Reinventing Goldman Sachs

Plus: Competition in Banking, Customer Germination, Affirm IPO

Welcome to another issue of Net Interest, my newsletter on financial sector themes. Net Interest now gets delivered to over 10,000 people each Friday, up ~700 on last week (thanks Josh!) If you’re reading this, but haven’t yet subscribed, join the discussion and sign up here.

Reinventing Goldman Sachs

Stripe is a company founded in 2010 in San Francisco. It builds software which allows companies to accept payments and manage their businesses online. Its mission is grand: “to increase the GDP of the internet.”

Goldman Sachs was founded much earlier, in 1869 to be exact, in New York City. It provides a wide range of financial services to companies, financial institutions, governments and more. Its mission is “to advance sustainable economic growth and financial opportunity across the globe.”

The two companies have a few things in common. They were both founded by immigrants from Europe. Their customers are typically businesses rather than consumers. [1] And, as their mission statements reflect, they are less in the business of economic production and more in the business of economic facilitation. 

Oh, and their market values are about the same. Stripe is reportedly in talks to raise funding at a valuation of between $70 billion and $100 billion; Goldman’s market cap is $84 billion. 

Last week the two announced a partnership: 

  • Stripe now enables platforms to embed financial services, enabling their business customers to easily send, receive and store funds

  • Stripe is enabling standardized access to banking capabilities via APIs by expanding its bank partner network to include Goldman Sachs Bank USA and Evolve Bank & Trust as US partners

While the announcement caused a buzz around Stripe, it fell flat around Goldman. Except for a solitary tweet: 

Throughout its history Goldman has moved forward on a culture of innovation. Yet in an environment where innovation is rife, can they compete?

Marcus Goldman, Banker and Broker

Before it was Goldman Sachs it was Marcus Goldman, Banker and Broker. Marcus Goldman had come to the US from Germany in 1848. He started out running a clothing business in Philadelphia, but pivoted into the money business after moving to New York. Every morning he would visit friends and acquaintances among the merchants of lower Manhattan and buy IOUs from them—bits of paper for which he would pay cash. In the afternoon he would make his way uptown to the commercial banks to sell these notes. Today, a fintech would eat his business for lunch, but the process of intermediation has always had value. 

The bits of paper Goldman traded would become known as commercial paper. After about ten years, Marcus Goldman was turning over $30 million in commercial paper a year. He decided to bring in a partner, his son-in-law, Samuel Sachs, and over the next few years other family members would join them. By the time of his death in 1904, the firm was the leading commercial paper house on Wall Street. 

As a new century dawned, the firm – now known as Goldman, Sachs & Co – had ambitions to enter new markets and in particular the market for underwriting debt and equity for corporate clients. It was a market controlled by three firms, among them JP Morgan, and it was difficult to break into.

Goldman’s entry was via a cigar company called United Cigar. Most capital raises at the time were anchored on hard assets, which is why the railroads were such important clients of the dominant firms. United Cigar didn’t have hard assets, but it did have earnings. Goldman developed the idea that companies such as United Cigar in the wholesale and retail sectors could be valued on the basis of their earnings power. Some might say the firm pioneered the method of valuing companies using discounted cash flow.

Goldman Sachs not only went on to break into the underwriting oligopoly, it went on to dominate it. A hundred years later it had a market leading position in debt and equity underwriting, becoming one of that rare breed of industry leaders “you don’t get fired for hiring”. In addition, it was still present in commercial paper, and it had amassed critical presence in equity and bond trading, M&A advisory and more. Underpinning its rise to prominence was a culture of innovation stemming from the United Cigar transaction.

Financial markets are a fertile ground for innovation. The money available creates the incentives and the distributed architecture of financial firms – where traders and bankers are given a great deal of autonomy – provides the environment (rogue traders aside). It is sometimes said that the porn industry finds use-cases for technology many years before they go mainstream (online payments, video streaming, video conferencing). If so, the finance industry isn’t far behind—it has been an early adopter of instant messaging, high speed communications and alternative data

Goldman’s innovation is less around fast adoption. Nor is it around new products. Other firms have their names pinned to certain products for posterity—JP Morgan and credit derivatives, Drexel Burnham Lambert and junk bonds, French banks and exotic equity derivatives. Not so, Goldman Sachs. Rather, its innovations have typically been around organisational structure and business model. 

