Help Understanding Swedroe's Minimizing Fat Tails Portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
ivesjl
Posts: 53
Joined: Wed Aug 01, 2012 5:54 pm

Help Understanding Swedroe's Minimizing Fat Tails Portfolio

Post by ivesjl »

I am trying to understand Larry Swedroe's investment approach. When I first found the strategy, quite frankly, I did not understand the approach. However, his goal of reducing volatility in a portfolio intrigued me; so I have spent the last week or so reading and here is what I understand:

◦ Bond allocation is typically very high and split between Short Term Investment Grade Bonds and TIPS (35%/35%)
◦ Equity allocation is typically very low and slit between Small Cap Value and Emerging Market Funds. (15%/15%)

Many Bogleheads advocate against this approach, claiming that it is not diversified and over tilted. In response to this, Mr. Swedroe is often quick to point out that this strategy is diversified to all 3 factors (beta, size, and value); furthermore, by selecting index funds, this portfolio will hold many thousands of funds both domestically and internationally.

First of all, please correct me if I am wrong with anything written above. Second, I am curious what others think of this strategy for a young investor like myself. From my perspective, stabilizing returns is very desirable. Down-turns in the market can have devastating effects on a portfolio, especially when you are making yearly contributions. Not to mention the emotional impact of holding a large proportion of equities when the market declines. Can this approach be modified to include a larger proportion of equities?

These examples start with $10,000 in 1985, and invest $15,000 annually. I had to slightly modify the Fat Tails portfolio due to my limited fund data, but I think this is very close and in keeping with the Fat Tails Portfolio. All funds are Vanguard Index funds. By slightly increasing the equity proportions held, the returns increase dramatically and the volatility slightly increases, but still less than holding 80% in the Total Stock Market Fund.

Image

And for reference, here is the comparison to an 80/20 Total Stock Market fund:

Image
"Successful investing is all about common sense." | ~ John Bogle.
wesleymouch
Posts: 237
Joined: Wed Dec 05, 2012 1:24 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by wesleymouch »

You might want to compare to the Permanent Portfolio which has a similar CAGR.
User avatar
Aptenodytes
Posts: 3786
Joined: Tue Feb 08, 2011 7:39 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Aptenodytes »

You won't learn much comparing performance over a specific, limited time period. If you could choose portfolios in this way, you could jack your returns into the triple digits.

I think the main reason that there is skepticism about the approach is that it places a very high bet that the small value and emerging market premiums will persist. Whether they will or not is unknowable, no matter how strong the evidence, which is why most people who believe in these premiums exist nevertheless don't go all-in like this but instead tilt toward them from a base that is much broader.

The logic of the fat-tail portfolio is very sound, but it requires a willingness to really concentrate your equity bet in a way that requires high resolve. The ups and downs will be much more pronounced, and you'll never know for sure that you made the right choice until you cash out of equities and see where you are.

Overlay the NASDAQ and S&P500 graphs from the past 50 years to get a sense of the higher volatility in small caps.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Random Walker »

Ivsjl,
I think you pretty much have it all understood well. The larger dose of bonds minimizes the likelihood of big positive or big negative outcomes. The equity tilt to small and value increases the expected return of the equity component and diversifies the equity component across weakly correlated risk factors. The smoother the ride of the portfolio, the closer the compounded annualized return is to the average annual return. Portfolio volatility is a killer! Smoothing the ride some should pay off. If you lose 50%, you have to gain 100% to get even!
I have an 80/20 portfolio with the equity component strongly tilted. I am depending a lot on diversification across risk factors helping. Nonetheless, I am probably making the mistake of a portfolio too much dominated by market risk.
I think Larry would agree that if you take on a heavy equity allocation you can extend bond duration because overall portfolio volatility is still dominated by equities. I think he wrote a paper on this.
Have you read the short paper by Larry entitled "Effective Diversification in a 3 Factor World"? It's a must read.
I would disagree with you on one point. These errors are SMALLER in early accumulation phase. You can make up for errors with new savings.

Dave
Topic Author
ivesjl
Posts: 53
Joined: Wed Aug 01, 2012 5:54 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by ivesjl »

wesleymouch,
I have seen the Permanent Portfolio, but I do not like the idea of holding Gold or other commodities. It is a personal choice, but I would prefer a portfolio made up of stocks and bonds.

Dave,
I have not read that paper, thanks for the recommendation! Based on your large Equity allocation, it would seem that you are also in the accumulation phase. I am just curious, did you choose to mix in a broader index fund (like the Total Stock Market)?

Apendotyes,
Based on my understanding, the volatility of the small cap market is what drives this sort of portfolio. In order to make up for fewer equities, this portfolio needs a higher volatility. Were you suggesting adding in more Large Caps? I went ahead and overlaid a small cap index fund (NAESX) and a large cap (VFINX) with The Minimize Fat Tails Portfolio shown above, here you go (again I only have data from 1985 onward):

Image
"Successful investing is all about common sense." | ~ John Bogle.
User avatar
baw703916
Posts: 6681
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by baw703916 »

ivesjl,

A few comments:

A backtest of this portfolio beginning in 1985 suffers from the huge limitation that TIPS didn't actually exist then. There have been various attempts to try and reconstruct how TIPS would have performed had they existed, but that approach has some obvious limitations.

