Importance of Asset Allocation

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
LHerr
Posts: 289
Joined: Sat May 17, 2008 7:47 am
Location: Portland, OR

Importance of Asset Allocation

Post by LHerr »

In a study by Ibbotson and colleagues, he states that it is time "for folklore to be replaced with reality" when it comes to the importance of asset allocation. While Ibbotson considers asset allocation to be very important, this latest research shows it is about as important as active management and the most important factor to portfolio movement is market movement. Putting aside the results of the Ibbotson et. al. research for a moment, I would like some interaction on the question of portfolio make up in these studies on the importance of asset allocation.

Am I wrong to question the validity of the Brinson and Ibbotson studies as they examine portfolios constructed of stocks, bonds, and cash? I've yet to come across a study where more complex portfolios are examined. Perhaps it is too difficult to conduct useful studies on portfolios that break the U.S. equities market into cap sizes as well as value and growth. In addition, many of us include developed international markets, commodities, REITs, emerging markets, and international REITs in our portfolios. Some investors will include international bonds as well as U.S. bonds.

My question is, does asset allocation take on more importance than Ibbotson gives credit if one builds a well-diversified portfolio with more than the three classic asset classes of stocks, bonds, and cash?

LHerr
rustymutt
Posts: 4001
Joined: Sat Mar 07, 2009 11:03 am

Who is Ibbotson?

Post by rustymutt »

Who is Ibbotson? I've done my own study over time and have to disagree with them. Rustymutt and colleagues study hasn't found anything other than better control of risk, as well as higher returns with correct asset balance.
Last edited by rustymutt on Tue Nov 02, 2010 7:27 am, edited 1 time in total.
Even educators need education. And some can be hard headed to the point of needing time out.
User avatar
Topic Author
LHerr
Posts: 289
Joined: Sat May 17, 2008 7:47 am
Location: Portland, OR

Re: Who is Ibbotson?

Post by LHerr »

rustymutt wrote:Who is Ibbotson? I've done my own study over time and have to disagree with them. Rustymutt and colleagues study hasn't found anything other than better control of risk, as well as higher returns with correct asset balance.
Google Roger Ibbotson and you will find many references. He is a well-known researcher in the investment world. His research firm was purchased by Morningstar a few years ago.

I agree with you that asset allocation (AA) is a way to control portfolio risk and I've yet to find a better way to construct a portfolio.

In the recent Ibbotson paper, 75% of a portfolio return is tied to market movement, 12.5% to asset allocation and 12.5% to active management. I assume active management includes both stock picking and market timing. My question remains, would the results be the same if one were analyzing portfolios with more than the basic three asset classes?

LHerr
RPS
Posts: 369
Joined: Thu Sep 09, 2010 7:58 am

Re: Who is Ibbotson?

Post by RPS »

LHerr wrote:In the recent Ibbotson paper, 75% of a portfolio return is tied to market movement, 12.5% to asset allocation and 12.5% to active management. I assume active management includes both stock picking and market timing.
Link?
User avatar
Opponent Process
Posts: 5157
Joined: Tue Sep 18, 2007 9:19 pm

Post by Opponent Process »

I agree. how much we make in the markets we invest in will be determined by the returns from those markets. asset allocation is best used to manage risk (e.g., "I have enough money, time to switch to more conservative investments"), not chase after added returns.
30/30/20/20 | US/International/Bonds/TIPS | Average Age=37
User avatar
ddb
Posts: 5511
Joined: Mon Feb 26, 2007 11:37 am
Location: American Gardens Building, West 81st St.

Re: Who is Ibbotson?

Post by ddb »

LHerr wrote:
rustymutt wrote:Who is Ibbotson? I've done my own study over time and have to disagree with them. Rustymutt and colleagues study hasn't found anything other than better control of risk, as well as higher returns with correct asset balance.
Google Roger Ibbotson and you will find many references. He is a well-known researcher in the investment world. His research firm was purchased by Morningstar a few years ago.
In my opinion, while Ibbotson Associates used to put out great research, they appear to have become a 'pay-for-research' organization. I've seen some papers from them on commodities, variable annuities, and active management which have questionable methodology and conclusions. It is typically useful to find who sponsored the paper, and you can quickly figure out the conclusions without even reading the paper.

I assume this is the article that you are referencing? If so, this article comes from Ibbotson the man, not Ibbotson the firm. Roger Ibbotson is chief investment officer and chariman of Zebra Capital Management, which is, of course, an active management shop.

