The aim of this blog is to compare modern strategic asset allocation optimisation methodologies with approaches historically used to optimise strategic asset allocation.

Modern approaches:

• Ensure realistic predictions of expected returns incl., capital market crises.
• Consider assets and liabilities (asset liability management) instead of an assets-only approach.
• Create bespoke and optimised asset allocations for each client rather than mapping risk profile outcomes to a limited number of pre-defined asset allocations.
• Incorporate all drivers of return into the asset allocation rather than using a static risk/return approach.
• Improve the client experience by advising clients with the right optimisation function, e.g., to optimise multiple goals instead of using a simple Sharpe ratio optimisation.

a) Ensure realistic predictions of expected returns incl., capital market crises
The “Modern Portfolio Theory” (MPT; also known as Markowitz/Sharpe ratio optimisation) is 70 years old and, despite its name, outdated. The same applies to the Black Litterman model, which is 33 years old. Both have multiple shortcomings but are still widely used. They, for example, assume that diversification can always reduce risk. In reality, this is not the case, as assets can become highly correlated in times of crisis. Outdated methodologies also assume that the past is a predictor of the future, ignoring the possibility of climate change and other structural changes. Finally, the models also underestimate crisis scenarios because they use normally distributed returns and assume that the upside and downside potential of markets is symmetric, which we know is not true as the equity market, for example, can have huge downside shocks.

In contrast to outdated methodologies, modern approaches are forward-looking and consider multiple scenarios (stochastic modelling). They assume that the upside and downside potential of markets is neither symmetric (fat tails) nor static (i.e., different returns over time). As a result, it is no longer as important to be 100% correct in predicting returns as in outdated methodologies, but it is important to model the scenarios and the downside as accurately as possible. This is particularly relevant when considering illiquid asset classes in holistic asset allocations such as real estate, infrastructure/renewable energy, and commodities, but also theme asset classes such as “water”, as these asset classes have a completely different skewness and kurtosis in their return distributions than “mainstream” asset classes.

Why it matters: Realistic scenarios of returns and risks, but also of the downside of asset classes, are relevant to optimise a holistic asset allocation in a high-quality manner. This ensures that the risk tolerance and capacity of the clients being advised on their optimal asset allocation and on their future wealth are in line with the real market risks.

b) Consider assets and liabilities (asset liability management) instead of an assets-only approach
When advising clients, it is important to consider not only their current liquid and illiquid assets, but also their current and future assets and liabilities – as these have a direct impact on their risk capacity. A client’s risk capacity will change significantly in the following cases: the client has limited assets but is expected to inherit a significant amount within the next two to three years (or is expecting a pension payout). Another example: the client has a mortgage that needs to be refinanced in five years (refinancing risk) or has other liabilities that are expected to come due. In an asset-only world, an adviser would not take this into account when allocating a pre-defined strategic asset allocation, nor would clients’ financial goals be considered. In modern approaches, the optimisation approach would find the best holistic asset allocation that achieves the specific client goals, constrained by the risk profile and fully considering, for example, the above-mentioned refinancing risk.

Why it matters: outdated methodologies of optimising asset allocation based on a subset of information (assets only) distort the true risk capacity of clients and are likely to result in sub-optimal advice. In addition, it is difficult for clients to understand the outcome of such recommendations as they cannot see how they align with their specific financial goals.

c) Create bespoke optimised asset allocations for each client, rather than mapping risk profiles to a limited number of pre-defined asset allocations
Modern solutions optimise the asset allocation for each client individually, taking full account of their constraints (no emerging markets, minimum 20% European equities, etc.). At the same time, they integrate the capital market assumptions of each financial institution in all scenarios, taking into account different returns per asset class over time as well as higher statistical moments (non-normally distributed, multi-periodic return and risk assumptions).

Why it matters: clients receive a bespoke and optimised asset allocation that optimises their specific situation while taking into account their constraints, which is rarely possible with pre-defined asset allocations that only map to a risk profile outcome.

d) Incorporate all drivers of return (e.g., climate change) into asset allocation rather than using a static risk/return approach
Economic factors such as inflation and interest rates need to be integrated when modelling scenarios, as asset classes behave differently in different situations (for example, all asset classes tend to lose value in times of crisis). Modern approaches consider these correlations and, additionally, factors that have only emerged in recent years – such as climate change and sustainability. For example, novel approaches incorporate the impact of climate change into holistic strategic asset allocation, which, according to 2018 Nobel Laureate Prof. Dr. W. Nordhaus, is essential as it will lead to lower investment returns in the future.

Modern, multi-period and scenario-based methods lead to much more robust optimisation results when parameters such as expected returns, volatilities, and correlations of assets change. With outdated methodologies, small changes in these estimates can lead to wildly different portfolio allocations.

Why it matters: A simple, single-period, and static risk-return approach does not take into account the complexity of our ever-changing world and leads to suboptimal asset allocations. Moreover, clients find it difficult to understand a complete shift in strategic asset allocation when individual parameters change little. Finally, robust allocations lead to fewer reallocations and reduced costs as far as asset management is concerned.

e) Improve the client experience by advising clients with the right optimisation function, e.g., to optimise multiple goals, instead of using a simple Sharpe ratio optimisation
Modern optimisation approaches allow holistic asset allocation to be optimised for, among other things, multiple client goals and future inflows and outflows with different durations, as they consider multiple periods (as opposed to a single-period model). In addition, different optimisation functions (max normal, max downside, max return, min risk, min CVAR) can be used to address specific use cases. We even use a function to minimise the carbon footprint when optimising a holistic allocation. All of this cannot be achieved with outdated methodologies.

Why it matters: Modern methodologies provide strategic flexibility for future use cases such as goal-based advice or carbon footprint optimisations.

Read the original article here.