I think making the bonds short-term investment grade isn't the best choice. Treasuries have a unique benefit of increasing in value in a financial panic, which saved this type of portfolio in 2008. In fact, going further out on the yield curve (to IT Treasuries, say) would better protect the portfolio against a 2008 event and increase the expected return (at the cost of increasing the inflation risk). Actually, I'm a little confused; in the figure it lists short term Treasury, but in the text it says short term investment grade.

Larry's portfolio benefits from having DFA funds, which provide a lot more size and value exposure than Vanguard's funds. Also, his portfolio is US SV, ISV, and EMV, compared to the Vanguard funds which are only US SV (with less tilt) and EM (with no tilt), w/o any Intl. developed markets. Probably there's not quite enough equity risk exposure to keep up with the 80% equities portfolio in the long run.

I think the best way to replicate the equity portion of Larry's portfolio is one doesn't have DFA access is to use ETFs such as RZV, DLS, and DGS for US, Intl, and EM respectively. REITs may also be good to include as an inflation hedge.

Brad
Most of my posts assume no behavioral errors.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Random Walker »

Lvesjl,
Yes I basically start with a whole markets approach. If one is to have a tilted portfolio with exposure to the risk factors of market, small, value, the most efficient way to achieve the tilt is to use whole market funds as the cornerstone of the portfolio and then use relatively small doses of the tilting asset classes to tilt to taste. This is because the whole market funds generally have lower ERs, are more tax efficient, and rebalance internally rather than between funds. For me, my whole market funds are the Tax advantaged US Core, World Ex US Core, and EM Core funds of DFA. These core funds are whole market funds but already have a tilt to small and value. Specific asset class funds tilt beyond these.
By the way, I think Effective Diversification in a 3 Factor World are 3 of the most important pages I've read in investing. Should be pretty easy to find on the Internet.

Dave
EDN
Posts: 528
Joined: Sat Dec 29, 2012 6:55 pm
Location: Oklahoma City, OK

A Few Comments on the Fat Tail and Permanent Portfolios

Post by EDN »

1. As a young investor, I'm not sure the pursuit of low volatility is necessarily ideal. It could be entirely beneficial to have bigger ups and downs for a given return if it allows you to put more money to work at temporarily depressed prices when dollar-cost-averaging and therefore earn the higher-expected returns.

2. I'm also not sure that someone who cannot stomach market volatility will do all that well with extreme amounts of market tracking error (a portfolio deviating wildly from the market return). For a young investor, I believe these are different sides of the same "behavioral coin" -- neither should be that problematic, and the sense that one will be is likely to lead to the other being an issue as well (but no assurances here).

3. You are probably better off having a portfolio like Dave (Random Walker) -- diversifying broadly across market/size/value on the equity side with a sizable equity component. Why? You can afford to, number one. And number two, I'm not sure the "high tilt, low equity" allocations will produce the same forward-looking efficiencies they once have, as the vast majority of the allocation resides in an asset class (bonds) with artificially low expected returns. Also, at really low equity levels, a balance of large/small and growth/value doesn't produce materially different returns than 100% SV -- there just isn't enough in stocks to move the dial (see the #s below).

4. The Permanent Portfolio has not had similar CAGR as tilted/balanced portfolios. From 1975-2011 ('75 is the first year to legally own Gold), a 33/67 portfolio (6.5% S&P 500, 10% US large value, 16.5% mid/small value, 67% 5YR T-Notes) matched the PP annual volatility with 2% per year higher returns. A 25/75 portfolio (25% mid/small value, 75% 5YR T-Notes) earned the same return as the 33/67 with 0.4 less annual SD. The comparison between diversified asset class portfolios and the PP are not even close when you adjust for risk (instead of arbitrarily comparing PP, with 25% in stocks, to generic 60/40 balanced portfolios).

Eric
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

First, never invest in anything you don't fully understand because when it underperforms you will be prone to confusing strategy and outcome and abandon the strategy

Second, I have backtested the strategy to 1970 and it has held up remarkably well in all environments.

Third, IMO it is the most diversified of portfolios when thought of in the new way of thinking about diversification, across risk factors, or sources of returns. Market like portfolio are undiversified in this way, having only exposure to beta, not size or value. And it certainly is globally diversified and has thousands of stocks so no idiosyncratic risks

fourth, it clearly cuts the bad left tail as high quality bonds tend to do well when stocks get hit badly in financial crisis, and that is only kind of bonds I recommend

fifth, the belief in small value premium is no different than belief in equity premium. In fact the value premium is more persistent.

