The Map Is Useful Until the World Changes
An eight-part journey through five centuries of uncertainty
In December 1997, Myron Scholes and Robert Merton received the Nobel Memorial Prize in Economic Sciences for work that had transformed the pricing of financial options. Less than a year later, Long-Term Capital Management, the hedge fund in which they were partners, was close to collapse.
The proximity of those events is striking, but it can also be misleading. The option-pricing work recognised by the Nobel committee had not suddenly become worthless, nor was LTCM simply a failed experiment in the same mathematics. The fund was caught in a different problem. It held large, leveraged positions whose relationships changed as markets came under stress. As losses mounted and lenders became uneasy, the time available for those relationships to recover began to disappear.
That distinction became the starting point for The Origins of Risk. A method can be valuable within the problem it was designed to solve and still leave its user exposed somewhere else. The failure is not always in the calculation. It may be in the financing, the liquidity, the size of the position or the behaviour of everyone using similar assumptions at the same time.
Across eight articles, I followed the development of risk management from the merchants of medieval Venice to the researchers now modelling financial markets as complex adaptive systems. This is not a story in which practical judgement was gradually replaced by mathematics, or one in which each generation repeated exactly the same mistake. It is a history of real advances, each of which improved our ability to act under uncertainty while introducing new boundaries that could be overlooked.
Capital goes to sea
For the Venetian merchant, risk began with a voyage. Capital was committed before the merchant knew whether the ship would return, what prices it would find or which hazards it would meet along the way. A successful voyage could be highly profitable. A lost vessel could destroy the person who financed it.
Partnerships allowed merchants to share the capital at stake and the proceeds that followed. Marine insurance later made it possible to transfer part of the financial loss to an underwriter willing to accept it for a premium. Neither arrangement made the sea safer. They changed how much of its uncertainty one person had to carry.
This early response to risk was practical rather than statistical. The future could not be calculated with much precision, so the transaction had to make room for an unfavourable outcome. Exposure could be divided. A price could be negotiated. A loss that might ruin one participant could be spread among several.
By the late seventeenth century, Edward Lloyd’s coffee house in London was bringing together merchants, shipowners, brokers and people prepared to insure voyages. Information about ships, captains, routes, weather, war and piracy travelled through the room. An underwriter who learnt that conditions had changed could charge more, accept a smaller share or decline the business.
The importance of that information was easy to misunderstand. It did not tell the underwriter whether a particular ship would sink. It helped him avoid accepting yesterday’s price for a risk that had already changed.
Lloyd did not invent marine insurance. His coffee house helped concentrate the people, information and capital required to make an underwriting market. Judgement remained imperfect, but it carried a financial consequence. The name written beneath a risk represented a claim against the underwriter’s own capital.
When uncertainty became measurable
At roughly the same time, mathematicians were learning how repeated events could be described through probability. The development changed the range of questions that could be answered. Uncertainty did not have to remain a collection of isolated stories. Under suitable conditions, observations could reveal a pattern.
Mortality tables brought that insight into eighteenth-century life insurance. The death of an individual could not be scheduled, but the experience of a sufficiently large group showed regularities. Combined with compound interest, those patterns gave insurers a more disciplined basis for pricing promises that might not be settled for decades.
The Equitable Life Assurance Society made this approach part of a new commercial practice. Its achievement was not that it predicted individual lives. It matched the method to a population and a long-term obligation for which repeated experience contained useful information.
That qualification matters. A table can reveal a stable pattern in the data used to construct it without containing every change that the future may bring. The model is useful because some features of the world persist. Its limitations appear when the relevant population, incentives or surrounding conditions change in ways the historical record does not capture.
Option-pricing theory later tackled a different problem. The work of Fischer Black, Myron Scholes and Robert Merton showed how, under specified conditions, an option could be valued by considering a changing hedge in the underlying asset. The insight linked pricing to replication and transformed derivatives markets.
The formula was never the whole practice. A hedge had to be financed, traded and adjusted. Prices could jump between adjustments. Liquidity could weaken when it was needed most. The theory clarified the exposure, but the market determined whether the hedge could be executed as assumed.
The 1987 crash exposed the importance of that feedback. Portfolio insurance was designed to reduce equity exposure as markets fell. When many investors tried to follow similar rules during the same decline, their selling added to the pressure. The crash had several causes, and portfolio insurance should not be treated as a complete explanation. Its role nevertheless showed that a protective action for one portfolio can alter the market faced by every other portfolio.
A position has to live long enough
Long-Term Capital Management pursued trades based on price relationships it expected to converge. The fund employed exceptional people, sophisticated analysis and large amounts of borrowed capital. In 1998, market stress pushed some of those relationships further apart. Losses grew, collateral demands increased and counterparties became less willing to extend time.
The episode demonstrated why an expected destination is not enough. A position may eventually prove correct and still fail if the owner cannot finance the route between today’s price and the expected one.
The mortgage crisis a decade later was far broader and arose from different failures. Mortgages of varying quality were pooled into securities, divided into claims with different priorities and, in some cases, repackaged again. Models and credit ratings shaped how those claims were priced and held. Borrowing allowed institutions to own more of them than their capital alone would support.
