Global Market Portfolio

Ask your investment, budget, and other money related questions here
User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Wed Jan 06, 2016 10:50 pm

FBeyer wrote:
BlueNote wrote:... I've personally been through 2 investing strategies and am currently on my 3rd (and hopefully last)...
How do you mentally separate your tendency to change strategies from market-timing and recency based scares, to analytically, theoretically well founded improvements on your current investing strategy?
brute would argue that even "picking and sticking to" one single strategy is market timing, unless one can somehow prove that this strategy works without making use of technical/fundamental insights.

IlliniDave
Posts: 1675
Joined: Wed Apr 02, 2014 7:46 pm

Re: Global Market Portfolio

Post by IlliniDave » Thu Jan 07, 2016 5:58 am

The question wasn't addressed to me but here goes anyway.

Initially I was what you'd call a performance-chaser when it came to stocks: invest in the funds that had strong recent past performance. My company's 401k essentially forced me in to this because they kept a selection of funds in the plan based on Morning Star rating, so any funds that faded they removed and replaced with something else in the 4- or 5-* range. Then a couple years ago they overhauled the plan to be primarily low cost income funds. Someone in a decision-making position turned boglehead, which in turn forced me to to look into the topic, and I was ultimately converted.

So my strategy now is to no longer try to pick "winners", either individual companies or fund managers. I just invest in publicly traded companies all over the world seeking only to gain the largest share I can (through low ERs) of their profit and growth in proportion to the money I put at risk. The folks who like to skim 1% or more of your wealth every year successfully perpetrated a myth that success was nothing less than "beating the market", which if you think about it is patently false, and ironically the pursuit of "winning", because of the cost involved, guarantees the majority of investors "lose".

So I have sort of an anti-strategy.

In the interest of full disclosure I do have small tilts towards US small cap and emerging markets (I hold them in slightly larger proportions than a pure market-cap weighting would call for). I hold 25%-30% in US bonds. And I have a tiny Roth IRA that is my play space where I hold an eclectic collection of things I play around with where I'm trying to assemble some things with relatively low correlation to see how I can do with tactical rebalancing.

But I'm 90%+ on the boglehead continuum and will probably stay there and just profit to whatever extent the world profits.

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Thu Jan 07, 2016 7:18 pm

BRUTE wrote:
FBeyer wrote:
BlueNote wrote:... I've personally been through 2 investing strategies and am currently on my 3rd (and hopefully last)...
How do you mentally separate your tendency to change strategies from market-timing and recency based scares, to analytically, theoretically well founded improvements on your current investing strategy?
brute would argue that even "picking and sticking to" one single strategy is market timing, unless one can somehow prove that this strategy works without making use of technical/fundamental insights.
The terms "market timing" or "timing the market" lack a clear shared definition. It's become an overused term IMHO. I think it would be better to simply describe the system you're using and why you like it. To some anything but the GMP is market timing, to others timing means something much more narrow like buying a stock because you think it's hit bottom.

I was thinking that the GMP might be a good base to start with and one could tweak this portfolio to match their personal needs.

User avatar
FBeyer
Posts: 678
Joined: Tue Oct 27, 2015 3:25 am

Re: Global Market Portfolio

Post by FBeyer » Fri Jan 08, 2016 4:03 am

BlueNote wrote:The terms "market timing" or "timing the market" lack a clear shared definition...
True, but buy-and-hold basically relies on historical market trends and autocorrelations, stretching back about 90 years. Buy low-now and sell high-now, kind of market timing, relies on recency and mixing with external factors, such as small political events as well as the agitation of other equally agitated 'would-be-quants'. We can't lump everything together in the market timing bin just because it has a temporal aspect to it.

Those of us who do no own a shred of economic envy will naturally choose long-term strategies, because we favour stability (or rather, unexciting investing) and income-per-hour-spent-on-managing-finances, compared to those who's like to make investing their day job. Managing portfolio stability is then another topic.

You can do a sort of hypothesis test of the two methods and find that the former passes up to a higher significance threshold. You can also choose to ignore the efficiency of the method and rely on the variance of the model swinging in your favour.

