Currency risk?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
asif408
Posts: 2613
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Currency risk?

Post by asif408 »

Can someone explain why currency risk is used to justify a lower allocation to foreign investments by US investors? And is this justification common among non-US investors for underweighting US equities? For instance, do Chinese, European, Japanese, or Brazilian investors overweight their home country stocks because of this currency risk? It seems that currency risk is cyclical. For example, when you look at mutual fund performance you see that:

- 1980s, 2000s: currency risk was a positive for foreign stocks, outperformed US stocks
- 1990s, 2010s (current): currency risk was a negative for foreign stocks, US stocks outperformed

I just don't see any evidence that currency risk is a one way street. Maybe things were different pre-1980, so if you have any data on currency fluctuations by country I'd be interested to see. It appears to me that sometimes it helps, sometimes it hurts, but in no case that I am aware of does one country's currency appreciate against other countries' currencies for decades. I assume currencies can go to zero for an individual country, but if you own a basket of countries' stocks, that would seem to me to be the best way to limit currency risk. I hear people also say their expenses will be in US dollars, but if they spend a lot of time traveling in retirement (which seems to be common), won't a decent chunk of expenses be in foreign dollars as well?

Curious to hear from those who use the currency risk argument as a reason to limit or have no foreign stock exposure.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

asif408 wrote: Mon Oct 16, 2017 8:03 am Can someone explain why currency risk is used to justify a lower allocation to foreign investments by US investors? And is this justification common among non-US investors for underweighting US equities? For instance, do Chinese, European, Japanese, or Brazilian investors overweight their home country stocks because of this currency risk? It seems that currency risk is cyclical. For example, when you look at mutual fund performance you see that:

- 1980s, 2000s: currency risk was a positive for foreign stocks, outperformed US stocks
- 1990s, 2010s (current): currency risk was a negative for foreign stocks, US stocks outperformed

I just don't see any evidence that currency risk is a one way street. Maybe things were different pre-1980, so if you have any data on currency fluctuations by country I'd be interested to see. It appears to me that sometimes it helps, sometimes it hurts, but in no case that I am aware of does one country's currency appreciate against other countries' currencies for decades. I assume currencies can go to zero for an individual country, but if you own a basket of countries' stocks, that would seem to me to be the best way to limit currency risk. I hear people also say their expenses will be in US dollars, but if they spend a lot of time traveling in retirement (which seems to be common), won't a decent chunk of expenses be in foreign dollars as well?

Curious to hear from those who use the currency risk argument as a reason to limit or have no foreign stock exposure.
Welcome to the academic phenomenon of "home country bias".

Generally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.

The rational position is therefore no currency risk. 100% of your portfolio hedged back to USD, say. Or to be more specific, one should be 100% in one's home currency *except* to the extent you have a propensity to spend in another currency (travel or imports, say).

In practice, the more volatile the currency pair, the greater the cost of hedging. And the longer the time period to hedge, the greater the cost of hedging. When you get to Emerging Market currencies, or even small country currencies like (?) SGD or NZD or SKR, then hedging can also get expensive-- the markets for those currencies are not liquid enough (big bid-ask spreads).

Equity exposure tends not to be hedged. The reason being in the long run capital market returns equilibrate-- currency adds and then subtracts from performance-- with Purchasing Power Parity (in theory) which prevents long term deviations from The Law of One Price (in theory). Eventually USD 100 in stocks buys about the same amount of goods and services as 100 EUR in stocks (assuming the starting exchange rate is 1:1). So you'd basically be taking on volatility for no gain. And paying a significant cost for it, potentially.

Bond exposure tends to be hedged. Otherwise a foreign bond fund is a play on currency more than it is on fixed income because currency pairs are much more volatile than bond yields.

The ideal investor is globally weighted as to equity markets- -thus gaining maximum diversification as to sectors, exposure to local economies** etc. And hedged 100% back to their home currency. For bonds, the argument for global weighting is much more difficult. Italy for example is the world's 4-5th largest government bond market, and is that a good case for an American investor to hold Italian government bonds to that proportion? Given that Italy has (some) credit risk? Needless to say, there's not complete agreement on that.

It's probably not a big error for a US investor to have zero weight in global equities. Given that c. 55% of equity value by market capitalization is US stocks. One must be wary of Japan, though, which was something like 40% of world markets in 1989, vs. less than 10% now. The gains from diversification into non US equities are present, but small. They would be greater if one invested in a good Small Cap (non US) fund and/or a good Emerging Markets (possibly EM small cap) fund-- DFA at least offers those products.

For non US investors

This is all more complex. From a purely sectoral viewpoint, do we really want to avoid owning Google Apple Amazon Microsoft say? These are 4 of the world's 10 largest companies, and also the 4 largest *tech* companies in the world. It's quite a loss of diversification if I just own the UK index (top 10 stocks include HSBC bank, tobacco companies, Oil & Gas (BP + Shell), pharmaceuticals (GSK + Astra Zeneca) etc.

So we can't safely ignore the US in our allocations (or any other international developed market; and EM are now c. 15% of world index, so we can't simply ignore those either). Our gains from international diversification are correspondingly large.