Even in the one product area with which they are historically associated – block trading – their innovation was around deployment rather than product design. When retail investors gave way to institutional investors as major participants in equity markets in the 1950s, demand for trading big blocks of stock grew. Many firms were cautious about the risks involved, put off by the perception it would tie up a lot of capital. Goldman employed a different model and between 1955 and 1965 had the market almost to itself. Its understanding of investor positioning in the markets allowed it to sell blocks as soon, or even before, they’d been bought. It turned a principal business into an agency business. 

To preserve its market position, the firm employed a bit of marketing on top. According to Charles Ellis in his book, The Partnership, “Goldman Sachs partners would often bemoan publicly how tough and costly the block-trading business could be—and never acknowledged how profitable it really was.”

Throughout its history, Goldman has excelled at such innovations:

  • Under senior partner Gus Levy, it crafted a new compensation model for its securities salesforce, moving away from the “eat what you kill” approach that was dominant at the time. Commissions were pooled and salespeople were rewarded based on team contribution as well as individual performance.

  • Under John Whitehead, it restructured its investment banking business to separate execution from sales. Copying from the manufacturing industry where the roles were split, Goldman directed its best salespeople to focus on sales and its best product people to focus on executing transactions. By making the two equal in stature, the firm was able to recruit skilled professionals into both functions.

  • It was one of the first firms to charge for M&A advice. In the 1950s and 1960s, most investment banks got paid only for underwriting debt and equity deals; they didn’t charge for advisory work on mergers and acquisitions. When Goldman's Sidney Weinberg advised on the merger that created Warner Lambert Pharmaceuticals, he charged a fee of $1 million.

The focus on organisational structure and business model meant that while Goldman was not usually the first to come up with new product innovations, it was able to deploy them quickly. This was especially relevant in the period of time before 1995 when the firm’s partners carried unlimited liability and shied from the risk of untried technologies. 

Underpinning this focus was a strong culture that was sustained even as the firm grew larger. In any organisation there’s an inherent conflict between an employee’s natural short-term bias and the institution’s long term bias; an employee thinks in terms of years, an institution in terms of generations. In the days of unlimited liability, those interests were aligned. Once Goldman went public in 1999, it created a challenge. 

It’s a problem Goldman was well-equipped to deal with. The employee/institution trade-off is yet another manifestation of the principal/agent problem. From its experience in block trading to building principal trading businesses, managing between principal and agent is a task Goldman is frequently called to resolve. Ironically, it was exposed along the employee axis by the very process of going public that enabled the firm to do more principal investing. 

Goldman addressed this challenge by retaining the veneer of its partnership model, if not the substance. Just a few weeks ago, as highlighted in More Net Interest, Goldman unveiled its latest cohort of partners—a process it undertakes on a two-yearly cycle. Today, the firm has around 440 ‘partners’, drawn from all corners of its business. The right number is probably around 150, consistent with the Dunbar number of stable relationships humans can typically sustain. But in a workforce of 38,000, that could be too elusive an incentive and so 3x that number is the next level of hierarchy, on a par with the size of an Oxbridge college, say. [2]

The partnership structure makes the firm a lot more navigable. In my dealings with Goldman, I could always get ‘the expert on X’ on the phone within a day; at other firms it would take too long to identify who that was. 

With this structure in place, the firm was able to sustain its franchise while protecting against short-term urgency. However, it doesn’t protect the firm from the tides of the market and over the past ten years two competitive challenges have emerged. First, technology is changing the economics of intermediation across all domains. It’s not just taxi dispatchers whose roles are being made redundant and Goldman Sachs is not immune. Second, in the aftermath of the financial crisis, Goldman Sachs was forced to become a bank. This imposed new regulatory burdens, limiting its scope in certain business areas. 

Nevertheless, as Charles Ellis notes:

“Redefining a business and reinventing the firm—often very substantially changing itself and its way of doing business—are themes in the extraordinary growth and expansion of Goldman Sachs.”