BTW, the article offers no help in how to use active management as a useful replacement for passive management.

- DDB
"We have to encourage a return to traditional moral values. Most importantly, we have to promote general social concern, and less materialism in young people." - PB
ScottW
Posts: 165
Joined: Tue Mar 25, 2008 8:13 am

Post by ScottW »

I read the paper ddb linked to, and didn't find it to have much substance. For one thing, he never really clarified what he meant by "asset allocation". Does he mean the domestic stock market in aggregate versus bonds or cash? Stocks by style or market cap? Domestic versus international?

I think the entire paper can be summarized in two sentences. Your split between stocks and fixed income is the largest, and therefore most important, driver of your return. Asset allocation (assuming he's referring where in the stock market you invest) affects a much smaller part.

Unless I've completely misunderstand what he's saying, this is kind of true. Fama/French's research is about using risk to increase/decrease your expected return relative to the market over an extended period of time. The "market" return is the main driver, and tweaking your allocation just changes return away from the market.

I also assume he's referring to return over extended periods of time, since clearly, where you invest has huge ramifications over the short term (large growth in the nineties, international and small value during the past decade).
hsv_climber
Posts: 3971
Joined: Tue Sep 22, 2009 7:56 pm

Post by hsv_climber »

Rebalancing is not supposed to increase return, just adjust the risk.

There is a very good example about it in Swedroe's "The Only Guide You'll Ever Need for the Right Financial Plan..." book.
According to Fama/French & others, the highest possible return can be obtained with more risk, e.g. SCV.

So, 100% SCV portfolio is expected to produce a better return than a 70/30 balanced portfolio over X years, assuming that investor will never sell it and would not worry about tracking error and such.
User avatar
Dick Purcell
Posts: 520
Joined: Tue Oct 26, 2010 1:58 am

Post by Dick Purcell »

The Ibbotson firm was one of the pioneers and missionary hucksters of the two-step malpractice best labeled mislead-and-fleece, which still dominates the training, certification, and regulation of “fiduciary” investment advisors.

The Ibbotson firm promoted and executed that two-step malpractice of mislead-and-fleece through a product named Portfolio Strategist and a companion product named Security Classifier. Using the theory of Markowitz, it misled the advisor and the investor to choose an asset-class mix based on comparison in technical measures for the mere volatility-dominated individual year, from which they cannot see what’s best or worst in prospects for the investor’s future – and cannot see the terrible smother-compounding long-term cost of fees. Then, said Ibbotson software encouraged them to switch from investment in the selected asset classes to any of thousands of gambles on individual actively managed funds or other speculations, which as we know have lower average net returns, greater uncertainty and risk – and higher fees. For the latter step, the switch, that Ibbotson firm's software presented the deceptive appearance that any fool gamble within an asset class is equivalent to investing in the whole diversified asset class.

The cited new paper features some of the same tricks. In attributing effects of (a) asset allocation and (b) active manager selection, it refers to effects of the two as equivalent, each “return.” Irresponsibly again, it fails to make the essential distinction. For investors’ longer-term prospects, asset-class allocation is of overwhelming importance – across the range from conservative to aggressive, the differences in upside likelihoods and uncertainty ranges are simply overwhelming. It’s based on the best (only) grounds we have for estimating future return-rate probabilities: whole asset classes’ decades of return-rate history, and widest diversification to minimize risks of speculating on particular businesses or “investment managers.” By contrast, active manager selection is gambling diversion from the asset-class allocation, lowering average net returns, due to extractions of higher fees – and increasing uncertainties and risks to such an extent that the very grounds for estimating long-term result probabilities are abandoned.

The CFA Institute publishes and takes ownership of this?? It’s this kind of stuff that keeps millions of investors and their “advisors” lost in the financial fog, misleading and fleecing, unable to see the wisdom of John Bogle and the Bogleheads.

Dick Purcell
Last edited by Dick Purcell on Fri Nov 19, 2010 8:47 am, edited 1 time in total.
User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 7:51 pm

Post by bob90245 »

Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.
User avatar
Topic Author
LHerr
Posts: 289
Joined: Sat May 17, 2008 7:47 am
Location: Portland, OR

Post by LHerr »

bob90245 wrote:I always thought Ibbotson set the record straight here:

Does Asset Allocation Policy Explain. 40%, 90%, or 100% of Performance? by Roger Ibbotson and Paul Kaplan
Bob,

I have the paper you reference above and the more recent work contradicts the "40%, 90%, or 100%" paper. How can one shift from asset allocation contributing 100% return to 12.5%? Is the research approach viable?

