The foundation most portfolios are missing

The case for market-implied expected returns over CAPM-derived priors — and how LUMIQ makes rigorous portfolio construction methodology accessible to wealth management firms of any size.

Every investment decision rests on a view about expected returns. The question is where that view comes from. For most of the last half-century, the dominant answer has been the Capital Asset Pricing Model (CAPM). It was a genuine intellectual achievement, and for its time, a practical one. Today, the data and technology that were once unavailable have changed what is possible. That changes what the right starting point looks like.

The Black-Litterman model, widely regarded as the most sophisticated approach to portfolio construction, uses CAPM to derive its prior equilibrium vector of expected returns, the baseline from which investor views are blended into a final allocation. This is where the model’s dependence on CAPM’s assumptions becomes consequential.

A better approach now exists, one that our team at LUMIQ has built into the foundation of our wealth management platform.

The CAPM: a landmark model built for a different era

When the Capital Asset Pricing Model was developed in the 1960s, it represented a major step forward in thinking about risk and return. For the first time, investors had a systematic framework for asking: given the risk of holding this asset relative to the market, what return should I expect? The answer: risk-free rate, plus a risk premium scaled by beta. It was elegant, teachable, and practically implementable with the tools of the time.

The constraints of that era were significant. Real-time market data was inaccessible to most practitioners. Computing power was limited. In that environment, using historical return averages to estimate expected returns, and a simple beta to measure risk, was a reasonable approximation.

Those constraints no longer exist. Markets today generate continuous, granular price signals across thousands of securities, sectors, and countries. The computational capacity to process those signals in real time is available. Relying on historical averages and assumed risk premiums, which was once a necessity, is now a choice. And it is increasingly a choice that leaves better information on the table.

Where the CAPM’s assumptions become a problem

The CAPM formula requires three inputs: a risk-free rate, a beta, and an equity risk premium. Each of these was workable as an approximation in a data-scarce environment. Each becomes harder to defend as a foundation for serious portfolio construction when better alternatives are available.

The risk-free rate is not, in practice, risk-free. There is no consensus on which rate to use, particularly when allocating across global markets. Beta is a statistical estimate of volatility relative to a benchmark, but the choice of benchmark is itself a judgement call, and beta is unstable during the periods when stability matters most. The equity risk premium cannot be directly observed while being the most consequential input. The standard approach is to use a long-run historical average, typically stretching back decades. But the equity risk premium estimated from different historical windows varies considerably, and the appropriate window is a matter of convention rather than analysis. Given that the implied duration of equity is roughly 20 to 30 years, a 1% change in the assumed equity risk premium can produce a 20 to 30% divergence in calculated intrinsic value.

When CAPM is used as the starting point for the Black-Litterman model, these input sensitivities propagate into the allocation. The prior equilibrium vector, i.e. the baseline the model returns to when investor views are weak or absent, is built on assumptions that are backward-looking and dependent on a benchmark construction that no single investor actually replicates.

Any portfolio constructed on this foundation carries the impracticalities of the model, the focus on historical returns, and unstable results for intrinsic value through every decision that follows.

A different starting point: what the market is already telling you

The alternative does not require a better model for estimating the equity risk premium. It requires a different question entirely: rather than asking what expected returns should be, ask what the market is already implying they are.

Every traded asset has a price. That price reflects the aggregated judgement of every market participant, incorporating all available information, all prevailing expectations, and all current risk appetites. The price already implies a discount rate: the return investors are collectively demanding for holding that asset today.

It is possible to reverse-engineer valuation models to extract that implied discount rate directly. The result is the market-implied cost of equity (Ke): a continuously updated, forward-looking measure of what the market is pricing in for any given security, sector, or country. Unlike CAPM, it requires no assumed equity risk premium, no reliance on a chosen historical return window, and no requirement to define an all-encompassing market benchmark.

This is the methodology at the core of LUMIQ’s approach to forward-looking expected returns, and the foundation on which its model portfolio construction is built.

From methodology to client portfolios: LUMIQ’s platform in practice

LUMIQ has built this methodology into the portfolio construction layer of its wealth management platform. Advisors and investment teams can use it to construct and maintain model portfolios grounded in forward-looking, market-implied return data – bringing a rigorous, market-derived approach to portfolio construction within reach of wealth management firms of any size.

Those model portfolios do not sit in isolation. Within LUMIQ’s platform, advisors utilise model portfolios directly in goals-based client wealth management. The rigour that informs the construction of a model portfolio flows through to how it is applied to an individual client’s circumstances, timeline, and objectives.

What this changes in practice for advisors and investment teams

Working from a Ke-based framework changes three things that matter most in practice.

First, it provides a comparable, continuously updated map of implied returns across global markets, ranked on a single consistent metric. A high Ke relative to history or peers signals that the market is pricing in elevated risk for that asset. If the investment team believes that risk is overstated, they have a well-grounded, data-supported basis for an overweight. This is a different quality of conviction than selecting an asset because a CAPM-derived expected return clears an assumed hurdle rate.

Second, it enables forward-looking monitoring of model portfolios. As Ke compresses on a held position, the opportunity is closing. As Ke expands on an asset already in the portfolio, risk is being priced in that was not present when the position was established. The Ke is an unbiased indicator to the team on when the conditions that justified the allocation have changed, something a backward-looking framework cannot provide.

Third, it gives every decision a defensible, transparent rationale. Committee views and model outputs are most credible when they are grounded in what markets are pricing in. Ke provides that grounding, as a continuously updated, observable reference point that connects the investment team's judgement to current market reality. When a client asks why the portfolio is positioned the way it is, that chain of reasoning is visible and examinable.

The starting point is everything

In portfolio construction, the prior matters. The equilibrium baseline you start from shapes every allocation decision that follows. The CAPM was the right prior for an era when real-time market data was inaccessible and computational tools were limited. It gave practitioners a rigorous, teachable framework for thinking about expected returns under those constraints.

Those constraints have lifted. The market now speaks continuously, through prices, about what it expects. The technology now exists to listen and to translate what the market is saying into a continuously updated, market-implied prior for portfolio construction.

This is what LUMIQ’s platform makes available: the infrastructure to build model portfolios on the foundation of market-implied cost of equity (Ke), and apply models in unique client engagements. Not a simplified allocation tool. The methodology itself made accessible for wealth management firms of any size, with direct usability in goals-based client wealth management.

LUMIQ provides investment analytics and wealth management technology to licensed advisors and institutions across Singapore and APAC. Learn more at lumiq.com

This article is for informational and educational purposes only. Nothing in this article constitutes financial advice, investment advice, or a recommendation to buy or sell any security or financial product. LUMIQ Pte Ltd does not hold a financial adviser’s licence under the Financial Advisers Act of Singapore and is not an exempt financial adviser. Readers should seek advice from a licensed financial adviser before making any investment decisions.