Chapter Four
Thick Data—Not Just Thin Data
This book is an essay in what is derogatorily called “literary economics,” as opposed to mathematical economics, econometrics, or (embracing them both) the “new economic history.” A man does what he can, and in the more elegant—one is tempted to say “fancier”—techniques I am, as one who received his formation in the 1930s, untutored. A colleague has offered to provide a mathematical model to decorate the work. It might be useful to some readers, but not to me. Catastrophe mathematics, dealing with such events as falling off a height, is a new branch of the discipline, I am told, which has yet to demonstrate its rigor or usefulness. I had better wait.
—Charles P. Kindleberger, Manias, Panics and Crashes: A History of Financial Crises
The Story of the Three Traders
One day in early September 1992, three currency traders were sitting in the dingy offices of their hedge fund on Seventh Avenue in New York City. With duct tape on the carpet and a shabby conference table in the middle of the room, this unassuming space seemed an unlikely scene for the most monumental—and profitable—bet of the decade.
The three men were in the midst of discussion, running through their various data analytics. This trove of data, however, did not consist of spreadsheets, benchmarks, or numerical models. The financiers were, in fact, unraveling a dense little nugget of thick data that required an empathic understanding of both bruised pride and aspirations of autonomy. Specifically, they were parsing out the nuances of a sparring match between Helmut Schlesinger, head of the German federal bank, the Bundesbank, and Norman Lamont, the British finance minister.
In the wake of the Maastricht Treaty, signed earlier in the year, the most powerful central bankers in Europe were working toward the goal of creating one single European currency—the euro. But before reaching this goal, they had to resolve all sorts of political, economic, and cultural impediments. Most of these obstacles involved the role of Germany’s central bank, the Bundesbank. Economists generally agreed that hyperinflation after World War I had left the Germans vulnerable to the Nazi regime. In the years following World War II, with the horrors of the Nazi era behind them, the main goal of Germany’s central bank was to keep inflation rates low to avoid creating another opportunity for a destabilizing political movement.
After Maastricht, however, the bank had another potentially conflicting agenda: the Deutschmark would now be responsible for anchoring the new exchange rate mechanism for the rest of Europe. The 1990 reunification between East and West Germany caused inflationary pressure inside the country, so the Bundesbank raised interest rates, a move they had made many times in the past. The higher interest rates, however, alongside the recession and low interest rates in some of the other European countries like Great Britain and Italy, caused liquidity to move into the Deutschmark. The lower currencies, specifically the pound sterling and the lira, were trading at the very bottom band of the exchange rate mechanism as a result of the Bundesbank’s monetary strategies. Suddenly, the sovereignty of Germany’s central bank—its historic priority to make policy in the best interests of the German people—was at odds with the overall vision of a unified European market. Where would the chips fall? Would the Bundesbank ultimately choose its own anti-inflation policy or would it maintain alliances with the other countries in Europe—keeping the dream of a unified currency alive?
After a contentious series of meetings in late summer and early fall, Helmut Schlesinger was incensed by the perceived audaciousness of Norman Lamont, who, banging his fist on the table, demanded that the Bundesbank take action. Instead, Germany retreated behind a wall of thinly veiled disdain. Schlesinger announced at a public forum that he did not want to guarantee any action on interest rates. Later, he told an audience that he had little faith in the notion of fixed currencies between European central banks.
Back in New York, our three traders were paying close attention to the unfolding drama. Had Lamont pushed Schlesinger too far? Were Schlesinger’s political ambitions with Europe, or with his own institution and the autonomy of Germany? How much stomach did the British government have for raising interest rates in the midst of their highly leveraged short-term mortgage economy? The head of the New York hedge fund—one of our three investors—had attended one of Schlesinger’s conferences in Germany; the hedge funder immediately approached the central banker after the talk to get more nuanced intelligence. He asked Schlesinger if he agreed, in general, with the goal of a European currency. The leader of the Bundesbank—a bureaucrat who had spent his entire career climbing the ranks at the central bank—replied that he did like the idea but that he was only interested in one name for that currency: the Deutschmark.
Despite the macroeconomic details, the plot points of this conflict might well have come from a reading list at a liberal arts university. Ego, political machinations, loyalty, wounded pride and ambition: these are the character traits you’ll find in a great Shakespearean drama or a work of sweeping history by Thucydides.
One of the New York traders stood up to chart out a probability tree of the situation on a blackboard. It seemed more than plausible—given the characters and their circumstances—that the Bundesbank would choose its anti-inflationary policy over saving the other currencies and their devaluations, particularly those of Italy and Great Britain. With the economy already in recession, higher interest rates in Great Britain would immediately hurt the British people. In order to keep this new exchange rate afloat, the men decided that three things could adjust: German prices could go up; British prices could go down, or the exchange rate would adjust.
