Analytical and Algorithmic Methods for the Efficient Frontier
Summary
The document asks why Harry Markowitz described algorithms for finding efficient portfolios without deriving a closed-form equation for the efficient frontier in the unconstrained case. It contrasts this approach with a later analytical derivation by Robert Merton, which expresses efficient portfolio return in terms of variance relative to the minimum-variance portfolio. The question suggests that a formula might seem easier to use than an algorithm, while recognizing that Markowitz addressed a broader portfolio-selection problem.
The accepted answer offers a historical interpretation rather than a definitive explanation. It suggests Markowitz focused on portfolio problems with additional constraints, including inequality constraints on weights, and developed the Critical Line Method to solve them. In that setting, an algorithm could handle cases beyond the simple analytical formulation. The answer further speculates that Merton’s preference for algebraic derivations helped him obtain a closed-form result. Since the explanation is conjectural and based partly on reported preferences, it does not establish Markowitz’s actual motivations or provide a comparison of the methods’ computational performance.
Key ideas
- Markowitz developed an algorithmic approach for finding efficient portfolios, including problems with inequality constraints on portfolio weights.
- Merton later derived an analytical expression for the efficient frontier in a simpler portfolio setting.
- The answer proposes that Markowitz’s focus on constrained optimization may explain his emphasis on algorithms.
- The suggested differences in research style are historical interpretation, not a demonstrated account of Markowitz’s intentions.
Tags
Full text
# Why did Markowitz not derive an equation for the efficient frontier?
# Why did Markowitz not derive an equation for the efficient frontier?
Currently, I´m studying portfolio management and portfolio selection. The founder of the MPT is Harry Markowitz, of course. But reading his famous article from 1952 and his book from 1959 (actually, I have the 2nd edition from 1991 at hand, but that shouldn´t make a difference), I realized that Markowitz never really derived an equation to calculate the efficient frontier.
He does describe what efficient portfolios are and introduces some algorithms to attain efficient sets. But if I´m not mistaken, the first who derived an equation to calculate the efficient frontier was Robert Merton in his paper "An Analytic Derivation of the Efficient Portfolio Frontier" from 1972. Starting with the expected return $\bar{E}$ of the minimum variance portfolio, Merton derives the following equation that yields the expected return of an efficient portfolio as a function of its variance:
$$E=\bar{E}+\frac{1}{C} * \sqrt{DC(\sigma^2 - \bar{\sigma}^2)}$$
My question is: Why didn´t Markowitz derive such an equation? I guess he could have done it, he is a genius in his field of research and the founder of this theory. Moreover, it seems much easier to calculate the efficient frontier using an equation rather than algorithmic approaches, so I suppose there must have been an early interest in finding such an equation like Merton did.
It would be very nice if someone can clarify this. Maybe I´m missing out on a pivotal element.
Thanks a lot in advance.
## Answer by Alex C (score 5, accepted)
https://quant.stackexchange.com/a/31431
It is surprising. What I think is: Markowitz became interested in the general problem when there are constraints (including inequality constraints) on the portfolio weights (in addition to the standard $\sum w_i = 1$ constraint). Once he devised a computer algorithm [the Critical Line Method] for solving this problem (he was a math programming whiz) he seems to have stopped there. Perhaps he did not realize the importance of an analytical formula in the simple case with only a fully invested constraint. Merton had a better math background I believe and saw this. Merton has also said [personal communication] that he does not like geometric arguments (like Markowitz used throughout his book) and preferred an analytic or algebraic method of deriving results from starting assumptions. He started out trying to derive Portfolio Theory in this manner, and it took him farther than Markowitz.Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.