Golden Butterfly Portfolio

[Gabe]

I think when I feel comfortable increasing the equity allocation (i.e. after we see some kind of correction), I'm going to split my assets into two Golden Butterfly portfolios, one in my Roth and one in my 401k/IRA. One will use VTI/VBR for the equity components and the other will use QVAL/QMOM (or something similar, pending further research). It will be fun to watch them side by side.

[Tyler]

So after letting the idea incubate for a while, I'm also sorta digging the same basic Golden Butterfly idea but with total stock market and small cap blend instead of large cap blend and small cap value.  Long story short -- SCV looks really nice historically but I have a tough time justifying it philosophically without pointing to back tests.

My strongest argument for SCV is its lower correlation to TSM than other domestic stock indices, which plays well into the rebalancing strategy. 

My strongest argument for SCB is that it complements TSM really well, almost perfectly balancing your typical 3x3 investment grid between small/medium/large and value/blend/growth (SCB funds have quite a few mid caps as well).  That basically shifts the stock allocation from purely cap-weighted to something resembling equal weight (but simpler with low fees and turnover), which does a nice job of diversifying beyond mega caps without fussing over value or growth biases.

Any thoughts on value vs. blend or market cap weighting vs. equal weighting?

BTW, hat tip to Koekebakker for discussing these issues all the way back on page 1!

[Reub]

Thanks for what you are doing. How does your new and improved Golden Butterfly compare mathematically with your original Golden Butterfly?

[Tyler]

Yes it's "new".  But "improved" is still TBD. 

Value certainly adds punch, but they're in the same ballpark.

Original Version
[img width=400]https://portfoliocharts.files.wordpress ... -20161.jpg[/img]

Alternative Version
[img width=400]http://i65.tinypic.com/f4fk94.jpg[/img]

[Reub]

Now that's what I call service!

So it seems to underperform by about about three tenths of a percent in terms of CAGR and average return. So how is this better again?

[Tyler]

It's slightly worse by the numbers, but perhaps better in terms of a Browne-like philosophy of building something based on the fundamentals rather than trend chasing.  Small cap blend seems like an effective agnostic alternative to small cap value for people who may not fully trust the value premium as a permanent feature of the market. 

That said, it's a relatively small nit and both are good options.  I just enjoy discussing the theory behind a portfolio just as much as the results. It's why I like the PP in the first place. 

[Mr Vacuum]

Any thoughts on value vs. blend or market cap weighting vs. equal weighting?

Yep, when I understood what the Golden Butterfly and MG's equal weight thread were saying, TSM + completion index made sense. I reread all the points on tilting in Craig's blog, tested some things on your web site, and moved to 20% FSTVX/FSTMX, 5% FSEVX/FSEMX.

You can make the backtesting work out pretty close to TSM+SCV if you play with the weights of the stocks, gold, and cash, but I reigned it in and stayed with 4x25.

[Tyler]

Man, the value vs. growth debate really is a rat hole!  Google is an adventure. 

Following up on my previous argument for SCV, I also appreciate the argument by Bill Bernstein that small cap growth is an inherently poor asset class because it (by definition) boots its best performers out of the index when they do really well and out-grow it.  I also can see the macroeconomic argument that in times of struggle for the stock market as a whole people may favor value stocks, which would explain the slightly lower correlations and work well in a PP-style AA.  So I can see how the value tilt for small caps may perhaps be intrinsically beneficial for this particular portfolio.

The Golden Butterfly definitely works.  Maybe SCV will continue to be the better choice over SCB and maybe not, but IMHO neither option is likely to steer you wrong.  As Meb Faber likes to say, maybe the best takeaway is to go enjoy your summer (well, at least for the Aussies this time of year). 

[koekebakker]

One thing that speaks in favor of SCB is that there will be less opportunities for behavorial mistakes.
The problem with SCV investing is that the premium is a moving target. Even if it still exists, you still have to catch it.
Is Vanguards SCV-fund 'valuey' enough to capture the premium? Or do you need the expensive special sauce from DFA or Bridgewater funds? And if your SCV-fund starts tracking a different index, or changes its methodology (which happens quite often) do you switch as well? Or is that performance chasing? And what about all the other new 'factors'? If you've decided there's a value premium, how about momentum, or quality? Where do you stop? And can you hold on to hose funds for 20+ years?

All of this seems pretty far removed from the simplicity of owning total market funds.
It might still be a good idea diversify beyond US megacaps but I would choose funds which are as widely diversified and as passive as possible.

[ochotona]

I came up with this variant of the GB, and I like it:

20% Gold
20% US Stocks
20% Small Cap Blend
20% Intermediate Term Treas
20% Total Bond Market

[Reub]

I came up with this variant of the GB, and I like it:

20% Gold
20% US Stocks
20% Small Cap Blend
20% Intermediate Term Treas
20% Total Bond Market

Numbers please?

[BearBones]

I came up with this variant of the GB, and I like it...

