TSM, Efficient Markets, and tilting

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Brian2d
Posts: 203
Joined: Thu Jan 20, 2011 9:04 pm

TSM, Efficient Markets, and tilting

Post by Brian2d » Sun Apr 08, 2012 9:42 pm

Hi,

This message is largely in response to Taylor's "Why I don't tilt message" and the accompanying link here: http://www.norstad.org/finance/tsmproofs.pdf from the post on small value tilting. I reposted in as a new topic because my question is more general about efficiency and TSM.

I am definitely overall a Boglehead, believe in minimizing cost and diversification, and TSM is the fund I have the most money in of any of the funds I own (and if it wasn't for realizing gains even more would be there).

However, I do tilt and am not sure I agree with the conclusion that the proofs above imply that one should not tilt.
First, these proofs are conditional on any one of the three definitions of market efficiency being true. If they are not true, then TSM (or lets say appropriate combination of TSM and TISM) may not be the most efficient allocation (I personally believe that markets are weakly efficient but not completely so. I can't make a risk free profit by deviating from the market portfolio but that doesn't mean that the market portfolio is definitely the best one out there IMO).

But lets for the moment presume that markets are efficient.

I may have a different risk and return objective than the market as a whole. If my risk tolerance is greater or less, I should adjust my portfolio to decrease risk or increase the return accordingly, even if it means lower returns or greater risk.

Second, different time horizons may be relevant. The risk-return profile over longer periods is different than that over shorter periods, as there does seem to be some evidence of long term mean reversion. This timing issue should mean what's efficient for one time horizon may not be efficient for another.

Am I missing anything here?...please convince me if I am wrong on anything here? Thanks.

BradMajors
Posts: 396
Joined: Sun Mar 18, 2007 1:56 pm

Re: TSM, Efficient Markets, and tilting

Post by BradMajors » Sun Apr 08, 2012 10:04 pm

The assumption that TSM is the "best" allocation assumes that everyone has the same risk tolerance. If you have a higher risk tolerance then tilting to higher risk assets would make sense.

User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 8:51 pm

Re: TSM, Efficient Markets, and tilting

Post by bob90245 » Sun Apr 08, 2012 10:04 pm

Efficient markets mean that the longer the time frame the less likely it is for active managers to outperform a benchmark index. Think Bill Miller as a recent example.

On the other hand, portfolio efficiency means that there is some combination of stock asset classes that will fall on the efficient frontier. Of course, we can never know that optimal combination ahead of time. But we can look at history to see how it turned out. Click on image for full size.

Image
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.

User avatar
ofcmetz
Posts: 2257
Joined: Tue Feb 08, 2011 8:09 pm
Location: Louisiana

Re: TSM, Efficient Markets, and tilting

Post by ofcmetz » Mon Apr 09, 2012 12:45 pm

I hold total market funds* whenever possible because it is a lazier way to invest and because I'm not willing to make the extra effort to grasp for what might be a slightly greater return. I get to avoid tracking error and know that no matter which asset class excels, that I will own it.

I believe that in some time periods you will be rewarded for the extra risk to take when you tilt. I also believe that certain investors enjoy a bit of complication. Tilting gives them more to tinker with as well as the chance to try to edge out an extra percent or so of return, and there is nothing wrong with this. I think that as more people try to tilt certain ways that it may drive up the price for certain asset classes and reduce the value of owning them. Not to mention that certain classes of equities are more expensive to own.

I think that whether you tilt or not should be about knowing yourself as an investor.

Brian2d wrote:But lets for the moment presume that markets are efficient.

I may have a different risk and return objective than the market as a whole. If my risk tolerance is greater or less, I should adjust my portfolio to decrease risk or increase the return accordingly, even if it means lower returns or greater risk.


If you hold a total market portfolio then this adjustment to increase or decrease risk is very easy. You just hold more of the total bond market fund than the total stock market fund or vice versa.

A good friend of mine who invests in the Thrift Saving Plan holds almost all of his equities in the S fund (extended market small cap fund). This is certainly a viable way to increase risk for those tilters with a strong stomach.


Brian2d wrote:Second, different time horizons may be relevant. The risk-return profile over longer periods is different than that over shorter periods, as there does seem to be some evidence of long term mean reversion. This timing issue should mean what's efficient for one time horizon may not be efficient for another.

Am I missing anything here?...please convince me if I am wrong on anything here? Thanks


I think equities are risky in general. Certain people give 5 year, 10 year, or even 15 year rules when it comes to determining how long you should be willing to let your money ride in equities before planing to withdraw it. I subscribe to the "decide how much to put in bonds and cash and then invest the rest for return rule". (just made that up).

