Golden Butterfly Portfolio

[Tyler]

I mentioned in another thread that I've been tinkering with my spreadsheets and have been studying an interesting PP variant.  Short story -- by adding just one new asset, it has the same long-term real CAGR as the total stock market but with the low volatility of the PP. 

http://portfoliocharts.com/2015/09/22/c ... butterfly/

Quick disclaimer -- I am currently 100% PP and this is just an idea I came across when looking for alternatives for cash in my IRA.  I'm tempted to move some money into SCV by the end of the year, and since the rest of the portfolio is basically a pure PP I consider it a VP idea.  But I'm not committed to anything.

Thoughts?

[koekebakker]

Well, any backtested portfolio will look a lot better if you add some SCV...

I'm pretty sceptical about the value premium going forward. So if I would add a fifth part to the vanilla PP it would be a broad international stocks fund or etf.

The risk adjusted return of a 40/20/20/20 portfolio has been really good over the last 40 years or so!

[dualstow]

I just read through the page, very interesting.

I've been thinking about moving the small cap value (SCV) portion out of of my pp and into my vp, replacing it with something that will get my other cobbled-together pp equity holdings closer to total market. The main reason for this would be to just never sell the SCV, to only buy more.

Looking at the Golden Butterfly, which is probably also the name of some lethal Bruce Lee move, I have to rethink that move.

[iwealth]

I think a prosperity tilted VP is a great idea. Given nearly guaranteed government/central bank intervention when things get rough, it's tough to see how the US stock market can stay down for too long. Millions of workers are depositing money into their 401k's every single paycheck. Most of it goes to stocks. Equity bull market tailwinds are strong.

But I agree with koekebakker. Whether or not small cap value outperforms like it has in the past...impossible to say. I do wonder if the machines have eliminated much if not all of the small value premium. Consider that it is EASY to create stock screens using historical data that obliterate the performance of the S&P 500 or any other index benchmark. For a couple decades momentum trading blew everything away. But those anomalies are getting more and more difficult to exploit.

[Tyler]

Small and Mid Cap Blend work almost as well.  Even if the "value" premium is a thing of the past, it still seems like there might be value in broadening exposure beyond large cap funds that are influenced by fewer companies than their high-level index count implies.  Even large cap international funds just don't have the same effect. 

But that's just an observation based on the data and not on any theory behind the asset class.  I'm curious if anyone disagrees. 

[koekebakker]

This was the case over the last 40 years. Small/Mid/Value outperformed Large.
It's highly unlikely that those asset classes keep outperforming on that same level. Just take a quick look at the current valuations of small cap stocks...

The main advantage I see of over-weighting small/midcap is the possible diversification benefit. Even though correlation with the general stock market is very high the magnitude of the returns can be very different between large-caps and small-caps. During the 'lost decade' for example for stocks small-caps did relatively well.

The higher volatility of small-caps could be another advantage when adding it to the stock portion of the PP. Small-caps tend to follow the general market very closely but with higher ups and downs. Sounds perfect .

I would split the equity part of my PP 50/50 between large-cap blend and small/mid blend if there was a single decent Euro or global small-cap fund. But so far I haven't found one...

Great site btw!

[lordmetroid]

Well, I am convinced. Going to get me some of those nice mid term bonds(2-10 years seems better compared to less than 1 year, past 10 years had no years of no negative returns which is good enough for me) and small cap growth(Could not find any small cap value fund with a low fee)

[dragoncar]

How about 20% private equity?

[Drewskers]

(Could not find any small cap value fund with a low fee)

VBR .09%

[lordmetroid]

(Could not find any small cap value fund with a low fee)

VBR .09%

Sorry, I should have been more detailed.
I can not find any small cap value fund with a low fee in SEK on the swedish market.

[Tyler]

By the way, how did you decide on large cap blend rather than TSM?  I assume just keeping small and large cap separate for rebalancing opportunities?