I hope that is helpful

Larry
Topic Author
ivesjl
Posts: 53
Joined: Wed Aug 01, 2012 5:54 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by ivesjl »

baw703916 wrote: Actually, I'm a little confused; in the figure it lists short term Treasury, but in the text it says short term investment grade.
Brad
Whoops, clicked the wrong fund there. Thanks for pointing that out:

Image
"Successful investing is all about common sense." | ~ John Bogle.
Topic Author
ivesjl
Posts: 53
Joined: Wed Aug 01, 2012 5:54 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by ivesjl »

larryswedroe wrote:First, never invest in anything you don't fully understand...
Larry
I completely agree with your first point, and don't worry, I will not make any changes in the immediate future. My major goal with this post is to collect as much information as possible, and not make an impulsive decision. Ultimately, I would like to have a straightforward and logical portfolio that doesn't require lots of tinkering. It has been my experience that rushing into large purchases is a very bad idea.
"Successful investing is all about common sense." | ~ John Bogle.
gc527
Posts: 16
Joined: Sat Feb 04, 2012 10:35 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by gc527 »

I have been very interested in this portfolio as it makes a lot of sense to me. However, I may be misunderstanding the 3-factor model. If you are focused primarily on value and small risks premiums, are you adequately covered for beta (market) risk? Or does the small/value provide a proxy for the beta? If not, then wouldn't it make more sense to have a wide but tilted portfolio?
User avatar
baw703916
Posts: 6681
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by baw703916 »

The idea is that any net ownership of equities automatically includes beta. In the academic literature, the value premium is usually defined by HmL--High (book value to market price ratio) minus Low (book to market). In other words, the net return you would get from buying the X% of equities with the highest book/market (lowest price/book) and shorting the X% with the lowest book/market. So by doing this, you would have no net equity exposure, and your return would purely reflect the difference in returns between value and growth stocks.

Obviously, this is more appropriate for a theoretical discussion of the value premium than as an investment approach in real life.

Similarly, the size premium is typically denoted by SmB (small company return minus big company return).
Most of my posts assume no behavioral errors.
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Ketawa »

gc527 wrote:I have been very interested in this portfolio as it makes a lot of sense to me. However, I may be misunderstanding the 3-factor model. If you are focused primarily on value and small risks premiums, are you adequately covered for beta (market) risk? Or does the small/value provide a proxy for the beta? If not, then wouldn't it make more sense to have a wide but tilted portfolio?
Any broadly diversified stock fund/ETF has a beta loading around 1.0. An 80/20 portfolio has an overall load of about 0.8 beta, regardless of whether the portfolio targets small value.

I am diversified across equity/size/value factors like the Swedroe portfolio, but I have different holdings for my fixed income. My fixed income is currently the G Fund, some promotional 1-year CDs at 3% and 4%, and a Mango Savings Account at 4%. With these holdings it is not possible for me to benefit from a "flight to quality", but I consider the advantages of the G Fund to be superior.

I also tilt heavily to SV in my domestic holdings, and less so in international. My domestic loadings are around 0.4-0.5 for both small and value. However, my first priority in my equities is to hold approximate market weights across regions, so I do not overweight emerging markets. I end up breaking up my equities into:

1) Domestic (S Fund in TSP, and VIOV/VBR)
2) Intl Developed (I Fund in TSP, VSS, may add EFV but it is low on the list of priorities)
3) Intl Emerging (VWO/IEMG, VSS)

VSS is convenient since it auto-rebalances across intl developed and emerging.

I'm 25 and have decades of investing to go, so I'm not too worried about reducing volatility.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by grap0013 »

I think Larry just uses 30% equity because he no longer has need to take risk. The is no reason you cannot employ a similar strategy with a higher equity allocation.

Larry,

I've always wanted to ask you this: If you were a young accumulator and had the willingness, the need, and the ability to take risk, would you go 100% global SCV and add bonds as circumstances change?

Thanks!

grap
There are no guarantees, only probabilities.
User avatar
fundtalker123
Posts: 901
Joined: Tue Feb 27, 2007 3:18 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by fundtalker123 »

larryswedroe wrote:First, never invest in anything you don't fully understand because when it underperforms you will be prone to confusing strategy and outcome and abandon the strategy

Second, I have backtested the strategy to 1970 and it has held up remarkably well in all environments.

Third, IMO it is the most diversified of portfolios when thought of in the new way of thinking about diversification, across risk factors, or sources of returns. Market like portfolio are undiversified in this way, having only exposure to beta, not size or value. And it certainly is globally diversified and has thousands of stocks so no idiosyncratic risks

fourth, it clearly cuts the bad left tail as high quality bonds tend to do well when stocks get hit badly in financial crisis, and that is only kind of bonds I recommend

fifth, the belief in small value premium is no different than belief in equity premium. In fact the value premium is more persistent.

I hope that is helpful

Larry
You're using present tense but you should use past tense. Does your back testing include any 20 year periods where SV and/or EM underperformed relative to TSM? If not perhaps you have not tested a suffiently large number of independent 20 year periods. Out of a hundred 20 year periods you would expect SV and/or EM to underperform in some fraction of those of 20 year periods, right? Otherwise you're saying you can safely assume that SV and/or EM will never underperform TSM in a 20 year period, and how can you know that?
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by FillorKill »

Random Walker wrote:Ivsjl,
Have you read the short paper by Larry entitled "Effective Diversification in a 3 Factor World"? It's a must read.
Dave
ivesjl wrote: Dave,
I have not read that paper
Here it is; no time like the present:

http://www.bamwebsites.com/581292.pdf
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

grap
there are many young investors who work at my firm as advisors and have the high tilt with higher equity allocations. That just uses the strategy to get higher expected returns, while I use it to keep returns at same expected level while cutting the tails. You can use a high tilt either way.


