Portfolio risk management for systematic strategies means evaluating risk at the level where capital is actually combined: across strategies, exposures and shared failure modes, not only inside each individual model.
Good strategies can still create a bad portfolio
Two strategies can each look robust in isolation and still create concentration when combined. They may trade different instruments but load on the same risk factor. They may have low average correlation that rises sharply during stress. A strategy may contribute little nominal capital but a disproportionate share of volatility or drawdown risk.
This is why portfolio risk is not simply the sum of strategy-level risk reports. The portfolio manager needs a common view of how strategies interact and whether a proposed allocation remains inside the mandate. That view becomes especially important in multi-strategy portfolio management.
Risk dimensions that allocation should see
Exposure
What market, factor and directional risks are actually present?
Concentration
Where is too much of the portfolio dependent on one strategy or common driver?
Correlation
How are relationships changing, especially when volatility rises?
Hard controls should stay explicit
Risk controls are a poor place for opaque automation. Prospect’s architecture separates contextual intelligence from deterministic portfolio rules. Models can help interpret strategy behaviour and propose allocations; explicit constraints define the permitted region in which those proposals can exist.
Design principle: the system should be able to explain not only why an allocation was proposed, but also which risk rules prevented other allocations from being considered.
Risk is dynamic even when weights are static
A portfolio can become riskier without trading. Volatility can rise, correlations can converge and the distribution of losses can change. The same 20% allocation can therefore represent a different risk contribution today than it did when the weight was set.
A systematic decision layer should continuously recalculate that context so allocation reviews are based on current risk rather than stale nominal weights. This is also why strategy monitoring and risk monitoring need to share the same evidence base.
Risk should be upstream of allocation, not a final check
If risk only appears after an allocation model has produced weights, the portfolio is effectively solving the wrong problem and then clipping the answer. A better architecture includes risk constraints inside the decision process: eligibility, reliability, risk and allocation are evaluated together.
Prospect is being designed around that integrated loop. The goal is not to replace an investment team’s mandate or governance; it is to make the evidence and constraints behind repeated portfolio decisions more consistent and traceable.
Related practitioner perspective
Dynamic risk-management research from Man Group likewise emphasises diversification and systematic management of portfolio risk. Prospect’s focus is the decision infrastructure connecting those portfolio-level considerations to changing strategy evidence and allocation governance.