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.
Thanks for referencing the Lake Wobegon article from the Janus Capitol Group. Many of the sentiments expressed in that article dovetail very nicely with a few insights that I have emphasized in recent postings. It’s a comfort to get professional reinforcements from the hard charging cavalry.
Jim Goff, the author of the piece, is definitely a member of that hard charging Janus research department, and a respected fund manager. He has been around the block more than a few times and is fully committed to his research. He has been with Janus for over ten years and heavily invests in his own products. That’s always a good signal.
I was particularly attracted to Goff’s statement about the inaccuracy of economists forecasting record. Here is the commentary that grabbed my attention:
“To find this in financial markets we have to look no further than track records of economists’ forecasts. When we look at GDP estimates over the past 30 years, we see that the “professionals” are only good at telling us what just happened.”
Jim Goff is a dedicated market participant and his perspectives should always be taken seriously, even if you take exception to them.
He adds professional depth to the observation that I reported when referencing a recent CXO statistical study. Economists are notoriously inept at forecasting, but so is everyone else. For completeness, here is the internal Link to my posting on the matter:
Janus seems to be recognizing the complex nature of the investment world at a rather late date. The fact that certain rare events are better characterized mathematically by a Power Law distribution rather than by a Normal Bell curve distribution has been known for decades. This is really not new stuff, but the Janus specific application of it might be.
I was disappointed that in the short reference list provided by Goff, no mention was made of Benoit Mandelbrot. He is often credited with originally appreciating and crudely modeling the Power Law nature of the marketplace. His 2004 book “The (Mis)Behavior of Markets” is considered a pioneering classic. The subtitle of the book is “A Fractal View of Risk, Ruin, and Reward”. Indeed it is. Fat tails can be wealth killers.
Since that time, complexity theory scientists, using virtual computer agents who obey simple decision rules in Monte Carlo-based computer simulations, have uncovered surprising group emergent properties that help to understand the interactive nature of the financial markets.
A little progress has been made; a lot more awaits discovery. Understanding market complexity is slowly being advanced, but the largely unpredictable behavior of investors makes modeling a challenging and uncertain chore.
Once again, thank you for providing access to this thought-provoking article.
I had one further thought relative to Complexity Theory that I failed to include in my initial reply.
Complexity Theory is a new, an exciting, and surely an evolving branch of science. It has much to contribute, but has much to learn in its evolving nature. Its bedrock tool is Monte Carlo computer simulations that are easily programmed, cheaply executed, but often fail to model the critical parameters that control complex problem outcomes. That’s mostly because these governing parameters are difficult to recognize.
But the subject itself is stimulating, fascinating, and relatively easy to understand.
I recommend you consider exposing yourself to its subtleties through a short introductory course offered by the Teaching Company. The title of the course is “Understanding Complexity”. The 12 half hour lectures are given by University of Michigan professor Scott Page. His style is relaxed and his lecture’s main goals are easy to grasp.
Surprisingly, the subject is listed in the Teaching Company’s catalogue in the Business and Economics section.
The subject matter incorporates discussions on tipping points, the wisdom of crows, the six degree of separation phenomenon, and emergent properties. The series defines a complex system as having the following characteristics: a population of diverse agents, connected interactions with feedback loops, interdependency of actions, and adaptability freedom.
The referenced course is a quick and relatively inexpensive pathway to understanding complexity. You might want to give it a try.
I enjoyed the covered material so much that I think I’ll do a second viewing.
Reply to @MJG: Thanks for mentioning Mandelbrot and the course recommendations...
Mandelbrot's power laws certainly model behavior of the market better but his formulations in that book are hard to use very susceptible to small changes in data, as far as I remember. So, it is good theory, not go great practical application there yet. At least not those of us that do not have the tools, computing capabilities and access to relevant raw data like quants. I expect this to change in time. You mentioned Monte Carlo simulations. I expect these simulations to incorporate these power distributions in the future more often.
Speaking of courses, an increasing number of leading universities are offering free courses online. Some charge a small fee if you want to get a certificate but if you simply want to learn the subject, these are great opportunities.
I agree. Power laws are especially sensitive to data input accuracy. The nature of the mathematical formulas makes them highly susceptible to small errors. That is likely one major reason why economists favor linear relationships.