Fundtalker
First, the most recent period is one where value has done poorly.
Second, if you think about it, the question is basically irrelevant. The reason is simple. If size and value premiums turn negative they can do so in two types environments.
a) periods when economic risks show up like Great Depression or 2008. If that happens then your much lower exposure to beta plus your much higher exposure to high quality fixed income will almost certainly offset the greater loss in your much smaller equity portfolio. This is exactly what happened in 2008.
b) periods like late 90s, when beta is way up but size and value are negative. You don't care, that is the trade off you made. You gave up the opportunity to have the big right tail to get rid of or reduce the left tail risk.

It's hard to imagine a scenario where beta is up big but size and value stocks actually have big negative returns (not negative premiums), which is how you could get hurt.

I hope that is helpful

Larry
User avatar
norookie
Posts: 3016
Joined: Tue Jul 07, 2009 1:55 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by norookie »

Interesting, :happy Thanks!
" Wealth usually leads to excess " Cicero 55 b.c
EDN
Posts: 528
Joined: Sat Dec 29, 2012 6:55 pm
Location: Oklahoma City, OK

More on Fat Tail vs. Broadly Balanced

Post by EDN »

There are a few scenarios I can think of where you would be much happier with a moderate tilt to small and value and equities vs. fixed income relative to an all-small value equity allocation with a heavy bond orientation.

First, lets define the portfolios. Since 1928, the following 3 allocations have had the same annualized return of +9.6%:

(a) 100% S&P 500
(b) 40% Small Value, 60% 5YR T-Notes rebalanced annually
(c) 16.5% S&P 500, 16.5% US large value, 22% US small value, 45% 5YR T-Notes rebalanced annually

(a) had an annual SD of 20.2, (b) had an annual SD of 13.3, and (c) had an annual SD of 14.0...so both (b) and (c) captured "market returns" with less volatility.

But, there have been long periods where:

#1 -- bonds have performed very poorly on a real basis. For almost 25 years starting in 1941 (thru '65), 5YR bonds had a -1.1% real return. Over that stretch, we saw the following returns:
(a) +14.5%
(b) +10.0%
(c) +10.9%

#2 -- bonds have performed very poorly relative to stocks. For the 16 years beginning in 1984, the S&P 500 outperformed 5YR bonds by almost 10% per year. Over that stretch we saw the following returns:
(a) +18.1%
(b) +11.4%
(c) +13.4%

#3 -- small value stocks do very poorly, and much worse than other stocks and especially bonds. From 1969-1973, we saw low but positive returns on the S&P 500 and large value stocks (+2% and +3% per year) and good bond returns (+6.8% for 5YR T-Notes), yet SV stocks got crushed (-9.4% per year). Over that stretch we saw the following returns:
(a) +2.0%
(b) +0.7%
(c) +2.1%

So give or take, these 3 periods cover the better part of 60 years (out of 85) where you'd be in a stretch where a more balanced portfolio (55/45) would be preferable to an extreme SV allocation (40/60). But what about the down years? Specifically the "big 3": '29-'32, '73-'74, and '08? Well clearly, during financial collapses, any portfolio that holds more in bonds will be better off than one that is more heavily invested in stocks. But the differences in returns aren't that substantial:

1929-1932 Total Returns
(a) -64.2%
(b) -39.9%
(c) -42.2%

1973-1974 Total Returns
(a) -37.2%
(b) -12.7%
(c) -16.1%

2008 Total Returns
(a) -37.0%
(b) -8.9%
(c) -17.0%

So other than 2008, the difference between (b) and (c) during the worst meltdowns was about 3%, not enough to make a difference one way or another. Over all negative periods, (b) and (c) had the same # of down years and lost almost the same amount (b lost 1% less than c on average).

Ultimately, it is up to each investor who they want their portfolio to behave. Personally, and when I council clients, I (and they) are happy to trade a bit lower returns or a bit more downside risk to own what I believe to be a more balanced portfolio that includes modest amounts of stocks and bonds as well as large/small & growth/value in the equity space. The data above seems to indicate that more balance provides a more consistently acceptable ride during many different environments. But if you are willing to endure long periods where you'd be better off with more balance to see a few % points less downside during one or two brutal bear markets in your lifetime, then maybe you are better off with the more tilted allocation.

Eric
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by FillorKill »

Eric,

Not to take away from your post...

I mention this distinction as the title of the thread refers to Larry's suggestions regarding a tail reduction/minimization portfolio. Your analysis is based on only US SV as the equity component. That isn't what Larry uses/recommends in the spirit of tail reduction:
by larryswedroe » Fri Jan 16, 2009 11:14 pm

For the record
My portfolio is TIPS (no RE due to limited space in tax advantaged accounts) and intermediate munis (about 5 years duration) for the fixed income and then about 50% US SV, 35% ISV and 15% EMV
His recommendation is a different animal - more diversified.
Last edited by FillorKill on Tue Feb 12, 2013 5:46 am, edited 1 time in total.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

eric
IMO your post only makes my case.
You don't care about underperformance when returns are positive, what you are trading off is the reduction in the left tail risk, and it showed up when needed most, in the worst bear market in the post war era.

