1WALTER WRISTON’S WORLD
In his time, Walter Wriston was one of the most powerful people on the planet, the chairman of the financial giant Citibank and its parent corporation, Citicorp. He was also a man with a vision. His book, The Twilight of Sovereignty, is nearly forgotten today, but it predicted how the information revolution would transform global politics. The sovereign power of states, which had grown since the waning of the Middle Ages, was now in decline. New technologies and market freedoms were about to “decentralize power,” rendering “once vital strategic ‘choke points’” irrelevant, and “shifting the tectonic plates of national sovereignty.” Global flows of information, money, and trade were not only flooding across national borders but washing them away, creating a true global marketplace in which agile individuals and businesses could evade government regulation.
When novel ideas succeed, they decay into stale clichés. Wriston’s arguments became the common wisdom of the business bestsellers stacked on tables in every airport bookstore. But he had begun speaking and writing about the irresistible challenge that global information technology and markets posed to government power in the 1970s, long before others were really paying attention. Back then, financial flows were largely restricted by national borders, and the Internet was an obscure government-funded experiment. Even when he published his book in 1992, it wasn’t clear that technology would transform everything. The Berlin Wall had fallen only three years before, and the Cold War still hadn’t yet relinquished its skeletal grasp on the world’s politics and markets.
Wriston inherited his zeal for economic liberty and deep distrust of the state from his father. Henry Wriston, a university president and president of the prestigious Council on Foreign Relations, was personally invited by the famous economist Friedrich von Hayek to help found the Mont Pèlerin Society, a highly influential group of libertarian and conservative thinkers who tended the flame of free market thinking after World War II. Like his father, Walter was a globalist who was deeply influenced by Hayek’s vision of a world in which market freedoms were the basis of individual liberty.
But Walter Wriston wasn’t an academic or a think-tank president. As Roy Smith, who worked for Citibank’s rival, Goldman Sachs, described it later, Wriston was “the most influential banker” of his time, the man who had turned Citibank into the “one bank that all others copy shamelessly.” When Wriston said in 1992 that he had written his book “from the perspective of a participant in the evolving global financial marketplace,” he was engaging in polite and perhaps ironic understatement. The histories of globalization pay close attention to the politicians and high officials who cleared the way for open markets, and the thinkers who argued for them, but they regularly ignore the business leaders who actually built them. Wriston was the Zelig of globalization. Look closely at the burgeoning of international finance, of information networks, of the logistical innovations that transformed trade. You’ll find him everywhere.
In his personal relations, Wriston was a little awkward and stiff, exemplifying the virtues and scruples of his Methodist upbringing. His business philosophy, in contrast, displayed a distinct piratical flair. He disdained borders and national rules in favor of the high seas of global markets, where Citibank and its rivals could outsail the grasping monarchs who ruled the land, creating their own freebooter’s republic on the waves.
Already as a trainee officer at National City Bank (now Citibank), Wriston had risked causing apoplexy in his superiors by lending $42 million to Malcom [sic] McLean, a trucking entrepreneur with a new and controversial idea about how to transport goods cheaply by water as well as land. McLean used the money to launch the containerization revolution, which dramatically lowered the costs of transporting goods around the world. Wriston’s financial innovations helped create the modern Eurodollar market—a vast offshore realm of financial transactions in U.S. dollars happening outside of U.S. borders. His efforts to build a private global payments system under Citibank’s control in the early 1970s prompted other banks to build their own collective system, so as to avoid being drawn inside Citibank’s gently smiling jaws.
Wriston’s willingness to put his ideas into action changed the world. As he explained in 1979, the “current banking network, with its Euromarkets and its automated payments system” seemed dull and technical, but it had immense political consequences. He believed that if money could move rapidly from country to country, it could no longer be mastered by states. Instead it might master them, replacing the whimsical tyranny of political rulers with the austere rigor of market discipline. Equally, the free movement of information across global telecommunications networks would prevent governments from halting the spread of ideas that they did not like. As Wriston explained later, telecommunications networks could transform high technology manufacturing, allowing multitudes of different producers in different countries to coordinate together on building a common product.
