Hi Guys,
First and most importantly, I like Brett Arends’ MarketWatch columns. He is well informed and mostly writes intelligent articles. But the referenced piece is not among his finest. It is arrogant, shortsighted, and pretentious. In competition with the Permanent Portfolio, I choose to call his the Pretentious Portfolio.
Allow me to explain why.
The perfect portfolio is a moving, dynamic target; it’s elusive because, in truth, it is an idealized concept. The theoretical Efficient Frontier line is not really a single, invariant line. It is actually a cloudy, murky zone that is constantly shifting. The zone migrates to address developing technologies, strong competition, evolving demographics, and emotions.
One major factor in that dynamic behavior is that category correlation coefficients are not nearly constant. They frequently change, and sometimes quite violently and unexpectedly. The perfect portfolio today will most certainly not be the perfect portfolio tomorrow. My guesstimate is that if Arends were to revisit his forever portfolio a year from now, some obvious amendments would be required.
Arends chosen time frame is rather limited. Fifteen years of back-testing is only a small fraction of the marketplace’s well documented history. Depending on how his portfolio was actually assembled, it could be a product of data mining.
At the very least, the robustness of his perfect solution should be tested against other time periods. Also, I hope he has the courage to monitor its future performance. Many of these perfect solutions dissolve almost as quickly as the print in the reporting article dries.
Arends portfolio is much more complex than his benchmarks which also seem to be very inappropriate as measurement yardsticks against the many components of the Arends discovery. This is yet another illustration of inappropriately selected straw-man benchmarks which were carefully chosen just to show superior rewards.
The portfolio is not the S&P
500 and is not a 60/40 Balanced fund. To be of any value whatsoever, a comparative benchmark measurement must be made against a real equivalent. As used in the Arends article, the comparisons simply demonstrate the merits of broad category diversification when cobbling together a portfolio.
Overall. I do endorse the fundamental logic and thinking that supports Arends’ construction. I simply take issue with his exaggerated claims and his short test time horizon. It is a Bridge to Far in claims and far to short in terms of data challenges. Usually, he’s better than that.
As an aside, if you want to explore the transient character of correlation coefficients, the “buyupside” website offers an attractive, easy tool for that purpose. Just input different dates for two investment products, and see the resultant change in their correlation coefficient valuation. Here is the generic Link to the site and to the specific correlation coefficient tool Link, respectively:
http://www.buyupside.com/http://www.buyupside.com/calculators/stockcorrelationinput.phpHuge changes in the correlation values are not uncommon. Remember, this is not an assignment. I don’t have that power or control. You do so as a learning lesson for your own benefit.
My answer to the opening question is a firm NO. A perfect portfolio doesn’t exist, and if it did, it would be very transient.
Best Wishes.