The Theory of Reflexivity
On the Unstable Dance Between Perception and Reality
George Soros, now in his 90s, was once one of the most successful hedge fund managers in the world. His flagship, the Quantum Fund, generated outstanding returns, averaging around 30% per year over decades.
However, Soros is probably best known as the man who “broke the Bank of England.” In his legendary bet in 1992, he and his partner Stanley Druckenmiller shorted the British pound with such enormous volume that the Bank of England simply couldn’t hold it anymore. It gave in, and the pound collapsed. Soros reportedly made over a billion dollars in a single day.
Soros was also a self-declared philosopher, writing about life, politics, and market theory. Many of his ideas are controversial and were dismissed by the academic mainstream. The theory your’re about to read included.
HOWEVER that does not mean it‘s useless! In fact, the opposite is true.
The theory of reflexivity, the central concept of Soros’ book The Alchemy of Finance, which he often described as his life’s work, is one of the most revolutionary theories on how psychology affects financial markets.
Once understood, you won’t see markets the same way again.
Before we get to the essence of this theory and how it plays out in markets, we need to take a step back and look at the assumptions of standard market theory. And more importantly, where those assumptions fall apart.
I. Anti-Equilibrium
In classical economics, you are told that markets tend toward equilibrium. Prices fluctuate only temporarily before gravitating back to their "fair value."
This notion, so often assumed by economists and value investors alike, rests on an elegant, comforting idea: that beneath the noise, there is order. That fundamentals are fixed points, and prices are mere shadows dancing around them.
But this is not how markets behave.
Anyone who trades in markets where prices move continuously knows: the participants themselves are not just neutral observers.
Rising prices attract buyers, falling ones repel them. Expectations don't just respond to market movements, they also drive them. In this way, what appears to be a path toward equilibrium is often a response to a moving target.
Soros challenged this core assumption. He observed that in financial markets, future expectations shape current decisions, and those current decisions, in turn, alter the future.
A rise in a stock price, for example, doesn’t just reflect optimism about a company’s prospects; it can also affect those very prospects. Rising valuations may lower borrowing costs, boost executive confidence, attract talent, or open doors to mergers and acquisitions. Therefore the stock price does not merely reflect reality. It becomes part of it.
Since such feedback loops are not compatible with mainstream economic theory, it often treats them as anomalies.
This is where Soros’ theory of reflexivity begins.
Once understood, it reshapes not only how we interpret market movements, but also how we behave within them.
II. The Problem of Imperfect Understanding
In order to understand reflexivity, we must understand an even more fundamental problem: the limits of our understanding.
Particularly in systems that contain thinking participants, such as markets, economies, and societies, where our perception of reality is never fully objective, and can never be.
Participants do not observe the system from the outside. They are within it. Their decisions shape the very reality they attempt to understand. This creates a fundamental distortion: there is no neutral ground, no fixed point of reference from which their expectations can be measured or verified.
The act of thinking is not detached from the system. It is part of it.
To make this clearer, we have to distinguish between two roles: natural scientists and participants.
The natural scientist studies phenomena that exist independently of human thought. Such as the behavior of atoms or the orbits of planets. These follow laws that are unaffected by whether or how they are observed. A scientist can test a hypothesis against an external reality and determine whether it is true or false. There is an objective standard against which knowledge can be measured.
Participants in social systems do not enjoy this privilege. Their thoughts and actions influence the very reality they are trying to understand. In economics, finance, or politics, there is no fixed backdrop. The canvas shifts as the brush moves. The system contains no neutral observer. It is populated entirely by participants whose beliefs and actions shape the environment in which they operate.
This gives participants the dual role of observer and actor as they fulfill two functions simultaneously:
Understand the situation.
Act on the basis of this understanding.
However, both roles hinder each other, as the understanding is never neutral and the action influences the reality one is trying to understand.
Therefore, you can say that there is no independent reality on which you could measure understanding. Reality bears the traces of imperfect understanding.
It’s a structure that is also found in quantum physics, with Heisenberg’s uncertainty principle. It describes the phenomenon that in subatomic systems, the act of observing a particle changes its position or momentum.
The described problem leads us to a central insight in Soros' theory: The Participant Bias.
Participant bias refers to the systematic distortion in perception and reasoning that arises when individuals act within a system that their own actions help to shape. In reflexive systems, this bias is not an anomaly, it is the default condition.
This is the gap that Soros seeks to address: Economic theory is built on assumptions of perfect knowledge and passive observation. It fails to grasp systems in which perception itself is a driving force.
