It is the MPT belief that only exposure to systematic risks can provide efficient and optimal risk adjusted returns, whereas the reality of the last decade and a ½ is that exposure to systematic risks have produced inefficient and sub optimal returns: the world has been turned on its head.
Since the late 1990s we have been living an efficient market nightmare: systematic equity risks have not been rewarded and clunky mean variance optimisers with way too high expected mean returns have produced inefficient forward looking portfolios.
While inappropriate inputs to MVOs have produced inefficient portfolios, greater damage is likely to have been caused by long term planning based on Monte Carlo simulation: many of these simulations have used expected mean returns based on historical data which would also have plotted a wildly incorrect course.
Perverse though it is, the reason why a dogma based on efficient markets and rationale investors has been instilled in financial services lore may be precisely due to the fact that most advisors, portfolio managers and ordinary investors in general lack the expertise and knowledge needed to more efficiently manage risk and return. MPT in a sense was a rational attempt to impose an efficient market framework.
MPT has unfortunately developed into a structure through which modern financial services and products are delivered without adequate thought and attention to structural and systematic risks. On the positive side it has instituted the need for rigorous, simple and disciplined frameworks that are difficult to manipulate by the behaviourally challenged.
It is more than likely that MPT adoption by the mainstream came about because of the coincidental development of computing power (and supporting software applications) as well as the upward trajectory of markets (systematic risks were seen to generate return, especially in the 1990s) that perversely confirmed the relationship between risk and return.
A key component of behavioural theory is, implicitly, that the processing ability of the human brain is insufficient to properly process all the information needed to make a rationale and efficient decision: computing power provided a framework to manage both information and computational algorithms.
MVO frameworks supposedly obviated rule of thumb processing deficiencies, which were further influenced by behavioural biases.
But MPT is a framework based on efficient markets, rationale investors and general equilibrium: it is a benchmark as to how a portfolio should be structured if prices were correct and future price movements random and independent. In an out of equilibrium world, it is behaviourally impaired in that its frame (the decision making rules that determine the decision) is based on a false reality.
To the roll call of behavioural biases we need to add, “belief in efficient markets and general equilibrium” bias.
The truth is that we live in a different world than the modern portfolio theory construct, a world in which significant structural imbalances and systematic risks have built up over the last 30 years or more, and particularly so since the mid 1990s. These have been unwinding since the end of the 1990s, and offset and subsequently reinforced numerous times since then. The steady denouement should be obvious.
While MPT structures have relevance in defining a portfolio in utopia, and in understanding risk and return at other times, MPT structures pose significant risks in their application, especially at the current point in time.