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
Annual Asset Class Returns: Version Of Callan Periodic Tables
FYI: The chart below shows several issues investors struggle with all the time. It’s difficult to pick the best performing investment year after year, yet for many investors it’s an annual event. They look for an encore, picking the best asset class last year with the hope of a repeat performance. Yet, betting on last year’s winner rarely works out. Regards, Ted http://novelinvestor.com/asset-class-returns/
Thanks for posting the Callan Periodic Table. I have printed a copy of it to put it in my monthly portfolio review and analysis record book. Good Stuff ... Thanks Again ... Old_Skeet
I want to start by thanking Ted for the Callan Periodic Table update. The investment Periodic Tables have been a staple at MFO and at FundAlarm for years. The Callan version is the most famous and most widely quoted.
However, a legion of alternate Periodic Tables are published, and each offers a slightly different data set over differing timeframes. Which is most applicable for your purposes depends on your specific portfolio construction and your representative time span.
For example, here are two Links to alternate Periodic Tables featuring 20 years and 10 years of market data respectively:
The 20-year data set is from BlackRock; the 10-year data set is from American Century.
The checkerboard pattern that all Periodic Tables exhibit demonstrate the complete randomness of annual investment class returns. What works one year is not commonly repeated the next.
A few years ago, I examined the long term portfolio outcome of using a strategy based on changing the portfolio holdings to the winner of the previous year. That simple strategy fails to capture average market rewards. It is a real world illustration of the maxim “a regression to the mean”.
Although the first reference contains more historical data (20 versus 10 years), I prefer the American Century shorter term format for several reasons. It includes separate tables for both Equity and Bond components, and it summarizes the reward/risk tradeoffs by providing a convenient list of average 10-year annualized returns to compare against the component 10-year standard deviations.
If you want a Periodic Table for Commodities, it too exists (from U.S. Global Investors) . Here is a Link:
When assembling a portfolio, these data (annual returns, standard deviation) provide baseline rates. You are encouraged to adjust these dependent upon the current economic and political environment, but also dependent upon your particular preferences, goals, and timeframe.
To design a portfolio with specific percentage asset allocations, you will also need some information concerning the correlation coefficients of these investment classes against one another. That critical set of inputs can be generated at the Portfolio Visualizer website. Here is the Link to that useful site:
Congratulations! You are asking the right questions before making a portfolio adjustment decision. Do the candidate funds really provide diversification benefits? If so, how much? What are the opportunity costs? A few numbers might help answer those issues and guide that decision.
Of course, much depends on your targets, both in terms of goal portfolio growth rates and timeframe. Defining the target is a preliminary necessary opening task. As Lucius Annaeus Seneca recorded: “When a man does not know what harbor he is heading for, no wind is the right wind”.
The Portfolio Visualizer website that I referenced earlier offers some nice tools that you might want to exercise to generate a guideline data set.
Just enter your current portfolio mutual fund symbols and the PIMCO All Asset, All Authority Institutional fund symbol (PAUIX) into the correlation coefficient calculator to generally assess any diversification pluses. You might want to test the stability of the correlations by repeating the computation for several timeframes of differing lengths.
I did a few calculations, and as usual, there are tradeoffs to consider. Indeed, Rob Arnott’s PAUIX responds to a different lead drummer. He is a smart active fund manager who is supported by an excellent research staff. These guys are superior number crunchers.
The PAUIX correlation coefficients offer significant diversification. But there is a price for it. Arnott’s approach has delivered reduced returns at higher volatility levels over the last 3 and 5 year periods relative to Balanced mutual funds (DODBX, VWELX, VWINX) during this same timeframe. Only you can judge if the diversification benefit is worth the implied risk.
Another deeper level of analysis is needed to help provide guidance when addressing these tradeoffs. The Monte Carlo option on the Portfolio Visualizer site might be deployed to advantage on this problem.
