Wednesday, February 13, 2008

Roger Gibson’s Asset Allocations

Most commentators agree that we should include some bonds in our long-term investments. My quest for a reasonable analysis to support this conclusion continues. Previously, I have discussed the ideas of Gordon Pape and Morningstar on this subject.

I’m starting to feel like I’m in some sort of boxing match. So, let’s do it right:

“In this corner ... Roger C. Gibson, esteemed author of ‘Asset Allocation: Balancing Financial Risk’ now in its fourth edition. He’s a well-respected expert whose ideas have been endorsed by Sir John M. Templeton and Don Philips, Managing Director, Morningstar.”

“And in this corner ... some guy who figured out how to use Blogger.”

Oh well. I lose on the credibility meter. My only chance is that people actually think about the arguments.

Gibson does an impressive amount of analysis and explains many important concepts clearly. When it finally comes time to figure out an optimal asset allocation, he tosses in an interesting assumption about our tolerance for risk. Here goes:

Suppose that our only investment choices are US Treasury Bills or stock in large US companies. Gibson assumes that for us to be willing to have a portfolio equally split between these two choices, the expected return from stocks would have to be 10% higher than the Treasury Bill return.

To understand the implications of this assumption, let’s assume that we have access to an investment that will either return 0% or 1000% on the flip of a coin. (Heads and you get your money back. Tails and you get 11 times your money back.)

Gibson’s assumption means that we would rather have a guaranteed 11% return than take a 50/50 chance on getting a 1000% return! This is a ridiculous level of aversion to risk.

In a previous post I explained that on Morningstar’s risk-aversion scale, I rate about a zero and Morningstar thinks that the average investor rates about a 2. Gibson’s assumption is that we rate a 7.1!

Despite all his analysis, in the end Gibson actually ignores most of it and focuses on investor psychology. He bases his portfolios on what the investor will tolerate rather than an analysis of what asset allocation has the best characteristics.

I understand that a major concern of investment advisors is whether their clients will stick with a plan. According to Gibson, it seems that investor psychology trumps the pursuit of investment returns.

However, this is not my concern. I seek a plan for myself that finds a rational balance between volatility and returns. I trust that I can control myself if I start having irrational feelings.


  1. So there's a 4th edition out? Wonder how it differs from the 3rd which I reviewed and found lacking in the same way as you. Obvisouly not enough in the right places.

  2. CanadianInvestor:

    I haven't compared this latest edition of Ginson's book to the 3rd edition, but some of the material is obviously new.

    One thing about the book that I find frustrating is that for many of his analyses, I can't tell whether the return figures he gives are instantaneous returns, arithmetic average returns, or compounded returns. I realize that this stuff would mean little to most of his readers, but he could put it into a footnote for the readers who are trying to judge the correctness of his conclusions.