III. The Concept of Reflexivity
The connection between a participant’s thinking and the situation they are thinking about can be broken down into two distinct processes.
The first is the cognitive function, which describes the attempt to understand reality. In this role, perception depends on the situation.
The second is the participating function, which describes the impact the actor‘s thinking has on the world. Here, the causal direction is reversed: the situation is shaped by perception. Beliefs trigger decisions, and those decisions alter the world they were meant to interpret.
„It can be seen that the two functions work in opposite directions: in the cognitive function the independent variable is the situation; in the participating function it is the participants' thinking.“
Ordinarily, a function needs an independent variable to produce a determinate outcome. But when both functions operate simultaneously, as they do in markets, societies, and other reflexive systems, each becomes the other’s input. The independent variable of one is the dependent variable of the other. As a result, the two functions begin to interfere with each other.
What emerges is a reflexive loop: an ongoing, recursive interaction in which both the situation and the participants‘ views are dependent variables. Hence perception shapes reality, and reality feeds back into perception.
Sketch of Soros:
Therefore:
„The two recursive functions do not produce an equilibrium but a never ending process of change. When a situation has thinking participants, the sequence of events does not lead directly from one set of facts to the next like in natural sciences; rather, it connects facts to perceptions and perceptions to facts in a shoelace pattern.“
In this way, reflexivity stands in direct opposition to the dominant concept of equilibrium. Equilibrium assumes that expectations remain stable until adjusted by real-world changes. Reflexivity suggests the opposite: that expectations are themselves world-altering forces. That the map redraws the territory, even as the territory reshapes the map.
Reflexivity, then, is not a special feature of rare market anomalies. It is the underlying structure of any system populated by thinking agents.
Soros himself notes that the implications of this idea stretch far beyond the realm of finance. And indeed, they do as we’ve already seen echoes of it in quantum physics. For now, however, we stay within the markets. Since that is where our interest as investors lies. And it is there that this concept reveals itself most clearly.
REFLEXIVITY IN THE STOCK MARKET
IV. Reflexivity in Markets: Mispricing as a Mechanism
Few environments are as well-suited for observing the dynamic between perception and reality as the stock market. Prices are tracked in precise, quantitative detail. Expectations are recorded in analyst reports and investor letters. Behavior is easily observable. And the data is widely available.
In fact, the stock market comes remarkably close to fulfilling the conditions of perfect competition as imagined by economic theory. There is a centralized marketplace, standardized instruments, near-instant information flow, and a vast crowd of participants. In theory, no single actor can move prices unilaterally.
And yet, paradoxically, it is precisely here that equilibrium theory fails most visibly.
At the heart of traditional valuation lies the belief that every stock has an intrinsic value, and that market prices tend to converge toward this value over time.
With the theory of reflexivity, Soros flips this on its head.
He argues that valuations are not passive reflections of underlying reality, they have a direct way of influencing underlying values. As share prices can affect management decisions, financing options, employee morale, and customer perception. A soaring stock price may enable acquisitions, unlock cheap capital, or boost brand prestige. A collapsing one may do the opposite. In both cases, the market doesn’t merely record the fundamentals. It helps write them.
In Soros’ words:
„I do not accept the proposition that stock prices are a passive reflection of underlying values, nor do I accept the proposition that the reflection tends to correspond to the underlying value. I contend that market valuations are always distorted; moreover, and this is the crucial departure from equilibrium theory, the distortions can affect the underlying values.“
This is the essence of reflexivity in markets: a feedback loop between expectations and outcomes becomes a structural driver of change. Prices influence the very companies they are supposed to measure. And this recursive interaction can push markets far from any theoretical notion of “fair value.”
In a moment, we’ll see how this dynamic plays out in short-term stock price movements.
Soros made another very interesting statement, which again emphasizes his fundamental skepticism toward academic market theory:
„Existing theories about the behavior of stock prices are remarkably inadequate. They are of so little value to the practitioner that I am not even fully familiar with them. The fact that I could get by without them speaks for itself.“
Quite interesting, considering his $7 billion net worth.
V. The Reflexive Spiral: Boom, Bust, Belief
Now, having laid the theoretical groundwork, we move from critique to construction.
Soros begins with a deceptively simple proposition: market participants are always biased.
Not occasionally, not only in times of greed or fear, but always. This is not meant as a moral critique, but as a structural feature of markets populated by thinking participants. Every decision is based on some belief about the future, and no belief can ever be free of distortion.
