Path Dependence

Why History Shapes What Comes Next

In economics and finance, the future is often treated as though it begins from a clean slate. Prices change, capital moves, firms adapt, investors respond, and new information enters the system. Yet the state of a market, institution, company, or economy at any given moment is rarely independent of the sequence of events that preceded it.

This is the essence of path dependence.

Path dependence describes situations in which outcomes depend not only on current conditions, but also on the particular sequence of events through which those conditions emerged. Two systems can arrive at apparently similar positions while possessing very different histories, and therefore very different future possibilities. This distinction matters because financial systems are not purely static optimisation problems. They are evolving systems shaped by accumulated decisions, institutions, expectations, technologies, relationships, capital structures, incentives, and beliefs. Some consequences of the past disappear quickly.; others become embedded in the architecture of the system itself.

A firm's capital structure can constrain its future strategic choices. A technology can become dominant because an early advantage generated adoption, which generated further adoption. A financial institution can become dependent upon a particular source of funding because previous decisions made alternative sources increasingly expensive. An investor can become anchored to a thesis because years of accumulated evidence, capital, reputation, and identity have become attached to it.

The past does not necessarily determine the future, but it can determine the set of futures that are realistically available. Understanding that distinction is central to understanding how complex economic and financial systems evolve.

What Is Path Dependence?

At its simplest, path dependence means that history matters to the present in a way that cannot be reduced to the present state alone.

In a memoryless system, knowing the current state is sufficient to determine the relevant dynamics. The route taken to reach that state is irrelevant; as if two systems occupy the same state, their subsequent behaviour should, all else equal, be equivalent.

A path-dependent system is different. The sequence of previous events can alter the system's structure, constraints, incentives, or expectations. Consequently, two systems that appear similar at a particular point in time may respond differently to the same future shock because they arrived there through different paths.

This can be expressed conceptually as:

Future state = f(current state, historical path, new information, shocks)

rather than simply:

Future state = f(current state, new information, shocks)

The distinction may appear subtle, but its implications are substantial. Path dependence does not mean that every historical event remains equally important forever; nor does it imply that the future is predetermined. Rather, it suggests that certain historical sequences leave persistent effects on the structure of the system.

The important question therefore becomes not merely where is the system now?, but:

How did it get here?

From Initial Conditions to Increasing Returns

One of the most important mechanisms through which path dependence develops is increasing returns.

When an early advantage produces consequences that reinforce the same advantage, relatively small differences in initial conditions can become amplified over time. Consider a simple example, a new technology gains a modest early user base. Its growing adoption encourages developers to build compatible applications, more applications make the technology more useful, greater usefulness attracts more users, a larger user base makes development even more attractive.

The resulting feedback loop can be represented as:

Initial advantage → adoption → complementary investment → greater utility → further adoption

Once such a feedback process becomes sufficiently strong, the system may move away from what would otherwise have been a competitive equilibrium. This is one reason why markets do not always converge neatly upon the theoretically superior solution. An alternative technology may be technically attractive, yet unable to displace an incumbent because the incumbent has accumulated complementary infrastructure, users, expertise, standards, distribution networks, and capital. The eventual outcome is therefore partly a consequence of the historical sequence.

The same logic can operate in financial markets; an asset class that attracts capital may develop deeper liquidity, better research coverage, more financial products, stronger institutional infrastructure, and greater investor familiarity. These improvements can attract still more capital. The original inflow becomes self-reinforcing.

What began as a relatively small difference can therefore become an enduring structural advantage.

Lock-In

The concept of path dependence is particularly important when historical processes generate lock-in.

Lock-in occurs when the cost of moving away from an established path becomes sufficiently high that previously adopted arrangements persist even when alternatives may appear attractive. Importantly, lock-in does not necessarily mean that the existing arrangement is optimal; a system can become locked into an outcome because changing it would require overcoming accumulated switching costs, sunk investment, coordination problems, institutional resistance, or behavioural attachment.

Financial systems contain numerous forms of lock-in; a company may have invested heavily in a particular production system. Its suppliers, employees, software, logistics, and customers may all be organised around that system. Replacing it could therefore be economically difficult even if another technology offers superior long-term characteristics. Similarly, an institution may develop a particular risk-management framework, reporting architecture, organisational structure, or investment process. Over time, employees become trained around it, data becomes structured around it, and performance is measured through it. Thus, over time the system becomes increasingly difficult to change; and the historical investment itself becomes a constraint on future choice.

