Knightian Uncertainty Explained
The Difference Between Risk and True Uncertainty
One of the most important distinctions in finance is also one of the most misunderstood.
Investors frequently speak about risk. Risk models attempt to measure it. Portfolio managers seek to manage it. Regulators attempt to monitor it. Financial theory often treats risk as something that can be quantified, estimated, and incorporated into decision-making frameworks.
Yet many of the most important events in financial history were not simply risky, they were uncertain in a much deeper sense.
The distinction was first articulated by the economist Frank Knight in his landmark 1921 work Risk, Uncertainty and Profit. Knight argued that economists and decision-makers often conflated two fundamentally different concepts.
The first was measurable risk, the second was immeasurable uncertainty. Today, this distinction is known as Knightian Uncertainty. The concept remains profoundly important because it challenges one of the central assumptions underlying much of modern finance: that uncertainty can always be converted into probability, Knight argued that this is not always possible.
Some future outcomes cannot be assigned reliable probabilities because the relevant information simply does not exist. At a deeper level, Knightian uncertainty reminds investors that the future is not merely difficult to predict, parts of it are unknowable.
Risk Versus Uncertainty
Knight's contribution begins with a simple distinction. Risk involves situations where probabilities can be estimated; uncertainty involves situations where probabilities themselves are unknown.
Consider a fair coin:
the probability of heads is 50%
the probability of tails is 50%
The outcome remains uncertain, but the probabilities are known; this is risk.
Now consider a completely new technology that may reshape global economic activity over the next twenty years:
What is the probability that it succeeds?
What is the probability that regulators intervene?
What is the probability that an unforeseen competitor emerges?
No reliable probability distribution exists; this is uncertainty.
Knight argued that these situations are fundamentally different; risk can be measured, uncertainty cannot.
Why the Distinction Matters
At first glance, the difference may appear largely academic. In practice, it is enormously important; most financial models are designed to manage risk.
They estimate:
volatility
correlations
probabilities
expected returns
default likelihoods
These models assume that uncertainty can be expressed mathematically.
Knightian uncertainty challenges this assumption. As certain future events may lie entirely outside historical experience; and without historical reference points, assigning probabilities becomes highly speculative.
The model may appear precise while failing to capture the most important uncertainties, this is one reason why major crises often surprise market participants despite sophisticated risk management systems.
Financial Markets and Unknown Futures
Markets operate continuously under uncertainty.
Investors make decisions regarding:
future earnings
future inflation
future growth
future interest rates
future technological change
Some of these uncertainties can be estimated reasonably well, others cannot. For example, historical inflation data may provide useful insight into future inflation behaviour. However, assigning a probability to an entirely novel geopolitical event or technological disruption is far more difficult. Knightian uncertainty exists precisely because the future contains possibilities that cannot be inferred reliably from the past.
The Limits of Probability
Modern finance relies heavily upon probability theory.
Probability is extraordinarily useful, it provides frameworks for:
portfolio construction
option pricing
risk modelling
expected value analysis
However, probability theory assumes that relevant outcomes can be described. Knightian uncertainty introduces a more challenging problem:
What happens when future possibilities themselves remain unknown?
One cannot assign probabilities to outcomes that have not yet been imagined, this creates a fundamental limit to quantitative modelling. The problem is not computational, the problem is epistemological; and there are limits to what can be known.
Black Swans and Knightian Uncertainty
Knightian uncertainty is closely related to the concept of Black Swan events. A Black Swan represents a highly consequential event that is difficult to predict using conventional models.
Examples often cited include:
the Global Financial Crisis
the COVID-19 pandemic
major geopolitical shocks
unexpected technological revolutions
Such events often emerge from regions of uncertainty that were not incorporated into existing probability frameworks. The issue is not merely that probabilities were estimated incorrectly, the issue is that the event itself was not properly represented within the model. Knightian uncertainty therefore helps explain why Black Swans remain so disruptive.
Why Forecasting Often Fails
Forecasting works best when environments remain relatively stable. Historical relationships provide useful information, patterns persist, probabilities remain meaningful. Knightian uncertainty becomes increasingly important when environments undergo structural change.