Soros uses this to explain one of the most persistent myths in finance: that markets have a near-magical ability to anticipate events. As it's often said that the stock market "predicts" recessions. (Perhaps it's more accurate to say that the market helps bring them about.)
Soros replaces the notion that “markets are always right” with two ideas:
Markets are always biased in one direction or another.
Markets can influence the very events they seem to anticipate.
Taken together, the combination of these two ideas may explain why markets so often appear to anticipate events correctly.
Using the participants’ bias as a starting point, Soros builds a model of how the participants views and reality interact. The difficulty, again, lies in the fact that these beliefs are themselves part of the situation to which they relate.
In practice, the stock market contains a variety of different views, optimistic and pessimistic.
Since many individual biases cancel each other out, Soros introduces the concept of a prevailing bias: a dominant collective tendency that emerges through aggregation.
Because these biases are expressed in a defined action, buying and selling, they find direct reflection in prices. A positive bias leads to rising prices; a negative one to falling prices. This makes the prevailing bias an observable phenomenon, not just a theoretical one.
At this point, Soros adds a second key concept: the underlying trend. This refers to any actual development, like earnings growth, macroeconomic shifts, or technological changes, that influences prices, whether or not it is recognized by the market.
„The trend in stock prices can then be envisioned as a composite of the "underlying trend" and the "prevailing bias."“
Through the cognitive function, the underlying trend influences how participants perceive the market. As a result, these altered perceptions affect the situation and feed back into the system through the participating function. Affecting not just prices, but potentially the trend itself.
Now, we have a reflexive relationship, where stock prices are not just determined by a single variable. Instead, they are shaped by two factors: the underlying trend and the prevailing bias. Which are both, in turn, affected by the stock price.
However, this system, the interplay between stock prices and the other factors, lacks a constant. There is no fixed point of reference. What is supposed to be the independent variable of one function becomes the dependent variable in the other. And without a constant, there can not be a tendency towards equilibrium.
What we are left with is not a mechanism moving toward balance, but a process of change. A looping sequence in which none of the variables remain stable. Instead, they reinforce one another, first in one direction, then in the other. This is the anatomy of a classical boom-and-bust cycle.
We’ve seen this structure repeatedly, and often at scale, in credit markets, in speculative bubbles, and in the lead-up to systemic crashes like those of 2000 and 2008.
To formalize this process, Soros offers a simple set of definitions.
When stock prices reinforce the underlying trend, we call the trend self-reinforcing.
When they work in the opposite direction, we call it self-correcting.
The same terminology applies to the prevailing bias. When it amplifies the divergence between expectations and reality, it is self-reinforcing. When it narrows the gap, it is self-correcting.
Prices themselves are described simply: as rising or falling. A positive bias helps drive prices upward. A negative bias reinforces their decline.
„In a boom-bust sequence we would expect to find at least one stretch where rising prices are reinforced by a positive bias and another where falling prices are reinforced by a negative bias. There must also be a point where the underlying trend and the prevailing bias combine to reverse the trend in stock prices.“
It’s important to understand that this is not a closed system that is able to sustain itself. It is a spiral. And once it gains momentum, upward or downward, it does not slow until it has exhausted itself.
VI. Anatomy of a Reflexive Cycle
Let us now take a look at what a rudimentary model of boom and bust might look like.
It starts with an underlying trend that has not yet been recognized by market participants. This trend may reflect improving fundamentals, like improving earnings, or some latent structural change. At the same time, there may also exist a prevailing bias among investors, maybe one that is initially negative and not yet reflected in prices.
As the trend begins to assert itself, the market eventually takes notice. The shift in perception causes stock prices to rise. What happens next depends on whether those rising prices affect the underlying trend. If not, the price reaction may remain limited. But if they do, if, say, higher valuations allow a company to raise capital, hire talent, or expand operations, then a self-reinforcing feedback loop begins.
This enhanced trend then feeds back into investor sentiment and the prevailing bias adapts. At this point, one of two things may happen: the market either expects continued acceleration, or it anticipates a correction. If the latter occurs, the price pullback may test the trend. Sometimes the trend survives, sometimes not.
But if the former happens, if expectations turn more bullish, a positive bias develops and prices rise further. This in turn reinforces the underlying trend, which now becomes not just stronger, but increasingly tied to stock prices themselves. And this interdependence breeds a new fragility: the system becomes vulnerable.