Assuming we believe our own government bonds to have no credit risk, then we probably can ignore the US Treasury and Japanese Govt Bond markets. In fact, I just bought an (unhedged) global government bond fund, because I believe the UK currency market has not fully priced in 2 possible eventualities in the next 12-18 months: hard Brexit (no deal with the EU) and the election of a Labour Party government (with the most socialist agenda since the 1960s, if not the 1940s). But if those things don't happen, then I am quite wrong ;-).



** oil companies are all exposed to the same basic factor-- the price of oil. It just depends which countries they operate in, and what fields. Whereas retail banking, or food retail, is much more a country-specific business. Tech is of course completely global. Consumer products it depends-- you get the P&Gs Unilever and Nestles of this world, but you also get local and regional brands. Reckitt Benkeiser, for example, used to make more than half of its profits from a heroin substitute used in US drug treatment programmes-- who knew?
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Currency risk?

Post by nisiprius »

asif408 wrote: Mon Oct 16, 2017 8:03 am...I just don't see any evidence that currency risk is a one way street...
Nobody ever said that it was. That is a straw man argument.

The argument is that currency risk a) does not add return, and b) does add risk. Therefore, it reduces the risk-adjusted return of foreign investments. I, as a U.S. investor, experience greater volatility/risk/standard deviation investing in, let's say European stocks than a European investor, because I am exposed to currency risk which a European investor does not need to take. There's no reason to expect me to be rewarded for taking unnecessary risk.

The basic idea that international stock mutual funds have higher "risk" (by most definitions) than U.S. stock funds for a U.S. investor is supported by just about everything, including Vanguard's categorization of the Total International Stock Index Fund in "risk potential 5" and Total [U.S.] Stock Market Index Fund in "risk potential 4;" by any prospectus for any international stock fund, etc. The idea that much of this additional risk is currency risk is also supported by most authorities.

The idea that international stocks as seen by a U.S. investor have lower risk-adjusted return than U.S. stocks is also well-supported. To take one number that happens to be easily available to me, from 1970 through 2015 the Sharpe ratio for the EAFE index was 0.299; for the S&P 500, 0.392.

If we accept that international stocks, as seen by a U.S. investor, have lower risk-adjusted return than U.S. stocks due to currency fluctuations, then they are in themselves a drag on risk-adjusted return. However, they also have a diversification benefit. In order to believe that international stocks will improve a portfolio we have to believe that international stocks will have enough of a diversification effect to overcome the drag.

Now, we get into some math that people have challenged, but I'll state what I believe and I'm pretty sure I'm right. If we believe that a financial quantity, such as currency fluctuation, has zero return, then in order for it to improve a portfolio it is not good enough for it to have low correlation, or even zero correlation with the portfolio. It needs to have negative correlation. I seriously doubt this is the case.

In other words, currency risk is bad because it adds risk/volatility/standard deviation without adding return. And it is good because it reduces correlations between foreign-currency assets and dollar assets. You have a balancing act, and it is not at all obvious which will win.
...The ideal investor is globally weighted as to equity markets...
I don't think this is true. I think that the textbook analyses that show that "the market portfolio is mean-variance optimum" assume a single currency.

If you imagine a Euro investor and a U.S. investor, each with a portfolio consisting only of a) a European stock mutual fund and b) a U.S. stock mutual fund, from the point of view of the European investor, currency fluctuation adds risk, but not return to the U.S. fund. For the U.S. investor, it adds risk, but not return, to the European fund. It does not make sense to me to believe that the optimum allocation for the two investors would be the same.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Topic Author
asif408
Posts: 2613
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Currency risk?

Post by asif408 »

nisiprius wrote: Mon Oct 16, 2017 9:28 am
asif408 wrote: Mon Oct 16, 2017 8:03 am...I just don't see any evidence that currency risk is a one way street...
Nobody ever said that it was. That is a straw man argument.
Nisi, thanks for your input. You always provide a lot to chew on. The reason I say this is that although it is not specifically stated, at least in my logic, an investor avoiding unhedged foreign stocks because of currency risk implies that the individual either doesn't think currencies will benefit at a minimum, or harm their returns significantly at a maximum.
nisiprius wrote: Mon Oct 16, 2017 9:28 amNow, we get into some math that people have challenged, but I'll state what I believe and I'm pretty sure I'm right. If we believe that a financial quantity, such as currency fluctuation, has zero return, then in order for it to improve a portfolio it is not good enough for it to have low correlation, or even zero correlation with the portfolio. It needs to have negative correlation. I seriously doubt this is the case.

In other words, currency risk is bad because it adds risk/volatility/standard deviation without adding return. And it is good because it reduces correlations between foreign-currency assets and dollar assets. You have a balancing act, and it is not at all obvious which will win.
It appears currency risk can possibly add return for a US investor if your investing timeframe is during times of the US dollar falling against other currencies more often than not (e.g., 1986-2006). It could also lower return of a US investor if you invest during periods of mostly US dollar outperformance (e.g., 1996-2016). And 1996 just so happened to be the start of Vanguard's Total International Stock Market fund, with many here citing its 20 year performance as a valid reason to limit or avoid foreign stocks.

