FICTION · An Unfair Advantage Intervention
Fiction. Real history. Sourced below.
Goldman Sachs was built on a reputation for honest dealing, and that reputation has been publicly broken at least three times — 1929, 2010, and 2020. The firm survived all three. Not because trust is an unbreakable moat, but because trust was never Goldman’s only moat. The proof is in the 2020 settlement, where a firm with its reputation in ruins still negotiated an outcome that preserved its licenses. That was scale talking, not character.
I’m Samantha Reynolds, and I build incentive systems, which is why this case interests me more than most people’s version of it does. The founding story is real and worth honoring. But the question everybody asks — how did a firm built on trust end up paying billions over bribery — is less interesting than the one nobody asks: why is it still here? Run it through Compounding Defense and both answers arrive together, and they are not flattering to anyone who thinks reputation alone will protect them.
Full fiction disclosure and sourcing note
This is a work of fiction. Samantha Reynolds is a fictional character from The Unfair Advantage by Todd Hagopian, and Cartwell is the book’s fictional company. Samantha analyzes the real, documented history of Goldman Sachs from the present day; she never met, advised, or spoke with any real person, and no dialogue is invented for anyone. This piece discusses the firm and its subsidiaries at an institutional level and describes only conduct admitted in settlement documents or established in public filings; individual criminal matters are not the subject here. The historical facts are accurate and sourced as of 2026; the framework analysis is invented.
- The Basement and the Silk Hat
- Compounding Defense: Why Three Breaches Didn’t Kill It
- Four Moats, Not One
- The Deferred Prosecution Agreement Is the Proof
- Why the Process Didn’t Stop It
- What Actually Happened
- From the Operator’s Chair
- The Monday Morning Playbook
- The Contrarian Reappraisal
- The Goldman Sachs Timeline
- By the Numbers
- Frequently Asked Questions
- About the Stagnation Assassin
The Basement and the Silk Hat
The founding is as good as the legend says. In 1869 Marcus Goldman, a Bavarian immigrant who had worked as a peddler, shopkeeper, and tailor, opened a one-room basement office at 30 Pine Street in Lower Manhattan and went into commercial paper: buying merchants’ short-term IOUs at a discount and reselling them to investors. By legend he carried the notes in the band of his silk hat as he walked between the merchants and the investors.
It filled a real gap. Post-Civil-War bank credit was tight and expensive, and small merchants needed liquidity the formal system underserved. And it ran entirely on trust: investors had to believe Goldman had selected creditworthy merchants, and he reportedly stood behind the paper personally to earn that belief. One carelessly sold default could have ended the business.
He scaled it three ways, and all three are genuinely transferable. Referral networks, where a trusted introduction carries the referrer’s credibility so a new relationship starts from somewhere better than zero. Carefully chosen partners — his son-in-law Samuel Sachs joined in 1882, later his son Henry — extending the firm through people selected as rigorously as the merchants whose paper he bought. And codified judgment, turning one man’s evaluation instincts into standards others could learn.
All of that is true and none of it is the interesting part, because those three mechanisms built one moat. The firm’s survival depended on three others it acquired later.
Compounding Defense: Why Three Breaches Didn’t Kill It
Compounding Defense holds that single-moat positions are vulnerable and multi-moat positions become structurally unattackable, because moats compound when types stack: when one is breached, the others carry the position. Goldman is the cleanest available demonstration, and it demonstrates the principle from the wrong side.
The reputation moat has been publicly breached three times. In 1928 the firm launched the leveraged Goldman Sachs Trading Corporation; it collapsed in the 1929 crash, inflicting heavy losses and damaging the firm’s standing for a generation. In 2010 Goldman paid $550 million — then a record SEC penalty against a Wall Street firm — to settle charges that it misled investors in a subprime mortgage product. In 2020 it resolved the 1MDB matter for more than $2.9 billion.
Three breaches of the founding moat, roughly forty and then ten years apart. Under a single-moat model, any one should have been terminal — that is what “reputation is everything” means in a relationship business. None of them were.
The explanation is not that the reputation moat is unusually resilient. It is that by 1929 Goldman was no longer a single-moat firm. It had accumulated scale, network position, and regulatory standing, and those three carried the weight each time the brand failed. That is Compounding Defense working — and it is why the founding legend is not merely incomplete but actively misleading. The story credits the survival to the trust. The survival was purchased by everything else.
