"The challenge of asset allocation now is no longer having too few ingredients to consider but rather selecting among an ever increasing array of sector-specific mutual funds and exotic ETFs"
A Seven Asset Portfolio out performed all other asset allocations, both prior to and during retirement.
This would consist of:-large-cap U.S. stock
-small-cap U.S. stock
-non-U.S. developed-market stock
-real estate
-commodities
-U.S. bonds
-cash
-in equal proportions, rebalanced annually.
and,
"The second part of this analysis compares three allocation models when used in a retirement portfolio — which is very sensitive to timing of returns, particularly large losses. This analysis assumed an initial nest egg balance of $250,000 — quite comfortable back in 1970, although fairly modest now — with an initial withdrawal rate of 5% (or $12,500 in year one) and an annual cost of living adjustment of 3%. Thus, the second-year withdrawal was 3% larger (or $12,875), and so on each year. The superior approach, however — with a median ending balance of over $2.1 million — is the model using seven different asset classes."
For retirees facing the future headwinds of rising rates this study found that:-during the inflationary periods of the 1970s, the seven-asset model had considerably better performance as a retirement portfolio — finishing with a balance of $2,086,863 for the 1970 to 1994 period, while the 60/40 model ended up at $1,090,081. The pattern recurs in the first four 25-year periods.
-an asset allocation model that has a large commitment to U.S. bonds (such as the classic 60/40 portfolio) may be at risk because if interest rates rise, bond returns will likely be far lower than over the past three decades.
-that a more broadly diversified portfolio is prudent — both in the accumulation years and in the retirement years.
Source:
which-asset-allocation-mix-outperforms?
Comments
Thanks for the reference to Dr. Craig Israelsen’s paper on the performance of portfolio asset mix options.
Based on precisely past performance data, he made a case for his equally divided 7 category portfolio.
Since I used the word “precisely”, and from my earlier submittals, you can easily guess where I’m headed with this post.
Israelsen’s work has a major shortcoming when using it for planning purposes. The results are perfectly tied to the exact schedule of returns recorded by past markets. They allow for no wiggle room. To expect identical results in the future requires that the order of returns must be precisely replicated. The chances of that happening are virtually zero.
The sequence of returns in any investment is significant to end wealth and portfolio survival. I’m sorry but once again, this uncertainty of the sequencing of future returns points to the use of Monte Carlo simulations to examine various portfolio options.
Although I favor the Flexible Retirement Planner for many Monte Carlo investment issues, I used the Portfolio Visualizer (PV) code to run a few sample cases because of convenience. I can run the PV version on my I-pad.
I examined 3 portfolios assuming a 1M dollar initial value with a 5.5% annual drawdown that was adjusted for inflation. To replicate Israelsen’s work as closely as possible, I assumed a 25-year retirement period. My analyses used the historical category returns formatted in a manner for random selections.
As a baseline, I inputted a simple 4 category portfolio with the standard 50/10/30/10 mix of US Equities, Foreign Equities, US Bond, and Money Market holdings. As a second portfolio, I duplicated the Israelsen 7 category portfolio that is equally divided. As a third case, I invented an 8 category portfolio which was more heavily weighted to US equities including Small Cap Value, TIPS, and a replacement of the money market holding with a Short Term Corporate Bond position. All three portfolios were basically a 60/40 split between equities and fixed income products.
I let the Portfolio Visualizer loose on all three portfolios.
The baseline portfolio had a median end wealth of 2.66M dollars with a survival probability of 83%. I’m not a happy warrior at that survival probability.
The Israelsen portfolio had a median end wealth of 3.73M dollars with an improve survival rate of 90%. So far, Israelsen wins.
But that winning record didn’t last beyond a single alternative option. The portfolio that I assembled had a median end wealth of 4.45M dollars with a much more attractive likelihood of survival at the 96% level. Note that I make no claims that my portfolio is optimum, but it is an improvement over the Israelsen construction.
This is yet another illustration of the powerful impact that Monte Carlo calculations can make when stress testing a portfolio designed for a long-term retirement period. The inputs are completed in minutes, the results are displayed in seconds, and a limitless set of what-if scenarios can be explored in a half-hour.
I urge all MFOers to become familiar with Monte Carlo tools. Your own analyses are superior to those reported by many financial advisors.
Best Wishes.
Thanks for chiming in.
You stated:
"As a third case, I invented an 8 category portfolio which was more heavily weighted to US equities including Small Cap Value, TIPS, and a replacement of the money market holding with a Short Term Corporate Bond position. All three portfolios were basically a 60/40 split between equities and fixed income products."