Goldman Today

Like any incumbent business, Goldman’s relationship with technology has been nervy. Charles Ellis writes that, “Senior management got its first briefing on the impact of the Internet in 1996, including answers to the basic question: What is the Internet?” However, as in other matters, once it grasped the issues, it was able to act quickly:

“By the end of the decade, it had more than twenty thousand personal computers and workstations backed by eight trillion bytes of storage. About eighteen thousand miles of cable and thirty thousand phone lines snaked through its Broad Street headquarters, linked to a fibre network that could sling data to its offices around the world at 6.1 gigabits a second.” (Excerpt from Dark Pools, by Scott Patterson.) 

Yet most of the value created directly via technology accrued outside Goldman’s legal structure. The firm was an early investor in market infrastructure plays like ICE, Markit and Archipelago (which merged with the New York Stock Exchange, now part of ICE). As discussed in a previous edition of Net Interest, The New Power Brokers, these companies are wrestling market power away from intermediaries like Goldman Sachs. 

The firm’s own tech spend is vast. In 2019 it spent $4 billion and its budget coming into this year was $4.3 billion, of which just under half is allocated to new investment and the rest to making the existing infrastructure tick. Of the firm’s 38,000 employees, around a quarter are classified as engineers. Given these numbers, the firm has in the past tried to pass itself off as a tech company. In 2015, then CEO Lloyd Blankfein said, “we are a technology firm.”

Goldman isn’t a technology firm. But then it’s difficult to define what is. Netflix isn’t a tech company according to Benedict Evans. And Amazon may have only become a tech company in 2010. But there’s a perception gap that Goldman’s own clients try to exploit (Lemonade, an insurance company, is covered by its tech analyst) so who can blame them for joining in? And to be fair, the audience may be on campus, where Goldman is forced to compete with tech firms for talent. 

What Goldman is, is a pretty good bank. Yet after being forced into becoming a fully-fledged bank in 2008, the firm suffered the associated burdens without capturing the advantages. David Solomon, the firm’s CEO, said earlier this year: “And so I think it's a big mistake that 5, 7, 10 years ago, when we became a bank, we didn't start thinking about the fact that we should be providing other services to corporations that we're deeply embedded with.”

Meanwhile, the firm was also spending on its consumer image, again with no payback. Lloyd Blankfein said in 2016: “In general, we never publicised what we did anywhere. We never saw Goldman Sachs commercials or advertisements, or even our name on our building. After the crisis, we realised that one of the things we needed to do is, we didn’t communicate enough with the public to let them know who we were...”

In the past few years, Goldman has sought to leverage the twin investments it made becoming a bank and raising its public profile. It launched a consumer banking brand, Marcus, in 2016 and a transaction banking business in 2020. By the end of September 2020 it had amassed $96 billion of customer deposits in Marcus and $28 billion of customer deposits in transaction banking. 

But the interesting thing is what Goldman is doing with the infrastructure supporting these businesses. By partnering with other companies, like Stripe and also Apple and others, it is expanding the distribution profile of its offering. It can combine its strengths in financial infrastructure and regulatory compliance with its relationships with big companies to help get banking products out there. Unlike other big banks like JP Morgan, it won’t end up cannibalising its own customer base because in retail and small business it doesn’t have one. In those markets, Goldman Sachs plus Apple may be the strongest competitor JP Morgan has.

Former CEO Hank Paulson is quoted in Charles Ellis’ book: “The best protection is to continually reinvent ourselves so someone else doesn’t do it to us. What keeps me up at night isn’t what our traditional competitors are doing, but that someone we didn’t foresee will use technology to emerge as a significant rival.”


The definitive books on Goldman Sachs are Money and Power: How Goldman Sachs Came to Rule the World, by William Cohan; and The Partnership: The Making of Goldman Sachs by Charles Ellis. For an excellent long-form piece on the firm, I would recommend this by Nicolas Colin. 

[1] Stripe’s co-Founder and President recently remarked, “one of the fundamental tensions at Stripe… is how do you explain us to your parents if you work here?” Goldman Sachs employees have suffered that tension for years.