In this instance, the Visualizer Monte Carlo code can only provide generic trend-lines since it does not do specific mutual funds; it only does its analysis using investment fund categories. But even with that compromise, the tool can be useful in the decision process,
Do a simplified baseline simulation without emerging market or commodities components. Select an appropriate time-span. Next, for several simulations, input different percentages of more complex Emerging Market, Precious Metal, etc, elements.
These calculations will allow you to develop a feeling for the impact of further diversification in terms of a probable end wealth and a minimum portfolio value that measure risk. By adjusting the input portfolio percentages you will experimentally test how much is needed to significantly influence the end game outcomes. These informative sensitivity studies will improve your decision making. Numbers help.
Of the 21 asset classes shown, junk corporates came in fourth but with only about half and less the volatility of the three that beat it. It was also like that volatility-wise in the 80s and 90s with junk. That's the allure of junk and it also makes it more amenable to various timing strategies where you don't have the whipsaw effect of the more volatile classes. I've seen various strategies with junk and very simple that show over 20% annualized long term returns. Then again, those are backtested results and I am more a "show me the real money results" kind of guy.
@Ted and @MJG, thank you for your posts. This is great information that I've seen before but never spent more than a short while analyzing (shame on me).
Anyway, having done some work to review the data, I find it most interesting that if you'd completely avoided developed international markets (MSCI EAFE) for the last 20 years or the last 10 or 15 years, you would have had a very good chance of doing better with that money almost anywhere else. Considering the demographics and the economic challenges that both the Europeans and the Japanese face, I'm wondering why we should think the next 10 or 20 years will be any different? Would anyone take the chance of an all US/Emerging Markets portfolio (considering just equity)?
Another interesting, but maybe not surprising, tidbit is that small-caps in developed international markets, and I only had data since 2001, were much better performers in the last 13 years and outperformed their large cap EAFE colleagues by far more than small-caps in the US or emerging markets outperformed those markets' large cap indices. So then a follow-up question... if you wouldn't avoid developed international markets totally, would you consider focusing investments in developed international markets on small-caps?
Comments
Good investing, Derf
P.S. Thanks for the post.
I want to start by thanking Ted for the Callan Periodic Table update. The investment Periodic Tables have been a staple at MFO and at FundAlarm for years. The Callan version is the most famous and most widely quoted.
However, a legion of alternate Periodic Tables are published, and each offers a slightly different data set over differing timeframes. Which is most applicable for your purposes depends on your specific portfolio construction and your representative time span.
For example, here are two Links to alternate Periodic Tables featuring 20 years and 10 years of market data respectively:
http://www.blackrock.com/investing/literature/investor-education/asset-class-returns-one-pager-va-us.pdf
https://ipro.americancentury.com/content/dam/americancentury/ipro/pdfs/flyer/Periodic_Table.pdf
The 20-year data set is from BlackRock; the 10-year data set is from American Century.
The checkerboard pattern that all Periodic Tables exhibit demonstrate the complete randomness of annual investment class returns. What works one year is not commonly repeated the next.
A few years ago, I examined the long term portfolio outcome of using a strategy based on changing the portfolio holdings to the winner of the previous year. That simple strategy fails to capture average market rewards. It is a real world illustration of the maxim “a regression to the mean”.
Although the first reference contains more historical data (20 versus 10 years), I prefer the American Century shorter term format for several reasons. It includes separate tables for both Equity and Bond components, and it summarizes the reward/risk tradeoffs by providing a convenient list of average 10-year annualized returns to compare against the component 10-year standard deviations.
If you want a Periodic Table for Commodities, it too exists (from U.S. Global Investors) . Here is a Link:
http://www.usfunds.com/interactive/the-periodic-table-of-commodities-returns-2013/#.U_Uw-kYvwQE
When assembling a portfolio, these data (annual returns, standard deviation) provide baseline rates. You are encouraged to adjust these dependent upon the current economic and political environment, but also dependent upon your particular preferences, goals, and timeframe.