Wriston was right that these changes would have immense political consequences, but he misunderstood what these consequences would be. He once told a friend that “centralization … is a fascist state,” and believed till his death that he and his peers were building a freer world with limited government. The irony was that he and other business leaders were centralizers by their nature: they sought to dominate markets, so that other businesses would have to use their systems and pay tribute to them. They built world-spanning networks that centered on a few key choke points. Eurodollar markets and global payment systems redirected the world’s financial flows through U.S. banks and U.S.-dominated institutions. Global information flowed through networks centered on U.S. territory and subject to U.S. surveillance. And as global manufacturing came to rely on information and financial networks, it, too, became concentrated in ways that made it vulnerable to U.S. authority. The tragedy of globalization was that men and women like Wriston built a world that seemed to escape the control of government but in fact was wide open to government power and its own undoing.
* * *
Wriston eventually came to believe that “international banking is a system designed by fate to exist in a certain state of economic tension, with all governments, including the most democratic.” But at the beginning of his career, international banking barely existed. The banking industry that Wriston and others confronted in the 1960s was sluggish, timid, and lazy. Banks were trapped inside national borders by complex and clashing rules put in place after the financial crash of the Great Depression. These regulations meant that most banks faced little international competition and had scant incentive to invest in new ways of doing things. It was nearly impossible to be a true international bank.
The 1960s banking industry was a Victorian survival into the modern era, a clattering steampunk engine of rusting pistons and gutta-percha-covered cables, with a few incongruously modern parts bolted on. Eric Sepkes, who helped build the European payments system, later recalled how Citibank’s London operation relied on a system of pneumatic tubes to communicate between its payments and authorizing offices. Staff had to handwrite payment instructions on a form, which they then stuffed into a canister and inserted into a partial vacuum conduit that ferried it to its destination (the City of London had built miles of pneumatic tube networks in the nineteenth century). One day, when the payments people didn’t hear back from authorization, they discovered that the tube had become blocked. Citibank had to call in a chimney sweep to fix the problem, re-enabling payments processing for the entire continent of Europe.
Global banking was a system of mysterious tubes on a much larger scale, with various portals that took money in, did expensive and incomprehensible things to it, and spat it out somewhere else. No one fully understood the machinery, least of all the people who were supposedly in charge. The gentlemanly activities of merchant banking, where men with excellent pedigrees built on their social connections to win deals, remained rigidly distinct from the mundane tasks of payments processing, which were performed by female clerks surrounded by vast piles of paper. It took a very long time to send money across borders. At one point, Citibank’s Argentinian branch had to convert its profits into cases of Scotch whisky, to prevent them from being inflated away before they could be sent to New York.
Wriston helped rebuild this clanking machine into an engine of transformation, welding disjointed national markets into a true world economy. His strategy was built around two insights. The first was that global markets could—if they were allowed—circumvent the labyrinthine systems of rules constructed by national regulators and eventually replace them. The second was that banking was a “branch of the information business.” Market prices provided one crucial source of information, summarizing as they did the decisions of millions of individuals over what to buy and sell. Technology provided another, allowing banks to discover the information they had buried within their own bureaucracies and to better exchange information with each other and their customers. With proper technology, the dull-seeming backroom activities of banking such as payments processing could become a source of profit and power.
When Wriston began to remake Citibank, money had already begun to seep through the seams in the ductwork. Businesses outside the United States desperately wanted U.S. dollars, which were needed, for example, to buy and sell oil, while businesses in the United States wanted to earn higher returns. U.S. regulators had capped interest rates for ordinary consumers and stopped interest payments altogether on corporate deposits. Bankers had already begun to figure out crafty means to connect demand to supply.