So far I've found that it's quite clear from context what kind of market timing were talking about, but it must be said that I am completely new to investing, so I'll have to trust you that the question of what kind of timing a given investor makes use of might be unclear.

User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Fri Jan 08, 2016 11:45 am

maybe it is of use to define market timing as a percentage instead of through a binary (unclear) term. someone who owns a single stock for a day might be at 100% market timing. someone who buys and holds the GMP forever might be 0%.

or categories might be more useful: day trader, hobbyist, boglehead, GMP. until humans invent something beyond the GMP.

interesting point: is it even possible to own securities in a way that actually completely eliminates any decisions by the investor, kind of like the indexing philosophy claims/is viewed? for this, consider that even picking a strategy is a decision, just as picking a fund manager is a decision. can there be a strategy to end all strategies?

can this be mathematically proven or disproven?

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Tue Jan 12, 2016 9:35 pm

I have updated the google spreadsheetwith a couple of new items based on some more research.

1: GMP Lower Cost version

This is a lower cost version of the original GMP (global market portfolio). The portfolio names are the same as the sheet names on the spreadsheet. It has an weighted average expense ratio of .20% which is 31% ($90 per $100k per year) lower than the original GMP.

2: Bid Ask Spreads

For those who don't know what bid and ask prices are, here is the short version. The bid price is the price which you can immediately sell shares, the ask price the price which you can immediately buy shares. If you paid no taxes or commissions and tried to buy some shares and sell them back to the stock market at the same time you'd still lose money because of the bid ask spread, therefore it is a real unavoidable cost. The bid ask spread is usually wider on low volume shares and thus makes trading stock in things like exotic ETF's, small/micro cap stocks and new issue ETF's more expensive.

Right now the bid ask spread data in my spreadsheet is wrong because I am getting market data from Yahoo! and the markets are closed. I'll lock in some bid ask spreads during tomorrows market and that'll give us an idea of the hidden cost of buying these securities. In particular the ITIP product seemed to have a very large bid ask spread compared to the other ETF's, it's thinly traded.

User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Thu Jan 14, 2016 1:27 am

brute tried to backtest both variants, but some of the funds only go back to 2014, so there really wasn't much to see. since 2014, the GMP definitely lost big time compared to the S&P500. it's still up, but not nearly as much. then again, that's a very short time frame to compare an indexing strategy.

brute thinks one could backtest the categories, i.e. us stocks, us bonds, emerging stocks, emerging bonds, .. instead of the actual, specific ones.

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Fri Jan 15, 2016 7:05 pm

BRUTE wrote:
brute thinks one could backtest the categories, i.e. us stocks, us bonds, emerging stocks, emerging bonds, .. instead of the actual, specific ones.
Quite a few index mutual funds have been around since the 70's and 80's. They work well for a backtest, but I think if you search around enough you'll find that someone has probably already done the GMP backtest and published results.

User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Sat Jan 16, 2016 12:14 am

brute could not find one just now.

User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Mon Jan 18, 2016 10:24 pm

brute has read Meb Faber's Global Asset Allocation. faber compares various well-known portfolios and backtests their performance from 1973 to 2013. among it the GMP (developed markets sovereign bonds 24%, emerging markets equities 9%, dm equities 32%, em corporate bonds 4%, dm corporate bonds 27%, em sovereign bonds 4%).

here are his results (1973-2013):

nominal return: 9.9%
volatility: 8.45%
sharpe ratio: 0.55
max drawdown: -26.87%

one thing that's interesting in this little book is that almost all the compared portfolios had almost the same returns and volatility. with the exception of the permanent portfolio (which did well during high inflationary times but terrible at all other times), all the nominal returns range between 9.5% and 10.45%. and even the PP does 8.53%, so not terribly far off.

fabers conclusion is that unless you just take your money out of any assets at all (which the PP does to 25%-50%, depending if you count gold as an asset), the specific asset allocation in different asset classes does not seem to matter much.

he emphasizes that international diversification and low fees are more important. a 1% fee will take you from the winning portfolio (10.45%) to below the losing portfolio (9.45% vs 9.5%, again excluding the PP at 8.53%).

