The Fossil Record of Finance

Capital at Risk — Paper 06

I. The Fossil Record

In evolutionary biology the fossils remain.

Extinct species leave traces. Their disappearance becomes part of the scientific record. The history of failure is visible.

Financial systems behave differently.

Companies disappear through bankruptcy or acquisition. Investment funds close, merge, or relaunch under new mandates. Strategies that fail are replaced by new ones.

Over time the observable record increasingly reflects the survivors rather than the full population.

This dynamic is commonly described as Survivorship Bias. The concept is widely recognised in performance studies, yet its broader implications for how financial systems appear to function are often underestimated.

Markets may appear more stable than they are partly because the evidence of instability has gradually been removed.

II. Editing the Record

The removal of failure occurs across multiple layers of the financial system.

At the company level, firms disappear through bankruptcy, restructuring, or acquisition. Index providers replace declining companies with rising ones. The composition of an index gradually evolves toward the surviving firms.

At the fund level, weak strategies rarely persist indefinitely. Funds with poor performance are closed, merged into stronger vehicles, or relaunched under revised mandates. Assets may remain invested, but the historical record becomes harder to observe.

Corporate actions within fund structures often accelerate this process. Mergers, renamings, and mandate changes allow assets to continue operating while the identity of the strategy evolves.

The result is a dataset that increasingly reflects the strategies that survived rather than the full set that originally existed.

III. The Illusion of Stability

Over time this process shapes the observable history of financial markets.

Indices tend to contain companies that survived previous downturns. Databases of fund performance often contain the strategies that avoided closure or merger. Long track records therefore become progressively selective.

One indication of this filtering process can be seen in the distribution of fund track records.

At any point in time the industry contains a large number of recently launched funds and a much smaller number with long operating histories. Thousands of strategies may have one or two years of performance data, while only a fraction remain after a decade.

Part of this pattern reflects the natural life cycle of investment strategies. Funds that fail to attract assets or deliver acceptable performance are often closed, merged, or restructured.

By the time a fund reaches ten years of history it has already passed through several rounds of selection. The distribution of long track records therefore reflects a population that has already been filtered.

The right side of the performance distribution remains visible. Much of the left side has quietly disappeared.

This filtering process can make markets appear more stable than they are. Periods of disruption are often described as sudden or unexpected, yet they frequently reveal structural pressures that had existed for some time.

When the record of past failures is incomplete, the recurrence of similar pressures can appear surprising.

The effect is not limited to perception.

When failure is removed from the observable record, the distribution from which risk is inferred becomes selectively truncated.

What is measured therefore reflects what remains, not what occurred.

IV. When Hidden Pressures Surface

Financial crises often reveal tensions that were already present within market structures.

A clear example emerged during the 2022 UK gilt crisis.

Defined benefit pension schemes had adopted liability-driven investment strategies designed to hedge interest-rate exposure using government bonds and derivatives. During stable conditions these arrangements appeared to reduce funding volatility.

When gilt yields rose sharply, however, collateral requirements associated with these hedging structures increased rapidly. Pension schemes were forced to obtain liquidity at short notice in order to maintain their positions.

Assets held for stability became the source of liquidity pressure.

The episode did not represent the sudden emergence of a new risk. Rather, it revealed structural tensions that had been manageable until market conditions changed.

V. Selection and Memory

Financial systems evolve through processes of survival and replacement.

Companies disappear through bankruptcy or acquisition. Funds with weak records close or merge. Strategies that fail are replaced by new ones.

Each cycle removes part of the historical record.

The system that remains increasingly reflects the survivors rather than the full population that once existed.

Periods of instability therefore appear unusual partly because much of the evidence of previous failure has already disappeared.