As far as reversion to the mean, John Bogle has excellent writing on this in his book Common Sense on Mutual Funds. I recall other authors writing about it as well. From what I remember it occurs to various mutual funds as well as individual asset classes. For me it confirms my decision to go for total market funds.

*All this said, I do invest in REIT funds.
Never underestimate the power of the force of low cost index funds.

chaz
Posts: 13601
Joined: Tue Feb 27, 2007 2:44 pm

Re: TSM, Efficient Markets, and tilting

Post by chaz » Mon Apr 09, 2012 12:48 pm

Tilting is not for everyone, including me.
Chaz | | “Money is better than poverty, if only for financial reasons." Woody Allen | | http://www.bogleheads.org/wiki/index.php/Main_Page

User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 8:51 pm

Re: TSM, Efficient Markets, and tilting

Post by bob90245 » Mon Apr 09, 2012 12:57 pm

ofcmetz wrote:As far as reversion to the mean, John Bogle has excellent writing on this in his book Common Sense on Mutual Funds. I recall other authors writing about it as well. From what I remember it occurs to various mutual funds as well as individual asset classes. For me it confirms my decision to go for total market funds.

Just remember that the data Bogle uses on mutual funds is notoriously subject to style drift. Hence, I would take any conclusions derived from there with a grain of salt.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Mon Apr 09, 2012 1:11 pm

bob90245 wrote:On the other hand, portfolio efficiency means that there is some combination of stock asset classes that will fall on the efficient frontier. Of course, we can never know that optimal combination ahead of time. But we can look at history to see how it turned out. Click on image for full size.


Keep in mind after Gibson posted that famous chart in his book, he then, in full disclosure, shifted the year range by just one year and got a completely different result. And of course, markets don't consult the history books when moving.

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: TSM, Efficient Markets, and tilting

Post by Rodc » Mon Apr 09, 2012 1:23 pm

BradMajors wrote:The assumption that TSM is the "best" allocation assumes that everyone has the same risk tolerance. If you have a higher risk tolerance then tilting to higher risk assets would make sense.


Actually no.

If you have a lower risk tolerance than average or the representative market participant, the theory that says TSM is still optimal, but should be diluted with cash (CDs, T-Bills, etc.) and if you want more risk you should use leverage (assuming you can borrow at the risk free rate).

I'm not say I believe this, just saying that the theory of the optimality of TSM does not suppose everyone has the same risk tolerance.

Added: see the figure in this link: http://en.wikipedia.org/wiki/Efficient_frontier
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

richard
Posts: 7961
Joined: Tue Feb 20, 2007 3:38 pm
Contact:

Re: TSM, Efficient Markets, and tilting

Post by richard » Mon Apr 09, 2012 1:30 pm

Brian2d wrote:But lets for the moment presume that markets are efficient.

I may have a different risk and return objective than the market as a whole. If my risk tolerance is greater or less, I should adjust my portfolio to decrease risk or increase the return accordingly, even if it means lower returns or greater risk.

Second, different time horizons may be relevant. The risk-return profile over longer periods is different than that over shorter periods, as there does seem to be some evidence of long term mean reversion. This timing issue should mean what's efficient for one time horizon may not be efficient for another.

Am I missing anything here?...please convince me if I am wrong on anything here? Thanks.

If markets are efficient, then the total market portfolio must be appropriate for the average/representative investor. However, if you are meaningfully different from the representative investor, another portfolio may be better for you.

What do you mean by risk tolerance? Many people mean they believe they have more psychological ability than others to stay the course. Others mean they have more economic ability to ride out bad times.

Different time horizons may suggest different allocations. However, I'd be very cautious about this. Whether or not stocks will mean revert over your relevant horizon is far from certain and does not, IMO, represent a sound basis for tilting.

Why would you believe tilting is better for you than changing your stock/bond mix?

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Mon Apr 09, 2012 2:46 pm

Brian2d wrote:I may have a different risk and return objective than the market as a whole. If my risk tolerance is greater or less, I should adjust my portfolio to decrease risk or increase the return accordingly, even if it means lower returns or greater risk.


Concentrating your portfolio in the name of "I want increased risk" may or may not be a good idea. The market portfolio certainly maximizes economic exposure - there are 20 candy bars at 7-11, and you're investing $1 on each of the 20. Any tilt is betting on specific bars. Tilting isn't diversifying into a new asset class with its own risk profile - it's distorting an existing class.