Maybe I am wrong, but I am of the opinion all these studies are flawed as the portfolios are limited to three large asset classes - stocks, bonds, and cash. Within those three asset classes portfolios vary to such a degree I'm not sure one can come to the conclusions attributed to Ibbotson. The variations in the conclusions are of such a magnitude as to automatically raise a level of doubt.

LHerr
User avatar
ddb
Posts: 5511
Joined: Mon Feb 26, 2007 11:37 am
Location: American Gardens Building, West 81st St.

Post by ddb »

Dick Purcell wrote:Ibbotson is a dangerous man, deserves to be ignored.

That firm of his was one of the pioneers and missionary hucksters of the two-step malpractice best labeled mislead-and-fleece, which still dominates the training, certification, and regulation of “fiduciary” investment advisors.
Nice to see you here, Dick. You and I participated on a different financial-related forum several years back, and I've always enjoyed your contributions.

- DDB
"We have to encourage a return to traditional moral values. Most importantly, we have to promote general social concern, and less materialism in young people." - PB
rai
Posts: 1342
Joined: Tue Apr 06, 2010 7:11 am

Re: Who is Ibbotson?

Post by rai »

LHerr wrote: In the recent Ibbotson paper, 75% of a portfolio return is tied to market movement, 12.5% to asset allocation and 12.5% to active management.
well, I don't understand this. We can not control market movement, so what can we do about it?

I mean can we predict 9/11 or BP oil spill or a thousand other events that are out of our control yet effect the market?

What is the use of 'market movement' if we can not do anything about it?

If active management means market timing? How can we know if we are timing the market correctly? Warren Buffett for example lost Billions with the poor market results in 2008, are we expected to do any better than that? How?

We can control Asset allocation, that to me makes it more important that those things that are un-knowable or out of my control.
Last edited by rai on Tue Nov 02, 2010 12:00 pm, edited 3 times in total.
"Life is what happens to you while you're busy making other plans" - John Lennon. | | "You say that money, isn't everything | But I'd like to see you live without it." - Silverchair
User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 7:51 pm

Post by bob90245 »

LHerr wrote:
bob90245 wrote:I always thought Ibbotson set the record straight here:

Does Asset Allocation Policy Explain. 40%, 90%, or 100% of Performance? by Roger Ibbotson and Paul Kaplan
Bob,

I have the paper you reference above and the more recent work contradicts the "40%, 90%, or 100%" paper. How can one shift from asset allocation contributing 100% return to 12.5%? Is the research approach viable?
You question demands an answer. I thought it was common practice in academic papers to cite relevant prior work and then explain how that prior work is now supplanted by current evidence. As your question raises, this is an important piece that is missing.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.
User avatar
Dick Purcell
Posts: 520
Joined: Tue Oct 26, 2010 1:58 am

Post by Dick Purcell »

It’s pure 100% deceptive foolishness. The “active management” component is the “make the prospects worse” component.

Picture a frontier of mixes of asset classes. If you invest in one of those mixes of asset classes, via index funds, you are essentially on the curve, except for the very low fees.

If after selecting the mix, one switches to “active management” things within the asset classes, he is moving into fog below and/or to right of the curve, generally both lower and to right. Lower because of the greater costs and fees, to right because of the uncertainty-increasing speculation on particular “investment managers.”

The extent to which the switch from asset classes to gambles on “active management” makes the prospects worse and dominates the prospects depends on how foolish the choices of “active management” things are. For example, at about the time Enron was crumbling, the Ibbotson firm's software showed me that for the Large Class Growth asset class, instead of the whole asset class I could put all my money in Enron and have the same prospects. If I had been foolish enough to follow this choice offered by the Ibbotson firm's software, my results would have been determined 100% by “active management.”

Dick Purcell
Last edited by Dick Purcell on Fri Nov 19, 2010 8:49 am, edited 1 time in total.
User avatar
SSSS
Posts: 1914
Joined: Fri Jun 18, 2010 11:50 am

Re: Importance of Asset Allocation

Post by SSSS »

LHerr wrote:and the most important factor to portfolio movement is market movement.
Since market movement can't be controlled, isn't this irrelevant?
User avatar
Charybdis
Posts: 242
Joined: Sun Apr 25, 2010 4:38 am
Location: Budapest, Hungary

Post by Charybdis »

If you do passive indexing, then asset allocation explains 100% of the level of your returns.

