Your International vs. U.S. Stock Allocation?

[trasmuss]

The expenses of Total International Admiral are three times higher than Total Stock Admiral. It is worth mentioning for those who feel expenses are important and especially those with large portfolios.

[nisiprius]

The expenses of Total International Admiral are three times higher than Total Stock Admiral. It is worth mentioning for those who feel expenses are important and especially those with large portfolios.

Let me put this in boldface: with expense ratios, what matters is the absolute difference, not the ratio.

If you had a checking account that was paying 0.05% interest, i.e. $5 a year on a $10,000 account, and the bank across the street put up a sign saying WE PAY 260% HIGHER INTEREST, i.e. 0.18%, i.e. $18 or $13 a year more, would you immediately run across the street and signing up, buy all-new checks, go home and enter the new routing and account numbers in every online account, etc? And that's an absolutely sure thing, $13 bucks extra every year--well, a sure thing unless the 0.18% is a teaser rate that's cut six months later.

Total International Admiral's expenses are 0.13% per year more than Total Stock Market Index.

That's peanuts.

That's chump change.

If it were a difference between 1.80% ER and 0.50% ER then sure, that should factor into your decision. But speaking as an international skeptic who only holds 20% international, the tiny cost difference has never factored into my allocation decision. Anyone who thinks international is going to do their portfolio any detectible amount of good at all thinks it's going to do more than 0.13% worth of good.

(And, yes, I can figure out what happens if you compound 0.13% for twenty years. It's 2.632%).

Of course we all have different portfolio sizes and personal "utility functions," and if minimizing costs is absolute top priority, it is what it is. But in that case, shouldn't we be talking Schwab and not Vanguard?

[Call_Me_Op]

Of course we all have different portfolio sizes and personal "utility functions," and if minimizing costs is absolute top priority, it is what it is. But in that case, shouldn't we be talking Schwab and not Vanguard?

..or Fidelity.

[Paul@]

35% international, 65% U.S. Am comfortable sticking with that although a good case can be made that the international ought to be higher.

[Noobvestor]

Obviously holding a global portfolio will give you somewhat greater diversification (at the expense of currency risk) than holding a domestic-only portfolio. It's equally true that Vanguard Total Stock Market Index Fund gives you greater diversification than holding Vanguard 500 Index. And yet, you'd be hard-pressed to show that holding 500 Index instead of Total Stock has ever ruined anyone's retirement. In the case of international, the question is whether the volatility reduction you get from the imperfect but high correlation of international and domestic outweighs the extra volatility you get from the currency risk of the international.

But Nis, this is an easy one: look at Japan. Other examples exist as well (you've seen the chart I tirelessly post showing the distribution of returns across markets over the last hundred years). Unless you want to argue that the US is unique and cannot suffer decades of global-market underperformance, I don't see any easy argument against the Japan example. Japan is also interesting because according to the data as presented by Larry, I believe, (don't have the citation handy, sorry) you actually would have been significantly better off holding small caps than large. I don't know how big a difference it would have made to hold the 500 largest vs. the total market, but if you had tilted small, you'd have weathered their bad decades better, anyway. Meanwhile, the Japanese market went up and down largely with the global market - i.e. the correlations were high, but it still had dismal returns.

Also, the comparison you make is a numerical stretch - we're talking about holding 45% of the global equity market (in terms of US/rest of the world) vs. holding 75% of the US market (in terms of large/small). Holding the last 25% intuitively seems less of a necessary step than holding the last 55% of the global market, not to mention geographical/political/economic/etc... diversification issues in the latter case.

But, to the point - two questions: Is Japan a valid example of the potential underperformance of a single country over a long period? Is it of retirement-ruining caliber?

If 'no' to either or both of those questions: why not?

PS Thanks for weighing in on the ER - I'm very glad we agree on that. I definitely would not let such a small ER difference effect, for example, the distribution of my bond holdings, so why anyone would let it play a significant role on the stock side is beyond me!

[trasmuss]

Nisprius; I think we agree that the definition of "peanuts" and "chump change" depends on the size of the portfolio. .13% can amount to hundreds of dollars a year.

Tom

[Noobvestor]

Nisprius; I think we agree that the definition of "peanuts" and "chump change" depends on the size of the portfolio. .13% can amount to hundreds of dollars a year.