From these three scenarios, the traders walked through the consequences: The Germans were not going to tolerate inflation because of their history; the British economy could not tolerate deflation because of its short-term mortgage market. The conclusion seemed very clear: the exchange rate would have to adjust. The British sterling was almost certain to sink in value and the German bank would do nothing to save it. It was obvious to all three of them: the best speculative opportunity for them was shorting the British sterling. It was a no-brainer.
“What is the probability that the British bank will have to devalue in about three months?” one of the men asked.
“I’d say it’s about 95 percent,” the man at the blackboard answered.
The three men sat in silence for a moment. Within the boundaries of currency speculation, if they shorted the British sterling and they were wrong, over time they might lose a percentage of their position. But if they shorted the British sterling and they were right, they would make 15 to 20 percent off the position.
“95 percent on a 20–1 bet…” Their silence said it all.
Finally, the leader, the manager of the entire fund, asked his fellow investor, an expert in foreign currencies: “What would you bet on this one?”
The trader paused. “Three times capital?”
The whole of the hedge fund was currently valued at five billion dollars. Three times capital was fifteen billion dollars. If they took a position to short the British pound and they succeeded—that is, the currency did indeed devalue—they would surely bankrupt the entire bank of Britain and, in turn, disrupt the fundamentals of macroeconomic policy across all of Europe. The repercussions of their actions would be felt across the entire financial industry for decades to come and bring them under the scrutiny of financial regulators all over Western Europe. It would also make them both the most infamous and the most admired currency speculators in the world.
A moment later, the leader of the three calmly made his decision: “Let’s do three times capital then.”
And then he stood up and walked out of the room.
Numerous investors made money on September 16, 1992, a day later dubbed “Black Wednesday,” but no one made as much money as our lead financial speculator—George Soros—his heir apparent, Stanley Druckenmiller, and his then–chief strategist, Robert Johnson. Soros became known as the man who “broke the Bank of England.” When the British government finally gave up and withdrew from the exchange rate mechanism—after weeks of attempting to buoy the sterling with artificially high interest rates—the British taxpayers lost $3.8 billion and George Soros’s personal fortune was enlarged to the tune of $650 million. By some estimates, Soros Fund Management walked away with a profit from the trade of over one billion dollars. The head of Fiat in Italy claimed that it was more lucrative to be a a mere shareholder in Quantum Funds in 1992 than to own the entire enterprise of Fiat automobiles.
What happened in that room on Seventh Avenue? How did the three investors know that this was the moment for a big bet? Every serious financial firm in the world was looking at this chain of events. How did George Soros extract so much more data from the context than most other investors, taking into account not just what was officially said, but what wasn’t said and why it wasn’t said?
Soros arrived at this particular moment in time with a long history of engaging in humanities thinking. Decades before he was making world-famous speculative market calls, he was a student of philosophy at the London School of Economics in the late 1940s and early 1950s. His intellectual hero and mentor there was philosopher Karl Popper. While Soros studied under him, Popper indoctrinated Soros with an intellectual rigor based on his concept of “falsifiability,” or the constant quest to prove how you’re wrong, rather than prove how you’re right.
As a philosopher of science, Popper’s concept was concerned mainly with the cult of certainty that surrounded the scientific method: if you cannot find a situation in which the theory does not stick, then the theory holds. But no amount of testing, Popper emphasized, was exhaustive. “All the scientist can do, in my opinion,” he wrote in his 1956 essay “Three Views Concerning Human Knowledge,” “is to test his theories, and to eliminate all those that do not stand up to the most severe tests he can design.”
Soros was originally captivated by these ideas as a philosophy student, but he soon learned that he could apply a version of “falsifiability” to great effect in market systems. He started to approach his market speculation with this same relentless drive to disprove his own sense of certainty about where the market was moving.
As a child who had grown up during an occupied Hungary in World War II, Popper’s philosophical worldview also resonated with Soros’s. “Where we believed that we were standing on firm and safe ground,” Popper wrote in his essay “The Logic of the Social Sciences,” “all things are, in truth, insecure and in a state of flux.” Having seen the destabilizing forces of war throughout his earliest years, Soros was especially sensitive to the nonlinear movements of history. He recognized that sweeping events in politics were often a result of seemingly trivial personal snubs—the turf wars, indignation, and bruised egos simmering just below the surface of more rational monetary policy and treaties.
Stanley Druckenmiller and Robert Johnson were more traditionally trained economists—Druckenmiller left academia to begin a career as an oil analyst, and Johnson studied economics at MIT and completed his PhD at Princeton—but they both thrived under Soros’s management style. The culture of humanities thinking at Soros Fund Management demanded that all three of them seek out the cultural context of their data.
