs on the advice of Henry Goldman himself, who by then had been forced out of the partnership over a bitter dispute related to his pro-German sympathies. Henry told Weinberg to go back to Goldman Sachs. That was where the opportunity lay. Weinberg started selling commercial paper for $28 a week plus a 1.8% commission. By 1927, he was a partner. By 1930, he was senior partner — a position he would hold for nearly four decades, during which he became a trusted advisor to presidents from Roosevelt to Eisenhower, sat on the boards of dozens of major corporations, and built Goldman Sachs into the most prestigious advisory franchise on Wall Street.
For many years, the firm was constantly in and out of trouble. I would say its reputation for pristine excellence — the envy of Wall Street, if you will — has really been in and around since the 1980s.
The Goldman Sachs that Weinberg built was not a trading house. It was not a lending institution. It was a relationship machine — a firm whose primary product was the judgment, discretion, and Rolodex of its senior partners. Weinberg once came up with the name of a man he thought should be Treasury secretary while riding the subway. He told Eisenhower, who had never heard of the man. The president appointed him anyway. That story — apocryphal or not — captures something essential about Goldman's operating model during the Weinberg era: the firm's power derived not from its balance sheet but from its proximity to decision-makers and its willingness to deploy that proximity as a form of capital.
June Breton Fisher's
When Money Was in Fashion captures the Goldman family's role in creating this template, while William Cohan's
Money and Power provides the definitive account of how subsequent generations of partners weaponized it.
The Partnership and the Problem of Greed
The structure that made Goldman Sachs different from every other Wall Street firm for most of its history was neither a product nor a strategy. It was a legal entity: the partnership.
Until 1999, Goldman Sachs operated as a private partnership — meaning its capital came directly from the personal wealth of its partners, who bore unlimited liability for the firm's losses. This was not a quaint artifact. It was a governance mechanism of extraordinary power. When your own house is on the line, you monitor risk differently. You scrutinize the trader who wants to increase his position size. You ask the questions that salaried employees at publicly traded banks do not ask, because at a partnership, the downside is not a bad quarter. It is personal bankruptcy.
The partnership also created a ferocious meritocracy — or at least the perception of one. Being "made partner" at Goldman was one of the most coveted achievements in American business, roughly equivalent to making partner at a white-shoe law firm but with dramatically larger financial consequences. Partners received a share of the firm's annual profits, and in good years, those shares made partners extremely wealthy. But the partnership class was deliberately small — typically 200 to 300 people out of a workforce that grew to tens of thousands — and entry was governed by a biennial review process that combined rigorous performance evaluation with the politics of any institution where incumbents choose their successors.
The tension embedded in this structure was straightforward: the partnership model demanded long-term thinking (your capital was locked up; you couldn't sell your stake on the open market), but it also created enormous pressure to generate short-term profits (your compensation was a direct share of this year's earnings). Gus Levy, who ran the firm's trading operations in the 1960s and 1970s, captured the resolution of this tension in the phrase that became Goldman's motto: "long-term greedy." The idea was that Goldman would sacrifice immediate profits — declining a deal, walking away from a fee, even investing in a client's success at Goldman's own expense — in service of relationships that would compound over decades.
📜
Goldman Sachs Business Principles
Written in 1979 under John Whitehead, these 14 principles codified the firm's culture
1979John Whitehead drafts Goldman's 14 Business Principles, including "Our clients' interests always come first" and "We stress teamwork in everything we do."
1999Goldman Sachs goes public at $53 per share, raising approximately $3.66 billion. 221 partners share $6.4 billion in proceeds.
2006Lloyd Blankfein becomes CEO, accelerating the firm's shift toward proprietary trading and principal investments.
2008Goldman converts to a bank holding company on September 21, gaining access to Federal Reserve lending facilities and TARP funds.
2018David Solomon succeeds Blankfein as CEO, inheriting a firm grappling with post-crisis identity and Morgan Stanley's ascent.
"Long-term greedy" worked beautifully — so long as the people doing the greeding were also the people bearing the risk. The IPO of 1999 broke that linkage. When Goldman went public on May 4 of that year, at $53 per share, the 221 existing partners collectively received approximately $6.4 billion for their stakes. It was, for them, the greatest single liquidity event in Wall Street history. But the transaction also fundamentally altered the firm's incentive structure. Partners were now employees with stock options. The capital at risk was no longer personal; it belonged to public shareholders. The brake on excessive risk-taking — the knowledge that a catastrophic loss would wipe out your own fortune — was replaced by a different mechanism: the stock price, quarterly earnings expectations, and the regulatory apparatus of a publicly traded bank holding company.
