Correlation Is Not Diversification

Stability, Synchronisation, and the Illusion of Risk Reduction

Diversification is a foundational principle of portfolio construction.

It is commonly understood as the practice of spreading capital across multiple assets in order to reduce risk. The underlying assumption is that not all assets will move in the same direction at the same time, thereby smoothing returns and limiting drawdowns.

This principle is often operationalised through correlation. Assets with low or negative correlation are combined in the expectation that their interactions will reduce overall portfolio volatility.

However, this interpretation rests on a critical assumption:

that correlations are stable.

In practice, they are not.

The Role of Correlation in Portfolio Construction

Correlation measures the degree to which two assets move together.

In traditional frameworks:

• low correlation implies diversification benefits

• negative correlation implies hedging potential

• stable relationships allow for predictable portfolio behaviour

These assumptions underpin modern portfolio theory and many risk management approaches. However, correlation is a statistical summary of historical relationships. It does not capture how those relationships may evolve.

The Instability of Correlation

Financial markets are dynamic systems.

Relationships between assets change over time, influenced by:

• macroeconomic conditions

• liquidity dynamics

• participant behaviour

• structural shifts

As a result, correlations are not fixed, they are state-dependent.

During periods of stability:

• correlations may appear low

• diversification appears effective

During periods of stress:

• correlations often increase

• assets move together

• diversification breaks down

This behaviour reflects a shift from independence to synchronisation.

Synchronisation and Systemic Risk

In stressed environments, markets often exhibit collective behaviour.

Participants respond to:

• shared information

• changes in risk appetite

• liquidity constraints

This leads to:

• alignment of positioning

• simultaneous buying or selling

• convergence of asset movements

The result is a rise in correlation across the system. Diversification, which relies on differences in behaviour, becomes less effective precisely when it is most needed.

The Illusion of Diversification

When correlations are measured during stable periods, portfolios may appear well diversified.

However, this can create an illusion, as:

• historical relationships may not persist

• diversification benefits may be overstated

• risk may be underestimated

This illusion becomes apparent during market stress, when:

• correlations converge

• drawdowns increase

• portfolio protection fails

Diversification based solely on historical correlation is therefore incomplete.

Structural Drivers of Correlation

Correlation is influenced by underlying structural factors.

These include:

• common exposure to macroeconomic variables

• shared dependence on liquidity conditions

• overlapping investor bases

• similar strategy allocation

Assets that appear distinct may, in reality, be linked through these common drivers. As a result, their behaviour may converge under certain conditions. Understanding these structures is essential for evaluating true diversification.

Correlation and Reflexivity

Correlation is not purely descriptive. It can be influenced by behaviour and feedback.

As participants observe rising correlations, they may:

• reduce risk

• adjust positions

• rebalance portfolios

These actions can reinforce the observed relationships. This reflexive process can amplify co-movement, particularly during periods of stress.

Beyond Static Measures

Effective diversification requires moving beyond static correlation measures.

This involves:

• analysing how relationships change across regimes

• identifying common risk factors

• considering behavioural and structural dynamics

Rather than relying solely on historical correlation, it is necessary to understand the drivers of co-movement.

Diversification Across Dimensions

True diversification extends beyond asset allocation.

It involves diversification across multiple dimensions, including:

• time horizons

• strategies and approaches

• sources of return

• underlying risk exposures

This reduces dependence on any single relationship or condition.

The MorMag Perspective

At MorMag, diversification is approached as a dynamic process rather than a static allocation. Correlation is treated as an indicator, not a guarantee.

The framework emphasises:

• monitoring changes in co-movement

• recognising periods of increasing synchronisation

• adjusting exposure in response to evolving conditions

Quantitative models provide insight into relationships, but interpretation accounts for:

• regime shifts

• behavioural dynamics

• structural dependencies

This ensures that diversification is evaluated in context.

From Correlation to Understanding

Correlation provides a useful starting point. However, it does not fully capture the complexity of market relationships.

Understanding diversification requires:

• identifying underlying drivers

• recognising when relationships may change

• managing exposure accordingly

This shifts the focus from measurement to interpretation.

Conclusion

Correlation is a valuable tool, but it is not synonymous with diversification. Relationships between assets are dynamic, influenced by behaviour, structure, and changing conditions. During periods of stress, correlations often increase, reducing the effectiveness of traditional diversification strategies.

At MorMag, diversification is understood as a function of structure, behaviour, and regime, rather than a static statistical property.

In financial markets, risk is not eliminated through correlation alone. It is managed through a deeper understanding of how assets interact within an evolving system.