As richard points out, tilting this raises a lot of questions. Why do you need to do this vs adjusting your stock/bond split? What do you mean by "tilt"? Value tilting is probably a waste of time in any case, small tilting might be ok.

Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Re: TSM, Efficient Markets, and tilting

Post by Jerry_lee » Mon Apr 09, 2012 3:25 pm

yobria wrote:
Brian2d wrote:I may have a different risk and return objective than the market as a whole. If my risk tolerance is greater or less, I should adjust my portfolio to decrease risk or increase the return accordingly, even if it means lower returns or greater risk.


Concentrating your portfolio in the name of "I want increased risk" may or may not be a good idea. The market portfolio certainly maximizes economic exposure - there are 20 candy bars at 7-11, and you're investing $1 on each of the 20. Any tilt is betting on specific bars. Tilting isn't diversifying into a new asset class with its own risk profile - it's distorting an existing class.

As richard points out, tilting this raises a lot of questions. Why do you need to do this vs adjusting your stock/bond split? What do you mean by "tilt"? Value tilting is probably a waste of time in any case, small tilting might be ok.


Not true at all. By spreading you portfolio dollars more gradually across a wide variety of names instead of betting heavily on a handful of the largest and highest priced companies, you are exposed to a wider variety of economic forces and positive expected returns. Since 2000, the large cap blue chip stocks of the s&P 500 have trailed inflation, but small companies and lower priced value stocks have produced significant excess returns.

Just like a 7-11, that sells everything from food to drinks to tobacco to pharmaceuticals to periodicals to gasoline (a diversified asset class portfolio) compared to a gas station with a vending machine outside (a total stock portfolio).

And, just to clarify, "tilting" raises approximately 3 questions:
a) your stock and bond split
b) your small vs. large split
c) your value vs. growth split

Why would you tilt on the equity side towards small and value instead of increasing your equity (vs. fixed income) allocation? Well, lets see, since 2000, the Russell 3000 Index compounded at +1.2% per year, whereas the Barclay's ST Bond Index produced +5.0% per year. If one were to try and increase their portfolio returns, how would they faired by going from 60% stock (Russell 3000)/40% bond (ST Bond Index) to 80% stock/ 20% bond? Despite higher risk, you'd have lost 0.9% per year of return.

What if, on the other, hand, the original investor instead diversified more broadly into 20% Russell 3000, 20% Russell Mid Value, 20% Russell 2000 Value, and 40% ST Bond Index? Both mid value and small value had higher risk than the market (remember the fall of '08?), but because their returns were also sensitive to value and size factors instead of just beta, and because size and value have low to negative correlations over time with the market factor, the asset class portfolio accomplished the goal of higher returns (+2.9% per year more than our Total Stock 60/40 mix). And despite adding more volatile equity components, their less than perfect correlation with the market meant that the 60/40 asset class portfolio only had 0.1% higher standard deviation than the 60/40 total stock portfolio.

So, it seems, modern portfolio theory, when applied to asset classes, has continued to hold up after almost 50 years. Other financial theories, like the Capital Asset Pricing Model, and it's direct decedent Total Market Investing, have long since proven to be woefully incomplete.
The most disciplined investor in the world.

User avatar
rustymutt
Posts: 3646
Joined: Sat Mar 07, 2009 12:03 pm
Location: Oklahoma

Re: TSM, Efficient Markets, and tilting

Post by rustymutt » Mon Apr 09, 2012 3:36 pm

I've got news for all you so called no tilt people out there, if you own VTI, or any other single ETF, or mutual fund that owns the markets, then you're tilted toward large cap stocks. So dump this argument about tilting being a bad thing. Historical those stocks with the higher risk premium, have produced better returns.
Last edited by rustymutt on Wed Apr 11, 2012 1:08 pm, edited 2 times in total.
Knowledge is knowing that the Tomato is a fruit. Wisdom is knowing better than to put the tomato in a fruit salad.

richard
Posts: 7961
Joined: Tue Feb 20, 2007 3:38 pm
Contact:

Re: TSM, Efficient Markets, and tilting

Post by richard » Mon Apr 09, 2012 3:42 pm

Jerry_lee wrote:Not true at all. By spreading you portfolio dollars more gradually across a wide variety of names instead of betting heavily on a handful of the largest and highest priced companies, you are exposed to a wider variety of economic forces and positive expected returns. Since 2000, the large cap blue chip stocks of the s&P 500 have trailed inflation, but small companies and lower priced value stocks have produced significant excess returns.