Tom

If it does, though, you have to ask: how much do those hundreds of dollars a year matter to the person holding that portfolio? It remains a very very tiny fraction of a percent in relation to the whole, no matter how much the person is holding, and I believe that is the core point here.

But really, just ask yourself: if your portfolio called for 50/50 TIPS/short-term treasuries, but TIPS had a .13% higher ER, would you give up your inflation-fighting portfolio component to save .13% in fees? I don't think it would make sense to, but that's just me!

For anything under .25%, I don't even let the differences register in my decision-making process since the *relative value* of diversifying to each sub-and-asset-class I chose is many factors larger than .25%, regardless. If it weren't, I wouldn't bother holding it.

[convert949]

75/25 US/Intl. Half way between Mr. Bogles 20% max and Vanguards 30%. FYI 63 and retired. Sleep fine...

Regards to all,

Bob

[Groundhog]

2/3 US, 1/3 International

Same here.

[MnD]

I just copy the US-vs-non-US split implemented in Vanguard Total World Stock Market. I live in the US but I think it makes sense to follow the overall global markets rather than hope that the US will outperform everyone else.

This what we do 45% US, 55% int overall and 13.5% EM. Both of our jobs are closely tied to the domestic economy.

[ruanddu]

Thanks again everyone. It's very interesting to get your outlook. I welcome more replies.

[staythecourse]

After reading numerous articles and books, discussing ad infintum on this site, and pondering for hours re: U.S vs. international allocation I have come up with some simple points that I can not refute:
1. There are times U.S. has outperformed in the past
2. There are times International has outperformed in the past
3. There is no way for ANYONE to predict which will do best going forward

These 3 points lead me to sit on 50/50. I have not read ANYTHING that refutes these points. In the end I agree with Vanguard's paper of anything from 20% to market cap is reasonable and should be based on the investor's preferences, i.e. frame of reference risk.

Good luck.

[backofbeyond]

50% US, 50% International here.

I live in New Zealand, before that Canada, Japan, Germany and the UK. With a few stops in the States every 3-5 years. Plan on continuing this international pattern on international leap frogging for the duration of my career..hopefully, with good health, another 20 years.

Then in retirement, will continue to spend 1/2 the year in Florida and 1/2 year in either Asia or Europe (with the added caveat of depending where the grandchildren are of course).

In retirement approx 38% of cash flow from pensions will be in British Pounds, the rest in US Dollars.

To me, it makes sense (or is it cents?) to be globally invested based on projected lifestyle. If I was staying in the States, I'd probably stay with a 20% exposure.

[Leif Eriksen]

I'm 50%/50% US vs Intl. It is not for any particular prediction, such as the US is a mess or Intl is a mess. Predictions are a fool's game. I expect their returns, in the long term, will be similar. I have 50/50 to give me diversification.

[stan1]

I'm about 1/3 international 2/3 U.S. equity.
Managing currency risk is what drives me to be somewhat under market weight.

[Tom_T]

I'm 50%/50% US vs Intl. It is not for any particular prediction, such as the US is a mess or Intl is a mess.

Another prediction is that the entire world is a mess, but we don't have a Emerging Planets ETF yet, so we'll have to make do.

[jeepie]

For equities, I follow the global market cap breakdown generally (roughly 55% non-US). This is accomplished with a slight tilt to small cap and value, and including REITs, which some consider to be a separate asset class. I currently use DFA Funds, but have used Vanguard for many years in the past. I don't believe there is one "correct" answer in this regard.

[Default User BR]

I'm 50%/50% US vs Intl. It is not for any particular prediction, such as the US is a mess or Intl is a mess.

Another prediction is that the entire world is a mess, but we don't have a Emerging Planets ETF yet, so we'll have to make do.

Then what was that Total Martian fund I just invested in?


Brian

[Valuethinker]

After reading numerous articles and books, discussing ad infintum on this site, and pondering for hours re: U.S vs. international allocation I have come up with some simple points that I can not refute:
1. There are times U.S. has outperformed in the past
2. There are times International has outperformed in the past
3. There is no way for ANYONE to predict which will do best going forward

These 3 points lead me to sit on 50/50. I have not read ANYTHING that refutes these points. In the end I agree with Vanguard's paper of anything from 20% to market cap is reasonable and should be based on the investor's preferences, i.e. frame of reference risk.