Robert Johnson explained their unique process to me: “The data was not numbers mostly. It was not all quantifiable on spreadsheets. It was experiences, newspaper articles, stories about how people were reacting, conversations. Narrative data.”
This is what I call thick data. What makes it thick? And why does it matter when machine learning gives us a surfeit of thin data? To answer these questions, I want to highlight four different kinds of knowledge, or ways of “knowing,” that the field of philosophy helps illuminate. This will help us to recognize our bias toward thin data—or numbers stripped of their contextual meaning—and the ways in which that bias can keep us from uncovering cultural insights.
The Four Types of Knowledge
How do we know what we know? And how can we be certain we know something? Philosophers have pondered this question for millennia. Do I truly know that I am sitting on a chair, that a2 + b2 = c2, or that Shakespeare is a great poet dealing with the topic of power? To people outside the world of philosophy, it seems absurd that such questions have created over two thousand years of debate. But philosophers really do spend much of their time wondering how we know that an object falls when we drop it or whether there really is a world when we close our eyes, and their insights should not be overlooked. Let’s begin with the abstract ways of “knowing” that characterize thin data: objective knowledge.
1. Objective Knowledge
Objective knowledge is the basis for the natural sciences. “I know two plus two is four,” “I know that this brick weighs three pounds,” and “I know that water is made out of two hydrogen atoms and one oxygen atom.” There is no real perspective involved in this kind of knowledge. This is why philosopher Thomas Nagel described it as “the view from nowhere” in his 1986 book of the same name. Objective knowledge can be tested and retested with the same results. Ants, atoms, and asteroids can all be observed and measured with objective knowledge because its claims are repeatable, universally valid, and correspond to observations in reality.
The proponents of objective knowledge have offered different frameworks over time, but we can trace its history back to positivism, the philosophical movement in the nineteenth century arguing that anything might be measured without the prejudice or value judgments of the observer. It’s no coincidence that the nineteenth century—the height of the Industrial Revolution—was also the age of the machine. It was an era in which people were riding a wave of optimism based on the belief that science was rational and objective and that the human mind could conquer practically anything. In many ways it did: our advances in science modernized agriculture and transportation, made possible the transfer of goods across countries and across continents, and automated manufacturing processes to mass-produce goods that were needed for an increasingly wealthy middle class.
The idea that you could objectively measure and guarantee results fed into the production-oriented culture that was developing as companies became more focused on sharpening productivity and increasing margins. Alongside this fixation on objectivity, a new aesthetic of “realism” began to emerge. Theater artists strove to re-create entire cityscapes on the stage—life as it truly occurred, not as it we might wish it to—while writers like Zola and Flaubert focused on the objective reality of the “everyman” and “everywoman.” Even the latter’s poor Madame Bovary, a mere hausfrau with a hankering for romance, was worthy of a meticulously detailed examination.
Of course, those who know a bit about the trajectory of art and philosophy in the twentieth century recall that this certainty, objectivity, and rational thought were soon overtaken by doubt, subjectivity, and the irrational—in the form of dreams and the subconscious. The humanities moved away from the scientific “realism” of objective knowledge, as did many of our natural science fields: Einstein’s theory of relativity, for example, was a major turning point in physics. Ironically enough, however, “management science” from the world of business continues to prioritize objective knowledge over all other types of knowing. This is why big data, with its ability to objectively measure quantities, outcomes, and iterations, is so appealing. It captures everything that occurs above the threshold of awareness: the clicks and choices and likes that characterize the reductionist versions of ourselves.
2. Subjective Knowledge
After objective knowledge, there is subjective knowledge: the world of personal opinions and feelings. It is the body of knowledge studied by cognitive psychologists, a reflection of our inner lives. We know certain things about ourselves that everyone around us respects as knowledge. When we say, “My neck hurts,” or “I am hungry,” people tend to defer to our own knowledge of our bodies and our selves. When a person experiences something belonging to the realm of the senses, it can only be accepted as true knowledge for them in that moment.
There are, however, surprisingly few examples of entirely subjective knowledge. When someone is at a ball game and they see everyone around them eating a hot dog, they are much more likely to say, “I am hungry.” This is the type of knowledge that happens in between the subjective and the objective. It is knowledge about the world we share, and it characterizes much of what makes thick data so powerful.
3. Shared Knowledge
Unlike objective knowledge, this third form is not something that can be measured like atoms or distances. And, unlike subjective knowledge, this type of knowledge is public and cultural. It involves a sensitivity to our various social structures, or, to use the concept that I introduced in Chapter One, our “worlds.”
