The Observer Problem

Why markets can’t be predicted — and what to do instead.

The Short Answer

Markets suffer from an observer problem: the act of measuring them changes them. Unlike physics, where electrons don’t read your research papers, market participants adapt to any pattern that becomes known. This is called reflexivity — and it’s why prediction is philosophically flawed in financial markets. The solution isn’t better prediction. It’s abandoning prediction for description.

The Quantum Mechanics Parallel

In 1927, Werner Heisenberg formalized the uncertainty principle: you cannot simultaneously know both the position and momentum of a particle with perfect accuracy. The act of measuring one disturbs the other.

This isn’t a limitation of our instruments. It’s a fundamental property of reality at the quantum scale. Observation affects outcome.

For decades, this remained a curiosity of physics — fascinating but seemingly irrelevant to everyday life.

Then George Soros applied the same logic to markets.

Soros and Reflexivity

George Soros isn’t just one of history’s most successful traders. He’s also a philosopher — trained at the London School of Economics under Karl Popper, the father of scientific falsifiability.

Soros developed a theory he called “reflexivity.” The core insight:

In markets, participants’ beliefs affect fundamentals, and fundamentals affect beliefs — creating a feedback loop that makes equilibrium impossible.

Traditional economics assumes markets move toward equilibrium. Supply meets demand. Prices reflect information. The system is self-correcting.

Soros argued the opposite: markets are inherently unstable because observation and reality are entangled.

Here’s a simple example:

Traders believe a stock will rise
They buy the stock
The price rises (confirming their belief)
More traders believe it will rise
They buy more
The price rises further
The belief becomes “true” — not because of fundamentals, but because of the belief itself

This works until it doesn’t. Eventually, the gap between price and reality becomes unsustainable. The bubble pops. But the timing is unpredictable precisely because the bubble was created by collective belief, not by measurable fundamentals.

Goodhart’s Law: When Measurement Becomes a Target

British economist Charles Goodhart formalized a related problem in 1975:

“When a measure becomes a target, it ceases to be a good measure.”

Originally applied to monetary policy, Goodhart’s Law explains why market patterns self-destruct.

Say you discover that stocks with low price-to-earnings ratios outperform. You publish this finding. Other investors read it. They start buying low P/E stocks. Demand increases. Prices rise. P/E ratios rise. The “cheap” stocks become expensive. The pattern disappears.

This isn’t a flaw in your analysis. The analysis was correct — until it was known.

Every published trading strategy contains the seeds of its own destruction. The act of observing the edge eliminates the edge.

The Replication Crisis in Finance

Academic finance is experiencing its own reckoning with this problem.

In 2016, researchers attempted to replicate 452 published trading strategies. The results were devastating: most strategies that looked profitable in the original papers showed no significant returns when tested on new data.

The reasons:

Data mining: researchers tested hundreds of patterns and published the ones that worked by chance
Lookahead bias: subtle leakage of future information into historical tests
Survivorship bias: only testing assets that still exist
Regime change: market conditions shifted between the study period and today

But there’s a deeper problem: even legitimate edges disappear once published because other traders pile in and arbitrage them away.

The observer problem isn’t just about measurement error. It’s about the fundamental nature of a system where the observers are also participants.

Why Prediction Is Philosophically Flawed

Let’s trace the logic:

Markets are moved by participant behavior
Participant behavior is influenced by predictions
Predictions, once known, change participant behavior
Changed behavior changes the market
The prediction changes what it was predicting

This is a self-referential loop — similar to the paradoxes that arise in logic and mathematics when a statement refers to itself.

“This statement is false.” “The market will go up because everyone expects it to go up.”

Both collapse under their own weight.

This doesn’t mean markets are random. Patterns exist. But patterns in an observed system are fundamentally different from patterns in an unobserved system.

The physicist can study electrons confident that the electrons aren’t reading physics journals. The trader cannot.

The Alternative: Description Over Prediction

If prediction is philosophically flawed, what’s left?

Description.

Instead of asking “Where will the price go?”, ask “What is happening now?”