When mortgage losses rose, uncertainty about the securities spread into the funding arrangements behind them. Lenders demanded more collateral. Holders needing cash sold assets. Falling prices damaged balance sheets and created further demands for cash. The effects travelled through connections among borrowers, securities, banks, investors and short-term lenders.
A loss estimate could describe part of this risk, but not the whole mechanism. The survival of a position also depended on who financed it, what collateral had been promised and how other participants would respond when prices began to fall.
The market reacts to the model
Complexity science begins from the observation that financial markets are populated by people and institutions that adapt. Their decisions affect prices, and those prices change the decisions that follow. The system does not stand still while it is being observed.
Researchers including J. Doyne Farmer have used agent-based models to study these interactions. Instead of assuming that every participant shares the same information and expectations, the models allow banks, investors and borrowers to follow different rules and respond to one another. A decline in price may trigger a margin call. The resulting sale moves the price again, placing pressure on someone else. What began as a local disturbance can travel through the network.
Such models do not recreate every institution or predict the date of the next crisis. Their value is more modest and more useful. They isolate mechanisms that conventional analysis can miss, particularly the feedback created by leverage, imitation, funding pressure and forced selling.
This perspective changes the questions a risk manager asks. The estimated loss still matters, but so do the conditions under which the estimate might cease to be relevant. Who else owns the position? What would make them sell? Is the apparent diversification dependent on relationships that may change under stress? Can the hedge be traded when many participants need it at once?
The response I can implement
I do not expect to identify the next catastrophe before it begins. My own response is a diversified systematic trend-following programme that can react to sustained price movement without first requiring an explanation for it.
The programme takes small initial risk across a broad range of liquid markets. Rules determine when positions are entered, reduced and exited. Many signals fail, producing small losses. When an unusual move persists, a profitable position is given room to develop rather than being closed at a predetermined target.
This is not automatic protection from crisis. A shock can occur before a trend signal appears. A rapid reversal can turn a gain into a loss. Different positions can offset one another, and markets that appeared independent can begin moving together. Results depend on the markets traded, the rules, costs, execution, financing and the trader’s ability to follow the process through an unproductive period.
The attraction lies in what the approach does not require. I do not have to know which market will produce the next large move, what event will cause it or how far it will travel. If the movement persists and the position remains manageable, the programme can adapt as evidence appears in the price.
Position size is central to that process. Ordinary losses must be small enough for the programme to remain active until an uncommon move arrives. Diversification creates more places in which such a move might develop, but it cannot make the portfolio immune to a common shock. Survival still depends on capital, liquidity, execution and discipline.
What five centuries leave us
The history of risk management is a record of problems solved. Marine insurance made hazardous ventures easier to finance. Mortality tables supported promises that stretched across generations. Option-pricing theory improved the valuation and hedging of contingent claims. Modern models of connected institutions help researchers examine how distress can spread.
None of these advances should be dismissed because it has limits. The useful response is to understand the limit well enough to know what must be handled elsewhere.
A model may estimate a loss without showing whether the position can be financed through it. A hedge may work in ordinary trading and become difficult when the market gaps. Diversification may reduce routine variation without protecting against a forced sale shared by many institutions. A sound long-term view may be of little help if the capital cannot survive the path.
This is why the articles belong together. They move from the division of a voyage to the measurement of repeated uncertainty, then from mathematical pricing to the behaviour of connected markets. The tools become more powerful, but the practical responsibility never disappears. Someone must decide how much capital to commit before the outcome is known.
The future will always extend beyond the map. The task is to use the map without mistaking it for the territory, and to build a portfolio that can continue when the two no longer match.
The Origins of Risk is an eight-part series at ATS Trading Solutions. Begin with the introduction: The Origins of Risk introduction. Continue through the Series archive.
Richard Brennan writes on systematic trading, complex adaptive markets, and the philosophical foundations of trend following at atstradingsolutions.com. His books include The Fractals of Finance, Complex Adaptive Markets, Carved by Impossibility and The Aussie Turtles Trend Following Guide.
Want to explore why structure exists at all?
Carved by Impossibility: What Remains When Everything Else Is Eliminated
The book explores the architecture of constraint, emergence, and reality itself, and what it means for how we understand markets, life, and the universe.
Available now on Amazon in paperback, hardcover, and Kindle.
Want the theoretical foundation for why trend following works?
The Fractals of Finance: Determinism, Adaptation and the Geometry of Markets
The book explores the full architecture of feedback, fat tails, and fractal structure in financial markets, and what it means for how we trade, invest, and understand risk.
Available now on Amazon in paperback, hardcover, and Kindle.
Want a practical field manual for trading trends and capturing outliers?
The Aussie Turtles Trend Following Guide: A Field Manual for Hunting Outliers adapts the timeless principles of the original Turtle traders into a systematic, rules-based approach for modern markets. Co-authored with Adam Havryliv.
Available now on Amazon in paperback, hardcover, and Kindle.