Portfolio Charts

Ask your investment, budget, and other money related questions here
BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

risk-adjusted return seems really hard to understand. brute thanks Tyler9000 for the examples. brute believes humans want to believe in the narrative that "risk == return", which is not necessarily true.

some humans also want to fetishize pain and its tolerance, maybe because of protestant work ethic? brute thinks, for example, that one popular narrative among indexers and especially 100% stock indexers (MMM, jhcollins, GCC) is that since risk == reward, pain is good. these humans seem to almost celebrate their stock losses, because they believe it is what enables them to get richer.

brute isn't so sure about fetishizing losses. he likens it to realizing that taking a band-aid off quickly is painful but better. but then they go out and seek pain, thinking it must be even better to feel more pain.

sure, humans would probably do good not to panic in a stock market cash, sticking to their strategy. but it's not necessarily true that the more volatile one's portfolio behaves, the richer one will become.

did brute get that right?

Tyler9000
Posts: 1758
Joined: Fri Jun 01, 2012 11:45 pm

Re: Portfolio Charts

Post by Tyler9000 »

"Fetishizing losses" is an astute observation that I think is absolutely true. This is easiest for young investors who have yet to actually experience a true bad year like 2008, but even experienced investors can fall into it as a natural defense mechanism. I'll also point out that believing you will have the money to purchase shares during a crash with 100% in any one asset is a very job-centric perspective.

Another phenomenon I see a lot is idealizing a vague and never quantified "long-term" where the promised superior returns are always right around the corner if you wait long enough. That one might actually need to use the money before then is too often ignored. Investors tend to fixate on maximizing their theoretical future net worth and lose sight of why they save to begin with.

And then there's the Robot Fallacy, where intelligent people rationally believe that emotions are stupid and assume they're much stronger than that and will never sell low. Sure it is best not to panic, but to assume you are personally above normal human psychology is naive. Even for BRUTE and HAL9000's cousin. ;)

IMHO, the most important takeaway is that one does not need to invest in a risky and volatile way to achieve satisfactory returns. Contrary to popular belief, the risk/reward relationship in investing is not necessarily linear for every portfolio. And once you start thinking of investing as an activity to support your important life goals rather than a maximization game of chance, different portfolios may appeal to you.

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

haha team AI :D

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

playing around with the calculators, brute has discovered that a GB substituting emerging markets for small value performs slightly better historically, but does not seem to have longer draw down.

if one is willing to accept 4 year draw downs, a GB with large cap, small cap, emerging, long term treasury, and gold seems to enable a SuWR of 6.6% and SWR of 7.5%(!) over 30 years

interestingly, this last portfolio cannot be found in the portfolio finder? brute is requiring those 5 assets and enters a 6% minimum, but the list comes up empty. edit: nvm, it has only 5% cagr and is thusly found.

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

new favorite: the emerging value butterfly

20% large cap value
20% small cap value
20% emerging
20% ltt
20% gold

cagr 8.12% since 1972
3 years longest drawdown
-18.7% worst year

enables a 7.3% SuWR and 7.7% SWR, i.e. FI in 8-12 years with a 50% savings rate starting from scratch, or 4-5 years at 75%

either brute is doing something wrong, or this thing is ERE on steroids. opinions?

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

feature request: comparison over time of how each asset class did during each year. this would enable brute to correlate different asset classes against each other and see which form good combinations.

also, comparing the rightmost (final) CAGR for each starting year when comparing 2 portfolios, to compare which portfolio would've won depending on start year.

FBeyer
Posts: 1069
Joined: Tue Oct 27, 2015 3:25 am

Re: Portfolio Charts

Post by FBeyer »

BRUTE wrote:new favorite: the emerging value butterfly
... opinions?
You don't use portfolio backtesting to find the portfolio that had the highest historical CAGR.
Whatever time period you choose you can always find THE ONE, that performed best. That means nothing in the future.
Japan is the novice example given in The Intelligent Asset Allocator. Portfolio backtesting from 1970 up to 1990 would have told you to BUY JAPAN NOW!!!
Check out how that would've fared: http://www.thebubblebubble.com/japan-bubble/

Portfolio backtesting is about gauging the stability of returns of your portfolio not the expected returns of your portfolio.

Tyler9000
Posts: 1758
Joined: Fri Jun 01, 2012 11:45 pm

Re: Portfolio Charts

Post by Tyler9000 »

FBeyer wrote: Portfolio backtesting is about gauging the stability of returns of your portfolio not the expected returns of your portfolio.
Very true.