As far as I can tell, when the US dollar has fallen against other countries, unhedged foreign stocks have outperformed, and vice versa. It doesn't appear to be anything magical, it just depends on your timeframe, and it appears currency risk is probably never equal to 0 in real investments for individuals, even over long periods. Sometimes you'll lose due to owning foreign investments, and sometimes you'll win. I admit if you took a large group of investors it is probably, on average, close to 0. But it would seem owning a variety of countries' investable stocks throughout the world would prevent you from being the biggest currency risk winner or loser.
nisiprius wrote: Mon Oct 16, 2017 9:28 am The basic idea that international stock mutual funds have higher "risk" (by most definitions) than U.S. stock funds for a U.S. investor is supported by just about everything, including Vanguard's categorization of the Total International Stock Index Fund in "risk potential 5" and Total [U.S.] Stock Market Index Fund in "risk potential 4;" by any prospectus for any international stock fund, etc. The idea that much of this additional risk is currency risk is also supported by most authorities.
So how do you define "risk"?
Topic Author
asif408
Posts: 2613
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Currency risk?

Post by asif408 »

Valuethinker wrote: Mon Oct 16, 2017 8:48 amGenerally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.
Do you have any idea why that is the case? Just looking at some historical performance, it appears to be very time dependent. I would agree that if you averaged in on many investors over very long time periods it is probably close to 0. But their appear to be some fairly long time frames (e.g., 20 years) where that is not true. So the general view of no extra return on average may be true, but individually it is very time dependent, particularly the shorter your time frame.

Valuethinker wrote: Mon Oct 16, 2017 8:48 am It's probably not a big error for a US investor to have zero weight in global equities. Given that c. 55% of equity value by market capitalization is US stocks. One must be wary of Japan, though, which was something like 40% of world markets in 1989, vs. less than 10% now.
So for US only investors, when level of market capitalization do you think would cause investors to look elsewhere? Or do you think investors would simply chase performance and begin moving into international based on past performance? Because if other countries had greater market caps and US had lower, that would imply those stock markets were performing better, similar to the way Japan was in the late 1980s.
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

asif408 wrote: Mon Oct 16, 2017 11:00 am
Valuethinker wrote: Mon Oct 16, 2017 8:48 amGenerally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.
Do you have any idea why that is the case? Just looking at some historical performance, it appears to be very time dependent. I would agree that if you averaged in on many investors over very long time periods it is probably close to 0. But their appear to be some fairly long time frames (e.g., 20 years) where that is not true. So the general view of no extra return on average may be true, but individually it is very time dependent, particularly the shorter your time frame.
Currency risk is bounded by Purchasing Power Parity (PPP). Think of the "Law of One Price". A currency might wander 20% above or below it "correct" price, but the further away a currency drifts from its correct price economic factors will push the price back. i.e., if currency prices fall too low then the country exports get a boost. The more exports there are the more demand for its currency, pushing prices back to the correct level.

Historical there has been a 1% volatility in inflation adjusted FX rates in developed markets. For equities that does not constitute one of the major risk factors.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

Valuethinker wrote: Mon Oct 16, 2017 8:48 am In practice, the more volatile the currency pair, the greater the cost of hedging.
I am going to have to disagree with you here. You hedge by entering into forward contracts and their cost is basically zero. The forward price is driven by interest rate differentials between the 2 currencies. The bid / ask spread might be higher on a more obscure pair.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

One other thought.

In theory one should hold a percentage of foreign stocks to get the maximum mean grievance portfolio. I believe his to be more or less the global market weight for 2 reasons. First, as international markets have increased integration the correlation between domestic and foreign markets have increased. So no reason here to under / overweight. Second, in a free float market capital weighted index most of the companies tend to be multinational companies. They are influenced by global factors, not local ones.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

alex_686 wrote: Mon Oct 16, 2017 11:23 am
Valuethinker wrote: Mon Oct 16, 2017 8:48 am In practice, the more volatile the currency pair, the greater the cost of hedging.
I am going to have to disagree with you here. You hedge by entering into forward contracts and their cost is basically zero. The forward price is driven by interest rate differentials between the 2 currencies. The bid / ask spread might be higher on a more obscure pair.
Ah, OK maybe I haven't properly understood that.

Generally things in finance cost more the more volatile they are. So I was thinking that it is costing the other side more to hedge their position if the currency pair is volatile, therefore they are going to demand a bigger spread for being the opposite side of your trade.

It may be that I have confused that notion with simply that of illiquidity i.e. USD : NZD is more illiquid than GBP : USD say.
protagonist
Posts: 9242
Joined: Sun Dec 26, 2010 11:47 am

Re: Currency risk?

Post by protagonist »

Valuethinker wrote: Mon Oct 16, 2017 8:48 am
asif408 wrote: Mon Oct 16, 2017 8:03 am Can someone explain why currency risk is used to justify a lower allocation to foreign investments by US investors? And is this justification common among non-US investors for underweighting US equities? For instance, do Chinese, European, Japanese, or Brazilian investors overweight their home country stocks because of this currency risk? It seems that currency risk is cyclical. For example, when you look at mutual fund performance you see that:

- 1980s, 2000s: currency risk was a positive for foreign stocks, outperformed US stocks
- 1990s, 2010s (current): currency risk was a negative for foreign stocks, US stocks outperformed

I just don't see any evidence that currency risk is a one way street. Maybe things were different pre-1980, so if you have any data on currency fluctuations by country I'd be interested to see. It appears to me that sometimes it helps, sometimes it hurts, but in no case that I am aware of does one country's currency appreciate against other countries' currencies for decades. I assume currencies can go to zero for an individual country, but if you own a basket of countries' stocks, that would seem to me to be the best way to limit currency risk. I hear people also say their expenses will be in US dollars, but if they spend a lot of time traveling in retirement (which seems to be common), won't a decent chunk of expenses be in foreign dollars as well?