Assume an economy with 20 factories, each with about the same value. Company A owns 19 factories. Company B owns 1 factory.

Would you own more of A than B? Which would be more diversified, more A or 50/50 A and B?

Put another way, should you diversify by economic resource or by organizational structure?

Jerry_lee wrote:So, it seems, modern portfolio theory, when applied to asset classes, has continued to hold up after almost 50 years. Other financial theories, like the Capital Asset Pricing Model, and it's direct decedent Total Market Investing, have long since proven to be woefully incomplete.

MPT is Markowitz's mean variance model introduced in 1952. It works to the extent you regard variance as an excellent proxy for risk. CAPM was introduced about 10 years later and is a direct descendant of MPT. The main utility of these models is to show that expected return and risk are related. There have been a number of subsequent models that try for better ways to proxy for risk. The Fama French three factor model is a prominent example of this, but there are a bunch of other multi-factor models.

Total market investing can be easily derived from the assumption that markets are efficient. It is not at all dependent on CAPM (other than MPT's and CAPM's insight about the relation of expected return and risk). As Gene Fama said, the total market is always on the efficient frontier.

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Mon Apr 09, 2012 3:44 pm

rustymutt wrote:I've got news for all you so called no tilt people out there, if you own VTI, or any other single ETF, or mutual fund that owns the markets, then you're tilted toward large cap stocks. So dump the this argument about tilting being a bad thing.


VTI et al are "tilted" to whatever the market says has greatest economic value. Since I don't have any information the market doesn't, I'll assume Apple has a bigger market cap than Joe's Hot Dog Stand for a reason, and will weight accordingly.

User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 8:51 pm

Re: TSM, Efficient Markets, and tilting

Post by bob90245 » Mon Apr 09, 2012 3:45 pm

yobria wrote:
bob90245 wrote:On the other hand, portfolio efficiency means that there is some combination of stock asset classes that will fall on the efficient frontier. Of course, we can never know that optimal combination ahead of time. But we can look at history to see how it turned out. Click on image for full size.

Image

Keep in mind after Gibson posted that famous chart in his book, he then, in full disclosure, shifted the year range by just one year and got a completely different result. And of course, markets don't consult the history books when moving.

He was likely looking at a short time period like ten years. Charts don't shift to something completely different incrementing by one year when the time period being examined is twenty years or more.

I even checked it out on my own four-asset-class spreadsheet that you can down load here:

http://bobsfinancialwebsite.com/download.html#Frontier

www.bobsfinancialwebsite.com wrote:Risk-Return Curve

This Excel file has a chart showing risk versus return for the S&P 500, Small-Cap Stocks, Intermediate-Term Government Bonds and 90-Day Treasury Bills. The risk-return curve is plotted between fixed income (bonds and bills) and equities (S&P 500 and small-cap stocks). You may vary the timeframe from 5 to 50 years. Data can be displayed in nominal values or inflation-adjusted (real) values. Data is from 1926 to 2010.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Mon Apr 09, 2012 4:01 pm

yobria wrote:
rustymutt wrote:I've got news for all you so called no tilt people out there, if you own VTI, or any other single ETF, or mutual fund that owns the markets, then you're tilted toward large cap stocks. So dump the this argument about tilting being a bad thing.


VTI et al are "tilted" to whatever the market says has greatest economic value. Since I don't have any information the market doesn't, I'll assume Apple has a bigger market cap than Joe's Hot Dog Stand for a reason, and will weight accordingly.


And Enron as well.
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

kontango
Posts: 114
Joined: Sun Dec 18, 2011 11:27 am

Re: TSM, Efficient Markets, and tilting

Post by kontango » Mon Apr 09, 2012 4:20 pm

If you have a higher risk tolerance then you should borrow money to buy the TSM (i.e., lever up). If you have a lower risk tolerance than TSM you should sell some TSM and buy risk-free securities (Treasury securities).

User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 8:51 pm

Re: TSM, Efficient Markets, and tilting

Post by bob90245 » Mon Apr 09, 2012 4:36 pm

Beagler wrote:
yobria wrote:
rustymutt wrote:I've got news for all you so called no tilt people out there, if you own VTI, or any other single ETF, or mutual fund that owns the markets, then you're tilted toward large cap stocks. So dump the this argument about tilting being a bad thing.

VTI et al are "tilted" to whatever the market says has greatest economic value. Since I don't have any information the market doesn't, I'll assume Apple has a bigger market cap than Joe's Hot Dog Stand for a reason, and will weight accordingly.