Examples include:
technological transformation
regulatory shifts
political upheaval
systemic crises
Under these conditions, historical data becomes less informative. The future begins diverging from the past and forecast accuracy deteriorates. This explains why many forecasting failures occur during periods of significant change rather than periods of stability.
Complexity and Adaptive Systems
Complex systems generate Knightian uncertainty naturally, financial markets are composed of millions of interacting participants.
These participants:
learn
adapt
innovate
compete
As behaviour changes, the system evolves and new possibilities emerge continuously. Because the future system differs from the current system, not all outcomes can be anticipated in advance. Complexity therefore creates uncertainty that extends beyond conventional risk. The future is not merely random, it is evolving.
The Problem with Historical Data
Financial research often relies heavily upon historical analysis.
Historical data provides valuable information regarding:
volatility
drawdowns
correlations
economic cycles
However, Knightian uncertainty highlights an important limitation. History contains only events that have already occurred; the future may contain events that have never occurred previously.
This means historical datasets are inherently incomplete; no matter how extensive a dataset becomes, it cannot include future innovations, future crises, or future structural transformations. The future always contains possibilities beyond the historical record.
Knightian Uncertainty and Portfolio Construction
The existence of Knightian uncertainty has profound implications for portfolio management. If the future cannot be described fully through probabilities, then portfolios cannot rely solely upon optimisation.
Traditional optimisation assumes:
expected returns are known
risk can be estimated
correlations remain meaningful
Knightian uncertainty weakens these assumptions.
This encourages a greater focus on:
robustness
diversification
liquidity
optionality
resilience
The objective shifts from maximising returns under expected conditions to surviving unexpected conditions.
The Importance of Margin of Safety
One of the most practical responses to Knightian uncertainty is the concept of margin of safety. Because forecasts may be wrong and unknown risks may exist, investors benefit from maintaining buffers against uncertainty.
These buffers may include:
conservative valuations
excess liquidity
limited leverage
diversified exposures
Margin of safety exists because the future contains surprises. Its purpose is not eliminating uncertainty, its purpose is absorbing uncertainty.
Decision-Making Under Knightian Uncertainty
When probabilities cannot be estimated reliably, decision-making becomes more challenging.
Investors must rely increasingly upon:
judgment
adaptability
resilience
scenario thinking
probabilistic humility
The objective is not predicting every outcome; instead it is avoiding excessive vulnerability to outcomes that cannot be predicted. This represents a fundamentally different approach to risk management; as rather than seeking certainty, investors seek robustness.
The MorMag Perspective
At MorMag, Knightian uncertainty is viewed as a central feature of financial markets rather than a temporary obstacle to forecasting.
Markets are interpreted as complex adaptive systems characterised by:
incomplete information
evolving behaviour
structural change
unknown unknowns
Within this framework, investment research extends beyond measurable risk.
Research seeks to understand:
fragility
regime transitions
systemic vulnerability
uncertainty itself
Portfolio construction emphasises resilience, diversification, adaptability, and probabilistic thinking. The objective is not eliminating uncertainty, it is operating intelligently within it.
Beyond Finance
Knightian uncertainty extends far beyond investing.
Many of humanity's most important decisions involve situations where probabilities are impossible to estimate accurately. Entrepreneurship, scientific discovery, technological innovation, and geopolitical decision-making all contain substantial elements of uncertainty rather than measurable risk. The future continually creates possibilities that have never existed before. This reality affects every complex human system.
Conclusion
Knightian uncertainty represents one of the most important concepts in finance because it distinguishes between measurable risk and true uncertainty. Risk involves known probabilities, whereas knightian uncertainty involves situations where probabilities themselves remain unknowable.
This distinction challenges many traditional assumptions regarding forecasting, modelling, and risk management. It reminds investors that not every future event can be quantified, anticipated, or incorporated into historical analysis.
At MorMag, this perspective forms part of a broader philosophy grounded in complexity science, adaptive systems thinking, probabilistic reasoning, and resilience-focused investing.
The future will always contain uncertainty, some of that uncertainty can be measured, some of it cannot. The most successful investors are often not those who predict the future most accurately. They are those who build portfolios capable of surviving futures that nobody predicted at all.