Curious to hear from those who use the currency risk argument as a reason to limit or have no foreign stock exposure.
Welcome to the academic phenomenon of "home country bias".

Generally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.

The rational position is therefore no currency risk. 100% of your portfolio hedged back to USD, say. Or to be more specific, one should be 100% in one's home currency *except* to the extent you have a propensity to spend in another currency (travel or imports, say).

In practice, the more volatile the currency pair, the greater the cost of hedging. And the longer the time period to hedge, the greater the cost of hedging. When you get to Emerging Market currencies, or even small country currencies like (?) SGD or NZD or SKR, then hedging can also get expensive-- the markets for those currencies are not liquid enough (big bid-ask spreads).

Equity exposure tends not to be hedged. The reason being in the long run capital market returns equilibrate-- currency adds and then subtracts from performance-- with Purchasing Power Parity (in theory) which prevents long term deviations from The Law of One Price (in theory). Eventually USD 100 in stocks buys about the same amount of goods and services as 100 EUR in stocks (assuming the starting exchange rate is 1:1). So you'd basically be taking on volatility for no gain. And paying a significant cost for it, potentially.

Bond exposure tends to be hedged. Otherwise a foreign bond fund is a play on currency more than it is on fixed income because currency pairs are much more volatile than bond yields.

The ideal investor is globally weighted as to equity markets- -thus gaining maximum diversification as to sectors, exposure to local economies** etc. And hedged 100% back to their home currency. For bonds, the argument for global weighting is much more difficult. Italy for example is the world's 4-5th largest government bond market, and is that a good case for an American investor to hold Italian government bonds to that proportion? Given that Italy has (some) credit risk? Needless to say, there's not complete agreement on that.

It's probably not a big error for a US investor to have zero weight in global equities. Given that c. 55% of equity value by market capitalization is US stocks. One must be wary of Japan, though, which was something like 40% of world markets in 1989, vs. less than 10% now. The gains from diversification into non US equities are present, but small. They would be greater if one invested in a good Small Cap (non US) fund and/or a good Emerging Markets (possibly EM small cap) fund-- DFA at least offers those products.

For non US investors

This is all more complex. From a purely sectoral viewpoint, do we really want to avoid owning Google Apple Amazon Microsoft say? These are 4 of the world's 10 largest companies, and also the 4 largest *tech* companies in the world. It's quite a loss of diversification if I just own the UK index (top 10 stocks include HSBC bank, tobacco companies, Oil & Gas (BP + Shell), pharmaceuticals (GSK + Astra Zeneca) etc.

So we can't safely ignore the US in our allocations (or any other international developed market; and EM are now c. 15% of world index, so we can't simply ignore those either). Our gains from international diversification are correspondingly large.

Assuming we believe our own government bonds to have no credit risk, then we probably can ignore the US Treasury and Japanese Govt Bond markets. In fact, I just bought an (unhedged) global government bond fund, because I believe the UK currency market has not fully priced in 2 possible eventualities in the next 12-18 months: hard Brexit (no deal with the EU) and the election of a Labour Party government (with the most socialist agenda since the 1960s, if not the 1940s). But if those things don't happen, then I am quite wrong ;-).



** oil companies are all exposed to the same basic factor-- the price of oil. It just depends which countries they operate in, and what fields. Whereas retail banking, or food retail, is much more a country-specific business. Tech is of course completely global. Consumer products it depends-- you get the P&Gs Unilever and Nestles of this world, but you also get local and regional brands. Reckitt Benkeiser, for example, used to make more than half of its profits from a heroin substitute used in US drug treatment programmes-- who knew?
Very good and concise explanation. Thank you.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

alex_686 wrote: Mon Oct 16, 2017 11:27 am One other thought.

In theory one should hold a percentage of foreign stocks to get the maximum mean grievance portfolio.
I am bedevilled by the autocorrect on my pad, so I am guessing you fell prey to the same gremlin?

I am loving the mean grievance portfolio though? The one that hurts you the least ;-) ?
I believe his to be more or less the global market weight for 2 reasons. First, as international markets have increased integration the correlation between domestic and foreign markets have increased. So no reason here to under / overweight. Second, in a free float market capital weighted index most of the companies tend to be multinational companies. They are influenced by global factors, not local ones.
You probably saw the recent piece in the CFA Journal (FAJ) that one could increase the diversification benefit by constructing a portfolio of stocks which were primarily domestic? For example US investors could buy UK housebuilders, which have nearly zero non UK exposure. Interesting idea.

For me it is sectoral. What sticks out like a sore thumb is that if you want technology exposure, you have to be in the US market. Because with a small number of exceptions like TSMC and Samsung, that's where the the world's tech stocks are. That's also true if you want exposure to US retail and domestic banking, say.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

asif408 wrote: Mon Oct 16, 2017 11:00 am
Valuethinker wrote: Mon Oct 16, 2017 8:48 amGenerally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.
Do you have any idea why that is the case? Just looking at some historical performance, it appears to be very time dependent. I would agree that if you averaged in on many investors over very long time periods it is probably close to 0. But their appear to be some fairly long time frames (e.g., 20 years) where that is not true. So the general view of no extra return on average may be true, but individually it is very time dependent, particularly the shorter your time frame.
See Alex's excellent response.