And Enron as well.

And all the tech-bombs and dot-bombs that were loaded into VTI (or it's equivalent) during the late 1990's. Their great economic value got sliced by half or more during the 2000-2002 bear market with very few recovering to their former peaks.

By the way, I don't understand where this notion that the market price is always right is still treated as doctrine here on Bogleheads. :shock:
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Mon Apr 09, 2012 4:40 pm

yobria wrote:The market portfolio certainly maximizes economic exposure - there are 20 candy bars at 7-11, and you're investing $1 on each of the 20.


Since when is TSM equal-weighted and not cap-weighted?
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Mon Apr 09, 2012 4:42 pm

richard wrote:
Jerry_lee wrote:Not true at all. By spreading you portfolio dollars more gradually across a wide variety of names instead of betting heavily on a handful of the largest and highest priced companies, you are exposed to a wider variety of economic forces and positive expected returns. Since 2000, the large cap blue chip stocks of the s&P 500 have trailed inflation, but small companies and lower priced value stocks have produced significant excess returns.

Assume an economy with 20 factories, each with about the same value. Company A owns 19 factories. Company B owns 1 factory.

Would you own more of A than B? Which would be more diversified, more A or 50/50 A and B?


Right or, continuing the example, Company A has 19% of the market cap of the total stock market, selling 95% of the products in its industry. One day it splits into 19 companies, each with 1% of the total market cap, each now selling 5% of the products consumers demand in that space. Would your weighting of Company A vary dramatically before and after the split? How big a bet would you place on any of the 19 companies?

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Mon Apr 09, 2012 4:45 pm

bob90245 wrote:
Beagler wrote:
yobria wrote:
rustymutt wrote:I've got news for all you so called no tilt people out there, if you own VTI, or any other single ETF, or mutual fund that owns the markets, then you're tilted toward large cap stocks. So dump the this argument about tilting being a bad thing.

VTI et al are "tilted" to whatever the market says has greatest economic value. Since I don't have any information the market doesn't, I'll assume Apple has a bigger market cap than Joe's Hot Dog Stand for a reason, and will weight accordingly.

And Enron as well.

And all the tech-bombs and dot-bombs that were loaded into VTI (or it's equivalent) during the late 1990's. Their great economic value got sliced by half or more during the 2000-2002 bear market with very few recovering to their former peaks.


And the thousands of savings and loans (those would be small value) that got more and more value-y until they disappeard in the 80s. Or the small caps that dropped 90% during the Great Depression. Unfortunately these are arguments for diversificaiton, not concentration.

Khanmots
Posts: 1178
Joined: Sat Jun 11, 2011 2:27 pm

Re: TSM, Efficient Markets, and tilting

Post by Khanmots » Mon Apr 09, 2012 4:51 pm

kontango wrote:If you have a higher risk tolerance then you should borrow money to buy the TSM (i.e., lever up). If you have a lower risk tolerance than TSM you should sell some TSM and buy risk-free securities (Treasury securities).


Investing on margin adds expenses and additional risks that alter the risk/reward ratio.

Also, one must consider how rebalancing between asset classes that aren't perfectly correlated affects the risk/reward ratio of the portfolio as a whole.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Mon Apr 09, 2012 4:52 pm

Image Image Image
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Mon Apr 09, 2012 4:53 pm

yobria wrote:.... Or the small caps that dropped 90% during the Great Depression.....


And the large-cap market which dropped 80%....
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Mon Apr 09, 2012 4:59 pm

yobria wrote:...these are arguments for diversificaiton, not concentration.


Agreed.

Rodc wrote:
Which is more diversified:

A) 100% S&P 500

B) X% S&P 500 + (1-X%) SC?

In A you have a few hundred stocks, at most, that matter. And pretty rapidly as you come down from the largest cap each new company matters less and less. By the bottom of the stack a company hardly matters at all. Compare to TSM and the bottom several thousand companies in TSM hardly matter at all.

In B you have those same few hundred stocks that matter, and if X% is greater than the 20% weight you get from TSM, you have a few more hundred stocks that matter.

By what definition is A more diversified?

Since TSM is not materially different from A, by what definition is TSM more diversified than A or B?

There is a nice bit of theory that says TSM weights are optimal by some return vs risk metric. Is that what most diversified means? If so what does the annoying fact that TSM is virtually the same as S&P 500 say about this theory? Or does it mean don't put all your eggs in one (or too few) baskets?
http://tinyurl.com/7khyo5g Emphasis mine.