Wriston and his colleagues built the institutional infrastructures that allowed this to happen at scale. They created financial instruments like the certificate of deposit, which provided a legal pipeline to smoothly convey the dollars owned by American businesses to the international banks that needed them for their customers. Citibank’s competitors, like J. P. Morgan and Warburg, adapted these instruments and came up with their own ideas. What was once a small and disconnected trade in Eurodollars, based primarily in London, became a vast marketplace for buying, selling, and lending in dollars outside the United States.
As the political economist Eric Helleiner describes it, the Eurodollar market became a legal gray zone where vast amounts of U.S. dollars circulated beyond U.S. borders. As the market grew, the U.S. dollar became established as the universal basis of international trade. If you were a Japanese company selling goods to a business in Italy, it was hard to convert the Italian lira that you were paid in directly into Japanese yen. The economic relationship between Japan and Italy wasn’t big enough to support a liquid market where the two currencies could be exchanged directly. The Eurodollar market provided an easy detour, where you could turn lira into dollars and then dollars into yen. And as the supply of Eurodollars grew, it increasingly made sense for companies to simply buy and sell in dollars, which they could then convert into their home currencies.
The result was that the dollar became a global currency, without anyone really planning it. More dollars circulated outside the United States than within it. U.S. officials at the Federal Reserve and elsewhere paid surprisingly little official attention to what was happening. This made the market attractive, for example, to the Soviet Union, which needed dollars for international trade but worried that they might be seized by the U.S. government if they were deposited directly in American banks. By using Eurodollars, which were bought and sold in London and Italy, they thought they could avoid this risk.
These markets all depended on an infrastructure of clever financial engineering. Banks weren’t trading physical stacks of hundred-dollar bills. Looked at closely, Eurodollars were an accounting fiction, imaginary dollars traded between real banks. They couldn’t be used for anything other than buying other currencies. But every single Eurodollar had to be backed by a real dollar, sitting in a U.S bank operating under U.S. law and responsible to U.S. regulators. As Wriston explained, “All the dollars in the world—except [physical] currency—are deposits in a bank in America, because that is the only place anyone can spend a dollar.”
This meant that transactions using Eurodollars had to be cleared through a U.S. bank’s internal processes (moving money from one customer’s account to another’s) or through a clearing institution run by U.S. banks, such as CHIPS, the Clearing House Interbank Payments System. Foreign banks had to maintain clearing accounts in U.S. financial institutions if they were to trade dollars and participate in global finance. The Eurodollar market might indeed have been a pirate kingdom, but it was one where the buccaneers had to regularly provision themselves in the monarch’s ports. The more that foreign banks came to depend on access to U.S. dollars, the more vulnerable they were to U.S. regulators, whenever those regulators finally woke up.
Gradually, the “dollar clearing system” run by U.S. banks like Citibank and J. P. Morgan and clearing institutions like CHIPS became the beating heart of the world’s financial system, circulating dollars around the world in a regular systole and diastole. The Eurodollar market, far from creating a decentered new realm of finance, had made the global financial system more fragile and more vulnerable to American jurisdiction.
This became clear in 1974, when Chase froze the dollar clearing account of a small German bank, Herstatt, which had run into financial difficulties. Flaws in the system meant that other banks’ transactions with Herstatt couldn’t be cleared, which meant that other banks’ transactions with those banks failed to clear, and so on, in an ever-growing cascade. Citibank pulled the plug on its automatic money transfer system to stop payments to banks that might no longer be creditworthy. The global financial system went into heart failure; Citibank’s decision “virtually brought the [world’s] payment system to a halt.” Wriston spent the next couple of days corralling elite bankers to put the system together again. He succeeded, but as he noted later, the episode showed how twenty to thirty private banks had become the “de facto payments mechanism of the world.” In his dry and understated description, this made central banks nervous, “and it also makes us nervous.”
* * *
Over the 1970s, the other key element of global banking—automated settlement messaging—became centralized, too. Again, Wriston’s Citibank played a crucial role in the story.