Whether the IPO ultimately weakened Goldman's culture is one of the great counterfactual debates of modern finance. What is observable is that the decade following the IPO was the period of Goldman's most aggressive expansion into proprietary trading, its deepest entanglement with mortgage-backed securities, and its closest approach to the abyss.
The Vampire Squid Earns Its Nickname
The financial crisis of 2008 nearly destroyed Goldman Sachs — though the firm's own narrative, burnished in the years since, tends to minimize just how close the edge was.
In September 2008, with Lehman Brothers already in bankruptcy and Merrill Lynch acquired by Bank of America in a shotgun deal, Goldman and Morgan Stanley were the last two independent investment banks standing. On September 21, 2008 — a Sunday evening, because existential decisions on Wall Street always seem to happen on Sundays — Goldman converted its legal status from a securities firm to a bank holding company, a move that gave it access to the Federal Reserve's discount window and the stability of a more regulated structure. The conversion was not voluntary in any meaningful sense. It was survival.
Within weeks, Goldman received a $10 billion injection from the U.S. Treasury through the Troubled Asset Relief Program (TARP), alongside a separate $5 billion investment from
Warren Buffett's Berkshire Hathaway — an investment structured with terms so favorable to Buffett (10% annual dividend on preferred stock, plus warrants to buy Goldman common stock at $115 per share) that it functioned simultaneously as a vote of confidence and a reminder that Goldman's negotiating leverage had evaporated.
Goldman repaid the TARP funds in June 2009, eager to escape the compensation restrictions and reputational taint that came with government money. But the deeper damage was already done. The SEC's fraud charges related to ABACUS — the synthetic CDO structured at the behest of John Paulson's hedge fund, which bet against the very mortgage securities that Goldman was marketing to other investors — became the defining scandal of the crisis era. The marketing materials, the SEC alleged, stated that the portfolio was "selected by" ACA Management, a third party, without disclosing Paulson's role in choosing the underlying securities or his short position against them.
This settlement is a stark lesson to Wall Street firms that no product is too complex, and no investor too sophisticated, to avoid a heavy price if a firm violates the fundamental principles of honest treatment and fair dealing.
The ABACUS settlement — $550 million, without admitting or denying guilt — was a rounding error on Goldman's balance sheet. But the reputational cost was immense. Matt Taibbi's "vampire squid" epithet, published in Rolling Stone in July 2009, entered the permanent lexicon. Goldman's own CEO didn't help matters; Lloyd Blankfein's remark to a journalist that the firm was "doing God's work" — intended, reportedly, as a joke — became the most quoted line of the entire financial crisis. Even Warren Buffett, who had invested $5 billion in the firm, captured the prevailing mood with characteristic dryness: "They're going to rewrite Genesis and have Goldman Sachs offering the apple."
By 2010, Goldman faced a challenge it had never encountered in its history: the need to explain itself to ordinary people. The firm launched its first-ever national advertising campaign — full-page ads in newspapers bearing the tagline "Progress is Everyone's Business," featuring photographs of wind turbines and smiling construction workers. The ads were managed by Young & Rubicam and received uniformly lukewarm reviews. Marketing professors noted that the Goldman logo occupied barely one square inch of a full-page ad. "If I just saw the ad, I might think it was for a utility company," observed Russell Winer, chairman of NYU Stern's marketing department.
We have a lot of work to do explaining to people what it is that we do. And we're starting from a hole.
The irony was total. Goldman Sachs — the firm whose power derived from operating in the shadows, from the quiet cultivation of relationships with corporate chieftains and heads of state, from being the firm that didn't need to advertise because every Fortune 500 CEO already had the number — was now buying banner ads on websites, trying to convince the general public that it was a force for good. The vampire squid was learning public relations.
Government Sachs and the Revolving Door
No institution in American history has produced more government officials per square foot of office space than Goldman Sachs. The list is so extensive it reads less like an alumni network and more like a shadow cabinet that never quite goes home.