Four Moats, Not One
The Moat Taxonomy names six types — cost, brand, switching cost, network, scale, and regulatory — and the discipline is identifying which ones you actually hold rather than which one you talk about. Goldman’s public story is entirely about brand. Its actual defensive position runs across four.
Brand is the founding moat and the one in every retelling: the honest broker at 30 Pine Street. It is real, and it is the only one of the four that has demonstrably failed.
Scale arrived over the following century — the balance sheet capacity to absorb a $550 million penalty in 2010 and billions more in 2020 without existential strain. A firm of Marcus Goldman’s size could not have written either check.
Network is the embeddedness: the firm sits inside the plumbing of global capital markets, counterparty to institutions that cannot easily route around it. Network moats are the hardest to attack because attacking them damages the attacker.
Regulatory is the licenses, the charters, and the systemic-importance status that makes the authorities’ calculus about consequences rather than punishment alone.
Read that stack and the puzzle dissolves. The brand moat failed three times and the firm held, because it was never the load-bearing one after about 1900. Which produces the uncomfortable inversion at the center of this case: for Goldman, betraying the trust was survivable. For almost any reader of this article, it would not be — because almost nobody else holds the other three.
The Deferred Prosecution Agreement Is the Proof
If you want to watch a compounded moat do its work in real time, look at what the 2020 settlement actually was, because the popular summary gets it wrong in a way that erases the whole mechanism.
Goldman Sachs Group Inc. did not plead guilty. According to the firm’s own 8-K filing, the parent entered a three-year deferred prosecution agreement under which one count of conspiracy to violate the FCPA would be filed and later dismissed if the firm abided by the terms. Only the Malaysian subsidiary, Goldman Sachs (Malaysia) Sdn. Bhd., pleaded guilty.
That distinction is the entire argument. Contemporary reporting noted that negotiations dragged on as the bank sought to avoid pleading guilty, and that the DPA should allow it to avoid exiting certain business operations. A parent-level conviction can jeopardize the licenses an investment bank runs on. Goldman avoided one.
Now ask what purchased that outcome. It was not the reputation — the reputation was the thing on trial. It was the regulatory and systemic-importance moat: the fact that the consequences of a parent-level guilty plea fall on more parties than the defendant. The moat you never mention in your founding story is the one that saved you.
Why the Process Didn’t Stop It
Here is the part that is my actual job, and the part the founding legend most needs correcting on. “Codify your judgment so it scales” is genuinely one of Marcus Goldman’s three scaling mechanisms, and it is true only under a condition nobody states: that incentives are aligned to want the judgment preserved.
The 1MDB transactions were not unvetted. Goldman’s compliance and legal architecture in that era was sophisticated, the bonds passed internal review, and senior people signed off. Prosecutors described bank staff who enabled the wrongdoing by disregarding signs that something was amiss. The controls existed and were defeated.
Look at the economics and you can see why. Goldman arranged three bond offerings raising approximately $6.5 billion for 1MDB and earned about $600 million in fees. Against a payout of that size concentrated on a small number of people, a control function staffed by people whose compensation does not depend on the deal closing — but whose careers are shaped by an institution that rewards closing — is not a real check. It is documentation.
The firm’s own subsequent action concedes the point. The board applied clawbacks, forfeitures, and compensation reductions to past and current senior management, including executives who were not involved in or even aware of the conduct. You do not adjust the compensation of uninvolved executives to punish individuals. You do it because you have concluded the incentive system, not merely the people, produced the outcome. That is the correct diagnosis, and it arrived roughly eight years late.
What Actually Happened
Marcus Goldman founded the firm in 1869, trading commercial paper from a one-room basement office at 30 Pine Street. Samuel Sachs joined as a partner in 1882 and the firm became Goldman, Sachs & Co.; it joined the New York Stock Exchange in 1896 and took Sears, Roebuck public in 1906. Goldman died in 1904.
The Goldman Sachs Trading Corporation, launched in 1928, collapsed in the 1929 crash. In 2010 the firm paid $550 million to settle SEC charges over a subprime mortgage product, then a record penalty against a Wall Street firm.
