The Problem With Modern Portfolio Theory (Video)

Let us say you believe all the science around investing – that Modern Portfolio Theory (MPT) and Asset Allocation can improve risk adjusted returns over the long-run. Let us also say you have studied and mastered the skills which incorporate advanced statistical analysis required to design a portfolio with the optimum asset class mix to achieve this result.

What can go wrong?  Well, quite a lot and we would group the errors into three categories:

Insufficient Diversification

Many advisors do not use all the available asset classes, sticking to a 60/40 portfolio as we discussed in our blog,  Is Diversification Dead.  Some advisors fail to understand the value alternative asset classes such as real estate, convertible bonds, commodities, and other market diversifiers add to a portfolio and they do not know how to select the right ones, so they do not use them.  This is the first execution error. Or they choose categories of investments which are not really asset classes and instead are strategies whose performance is highly dependent on the manager of the investment or the category does not really improve the efficiency of the portfolio. (Think high yield bonds, for example).

Inadequate MPT Process

Many advisors do not use portfolio optimization software, so their analysis is flawed.  For the small percent who do use software, many of these advisors use weak inputs, relying on static returns, static standard deviations and static correlations which do not account for changes over time and varying market conditions.   They also may have a poor understanding of the original theory and the need to create “views” on future returns. In some cases when they do create views, they rely on too short of a time for the view. That problem is most profound for asset classes that are highly volatile. Additionally, using the arithmetic mean to capture return, standard deviation to measure risk and the Pearson Correlation Coefficient to measure the link between asset classes are incorrect statistical applications – a discussion for a future blog.   The second execution error is created by not using more advanced statistical tools.

Inadequate Asset Class Representation…FUND STYLE

Assuming the advisor uses sufficient diversification, an adequate MPT process using appropriate statistical tools, and a solid process to determine which asset classes improve portfolio efficiency, the next step is to select investments (stocks, bonds, mutual funds, ETFs) to represent the “style” of each asset class.


Style is one of the key characteristics used to delineate an asset class.  The style of the investment refers to the investment approach used to pick securities, manage risk, and direct the course of the fund.  Examples of investment style are large cap growth, large cap value, medium cap growth, etc.  The style can sometimes be determined by the name of the fund and it is also outlined in a fund’s prospectus.  But since the prospectus is a snapshot in time, not reflecting changes over time, further analysis is required.

For example, if the asset allocation mix calls for large-cap value exposure, the advisor might select a fund called “Large-cap Value Fund” to r|epresent the asset class.  Often standard names that describe an investment style, like large-cap value will accurately describe a fund’s investment strategy but not always, creating the needto monitor the underlying approach.  Funds with vague names can also be deceptive and misleading.  If a selected fund does not adequately represent the asset class, the MPT analysis is overridden.  This is the third execution error.


A recent Bloomberg Wealth article  highlights the danger of not monitoring the underlying style of a fund.  The article discusses the most recent widespread tilt towards tech stocks providing examples such as the Vanguard Institutional Index Fund and Fidelity Contrafund which hold 29% and 31% respectively of the top five tech stocks- Apple, Microsoft Corp., Amazon, Facebook Inc., and Google parent Alphabet Inc.  This high concentration in five stocks amplifies the risk of being heavily invested in one volatile sector (i.e., technology), and in one style (i.e., growth) which is not apparent by looking at the name of the fund and even the prospectus.  In this article Dr. Laura was quoted as she explains, “the fund managers know better, but they are being measured on short-term performance every day, and if they aren’t living up to client expectations, they will be out of a job”.  Unfortunately chasing short-term returns to avoid scrutiny is still another example of the conflict of interest, rampant in the financial services industry.

When executing a portfolio designed using MPT principals, advisors need to monitor the style and the fund’s underlying holdings.  Using Vanguard institutional Index Fund to represent the large-cap value asset class is an error.  Perhaps it was the right fund in the past but if the advisor is not monitoring ongoing changes, they will be unaware that it is wrong now and the portfolio design, created using MPT tools has been corrupted.


In summary there is no problem with Modern Portfolio Theory.  It works to improve risk-adjusted returns.  The problems exist due to a lack of understanding about the body of knowledge behind the science and a lack of skill in executing this knowledge.  Harry Markowitz, the father of Modern Portfolio Theory is right.  Diversification using MPT is the only free lunch in investing and it would be unwise to ignore its benefits.