One tool I particularly like for this sorta detail is the Portfolio Growth calculator. It's really good at showing both stability and trends. An extreme example is Emerging Markets:

Image

Note that the lighter colors are from more recent timeframes. Also note how many endpoints you can see at the bottom of the chart. These are both signs that the high historical returns that one sees with Emerging Markets are generated mostly by past periods that aren't around anymore, and that recent history is setting new lows for the data set. Both returns and SWRs may not remain as advertised.

@Brute: I'm looking at ways to make it easier to directly compare portfolios. I'll see what I can do.

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

FBeyer wrote:Whatever time period you choose you can always find THE ONE, that performed best. That means nothing in the future.
Japan is the novice example given in The Intelligent Asset Allocator. Portfolio backtesting from 1970 up to 1990 would have told you to BUY JAPAN NOW!!!
Check out how that would've fared: http://www.thebubblebubble.com/japan-bubble/
this is not betting on a single country, or asset class. this portfolio contains stocks, gold, and long term bonds, and is therefore more diversified than most of the common stock/bond slide portfolios - and still beats them in overall returns?

brute is mainly excited about the idea of the CAGR - not knowing about this before, average returns seem a complete scam now. just comparing this measure in assets completely changes the game, doesn't it? all of the sudden a Total Stock portfolio doesn't seem that great, whereas the PP seems pretty good.

the way brute sees it, it's kind of common wisdom among many humans that "volatility + time == wealth", leading them to very stock centric portfolios. but if measured with actual compound returns, or CAGR, this is not true, as volatility can drain gains quickly. it becomes more important to avoid losses, even if this means not grabbing all the rallies.

in CAGR terms since 1972, a Total Stock portfolio and the PP are a measly 1.1%pa apart. TS and Golden Butterfly are 0.1% apart - for just about the riskiest vs. just about the safest/most diversified portfolio one could imagine.

steveo73
Posts: 1733
Joined: Sat Jul 06, 2013 6:52 pm

Re: Portfolio Charts

Post by steveo73 »

Brute - this is discussed on the MMM forum. It's not as simple as what you are stating. I think that there is a massive chance that the PP and Golden Butterfly portfolios are the result of data mining. Yeah the results some impressive but will those some results hold true for the next 40-60 years.
Last edited by steveo73 on Mon May 09, 2016 4:48 am, edited 1 time in total.

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

feature request: predict future ;)

of course that's an unknown. does steveo73 have a link to the relevant MMM discussion? nvm: http://forum.mrmoneymustache.com/invest ... butterfly/

Tyler9000
Posts: 1758
Joined: Fri Jun 01, 2012 11:45 pm

Re: Portfolio Charts

Post by Tyler9000 »

steveo73 wrote: I think that there is a massive chance that the PP and Golden Butterfly portfolios are the result of data mining.
Rather than rehash that conversation, I would recommend that people read the works by Harry Browne before passing judgment on the origins of the Permanent Portfolio. One is free to not share his perspective, but he was not a data miner.
Last edited by Tyler9000 on Sun May 08, 2016 5:35 pm, edited 1 time in total.

theanimal
Posts: 2628
Joined: Fri Jan 25, 2013 10:05 pm
Location: AK
Contact:

Re: Portfolio Charts

Post by theanimal »

I appreciate the reminder why I don't visit the MMM forums.

As a fan of Harry Browne and the PP, the golden butterfly is very appealing. I'll be looking into this a lot more. Thanks, Tyler! Love the site.

Dragline
Posts: 4436
Joined: Wed Aug 24, 2011 1:50 am

Re: Portfolio Charts

Post by Dragline »

This is a great discussion, at least since my last comment. I agree HB was limited by his times, but his views were not built on data per se, but on fundamental ideas of how he thought different asset classes would perform in different conditions.

FBeyer
Posts: 1069
Joined: Tue Oct 27, 2015 3:25 am

Re: Portfolio Charts

Post by FBeyer »

BRUTE wrote:
FBeyer wrote:Whatever time period you choose you can always find THE ONE, that performed best. That means nothing in the future.
Japan is the novice example given in The Intelligent Asset Allocator. Portfolio backtesting from 1970 up to 1990 would have told you to BUY JAPAN NOW!!!
Check out how that would've fared: http://www.thebubblebubble.com/japan-bubble/
this is not betting on a single country, or asset class....
You're missing the point. The faulty use of portfolio backtesting chooses what was best in the past, and applies that to the future. It does not matter whether the 'past best' was a single asset or several. Japan is just the easiest thing to point to in that regard.