This is why sunk costs can have effects even though, in a narrow economic sense, sunk costs should not influence rational forward-looking decisions. The expenditure may be economically irreversible, but the organisational, technological, and behavioural structures created by that expenditure remain very real.

Hysteresis: When the Past Changes the Present

Path dependence is closely related to the concept of hysteresis.

Hysteresis occurs when the effect of a previous shock persists after the original shock has disappeared, a temporary disturbance can therefore produce a permanent or semi-permanent change in the system, this matters enormously in finance. Suppose a company experiences a severe liquidity crisis. Even if market conditions subsequently normalise, the company may emerge with damaged credit relationships, higher financing costs, reduced investment capacity, weaker supplier confidence, and a more conservative balance sheet. The original shock is gone, yet its consequences are not.

Likewise, a financial institution that suffers a major loss may permanently alter its risk appetite. An investor who experiences a severe drawdown may subsequently maintain a lower allocation to risky assets; a country that experiences a banking crisis may introduce regulatory institutions that remain in place for decades. Therefore, the system has changed because of what happened to it.

This creates an important distinction between shock persistence and structural persistence. A shock can disappear while its consequences remain embedded in the system.

Path Dependence and Financial Markets

Financial markets are particularly fertile environments for path dependence because they combine feedback loops, expectations, institutional constraints, heterogeneous agents, and adaptive behaviour.

Prices influence beliefs, beliefs influence positioning, positioning influences prices. Capital flows influence liquidity, liquidity influences transaction costs, transaction costs influence capital flows. Performance influences allocations, allocations influence prices, prices influence subsequent performance. These relationships can create feedback mechanisms in which yesterday's outcomes affect today's behaviour, which then alters tomorrow's outcomes.

Consider momentum.

An asset that has performed strongly may attract additional capital from trend-following strategies, discretionary investors, institutional allocators, and investors responding to improving narratives. The additional demand can reinforce the price trend, and the resulting performance then provides further evidence for the original thesis. The process does not continue indefinitely, of course; feedback can reverse, weaken, or become destabilising. But the important point is that the current price cannot always be understood independently of the sequence of previous price movements and investor responses.

Fundamentally, the path matters.

Corporate Strategy and Capital Allocation

Path dependence is equally important at the company level.

Corporate decisions accumulate. A firm chooses a market, it builds expertise in that market, it hires people with relevant skills, it develops relationships with suppliers and customers, it invests in infrastructure, it establishes internal processes, it develops a reputation. Those decisions increase the firm's capabilities in some areas while reducing the relative attractiveness of others; the firm's future opportunity set is therefore partly endogenous to its previous choices.

This creates a fundamental asymmetry between option creation and option destruction. Whereby, some investments create future choices, whilst, others narrow them. A company that maintains a strong balance sheet, diversified capabilities, flexible technology, and valuable organisational knowledge may preserve strategic optionality. Conversely, a company that accumulates excessive leverage, commits to inflexible infrastructure, or becomes dependent upon a narrow customer base may progressively reduce its future choices.

Capital allocation should therefore not be evaluated solely through immediate returns, it should also be evaluated through the future state it creates. As an investment that generates a modest return today but materially expands future strategic options may possess a different economic value from an investment that produces a higher immediate return while creating significant future constraints.

Path Dependence and Competitive Advantage

Competitive advantages are often themselves path-dependent.

A company's position can be strengthened through cumulative processes that competitors cannot easily replicate. Brand recognition compounds, organisational knowledge compounds, customer relationships compound, distribution networks compound, data can compound, research capabilities compound, reputation compounds.

These assets frequently derive their value from accumulated history; this creates an important distinction between static advantage and dynamic advantage. A static advantage exists because a firm possesses something valuable today, whereas, a dynamic advantage exists because the firm's current position allows it to become stronger tomorrow. The latter can be particularly powerful; as a company with a strong research organisation may discover products more effectively, attracting more customers and generating greater cash flow. Those cash flows can finance further research, which improves future products and reinforces the company's position.

The competitive advantage is therefore not simply an asset on the balance sheet, it is primarily a process.

The Investor Is Part of the Path

Investors themselves are also path-dependent.