Robert Rubin, co-senior partner, became Treasury Secretary under Clinton. Hank Paulson, CEO, became Treasury Secretary under George W. Bush — and in that role, presided over the bailout of the financial system that included Goldman's own rescue. Steve Mnuchin, a Goldman partner for 17 years, became Treasury Secretary under Trump. Mark Carney, who spent 13 years at Goldman, became Governor of the Bank of England and then the Bank of Canada. Steve Bannon, who worked at Goldman in the 1980s, became Trump's chief strategist. Gary Cohn, Goldman's president, became director of the National Economic Council. Mario Draghi, who spent time at Goldman Sachs International, became president of the European Central Bank and then Prime Minister of Italy.
The pattern is so consistent — and so bipartisan — that "Government Sachs" functions not as an insult but as a description of a structural feature of the American political economy. Goldman's relationship with government is symbiotic in the most precise sense: each organism benefits from the other's existence. The firm gains access to policy formation, regulatory intelligence, and a network of former colleagues in positions of power. The government gains access to people who actually understand how capital markets work — a surprisingly scarce commodity in Washington.
Henry Goldman himself was consulted by the Cabinet members who created the Federal Reserve System in 1913, and they mostly followed his lead. Sidney Weinberg advised every president from FDR to Eisenhower. The pattern is not corruption, exactly — though critics have made that case with some force. It is something more interesting: an institutional DNA that treats the boundary between public and private power as a permeable membrane, and views the cultivation of governmental relationships as a core competency on par with securities underwriting or risk management.
The revolving door also creates a specific form of institutional knowledge. Goldman alumni in government understand how Goldman thinks, which means they understand how Wall Street thinks, which means they can design policy with — or without — that knowledge as a guide. The question of whether this produces better or worse policy outcomes is genuinely unresolvable. What is clear is that it produces a firm whose tentacles, as the BBC's Simon Jack noted, "are still very long."
The DJ CEO and the Consumer Experiment
David Solomon became CEO of Goldman Sachs in October 2018, succeeding Lloyd Blankfein after a succession process that was — by Goldman standards — remarkably public and somewhat brutal. Solomon, a leveraged finance banker who had spent most of his career in the firm's investment banking division, brought a different sensibility than the trading-floor warriors who had typically led the firm. He was, famously, an amateur DJ who performed at nightclubs and music festivals under the name "D-Sol," releasing dance remixes on his own label, Payback Records, through a partnership with Atlantic Records. His Spotify profile shows 269,388 monthly listeners — a metric that most Goldman partners find either endearing or mortifying, depending on their vintage.
The DJ hobby was a sideshow, but it revealed something real about Solomon's strategic instincts: he wanted Goldman to be visible in ways it had never been before. Under his leadership, the firm dramatically expanded its marketing presence, investing in brand-building efforts that would have been unthinkable under Blankfein or his predecessors. Goldman launched "Talks at GS," a podcast and video series featuring interviews with CEOs, founders, and cultural figures. The firm's YouTube channels proliferated. Solomon himself became a regular on the Davos circuit, making pronouncements about diversity and corporate responsibility that positioned Goldman as a progressive institution rather than a shadowy intermediary.
But the signature strategic bet of the Solomon era was not marketing. It was Marcus.
Launched in 2016 — before Solomon became CEO, but elevated to a strategic priority under his leadership — Marcus by Goldman Sachs was the firm's first foray into retail consumer banking. Named after founder Marcus Goldman, the platform offered high-yield savings accounts and unsecured personal loans, targeting the mass affluent consumer segment that Goldman had historically ignored in favor of institutions and the ultra-wealthy. The logic was seductive: Goldman's core businesses — investment banking, trading, asset management — were cyclical and capital-intensive. Consumer banking offered a potential source of stable, recurring deposits and fee income that could smooth the earnings volatility that Wall Street analysts perpetually punished.
In 2019, Goldman partnered with Apple to launch the Apple Card, a consumer credit card that represented the firm's most visible consumer product. The partnership gave Goldman immediate access to Apple's enormous customer base, but it also dragged the firm into a business — consumer credit card servicing — that was operationally alien to everything Goldman had ever done. The firm had no experience with customer service call centers, consumer compliance requirements, or the granular unit economics of credit card portfolios.
The Marcus experiment would eventually become Solomon's most painful lesson. By 2022, the consumer business had accumulated losses reportedly exceeding $3 billion, and Goldman was actively seeking to exit the Apple Card partnership. The firm's retail ambitions, launched with considerable fanfare, were being quietly dismantled. Solomon reorganized the firm's business segments in late 2022, folding consumer operations into a broader "Platform Solutions" segment that de-emphasized the retail strategy. The Wall Street consensus was brutal: Goldman had attempted to become a consumer bank and had discovered, expensively, that the skills that make you the world's best M&A advisor have almost no transferability to the business of issuing credit cards to people who shop at Target.