it kind of makes sense to brute. there are only so many asset classes (equities, gov bonds, corp bonds, real estate, cash equivalents, precious metals, commodities) that one can diversify across. almost all of them (except cash) thrive in certain environments, and fail in others. how these environments occur throughout a period of time is unknowable ex-ante. so as long as one is even reasonably diversified across the asset classes (let's say 3+ classes), that's about as well as one can do.

the tiny differences in volatility and average return can easily explained by the random occurrence of the investment environments of the period one is examining.

brute concludes that yes, theoretically, the GMP is optimally diversified, especially because of the G part, which many other portfolios lack. but in practice, the advantage compared to most simple all weather portfolios is smaller than statistical noise.

if there exists a pre-packaged low-cost GMP etf, brute would probably buy it as his only investment source.

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Sun Jan 31, 2016 12:40 pm

@Brute. I knew I saw those figures somewhere before.

Have you also read Meb's paper where he looks into market timing filters as a layer on top of a globally diversified portfolio?: A Quantitative Approach to Tactical Asset Allocation

He also wrote a book about portfolio construction that uses broad internationally diversified asset classes: The Ivy Portfolio


After reading these works I ended up getting into other white papers and books around the topics of diversification, momentum, timing filters, etc.

It's pretty interesting stuff if you're into that sort of thing.

jacob
Site Admin
Posts: 9033
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: Global Market Portfolio

Post by jacob » Sun Jan 31, 2016 1:10 pm

Also consider that the market is now very much globalized while MPT and rebalancing has become dominantly popular strategies. One should, therefore, naturally expect a convergence to similar performances in all asset classes. It's a consequence of diminishing strategy-diversification. See the recent rebalancing thread for TMI along those lines.

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Tue Jan 31, 2017 7:06 pm

I updated the link to the example GMP I made in google sheets: https://docs.google.com/spreadsheets/d/ ... sp=sharing

ThisDinosaur
Posts: 576
Joined: Fri Jul 17, 2015 9:31 am

Re: Global Market Portfolio

Post by ThisDinosaur » Wed Mar 08, 2017 10:28 am

If the price change component of total return is optimized by owning an asset *before* the unwashed masses want to buy it, wouldn't it make sense to own mainly the smallest components of the investable GMP? Like, if GMP represents the aggregate average investor's preferences, sooner or later they will rush out of stocks and bonds and into some other investable asset. So allocating to the *underrepresented* REITs, TIPS, EM debt like a contrarian should ultimately win, yes?

User avatar
BRUTE
Posts: 2520
Joined: Sat Dec 26, 2015 5:20 pm

Re: Global Market Portfolio

Post by BRUTE » Wed Mar 08, 2017 1:08 pm

all on small cap growth?

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Wed Mar 15, 2017 9:19 pm

@ThisDinosaur

The GMP strategy assumes:

- Stocks, bonds and real estate are investment assets, everything else isn't or is simply a derivative of those asset classes.
- You can't pick stocks/bonds/real estate and expect to beat the market (on a risk adjusted basis) because the market is the sum total of all investment decisions and you're never going to be as smart, adaptable and sophisticated as the market.
- The more diversification the better
- You can't even do an asset allocation better then the market so you should just let the market do that too.

The GMP has all sorts of implementation issues for the 'average' investor though. For example:

- Exchange rates are way more volatile than bond values and can also overwhelm stock and real estate values. Currency hedging comes at a cost of the effort/time of implementing it, the cost of the hedging securities and the risk that it won't work as planned.

- Taxes vary in different countries , in some cases taxes are so onerous on certain investment areas (foreign stocks, bonds, foreign bonds etc.) that investing in those asset classes make no sense.



Therefore according to the GMP strategy it's bad to be a contrarian and good to be smack dab in the middle of the herd where you can ride for free on the markets supposedly superior opinion of prices and asset allocation.


BTW small cap growth is, historically, one of the worst performing stock screens. Small cap works well, historically, with a value screen though. That doesn't mean that those results will hold for the future though.