Investment decisions do not occur in isolation. Investors accumulate experiences, beliefs, relationships, reputational commitments, portfolio exposures, tax considerations, and emotional associations with previous outcomes. A portfolio that has evolved over ten years is not equivalent to a portfolio assembled from scratch today, even if its current holdings are identical. Intrinsically, the route matters.

An investor who bought an asset at £20 and watched it rise to £100 may behave differently from an investor who purchased the same asset at £100. Their current positions are identical in one sense, but their psychological reference points, realised gains, perceived risks, and willingness to sell may differ substantially.

This is one reason behavioural finance and path dependence intersect. Anchoring, endowment effects, loss aversion, commitment escalation, and narrative attachment can all transform historical experience into present-day decision constraints. As the investor does not simply observe the market. they also carry a history into the market.

Path Dependence and Market Narratives

Narratives can also become path-dependent.

A particular interpretation of an asset, company, sector, or economy can gain traction through repetition. Analysts publish research, investors cite previous research, management teams reinforce the narrative, media coverage expands, new investors encounter the existing interpretation before forming their own.

Eventually, the narrative can become part of the informational environment itself. This does not necessarily make the narrative false, instead it means that the history of the narrative influences how new information is interpreted; the same piece of information can therefore produce different reactions depending upon the expectations that preceded it. A disappointing earnings result may be interpreted as confirmation of an already-established negative thesis. The identical result may be interpreted as a temporary setback if expectations were extremely pessimistic beforehand.

Information enters a system with a history.

Multiple Equilibria

One of the deeper implications of path dependence is the possibility of multiple equilibria.

If different historical sequences can lead to different stable outcomes, then there may be no single inevitable equilibrium toward which the system converges. For example, two economies with similar resources can develop different industrial structures; two companies operating in similar markets can develop different organisational capabilities; two investors receiving similar information can construct radically different portfolios.

The difference may originate in relatively small historical divergences that subsequently compound, this makes economic systems fundamentally different from many mechanical systems. As such, the final state may depend upon the route taken.

Critical Junctures

Path dependence does not imply that systems are equally constrained at all times.

Periods of stability can be interrupted by critical junctures: moments when existing structures become unusually susceptible to change. Crises are often such moments. A financial crisis can force institutions to reconsider leverage; a technological breakthrough can undermine established infrastructure; a regulatory change can alter incentives; a management transition can redirect corporate strategy. During these periods, previously stable paths can diverge.

This produces a useful conceptual pattern:

Accumulation → reinforcement → lock-in → disruption → branching → new path

The significance of a shock therefore depends not only on its magnitude, but on the structure of the system it enters. A relatively small disturbance can have substantial consequences if the system is near a critical transition. Conversely, a much larger disturbance may produce relatively little lasting change if the system possesses substantial resilience. This is one reason why analysing shocks independently of system structure can be misleading.

Path Dependence and Financial Fragility

The relationship between path dependence and fragility is particularly important.

Fragility often accumulates gradually. Leverage increases, liquidity buffers decline, risk becomes concentrated, correlations change, institutions become dependent upon particular funding mechanisms, and investors increasingly rely upon similar models and assumptions. None of these developments necessarily produces an immediate crisis.

Instead, they alter the structure of the system; with the system becoming increasingly dependent upon a particular path continuing; as when that path is interrupted, vulnerabilities that were previously hidden can become visible simultaneously. This helps explain why financial crises can appear sudden even when the underlying fragility accumulated over many years.

The crisis is sudden, the path to the crisis was not.

Path Dependence and Risk

Traditional risk analysis often focuses on the distribution of possible future outcomes.

Path dependence adds another dimension: the sequence through which those outcomes emerge. Two scenarios can produce the same final portfolio value while having radically different risk characteristics.

A portfolio that falls 30%, recovers, and ends flat is economically different for an investor facing liquidity constraints from a portfolio that remains stable throughout. A company that reaches a given debt ratio after decades of conservative financing may have a different risk profile from one that reaches the same ratio after rapid debt accumulation. The endpoint is identical, the histories however, are not; this suggests that risk analysis should consider not only distributions of outcomes, but also trajectories, transitions, and state dependence.

For complex portfolios, the relevant question may therefore be less:

What could happen?

and more:

What sequences of events could take us there?

Why Historical Context Matters

Path dependence provides a strong argument for historical analysis.