Also remember for those in withdrawal phase, cutting that tail risk is especially valuable, reducing SD helps portfolio survival rates

Finally, once you start to tilt portfolios to the kind of levels you have in Portfolio C the tracking error risk you take on to move to higher tilt is relatively minor relative to the tracking error risk you take when moving from market like portfolio to tilted portfolios. Thus, if you are going to lose discipline with B, almost certainly you'd lose it with C anyway. And tracking error, though real issue, is a purely psychological problem faced by people who are prone to confusing strategy with outcomes. If that is problem you should not tilt anyway, at least IMO,

Best wishes
Larry
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: More on Fat Tail vs. Broadly Balanced

Post by grap0013 »

EDN wrote:
1929-1932 Total Returns
(a) -64.2%
(b) -39.9%
(c) -42.2%

1973-1974 Total Returns
(a) -37.2%
(b) -12.7%
(c) -16.1%

2008 Total Returns
(a) -37.0%
(b) -8.9%
(c) -17.0%


Eric
Speaking of strictly the equity portion, you could even throw a bunch more date ranges in there to make your case even stronger. 1966-1982 and 2000-2012. Heck, even SCV Japan stocks have done much better than their total Japan market counterparts 1989-2012. I have yet to find a time frame where large blend saved your bacon when SCV had flat real returns. I can find multiple periods where you are very happy you have the tilts.
There are no guarantees, only probabilities.
EDN
Posts: 528
Joined: Sat Dec 29, 2012 6:55 pm
Location: Oklahoma City, OK

Globalizing the Fat Tail and Balanced Portfolios

Post by EDN »

BBL wrote:Eric,

Not to take away from your post (and thank you by the way for that and many of your others - very useful).

I mention this distinction as the title of the thread refers to Larry's suggestions regarding a tail reduction/minimization portfolio. Your analysis is based on only US SV as the equity component. That isn't what Larry uses/recommends in the spirit of tail reduction:
by larryswedroe » Fri Jan 16, 2009 11:14 pm

For the record
My portfolio is TIPS (no RE due to limited space in tax advantaged accounts) and intermediate munis (about 5 years duration) for the fixed income and then about 50% US SV, 35% ISV and 15% EMV
His recommendation is a different animal - more diversified.
BBL,

You are correct, but in order to observe the historical behavior of different approaches over the longest possible periods, we are confined to US-only asset classes. If I were to globalize the equity portion of portfolio (c) above, we'd go something like 20% S&P 500, 20% US large value, 30% US small value, and 10% each Int'l large value, Int'l small value, and EM value (70/30 also helps to address some of the psychological issues with being too heavily invested in non-US stocks during a strong domestic market). I don't think we'd come to different conclusions that the longer, US-only data led us to above: for most, a more balanced portfolio is a better approach -- those looking only to preserve 3% or so more of portfolio principal during the most severe market sell-offs at the expense of the other issues I raise will find the more titled portfolio a better option.

Eric
EDN
Posts: 528
Joined: Sat Dec 29, 2012 6:55 pm
Location: Oklahoma City, OK

All Portfolio Behavioral Biases Matter

Post by EDN »

larryswedroe wrote:eric
IMO your post only makes my case.
You don't care about underperformance when returns are positive, what you are trading off is the reduction in the left tail risk, and it showed up when needed most, in the worst bear market in the post war era.

Also remember for those in withdrawal phase, cutting that tail risk is especially valuable, reducing SD helps portfolio survival rates

Finally, once you start to tilt portfolios to the kind of levels you have in Portfolio C the tracking error risk you take on to move to higher tilt is relatively minor relative to the tracking error risk you take when moving from market like portfolio to tilted portfolios. Thus, if you are going to lose discipline with B, almost certainly you'd lose it with C anyway. And tracking error, though real issue, is a purely psychological problem faced by people who are prone to confusing strategy with outcomes. If that is problem you should not tilt anyway, at least IMO,

Best wishes
Larry
Larry,

The issue is you are picking and choosing which psychological biases matter and which do not. You say "missing out on some upside gains" doesn't matter, nor does having the most extreme equity portfolio possible(and associated tracking error) , but having 2% to 5% less in losses during 3 severe bear markets in 85 years makes all the difference. The reality is, both mixes (b) and (c) had the same returns and SDs within 0.5 of one another -- not a big enough difference to have any impact on retirement outcomes. I'd actually give the nod to the 55/45 balanced portfolio when taxes are considered (lower after-tax returns on portfolios with greater amounts of taxable debt or even lower-returning tax-free bonds) and when we acknowledge that the periods I highlighted above (bad real or relative bond returns) are probably more likely than not going forward.

Finally, as it relates to behavioral biases, you cannot say on one hand "once you've tilted somewhat --- all additional tilts and added tracking error are the same", yet when your portfolio declines, 2% to 5% differences in loss (in the midst of 30%+ declines) makes a huge difference. We are talking about the same percentages, just applied to different observations. If cutting out a few hundred basis points of negative returns every few decades is important, reducing portfolio tracking error relative to the market for a give return should also be worthwhile.

