Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
Support MFO
Donate through PayPal
Here's Rick, Ferri That Is !: 3-To-1 Odds Favor Index Investor
I had never heard of Rick Ferri until last year, but I find his articles to be well reasoned and logical. One of the good guys out there as far as I'm concerned
I too am a Rick Ferri fan, and have been so for a long time.
I particularly like his use of Monte Carlo techniques and tools to support his positions and decision making. Monte Carlo is indeed a powerful decision making tool under uncertain circumstances.
Also, I do trust Ferri’s honesty when applying this tool set. He does organize and conduct his work with appropriate fair modeling, and reasonable inputs such that his findings are both trustworthy and reliable.
But, regardless of the credibility of his conclusions, ultimately they are the product of modeling limitations, assumptions, and input data. By definition, a model is an imperfect simplification of the real world, even plausible assumptions add constraints, and input data sets are time dependent. As impressive as Ferri’s work is, it does not discover universal truths.
The Monte Carlo simulations that are the subject of many recent Ferri references are comprehensive, and individual investors should carefully review, and consider integrating his findings into their portfolio construction. But, also an investor should recognize some of the shortcomings in any such study.
The major shortfall in the present Ferri study is embedded in the Monte Carlo procedure itself. The basic operational mechanism for Monte Carlo is a random selection process. A set of fair dice are independently rolled for each and every simulation completed.
The dice are dumb slaves in the process. When applied to investment mutual fund decisions, the dice rollers, the individual investors, are modeled as having zero fund choice preferences. This is equivalent to the famous cartoon of a monkey tossing darts at a stock listing newspaper target.
It is true that many investors are know-nothing mutual fund buyers. In that instance, the Ferri studies are fair representations of the buying public population. However, there is a sizable population of mutual fund participants who have a much more complex investment profile.
MFO membership provides an excellent illustration of this smaller cohort. It is highly likely that the Ferri findings are NOT representative of this group. MFO members do not randomly throw darts at a mutual fund listing, but are far more discriminating in their choices.
MFO members are more discerning market players. Various selection criteria are used to improve the odds of success. Each investor formulates, experimentally tests against a benchmark, and continuously modifies his own personal set of selection rules. One universal observation is to keep execution costs to a minimum. As John Bogle succinctly said “costs matter greatly”.
One obvious recommendation that I would advance to Rick Ferri is that he should repeat his Monte Carlo simulations using a single cost parameter to adjust his constructed portfolios. A few lines of additional computer programming can easily accomplish this task.
After a random selection of a mutual fund is made within a given run, that mutual fund is challenged against a below average expense ratio test. If the fund does not satisfy that simple test, it is discarded and another random selection is made before proceeding. In this manner, a portfolio of below average expense ratio mutual fund components is assembled.
My guesstimate is that an actively managed portfolio cobbled together in this manner will have a much higher likelihood (higher odds probability) of outdistancing an Index benchmark than the outcomes now cited in Ferri’s frequent references.
I wish Ferri would accept this challenge; it is not a labor intensive effort. As a minimum, such an effort will permit a measure of performance sensitivity to costs in a global sense.
In general, I’m a big believer in the Pareto principle. It is also known as the 80–20 rule. It states that, in many instances, roughly 80% of the output comes from 20% of the input. In a factory setting, 80 % of the output is generated by 20 % of the workers. In baseball., 80% of the runs scored are produced by 20% of the hitters. It’s something like a diminishing returns observation.
The Pareto rule is one reason why, although I do believe that it is extremely difficult to choose outperforming fund managers, I will not totally abandon my search. I plan to keep approximately 20% of my fund positions with active fund managers. Hope springs eternal, especially when reinforced by a few well chosen fund selection rules.
Sorry for my late response to this posting; I had some local issues to attend. I do like Monte Carlo work, but it is not the only weapon in an investor’s toolbox, and it certainly is not perfect.
Comments
I too am a Rick Ferri fan, and have been so for a long time.
I particularly like his use of Monte Carlo techniques and tools to support his positions and decision making. Monte Carlo is indeed a powerful decision making tool under uncertain circumstances.
Also, I do trust Ferri’s honesty when applying this tool set. He does organize and conduct his work with appropriate fair modeling, and reasonable inputs such that his findings are both trustworthy and reliable.
But, regardless of the credibility of his conclusions, ultimately they are the product of modeling limitations, assumptions, and input data. By definition, a model is an imperfect simplification of the real world, even plausible assumptions add constraints, and input data sets are time dependent. As impressive as Ferri’s work is, it does not discover universal truths.
The Monte Carlo simulations that are the subject of many recent Ferri references are comprehensive, and individual investors should carefully review, and consider integrating his findings into their portfolio construction. But, also an investor should recognize some of the shortcomings in any such study.
The major shortfall in the present Ferri study is embedded in the Monte Carlo procedure itself. The basic operational mechanism for Monte Carlo is a random selection process. A set of fair dice are independently rolled for each and every simulation completed.
The dice are dumb slaves in the process. When applied to investment mutual fund decisions, the dice rollers, the individual investors, are modeled as having zero fund choice preferences. This is equivalent to the famous cartoon of a monkey tossing darts at a stock listing newspaper target.
It is true that many investors are know-nothing mutual fund buyers. In that instance, the Ferri studies are fair representations of the buying public population. However, there is a sizable population of mutual fund participants who have a much more complex investment profile.
MFO membership provides an excellent illustration of this smaller cohort. It is highly likely that the Ferri findings are NOT representative of this group. MFO members do not randomly throw darts at a mutual fund listing, but are far more discriminating in their choices.
MFO members are more discerning market players. Various selection criteria are used to improve the odds of success. Each investor formulates, experimentally tests against a benchmark, and continuously modifies his own personal set of selection rules. One universal observation is to keep execution costs to a minimum. As John Bogle succinctly said “costs matter greatly”.
One obvious recommendation that I would advance to Rick Ferri is that he should repeat his Monte Carlo simulations using a single cost parameter to adjust his constructed portfolios. A few lines of additional computer programming can easily accomplish this task.
After a random selection of a mutual fund is made within a given run, that mutual fund is challenged against a below average expense ratio test. If the fund does not satisfy that simple test, it is discarded and another random selection is made before proceeding. In this manner, a portfolio of below average expense ratio mutual fund components is assembled.
My guesstimate is that an actively managed portfolio cobbled together in this manner will have a much higher likelihood (higher odds probability) of outdistancing an Index benchmark than the outcomes now cited in Ferri’s frequent references.
I wish Ferri would accept this challenge; it is not a labor intensive effort. As a minimum, such an effort will permit a measure of performance sensitivity to costs in a global sense.
In general, I’m a big believer in the Pareto principle. It is also known as the 80–20 rule. It states that, in many instances, roughly 80% of the output comes from 20% of the input. In a factory setting, 80 % of the output is generated by 20 % of the workers. In baseball., 80% of the runs scored are produced by 20% of the hitters. It’s something like a diminishing returns observation.
The Pareto rule is one reason why, although I do believe that it is extremely difficult to choose outperforming fund managers, I will not totally abandon my search. I plan to keep approximately 20% of my fund positions with active fund managers. Hope springs eternal, especially when reinforced by a few well chosen fund selection rules.
Sorry for my late response to this posting; I had some local issues to attend. I do like Monte Carlo work, but it is not the only weapon in an investor’s toolbox, and it certainly is not perfect.
Best Regards.