The retreat was a humbling moment for a firm that prided itself on strategic omniscience. But it was also, in a perverse way, a validation of Goldman's deeper culture: the firm recognized the mistake, absorbed the losses, and returned to its core franchise with a speed and discipline that more bureaucratic institutions could not have matched.
The Talent Machine and the Culture of Up-or-Out
Goldman Sachs receives more than one million lateral job applications per year. It accommodates far fewer than 1% of them. The selectivity is not a boast but a structural feature: Goldman's business model depends on human capital in a way that even other professional services firms do not quite match. A manufacturing company can survive a bad hire on the assembly line. A technology company can ship mediocre code and patch it later. When a Goldman banker gives bad advice on a $50 billion merger, the consequences are measured in reputational damage that compounds over decades.
We continue to see incredible demands for people who want to come and work at Goldman Sachs, more than a million people asking to move in laterally to the firm. We can accommodate far less than 1%, so we're still in a position to be extremely selective on the people that we hire.
The firm's hiring philosophy underwent a significant transformation in the years following the financial crisis, as Dane Holmes, Goldman's then-head of human capital management, detailed in a 2019 Harvard Business Review article. The problem was straightforward: the crisis had "taken some of the sheen off" investment banking as a career, and simultaneously, the battle for talent had intensified as candidates headed to Silicon Valley, private equity, and startups. Goldman was no longer principally looking for accounting and finance majors — "new skills, especially coding, were in huge demand at Goldman Sachs and pretty much everywhere else." The wind, as Holmes put it, had shifted from their backs to their faces.
Goldman's response was to fundamentally rethink its recruiting pipeline. The firm expanded its campus recruiting beyond its traditional Ivy League feeder schools, invested in video interviewing technology, and began actively recruiting computer science and engineering talent in direct competition with Google and Facebook. The annual performance review process — which typically targets the lowest 3% to 5% of performers for termination — remained intact, ensuring that the culture of up-or-out continued to function as both a motivational tool and a filtration system.
The partnership tradition, even in a publicly traded company, still structures Goldman's internal status hierarchy. "Making partner" remains the firm's most powerful incentive, conferring not just compensation but a kind of institutional identity that alumni carry for the rest of their careers. Whether they go to government, start hedge funds, or run the World Bank, former Goldman partners are always introduced as former Goldman partners first. The brand is a credential that depreciates less than any degree, any title, any other line on the résumé.
The Machine Beneath the Machine
If the public narrative of Goldman Sachs is about relationships, prestige, and political power, the financial reality is more prosaic and more interesting. Goldman is, at its core, a risk intermediation machine — a complex system that earns money by standing between parties who have different views on the price of risk and collecting a toll for the service.
The firm operates through four main business segments, though the nomenclature has shifted repeatedly as successive CEOs reorganize the deck chairs. The essential architecture is this:
Investment Banking generates fees from advising corporations on mergers and acquisitions, underwriting equity and debt offerings, and providing strategic counsel. This is the business that gives Goldman its prestige and its political connections — but it is also intensely cyclical and represents a minority of the firm's total revenue.
Global Markets — the trading operation — is where the bulk of Goldman's revenue and risk live. The division makes markets in equities, fixed income, currencies, and commodities, serving institutional clients who need to buy or sell securities. The revenue comes from bid-ask spreads, commissions, and — crucially — the firm's willingness to commit its own balance sheet to facilitate client trades. This is not, technically, the same as the proprietary trading that Goldman engaged in before the Volcker Rule restricted such activities after the crisis. But the boundary between "market-making" and "prop trading" is famously blurry, and Goldman's skill at navigating that boundary has been both its greatest source of profit and its greatest source of controversy.
Asset and Wealth Management oversees more than $3 trillion in assets under supervision, managing money for institutions, sovereign wealth funds, pension funds, and ultra-high-net-worth individuals. Under Solomon's strategic refocusing, this division has become the firm's most important growth engine — a source of recurring management fees that are less volatile than trading revenue and less cyclical than banking fees.
Platform Solutions houses the remnants of the consumer banking experiment, including the Apple Card partnership and Goldman's transaction banking business. It is the segment that Solomon would prefer analysts not ask about.
The interplay between these segments creates Goldman's distinctive economic profile: high return on equity in good years (often above 15%), punishing cyclicality in bad ones, and a constant strategic debate about where to allocate incremental capital.