Returns for TSM and VG's S&P 500 index fund from VTSMX's inception:
Image
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

User avatar
Clearly_Irrational
Posts: 3087
Joined: Thu Oct 13, 2011 3:43 pm

Re: TSM, Efficient Markets, and tilting

Post by Clearly_Irrational » Mon Apr 09, 2012 5:04 pm

Beagler wrote:
yobria wrote:.... Or the small caps that dropped 90% during the Great Depression.....


And the large-cap market which dropped 80%....


Prior to indexes and the ability to access them through mutual funds/ETFs I would have been a lot more hesitant to own small cap stocks due to individual security risk.

pingo
Posts: 2571
Joined: Sat Sep 19, 2009 8:24 pm

Re: TSM, Efficient Markets, and tilting

Post by pingo » Mon Apr 09, 2012 6:06 pm

Brian2d wrote:Second, different time horizons may be relevant. The risk-return profile over longer periods is different than that over shorter periods, as there does seem to be some evidence of long term mean reversion. This timing issue should mean what's efficient for one time horizon may not be efficient for another.


I'm on the fence, so I shall continue to be market-weight biased in the execution of my own portfolio until I am no longer a fence sitter. But the OP provokes the following question (for me):

What is one supposed to do with the tilt(s) as my longer time horizon becomes a shorter one?

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: TSM, Efficient Markets, and tilting

Post by Rodc » Wed Apr 11, 2012 7:03 am

richard wrote:
Jerry_lee wrote:Not true at all. By spreading you portfolio dollars more gradually across a wide variety of names instead of betting heavily on a handful of the largest and highest priced companies, you are exposed to a wider variety of economic forces and positive expected returns. Since 2000, the large cap blue chip stocks of the s&P 500 have trailed inflation, but small companies and lower priced value stocks have produced significant excess returns.

Assume an economy with 20 factories, each with about the same value. Company A owns 19 factories. Company B owns 1 factory.

Would you own more of A than B? Which would be more diversified, more A or 50/50 A and B?

Put another way, should you diversify by economic resource or by organizational structure?

Jerry_lee wrote:So, it seems, modern portfolio theory, when applied to asset classes, has continued to hold up after almost 50 years. Other financial theories, like the Capital Asset Pricing Model, and it's direct decedent Total Market Investing, have long since proven to be woefully incomplete.

MPT is Markowitz's mean variance model introduced in 1952. It works to the extent you regard variance as an excellent proxy for risk. CAPM was introduced about 10 years later and is a direct descendant of MPT. The main utility of these models is to show that expected return and risk are related. There have been a number of subsequent models that try for better ways to proxy for risk. The Fama French three factor model is a prominent example of this, but there are a bunch of other multi-factor models.

Total market investing can be easily derived from the assumption that markets are efficient. It is not at all dependent on CAPM (other than MPT's and CAPM's insight about the relation of expected return and risk). As Gene Fama said, the total market is always on the efficient frontier.


To the first point, this is entirely valid and to the extent that if I was forced to only own one or a few companies I would own large caps that have interests in many sectors. Think Berkshire Hathaway. And I would shy away from the Apples (Enron, etc.) that are highly concentrated in one industry. On the other hand if I can own two funds at low cost, one concentrated in large caps (note that of the top five companies Apple, Exxon, Microsoft, IBM and Chevron in US TSM all are fairly concentrated each in one sector either tech or oil, not until #6 General Electric do you get to a more broadly based company) and one in smaller companies (like the mid-small Vanguard Small Cap) it is unclear why I would want to skip having more companies playing a meaningful role, especially if I am not too worried about risk showing up at the wrong time (job security largely independent of whatever the market is doing).

As to MPT, as far as I know it never really "works". Not only do you need to fully believe standard deviation really is risk, you have to believe you can accurately estimate the 5000x5000 or so covariance matrix for stocks if you are looking to cover the stocks in TSM (include international and you need approximately twice the size matrix or 4 times as many individual covariance values). And we can't even do a decent job of estimating the underlying true mean return (if such a thing even exists which it almost certainly does not). In practice MPT is numerically pretty unstable and thus while a nice theory for thinking about how markets work is not useful to investors.