Good luck.

Your retirement liabilities are in USD. So any currency outside of that exposes you to risk.

However you probably have US Social Security and US housing equity. That's an implicit weighting back into USA (ditto any pensions etc.).

the theoretical optimum is by market cap (ie 55% non US, 45% US) with currency hedged back to USD. Since that latter is expensive (and reduces the diversification offered by holding assets in foreign currencies) one probably doesn't do it.

So *as long as you pay attention to total currency risk* then 45/55 is defensible on your equity portfolio.

The historical best percentage in terms of minimizing risk/ maximizing return for a USD investor has been 20-30% of portfolio in international. There's no way of telling whether more will be better in the future, or less. If you look at the historic numbers, you bought more volatility going beyond 30%, without meaningfully adding to return.

On a 60/40 equity bond portfolio I would suggest anywhere from 1/3 to 1/2 of equities in international (ie 20-30% of total portfolio) works best for most USians.

Rather than holding international bonds I would suggest holding I Bonds and TIPS, however yields on the latter are currently very unattractive. The US CPI hedge proxies for a currency hedge (if Purchasing Power Parity holds-- which it doesn't of course ).

For those of us from smaller countries with less deep stock markets the case for having much greater degrees of international diversification is stronger. Canada is only 4.5% of world stock markets (c.) and is NOT a diversified index (80% natural resources + financials).

I suspect the 'low volatility' international portfolios add even more diversification. The small cap international value certainly do (if you can get access to DFA product, or find a similar proxy-- the trick is dealing costs really really count in this space, liquidity is very low in the stocks, so you have to be really clever to get the performance, also the accounting numbers are very imperfect so 'cleaning' those numbers is rewarded).

However the low volatility ETFs I have seen (Toronto) have tiny market caps-- it's too early. And this could be more data mining (or, we have discovered the 'Buffett Anomaly' in which case it is about to be arbitraged away).

Larry Swedroe stuff good on all this in layman's language.

[winguy]

So out of the total portfolio (equities + fixed income + others), <30% total currency risk is acceptable, >30% is not?

[FNK]

So out of the total portfolio (equities + fixed income + others), <30% total currency risk is acceptable, >30% is not?

Given that the standard advice is 20% bonds at the minimum, an aggressive portfolio could have 40% intl (80% * 50%).

Now, let's look and the two risks of international equities a little closer. Again, let's separate corporations from markets. I claim that currency risk arises primarily from the markets a company operates in, and political risk arises primarily from the nations in which the corporation is registered.

A US corporation doing a lot of business abroad has incomes in various currency and those fluctuations will surface in the bottom line. A foreign corporation doing a lot of business stateside will experience a lot of dollar hedging by virtue of, well, collecting dollars.

On the other hand, a company getting nationalized or overtaxed by its government is a problem for its investors. On the other hand, a local company getting an unfair advantage over a US company is receiving a boost that you might as well capture.

So... I guess the stock market has already internalized all this stuff and you don't want to diverge too far from it.

[Valuethinker]

So out of the total portfolio (equities + fixed income + others), <30% total currency risk is acceptable, >30% is not?

I would not use 'acceptable' but perhaps 'advisable' or 'accords with historical experience'.

The existence of US housing equity and US SS benefits implies a higher weighting to overseas assets.

The optimal porfolio (equity) is going to be 45 US/ 55 non US (or whatever the current market cap weightings of the world indices are).

But that could get you to a lot of currency risk. If currency risk is uncorrelated with equity risk (it's not) then the currency weighting should be completely separate.

All we can say is that, historically, much over 30% international equities has bought a US investor higher volatility for minimal or no higher return. Similarly below 20% has left diversification gains 'on the table'.

Note there are a couple of 'cheats':

- international small value is more diversifying than just international
- a proxy to a worry about USD exposure is TIPS & I Bonds, because to the extent Purchasing Power Parity (PPP) holds, higher US inflation than international inflation will be reflected in a fall in the USD. So if you hold your buying power in USD then currency fluctuation is just less of a worry.

[AlohaJoe]

Your retirement liabilities are in USD. So any currency outside of that exposes you to risk.

What does one do if they don't know what currency their retirement liabilities are going to be in?

[winguy]

How much total currency risk (stocks + bonds) should a Canadian/British/Japanese/etc limit to then? <40%?