ThisDinosaur
Posts: 576
Joined: Fri Jul 17, 2015 9:31 am

Re: Global Market Portfolio

Post by ThisDinosaur » Thu Mar 16, 2017 6:38 am

BlueNote, are you aware of any information on the past performance of the GMP? The closest I can find is past projections of the current (approximated) allocation projected backwards and rebalanced as a fixed allocation. But that's obviously flawed.

Shouldn't a non-rebalanced GMP have minimal volatility? Shouldn't it have essentially zero volatility if you include correct allocations of gold, cash, and commodities?
You can't pick stocks/bonds/real estate and expect to beat the market (on a risk adjusted basis) because the market is the sum total of all investment decisions and you're never going to be as smart, adaptable and sophisticated as the market.
I've always had trouble with this assumption. It seems to claim that every single investor is below average. But I read something by Sharpe recently where he says the Market price implicitly includes a weighted statistical insight of all predictions. I suppose that makes sense. Even more so if the investors with the most money invested are also the most well informed.
- Exchange rates are way more volatile than bond values and can also overwhelm stock and real estate values. Currency hedging comes at a cost of the effort/time of implementing it, the cost of the hedging securities and the risk that it won't work as planned.
If PPP generally holds, shouldn't exchange rate volatility cancel out as noise over multi-decade investment horizons? Especially if you hold GMP and never have to rebalance?
- Taxes vary in different countries , in some cases taxes are so onerous on certain investment areas (foreign stocks, bonds, foreign bonds etc.) that investing in those asset classes make no sense.
Can you talk more about this? Any specifics?
according to the GMP strategy it's bad to be a contrarian
Popular sources of investing advice are flooded with the idea that contrarian is good. Be greedy when others are fearful etc. Rebalancing a multi-asset portfolio is cyclical contrarianism, as I see it. When we optimize our way up to the "perfect" index portfolio, all of a sudden that whole line of thought gets abandoned? *This* is where I feel like I'm missing something important. When I see a clear picture of where every other investor in aggregate is putting their money, it seems like that's exactly where I *shouldn't* be following the herd.

@BRUTE
I'm not trying to be argumentative. I genuinely feel like thinking about this portfolio is exposing some very fundamental flaw in my understanding.

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Thu Mar 16, 2017 9:55 pm

@ThisDinosaur

Lot's of questions there. I'm not an expert or anything so here's my opinion based on my current base of knowledge.

I want to note that I don't recommend the GMP investment strategy, nor do I use it. My portfolio is super diversified but I don't invest in foreign bonds due to the currency volatility and tax issues for me in my country. However it is interesting and may work well in pure form or as a template for some people.

Read the Meb Faber document cited by Brute earlier in this post for historical performance, the works cited in that document will lead you to yet more research material. Researching back through citations like this is a great way to learn more.

Regarding elimination of volatility

You generally can't eliminate that because all assets are priced relative one another. Even cash is relative, the value of cash one minute ago isn't the same as cash now because what it can buy is forever changing. Diversification tends to dampen it because of the un-coorelated nature of different asset classes but it's always going to be there. There may be some highly theoretical ideas about how to eliminate it, I'm not aware of any, but that's just pie in the sky.

William Sharpe is good background reading for the GMP, I think his ideas are behind a lot of it.

Will PPP hold?

Time is sort of like a form of diversification in that the longer the duration of time you use for comparisons the less volatility you'll see. For example daily stock returns are more volatile then weekly which is more volatile then monthly etc. This holds for exchange rates too. Also the more countries you're invested in the less volatility you'll expect to see. However I don't think the currency volatility is ever totally cancelled out.


Taxes.

For example withholding taxes. Many countries will withhold a certain portion of dividends and interest paid to foreigners before it ever leaves the home country. Foreigners often have no control over this and it is taken at the source. Then after the foreigner gets their cut of the distribution the home country will often tax the remaining foreign income higher then they do distributions from home country securities. So any 'free lunch' from international diversification is often wiped out (and then some) due to discriminatory taxation policy. You'd need to research the relevant tax code in your country to see how it works there, usually a google search on the topic is helpful.

Contrarianism.

In my opinion contrarianism alone isn't going to allow you to beat the market, it is a necessary component though. Also would the general public ever buy a book about an investing strategy that sometimes does well?

