Strategic Asset Allocation: Analyzing Risk to Drive Return

Long-term investors typically use strategic asset allocation (SAA) to help them determine how to split investments among different asset classes. But traditional SAA often fails to consider both investor-specific needs and all feasible risks. 

SAA is supposed to produce a relatively stable portfolio allocation that should prevent panicky selloffs during turbulent times. However, conventional SSA does not consider the full range of scenarios, and unless all feasible uncertainties are incorporated into such portfolio analysis, you won’t know your true return - regardless of the strength of your capital market assumptions.

Alla Gil

What’s more, investors’ understanding of their own risk tolerance can change. People who think they prefer the most aggressive strategy might become very unhappy if they see they lost 70% of the value of their investments.

For all of these reasons, it’s very important to make SAA fit specific investor needs. This need for customization, moreover, applies equally to wealth managers (who design portfolios for individual investors in various risk categories) and to institutional investors.

Traditional SAA, as we’ve mentioned, lacks these customization capabilities.

Positive, But Insufficient, Steps

The inadequacy of static SAA has been reinforced by the tremendous uncertainty caused by the pandemic, supply-chain disruptions and follow-on inflation. Realizing this problem, some investment management firms have tried to develop alternative SAA approaches.

BlackRock, for example, is using an approach that generates multiple forward-looking scenarios of potential returns, with a focus on adverse pathways. This allows them to tailor portfolio allocations to specific points in time and to investors’ risk appetite.

To perform portfolio optimization that achieves return goals and protects against negative shocks, a firm could also incorporate correlations between capital markets views and the economic environment.

But while the aforementioned approaches are definitely big improvements over the traditional SAA methodology, suggested uncertainty paths and time-dependent capital markets assumptions are not sufficient for robust portfolio optimization.

The problem is that the described uncertainty paths are based on historical assumptions, and therefore rely on artificially-created data that merely replicates data patterns and may omit outliers. Consequently, potential futures generated this way may miss a key objective: protecting against unprecedented shocks.

Addressing Uncertainty

Robust SAA requires that uncertainty paths must be generated with the dynamic relationships between asset classes. Starting from capital market views informed by economists and market experts, one has to look at pathways that have been generated with overlaying historical and unprecedented shocks.

No one can predict whether a market-changing event will actually happen. But the probabilities and severities of potential market shocks can be assessed based on historical occurrences (for repeatable events) and expert views (for unprecedented ones). Moreover, anticipated events can be implied from the market data.

The generation of uncertainty paths must therefore include these Poisson-style shock events, along with their probability and severity. These types of shocks simultaneously impact all markets and asset classes, and can help produce not only a full range of scenarios (FRS) but also a comprehensive picture of portfolio outcomes.

Through this approach, investors can construct a shock-resilient SAS that can estimate returns as an expectation (or mean) of all feasible outcomes.

Advanced SSA: An Example

Let’s now consider an example of how an advanced SSA approach could potentially be used for high-yield asset classes. These assets have wider credit spreads than investment-grade bonds, to compensate for a higher probability of default. But how much extra spread is enough?

It is important to remember that credit risk is asymmetric – it has two-sided volatility risk that affects its present value, and one-sided default risk when losses are mostly irrecoverable.

Given that interest rates are still low, there is an incentive to add a portion of high-yield assets to your SAA. The FRS analysis shows that unless high-yield spreads are above 3.5%, the expected return on high-yield assets will be lower than the risk-free treasury yield. Of course, if high-yield assets do not default, they’ll yield a better return than treasury assets.

But the full distribution of returns for high-yield assets is tail-heavy and includes potential irreversible losses. So, expected returns for such assets are dragged down by a high probability of possible defaults. That’s why a spread of less than 3.5% on high-yield assets would not cover for potential losses at expected or median levels. (Indeed, starting from spread levels around 2.5%, high-yield assets perform better than treasuries only about 30% of the time.)

That’s why it is critical to generate the full range of pathways around your capital market assumptions. To capture the optimal risk-return trade-off in the SAA, and to match investors’ risk appetite with a much higher level of granularity, these pathways must consider all tail risks.

Once this methodology is used to solve for credit spreads in both high-yield and investment-grade asset classes that serve respective risk-return targets, it can also be applied to select asset subclass weights. Again, one needs the FRS to understand how risk could influence the return assumptions.

When constructing their credit portfolios, to enhance yield and to boost returns, investors often use credit “barbells” – i.e., a combination of the safest AAA-rated segment and a small percentage of high-yield investments.

But our FRS analysis demonstrates that a “bell-shaped” credit allocation (i.e., a combination of A-rated and BBB-rated segments) is more efficient than the barbell one. When the weights in both strategies are selected to produce equivalent returns, the barbell SAA produces fatter tails that can also impact expected returns.

Parting Thoughts

Advanced SAA calls for more granular quantification of risk, specific to each investor’s profile. Rather than introducing hard constraints (e.g., 60% equity/40% fixed income), the asset class weights should be selected based on real constraints – e.g., “net profit should not drop by more than X in a single quarter,” or, “at target date T, portfolio value must be at least Y with 95% probability.”

This type of resilience and flexibility can be incorporated into SAA design only under FRS, which considers the complete range of long-term uncertainty.

Strategic foresight cannot precisely predict the future. But it can help us determine how to think about it.

Alla Gil is co-founder and CEO of Straterix, which provides unique scenario tools for strategic planning and risk management. Prior to forming Straterix, Gil was the global head of Strategic Advisory at Goldman Sachs, Citigroup, and Nomura, where she advised financial institutions and corporations on stress testing, economic capital, ALM, long-term risk projections and optimal capital allocation.