Everything in moderation.

Eric
User avatar
Hub
Posts: 480
Joined: Fri Jul 13, 2012 8:56 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Hub »

Just wanted to thank everyone for this very thoughtful thread. I commonly have the conversation in my head about going "all in" on small and value, though it does seem to make more sense as a plan once you're in a position to need significantly higher percentages of fixed income than this young accumulator currently possesses. With 30 more years on the table before retirement, I think my left tail risk is nonexistent at the exact moment, or at least it is in any way that I can avoid it.
bikenfool
Posts: 99
Joined: Fri Jul 22, 2011 9:05 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by bikenfool »

Perhaps it's too radical for some bogleheads, but I think the Swedroe Short Fat Tail porfolio should be in the wiki.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Browser »

If you want a "risk parity" portfolio - on that matches the impact of asset volatility (risk) on overall portfolio volatility - you generally need about a 2:1 ratio of bonds vs. stocks unless you leverage the bond component. If the equity component is more volatile than TSM (as is SCV), then the ratio moves closer to 3:1. Isn't the Swedroe portfolio just one representative of a risk parity based allocation strategy?
We don't know where we are, or where we're going -- but we're making good time.
bogle2013
Posts: 87
Joined: Tue Jan 22, 2013 2:56 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by bogle2013 »

Is it possible to use this portfolio if most of the money is in a taxable account with little tax-advantaged $?
vesalius
Posts: 794
Joined: Tue Jul 13, 2010 7:00 pm
Location: Texas

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by vesalius »

bogle2013 wrote:Is it possible to use this portfolio if most of the money is in a taxable account with little tax-advantaged $?
From reading prior Swedroe post this is his situation as well. He will likely chime in, but from what I recall he utilizes high quality, AAA/AA, municipal bonds in his taxable accounts. Larry uses individual bonds, of course. I utilize BMBIX (Baird Intermediate Muni Institutional) in taxable for this purpose because it has a significantly higher % of AAA/AA bonds than any of vanguards low to intermediate duration bond funds.
gc527
Posts: 16
Joined: Sat Feb 04, 2012 10:35 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by gc527 »

@baw703916, that explains it rather well – even if theoretical! Thanks.

@Ketawa, I am impressed that you have a plan and executing on the plan at this young age. Good luck!

I have a few more portfolio construction questions for Larry and those who understand the Larry’s fat tail portfolio:
  • I have access to only Vanguard funds and obviously Vanguard funds don’t have pure asset classes such as DFA funds. Is that acceptable? Or does this break the fundamental assumption behind such a portfolio construction?
  • Do REITs have a place in the portfolio?
  • For the Bonds section, I assume TIPS would give inflation protection. But for the other bond holding should the focus be on intermediate-term or long-term treasuries?
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by grap0013 »

gc527 wrote:@baw703916, that explains it rather well – even if theoretical! Thanks.

@Ketawa, I am impressed that you have a plan and executing on the plan at this young age. Good luck!

I have a few more portfolio construction questions for Larry and those who understand the Larry’s fat tail portfolio:
  • I have access to only Vanguard funds and obviously Vanguard funds don’t have pure asset classes such as DFA funds. Is that acceptable? Or does this break the fundamental assumption behind such a portfolio construction?
  • Do REITs have a place in the portfolio?
  • For the Bonds section, I assume TIPS would give inflation protection. But for the other bond holding should the focus be on intermediate-term or long-term treasuries?
Why only Vanguard? Just use their brokerage and you can buy ETFs with better tilts for relatively low fees (ie. up to 25 trades per year at $7/trade)

Probably don't need REITS as SCV funds typically already have around 10% REITS in them.

SCV has performed great during inflationary times. These companies have lots of debt which is great during inflation. The REITS help too. Plus the overall higher real returns vs. other asset classes help as well. For higher equity allocations you don't need much inflation protection from bonds, but at 30% equity you probably need a little bit more. The data shows that you should only hold long treasuries with this strategy at 70%+ equities otherwise you should go short or intermed term. See link for one of my favorite threads of all time: http://www.bogleheads.org/forum/viewtopic.php?t=47240

Full disclosure:

I'm 100% equity with targets of:

50% PXSV
25% PDN/SFILX
25% DGS/DFEVX

grap
There are no guarantees, only probabilities.
User avatar
LadyGeek
Site Admin
Posts: 95696
Joined: Sat Dec 20, 2008 4:34 pm
Location: Philadelphia
Contact:

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by LadyGeek »

bradrh wrote:Perhaps it's too radical for some bogleheads, but I think the Swedroe Short Fat Tail porfolio should be in the wiki.
Caution, there are a LOT of portfolios that "work." The intent of the wiki is to first teach the basics of investing in the entire market, which is the Bogleheads investment philosophy. If an investor wants to deviate from the total market approach, which is called tilting, they should be prepared to understand the risks involved with this approach.

The portfolios in this thread intentionally deviate from the total market approach, and therefore have additional risks which must be understood by the investor. (Note one of Larry's post was "never invest in anything you don't understand" - which is also one of the Bogleheads' tenets.)