Like it too! More of a bogleheads-tilted GB. I've been playing with the charts and am favoring this for an IRA. Not as SHTF/1970s-repeat protection, but seems pretty safe with higher CAGR.

                  CAGR    Worst Yr      Longest DD
PP              4.8        -11.9            2 (really closer to 3 )
GB              5.8        -8.8              2 (really closer to 2.9 )
Mofified GB  5.5        -11.1            3 (really closer to 3.8 )
Displayed    6.4        -10.2            3 (really closer to 3.7 )

[ochotona]

Yes, a Bogle-type lazy might the bonds all as Total Bond Index (US Aggregate Bond Index). This portfolio makes a concession to safety, and has 1/2 of bonds in Intermediate US Treasuries. It has 20% Gold. Large Cap and Small Cap blend as discussed, rather than Small Cap Value. Bond durations kept shorter... I just hate the interest rate risk; there will be tears someday. I think it would be a pretty robust portfolio for an IRA. CAGR is 1.2% more than the PP, huge over a long period of time.

[img width=590]http://i779.photobucket.com/albums/yy76 ... lsz5g1.png[/img]

MT edit: fixed image width

[Tyler]

I came up with this variant of the GB, and I like it:

20% Gold
20% US Stocks
20% Small Cap Blend
20% Intermediate Term Treas
20% Total Bond Market

I think that's just fine, but total bond market and intermediate treasuries are extremely similar.  So I'm not sure what that's buying you over just holding 40% of one or the other.

[ochotona]

I came up with this variant of the GB, and I like it:

20% Gold
20% US Stocks
20% Small Cap Blend
20% Intermediate Term Treas
20% Total Bond Market

I think that's just fine, but total bond market and intermediate treasuries are extremely similar.  So I'm not sure what that's buying you over just holding 40% of one or the other.

The correlation between IEF and BND is high, but not 1.00. It's actually 0.75. You get a bit of diversification.

[Jpeizie]

Tyler et al,

I've finally decided to stop lurking and post something here! I wanted to pick the board's brain about the GB. Specifically, why do you like small cap? It seems to me that this tilts the portfolio really far in the direction of the stock market in general and toward the idea that the small cap premium will stay intact specifically. I read through the thread so I know you're aware there's a lot of uncertainty about that last point. So why make such a huge bet on it with 20% of the portfolio? In addition, I am really starting to buy into a lot of the demographic worriers saying that Japan is a more extreme preview of what we can expect from the US stock market. All of that makes me really uncomfortable about placing such a relatively large bet on the US market continuing its historic performance going forward.

All that said, I'm considering a portfolio that is basically the GB but instead of 20% small caps, it would use 20% residential REITs. I like this basically because sitting back and collecting rent as a landlord has and probably always will make good money, REIT's are highly volatile so from a PP perspective it seems like they should be a good addition, and residential REITs should be less highly correlated to the market than REITs in general. In other words, I expect residential REITs to weather a recession much better than REITs with heavy business/retail exposure. 

Thoughts?

[Tyler]

Hi JPeizie.  Welcome to the forum!

Good questions.

I like small cap because it provides good diversification away from large cap, not just in correlations but also in company exposure.  Because of cap weighting, the price of large cap funds (or even total market funds) is really driven by a relative handful of mega caps.  Honestly, mid-caps or total international would also be reasonable choices for the same reason.

Beyond the simplified GB version, I also am attracted to the specific combination of VB/VBR with VTI.  They all also include a good amount of mid caps as well, and the resulting combined portfolio has a relatively even spread between small, mid, and large.  Just something to think about.

To be honest, I like small cap value because it works.  The historical performance over blend or growth is pretty striking.  Now opinions differ on why it works and whether it will continue to work, and I've discussed that in the thread a bit.  I don't know the answer.  The most vocal supporter of SCV is probably Larry Swedroe.  So for the pros, google some of his articles.

REITs are also a tempting choice, but I don't like them for myself because between my house and the REITs already included in the stock market index, I'm already plenty exposed to real estate.  I didn't include them in the GB because many others are in the same boat, and for those people a portfolio with REITs may not be as diverse as it seems on the surface.  But everyone's situation is different, and they may also be a decent option for you.

And don't forget -- if you're worried about the US going the direction of Japan, the good ol' Permanent Portfolio is still an excellent option.  I've been very happy with it.  No 5th component or tilting required. 

Hope that helps.

[ochotona]

What about a PP with 25% SCV?

(Took a quick look... I like it)

[Austen Heller]

I first learned about the Fama-French small-cap value advantage back in 2004.  This is how VBR, the small-cap value ETF, has done since then, compared to total stock market (VTI):



I know these dates are cherry-picked, but as long as I've been investing, SCV has not been worth the effort.

The following chart includes dividends:
VISVX (small-cap value) vs VTSAX (total stock market)

[ochotona]

Interesting, over the last 10 years, a PP with small cap blend has had higher CAGR than one with large cap blend or small cap value.

[Dieter]

The SCV premium may or may not show up for a decade or two (or more.)