To design a portfolio with specific percentage asset allocations, you will also need some information concerning the correlation coefficients of these investment classes against one another. That critical set of inputs can be generated at the Portfolio Visualizer website. Here is the Link to that useful site:
http://portfoliovisualizer.com/asset-correlations
The inputs are simple and the correlations can be calculated between input dates or for a prescribed number of days.
I hope this is helpful. Enjoy and prosper.
Best Wishes.
Just not sure I am willing to conseed versus say an actively managed all-asset, all-authority fund.
Congratulations! You are asking the right questions before making a portfolio adjustment decision. Do the candidate funds really provide diversification benefits? If so, how much? What are the opportunity costs? A few numbers might help answer those issues and guide that decision.
Of course, much depends on your targets, both in terms of goal portfolio growth rates and timeframe. Defining the target is a preliminary necessary opening task. As Lucius Annaeus Seneca recorded: “When a man does not know what harbor he is heading for, no wind is the right wind”.
The Portfolio Visualizer website that I referenced earlier offers some nice tools that you might want to exercise to generate a guideline data set.
Just enter your current portfolio mutual fund symbols and the PIMCO All Asset, All Authority Institutional fund symbol (PAUIX) into the correlation coefficient calculator to generally assess any diversification pluses. You might want to test the stability of the correlations by repeating the computation for several timeframes of differing lengths.
I did a few calculations, and as usual, there are tradeoffs to consider. Indeed, Rob Arnott’s PAUIX responds to a different lead drummer. He is a smart active fund manager who is supported by an excellent research staff. These guys are superior number crunchers.
The PAUIX correlation coefficients offer significant diversification. But there is a price for it. Arnott’s approach has delivered reduced returns at higher volatility levels over the last 3 and 5 year periods relative to Balanced mutual funds (DODBX, VWELX, VWINX) during this same timeframe. Only you can judge if the diversification benefit is worth the implied risk.
Another deeper level of analysis is needed to help provide guidance when addressing these tradeoffs. The Monte Carlo option on the Portfolio Visualizer site might be deployed to advantage on this problem.
In this instance, the Visualizer Monte Carlo code can only provide generic trend-lines since it does not do specific mutual funds; it only does its analysis using investment fund categories. But even with that compromise, the tool can be useful in the decision process,
Do a simplified baseline simulation without emerging market or commodities components. Select an appropriate time-span. Next, for several simulations, input different percentages of more complex Emerging Market, Precious Metal, etc, elements.
These calculations will allow you to develop a feeling for the impact of further diversification in terms of a probable end wealth and a minimum portfolio value that measure risk. By adjusting the input portfolio percentages you will experimentally test how much is needed to significantly influence the end game outcomes. These informative sensitivity studies will improve your decision making. Numbers help.
Best Wishes for a Successful Decision.
https://ipro.americancentury.com/content/dam/americancentury/ipro/pdfs/flyer/Periodic_Table.pdf
https://ipro.americancentury.com/content/americancentury/ipro/en/landing-pages/rising-rates-video.html
Anyway, having done some work to review the data, I find it most interesting that if you'd completely avoided developed international markets (MSCI EAFE) for the last 20 years or the last 10 or 15 years, you would have had a very good chance of doing better with that money almost anywhere else. Considering the demographics and the economic challenges that both the Europeans and the Japanese face, I'm wondering why we should think the next 10 or 20 years will be any different? Would anyone take the chance of an all US/Emerging Markets portfolio (considering just equity)?
Another interesting, but maybe not surprising, tidbit is that small-caps in developed international markets, and I only had data since 2001, were much better performers in the last 13 years and outperformed their large cap EAFE colleagues by far more than small-caps in the US or emerging markets outperformed those markets' large cap indices. So then a follow-up question... if you wouldn't avoid developed international markets totally, would you consider focusing investments in developed international markets on small-caps?
http://portfoliovisualizer.com/factor-analysis