Like MPT, market efficiency leading to the Market Portfolio of all risky assets being optimal has many assumptions about risk, investor behavior, etc etc. To apply this logic it to one slice on the order of 20% or less of the Market Portfolio (US TSM) is a further stretch. TSM is a great holding which has a lot going for it. But unclear it has all the magical properties assigned from applying simple models that really are about all risky assets with one particular smallish slice held by individuals not the mythical "representative" investor. I would not want to stray too far from this theory, so I stick to cap weighted (helps keep cost low too) funds, I don't buy individual stocks, I widely diversify across all sectors of the global economy, etc.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

User avatar
1210sda
Posts: 1239
Joined: Wed Feb 28, 2007 8:31 am

Re: TSM, Efficient Markets, and tilting

Post by 1210sda » Wed Apr 11, 2012 9:58 am

richard wrote:Why would you believe tilting is better for you than changing your stock/bond mix?

As Gene Fama said, the total market is always on the efficient frontier.


Richard, we are in agreement.

However, Larry S in one of his books has said that increasing small and value (i.e. tilting) results in a more effective form of diversification than just increasing the overall allocation to stocks. That it should reduce "fat tail" risk.

I use the Three Fund portfolio. To tilt to small and value would add more complexity to my portfolio than I (and eventually my surviving spouse :happy ) want or need. Before taking this step, I would have to be convinced that the additional "diversification effectiveness" outweighs the greater complexity. Not sure how to do that.

1210

Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: TSM, Efficient Markets, and tilting

Post by Rodc » Wed Apr 11, 2012 10:17 am

1210sda wrote:
richard wrote:Why would you believe tilting is better for you than changing your stock/bond mix?

As Gene Fama said, the total market is always on the efficient frontier.


snip

I use the Three Fund portfolio. To tilt to small and value would add more complexity to my portfolio than I (and eventually my surviving spouse :happy ) want or need. Before taking this step, I would have to be convinced that the additional "diversification effectiveness" outweighs the greater complexity. Not sure how to do that.

1210


Unfortunately there is no way to really settle the debate. Nowhere near enough data, nowhere near solid theory (on either side IMHO). This is not chemistry or physics. Much more like sociology at best. The key really, I think, is to pick a reasonable low cost widely diversified portfolio that hits the biggies: invest a good slug of money regularly, pick a more or less appropriate allocation between fairly risky assets (stocks, REIT, etc.) and fairly safe assets (nominal and inflation protected bonds, cash), widely diversity at low cost, ignore market noise to stick with your plan. I think we can all (mostly) agree with that and feel pretty confident those are good ideas. Three funds, six funds, not so much. But if you get the biggies either approach will work well enough, I believe.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.

yobria
Posts: 5978
Joined: Mon Feb 19, 2007 11:58 pm
Location: SF CA USA

Re: TSM, Efficient Markets, and tilting

Post by yobria » Wed Apr 11, 2012 11:30 am

Rodc wrote:Unfortunately there is no way to really settle the debate. Nowhere near enough data, nowhere near solid theory (on either side IMHO). This is not chemistry or physics. Much more like sociology at best. The key really, I think, is to pick a reasonable low cost widely diversified portfolio that hits the biggies: invest a good slug of money regularly, pick a more or less appropriate allocation between fairly risky assets (stocks, REIT, etc.) and fairly safe assets (nominal and inflation protected bonds, cash), widely diversity at low cost, ignore market noise to stick with your plan. I think we can all (mostly) agree with that and feel pretty confident those are good ideas. Three funds, six funds, not so much. But if you get the biggies either approach will work well enough, I believe.


Indeed. Everyone, from Fama and French to Bogle, agrees with those wise words. I think the debates here are often over miscommunication between what the experts say, and what people actually undertstand, when it comes to concentrating their portfolios. For example:

They said --> "You can modestly tilt toward small or value". Investor heard --> "Bet it all on red 23!"

They said --> "A historical premium has been found in small, illiquid, high book to market stocks". Investor heard --> "I should buy a bunch of funds with "value" in the title."

User avatar
Clearly_Irrational
Posts: 3087
Joined: Thu Oct 13, 2011 3:43 pm

Re: TSM, Efficient Markets, and tilting

Post by Clearly_Irrational » Wed Apr 11, 2012 2:26 pm

1210sda wrote:Richard, we are in agreement.

However, Larry S in one of his books has said that increasing small and value (i.e. tilting) results in a more effective form of diversification than just increasing the overall allocation to stocks. That it should reduce "fat tail" risk.

I use the Three Fund portfolio. To tilt to small and value would add more complexity to my portfolio than I (and eventually my surviving spouse :happy ) want or need. Before taking this step, I would have to be convinced that the additional "diversification effectiveness" outweighs the greater complexity. Not sure how to do that.