Historical research is sometimes treated as descriptive background: useful for understanding where a system came from, but secondary to quantitative analysis of the present. For path-dependent systems, that distinction becomes much weaker.

History can contain information about the mechanisms that generated current conditions. The evolution of a company's margins can reveal the development of its competitive structure; the history of capital expenditure can reveal why a business has its current cost base; the history of regulation can explain why certain market structures exist; the history of investor positioning can help explain why apparently modest information produces unusually large price movements.

History is therefore not merely context, it can also be viewed as apart of the causal model.

Implications for Investment Research

For investors, path dependence changes the way research should be conducted.

A company should not be assessed solely on its current financial statements; its current position should be understood as the consequence of accumulated capital allocation decisions, strategic choices, technological developments, competitive interactions, and external shocks. Similarly, a market should not be evaluated solely through its current valuation, volatility, liquidity, or positioning. Researchers should ask how those conditions developed and whether the mechanisms generating them are reinforcing or weakening.

This leads to several practical questions:

What historical decisions created the company's current structure?

Which advantages are self-reinforcing?

Which constraints are becoming increasingly difficult to reverse?

Where have temporary shocks created permanent changes?

Which current conditions depend upon continuation of an established path?

What would have to change for the system to move onto a different trajectory?

Thus, the objective is not to predict the future by extrapolating the past, instead it is to understand how the past has shaped the future's available possibilities.

Beyond Prediction

Path dependence also challenges simplistic notions of forecasting.

If the future depends upon endogenous feedback, adaptive behaviour, and sequence-dependent developments, then prediction becomes inherently conditional.

A forecast may be correct under one path and irrelevant under another; this does not make forecasting useless, instead it changes what useful forecasting looks like. As rather than attempting to identify one deterministic future, investors can examine multiple plausible trajectories and the conditions under which each becomes more likely. This is closer to scenario analysis than conventional point forecasting, therefore, the emphasis shifts from predicting what will happen to understanding what could cause the system to move from one state to another. That distinction is crucial in environments characterised by uncertainty.

The MorMag Perspective

At MorMag, path dependence is best understood as a reminder that markets are evolving systems rather than collections of independent observations.

Prices, fundamentals, expectations, capital structures, institutions, and investor behaviour are continuously shaped by what came before them. The present is therefore not simply a snapshot, it is the accumulated result of previous interactions; this has direct implications for investment research.

A useful research process should examine not only the current state of an asset or company, but the trajectory that produced it. It should distinguish temporary deviations from structural changes, identify feedback mechanisms, recognise emerging lock-in, and examine where historical decisions have expanded or constrained future optionality.

The objective is not to assume that history will repeat itself, history rarely does so neatly, instead the objective is to understand how history has altered the system in which the future will unfold. This perspective also reinforces a broader principle of disciplined capital allocation: value cannot always be understood independently of process. The quality of an investment depends not only on the expected outcome, but on the sequence of decisions, constraints, feedback loops, and risks through which that outcome must be reached.

For a probabilistic investor, this means thinking in terms of trajectories rather than isolated endpoints; for a risk manager, it means recognising that fragility can accumulate before it becomes visible; for a capital allocator, it means considering whether today's decision creates future optionality or future constraint; and for a researcher, it means treating history as part of the mechanism rather than merely the introduction. Markets move through time, capital moves through time, institutions evolve through time.

The consequences of decisions therefore do not simply disappear when the decision is made, they become part of the system from which subsequent decisions emerge. That is the central insight of path dependence:

The future is shaped not only by where a system is, but by the path it took to get there.

Conclusion

Path dependence provides a powerful framework for understanding why economic and financial systems frequently behave differently from simplified models of equilibrium and optimisation.

Historical events can generate feedback loops. Feedback loops can produce increasing returns. Increasing returns can create lock-in. Lock-in can constrain future choices. Shocks can then disrupt established paths and produce new trajectories. Thus, the resulting system is neither completely deterministic nor completely unconstrained, it is conditional. Understanding that conditionality is particularly important for investment; as current valuations, financial statements, market structures, and investor expectations are not independent facts floating outside history. They are outcomes produced by processes that have unfolded over time.

The investor's task is therefore not simply to ask where the system stands today. It is to understand how it arrived there, which mechanisms are sustaining its current trajectory, where those mechanisms may break, and what alternative paths could emerge when they do. In complex financial systems, the past may not dictate the future; but it can profoundly shape the map of what comes next.