Prediction: “BTC will hit $150K by October”
Description: “BTC momentum is strengthening, volatility is compressing, volume confirms the trend”

The prediction claims to know the future. The description captures the present state with accuracy.

Here’s why this matters:

Description can be verified immediately — it’s either true or false right now
Prediction can only be judged in hindsight — and by then it’s too late
Description is robust to regime change — you’re not betting on patterns persisting
Prediction creates liability — you’re wrong by definition half the time

The best traders don’t predict. They read conditions and respond. They’re like surfers — they don’t predict where the wave will go, they feel its current state and adjust.

Implications for Systems and Models

This doesn’t mean quantitative analysis is useless. It means the goal of analysis should shift.

Old goal: Find patterns that predict future prices New goal: Measure current market state with clarity

Useful questions:

Is momentum increasing or decreasing?
Is volatility expanding or contracting?
Is volume confirming or diverging from price?
Are we in a trending or mean-reverting regime?
Where is risk concentrated?

These are questions about the present. They can be answered with data. And they don’t self-destruct when published because they’re not claiming to predict — they’re claiming to observe.

The MarketCrystal Philosophy

This is why we built MarketCrystal the way we did.

We don’t tell you where the market is going. We tell you where the market is.

Current trend direction and strength
Momentum acceleration or deceleration
Volatility regime
Volume confirmation
Key levels and inflection points

Armed with a clear view of the present, you make your own decisions about the future.

We’re not selling prophecy. We’re selling clarity.

Key Takeaways

Observation affects outcome — unlike physics, market participants change the system by observing it
Reflexivity creates instability — beliefs affect fundamentals, fundamentals affect beliefs, in an endless feedback loop
Goodhart’s Law kills edges — any pattern that becomes a target ceases to be a reliable pattern
Prediction is self-referential — the prediction changes what it’s predicting
Description is more robust — measuring current state is verifiable and doesn’t self-destruct

Part 1: “When Physicists Invaded Wall Street” — how physics built modern finance
Part 2: “The Map Is Not the Territory” — when the models met reality
Part 3: “The Observer Problem” — you are here

Stay tuned for Part 4: “Reading the River, Not Predicting the Rain” — where we get practical about implementing a description-first approach to markets.

Frequently Asked Questions

What is reflexivity in financial markets?

Reflexivity is a theory developed by George Soros describing how market participants’ beliefs affect market fundamentals, and how those fundamentals in turn affect beliefs. This creates a feedback loop that makes markets inherently unstable and unpredictable. Unlike physical systems where observation doesn’t affect outcome, markets change based on what participants believe about them.

What is Goodhart’s Law in trading?

Goodhart’s Law states that “when a measure becomes a target, it ceases to be a good measure.” In trading, this means that once a profitable pattern is discovered and widely known, traders exploit it until it disappears. The act of targeting an edge eliminates the edge, which is why published trading strategies often stop working.

Why can’t markets be predicted?

Markets can’t be reliably predicted because predictions change what they’re predicting. When traders believe prices will rise, they buy, which causes prices to rise. This self-referential loop means that any prediction that becomes known affects the outcome. Additionally, markets are open systems where participant behavior constantly adapts to new information, making historical patterns unreliable guides to the future.

What is the observer problem in markets?

The observer problem in markets refers to the fact that market participants are both observing and participating in the system they’re trying to measure. Unlike physics, where observation doesn’t affect the phenomenon being studied, in markets the act of measuring (and acting on those measurements) changes the thing being measured. This is why strategies work until they’re discovered, then stop working.

What is the difference between market prediction and description?

Market prediction claims to know future prices or directions (“BTC will reach $100K”). Market description measures current conditions without claiming to know the future (“Momentum is weakening, volume is declining”). Description is immediately verifiable, robust to regime changes, and doesn’t self-destruct when published. Prediction creates liability and is wrong roughly half the time by definition.

How does quantum mechanics relate to financial markets?

The connection is philosophical rather than literal. In quantum mechanics, observation affects outcome (the Heisenberg uncertainty principle). Similarly, in markets, observation and participation change the system being observed. Both systems exhibit a form of observer-dependence that makes perfect prediction impossible. George Soros drew on this parallel when developing his reflexivity theory.