Financial transactions between banks in different countries have always been difficult. In the eighteenth and nineteenth centuries, banks relied on close relationships with their “correspondent” banks in other countries. They would issue physical letters of credit, signed instructions telling a correspondent bank to provide money to the bearer on the promise that the bank would be paid. However, such letters could be forged and were difficult to verify in a world where sea voyages could take weeks or months (much of the plot of Golden Hill, Francis Spufford’s novel of pre-Revolutionary New York, turns on the difficulties of establishing whether an individual carrying a large letter of credit is who he says he is). The telegraph of the nineteenth century and the telex machines of the twentieth century made communication faster, but they were still awkward and cumbersome. In the 1960s, telex payments required the operators for both banks to perform logarithmic calculations using shared code books to ensure that security had not been breached.
As Citibank became the most internationalized bank in the world—it had branches in over ninety countries—Wriston recognized an opportunity to standardize the ways in which banks talked to each other across borders. Every important international bank had to do business with Citibank. That meant that if Citibank decided on a new technological standard for payments messages, there was a good chance that it would stick and spread. And if it did spread, Citibank would become “the nexus of the world’s payments system,” giving it a permanent advantage over its competitors. Every time that money went from one country to another, it would have to go through Citibank’s system, providing it with a potential stranglehold on the market.
Wriston and his eventual successor, John Reed, were both obsessed with technology and had set up their own skunk works, Transaction Technologies, Inc., to build specialized hardware and software. Transaction Technologies was tasked with building MARTI (Machine-Readable Telegraphic Input), a kind of secure post office system just for banks. After MARTI was launched, Citibank’s senior operations officer Richard Matteis reportedly demanded that all its correspondent banks use MARTI to communicate with it or have their telexes returned. As banker Renato Polo described it decades later, Citibank more or less said, “We advise you to use MARTI from now on. If you don’t use it, we will not execute your instructions.”
The problem was that Citibank was too big to be trusted. Citibank might want to control a messaging system that every other bank depended on, but those other banks had excellent reason to be wary. They had no desire to depend on a technology that was outside their control and that Citibank could change at will to squeeze them. As Polo went on to explain, “[E]ither you make yourself captive to one correspondent, which no one in his right mind would ever do, or you say no.” Many indeed said no, leading to utter chaos as Citibank’s backroom operations struggled to deal with a blizzard of failed transfers and the temporary near collapse of Citibank’s network of correspondent banks.
Citibank’s MARTI wasn’t the only possible system for payments messages. A Dutch banker, Johannes (Jan) Kraa, had persuaded a group of European banks to found the Society for Worldwide Interbank Financial Telecommunication (SWIFT) in 1973 to create an alternative system for secure communications between banks. SWIFT was based in Belgium, to sidestep the rivalry between the two financial centers of London and New York. However, it had a hard time getting European banks to agree to participate. Each wanted its own country’s standards to prevail, making it hard for SWIFT to reach consensus or attract enough banks to make the system workable. By 1974, it looked increasingly likely that SWIFT would fail. Then the MARTI debacle persuaded European banks that if they didn’t come up with a shared standard, someone else, like Citibank, would come up with a standard for them.
Citibank’s campaign to force MARTI on its correspondents led to a “decisive landslide” in SWIFT adoption. By the end of 1975, SWIFT had 270 member banks located across fifteen countries. Although Wriston blamed MARTI’s failure on the difficulties of getting banks to agree with each other, Matteis later admitted that “people’s resistance to MARTI made SWIFT a success.”
Soon, SWIFT membership was a necessary condition for participating in the global financial system. And the more that SWIFT grew, the more essential it became to the world’s financial system, including America’s. Eleven years after its founding, Robert Moore of Chemical Bank became the first American chairman of SWIFT’s board, while Citibank’s Yawar Shah became chairman in 2006. Today, SWIFT’s messaging system carries over ten billion messages annually, facilitating 1.25 quadrillion dollars in transactions. Like the dollar clearing system, it plays a central role in global finance. As the writers of SWIFT’s official history acknowledged, SWIFT has become an “‘obligatory passage point’ … if you want to participate in financial services you must join because there is no real alternative.”