I'll concur that PP and the GB look absolutely wonderful for all sorts of reasons(*), but those are the reasons stated in Fail-Safe Investing (Harry Browne) not the color of the right-most pixels on the portfoliofinder pixel chart.

(*) If the Danish taxation laws were slightly different, and if private persons could invest in REIT I would build something that looks a bit like the GB. I am very big fan of the idea of a Permanent type portfolio, especially the butterfly like one with REIT rather than gold. Don't get me wrong. I ONLY arguing about how to apply backtesting and what to conclude from it.

steveo73
Posts: 1733
Joined: Sat Jul 06, 2013 6:52 pm

Re: Portfolio Charts

Post by steveo73 »

BRUTE wrote:feature request: predict future ;)

of course that's an unknown. does steveo73 have a link to the relevant MMM discussion? nvm: http://forum.mrmoneymustache.com/invest ... butterfly/
That is the one. It's a great thread.

steveo73
Posts: 1733
Joined: Sat Jul 06, 2013 6:52 pm

Re: Portfolio Charts

Post by steveo73 »

Tyler9000 wrote:
steveo73 wrote: I think that there is a massive chance that the PP and Golden Butterfly portfolios are the result of data mining.
Rather than rehash that conversation, I would recommend that people read the works by Harry Browne before passing judgment on the origins of the Permanent Portfolio. One is free to not share his perspective, but he was not a data miner.
Dragline wrote:This is a great discussion, at least since my last comment. I agree HB was limited by his times, but his views were not built on data per se, but on fundamental ideas of how he thought different asset classes would perform in different conditions.
I agree that the idea of the PP didn't come up from backtesting. The point is though will that portfolio perform in a similar fashion in the future as it has within the past. We don't know. It might but it might not.

So I should clarify my comments to state that the good performances of the PP and the GB may be statistical anomalies.

In stating that I am a fan of a diversified portfolio across asset classes and within asset classes.

BRUTE
Posts: 3797
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE »

wow, in reading that MMM thread, brute has learned a lot:
1)there are some incredibly rude humans on the MMM forums, brute will stick with the polite anti-social ubernerds here
2)they don't have very good arguments against the PP/GB

the main argument against PP/GB seems to be that no human can predict the future. therefore, all humans should invest in 100% stocks. lol.

uncertainty is a thing, and a good argument. but arguing with uncertainty for a single asset class or just 2, vs. a more diversified portfolio should be included as the definition of irony. if a human is afraid of uncertainty, he or she should be investing in the Global Market Portfolio, not in 100% domestic stocks or 70/30 or 50/20/30 domestic/foreign/bonds.

the humans over at MMM seem extremely invested in their ideas, and not open to new ones. brute was laughing out loud at times, reading statements like "clearly you don't understand.." and "ah, the old rhetorical trick of making me look dumb".

what brute finds interesting about PP/GB is that apparently, a diversified portfolio can beat a more stock-centric portfolio at WR, which is what counts for brute (and probably most FI types). sure, the exact GB might be a result of backtesting and survivorship bias, but the basic idea works if you substitute any of the parts.

brute likes the slightly more aggressive butterfly (is it still a butterfly?), substituting another aggressive stock fund for the cash. in effect, this makes it a 20% conservative stock, 40% aggressive stock, 20% bonds, 20% other portfolio. one can substitute many broad index funds here. REITs or commodities for gold. TSM funds. international funds. whatever.

if this portfolio worked exactly like described, but any small changes made it break down, brute would attribute this to backtesting error. but the fundamental strategy seems sound.

and certainly the logical conclusion of "the future is uncertain" is not "100% stocks, because stocks always go up".

side rant: brute hates how humans on MMM always talk about "the market", "beating the market", "market returns", but really mean "domestic stock market". "owning the market" in any reasonably sense would mean owning the GMP, not 100% domestic stock. making fun of other humans for "trying to beat the market" if they disagree with 100% domestic stock is absurd.

FBeyer
Posts: 1069
Joined: Tue Oct 27, 2015 3:25 am

Re: Portfolio Charts

Post by FBeyer »

BRUTE should develop a better assholes-on-the-internet filter. Perhaps BRUTE should check into a shop for a software upgrade :)

Toska2
Posts: 420
Joined: Fri Nov 20, 2015 8:51 pm

Re: Portfolio Charts

Post by Toska2 »

I like my "Brick and Motar Moth".

30% TSM 30% International 20% REITS 10% Small cap value 10% Emerging markets. My elevator speech on it, " It's no Icarus and it gets a little hairy but it still flies." I admit my speech needs work. ;)

Post Reply