Portfolios which deviate from the total market approach are sometimes called: Slice and Dice

For new investors, consider the portfolios here: Lazy Portfolios. This thread may not be what you are looking for. I would not recommend changing anything based on the discussion in this thread. The fact that a lot of members are interested in this portfolio is simply because this is an investing forum and discussing portfolios is what we do.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
User avatar
tarnation
Posts: 2544
Joined: Thu Apr 26, 2007 12:36 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by tarnation »

bradrh wrote:Perhaps it's too radical for some bogleheads, but I think the Swedroe Short Fat Tail porfolio should be in the wiki.
I noticed several variants of the name in this thread. We need a standardized name for it. Maybe "No Fat Tails" portfolio?
Image
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

gc527

Hope this helps

The smaller the market cap and the higher the BtM or lower the P/E the higher the expected return of the fund. The higher the expected return the less beta exposure you need to get the same expected return, and that cuts tail risks. And of course the reverse is true.So all else equal you want funds with the lowest market cap and highest BtM. Which is why we now use the Bridgeway Omni Funds for US SV.

Best wishes
Larry
User avatar
Random Musings
Posts: 6771
Joined: Thu Feb 22, 2007 3:24 pm
Location: Pennsylvania

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Random Musings »

larryswedroe wrote:gc527

Hope this helps

The smaller the market cap and the higher the BtM or lower the P/E the higher the expected return of the fund. The higher the expected return the less beta exposure you need to get the same expected return, and that cuts tail risks. And of course the reverse is true.So all else equal you want funds with the lowest market cap and highest BtM. Which is why we now use the Bridgeway Omni Funds for US SV.

Best wishes
Larry
Larry, it looks like Bridgeway has the characteristics you like, including the fact that it holds no (nominal) real estate and utility companies. However, compared to the benchmark they use (Russell 2000, for whatever reason :oops: ), the sector holdings Omni SCV uses is done in a non-proportional manner relative to the benchmark they chose, the consumer discretionary group is highly weighted relative to other holdings. What is the rationale Bridgeway uses for chosing what (roughly) 600 stocks they are holding in that fund?

RM
I figure the odds be fifty-fifty I just might have something to say. FZ
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by grap0013 »

Random Musings wrote:Larry, it looks like Bridgeway has the characteristics you like, including the fact that it holds no (nominal) real estate and utility companies. However, compared to the benchmark they use (Russell 2000, for whatever reason :oops: ), the sector holdings Omni SCV uses is done in a non-proportional manner relative to the benchmark they chose, the consumer discretionary group is highly weighted relative to other holdings. What is the rationale Bridgeway uses for chosing what (roughly) 600 stocks they are holding in that fund?

RM
My money is on PXSV for SCV. Literally. It is kinda fun to look under the hood at the Omni SCV (BOSVX) at morningstar. It is REALLY small and valuey. Wow.
There are no guarantees, only probabilities.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

Bridgeway uses the same type screens as DFA--numeric. So say bottom 20% by P/E.
Larry
User avatar
Bustoff
Posts: 2033
Joined: Sat Mar 03, 2012 5:45 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Bustoff »

Larry - does the small tilt strategy work for those of us that are retired with more conservative portfolios, say 30/70 ?
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

Bustoff
It works for all investors, regardless of AA but especially helpful for those in withdrawal phase becuase that is when downside volatility risk is greatest, and the high tilt allows one to hold less beta risk, which is the greater risk, and more safe fixed income, which tends to do well in financial crises, and that cuts the downside risk, the left tail

So if you would normally hold a 30/70 portfolio with a TSM type equity allocation, you can get the same expected return with a lower equity allocation if you tilt, and the more you tilt the less beta risk you need.

I would add one more thing, the value tilt creates a portfolio that tends to do bit better in inflation times as value companies tend to be more leveraged, and retirees are typically more exposed to inflation risk.

Best wishes
Larry
User avatar
Bustoff
Posts: 2033
Joined: Sat Mar 03, 2012 5:45 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Bustoff »

larryswedroe wrote:Bustoff
It works for all investors, regardless of AA but especially helpful for those in withdrawal phase becuase that is when downside volatility risk is greatest, and the high tilt allows one to hold less beta risk, which is the greater risk, and more safe fixed income, which tends to do well in financial crises, and that cuts the downside risk, the left tail

So if you would normally hold a 30/70 portfolio with a TSM type equity allocation, you can get the same expected return with a lower equity allocation if you tilt, and the more you tilt the less beta risk you need.

I would add one more thing, the value tilt creates a portfolio that tends to do bit better in inflation times as value companies tend to be more leveraged, and retirees are typically more exposed to inflation risk.

Best wishes
Larry
Thanks Larry - so would a typical 30/70 tilt be accomplished by reducing TSM to 20% and adding 10% Small/Value to make up the 30% ?
Clive
Posts: 1950
Joined: Sat Jun 13, 2009 5:49 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Clive »

larryswedroe wrote:value tilt creates a portfolio that tends to do bit better in inflation times as value companies tend to be more leveraged, and retirees are typically more exposed to inflation risk.
Hi Larry.

