Frankenstein is often portrayed as frightening—and at times he is. What unsettles us most, though, has less to do with the monster himself and more to do with the way he is constructed: assembled from parts that were never meant to function together. Awkward. Clumsy. Misaligned. When those flaws become visible, discomfort turns into fear.
We see the same dynamic in markets. In an earlier post, we explored why markets operate in distinct regimes—and why recognizing those regimes matters even without trying to time them. What follows is the natural next step.
As an industry, we have been limited by our tools. For decades, we have built equity portfolios for a world that exists only in Excel—not because we lacked discipline, but because we lacked the resolution to see the alternative. We plug decades of returns, volatilities, and correlations into an optimizer and accept the output as “balanced.” On paper, the process conveys discipline and precision. In practice, though, we often engineer a Frankenstein portfolio: an equity construction assembled from disparate parts.
Link: Why Regimes Matter
The notion that equity markets behave consistently enough over long periods for averages to be trustworthy is a core assumption that is rarely questioned. Mean returns, correlations, and volatility are treated as stable features of the landscape. Smooth them, optimize them, and the result is a single broad-market equity portfolio designed to work “over the long term.”
But the evidence is clear: markets do not work that way.
Risk is episodic. It does not arrive as a gentle drizzle; it comes in clusters, in storms, and in long stretches of calm punctuated by sharp breaks. Periods of steady compounding alternate with periods of stress, volatility persistence, and correlation shocks. These episodes are not small deviations around a stable mean. They are distinct regimes with their own internal logic.
By blending all regimes into a single dataset and forcing one optimization, traditional equity portfolio construction effectively ignores this reality. The result is a portfolio designed for a mythical middle ground—an architecture that is structurally misaligned with the environment it actually inhabits.
We see the cost in both directions.
In durable, low-correlation markets, the portfolio drags around too much protection. Defensive tilts calibrated for “all seasons” become unnecessary ballast, quietly taxing compounding when equity risk is being rewarded.
In fragile, high-correlation markets, the same portfolio is under-defended. Its protection is tuned to average conditions—precisely when equity markets stop behaving like averages and diversification within equities begins to fail.
The subtle but crucial point is this: optimizing on averaged equity behavior is not the same as averaging equity portfolios that are each optimized for specific regimes. Portfolio construction is non-linear. Once regime-specific risk structures are blended before optimization, conditional information is lost—when diversification holds, when it fails, and when convexity within equities suddenly becomes the dominant concern.
Ultimately, this is not about discarding the foundational principles of diversification or risk premia. Advanced analytics now allow us to see long-understood truths in much higher definition. We no longer have to rely on static, averaged approximations of market behavior when we can observe the distinct, episodic nature of risk as it unfolds. By moving beyond a “Frankenstein” assembly of blended regimes, portfolios can be constructed not in hope of a mythical average, but with clear intent—purpose-built for the environments they actually inhabit.
A more resilient equity architecture, therefore, starts from a different premise: markets spend time in a small number of persistent risk states, and those states reward very different equity behaviors. Some environments compensate investors for owning growth, momentum, and higher-beta exposures. Others punish those exposures and reward quality, defensiveness, and structural resilience.
The bottom line is simple. Portfolio resilience does not come from compromise. Resilience springs from specialization. Instead of searching for a single equity portfolio that works “on average,” investors can design a structure built around clear mandates and explicit, regime-oriented roles.
The Growth Sleeve is optimized for compounding when conditions are durable and correlations within equities are stable. Its role is not to hedge every possible shock, but to express equity risk premia cleanly.
The Defensive Sleeve is optimized for capital preservation and drawdown control when conditions are fragile. Its role is not to keep up in bull markets, but to keep investors engaged—financially and psychologically—during periods when correlations rise and volatility becomes sticky.
Notably, the value of this framework does not depend on perfectly timing regime transitions. Regimes are powerful even when treated as structural states rather than trading signals. Simply recognizing that different equity behaviors dominate in different environments—and designing portfolios that respect each state—creates a more resilient architecture than any single, averaged solution. Timing may refine outcomes, but structure alone improves alignment.
Together, these sleeves form a deliberate two-state equity architecture rather than a single-state broad-market portfolio built on long-term averages. The growth sleeve does not apologize for being pro-risk. The defensive sleeve does not apologize for being return-inefficient at times. Each exists for a reason, and the portfolio’s resilience comes from that clarity of purpose.
To benefit from regime thinking, investors need not abandon discipline or embrace ad-hoc discretion. On the contrary, successful regime-based investing requires greater discipline and systematic rigor, precisely because it seeks to align portfolios with how risk actually behaves: clustered, regime-dependent, and often discontinuous.
One important clarification is worth making. Not all regime frameworks meet this standard. Regimes defined after the fact—retroactively aligned to economic narratives such as recessions, inflation cycles, or policy shifts—can be intuitive and explanatory. But a model that explains the past is not the same as one that provides informative, forward-looking structure.
When regimes are shaped to fit outcomes already observed, bias enters the model, and portfolios constructed on those labels inevitably inherit those biases. In a non-stationary world, resilience comes not from smoothing reality or imposing hindsight-driven labels, but from constructing equity portfolios that respect regimes as they are—through purpose-built, regime-aware components, each designed for distinct risk states.
In the next post, we’ll explore the discipline required to implement regime-based constructs—and begin the transition from static multi-state allocations to dynamic exposure across these portfolio components.
About CrestCast™