Put another way, the third knowledge is the realm of shared human experience. In Chapter Four, we will look at how this type of experience can be analyzed using the study of phenomenology: What is the Jewish experience? What does it mean to be a working woman in America? What does it feel like to migrate to the city in a rapidly urbanizing China?
For Soros and his colleagues, the third knowledge was an integral part of their big bet. They “knew” the German experience and how it was made manifest in the monetary policy following World War II. They “knew” the feel on the streets in London and how strapped the Brits felt by increases in interest rates. This is not universal knowledge; it is necessarily situational. And it is not inner knowledge but rather a shared codex. It is the being and knowing that we experience together. They were following the main event—the devaluation of the pound—but the speculative opportunities existed in the secondary and tertiary waves of events that would then occur. How might investors react to the drama that would unfold? And how would scenarios of greed or fear play out after that—what would be the response to the response?
Understanding moods, a form of thick data, was an essential element in this analysis. Moods are bigger than we are: they can take over a room, a city, or a country. We say, “I am in an anxious mood,” not “An anxious mood is in me.” This is an important distinction, because moods are inherently social. There is nothing objective about a mood, nor is there anything entirely subjective about them. Moods capture the way we all feel together, or how the way one person feels can affect those around him or her. The Soros team used their nuanced sensitivity toward mood to analyze the waves of excitement and panic that followed certain market moves after Maastricht.
4. Sensory Knowledge
The three investors also stayed attuned to a fourth type of knowledge coming from the body. This fourth knowledge tells us something about how we navigate through a lower-level understanding of our world. We see this knowledge when experienced soldiers in Iraq describe “feeling” the booby traps in their bodies upon getting near to them. Veteran firefighters recall understanding the movements of a fire through a “sixth sense,” and expert paramedics are seen grabbing for the defibrillator before any explicit signs of a cardiac arrest are visible.
Soros describes his own body as inside the market system. Like a surfer becoming one with the surfboard and, by extension, the movement of the ocean waves, he experiences market data as a kind of stream of consciousness, inextricably linked to his own perception. He asks his colleagues and employees where they experience their best calls—the neck, back, head, or stomach—and he is known for making major investment calls based on a pain in his back or a bad night’s sleep. Another investor at his firm described respiratory infections as a valuable piece of soft data about possible over-leverage in a position. When he started coughing at meetings, Soros immediately asked him: “Is it time to take some risk off?”
Synthesizing Knowledge and Pattern Recognition
While you might scoff at any correlation between bodily sensations and market knowledge, the myth of Soros as a shaman speculator should not distract us from appreciating a far more analytically rigorous process in his work. George Soros and his team were able to make killer calls not because they relied solely on a back pain but because they artfully synthesized all four types of knowledge. Most important for sensemaking, they did not prioritize any one as more valid than another.
By extracting more knowledge from the given context—using benchmarks and modeling as mere guideposts—the three traders in 1992 had access to significantly more information. Since most investors base their decision-making on models of rational behavior and equilibrium, the Soros team could predict the actions of the rest of the market players. If you know your enemy’s world—if you are empathically engaged in understanding their perspective—you can take advantage of them with great market calls.
Let’s take a moment to discuss just how radically different the information-gathering process was with Soros and his group compared to a more traditional bank or investment firm. While the group around the table at Soros Fund Management was keenly attuned to all four types of knowledge—synthesizing a data point like the German national pride along with an understanding of the British stomach for severe austerity measures—the people employed at a bank like Goldman Sachs or Morgan Stanley would have been far more likely to be working with a mathematical model designed by a highly intelligent and impeccably educated mathematician or physicist. Such models have absolutely no use for an unquantifiable data point like Norman Lamont’s sense of indignation. These models optimize only one type of knowledge—objective knowledge. They claim to take risk out of the system by mapping out various scenarios on a global scale with one core assumption underneath all the activity: markets are rational and will always—eventually—return to equilibrium. Risk and reward are balanced and therefore fair and predictable. In this optic, human actors are also always rational, with clear and preset goals.
It’s worth considering that almost all of this modeling takes place on the highest floors of glass buildings in places like London, New York, and Frankfurt, far away from the real world. The data and the models become a sort of window into the world through which the financial firm’s employees may well feel that they can understand the entire global economy. Because of the underlying assumptions about rational behavior and equilibrium, there is no real need to get out of the office. It is a clean world: clean offices, clean assumptions, and clean intentions. Millions of dollars are lost and gained every moment based on a body of knowledge that isn’t situated in any real or concrete place in the world—knowledge that only refers back to its own mathematical beauty.