Kenneth French's small, high book to price yearly data for Japan indicates that SCV performed well since 1991 whilst the Nikkei 225 performed poorly, and inflation was low (zigzagging either side of +/- 0%).

Might it not be that both small and value are higher risk. Either more likely to crash to zero or jump significantly in value - and its more of a coin flip as to when a portfolio of both/either small or value is likely to spike upwards or downwards - independently of what the larger market might be doing.

If (pulling figures out of thin air) TSM averaged 6% yearly real gains with a 20% standard deviation - that approximates to a 4.1% annualised. If SCV yearly averaged 7.5% yearly real gains with a 27% standard deviation (more 'risk' (volatility), more reward), that also approximates to a 4.1% annualised. Hold 25% less SCV with the remainder invested in inflation bonds and you'd have a similar yearly average and standard deviation to that of TSM (similar gain with less 'at-risk' exposure), supplemented with any additional benefit if the inflation bonds perhaps earned >0% real.
gc527
Posts: 16
Joined: Sat Feb 04, 2012 10:35 am

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by gc527 »

larryswedroe wrote:gc527

Hope this helps

The smaller the market cap and the higher the BtM or lower the P/E the higher the expected return of the fund. The higher the expected return the less beta exposure you need to get the same expected return, and that cuts tail risks. And of course the reverse is true.So all else equal you want funds with the lowest market cap and highest BtM. Which is why we now use the Bridgeway Omni Funds for US SV.

Best wishes
Larry
Thanks Larry.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

Bustoff

Unfortunately it will depend on type of fund you use to implement. But I think that sounds reasonable

Clive
First, you want to include international small value and EM value as well. Second you also have to consider that in the left tail events where the risks of small value are likely to show up are exactly when the high quality bonds are likely to do very well. So you would own more of the highly performing asset class. That is exactly what happened in 08.
Low beta and high tilt portfolios with high quality FI far outperformed TSM portfolios with higher beta as much higher beta is needed. so while your equity allocation did do worse, you owned much less and your high quality bonds did real well and you owned much more

Best wishes
Larry
Jebediah
Posts: 805
Joined: Tue Aug 28, 2012 9:19 pm
Location: Austin TX

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Jebediah »

Larry, what do you mean when you say high quality bonds "did very well" in 08? As far as I can tell, short term Treasury and IG had the tiniest of upticks in NAV. You just mean, compared to losing half your money in stocks?
User avatar
baw703916
Posts: 6681
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by baw703916 »

Jebediah wrote:Larry, what do you mean when you say high quality bonds "did very well" in 08? As far as I can tell, short term Treasury and IG had the tiniest of upticks in NAV. You just mean, compared to losing half your money in stocks?
Short term Treasury has such a short duration that it isn't going to move much no matter what happens. But if you look at something with more volatility (but still no credit risk) such as EDV (zero-coupon Treasuries ETF)...it compensated for the 50% drop in equities in 2008, and then some.
Most of my posts assume no behavioral errors.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

jebediah
Obviously even if you own one month Treasury bills they certainly did relatively well compared to stocks and risky bonds. The longer the maturity of the high quality bonds the better they did as well. So how well they did depended on your own duration of the portfolio.
Best wishes
Larry
jmk
Posts: 643
Joined: Tue Nov 01, 2011 7:48 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by jmk »

larryswedroe wrote:I have backtested the strategy to 1970 and it has held up remarkably well in all environments
Larry, in your back testing what did you use for "synthetic tips" before 1999?
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by larryswedroe »

jkandell
Did not use TIPS in the testing of the high tilt portfolios. Either 1 or 5 year Treasuries. Normally we build roughly 10 year ladders so a 5-year would be good proxy, but I did not want to be accused of data mining in period when bond yields fell so I often showed 1 year Treasuries, penalizing myself if you will

Larry
Malkielino
Posts: 43
Joined: Sun Jan 23, 2011 4:57 pm

Re: Help Understanding Swedroe's Minimizing Fat Tails Portfo

Post by Malkielino »

When using the website:

http://www.portfoliovisualizer.com

I inadvertently rediscovered the fat tails portfolio.

The route was not the same as mentioned here, but in some ways works out the same.

I ran the MC simulations and studied the `Safe Withdrawal Rate' (inflation included)
for a 35-year withdrawal period (suppose I live from 65 to 100 years old). I study
the maximum withdrawal rate that will give a chance of not depleting the funds
of >99%.

Initially 15-15-70 EM-SCV-Total Bond did the best. Then a bunch of googling made me
realize Larry Swedroe got there first. Then it turned out that his fat tails portfolio was
already programmed into the Backtest option on the Portfolio Visualizer! I found it
via a completely roundabout journey.

Replacing Total Bond with 2 year treasuries
and TIPS does a bit better (measured by the Safe Withdrawal Rate)
than I did, Larry Swedroe nailed that.

Replacing SCV with Mid Cap Value actually does even a little better.

Not sure if I can trust that MC simulation out in the tails. But in any case, my
inadvertent recreation of Larry Swedroe's fat tails makes me think it is probably
a very interesting choice.
Post Reply