Harvard Management Company (2001)

This essay refers about the case study: Harvard Management Company (2001), provided by Harvard business School.

The framework

The endowment plays an important role in the financial structure of Harvard University. The annual endowment spending represents a significant part of the budget of all schools and the idea was to maintain its real value and its real income distribution. According to this goal, the HMC had been founded in 1974. Since 1990, Jack Meyer has been the CEO of HMC and has implemented a lot of measure among which the Policy Portfolio is the most important. The Policy Portfolio consists of the HMC strategic benchmark and is also the core of this discussion.

Advantages and limitations of optimal portfolio allocation

The theory of optimal portfolio allocation is the cornerstone of the modern finance. It is based on the diversification principle and emphasizes the importance of the correlation: the dependence that the fluctuation of an asset has on some others. This word is everywhere in finance and without it; the optimal portfolio theory would mean nothing.

The main advantage of this theory is that it gives to anyone who wants to invest, a methodology. To reach higher expected return without taking too much risk, one has to diversify his portfolio. The optimization procedure is initiated by giving some inputs (the expected returns, volatility and correlation). Then you do simple calculation and magically, you get the portfolio weights. We have a simple but powerful tool: we don’t need any regularity conditions; our optimizer will always be the best output we can get under our input assumptions.

Despite these pros, this optimal portfolio theory fails to answers a lot of questions based on the input assumptions. The first limitation is that we do static optimization: the optimal portfolio we get is the optimal one under the assumption that the inputs are static over time. This generates another limitation, which seems to be the main one: what are the assumptions of the theory? The theory is forward looking, which imply that all the inputs are forecasts. How do we forecast return and volatility? Are returns and volatility constant over time?

Is the correlation, which is so important for optimal portfolio calculation fixed, or does it move?

The answer for both questions is that returns, volatility and correlation are stochastic. This would have been meaningless without the diversification principle based on correlation. The optimal portfolio theory has a meaning since the diversification reduce the global risk of the portfolio. The correlation is stochastic which imply that we don’t know if tomorrow our risk is well diversified. Worse, according to empirical studies, it seems that in stress conditions,

...