Now consider how Robert Johnson, Soros’s main expert in currency trading, prepared for his role in “breaking the bank of England.” In the fall of 1991, he could sense the growing pressures of German unification, which were only augmented by the collapse of the U.S.S.R. Everyone knew that the Maastricht Treaty was going to cause the already stressed European system to unravel in some way. Johnson had a long position of approximately two billion dollars on the stability of the Finnish currency (markka), but he had started to doubt that investment. Although he might have sat in his office in New York or Paris and run the numbers through additional models, he decided instead that the best way to discern the next move was to live out of a hotel in Helsinki for the winter.
“They like to drink in Finland, if you haven’t heard,” he told me, “so I went out with them every night to a place called Café Mozart. Everyone had a lot to drink and, over the course of the winter, I got to know these people quite well. One night, they all started telling me about their simulations. These people were preparing for devaluation of the markka. I could hear it in their voices.”
The next day, Johnson walked into the Central Bank of Finland and told them that he wanted to get out of his two-billion-dollar position. They made the transaction at the Finnish Postal Bank so as not to alarm the markets and, at ten o’clock that morning, Johnson was out of his bet. He caught the next plane home to New York and upon arrival shorted—or switched his position—to bet against the Finnish currency. Only days later, the Finnish currency fell by about 18 percent. Johnson walked away with substantial gains while almost every other investor in that market took a huge hit.
“Everyone was walking up to me and saying: ‘How did you see that? How did you know they were planning to let their currency go?’” Johnson told me. “I knew because I was there. I could feel it in the room when I talked to the Finnish people—not just the officials at the central bank, but when I talked to the financial investors and the trade guys and the labor union negotiators. I learned it from real conversations with real feelings, not from fundamental mechanical economics.”
Johnson compared this type of thick data with the thin data considered “legitimate” at his alma mater, MIT: “When I took a mathematical formula from economics over to the engineering department and I put it in on the oscilloscope, it fit like a glove. When I applied it to labor markets, however, it never made any sense. ‘What are you guys doing?’ I said. ‘You’re pretending. These are human systems, not mechanical ones.’”
Johnson wistfully described his encounter at MIT with one of his mentors, famed economist and historian Charles Kindleberger: “This guy would invite us all to go the last rehearsal of the Boston Symphony on Friday mornings before the weekend performance. He would take us all out for coffee and a muffin and we’d go listen and then we’d have a sit around afterwards. This guy lived economic history: He worked on the Marshall Plan and he wrote the famous book Manias, Crashes and Panics. His pattern recognition came from circumstance, history, human stories. He told bankers that books by Defoe, Balzac, and Dickens were all important books for genuine sophistication in the field.”
A Return to Literary Economics
Economics, as a discipline, is a perfect example of an activity that benefits most from all four types of knowledge—data thick and thin. Why, then, do so many insist that it should reside solely in the realm of objective knowledge? The great economist Paul Samuelson spoke to this conflict in an interview on PBS’s NewsHour in the late nineties:
Economics is not an exact science, it’s a combination of an art and elements of science. And that’s almost the first and last lesson to be learned about economics: that, in my judgment, we are not converging toward exactitude, but we’re improving our databases and our ways of reasoning about them.
Historian Isaiah Berlin had a unique insight into this particular kind of lesson. He spent much of his academic life studying politics, seeking out a description for political insight and leadership. During the period in which he was writing his books and essays—the mid-to-late twentieth century—political scientists and economists were fixated on finding the universal laws and frameworks for all political systems. They argued that these types of theoretical concepts could guide entire social and political bodies to move forward in the manner of scientific progress. They wanted politics to play a rational game.
Berlin set out to investigate these arguments in his 1996 essay collection The Sense of Reality. Was this rational game an accurate reflection of reality? Is this the way that politics actually played out? What he discovered, throughout his investigation, was quite the opposite. Much as a master investor like George Soros is able to simultaneously synthesize inputs of inconceivable complexity, Berlin found that great political leaders had a set of personal skills that he called “perfectly ordinary, empirical, and quasi-aesthetic.” These skills were characterized by an engagement with reality founded on experience, empathic understanding of others, and sensitivity to the situation. It’s the extraordinary ability to synthesize “a vast amalgam of constantly changing, multicolored, evanescent, perpetually overlapping data, too many, too swift, too intermingled to be caught and pinned down and labeled like so many individual butterflies.”
If we follow Berlin’s argument, the gift of these investors is to be able to see patterns in a vast ocean of data, impressions, facts, experiences, opinions, and observations and to then connect these patterns into a single unifying insight. In his mind this requires a “direct, almost sensuous contact with the relevant data,” an “acute sense of what fits with what, what springs from what, what leads to what.”
