When Uncertainty Becomes the Baseline

Equity markets are rising through a world that looks unstable by every conventional measure. The usual explanation is complacency. The more accurate one is that markets have repriced what counts as uncertainty worth reacting to.

Geopolitical conflicts remain unresolved. Inflation has not disappeared cleanly. Public debt has returned as a macro-financial concern. Bond yields remain capable of unsettling valuations. Elections increasingly function as market events. And yet equities continue to advance, sustained in part by one of the largest corporate investment cycles in modern history around artificial intelligence.

The contemporary market is not indifferent to instability. It is selective about it. It is less sensitive to instability as atmosphere and more sensitive to instability as transmission. The question is no longer, Is the world uncertain? It is, Does this uncertainty alter earnings, liquidity, rates, or market leadership?

If the answer is no, the market often absorbs it. If the answer is yes, repricing can still be abrupt.

The market is not calm. It is conditional.

Knight: markets convert uncertainty into tradable risk

Frank Knight’s Risk, Uncertainty and Profit (1921) drew a distinction that financial markets routinely blur. Risk refers to situations where probabilities can be estimated. Uncertainty refers to situations where probabilities are fundamentally unknowable.

A change in next quarter’s earnings estimate is risk. A geopolitical rupture that may alter energy flows, inflation expectations, trade routes, and alliances is uncertainty. A known interest-rate path is risk. A new technological regime that may reshape margins, labour, productivity, and capital expenditure is uncertainty.

Modern finance is built to price risk. It is less comfortable with uncertainty. Yet markets cannot trade uncertainty in the abstract. They must translate it into variables. A war becomes an oil-price scenario. A tariff threat becomes a margin assumption. Political instability becomes a regulatory or fiscal probability. Inflation anxiety becomes a central-bank reaction function.

This translation is the market’s way of domesticating disorder. It turns the unknowable into something that can be modelled, hedged, discounted, and traded. The process is necessary, but it can produce false precision. Some events do not merely change the inputs of the model. They change the model itself.

The market no longer trades the headline

One reason markets appear less reactive is that they have become more hierarchical in their treatment of fear.

At the lowest level is narrative uncertainty: political rhetoric, diplomatic tension, institutional dysfunction, general geopolitical anxiety. These may influence sentiment but do not always trigger repricing. Above that is macro uncertainty: inflation, employment, growth, fiscal deficits, central-bank policy. This matters more because it affects discount rates and liquidity. Above that is earnings uncertainty: revenues, margins, demand, pricing power, capex returns. This is central, because equity valuation ultimately depends on future cash flows. At the top is liquidity uncertainty: the possibility that buyers disappear, credit tightens, volatility becomes self-reinforcing, or funding conditions break.

This hierarchy is why markets can rise while the news environment deteriorates. The market is not asking whether disorder exists. It is asking whether disorder has found a channel into cash flows, discount rates, or liquidity.

That is a more sophisticated market. It is not necessarily a safer one.

Keynes: confidence survives where calculation cannot

Investment is never purely mechanical. In The General Theory of Employment, Interest and Money (1936), Keynes argued that enterprise depends not only on mathematical expectation but on “animal spirits”: the spontaneous confidence required to act despite an uncertain future.

Markets do not require certainty in order to rise. They require enough confidence to remain invested despite uncertainty. Today, that confidence is supported by three forces: resilient corporate earnings, the expectation that policy will eventually adjust if financial conditions deteriorate, and the belief that artificial intelligence may generate a durable productivity and profit cycle.

In Keynesian language, AI has become a major source of contemporary animal spirits. It gives investors a future large enough to discount.

But confidence is not certainty. It is a bridge over uncertainty, not its elimination. And bridges can carry too much weight.

AI is not a bubble of absence. It is a test of extrapolation.

The AI trade is often described as a bubble. The comparison is analytically imprecise.

A classical speculative bubble asks investors to believe that profits will eventually appear. The AI valuation question is different. It asks whether today’s extraordinary profits can remain extraordinary for long enough to justify the prices already paid.

The leading AI beneficiaries are not cash-flow-less speculative shells. Nvidia, Microsoft, Alphabet, Amazon, and Meta are among the most profitable companies in modern economic history. Their valuations are built not on the absence of earnings but on extraordinary earnings, dominant market positions, and the expectation that current growth can persist.

Nvidia illustrates the point. In the first quarter of fiscal 2027, the company reported record revenue of $81.6 billion, up 20% sequentially and 85% year-on-year, at a GAAP gross margin of 74.9%. These are not imaginary fundamentals. They are exceptional fundamentals.

That is why the bubble framing falls short. It is not a bubble of absence. It is a risk of extrapolation. The market reaction to recent earnings has reinforced the point: exceptional performance may no longer be enough to surprise. This does not mean the AI thesis is weak. It means the market’s expectations have become demanding.

The issue is no longer whether AI is real. It is how much of AI’s future has already been capitalised into today’s prices.