One of the ideas I wanted to communicate is that large cap and small cap are two different stock indices that track different things, and that having some of both can be helpful.  Since a TSM fund also contains small caps, choosing large cap blend for the example was less likely to trip someone up.  Practically speaking, because of the way TSM funds are market weighted they act a lot like LCB anyway and substituting one for the other makes very little difference in this particular example. 

Very nice portfolio, BTW.  I can see why you like it.  One can apply good investing principles in a variety of productive ways. 

[mathjak107]

my opinion is i think you give the portfolio a bit more lifting ability  . especially because after the s&p 500 has been the only game in town like last year , it has always taken 5 years of under performance by the s&p 500 to  get back to normal .  having more midcaps and small caps while making things more volatile  can add to the pulling ability  if you insist on keeping gold and LT  treasury bonds which at some point will be fighting the fed .


those midcaps and small caps with the bigger equity allocation just may provide the forward traction you will need .

[murphy_p_t]

comparing VBR to VTI over, say, 10 years...charts overlay very closely. Shouldn't we attribute the outperformance of the Golden Butterfly vs the PP *mostly* to more $$ invested, kinda like a leveraged PP of 30/30/30/10?

[Tyler]

comparing VBR to VTI over, say, 10 years...charts overlay very closely. Shouldn't we attribute the outperformance of the Golden Butterfly vs the PP *mostly* to more $$ invested, kinda like a leveraged PP of 30/30/30/10?

The 30/30/30/10 results still don't really match the GB in returns or withdrawal rate, and fall below both the GB and PP in downside protection.  So IMHO it's not simply about reducing cash, but also about where that money is placed.  I personally think there might be something to be said for the large/small cap split, as well as for tilting a little more towards prosperity.  But to be honest those are just guesses, and I prefer to let the data speak for itself rather than to speculate on market "logic" too much. 

[lordmetroid]

Why do Small Cap Value make such a great suppelemt in the golden butterfly and why doesn't Small Cap Growth, Blend or Micro Caps do that same?

[koekebakker]

The most basic answer is because SCV outperformed the other asset (sub)classes and on top of that its correlation with the general market was relatively low. Adding SCV would help any backtested portfolio over this timeframe.

Wether the small/value premium persists going forward is a whole different question. There are an insane amount of threads about this on the bogleheads forum.

[MachineGhost]

Wether the small/value premium persists going forward is a whole different question. There are an insane amount of threads about this on the bogleheads forum.

Value actually has to be value to outperform.  The median stock is now more overvalued than in 2000.  Is there still pockets of undervaluedness by buying the bottom 20%-30% by price/book and the kitchen sink?

[Kbg]

My apologies for not remembering the details but I believe the latest academic research shows there is a small cap value premium but only for top decile value stocks in terms of quality measures. And what those are is what I can't remember.

[Cortopassi]

Desert,

Just looking back through some older trails involving Tyler's portfolios, and I see the allocation below from one of your posts.  What is Int Tr?

10% TSM
10% SCV
10% EM
60% Int Tr
10% Gold

I am liking the Golden Butterfly allocation and results, and the reason for all this is I finally got some time to look more in depth at Tyler's site starting with the latest Merriman portfolio, which is a bit too complex for me to want to pursue.  But the GB looks like a reasonable way to diversify past the PP (in my opinion).

Tyler, thanks for all the great work. 

[Cortopassi]

Tyler, at the risk of hijacking your thread here, I've been thinking more about the portfolio in my previous post:

25% Gold
25% LTT
50% Equity, split equally between TSM, SCV and EM

This portfolio is essentially a hybrid between Larry's Minimize Fat Tails portfolio and the PP.  And, in case it looks like these returns were merely a result of historical excess returns from equity tilts that have long since disappeared, this mix returned 10% over the last 10 years.  Tilting even smaller, by substituting mid cap blend for TSM, bumped the CAGR to 12.6%.  Your withdrawal rate calculator shows one could have withdrawn more than 7% from this portfolio over the past 40 years. 

Again, I recognize the hazards of backtesting, e.g., identifying mere artifacts in the historical returns through hindsight bias.  But this portfolio appears to provide some of the same things the HBPP does, but with a much larger prosperity bet, and with a Larry-style heavy tilt to higher risk equities. 

