The Harvard endowment is restructuring its asset allocation process with a new approach called Flexible Indeterminate Factor-based Asset Allocation (FIFAA). This new approach will try and address the issue that the endowment has underperformed many of its rivals since the Financial Crisis. The question is whether this is going to have a large impact on the thinking of other toward asset allocation like the Yale endowment model.
The key idea that your core approach is "Flexible" and "Indeterminate" tells you a lot about how much Harvard wants to be constrained by any view or philosophy concerning asset allocation. The endowment describes it as a process.
The allocation process has four steps: 1. Determining the underlying factors that drive performance; 2. Measuring the link between those factors and asset classes; 3. Finding the desired factor exposures; 4. Building a portfolio based on those factor exposures.
The factor are said to be flexible but include world equities, US Treasuries, high yield credit, inflation and currency. There is no universal set of factors and there has been discussion of adding liquidity and volatility to this mix. This mix is a fairly straight-forward set of betas for equities, rates, credit, inflation, and currency risks. Each traditional asset class can be measured against the underlying factors. For example, an exposure to credit will also have some exposure to equity, rates, and inflation. These betas can be quantitatively measured from past data with a qualitative overlay based on the impressions of the portfolio managers.
The allocation process will then choose the desired factor exposures based on past risks and insights on the future. A simple question would be, for example, "How much inflation exposure should the portfolio have?" From this decision, the endowment will select the asset class allocations that will translate into the desired factor exposures. The process attempts to move away from choosing asset classes directly and focusing on the factor exposures. This is a hybrid between traditional asset allocation decision-making and the clear movement to factor-based analysis.
The current asset class ranges for fiscal year 2016 are listed below.
Is this a better approach to asset allocation? There is a well-defined process of mapping factors to asset classes and discussing what the appropriate exposures should be for the endowment. This is a positive development and makes sense. The question is whether this process allows for too much flexibility and thus becomes ill-defined over time. Can a regular investor implement this approach? This usability is not clear, so it is hard to say whether this approach will filter down to more endowments or to individual investors. It will likely be discussed by many investors and consultants will have to have an opinion on this approach. We should start to see in impact on other endowments in the next year.