Just over 30 years ago, a seminal paper was published about the various contributions to portfolio returns. The authors examined US pension plans to try and ascertain the contributions of asset allocation, market timing and security selection. Their conclusion was that on average close to 94% of a portfolio’s variance (risk) can be explained by asset allocation while the other two factors made only a minor contribution. In addition, they discovered that the results of market timing and security selection were on average both negative, -0.66% and -0.36% annualized respectively. Even the best market timers only added 0.25% while the best stock pickers added 3.6% and the worst detracted -2.9%.
These conclusions have been verified by other studies and more recent evidence has reinforced them further.
In terms of stock picking or active management, the most comprehensive, regular analysis on investment funds globally is carried out by Dow Jones SPIVA. They report that over the last 3 and 5 years, 82% of US large cap active managers have underperformed the S&P 500. These figures worsen as the time period lengthens. Over 10 years, the figure is 85% and this rises to 93% over 15 years.
Other regions also show that the majority of managers also underperform their benchmarks: Europe shows 73% underperforming over 5 years and 59% over 3 years while in Japan the figures are 60% and 52%. Most fixed income categories also show similar results. Even in 2017 when conditions have been far better for stock pickers as single stock and sector correlations have fallen and dispersion has increased, 57% still underperformed over a one year period to mid year.
Market timing is just as difficult to achieve successfully. Numerous academic studies have been written on this issue including one recent one that analysed forecasts of 28 market timing gurus over the period 2000-2012 and discovered that not one of them achieved the 74% accuracy needed to make timing profitable.
Some managers will claim that while most people struggle to pick the best funds, their expertise allows them to add value in this area. These claims should be examined very carefully especially as most fund analysis is heavily based on past performance. Dow Jones SPIVA does a quarterly analysis of persistence in fund performance. They found that of all the US domestic equity funds that were in the top quartile in 2015, only 1.94% were still there just two years later. Over five years there were no funds that remained in the top quartile.
In recent times, many large institutions have acted on this research and moved towards emphasizing asset allocation, using mainly passive investment vehicles and avoiding market timing. However others in the portfolio management business, especially in the retail space, have yet to internalise these conclusions and still attempt to find the best active funds and/or time the market.
Besides the damage to performance that is likely to occur, there is another important issue at stake. Apart from the largest institutions, almost everyone else is constrained by limited resources available for research and portfolio management. One would therefore expect that these resources would be devoted to the areas that can make the most difference. But in many cases, large amounts of time are devoted to the areas where the chances of adding value are the smallest while more important areas are neglected. It was this observation that led to the research that was done for the paper mentioned at the start of this article. Unfortunately in many cases not much has changed since then.
So what should a manager with limited resources do? Firstly, study the example of Steve Edmundson, the head of the Nevada Public Employees’ Retirement System. Mr Edmundson runs a $35 Billion pension fund all by himself and has time left over. His performance has been superior to many other large US pension funds, all of which have far greater resources. How does he do it? All investments are in passive vehicles and he doesn’t try and time the market. He makes changes to the portfolio no more than once a year. Other US pension funds are trying to emulate this model.
His results should not come as a surprise to those familiar with Modern Portfolio Theory (MPT) and the Capital Asset Pricing Model (CAPM). The former stated that one can create the best risk adjusted portfolio by combining uncorrelated assets while the latter defined this portfolio as being the market ie all the investable securities available globally. While CAPM dates to the 1960s, it has not been realistic to implement such a portfolio until recently as global markets have expanded, opened up and low cost vehicles have been created to invest in them. While it is still impossible to invest in every security in every market, one can construct a benchmark that covers a large percentage of the market while maintaining liquidity and low cost.
It is true that there are many shortcomings to the CAPM and I hope to detail the many ways in which one can improve on this portfolio, such as tactical asset allocation, in a follow up article. Nevertheless some form of this should be the benchmark that every manager uses as a basic reference point, since it is cheap, easy to implement and can be expected to deliver reasonable returns. If one is not keeping up with this benchmark then the investment process probably needs to be changed.