If you do active management, then the importance of asset allocation depends on how active you are. The more active you are, the less impact has the asset allocation on your returns.
User avatar
Topic Author
LHerr
Posts: 289
Joined: Sat May 17, 2008 7:47 am
Location: Portland, OR

Post by LHerr »

"Well, I don't understand this. We can not control market movement, so what can we do about it? "

In two words - Not Much. On any given day, if the broad market moves up, my index oriented portfolio also moves up. When the broad market declines, my portfolio usually declines. I define "broad market" by the VTSMX Vanguard Total Market Index Fund.

I use the VTSMX as a portfolio benchmarks. In addition to measuring portfolio return vs. the benchmark, I calculate the Sortino Ratio (SR) or a slight modification of the SR to determine portfolio risk. When updating a portfolio, I am interested in what the SR is doing as it takes both the portfolio return and portfolio risk into consideration.

Even if a portfolio includes a wide variety of asset classes, I observe that portfolio movement is highly correlated with market movement.

LHerr
sscritic
Posts: 21853
Joined: Thu Sep 06, 2007 8:36 am

Post by sscritic »

Charybdis wrote:If you do passive indexing, then asset allocation explains 100% of the level of your returns.
The answer to a question depends on what the question is. When the question is not clearly defined, any answer can be the answer to the question.

I own TSM. In year one, I lost 40%. In year two, I gained 50%. My allocation explained zero percent of my returns. I got two totally different returns with exactly the same allocation.

I own TSM. My twin owns IPS. In one year, I lost 40%. In the same year, my twin gained 10%. The difference in our allocations explained 100% of the difference in returns. Clearly, I should have owned Tips that year.

Two different questions, two totally different answers.
User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 7:51 pm

Post by bob90245 »

sscritic wrote:I own TSM. In year one, I lost 40%. In year two, I gained 50%. My allocation explained zero percent of my returns.
You baffle me. Your returns are explained 100% (not zero percent) to your allocation completely composed of TSM.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.
sscritic
Posts: 21853
Joined: Thu Sep 06, 2007 8:36 am

Post by sscritic »

bob90245 wrote:
sscritic wrote:I own TSM. In year one, I lost 40%. In year two, I gained 50%. My allocation explained zero percent of my returns.
You baffle me. Your returns are explained 100% (not zero percent) to your allocation completely composed of TSM.
No. I held the same allocation and got two different returns. If I plot holdings (in some fashion) on the X axis and returns on the Y axis, I get two points, one directly above the other. The correlation is undefined as the the "values" of X are identical. If asset allocation explained my returns, then keeping my allocation constant would produce constant returns. My returns weren't constant, so asset allocation can't be the explanation for my varying returns (the market is the explanation).

The issue is that my one example held time constant and I changed the asset allocation. In the other, I held the allocation constant and changed time. Those represent two different questions with two different answers.

Now you could change the question to how do my returns differ from the market returns over time, but that is a third question with a third answer. But if the question is the actual level of my returns relative to zero and not to some benchmark, then asset allocation isn't 100% the answer.
User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 7:51 pm

Post by bob90245 »

sscritic wrote:
bob90245 wrote:
sscritic wrote:I own TSM. In year one, I lost 40%. In year two, I gained 50%. My allocation explained zero percent of my returns.
You baffle me. Your returns are explained 100% (not zero percent) to your allocation completely composed of TSM.
No. I held the same allocation and got two different returns. If I plot holdings (in some fashion) on the X axis and returns on the Y axis, I get two points, one directly above the other. The correlation is undefined as the the "values" of X are identical. If asset allocation explained my returns, then keeping my allocation constant would produce constant returns.
You're not going to get constant returns holding 100% TSM which you already explained from you example: year one had a loss of 40% and year two had a gain of 50%.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.
sscritic
Posts: 21853
Joined: Thu Sep 06, 2007 8:36 am

Post by sscritic »

Bob:

Exactly. They were silly examples to show the "level" of return is determined by the market, not by asset allocation.

The problem is that most people reading these studies don't understand R-squared. R^2 gives a number that should be interpreted as the fraction of the variation in Y that can be explained by variation in X. It does not have anything to do with explaining the level of Y by the level of X.

Example: Take your favorite set of X and Y. Find R^2. Add 100 to every Y. R^2 is exactly the same. R^2 tells you nothing about whether you are at the original Y level or 100 points higher. R^2 tells you nothing about the level of Y.

Now subtract 10 from every X value. Find R^2. It is unchanged, exactly the same as for the original numbers or the original numbers with the 100 added to every Y. R^2 tells you nothing about what level of X will get you a certain level of Y.