Are you re-balancing and how often? Have you personally implemented this portfolio for your own retirement and have you explored the cost (Fees, ER, trading costs, etc.)? Are you selecting active or passive investments? Are you able to simulate a rising rate environment?
Does your simulator provide a way to examine different inflation trend scenarios?
I believe that part of managing a well diversified portfolio is to be responsive to inflation trends, either rising or falling, as well as prolong periods of stagflation.
Thank you for reading my post and for the questions that it motivated.
I’ll try to address each and every one of them. If I inadvertently fail to do so, please ask again.
The portfolio that I postulated was a somewhat arbitrary attempt to capture the elusive and non-stationary Efficient Frontier that even its inventor, Harry Markowitz, doesn’t match in his investments.
For the record, it included a 30% commitment to US stocks and 10% equal allocations to Foreign Equities, US Small Cap Value, REIT, Treasury Bonds, Corporate Bonds, TIPs, and Short Term Corporate Bonds. I tried to make the portfolio rather generic in the spirit of Professor Israelsen’s recommendation.
This single attempt at a competitive portfolio does not directly reflect my portfolio which is slightly more nuanced and more numerous in holdings. My current portfolio contains both actively managed and passively managed products, although I am shifting resources in the passive direction and am attempting to simplify.
The portfolio that I tested does reflect some of my preferences and biases. I did not include any Commodities. I used Short Term Corporate Bonds as a Money Market equivalent.
The Portfolio Visualizer code option that I exercised used historical category return data to estimate future returns. A user can override that option with his own set of anticipated statistical returns. Since I deployed the historical data sets, the analysis results should be interpreted as the equivalent (minus the minor costs) of Index products.
The code makes inflation adjustments. An inflation model that incorporates historical data is a default option that I used. The user can override with his own statistical model if he so chooses. Portfolio rebalancing is done annually.
I do not believe that you can force the Portfolio Visualizer (it’s not my product) to specifically do a rising inflation rate. You can input a statistical representation of a higher mean inflation level with whatever standard deviation you deem appropriate. Now that’s hazardous duty and really getting into the weeds from a projection prospective.
Certainly costs are important, especially if future market returns are muted so that costs absorb a higher percentage of gross returns. That’s one of my motivators to increase my Index holdings.
One of the chief benefits of any respectable Monte Carlo code is that it permits rapid turnarounds for postulated what-if scenarios. By running a patch of these scenarios, a user can develop a feel for what is important and what is noise.
Since the codes use Monte Carlo methods, results are always displayed in probabilistic format. There are no guarantees that the likeliest events will happen. But for most folks, that should be the chosen route: take the most favorable odds. In the game of Blackjack, it’s not a good idea to hit when you hold a 17 hand. In roulette, its better to play a wheel with a single green zero and not with a wheel that contains both a single green zero and a double green zero.
Life is a series of challenges. Monte Carlo is an imperfect tool that can help to quantify the odds coupled to some of those challenges. Retirement decisions are one such task.
Thanks again for your interest and your questions.
Best Wishes.
thanks
Thank you for asking. You are the first person on MFO to specifically ask for a few Monte Carlo code addresses. Here are three that offer different user options.
My favorite because of its flexibility for my needs is from Flexible Retirement Planner. Here is a Link:
http://www.flexibleretirementplanner.com/wp/
The code I used for these submittals is from the Portfolio Visualizer website. It runs on my I-Pad and also offers some nifty other investor tools. Here is the Link:
https://www.portfoliovisualizer.com/monte-carlo-simulation
A third source that is the easiest input, but has the fewest user options is from MoneyChimp. This code requires that you estimate the mean and standard deviation returns from your pre- and post-retirement portfolios. Here is the Link:
http://www.moneychimp.com/articles/volatility/montecarlo.htm
My primary purpose for participating on the MFO Discussion Board is to be helpful to mutual fund investors. I’m way beyond salvation myself. So I’m extremely pleased that you requested this information. No trouble whatsoever.
Best Wishes.
I too have access to Bill Sharpe's Financial Engines website. I elected not to mention it because of its likely limited access for many MFOers.
I have a true anecdotal story involving Sharpe and me. In the early 1990s I was planning retirement and consulted with advisors asking for Monte Carlo analyses to support their opinions. They thought I was nuts.
So I was motivated to do my own programming. I ran into some stumbling blocks and sought help from the academic world. Professor Sharp rescued me with great advice on Monte Carlo issues and Gene Fama sent me tons of data. Both professors were extremely generous, and both were friendly. It never hurts to ask for help.
I don't understand the reluctance of some MFOers to even explore the potential benefits that Monte Carlo offers. Open mindedness when investing is an essential element to enhance the odds of success.
Many thanks for your contribution.
Best Wishes.