This skill combines reason, emotion, judgment, and analysis, or as Berlin put it: “so many individual butterflies.” And, in the context of financial speculation, it also requires the nerve to act on all four types of knowledge.
Discipline and Action
On October 19, 1987, the Dow Jones Index lost 22.6 percent of its value, the biggest drop for the index since its incarnation in 1896. Market watchers were trying to read the signs in the precipitous drop of the day, soon dubbed “Black Monday.” Druckenmiller thought he could discern a pattern familiar from previous crashes: based on experience and historical knowledge, he theorized that a sharp fall in the markets would be followed only days later by a rally and then a dramatic fall. Soros, on the other hand, read the market signals as a symptom of financial engineering. He was convinced that the newly developed financial product called portfolio insurance had produced a destabilizing feedback loop in the markets. The product was meant to protect investors against significant losses in the case of a market fall, but when thousands of investors purchased the product, it triggered a free-for-all in futures selling. Extreme movements in this seemingly small corner of the market—like the wings of a butterfly in Japan—ultimately created enormous market volatility.
Because Soros was convinced that Black Monday had been caused by these complex feedback loops and not from a fundamental shift in the market bottoming out after the bubble, he kept his position and remained bullish. On Wednesday, October 21, however, Soros watched his bet to short the Japanese yen fall flat as the market in Tokyo was starting to rise. He realized that the U.S. market was, in fact, bottoming out and that the oscillations caused by portfolio insurance were only a small part of a bigger bust cycle. His fund was hemorrhaging money and he would soon lose credibility with his investors. Only weeks before, his Quantum Fund had been up 60 percent for the year, and now it was down 10 percent. Soros got out of his positions in a fire sale so large it shifted the entire movement of the market. In a matter of days, $840 million from his fund went up in smoke.
He was not alone. Storied hedge fund managers were taken down right and left as the market made a mad dash for the bottom. Soon after the crash, Soros attended a party with several prominent investors. Unsurprisingly the atmosphere in the room was morose. Billionaire investors—once masters of the universe—were despondent over their inability to predict market volatility. Legendary hedge fund manager Michael Steinhardt greeted guests while lying in the fetal position. He told colleagues that he was ready for a career and lifestyle change.
Soros was keenly aware of the mood of the hour—and the markets—but he did not allow himself to succumb to them. He used his intellectual acumen to act in the midst of the funereal atmosphere. He could see that Alan Greenspan, then-head of the Federal Reserve, would ease credit massively to allow for more market movement. While most shaken investors remained in recovery mode that week, Soros never lost his nerve. He made a swift move shorting the dollar in the Foreign Exchange Market. It was a position of pure bravado, another bet on other people’s bets just days after the losses of Black Monday. But, he later described, his mouth literally watered “like one of Pavlov’s dogs” at the speculative opportunity. As anticipated, the dollar fell and his move was a success. By the end of 1987, Quantum was up again, this time by 13 percent.
Most of the other investors at that party in 1987 had access to the exact same information and economic models as George Soros. Everyone knew that the Federal Reserve would ease monetary policy to avoid a complete shutdown of the market. If you had asked any one of the power players—whether in the fetal position or drowning their sorrows in a drink—they would have predicted the exact same series of events. The difference is that Soros had cultivated a disciplined mode of practice: he was able to remain dispassionate toward his recent losses because he was focused on the new speculative opportunity. By removing himself and his wounded ego from the context, he only felt what the market felt. And that feeling was one of tremendous opportunity.
Soros trained himself to rigorously stay open to all types of knowing. In using his body and the cultural context, he kept himself from growing too attached to the objective knowledge of the numbers: “Confidence is not a choice, but what George Soros does is a choice,” Johnson told me. “It’s his underlying belief in his ability and the choice he makes to act on that ability. That’s what I marvel at: his ability to act. It’s not that he can logically recognize the most lucrative actions to take; many highly skilled people in our field can do that. It’s the fact that he actually takes the action. He doesn’t stand aside from the battle to observe. He acts from inside it.”
From the Atmosphere to the Stratosphere
Chris Canavan, who has a PhD in economics from Columbia University and currently works with Soros, differentiated his own tendencies from those of his boss: “When I was a trader, like Soros, I remember having hunches or an inchoate sense of how a market might move or would move. And in retrospect, I would realize that I was right and I was responding to something that was going on around me. But I wasn’t willing to listen to it; I wasn’t willing to allow it to alter what I was thinking. And I say ‘thinking’ in a kind of empirical way. You have the skill to intuit that you’re at an inflection point and you may even have a sense about which way something is going to break, but I was still too timid to take advantage of that information or knowledge, because I couldn’t express it numerically. And few of us have the ability to tell others around us that we’re betting on a position not because the numbers are telling us to do so but because we just know it’s the right thing to do.”