1210


You've already captured the vast majority of the investing benefit and if you're comfortable with what you're doing now there is no reason to change. For me, once I accepted that statistics were valuable for constructing portfolios it naturally led to an examination of the more complex topics like beta, risk factors, skew, kurtosis, etc. I find that my utility function is not completely satisfied with a normal portfolio as I have a preference for lower beta, more positive skew and more negative kurtosis. (hate those fat tails, especially the left one!) Once I reviewed the three factor model it seemed pretty sound and had withstood a number of challenges over the years. By taking advantage of the ability to select MORE risky components I actually need to allocate a lower percentage to overall risk assets. So instead of the bog standard 61/39 portfolio that is recommended for my age, I get a 50/50 portfolio that actually has a higher expected return. Here are some backward looking stats:

Reference 61/39 Portfolio (30.5% TSM, 30.5% Total International, 39% Total Bond)
mean 5.87
standard deviation 12.174
skew -0.474
kurtosis 0.088
maxDD 22.78%

My Portfolio (30% SCV, 20% EM, 30% Long US Treasuries, 20% Gold alternated with cash based on real interest rates)
mean 10.13
standard deviation 10.404
skew -0.264
kurtosis -0.549
maxDD 12.09%

Will it perform exactly that way in the future? No one knows for sure, but I think the odds are in my favor. Should you switch to this portfolio? Heck no, this is an aggressive and complex setup despite the fact that it only uses four funds. Tracking error will be significant. Lets look at a less extreme example though, suppose we switch half the domestic allocation to SCV instead of TSM:

Less extreme example (15.25% SCV, 15.25% TSM, 30.5% Total International, 39% Total Bond)
mean 6.37
standard deviation 11.917
skew -0.546
kurtosis 0.234
maxDD 22.02%

If you only look at the Sharpe ratio, this setup looks like it's a free lunch, but by looking at the higher order moments we've actually increased the "riskyness" due to a longer fatter left tail. Adjust it to the same mean though and things get interesting:

Mean adjusted example (12.65% SCV, 12.65% TSM, 25.30% Total International, 49.40% Total Bond)
mean 5.87
standard deviation 10.376
skew -0.454
kurtosis 0.159
maxdd 17.41%

At this point we've moved from a 61/39 to a 50/50 portfolio. The left tail is slightly shorter, though it is still fatter. Overall our sharpe ratio has risen significantly and we have less beta. Best of all though, the max historical drawdown has declined by 5.37, a significantly less scary event.

To me, the adjusted portfolio looks like a significant improvement and I feel like the odds are good that it will act similarly going forward. Obviously, not everyone agrees with that analysis so you'll just have to look at the data and come to your own conclusions.

User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 8:51 pm

Re: TSM, Efficient Markets, and tilting

Post by bob90245 » Wed Apr 11, 2012 2:42 pm

Rodc wrote:Like MPT, market efficiency leading to the Market Portfolio of all risky assets being optimal has many assumptions about risk, investor behavior, etc etc.

And some people like me scratch their heads and ask why two seperate concepts have been merged in the same sentence. No need to come up with different words. I'll just quote what I wrote upthread.

bob90245 wrote:Market Efficiency means that the longer the time frame the less likely it is for active managers to outperform a benchmark index. Think Bill Miller as a recent example.

On the other hand, portfolio efficiency means that there is some combination of stock asset classes that will fall on the efficient frontier. Of course, we can never know that optimal combination ahead of time. But we can look at history to see how it turned out. Click on image for full size.

Image

I'll just add that I'm reading the latest edition of Burton Malkiel's A Random Walk Down Wallstreet. The chapter I just finished explains Market Efficiency very well.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.

Beagler
Posts: 3442
Joined: Sun Dec 21, 2008 7:39 pm

Re: TSM, Efficient Markets, and tilting

Post by Beagler » Sat Apr 14, 2012 12:40 pm

TSM is a representation of "the stock market," (Wall Street) which is different than "the US economy."

Less than half of all US business profits are generated by publicly traded companies. http://tinyurl.com/6rjuqp3

TSM sports a dwindling number of stocks (VTSMX holds just 3,302 stocks today, down dramatically from just shy of 7,500 in 1997). IMHO, as the number of stocks in TSM shrinks, its claim to diversification is similarly diminished.
“The only place where success come before work is in the dictionary.” Abraham Lincoln. This post does not provide advice for specific individual situations and should not be construed as doing so.

Post Reply