A Taxonomy for Multi-Asset-Class Portfolios
In a Multi-Asset-Class portfolio, the taxonomy you use is the palate for the picture you paint for your investors and board with your performance reports. The reports showcase your work – they should provide clarity and eliminate any distractions. Your audience should be able to use them both for an overview of your success in running the portfolio and to drill down into details, extrapolating facts about your process. A first-class portfolio taxonomy will help you do all this.
Just as in science, where taxonomy encompasses description, identification, nomenclature, and classification, the taxonomy you use for your portfolio needs to be consistent and clear, and provide a working vocabulary. If it lacks in any of these categories, it won’t be as useful as you need it to be. Yes, part of the portfolio taxonomy is standard in the industry, but much is left to a portfolio manager’s discretion. You create the taxonomy to suit the way you classify assets and income. A clear taxonomy gives you the tools and the vocabulary to communicate efficiently with your stakeholders.
Some attributes of a powerful taxonomy are as follows:
Morphs over time, yet retains its history
You will reclassify positions within the hierarchy on a regular basis, and you may need to readjust the ranks and hierarchy from time to time. When you do, the taxonomy should retain a long-term memory of what it looked like at any point in time. For example, you may re-classify under-performing positions at times so their performance and attribution is visible but doesn’t skew a category. You should be able to change the classification without needing to re-book the position, and without losing the history of how you defined it in any prior period.
It has multiple levels that are clearly defined and applied consistently
For a given dimension, you need multiple levels to distinguish almost-alike things. For cash items, you need to distinguish between hard cash, pending cash (other assets and liabilities), and assets in your portfolio with a cash strategy (e.g. T-bills). You should set up different types of 'cash' to reflect the returns on the assets, remembering that you control the attribution results by classifying the positions appropriately. Liquid ‘cash’ should be distinguished from illiquid 'cash.' Assets tied up in holdbacks, whose actual payout date and amount is unknown or at best unreliable, shouldn't be lumped into the same grouping with cash that earns interest, or with a portfolio position in a cash-like instrument.
Multi-dimensional, each dimension used for a different purpose
You need to tag assets and income within different dimensions for different purposes. And you need to be able to mix and match the hierarchies when making presentations for different audiences. Positions lumped together for performance reports may require different operations. For example, you need to distinguish voluntary investments from involuntary ones (such as side pockets) for the purposes of expectation of returns, timing of re-valuation, and liquidity analysis.
Rule-based, yet supports exceptions
You need the flexibility to make rules and have exceptions to your rules. For example, positions and returns on foreign exchange have many levels of complication - how and where you report FX gains and losses on non-base-currency positions depends a lot on the currency, your hedging strategy, the agreement with the counterparty, and the nature of the underlying asset. Even the question of when you report FX gains has to be evaluated for each position – unrealized FX on at-risk receivables may need exception handling. These types of exceptions may not warrant new rules – you just need a way to explicitly tag certain positions.
Has a workflow-aware time dimension
Alternative investment portfolio positions can have multiple prices on the same date, so your taxonomy must provide a vocabulary for distinguishing between them. You need to be able to capture the relationship between the timing of position valuation, and the timing of when you report your returns to your investors and board. If the disparity is significant, you need to distill your own net returns from gross returns reported by the counterparty. And you need to be able to account for post-period-close changes to a position’s valuation, and distinguish between accounting and performance returns.
Asymmetrical, with deeper classification on specific branches
Different branches of a multi-asset-class portfolio have very different characteristics that require different processing or different reporting. For an attribute such as cost basis, you just need to support different behavior for different branches of the hierarchy. But in other cases, you need to have custom reporting for a specific branch of the hierarchy – for private equity positions you must report on attributes such as unfunded commitments, and the quality and timing of valuation.
These are just some of the portfolio and performance reporting scenarios that require a rich and flexible taxonomy. A host of additional complexity, also mitigated by a rich taxonomy, surrounds operational processes such as cash projection, liquidity, portfolio modeling, and transaction processing. The multi-asset-class portfolio manager makes many decisions to illuminate their work – and much light is embedded within the taxonomy they use.