How changing regimes and asset allocation approaches demand better CMAs

In our previous Insights piece, we evaluated the historical performance of freely available capital market assumptions against subsequently realised returns. The finding was stark: across six major providers, the correlation between expected and realised equity returns was negative. Investors following these CMAs would have systematically underweighted the markets that went on to outperform.

That piece addressed a backward-looking question: do freely available CMAs work? In this piece, we look forward. The current investment environment is placing new demands on capital market assumptions that go beyond historical accuracy. We set out what those demands are, and why the gap between what is freely available and what is needed has widened.

We have previously outlined a set of checks that investors should apply to their CMA source, covering predictive power, risk modelling, tailoring, and internal consistency. The themes we discuss here build on those checks, focusing on how CMAs need to adapt to changing macro and geopolitical regimes, and shifts in the asset allocation industry.

Changing regimes and the case for CMA granularity

Today’s markets are having to digest a confluence of large shifts in economic, political, and geopolitical regimes alongside AI-driven disruption. For example, shifting geopolitical dynamics and the renewed prominence of geoeconomics are driving higher volatility across asset classes and market segments. Similarly, AI is cutting across sectors, facilitating dispersion across sectoral outcomes.

A by-product of these shifts is that country-level policy is becoming a more important driver of asset returns -- from FX through to equities -- and the sector composition of an equity index is increasingly consequential for expected performance. In this environment, capital market assumptions at the level of broad regional aggregates such as ‘Developed Market Large-caps’ or ‘US Aggregate Fixed Income’ are no longer sufficient. Being able to measure expected returns and risk at the country, sector, and segment level is becoming more and more essential.

The Korean equity market is one concrete illustration of this. The outsized expected return we flagged in one of our Insights pieces last year, was only visible at the country level, and would have been nearly invisible in any aggregate regional estimate. This is not an isolated case. As geopolitical events increase in both frequency and market impact, the most consequential return differentials are increasingly opening up within regions and asset classes, not just across them.

There is also a growing need for analytics that help investors distinguish between temporary dislocations and permanent regime shifts. A tariff announcement, a sanctions package, or a shift in trade policy can move markets sharply — but the question for asset allocators is whether the repricing reflects a short-lived shock or a lasting change in the macro trajectory for a country or sector. Static CMA publications are not designed to support this kind of assessment. What is needed are expected returns modelled across multiple horizons and updated with sufficient frequency. This allows investors to act on the distinction between dislocation and regime shift, rather than guess at it.

CMAs connected to scenarios

In a world where established investment playbooks are being rewritten, scenario analysis is becoming a more important part of the asset allocation toolkit. Freely available CMAs are by construction point estimates, published infrequently and disconnected from any scenario framework. What investors increasingly need are expected returns that can be regenerated under alternative macro and geopolitical scenarios, so that a shift in the macro or geopolitical outlook translates directly into a revised set of expected returns and, from there, into portfolio positioning.

Relatedly, what is increasingly valuable is the ability to test alternative portfolio constructions under a range of scenarios and iterate on them in real time. This requires a user-friendly front-end connected directly to the underlying macro-finance engine — something that only a modern, purpose-built technology stack can support.

Most scenarios tend to resolve differently across countries, sectors, and asset classes. A trade policy escalation between the US and China has different implications for European industrials than for US technology, and different implications again for emerging market FX. Each of these requires expected returns that can be regenerated under alternative paths, not a single point estimate produced once a year. We have described our approach to this in a previous Insights piece on geopolitical scenarios.

The practical requirement is a framework that connects scenarios to expected returns in a disciplined way — one where changes in macro drivers propagate coherently across asset classes and market segments, and where the scenario output is a full set of capital market assumptions and realised portfolio impact rather than a collection of ad hoc judgments. This connects directly to one of the checks we have previously outlined: whether CMAs are jointly determined through common macro drivers. A scenario framework only works if the underlying expected returns already share that structure.

Industry trends, Total Portfolio Approach, and the demand for CMA timeliness

For asset allocators across the institutional investment landscape, it has been hard to avoid debates around the 'Total Portfolio Approach' (TPA) to allocation. It is not a coincidence that renewed interest in TPA has coincided with a period of heightened geopolitical and macro uncertainty. TPA implementation has gained further traction as traditional strategic asset allocation, as commonly implemented, with its fixed asset-class silos, infrequent rebalancing cycles, and reliance on long-horizon assumptions that can be slow to reflect regime shifts, has found it increasingly difficult to keep pace with an environment where the most consequential return differentials are opening up within asset classes, not just across them.

The core insight behind TPA is that investment opportunities should be evaluated based on their contribution to the portfolio as a whole, rather than within fixed asset-class silos. The approach involves looking through broad asset-class labels to the underlying risk-return factors that drive performance and using these to construct portfolios that are more deliberately diversified and more responsive to changing conditions. Critically, TPA is as much a governance framework as an analytical one. It works by collapsing the distance between investment insight and portfolio action, giving the investment team authority to act across the whole portfolio without waiting for asset-class-specific mandates to be renegotiated.

But TPA creates a new class of demand for expected returns. The analytics need to be timely enough to act between governance cycles, not only at the next annual committee review. Traditional SAA processes, which tend to reset infrequently, have been slow to respond to changes in the macro environment. TPA, by contrast, requires expected returns and risk estimates that are available across multiple horizons and updated frequently. Its effectiveness is strengthened further when those estimates are disaggregated at the country and sector level, particularly in the current environment, where geopolitical and structural shifts are creating dispersion at precisely those levels.

A common concern is that TPA and SAA are inherently in tension, i.e. that the dynamism of TPA undermines the discipline of a long-term strategic allocation. We do not think this is right. The features that make TPA effective such as granular expected returns, multi-horizon risk estimates, and a factor-based view of portfolio exposures, also strengthen the foundations of SAA. A regime shift in trade policy, for example, might trigger a dynamic portfolio tilt under TPA while simultaneously updating the long-horizon assumptions that inform the next strategic review. The analytical infrastructure is the same; the decisions it supports operate at different speeds. When that infrastructure supports both strategic and dynamic decision-making, the two approaches reinforce each other rather than compete.

The gap and what fills it

Adding up these requirements — granularity, timeliness, scenario integration — it becomes clear that currently available CMAs fall well short of what institutional investors need in the current environment. They are too aggregated to capture the country and sector dynamics that increasingly drive returns. They are too infrequent to support the decision-making cadence that TPA and modern governance structures demand. And they are disconnected from the scenario frameworks that investors need to navigate a more volatile and uncertain world. The institutions that produce freely available CMAs have other priorities. Our mission is to build the new generation of CMAs, scenario analytics, and asset allocation tools that meet the demands of institutional investors in today's environment.