Time dependence is a warning that you can't make money from this. What it means is that things drift away from a PPP mean, and then drift back. For countries that do not have currency controls or capital controls, at least.

China is an interesting case because they do. Market forces want the Renimbi to devalue against USD. Chinese government for trade & prestige reasons badly do not want this to happen. The result is the Renimbi is overvalued (because it is not a free floating currency) and capital is trying to flow out by various channels (Bitcoin!). So the Chinese government is cracking down on people shifting money out of the country.
Valuethinker wrote: Mon Oct 16, 2017 8:48 am It's probably not a big error for a US investor to have zero weight in global equities. Given that c. 55% of equity value by market capitalization is US stocks. One must be wary of Japan, though, which was something like 40% of world markets in 1989, vs. less than 10% now.
So for US only investors, when level of market capitalization do you think would cause investors to look elsewhere? Or do you think investors would simply chase performance and begin moving into international based on past performance? Because if other countries had greater market caps and US had lower, that would imply those stock markets were performing better, similar to the way Japan was in the late 1980s.
[/quote]

I only meant that we might look back in 25 years and say "of course the US was badly overvalued". I don't believe the situation is the same e.g. there's not the same crossholdings etc, nor an apparent property bubble of that scale.

Investors have to make their own call. I agree that they will performance chase-- if other markets start getting bigger again, then that would imply outperformance.

As long as US companies keep their prodigious rate of stock buybacks (retiring equity) I don't see it happening. If they stopped doing that then we shall see.
Topic Author
asif408
Posts: 2613
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Currency risk?

Post by asif408 »

alex_686 wrote: Mon Oct 16, 2017 11:15 am
asif408 wrote: Mon Oct 16, 2017 11:00 am
Valuethinker wrote: Mon Oct 16, 2017 8:48 amGenerally investing in currencies is held to be zero sum. You get increased volatility, for no extra return.
Do you have any idea why that is the case? Just looking at some historical performance, it appears to be very time dependent. I would agree that if you averaged in on many investors over very long time periods it is probably close to 0. But their appear to be some fairly long time frames (e.g., 20 years) where that is not true. So the general view of no extra return on average may be true, but individually it is very time dependent, particularly the shorter your time frame.
Currency risk is bounded by Purchasing Power Parity (PPP). Think of the "Law of One Price". A currency might wander 20% above or below it "correct" price, but the further away a currency drifts from its correct price economic factors will push the price back. i.e., if currency prices fall too low then the country exports get a boost. The more exports there are the more demand for its currency, pushing prices back to the correct level.

Historical there has been a 1% volatility in inflation adjusted FX rates in developed markets. For equities that does not constitute one of the major risk factors.
Thanks, alex, your explanation is very helpful.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

nisiprius wrote: Mon Oct 16, 2017 9:28 am
...The ideal investor is globally weighted as to equity markets...
I don't think this is true. I think that the textbook analyses that show that "the market portfolio is mean-variance optimum" assume a single currency.

If you imagine a Euro investor and a U.S. investor, each with a portfolio consisting only of a) a European stock mutual fund and b) a U.S. stock mutual fund, from the point of view of the European investor, currency fluctuation adds risk, but not return to the U.S. fund. For the U.S. investor, it adds risk, but not return, to the European fund. It does not make sense to me to believe that the optimum allocation for the two investors would be the same.
Hi.

Here's the argument as I understand it:

- investors should seek maximum diversification
- by holding the global equity portfolio they have maximum diversification as to stock risk - all the factors which go into stock performance they are fully diversified over
- however there is then more currency risk. But 100% of currency risk can be hedged away (see Alex above) at fairly low cost

Thus the optimal portfolio is 100% globally diversified with the currency risk hedged. You have diversified away all the unsystematic (diversifiable) risk in your equity portfolio that you can and have only equity risk (which is plenty in and of itself).

The picture changes somewhat if you have unrecoverable taxes on dividends in some foreign markets. That creates a performance drag arising from that. So too do greater bid ask spreads in foreign markets than the US ones.

I would summarize the above as:

- an American investor can hold only the US Total Stock Market, and achieve adequate diversification

- a non US investor cannot afford to only be domestically invested. Even if in the UK (say 10% of world stock markets by value). The best guideline then as to target weightings is global equity diversification by market capitalization.
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

Valuethinker wrote: Mon Oct 16, 2017 11:28 am
alex_686 wrote: Mon Oct 16, 2017 11:23 am
Valuethinker wrote: Mon Oct 16, 2017 8:48 am In practice, the more volatile the currency pair, the greater the cost of hedging.
I am going to have to disagree with you here. You hedge by entering into forward contracts and their cost is basically zero. The forward price is driven by interest rate differentials between the 2 currencies. The bid / ask spread might be higher on a more obscure pair.
Ah, OK maybe I haven't properly understood that.

Generally things in finance cost more the more volatile they are. So I was thinking that it is costing the other side more to hedge their position if the currency pair is volatile, therefore they are going to demand a bigger spread for being the opposite side of your trade.