I find it difficult to find really obviously good investments that are better then the market index. I find it really , really hard . I do have an active portfolio (10% of my wealth) which has 2 investments right now. So far I am beating the markets but not by much and it could just be luck.

ThisDinosaur
Posts: 576
Joined: Fri Jul 17, 2015 9:31 am

Re: Global Market Portfolio

Post by ThisDinosaur » Fri Mar 17, 2017 2:55 pm

Read the Meb Faber document cited by Brute
I became aware of GMP after reading this post on Meb's site:
http://mebfaber.com/2015/05/30/chapter- ... portfolio/
Figure 32b compares GMP and GAA(which is GMP plus allocations to Gold and Commodities) to stocks, bonds and T bills.
Image
Notice how GAA has a max drawdown of 26.72%.

Here is where my brain goes all logic bomb:
You hold a perfectly sampled selection of all investable assets and hedges, i.e., one that matches the actual market. When investors leave one asset class and go into another, the change in values should exactly match eachother (your total portfolio value stays the same + earnings, dividends, and interest) UNLESS they are investing in a tertiary asset that you aren't (like cash in their mattress.) If you hold the right proportion of cash, then the purchasing power of that cash goes UP, you rebalance (or not?), and you should have a Max Drawdown as low as the error between your GMP sample and the precise actual market.

Why/How is this wrong?

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Fri Mar 17, 2017 11:16 pm


ThisDinosaur
Posts: 576
Joined: Fri Jul 17, 2015 9:31 am

Re: Global Market Portfolio

Post by ThisDinosaur » Sat Mar 18, 2017 7:07 am

"The market portfolio is unobservable."
So the GMP can only ever be an incomplete, inaccurate approximation. But shouldn't this approximation be significantly less volatile than other Fixed Asset Allocations? Brute quotes Faber's book (which i haven't read yet) as saying the performance of multi asset portfolios is essentially identical.
http://mebfaber.com/2016/05/18/institut ... on-models/
http://mebfaber.com/2015/08/23/your-buy ... nt-matter/

So why should it be that GMP performs so similarly to these other fixed asset models, rebalanced annually? Jacob offers one compelling explanation. That this period studied is the MPT era.

Does rebalancing function to 1)reduce volatility by better approximating the GMP, 2)maintain a specific "risk profile", or 3)sell assets at their peak/buy at the trough in order to increase your returns above those of other investors.

I've always preferred explanation#3, but Sharpe points out that, if this is what all investors (active and passive) are trying to do, it is mathematically impossible for all of us to accomplish.
http://web.stanford.edu/~wfsharpe/aaap/wfsaaap.pdf

User avatar
BlueNote
Posts: 389
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Global Market Portfolio

Post by BlueNote » Mon Mar 20, 2017 5:54 pm

@ThisDinosaur

Re-Balancing Favours Sideways markets , Adaptive Asset allocation (from the aforementioned Sharpe paper) favours trending markets.
over an investors lifetime as the market values of the underlying securities change. The rationale being that the market is the best arbiter of value and relative riskiness so the portfolio construction should use markets as a guide to asset allocation in order to maintain a desired level of risk.

Here is a quote from his paper:

"Figure 2 shows that over this period our fund varied from being significantly riskier than the U.S. bond+stock market to being considerably less risky. At the end of March 2000, the proportion of market value in stocks was 75.06%, leading to the lowest relative risk for the fund in the entire period. At the end of February 2009, the situation was just the opposite. The proportion of market value in stocks fell to its nadir of 43.18%; making the fund, at 60%, much riskier than the overall U.S. bond plus stock market."

So it appears that adaptive asset allocation would have suggested a heavier stock weighting right before the dotcom bust and a lighter weighting on stocks right before the current mega-bull market we've been experiencing. I'll just leave things at that, I think both re-balancing and adaptive asset allocation can work. Adaptive asset allocation is lazier and probably easier to implement. Re-Balancing is contrarian, buy into weakness and sell into strength, maintain the fixed allocation approach. Either approach is basic enough that they're both still probably within the spectrum of investing strategies that people consider passive, adaptive asset allocation is arguably more passive.

Post Reply