In 2020 Goldman resolved the 1MDB matter. The firm arranged three bond offerings raising approximately $6.5 billion for the Malaysian sovereign wealth fund and earned around $600 million in fees; roughly $2.7 billion was subsequently misappropriated, and settlement documents describe over $1.6 billion in bribes promised and paid to officials in Malaysia and Abu Dhabi to win the business. The Malaysian subsidiary pleaded guilty and Goldman agreed to pay about $2.9 billion to various authorities, including the largest-ever penalty under the Foreign Corrupt Practices Act. The parent entered a deferred prosecution agreement rather than pleading guilty. Separately, the firm had settled with the Malaysian government for $3.9 billion — $2.5 billion in cash plus a $1.4 billion asset-recovery guarantee — and Hong Kong imposed a record $350 million fine largely outside the $2.9 billion total.
The firm remains one of the most powerful financial institutions in the world, overseeing trillions in assets.
From the Operator’s Chair
I’m Todd Hagopian, and reputation as an asset, moats as a portfolio, and the gap between process and integrity are things I’ve worked with across Berkshire Hathaway, Illinois Tool Works, Whirlpool, and now as VP of Product Strategy and Marketing. Three disciplines out of this one.
First, count your moats honestly, and notice which one you talk about. Most companies have a founding story built on one moat — quality, service, integrity, craftsmanship — and a real defensive position that either includes others or doesn’t. Goldman talks about brand and survives on scale, network, and regulatory standing. If you audited your own business and found that your only genuine moat is the one in your founding story, you are single-moat dependent, and you should treat that as the strategic finding it is rather than a point of pride.
Second, do not read Goldman’s survival as permission. The honest version of this case is not “reputation can be spent and you’ll be fine.” It is “Goldman could spend it because Goldman had three other moats, and you have none of them.” A mid-market industrial company that betrays a major customer’s trust does not get a deferred prosecution agreement and a return to business as usual. It loses the account, then the referrals, then the segment. For almost every operator reading this, reputation genuinely is close to the only moat — which makes it more fragile in your business than it is in Goldman’s, not less.
Third, fix the incentive before you trust the control. This is the one I’d put on the wall. A documented standard is only as strong as the organization’s willingness to enforce it against its own profit motive. If your compensation rewards volume and your process rewards caution, the compensation wins — every time, in every company, regardless of how good the process document is. When I want to know whether a control will hold, I don’t read the control. I read the comp plan of the person who can override it.
The Monday Morning Playbook
- Audit your moats against all six types. Cost, brand, switching cost, network, scale, regulatory. Write down which ones you genuinely hold. If the honest answer is one, you are single-moat dependent and a single breach is terminal.
- Notice the gap between your story and your defense. The moat in your founding legend and the moat actually carrying your position are often different. The one you never mention is usually the one doing the work.
- Read the comp plan, not the control document. Before trusting any process to protect judgment, find out how the person who can override it gets paid. If the incentive points at yes, the control is documentation.
- Name your fork moments in advance. The deals lucrative enough to make a quiet betrayal look worth it are the ones that define your reputation. Decide how you’ll act before the money is on the table, because you will not decide well with it there.
- Don’t confuse survival with vindication. Enduring a breach is not the same as honoring the trust, and an organization that learns it can spend reputation profitably has begun spending its foundation.
The Contrarian Reappraisal
The consensus reading — Marcus Goldman built Goldman Sachs on trust, proving honest dealing creates a permanent moat — is half a real lesson and half a comforting myth. Three corrections.
The reputation moat broke, repeatedly. The 1929 Trading Corporation collapse, the $550 million subprime settlement in 2010, and the 1MDB resolution in 2020 are three public breaches of the founding promise. Trust is a real moat and a breakable one.
Survival came from the other moats, not the broken one. Goldman held scale, network position, and regulatory standing alongside brand, and those three carried the firm each time the brand failed. The 2020 deferred prosecution agreement is the visible proof: with its reputation in ruins, the parent still avoided a guilty plea and preserved its licenses. Scale bought that, not character.
Which makes this a warning rather than a template. The dangerous misreading is “reputation can be spent and survived.” It could be, by an institution holding three other moats. Most businesses hold one. If reputation is your only moat, you are the single-moat case the framework warns about, and Goldman’s history tells you nothing reassuring about your own.
The Goldman Sachs Timeline
- 1869 — Marcus Goldman opens a commercial paper business in a one-room basement office at 30 Pine Street, built on his reputation as an honest broker.
- 1882–1885 — son-in-law Samuel Sachs joins as a partner; the firm becomes Goldman, Sachs & Co., growing through carefully chosen partners.
- 1896–1906 — the firm joins the New York Stock Exchange and enters the IPO market by taking Sears, Roebuck public; Marcus Goldman dies in 1904.