It may be that I have confused that notion with simply that of illiquidity i.e. USD : NZD is more illiquid than GBP : USD say.
For forward contracts, and hence for exchange traded forward contracts, interest rate yields dominate. Volatility is not one of the variables. See the "No Arbitrage Principle". This is a classic common intuitive mistake than many people stumble into. Volatility is a very important variable in options.
Valuethinker wrote: Mon Oct 16, 2017 11:32 am I am bedevilled by the autocorrect on my pad, so I am guessing you fell prey to the same gremlin?

You probably saw the recent piece in the CFA Journal (FAJ) that one could increase the diversification benefit by constructing a portfolio of stocks which were primarily domestic? For example US investors could buy UK housebuilders, which have nearly zero non UK exposure. Interesting idea.
Yes, darn auto-correct. As for the CFA article I had reached that conclusion long ago. However I did find the piece insightful - another missing piece of the puzzle, another angle to be considered.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Currency risk?

Post by nisiprius »

Valuethinker wrote: Mon Oct 16, 2017 8:48 am...It's probably not a big error for a US investor to have zero weight in global equities...
Now, that is the key point, I think. John C. Bogle has written that "Successful investing involves doing just a few things right and avoiding serious mistakes." Nobody has ever been able to convince me that anything in the entire range of 0% international up to full cap-weighted international is a "serious" mistake.

All of the charts I've ever seen of risk, or risk-adjusted return, as a function of percentage international are extremely shallow and extremely broad. In the charts I've seen, the minimum volatility... or the "tangent portfolio" on the "efficient frontier..." have been noticeably below full cap-weight. But the data is extremely dependent on the range of years chosen, and recency effects are ten times stronger than anything that looks robust or persistent.

Another weirdness is that the widely-recommended percentages for international have done nothing but go up, up, up and there is very little obvious to explain why this should be so--why Burton Malkiel should have recommended that 16% of stocks be international in 1990 and 50% e international nowadays. Basically I let myself get pushed by the consensus wisdom up to about 20% or 25% international, and then my sales resistance kicked in and I said "to heck with it, I'm stopping here."
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Currency risk?

Post by lack_ey »

nisiprius wrote: Mon Oct 16, 2017 1:39 pm ...
Another weirdness is that the widely-recommended percentages for international have done nothing but go up, up, up and there is very little obvious to explain why this should be so--why Burton Malkiel should have recommended that 16% of stocks be international in 1990 and 50% e international nowadays. Basically I let myself get pushed by the consensus wisdom up to about 20% or 25% international, and then my sales resistance kicked in and I said "to heck with it, I'm stopping here."
Some people quote costs decreasing as the reason, not just mutual fund fees but underlying trading costs.

But I think it has significantly to do with the the speed and mechanism by which conventional wisdom changes. There's also the natural tendency for many to anchor on to what they already have, and it furthermore makes sense for academic writers and practitioners to think from the perspective of the audience of the time. It takes some nudging or shoving to change the status quo. When most investors had no international, it was a stretch to convince people to own any. Now that it's accepted and the case that many investors have some, it's now a stretch to convince to convince them to increase the allocation. Many of the prior investors who never had international are now pushing the daisies, so it's not even the same group of people as before.
rnitz
Posts: 282
Joined: Thu May 10, 2007 12:47 pm
Location: San Diego

Re: Currency risk?

Post by rnitz »

I think one of the reasons for differences in opinions on international and currency exposure is their base frame of reference. What do I mean by this - those that see exposure to other currency variation as of no value, and only negative added volatility are measuring their situation based on a view that their currency ( the dollar) is rock solid and all measurements can be absolutely measured against this.

I think this is false. If the dollar goes into the sh*tter US investors will be impacted negatively, as imports will cost more, inflation will increase, even non-imported products will be impacted. Yes, some companies will benefit from better exports, but the US consumer will suffer. Dollar only investors and consumers ALREADY HAVE CURRENCY RISK. They just don't think that they do because they are measuring with a moving measuring stick. I think there is an argument that exposure to other currencies is diversification, rather than dead weight loss of additional volatility.

To a man on a ship, the horizon is adding volatility.
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

rnitz wrote: Mon Oct 16, 2017 8:53 pm I think one of the reasons for differences in opinions on international and currency exposure is their base frame of reference. What do I mean by this - those that see exposure to other currency variation as of no value, and only negative added volatility are measuring their situation based on a view that their currency ( the dollar) is rock solid and all measurements can be absolutely measured against this.

I think this is false. If the dollar goes into the sh*tter US investors will be impacted negatively, as imports will cost more, inflation will increase, even non-imported products will be impacted. Yes, some companies will benefit from better exports, but the US consumer will suffer. Dollar only investors and consumers ALREADY HAVE CURRENCY RISK. They just don't think that they do because they are measuring with a moving measuring stick. I think there is an argument that exposure to other currencies is diversification, rather than dead weight loss of additional volatility.

To a man on a ship, the horizon is adding volatility.
From an economic point of view risk = risk that your financial assets will not meet your future consumption requirements

So from an economic point of view you should hold enough foreign currency assets to meet future consumption needs that have to be paid for in foreign currency.