Now triple all the Y values. R^2 is still unchanged. Instead of getting returns of -10, -3, 5, and 20 with your different asset allocations, you get returns of -30, -9, 15, and 60. R^2 not only doesn't tell you anything about the levels of your returns, it doesn't tell you anything about the actual value of the variation in your returns (the standard deviation or variance of the returns). R^2 tells you nothing about the variance of the levels of returns.

These examples all have the same R^2, but different levels and different variances.

An R^2 of 40% doesn't tell you that 40% of the your return levels is explained by X nor does it tell you that 40% of the variance of your return levels is explained by X. It does tell you that 40% of the variation in your return levels (whatever that value is) is explained by the variation in X.
lazyday
Posts: 3849
Joined: Wed Mar 14, 2007 10:27 pm

Re: Importance of Asset Allocation

Post by lazyday »

LHerr wrote:My question is, does asset allocation take on more importance than Ibbotson gives credit if one builds a well-diversified portfolio with more than the three classic asset classes of stocks, bonds, and cash?
Gibson and others have written about portfolios containing REITS and other classes. In Gibson's Asset Allocation book, a quick easy read, he shows how using very different, uncorrelated classes including REITS (and commodities IIRC) can lead to drastically different returns. Mostly with graphs, not statistics. Traditional portfolios are also included, so you might get some good hints to answer your question.


Stepping back, I don't know if any of these studies address the failure of many investors to stick to their AA. According to some studies, and data shown in Morningstar, the average individual investor manages to underperform by chasing the latest hot asset class repeatedly. Maybe this gets muddled into the data.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Post by grap0013 »

hsv_climber wrote:Rebalancing is not supposed to increase return, just adjust the risk.
As long as your talkin' Swedroe, I think he has also commented on this somewhere, but if all your assets have the same expected returns, I believe he states (I agree), you can juice your returns and get a rebalancing bonus. i.e. A 100% equity portfolio of large blend both domestically and internationally or better yet, a SCV domest/international rebalanced on trigger bands. There will be times when they are not priced as equals and your rebalancing will force you to take advantage of the "unfair" pricing.
There are no guarantees, only probabilities.
User avatar
Topic Author
LHerr
Posts: 289
Joined: Sat May 17, 2008 7:47 am
Location: Portland, OR

Post by LHerr »

"Gibson and others have written about portfolios containing REITS and other classes. In Gibson's Asset Allocation book, a quick easy read, he shows how using very different, uncorrelated classes including REITS (and commodities IIRC) can lead to drastically different returns. Mostly with graphs, not statistics. Traditional portfolios are also included, so you might get some good hints to answer your question."

Yes, I've read Gibson and highly recommend his book on Asset Allocation. He makes a case for including commodities in a portfolio.

LHerr
rustymutt
Posts: 4001
Joined: Sat Mar 07, 2009 11:03 am

Post by rustymutt »

bob90245 wrote:
sscritic wrote:
bob90245 wrote:
sscritic wrote:I own TSM. In year one, I lost 40%. In year two, I gained 50%. My allocation explained zero percent of my returns.
You baffle me. Your returns are explained 100% (not zero percent) to your allocation completely composed of TSM.
No. I held the same allocation and got two different returns. If I plot holdings (in some fashion) on the X axis and returns on the Y axis, I get two points, one directly above the other. The correlation is undefined as the the "values" of X are identical. If asset allocation explained my returns, then keeping my allocation constant would produce constant returns.
You're not going to get constant returns holding 100% TSM which you already explained from you example: year one had a loss of 40% and year two had a gain of 50%.
You could reduce risk of this investment by holding something that is not correlated to TSM, such as bonds.
Even educators need education. And some can be hard headed to the point of needing time out.
rustymutt
Posts: 4001
Joined: Sat Mar 07, 2009 11:03 am

Post by rustymutt »

LHerr wrote:"Gibson and others have written about portfolios containing REITS and other classes. In Gibson's Asset Allocation book, a quick easy read, he shows how using very different, uncorrelated classes including REITS (and commodities IIRC) can lead to drastically different returns. Mostly with graphs, not statistics. Traditional portfolios are also included, so you might get some good hints to answer your question."

Yes, I've read Gibson and highly recommend his book on Asset Allocation. He makes a case for including commodities in a portfolio.

LHerr
Larry Swedroe makes this point in one of his books also. While risk is reduced with commodities, so is return. Go figure.
Even educators need education. And some can be hard headed to the point of needing time out.
Post Reply