- 1928–1929 — the leveraged Goldman Sachs Trading Corporation collapses in the crash, damaging the firm’s reputation for a generation. The first breach; the firm survives.
- 2010 — Goldman pays $550 million to settle SEC charges over a subprime mortgage product, then a record penalty against a Wall Street firm. The second breach; the firm survives.
- 2012–2013 — the firm arranges three bond offerings raising approximately $6.5 billion for 1MDB, earning about $600 million in fees.
- July–October 2020 — a $3.9 billion settlement with Malaysia is followed by a resolution of more than $2.9 billion with U.S. and other authorities; the Malaysian subsidiary pleads guilty while the parent enters a deferred prosecution agreement.
By the Numbers
- 4 moats, 1 broken — brand, scale, network, and regulatory; only the brand moat has publicly failed, and it has failed three times.
- $6.5 billion raised, ~$600 million in fees — the three 1MDB bond offerings and what Goldman earned arranging them.
- $2.9 billion+ — the 2020 resolution with U.S. and other authorities, including the largest-ever FCPA penalty, plus a separate $3.9 billion Malaysian settlement and a record $350 million Hong Kong fine.
- A DPA, not a plea — the parent company avoided a guilty plea and preserved its business licenses; only the Malaysian subsidiary pleaded guilty.
- $550 million (2010) — the SEC settlement over a subprime product, then a record penalty against a Wall Street firm.
- 1869, one basement room — where the reputation moat began, and the scale of firm for which it was the only moat there was.
Frequently Asked Questions
How did Goldman Sachs start?
Marcus Goldman, a Bavarian immigrant, founded the firm in 1869 in a one-room basement office at 30 Pine Street in Lower Manhattan, trading commercial paper — buying merchants’ short-term IOUs at a discount and reselling them to investors. The business ran on trust; he reportedly stood behind the paper personally. His son-in-law Samuel Sachs joined in 1882, and the firm grew through carefully chosen partners and referral networks.
Is reputation really a permanent competitive moat?
No. Goldman’s own history shows the reputation moat breaking three times: the 1929 Trading Corporation collapse, the $550 million subprime settlement in 2010, and the 1MDB resolution in 2020. The firm survived each breach not because trust proved durable but because it also held scale, network position, and regulatory standing. Reputation is a real moat and a breakable one.
Did Goldman Sachs plead guilty in the 1MDB case?
The parent company did not. Goldman Sachs Group Inc. entered a three-year deferred prosecution agreement under which an FCPA conspiracy charge would be filed and later dismissed on compliance with the terms. Only the Malaysian subsidiary, Goldman Sachs (Malaysia) Sdn. Bhd., pleaded guilty. Reporting at the time noted the bank sought to avoid a parent-level plea, which can jeopardize the licenses an investment bank operates on.
How much did Goldman pay over 1MDB?
More than $2.9 billion to U.S. and other authorities in October 2020, including the largest-ever penalty under the Foreign Corrupt Practices Act. Separately, the firm had settled with the Malaysian government for $3.9 billion — $2.5 billion in cash plus a $1.4 billion asset-recovery guarantee — and Hong Kong imposed a record $350 million fine largely outside the $2.9 billion figure.
Why didn’t Goldman’s compliance process stop the 1MDB deals?
Because process preserves judgment only when incentives reward preserving it. The transactions were vetted and passed internal review; prosecutors described staff who disregarded signs something was amiss. Against roughly $600 million in fees, a control function whose members are shaped by an institution that rewards closing is documentation rather than a check. The board’s subsequent clawbacks and pay reductions — extending to executives who were not involved or aware — amount to an admission that the incentive system produced the outcome.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B+ in shareholder value across Berkshire Hathaway, Illinois Tool Works, and Whirlpool. He is VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he turns proprietary frameworks — including Compounding Defense and the Moat Taxonomy, which sit under the LEAD doctrine’s Defense principle — into field-tested tools for operators who need to know which moats they actually hold rather than which ones they talk about. He is the author of The Unfair Advantage and the Stagnation Assassin series, a Forbes contributor, and an award-winning keynote speaker. Connect with Todd on LinkedIn or bring him to your stage via his speaking page.
Goldman’s reputation moat has broken three times and the firm is still here — because reputation was never the moat carrying the weight. Read that as permission and you’ve learned the wrong lesson. Goldman could spend its trust because it held three other moats. Count yours honestly, and if the answer is one, act like it.
The discipline of knowing which moats you actually hold is in The Unfair Advantage.