Canavan likens it to the mastery he witnessed as a young golfer immersed in the game. “The better I got at golf,” he told me, “the more I began to appreciate just how truly great the great golfers really were.” He explained the startling perception shift he experienced in the presence of golf masters. Though he was moving closer to the elite athletes in terms of his skill level, his advancement only showed him how truly far ahead they were in their understanding of the game.
“They achieved this exponential lift-off and they moved into a stratosphere. But unless you were a good enough golfer, you didn’t really appreciate that they were in a stratosphere and not just the atmosphere.”
This “lift-off” was only made possible through the courage to act on all types of knowledge. “An average player,” Canavan explained, “just doesn’t realize how complicated it is to factor in all of these variables—wind, temperature, the grain of the grass, all of this—and then decide how to hit a shot. But, perhaps most important of all, to have the courage to hit that shot knowing how narrow your margin of error is.”
Not all of us can personally relate to mastery in golf or, say, jazz piano, but most of us have seen mastery at work in the way we acquire knowledge of a new language. The newcomer to German will need to start with learning the rules of the language, or the grammar. The student is concerned with questions like, How is this language organized? What are the rules, and what are the exceptions to the rules? After some time immersed in the vocabulary, the student begins stringing sentences together, working diligently to avoid mistakes of grammar. Soon enough, the assiduous student becomes accustomed to rules like always putting the verb at the end of those (very long) German sentences. And then, just as Canavan described, the student leaves the atmosphere and zooms into the stratosphere. Like magic, she lets go of all thought and relaxes into fluency. The rules themselves—those principles of abstraction—withdraw and she speaks in the new language with a kind of fluidity and porousness: she bends the language to her will, plays with it, owns it as a tool of expression. The experience of searching for a vocabulary word is now entirely in the background, and a focus on the meaning of what she wants to say is front and center of her attention.
In this example of mastery, as in so many others I will turn to throughout the book, we can refer, again, to philosopher Hubert Dreyfus. For Dreyfus, and for other philosophers working in the phenomenological tradition, our greatest skills and innovations are not the result of conscious thought. Though this seems obvious when we discuss acquiring a new language, it is actually in radical contradiction to the prevailing norms in many corporations, institutions, and even education systems that explicitly and tacitly assert that our greatest skills are exhibited when we are sitting alone thinking abstract thoughts.
The Power of Practical Wisdom
Aristotle’s notion of phronesis—or practical wisdom—helps us understand this better. He argued that the man of practical wisdom is able to transcend the “grammar” of the field in which he works. He no longer needs the training wheels of rules or models; instead he reads the context as concrete and situated. He does the appropriate thing in the appropriate way at the appropriate moment. When Canavan describes a master golfer, he is really describing the concrete reaction of a player to a specific situation. We might even say that his context pulled the action out of him. And, because he was a master, he was skilled enough to stay receptive to the meaning of the moment.
As Canavan put it, “The more you become a slave to this Cartesian way of thinking, the less willing you are to listen to your hunches and your neck aches and your upset stomachs. And that is going to make you a less successful, not more successful, trader or speculator. We operate with this paradox: the prevailing narrative is that the more scientific the information the better, and yet it shuts us off to all sorts of information that comes in a different form. But it’s the people who don’t shut themselves off to that information who, many times, outperform the others. And outperforming very often means it’s a zero-sum game: my gain is your loss.”
A senior trader I spoke with at the Soros Fund described his own experience of pulling from all four data streams. The trader was an extremely skilled advisor to the Brazilian government as well as a decades-long veteran in the South American markets. He used both of these experiences to take a risky leveraged position on Brazil in 2001. Things were looking good for him in the first quarter, but then his position started to lose money. And then it lost a lot of money. After six months had passed and several hundred million dollars had been lost, Soros called the trader up on the phone and asked him to stop by.
“My first thought was, well, I’ve had a good run and at least now I can say that I used to work for George Soros,” he told me, laughing.
But when he met with Soros in person and tried to explain his reasoning for the disastrous position, something altogether different occurred. Soros stopped him mid-explanation after only fifteen seconds. He told the trader that he should keep his position; he said, “Wait until it can’t get any worse, and at that moment, double your position.” He was relieved but also flabbergasted at this directive from Soros. How was he to “know” when it couldn’t get any worse?
He waited and watched the markets. Two weeks later, the trader looked at his position and there it was: he was filled with a certain understanding that the market really had bottomed out. At that moment, he turned back to the advice given to him by George Soros. At the very bottom of the market, when everyone was running from the position, this trader took a long breath and doubled down. He waited.