Comments
Thanks for referencing the Lake Wobegon article from the Janus Capitol Group. Many of the sentiments expressed in that article dovetail very nicely with a few insights that I have emphasized in recent postings. It’s a comfort to get professional reinforcements from the hard charging cavalry.
Jim Goff, the author of the piece, is definitely a member of that hard charging Janus research department, and a respected fund manager. He has been around the block more than a few times and is fully committed to his research. He has been with Janus for over ten years and heavily invests in his own products. That’s always a good signal.
I was particularly attracted to Goff’s statement about the inaccuracy of economists forecasting record. Here is the commentary that grabbed my attention:
“To find this in financial markets we have to look no further than track records of economists’ forecasts. When we look at GDP estimates over the past 30 years, we see that the “professionals” are only good at telling us what just happened.”
Jim Goff is a dedicated market participant and his perspectives should always be taken seriously, even if you take exception to them.
He adds professional depth to the observation that I reported when referencing a recent CXO statistical study. Economists are notoriously inept at forecasting, but so is everyone else. For completeness, here is the internal Link to my posting on the matter:
http://www.mutualfundobserver.com/discussions-3/#/discussion/6434/more-forecasting-follies
Janus seems to be recognizing the complex nature of the investment world at a rather late date. The fact that certain rare events are better characterized mathematically by a Power Law distribution rather than by a Normal Bell curve distribution has been known for decades. This is really not new stuff, but the Janus specific application of it might be.
I was disappointed that in the short reference list provided by Goff, no mention was made of Benoit Mandelbrot. He is often credited with originally appreciating and crudely modeling the Power Law nature of the marketplace. His 2004 book “The (Mis)Behavior of Markets” is considered a pioneering classic. The subtitle of the book is “A Fractal View of Risk, Ruin, and Reward”. Indeed it is. Fat tails can be wealth killers.
Since that time, complexity theory scientists, using virtual computer agents who obey simple decision rules in Monte Carlo-based computer simulations, have uncovered surprising group emergent properties that help to understand the interactive nature of the financial markets.
A little progress has been made; a lot more awaits discovery. Understanding market complexity is slowly being advanced, but the largely unpredictable behavior of investors makes modeling a challenging and uncertain chore.
Once again, thank you for providing access to this thought-provoking article.
Best Wishes.
I had one further thought relative to Complexity Theory that I failed to include in my initial reply.
Complexity Theory is a new, an exciting, and surely an evolving branch of science. It has much to contribute, but has much to learn in its evolving nature. Its bedrock tool is Monte Carlo computer simulations that are easily programmed, cheaply executed, but often fail to model the critical parameters that control complex problem outcomes. That’s mostly because these governing parameters are difficult to recognize.
But the subject itself is stimulating, fascinating, and relatively easy to understand.
I recommend you consider exposing yourself to its subtleties through a short introductory course offered by the Teaching Company. The title of the course is “Understanding Complexity”. The 12 half hour lectures are given by University of Michigan professor Scott Page. His style is relaxed and his lecture’s main goals are easy to grasp.
Surprisingly, the subject is listed in the Teaching Company’s catalogue in the Business and Economics section.
The subject matter incorporates discussions on tipping points, the wisdom of crows, the six degree of separation phenomenon, and emergent properties. The series defines a complex system as having the following characteristics: a population of diverse agents, connected interactions with feedback loops, interdependency of actions, and adaptability freedom.
The referenced course is a quick and relatively inexpensive pathway to understanding complexity. You might want to give it a try.
I enjoyed the covered material so much that I think I’ll do a second viewing.
Best Wishes.
Mandelbrot's power laws certainly model behavior of the market better but his formulations in that book are hard to use very susceptible to small changes in data, as far as I remember. So, it is good theory, not go great practical application there yet. At least not those of us that do not have the tools, computing capabilities and access to relevant raw data like quants. I expect this to change in time. You mentioned Monte Carlo simulations. I expect these simulations to incorporate these power distributions in the future more often.
Speaking of courses, an increasing number of leading universities are offering free courses online. Some charge a small fee if you want to get a certificate but if you simply want to learn the subject, these are great opportunities.
I agree. Power laws are especially sensitive to data input accuracy. The nature of the mathematical formulas makes them highly susceptible to small errors. That is likely one major reason why economists favor linear relationships.
Thanks also for the college course tip.
Best Wishes.