One of the most fun parts of writing this blog is that–for whatever reason–it seems to have attracted a group of readers most of whom appear to be particularly well informed and insightful.

Makes for a some great discussions. Much better than the typical “Hey, awesome post!” fare that you see on many other blogs.

Also, I fully agree that a lump sum invested for 39 years isn’t exactly a common real-life scenario. My only point was simply that a seemingly small decrease in return can lead to a surprisingly large decrease in accumulated wealth.

To be clear (as I mentioned above in my comment to Frank), I’m not “promoting” a 100% equity portfolio. I’m just arguing that the primary benefit of reducing the equity allocation (if you still have at least a few decades until retirement) is a psychological one–not that that means it isn’t a real benefit. Clearly, mental comfort has a very real value.

As to your question about rebalancing, I intend to address that in an upcoming post in the near future. The main idea being that one of the points in favor of holding multiple asset classes in a portfolio is that–if rebalanced regularly–the annual return for the portfolio tends to be greater than the weighted-average return of the various asset classes.

thanks for your response: it is a very sensible and ambitious, if perhaps not always optimal, policy to avoid quantifying (past) investment returns, and to focus on explaining the mechanisms and concepts, rather than mining historical data.

I’m quite sure you will agree, however, that before a “story”, mechanism, or concept is suggested/accepted as good practice in investment (e.g., your claim that the more stocks you own in your portfolio the better return you can expect in the long run), we must make serious attempts to “falsify” it (or, stress-test it, lol) by rigorous and robust back-testing. (Note the focus on “falsification” rather than “validation”. Good past performance does not imply that an idea “works”, but consistently bad past performance is strong evidence that it doesn’t.)

Since neither the Merriman data nor the article seem to mention regular rebalancing, my question to you is: do you know what relative performance of the 11 Merriman portfolios (the 11th being the one that you’re promoting: 100% equities) has been (over long periods), given that, say, the portfolio is rebalanced at every year-end?

And what about a more realistic scenario in which an investor is not dumping a lump sum and holding the same portfolio forever, but is instead dollar-cost averaging for a typical 40-year working-life period and rebalancing once every year?

Thanks for sharing that chart. Pretty neat compilation of data there.

A few thoughts on it:

I’m reluctant to describe something as having “only 10% smaller annualized return” for the reason that I suspect some readers might come to the conclusion that the 10% decrease in return means a 10% smaller ending value. In reality, that 10% decrease in return means an ending value of only 2/3 the ending value provided by an all-equity portfolio–if we assume a lump-sum invested at the beginning.

Also, (for better or worse) I actually tend to go out of my way to avoid quoting data to explain a point. Whenever possible, I’d rather try to explain the reason behind why something happened rather than quoting some figures and saying “here’s what happened.”

In my opinion, an understanding of the concepts behind investing is more useful than historical performance data, which can tempt us into believing we can predict things more precisely than we can.

That said, that chart is great–showing year by year returns rather than simply “here’s how portfolio X did over the entire period.”

http://www.fundadvice.com/articles/buy-hold/fine-tuning-your-asset-allocation.html

http://www.fundadvice.com/images/stories/fundadvice_images/fine_tuning_2008.pdf

You can derive from there, e.g., that in the (long enough to be indicative) period from 1970 to 2008, the 70% equity/30% bonds portfolio had only 10% smaller annualized return than the 100% equities portfolio (10.6% vs 11.8%), while reducing the (standard deviation measure of) volatility by 30% (10.4% vs 14.6%).

I didn’t mean to imply any psychological “failings” as such. For instance the “having-trouble-sleeping” scenario that I brought up is fairly normal in my opinion and can be experienced by somebody who is both “calm and rational” (or at least as calm and rational as any of us are).

Regardless, I do think that the difference between a 25-year old who is 80% in stocks vs 100% or 120% in stocks is a psychological one. (And I could see any of those 3 positions as being reasonable for various people who I’d include in the category of “calm and rational.”)