As long as the bias remains self-reinforcing, expectations continue to climb, often even faster than prices. Eventually, prices can no longer sustain the level of belief they once inspired and a correction sets in. Disappointed expectations weigh on sentiment, stock prices falter, and the underlying trend begins to weaken. If that trend has become too dependent on the momentum of prices, the correction may deepen into a full reversal, which causes expectations to fall even further.
At that point, a self-reinforcing downturn begins. The system enters a negative spiral that mirrors the optimism that preceded it. Eventually, this decline also reaches an endpoint, a point of exhaustion, or perhaps recognition that the pessimism has overshot. The cycle stabilizes, new opportunities arise, and the loop resets.
Soros visualises this process in a diagram of two curves. One represents stock prices. The other tracks earnings per share, as a rough proxy for fundamentals and thus for the underlying trend. The divergence between the two curves can be envisioned as an indication of the underlying bias. Even if they appear to move in parallel, their relationship is far from straightforward.
A typical path for the two curves may be as follows.
AB: The underlying trend begins to materialize in earnings, even before the market notices.
BC: Recognition sets in. The trend is reinforced by rising expectations.
CD: Doubts emerge, but the trend holds. A correction may occur, but is ultimately shrugged off.
DE: Confidence solidifies. Expectations become deeply embedded, even resistant to temporary disappointments.
EF: The system overheats. Expectations detach from reality, and fail to be sustained by reality.
FG: Reality asserts itself. The bias is recognized as such and expectations are lowered.
G: Prices collapse. The final prop of sentiment gives way.
GH: The underlying trend reverses. Decline reinforces decline.
HI: Exhaustion sets in. Pessimism becomes overdone and the market stabilizes.
→ Adapted from a diagram in George Soros’ The Alchemy of Finance
This sequence does not capture every possible nuance. There are many more possible ways the interplay of an underlying trend and a prevailing bias could turn out. But the model beautifully illustrates the anatomy of reflexive cycles and shows how expectations and fundamentals interweave, sometimes resulting in vicious loops.
Of course, the model remains partial. The notion of an "underlying trend" is a vague placeholder for the fundamentals. It is very hard, in this context, to define what fundamentals really are. Whether its earnings, dividends, asset values or free cash flow, all of these matter. Yet their importance varies depending on context, sentiment, and prevailing bias.
Even the process of measuring them is subjective. Security analysts have debated these definitions for decades, without ever reaching a clear resolution.
Fortunately, we don’t need to settle that debate here. Soros is not attempting to replace fundamental analysis, only to complete it. Even without a clear definition of fundamentals, the reflexive model allows for important generalizations. The most critical of which is this:
Stock prices, through reflexive dynamics, influence fundamentals. This is what allows boom-bust cycles to unfold.
That’s the whole point. If prices had no feedback effect, they were merely shadowed fundamentals, and we wouldn’t see the exaggerated cycles that characterize real markets.
So yes, the model is rough. It lacks mathematical precision and it does not capture every nuance. But that is beside the point.
VII. A Final Reflection: The Map Is the Territory
In the end, the theory of reflexivity does not aim to replace fundamental analysis. It simply provides something that has been missing and is useful for every investor: a recognition that the observer is also a participant.
While fundamental analysis offers a static snapshot of value, reflexivity reveals the feedback loops that shape that value over time.
At its core, reflexivity challenges the deepest assumptions of economic thought. For decades, economics modeled itself after Newtonian physics, searching for timeless, universal laws to govern behavior. But markets are not mechanical systems. They are human systems. They do not operate on perfect information, but on belief, bias, narrative, and noise.
As Soros puts it:
“People’s thinking affects reality, and reality affects their thinking.”
In markets, in politics, in societies, we are never outside the system we are trying to understand. We live inside the loop. And the loop is unstable.
This recognition is both humbling and empowering.
It is humbling because it reminds us that no theory, no matter how elegant, can fully capture the world. All knowledge is provisional. Every model is a simplification. Every map omits more than it shows.
But it is also empowering. Because if reality is shaped by perception, then our actions matter more than we think. Our views are not just passive reflections. They are instruments of change.
The map is not the territory. In reflexive systems, sometimes, the map redraws the territory. And sometimes, it burns it down.
Disclaimer: This essay is not intended as an original contribution. It is a condensed and interpretive summary of George Soros’ work, particularly his theory of reflexivity as developed in The Alchemy of Finance. All core ideas, concepts, and arguments presented here are derived directly from Soros’ writings. This text should be understood as an extended quotation, an attempt to restate his thinking in accessible form, not to claim authorship of it. All credit belongs to George Soros.