Mark Barry, Head of Asset Allocation at Milestone Group, explores how a simple question can highlight opportunities to automate and streamline your asset allocation investment process. Asset allocation is a well-recognised investment discipline, yet arguably the least understood and most overlooked, in terms of the application of technology to support the investment process. Over the years, organisations have launched many technology initiatives to improve investment outcomes. These often devolve to the lowest common denominator functions as they become projects - such as a data management project, a performance or risk project, or an order management system implementation. These are all necessary functions of investment businesses, but the basic nature of these initiatives calls into question their ability to effectively support the full investment decision-making process undertaken in pursuit of a stated investment outcome. Ask an investment professional, “How and where do you capture your asset allocation decisions?” and you may be surprised that the primary system for recording asset allocation decisions is typically a spreadsheet. There is a certain degree of counterintuitiveness to this answer. We know that “asset allocation policy, however determined, is the overwhelmingly dominant contributor to total return¹” – so it raises the question as to why something so important is captured and managed in a personal productivity tool rather than a built-for-purpose integrated platform. Is it because organisations don’t expect that their asset allocation process can be encoded and systemised? Is it because some other process in this era of transparency and compliance is deemed more important? Or is it because technology procurement processes for large firms gravitate to technology that has well-known functionality, rather than search for a platform that is specific and complete for the holistic needs of the asset allocation investment process? Investment processes are like any other process; technology can make the process repeatable and transparent and facilitate scalability, safety and continuous improvement. The investment process is grounded in data management to support the generation of insight into sources of risk and return, while also creating operational control and efficiency in transaction management. However, between these two aspects are several decision-related requirements that are often underrepresented when technology is being applied to the investment process. The first of these requirements is decision analysis. This requirement is not purely for performance and attribution or risk analysis, although these have a bearing here. The requirement that is often most needed, is the ability to form a total portfolio view from the perspective of the way that capital is allocated in an asset allocation model. If the investment strategy demands a highly bespoke asset allocation model, with capital being allocated to a mix of direct assets, pooled funds, sub-advised managers and alternative assets, then that portfolio requires a system with capabilities that exceed those of systems designed for liquid equity markets. The important point here is that the ability to form this total portfolio view over all assets – whether they be look-through based assets, opaque risk category-based exposures or interfund cross-invested vehicles – requires investment classification structures , that are utilised in exposure reporting, performance, attribution or risk analyses. These considerations often get overlooked in data management initiatives. The second requirement that is often underrepresented is decision capture. To emphasise and clarify: this is not referring to trade capture but rather the recording of all asset allocation decisions and their intent. These should be captured relative to the total portfolio view in a manner that is both robust from an asset allocation model compliance perspective, as well as auditable to satisfy clients, internal governance, as well as servicing decision analysis requirements. Importantly, the key factor that determines capability is the way that different stages of the asset allocation process are recorded. For example, if the organisation is employing strategic asset allocation targets and multiple tactical tilts, then it is important that each decision stage is recorded as a discrete decision. Similarly, if the organization is operating target date fund products with separate investor cohort series adjustments, again each cohort should be captured independently. This is critical in order to maintain a record of “the way you actually make decisions” and avoid compromises that limit the prior decision analysis stage. The third requirement is decision implementation. This is about recognising that there are choices in the way that allocation decisions can be implemented. In some organisations, decisions may be implemented algorithmically, with or without discretionary approvals; in others, portfolio managers may have full discretion either via physical assets or synthetic overlays. Some portfolios have untradeable assets or very specific constraints when raising liquidity. Similarly, where investment products are being assembled, implementation may require automation of glidepaths or de-risking strategies. These are all elements of decision implementation that can easily be overlooked when defining scope of requirements for systems technology. Principles
At Milestone Group, three principles are considered as paramount:

• Ensure that the total portfolio view is transactional. Without a transactional total portfolio view, it is challenging at best, and often not achievable, to trace back from implementation through decision capture to decision analysis.
• Ensure that the transactional total portfolio view supports the actual lifecycle of the required transactions. Practitioners often speak of IBOR technologies that cannot support the settlement arrangements of alternatives or have been deployed with a “kill and fill” model. In both cases, these limitations call into question how trading technologies support the asset allocation process.
• Finally, ensure all implementation steps are captured and managed via an auditable and compliant workflow. While this may appear obvious, the ability to capture your actual investment process may not be complete. It is your investment process and your intellectual property, so the workflows required to manage decision implementation are likely to be highly specific to you.

This is challenging ground for investment organisations, whether they are plan sponsors, pension and annuity companies, or delegated solutions providers such as outsourced CIOs. Irrespective of your organisation’s focus, there is always value in starting with the simple question: “How and where do you capture your asset allocation decisions?”

¹ – Determinants of Portfolio Performance II: An Update. G. Brinson, B. Singer and G. Beebower. Financial Analysts Journal, May/June 1991. © 2020 funds global asia