Thoughts?

Been playing all night with Tyler's calculators (put a donate button up, I WILL donate!), and this 17/17/16/25/25 TSM/SCV/EM/LTT/Gold really is shining for me.  I like the additional risk/reward of the stock split, while keeping my gold level to where I really like it (20-25%)

The only thing that is obvious in all these mixes is that there is no money market/cash component.  I need to mentally adjust for that, and just separate out a rainy day cash reserve separate from this.

I think Jan 1 I will start rebalancing to this allocation.  I also am open to thoughts.  I have no specific loyalty to the PP, it is still better than what I was doing, and this setup allows a lot of my current holdings to remain similar without having to trade out.

[ochotona]

If you get rid of your cash, your LTTs alone will expose you to rate risk. Consider 10 year Treasuries.

[mathjak107]

that is the issue spending down from the pp.

all well and good you have 25% cash but now that you spent it down  below your rebalance point you have to rebalance and refill selling volatile assets . it may be  no different from leaving the cash and spending it down equally from the  other parts  right from the get go ..

in a typical bucket system you exhaust the cash , then refill from short and  intermediate term bonds which are no where near as volatile as stocks would be like long term bonds are. finally many years later sell equity's to refill bonds and cash

so instead of refilling from short and intermediate term  bonds which may be down a little you are selling long term bonds which are  something as volatile as the stocks  or selling stocks or gold which run an equal chance of being down as much you are trying to avoid selling at a bad time .

with the pp you do not really have a non volatile 2nd line of defense to draw from if rates rise . .

i think anyone spending down from the pp has to examine this and find away to provide a secondary source for spending without selling assets as volatile as stocks are .

perhaps an income annuity may add time to allow other assets to  at least recover before they are needed  and prolong rebalancing to cash .

i don't know , how it would shake out as i never looked at the pp in that regard .

[Cortopassi]

I have probably 17 years till retirement.  I understand the spend down problem, but do you guys think that's something I even need to consider at this point?  Assuming I have a decent rainy day cash reserve outside of the this modified PP and will figure out spend down changes as I get closer to retirement?

[mathjak107]

not a thought for you yet .

[Tyler]

i think anyone spending down from the pp has to examine this and find away to provide a secondary source for spending without selling assets as volatile as stocks are .

I don't think I still have the spreadsheet, but I looked at this once.  A 1972 retiree using the PP with a 4% SWR and the 15-35 rebalancing bands would have rebalanced something like 8-10 times over the years (I don't remember the exact number).  Less than half of those times would it have been because cash dropped below 15%,  The rest were because stocks, gold, or LTTs all exceeded 35%. 

A PP retiree typically turns off automatic reinvestment of dividends and interest and uses that first for expenses before selling assets.  That usually does not completely cover expenses unless the portfolio is very large, but it definitely slows the rate of drawing down cash beyond what you'd initially think.  And when it does sell volatile assets, it only sells the highest ones as part of a normal rebalance.  I don't see that as a bad thing at all. 

In any case, that's the academic method and there are other good ways to approach it.  Personally, my PP drawdown method is sorta hybrid anyway.  I first live off of the dividends and interest, which funds almost half of my needs.  Each year, I rebalance as close to 4x25 as I can get while staying in the 0% tax bracket for capital gains and skim off enough in the process to cover the rest of my expenses.  That's a free rebalance most years, with tax-free funds coming from appreciated assets rather than cash.  In years where it is beneficial (like markets are struggling or there's a tax benefit to sitting tight) the cash is there to support me for years if necessary.  Cash is just one tool of many in the arsenal.  It just happens to be a highly flexible one. 

Been playing all night with Tyler's calculators (put a donate button up, I WILL donate!), and this 17/17/16/25/25 TSM/SCV/EM/LTT/Gold really is shining for me.  I like the additional risk/reward of the stock split, while keeping my gold level to where I really like it (20-25%)

The first hit is always free.    Thanks for the donate button idea.  I'll look into that.