SWIFT’s central importance created its own problems. Willie Sutton is supposed to have said that he robbed banks because “that’s where the money is.” Now, the money flows through complex technical networks such as SWIFT, and criminal minds have taken notice. North Korean hackers exploited weaknesses in SWIFT’s system to transfer $81 million out of the Bank of Bangladesh’s accounts at the New York Federal Reserve Bank in 2016. If they hadn’t made a typo, they might have gotten away with a billion.
The cops noticed, too: SWIFT came under political pressure from the United States, which wanted it to do more to prevent and detect crime. For a long time, SWIFT’s officials managed to fend off these requests, arguing that SWIFT was a purely technical organization, dedicated to managing the world’s financial plumbing.
Even though SWIFT was based in Belgium, it was run by a consortium of international banks. It had succeeded in replacing a Dickensian tangle of pneumatic tubes with a system for sending money that actually worked. SWIFT had defeated an alternative system that would have been controlled by a single American bank if Walter Wriston had had his way. But some of the banks were American, and all of them needed access to the dollar clearing system to carry out their work. Most people didn’t pay attention to SWIFT, which seemed to be a boring but useful organization dedicated to maintaining a boring but useful part of the world’s financial plumbing. But SWIFT’s ability to fend off politics depended on U.S. willingness to tolerate its apparent independence. If the U.S. government ever really used its muscle, SWIFT would have to cave.
* * *
In 1996, Walter Wriston enthused that “there is no way on God’s green Earth government can exercise censorship on the Net in any meaningful way.” But The Twilight of Sovereignty, which he wrote just a few years earlier, had nothing to say about the Internet. Its index jumped straight from “International Monetary Fund” to “Investment spending.” Back when the Internet was one specialized network among many, the most important change seemed to be the physical transformation of the world’s telecommunication network, which the Internet and its competitor networks rode upon. A few years later, when technological evangelists celebrated the liberatory potential of the Internet, they paid little attention to the wires and cables that carried it.
Wriston had intimate and painful experience of the world before it had been transformed. The first transatlantic cable capable of carrying voice transmissions had been laid in 1956. It allowed a grand total of thirty-six phone calls to happen at once. There were so few phone circuits available between Brazil and New York that Citibank’s local branch hired “squads of Brazilian youths” as dialers, who kept calling and calling for days until they got an open line. When they succeeded, Citibank’s employees stayed on the phone, reading books and newspapers to keep the line open until it was needed.
Things began to change in the 1970s and 1980s. Electronic switches—specialized computers—began to replace mechanical arrays in which human operators had connected one caller to another by plugging their cables together. These switches could handle many conversations at once. Digital fiber-optic technology allowed human voices to be turned into information, and information into rapid pulses of light, so that a thin strand of flexible glass could carry a myriad of individual conversations. The first transatlantic fiber-optic cable, which was laid in 1988, could carry forty thousand simultaneous phone calls. As other fiber-optic cables were laid, flows of money, information, and ideas became “utterly dependent on the new world communications network” and the “path to prosperity” it offered. Wriston believed that this new, apparently decentralized system of communication required governments to “surrender control over the flow of information.”
As the Internet swallowed up other networks over the 1990s, it seemed to reinforce Wriston’s arguments about decentralization. After all, the Internet was designed from the ground up to avoid central control. As a popular dictum described it, the Internet “interprets censorship as damage and routes around it.” But something funny started happening with the physical networks that the Internet relied upon. The more that they spread around the world, notionally escaping their American origins, the more centralized they became, so that their key flows were channeled through switches and exchange points on American soil.
Copyright © 2023 by Henry Farrell and Abraham Newman