The relevant historical analogy is British Railway Mania in the 1840s. Railways genuinely transformed the economy — transport, trade, industrial logistics, geography, and the standardisation of time itself. Capital rushed in, Parliament authorised hundreds of new projects, and railway shares rose sharply through the mid-1840s before collapsing as financing strained and projects failed to meet expectations. The episode is useful precisely because it was not a story of fantasy. It was a real technological transformation whose financial expectations moved faster than its economic absorption.

The lesson is not that technological revolutions are illusions. It is that even real revolutions can become bad investments at the wrong price.

AI therefore plays a double role. It explains why markets can look through macro and geopolitical uncertainty, because it provides a credible growth engine. But because that growth engine is now central to index performance and investor psychology, it also concentrates fragility.

AI is not the market’s hallucination. It is the market’s most profitable conviction — and therefore one of its most dangerous.

Kahneman and Tversky: the reference point has moved

Investors are no longer comparing events against a stable world. They are comparing them against a world already assumed to be unstable. Prospect theory anticipates this. In their 1979 paper, Kahneman and Tversky showed that people evaluate outcomes relative to a reference point rather than in absolute terms. Gains and losses are judged against what has become normal.

The implication for markets is direct. The question is not, Is there uncertainty? The question is, Is the uncertainty worse than the uncertainty already priced?

A geopolitical headline that once would have looked like a rupture may now be treated as background. Inflation volatility, political dysfunction, trade tension, and technological disruption are no longer interruptions to normality. They are increasingly perceived as normality itself. The market’s threshold for surprise has risen.

This is not necessarily irrational. It reflects learning. But it can also produce desensitisation. A market repeatedly exposed to instability may become better at filtering noise and worse at recognising when familiar noise has become a genuine signal.

Minsky: resilience can manufacture fragility

In Stabilizing an Unstable Economy (1986), Hyman Minsky argued that financially sophisticated capitalist economies can become unstable from within. Periods of calm encourage risk-taking, reduce margins of safety, and create the belief that the system is more robust than it really is.

Today’s market may contain a Minskyan dynamic, but not only through credit or leverage. It may also operate through psychology. Repeated recoveries have taught investors that shocks are temporary, liquidity eventually returns, dominant firms adapt, and volatility can be bought. This learning is grounded in experience. It is also the beginning of complacency.

The risk is not that investors fail to see uncertainty. The risk is that they see it, recognise it, and still assume it will behave like previous shocks.

That is the uncomfortable possibility: the market may be processing uncertainty more intelligently while simultaneously becoming more fragile.

It can be rational to rise on strong earnings. It can be rational to discount geopolitical noise that lacks a transmission channel. It can be rational to reward companies with exceptional cash flows. But when investors stop treating uncertainty as a reason for caution and begin treating it as a routine condition to be bought, the market becomes less prepared for the kind of shock that cannot be easily absorbed.

Shiller: narratives rank reality

Robert Shiller’s work on narrative economics argues that economic behaviour is shaped not only by data but by contagious stories that organise how people interpret data.

The current market narrative is more structured than “stocks go up”. It runs roughly as follows: AI will create a new productivity regime; dominant firms will capture a disproportionate share of the gains; large capital expenditures should be read as infrastructure rather than excess; volatility is temporary; and uncertainty is permanent but high-quality companies can compound through it.

The narrative is powerful because it does not deny instability. It absorbs it. Disorder becomes a reason to own the strongest companies. Volatility becomes a reason to buy future cash flows at a temporary discount. Concentration becomes evidence of quality rather than a warning sign.

That is what strong market narratives do. They do not merely describe reality. They rank it. They tell investors which facts matter and which can be ignored.

The risk is not that the AI narrative is false. The risk is that it becomes too total. When one story organises too much capital, too much valuation, and too much confidence, the market becomes vulnerable not only to bad news but to insufficiently good news. A company can report exceptional results and still see its share price fall. At a certain point, excellence is no longer enough. The market requires surprise.

The real risk is correlation

The greatest danger is not that one uncertainty materialises. Markets can often absorb isolated shocks. The greater danger is correlation: the moment separate risks begin reinforcing one another.

Oil prices rise because geopolitical risk escalates. Inflation expectations move higher. Bond yields rise. Expected rate cuts disappear. Equity multiples compress. AI valuations become harder to justify. Consumer weakness spreads. Margins decline. Credit conditions tighten. Volatility rises. Investors who were crowded into the same trades begin selling at the same time.

Each risk may be manageable alone. Together, they become a regime change.

This is the central vulnerability of a market adapted to permanent uncertainty. It can handle disorder as long as disorder remains fragmented. It struggles when separate uncertainties begin to reinforce one another.

Markets do not usually break because uncertainty exists. They break when uncertainty becomes connected.

The desensitised market

The current equity market is not one that has stopped caring about risk. It is one that has changed its definition of relevant risk. Uncertainty alone no longer commands a strong reaction. It must pass through one of the market’s critical channels — earnings, liquidity, rates, or leadership — to become actionable. If it does not, it is treated as a background condition.

This is adaptation before it is complacency. But adaptation has a cost.

When uncertainty becomes normal, markets stop reacting to its presence. They react only to its consequences. That works while consequences remain isolated, measurable, and delayed. It fails when they arrive together.

The market has adapted to disorder. The risk is that adaptation becomes a theory of immunity.