Consider for a moment how hard it was for a trader to double down and wait in the midst of the herd mentality to escape. Everything in his being was screaming out in panic: “Get out now.” Instead, however, he used discipline to stay calm. He attempted to detach his own emotions from his involvement with his position.
And then, almost like magic, the market turned. No one was surprised by the fact that the markets turned: what goes down must eventually go up. What was shocking, however, was the way the market turned immediately after his experience of hitting the bottom.
“My years of looking at the markets taught me how to recognize it,” he told me. “But George Soros was the one who showed me how to make big bets against the grain on that feeling. Our bet wasn’t based on analytics but on the years and years of seeing the patterns in the market. This was something I could never codify or even explain really. But we got back what we lost several times over.”
Chris Canavan described to me his most profound encounter with these alternative ways of knowing, or soft data. In 1997, he left his academic career behind and landed at Goldman Sachs, where we often spent time on the commodities trading floor. It was there that he started to puzzle over some of the truths of his world: the average commodities traders—traders in goods like gold, oil, natural gas, and palladium, for example—were generally about three to five years older than the traders on the foreign exchange floor. The commodities traders were also generally perceived to be less intelligent than the foreign exchange traders. They didn’t have the same pedigrees: the Ivy League schools, the prestigious internships. And yet, it was also well known that the commodities trading unit was a veritable powerhouse at Goldman; it had been one of the most profitable commodities businesses in the field for decades. Why were these guys, objectively not the smartest guys in the room, systematically landing the better trades?
Then one summer, a hurricane ravaged the Gulf Coast and headed inland toward Louisiana. Canavan watched as the commodities market went haywire. And yet, in the midst of this incredible volatility, he noted that the commodities traders were not just continuing to trade, they were trading at their best.
“These were guys who had been doing nothing but trading crude oil or refined products—stuff going into and out of a refinery—for the last twenty years,” Canavan told me. “And they were in their element because they could summon up relevant information—without even looking any of it up—about all of the rigs along the path of the hurricane. But not only that, they knew which refineries along the Gulf coast and the Atlantic coast those rigs supplied.”
Canavan realized that none of the traders were working from mathematical models. They weren’t numerates searching for the best equation to objectively make sense of the market activity. Instead, they were completely immersed in the movements of the markets.
“They would say, ‘Because I know where everything is and where everything is flowing, I can picture in my mind what is about to happen because I can just see the hurricane and its consequences unfolding. From there, I can map out the consequences that will give me a sense of which way the prices of crude and other kinds of products will move. And then I can take advantage of those anticipated moves by trading them.’”
During that period, he watched as the commodities traders made a profit throughout one of the most volatile periods of the decade. The traders on the foreign exchange desks, on the other hand, fared quite differently in the midst of instability.
“The general idea was to teach a sharp twenty-three-year-old college graduate a few basic mathematical concepts about how foreign exchange rates get set—or at least how we assume that they get set—and to give them some powerful models to compute these things pretty quickly. That person can start trading and make some money without having to accumulate bespoke and difficult-to-quantify information because of the nature of the market.
“I used to think: these young whippersnappers will take down the crude oil traders any day of the week.” After the hurricane, however, he changed his mind about almost everything he thought he knew.
“Up until that moment, I believed the orthodoxy. Here I was: I had a PhD in economics and I had been in academia for a while. I thought a good trader needed to find a better model than the next guy. And I thought all of the work was in inventing and improving and refining the next sophisticated statistical technique to make that better model.
“After that hurricane, I realized that all the skills of this kind of foreign exchange trader get frozen the minute something unexpected occurs. I thought, ‘I have to completely invert my sense of what a good trader is.’
“Someday soon we will approach a point where all quantifiable information will be processed by all market participants simultaneously and instantaneously. All that requires is faster fiber optic cable, more memory, and models. These models are proprietary for a day and a half and then everyone else has them. At that point, what distinguishes a systematically good trader from a systematically bad trader?”
The Cartesian prediction—based entirely on a rational market—is that all traders will begin flipping coins, no one any better than another. Canavan’s prediction is different. He feels that, even in the midst of utter transparency, some traders will still be better. And some will achieve extraordinary results over and over again. These will be the ones working from real life—pulling and synthesizing data thick and thin. They will draw insights from understandings of culture, firmly grounded in the reality of each and every situation. After all, human intelligence at its best is never just a flip of the coin.
In the next chapter, we’ll look at where we can access the type of thick data that matters most. Cultural study requires a methodology that acknowledges the complexities of our world. It’s time to turn away from the promises of false abstractions and immerse ourselves fully in the richness of reality. We’ll start with the study of an apricot cocktail.