The War with Morgan Stanley
For decades, the hierarchy of Wall Street was clear: Goldman Sachs was the most prestigious investment bank, and everyone else was chasing it. Morgan Stanley, Merrill Lynch, Lehman Brothers, Bear Stearns — they were all formidable competitors, but Goldman occupied the summit.
The financial crisis shattered that pecking order. Lehman and Bear disappeared. Merrill was absorbed by Bank of America. And Morgan Stanley, under CEO James Gorman, began executing a strategic pivot that would eventually challenge Goldman's supremacy in ways the firm had not anticipated.
Morgan Stanley's bet was wealth management — specifically, its 2009 joint venture with Citigroup's Smith Barney brokerage, which it eventually acquired outright for $13.5 billion. The deal gave Morgan Stanley a massive retail distribution platform and a source of stable, fee-based revenue that Goldman lacked. While Goldman was still generating a disproportionate share of its earnings from the volatile trading floor, Morgan Stanley was building a recurring-revenue franchise that Wall Street's public-market investors — who above all prize predictability — valued at a premium.
The results were visible in a metric that Goldman had traditionally dominated: market capitalization. In 2020, Morgan Stanley's market cap surpassed Goldman's for the first time in history — a symbolic moment that signaled a genuine shift in how the market valued the two franchises. Morgan Stanley also made inroads in equities trading, at one point overtaking Goldman in that business, which Goldman had long considered its crown jewel.
Goldman's response under Solomon has been to pivot aggressively toward asset and wealth management — essentially playing catch-up on a strategy that Morgan Stanley implemented a decade earlier. The firm has been raising third-party capital for its alternatives business, growing its wealth management platform for ultra-high-net-worth clients, and attempting to build the kind of stable, fee-generating franchise that investors will reward with a higher multiple.
Whether Goldman can close the gap is the central strategic question of the Solomon era. The firm retains its dominance in M&A advisory — it ranked #1 globally in 2024 — and its trading operation remains among the most profitable on Wall Street. But the market is asking a question Goldman has never had to answer before: Is the world's most prestigious investment bank actually the best-structured business?
OneGS 3.0 and the AI Wager
In late 2025, Goldman Sachs announced OneGS 3.0, a multiyear initiative to integrate artificial intelligence throughout the bank's operating model. The program, described by CFO Denis Coleman as "a fundamental rethinking of how we expect our people to operate at Goldman Sachs," represents the firm's most ambitious operational transformation since the post-crisis restructuring.
We're asking all of our people to rethink the human processes they go through. And then we're making investments in AI and agentic AI to accelerate change across these processes and platforms.
The initiative identifies six discrete workstreams across every division and function — from business lines to control functions to engineering — with dedicated teams tasked with reviewing key activities, analyzing pain points, and presenting formal investment cases for leadership review. "We'll fund some of those investments and hold teams accountable for the productivity outcomes that follow," Coleman said. "We don't want to simply add more manual processes to drive growth. We need to convert some of that effort into digitized and automated systems — and rethink how those engines work."
Goldman's approach to AI is characteristically top-down and systematic — the antithesis of the Silicon Valley model where engineers build tools and hope the organization adopts them. At Goldman, the emphasis is on "the quality, availability, accuracy, and timeliness of data," and the firm is investing in shared platforms that span the organization rather than allowing individual teams to build bespoke solutions. The investment is also flowing outward: Goldman's growth equity group led a $75 million funding round for Fieldguide, an AI-native accounting platform, in early 2026, signaling that the firm sees AI transformation as both an internal priority and an external investment theme.
The AI bet matters for Goldman more than for most firms because the bank's core product — the judgment and execution of highly paid professionals — is precisely the kind of knowledge work that large language models threaten to commoditize. If an AI agent can draft the first version of a fairness opinion, generate the preliminary analysis for an M&A pitch, or execute the routine components of a derivatives trade, then the question becomes: what is the Goldman premium actually paying for? The answer, presumably, is the same thing it has always been: relationships, discretion, and access to decision-makers at the highest levels of business and government. Those are harder to automate. But the margin between Goldman's execution and a competitor armed with the same AI tools narrows with every improvement in model capability.
The Superego of Capital
Goldman Sachs produced a 150-minute documentary series to celebrate its 150th anniversary. The film, directed by Ric Burns and available on Amazon Prime, cost what was reportedly an eight-figure sum — paid for entirely by the firm itself. As Vanity Fair's William Cohan observed, watching it was like visiting Lake Wobegon on the Hudson: "all the women are strong, all the men are good-looking, and all the children are above average."