[2] Geoffrey West in his book, Scale, highlights that optimal social groupings scale at a constant factor of around three; Priya Parker in The Art of Gathering references similar ‘magic numbers’ in groups.

Forwarded this? Sign up here

Share Net Interest

More Net Interest

Competition in Banking

When firms complain about competition, particularly when they are leaders in their industry, it’s difficult to know whether to believe them. No-one’s going to put their head above the parapet, not even a monopolist, and say, “there’s no competition in my market”. This week, Jamie Dimon was extremely vocal about competition in banking:

And recently, you saw Google, and they're entering the banking business kind of as a marketplace. Look, that’s a real competitor… And if you have any complacency about that, you’re a little crazy.

And I’ve also mentioned there are some great competitors out there not in the banking system. PayPal is worth more than most other banks. Stripe is worth more than most banks. Square has done an unbelievable job of Square cash… I look at some of those things very often to say, “We could have done that, too,” and we didn't.

I mean you’ve got -- everyone’s out there. Everyone’s -- there’s a lot of very good people in digital, and I've mentioned some who are not banks… It's trench warfare. You got to fight for that share, and it’s like -- it’s an ongoing battle with lots of competitors out there, and there are some invisible ones…  So everyone is coming through. And I mentioned Google, that’s not going to stop -- some of the other big ones from doing almost the same thing. So no, I think the competition is going to be very tough, and we hope to be able to eke out continued gains.

Although he’s a market leader, banking remains a highly fragmented industry. So I think we can believe Jamie Dimon that competition is tough. But just as the monopolist’s audience is the regulator, so too might be Jamie’s:

The banking system as a size relative to the global economy gets smaller and smaller and smaller and smaller… So you've had 80% of the mortgage go outside of banking. Tremendous amount of private credit is going outside of banking. And you can go down one after another, things that are leaving banking because they get more favorable treatment outside. And the regulator... eventually they're going to regulate the banking system out of business.

Competition in banking may be running ahead of regulators. This is Dimon’s call to them to fix it. 

Customer Germination

A popular hypothesis behind some fintech strategies is that you can grow with your customers. By acquiring them young, even when they don’t have much money, you can lock in their loyalty and become more profitable as they grow more affluent. (Assuming that more affluent customers tend to be more profitable, which in wealth and savings, although not in credit, is typically the case). 

The median age of an entry level Lemonade customer is 30, the median age of a new Robinhood customer is 31 and median age of a Monzo customer is 31-32. According to the Federal Reserve, the median net worth of an adult American under the age of 35 is about $11,000. However, by their 40s and 50s, their net worth has grown by 10 to 15 times, and that growth peaks at 25 times after the age of 75. Keep them through to their 40s and 50s, then, and the economics start to look a lot more interesting. 

It’s a difficult hypothesis to test because not enough time has transpired from the launch of these companies to see what happens when their customers hit 40-50. The alternative, of course, is that once they have greater wealth, they shift it over to incumbents. 

Chime may have short-circuited this process. I saw the Co-founder and CEO, Chris Britt, present this week, and he commented on how, since Covid, their average customer age is increasing. Having skewed younger before Covid, new customers are being acquired in their 40s and 50s. He didn’t give an indication of profitability, but this must make a difference. Chime’s target market is Americans earning less than $100,000 a year. Significantly less and they may not be profitable; the 40-50 age demographic may be the key to some fintechs’ long term success. 

Affirm IPO

Affirm has chosen a good time to go public. With AirBnB up 120% from its IPO price and DoorDash up 75%, anticipation is building. I contributed to a detailed analysis of the company as part of Mario Gabriele’s S-1 Club. 

The key question is whether Affirm is seen as a merchant network and payment platform like Paypal or a lending business like… well, a bank. Parsing its income statement, and most of the money Affirm makes comes in payments. It processed just over 8.5 million transactions in the past year and took a ~6.0% slice out of each of them.

The trouble is, the costs of this business appear over on the banking side, where it gives away free loans (literally, 0% APR). In the past twelve months, the loans it has kept on its balance sheet have just about washed their face after credit and funding costs. And the loans it has sold have been sold at a loss. 

To help untangle the business, the writing crew and I will be hosting a call next Tuesday. Look out for details on Twitter.