This gets complicated:

- companies typically hedge their transactional exposure to FX at least within the year. If they know they have a big import bill to pay they lock in the forward rate. Same if they have a big inflow in a foreign currency

- they also tend to hedge their balance sheet exposure. A UK company with large assets in USA will also borrow in USD so a change in XR does not change their balance sheet gearing

Thus the degree of hedging in our stock portfolios is not clear. Apple is a huge USD company, but it also makes a lot of money outside the USA. Exxon gets almost all of its revenues worldwide in USD, by contrast, because oil is priced in USD. A UK commodity producer is basically a play on the USD for that reason-- Shell, BP, Rio Tinto, BHP Billiton etc.

In terms of our consumption:

- our own direct propensity to consume imports can be estimated-- sort of. We can probably figure out what percentage of your BMW is American made (parts + assembly in Atlanta). Mine? Even harder-- the Mini is made in Oxford but with imported parts (also a BMW car). X3? Made in Germany.

Travel is a fairly easy one and one of the best examples of direct exposure. The -14% drop in sterling against the Euro post Brexit vote has hurt ;-)

- the economy's own propensity to consume imports is also available, as a crude measure

- falls in the USD will also be reflected in higher CPI inflation so things like SS will cover that

BUT

The US generally has very low exposure to foreign imports and exports for a developed country:

- it's a huge economy and most of what you consume is not internationally traded -- your housing services (own/ rent) + healthcare + the domestic cost of products even ones that are imported (labour, shipping, retailer markup etc.)

- commodities are priced in USD. There is some negative correlation (dollar falls, commodity prices rise) but it's not the main factor. Thus you are naturally hedged i.e. you consume in the currency in which these things are priced

- much of that "foreign" trade is with Canada and Mexico. I realize that NAFTA is under threat, but just at the moment Canada: 1). runs a trade surplus with the USA but only if you include energy exports to USA (oil, gas, electricity) 2). second largest export item is auto & parts exports to USA but it is also the 2nd largest import from USA in other words there is perfect integration between production in the 2 countries (and w Mexico) 3). Canadians travel a lot and the biggest destination is far and away the USA. Similar issues pertain with Mexico- -economically this is not three countries but one country + 2 politically independent territories

Since this is all hard to estimate the best one can fall back on is historic portfolio performance.

The minimum volatility/ maximum return "efficient frontier" portfolio has tended to have 20-40% foreign equities, if you are a US based investor, based on what Vanguard has published.

Conversely if you are based in Canada, Australia, UK etc. then there's really no case for anything other than a globally weighted portfolio by market cap for equities. Otherwise you have too much concentration on the risk of your own stock market. And a significant underweighting in the world's most successful companies of recent years (i.e. certain technology companies).

Tax is the only important exception to the above for equity portfolios as well as other portfolio management costs.

The case for bonds invested internationally is less clear but a currency hedged global government bond fund diversifies across credit risk and real interest rate cycles- -there's some gain in that. I would argue as a US investor (if I was one) that US Treasury bonds are safe and every other government bond (bar a handful) is probably less safe so it's not something I need (and bank CDs may be a higher yielding alternative, if within FDIC limits). As a UK investor I tend to hold gilts (UK govt bonds) and that has hurt quite a bit.
Last edited by Valuethinker on Tue Oct 17, 2017 7:50 am, edited 1 time in total.
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

alex_686 wrote: Mon Oct 16, 2017 11:15 am Historical there has been a 1% volatility in inflation adjusted FX rates in developed markets. For equities that does not constitute one of the major risk factors.
Where did you find that 1% volatility in currency?

I found previously in this thread (viewtopic.php?f=10&t=222809&start=50#p3440628) that the standard deviation of the annual percentage change in the dollar index is nearly 9%.

However, that doesn't mean that the international stock fund's volatility is 9% higher than it otherwise would be. Assuming that international stocks have the same intrinsic equity or market beta based return and standard deviation when averaged over many decades as domestic, and assuming that currency risk can be modeled as having zero mean and zero correlation to return, then currency risk and international stock risk add in quadrature, such that total international stock standard deviation is SQRT(SD_international_market_beta^2 + SD_Currency^2). This assumes I did the math right in addition to assuming the assumptions I listed. Since market standard deviation is probably 16-20%, the fact that the risks add in quadrature means that currency risk is meaningful but not enormous for equities. (For international bonds, it is a large component of overall risk, which is why international bond funds often hedge)

I discussed later in that thread a more complicated model (viewtopic.php?f=10&t=222809&start=50#p3442332) that further attempts to decompose currency risk, international market risk, domestic market risk, and the correlation of international market returns to domestic market returns. I feel that the ~0.66 correlation between domestic and international provides substantial benefits in theory and makes it such that 20-50% international is probably most reasonable, in my opinion.

rnitz wrote: Mon Oct 16, 2017 8:53 pm I think one of the reasons for differences in opinions on international and currency exposure is their base frame of reference. What do I mean by this - those that see exposure to other currency variation as of no value, and only negative added volatility are measuring their situation based on a view that their currency ( the dollar) is rock solid and all measurements can be absolutely measured against this.

I think this is false. If the dollar goes into the sh*tter US investors will be impacted negatively, as imports will cost more, inflation will increase, even non-imported products will be impacted. Yes, some companies will benefit from better exports, but the US consumer will suffer. Dollar only investors and consumers ALREADY HAVE CURRENCY RISK. They just don't think that they do because they are measuring with a moving measuring stick. I think there is an argument that exposure to other currencies is diversification, rather than dead weight loss of additional volatility.