Probably shows up a few years after most give up on it and large caps are perceived to be more risky....

I do some tilting to SCV, but got enough tracking error vs Boglehead 60/40 that I can't see all my stock in SCV...

And the. There are Mel's unloved Midcaps.... (Another BH reference.)

[Jpeizie]

Hi JPeizie.  Welcome to the forum!

Good questions.

I like small cap because it provides good diversification away from large cap, not just in correlations but also in company exposure.  Because of cap weighting, the price of large cap funds (or even total market funds) is really driven by a relative handful of mega caps.  Honestly, mid-caps or total international would also be reasonable choices for the same reason.

Beyond the simplified GB version, I also am attracted to the specific combination of VB/VBR with VTI.  They all also include a good amount of mid caps as well, and the resulting combined portfolio has a relatively even spread between small, mid, and large.  Just something to think about.

To be honest, I like small cap value because it works.  The historical performance over blend or growth is pretty striking.  Now opinions differ on why it works and whether it will continue to work, and I've discussed that in the thread a bit.  I don't know the answer.  The most vocal supporter of SCV is probably Larry Swedroe.  So for the pros, google some of his articles.

REITs are also a tempting choice, but I don't like them for myself because between my house and the REITs already included in the stock market index, I'm already plenty exposed to real estate.  I didn't include them in the GB because many others are in the same boat, and for those people a portfolio with REITs may not be as diverse as it seems on the surface.  But everyone's situation is different, and they may also be a decent option for you.

And don't forget -- if you're worried about the US going the direction of Japan, the good ol' Permanent Portfolio is still an excellent option.  I've been very happy with it.  No 5th component or tilting required. 

Hope that helps.

Hi Tyler, thanks for your thoughtful response. I checked out Swedroe and it turns out just a few days ago he posted a new paper on this very topic: http://www.etf.com/sections/index-inves ... nopaging=1.

His take makes me really uncomfortable though... It looks to me like he's admitting that the SCV premium is gone, but he chalks it up to statistics and says it'll be back. When I look at it, I see a sector that historically overperformed, but once that overperformance was became widely known, the market corrected and its not coming back. What attracts me to the PP is that there are good solid macro-economic reasons to think that as the big money flows out of one of these assets (eg. a stock market crash) we can expect it to flow into one of the others, so end of day the portfolio is OK. I just don't see SCV working like that in the future. (Though by that logic, I guess my idea about residential REITs doesn't help much either)

I hear your argument about correcting for over-exposure to mega-caps, but that just makes me want to take my existing stock allocation and divide it between LCB and SCB, not double the size of my stock allocation.

End of the day is I'm a real beginner at this stuff and trying to figure out what's going to work best for me as I save for retirement. I know that I need to feel comfortable with the theory behind the portfolio to stop myself from second guessing my allocations if one of the assets underperforms for a decade. That's why I'd be comfortable holding under-performing gold for a decade but totally uncomfortable holding under-performing SCV for a decade. This is what draws me to the PP. But I cant help from wanting to juice the returns!

Re: REIT's, I think your reasons for not liking them make sense. So far though, I don't look at a house as an investment, but rather a depreciating illiquid asset, so I don't include my own house in my AA. And while REITs are part of the overall stock market, I think theres a reasonable argument to make that they're a different breed and perhaps I should include them as a substantial part of my portfolio in their own right. Their size as a percent of the overall market is much too small to make a difference if I only included them as part of the stocks portion of my allocation.

Thanks again for your thoughts. I appreciate the discussion. I've learned an incredible amount from portfoliocharts.com.

[ochotona]

I hear your argument about correcting for over-exposure to mega-caps, but that just makes me want to take my existing stock allocation and divide it between LCB and SCB, not double the size of my stock allocation.

This is what draws me to the PP. But I cant help from wanting to juice the returns!

Have a look at 12.5% mid-cap blend, 12.5% small-cap blend. It outperforms, and has good risk measures.

Well, I'd say if you are in your 20s, 30s, the PP isn't hot enough, so the returns do have to be juiced for young investors, unless they are OK with scaling-down and being more frugal in retirement than they would be otherwise.

[Jpeizie]

Well you guys are certainly giving me something to think about. Don't get me wrong - I LOVE the way those allocations backtest. But it's hard for me to commit to that as a strategy when according to Larrys article the premium only existed for 6 out of the last 21 years and I'm not sure there's a real fundamental economic reason they would outperform consistently going forward. Choices choices...

[Tyler]

Now that I've figured out how to map the results of all possible combinations of assets, the explanation for why the Permanent Portfolio and Golden Butterfly are so appealing is a little more obvious:

[img width=600]http://i65.tinypic.com/125ho9t.jpg[/img]

Both are among the most efficient portfolios available in terms of balancing risk and return.  The GB, specifically, lies right on the efficient frontier as one of the least stress inducing available portfolios for its return class.  Of course that position may change depending on the timeframe chosen, but I think there's something to be learned there about the power of true diversification.