The documentary's existence reveals something essential about Goldman that transcends any individual strategic decision or product line. Goldman Sachs has a superego. It is not simply in the business of making money — though it does that with formidable efficiency — but in the business of believing that making money, the Goldman way, is itself a form of civic contribution. The firm's brief, as Cohan wrote, "is about being able to make money while also providing capital and advice to corporations, governments, institutions, and wealthy people who need it and can pay a fair price for getting the best."
This self-regard is not unique to Goldman — every institution has a creation myth — but at Goldman, it is unusually persistent, unusually coherent, and unusually integrated into the firm's operating culture. The 14 Business Principles drafted by John Whitehead in 1979, which open with "Our clients' interests always come first," are not just words on a wall. They function as a coordination mechanism, a shared language that allows 46,000 people spread across dozens of countries to make decisions that are roughly consistent with what the firm's senior leadership would want. Culture, at Goldman, is not a perk. It is the product.
The tension — and it is a genuine tension, not a rhetorical one — is that the same self-regard that produces disciplined client service also produces a capacity for self-delusion. Goldman's partners genuinely believe they are providing an essential service to the global economy. They are, in many cases, correct. But the belief also creates blind spots — the conviction that Goldman's interests and the market's interests are naturally aligned, that what is good for Goldman is, in some structural sense, good for capitalism. The ABACUS scandal was, at its core, a failure of that assumption: what was good for Paulson & Co. and what was good for Goldman's fee income was not, it turned out, good for the investors on the other side of the trade.
Goldman's CEO David Solomon captured the firm's current strategic posture at Davos in January 2026, where he described American CEOs as "bullish but nervous" — a phrase that could serve as Goldman's own epitaph if things go wrong, or its motto if they go right. The firm is betting that its traditional strengths — advisory prestige, trading acumen, an unmatched alumni network — can be augmented by new capabilities in asset management, technology, and AI to create a more durable and less volatile franchise. The market, as of early 2026, is cautiously willing to pay for that story, with Goldman's stock trading near all-time highs and its market capitalization approaching $200 billion.
But Goldman has always been a bet on a specific theory of how capital markets should work — a theory in which the smartest, most connected, most disciplined intermediary captures a disproportionate share of the value created by the movement of capital. Whether that theory holds in a world of AI-powered analysis, zero-cost trading, and democratized financial information is the question that the next chapter of Goldman's story will answer.
The Number That Explains Everything
There is a number that Goldman Sachs tracks internally with the devotion that a cardiologist reserves for a patient's resting heart rate. It is not revenue, or profit, or even return on equity — though all of those matter. It is the compensation ratio: the percentage of net revenues that the firm pays out to its employees.
For most of Goldman's modern history, this number has hovered around 35% to 45%, depending on the year and the competitive environment. The number encodes Goldman's entire business philosophy. It says: we are a people business. Our product is judgment. Our competitive advantage is talent. And we will pay whatever is necessary — but not a dollar more — to attract and retain the best people in the world.
When the compensation ratio rises, it means Goldman is investing in human capital at the expense of shareholder returns. When it falls, it means the firm is either finding operational efficiencies or squeezing its workforce. The OneGS 3.0 initiative, with its emphasis on AI and automation, is explicitly designed to push this ratio down — to generate more revenue per dollar of compensation by replacing routine human processes with digital ones. "We don't want to simply add more manual processes to drive growth," Coleman said. The subtext is unmistakable: we want to grow the numerator without growing the denominator.
Whether Goldman Sachs can pull this off — whether it can remain a people business while simultaneously reducing its dependence on people — may be the defining strategic question of the next decade. It is a question that every professional services firm, every knowledge-work institution, every organization that sells human judgment for a premium, will eventually have to answer. Goldman, characteristically, intends to answer it first.
On a recent earnings call, Solomon noted that 2025 was shaping up to be the second-biggest year in history for announced mergers and acquisitions industrywide. Goldman's economists expected a 25-basis-point rate cut from the Fed, followed by a pause, then possibly two more cuts. The M&A pipeline was full. The trading floor was humming. The firm was still, after 155 years, doing what Marcus Goldman did with his top hat on Pine Street: standing between those who had capital and those who needed it, and taking a cut.
The compensation ratio for the most recent quarter was not publicly disclosed at the time of this writing. But somewhere in 200 West Street, someone was watching it.