To a man on a ship, the horizon is adding volatility.
I agree that there is also a "currency diversification" aspect. However, could it be better to separate out currency risk in "normal" times (which would be expected to be zero-mean and thus bad) and economic catastrophe risk where the home country has hyperinflation or a massive black swan decline in its currency or economy? Having some assets in other economies could provide the benefit of this protection.



Also, amusingly, Morningstar allows one to plot returns of Vanguard Total World in SDRs! Maybe that's a better "neutral" reference!
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

nisiprius wrote: Mon Oct 16, 2017 1:39 pm
Valuethinker wrote: Mon Oct 16, 2017 8:48 am...It's probably not a big error for a US investor to have zero weight in global equities...
Now, that is the key point, I think. John C. Bogle has written that "Successful investing involves doing just a few things right and avoiding serious mistakes." Nobody has ever been able to convince me that anything in the entire range of 0% international up to full cap-weighted international is a "serious" mistake.

All of the charts I've ever seen of risk, or risk-adjusted return, as a function of percentage international are extremely shallow and extremely broad. In the charts I've seen, the minimum volatility... or the "tangent portfolio" on the "efficient frontier..." have been noticeably below full cap-weight. But the data is extremely dependent on the range of years chosen, and recency effects are ten times stronger than anything that looks robust or persistent.
I respectfully feel that anything below about 20% international is a serious mistake. Not as serious as living above one's means, market timing, or high fee active funds, but still bad. I have seen the efficient frontier charts, and I recognize that if you zoom out and plot both axes from zero, the formerly big squiggle looks tiny. However, several things aren't taken into account in this backtesting:

First, the backtesting relies often on data going back to the 1870s. Yes, historical data shows long-term outperfomance of US equities. However, this was during a time of transformation of the US from a emerging/frontier markets to the large power it is today. The US won two world wars during this time. Britan and France were colonial empires a hundred years ago, were seriously damaged in both world wars, and now have far fewer colonies. Germany and Japan lost two world wars. Unless one assumes that this history will exactly repeat itself, I think that it is reasonable to chalk differences seen over long periods in domestic vs international returns to historical accidents - and assume that going forward, "stocks are stocks" no mater where they're from. Thus, one would need to project the same returns for domestic vs international stocks going forward instead of the lower international returns seen in backtesting.

Second, those with zero percent international do not fully take into account the "Japan scenario". In the 1980s, Japan (an advance, diversified economy) experienced a massive stock crash that it has not fully recovered from. Losses to Japanese investors could have been mitigated with international allocations. While it is doubtful that this sort of scenario will happen to any single country, and particularly to the US, if it's happened before to a large developed nation, there's no garuntee it won't happen again. For protection from this, I feel at least a 20% international allocation is needed, and a 0% international allocation is a serious mistake.

Third, this analysis doesn't take into account the equity portion of human capital. If human capital has a bond-like component and an equity-like component, and if the equity-like component is taken to be domestic equity, then the investor could be better off with a higher international allocation within investments because of this.

Fourth, I question the motives of those believing in a 0% international allocation (or 5% or 10%). While many are simply ignorant, and others cite reasons such as currency risk for their home country bias, for a few, is their bias ultimately... bias, a behavioral error? Are some inherently biased against stocks from other countries because of personal beliefs unrelated to investment theory?
nisiprius wrote: Mon Oct 16, 2017 1:39 pm Another weirdness is that the widely-recommended percentages for international have done nothing but go up, up, up and there is very little obvious to explain why this should be so--why Burton Malkiel should have recommended that 16% of stocks be international in 1990 and 50% e international nowadays. Basically I let myself get pushed by the consensus wisdom up to about 20% or 25% international, and then my sales resistance kicked in and I said "to heck with it, I'm stopping here."
I feel that this change is largely due to international costs falling. The logical conclusion to this progression of "conventional wisdom" allocations going up and up is global cap weight - or at a minimum 30-40% international, where many MPT-based analyses suggest. Recall that often the domestic vs international stock allocation is done in isolation, not taking into account anything but stocks. Many investors have all of their bonds and human capital in domestic. If one looses one's domestic job (and thus human capital) in a domestic downturn, international investments may be helpful. I feel that in my opinion, Vanguard Total world is underappreciated, and could be reasonable for an investor who wants cap weight in international stocks, or wants near cap weight and wishes to construct a Total World + Total US Stock (or Total World + US Small Value, or Total World + US REIT, etc) portfolio.
Topic Author
asif408
Posts: 2613
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Currency risk?

Post by asif408 »

TD2626 wrote: Tue Oct 17, 2017 7:52 amFourth, I question the motives of those believing in a 0% international allocation (or 5% or 10%). While many are simply ignorant, and others cite reasons such as currency risk for their home country bias, for a few, is their bias ultimately... bias, a behavioral error? Are some inherently biased against stocks from other countries because of personal beliefs unrelated to investment theory?
It appears at least some of this is due to simple performance chasing. I saw a survey recently (can't find it right now, but when I do, will post) that showed ownership of international stocks among US investors in 2008 and 2015. Although they were both low overall (in the 10-30% range), the amount held in international overall had fallen from 2008 to 2015, in line with the falling performance, and the biggest decrease was among younger investors. To me, this showed that performance chasing is still alive and well, and appears to be a behavioral error that gets repeated with regularity over the generations.
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

TD2626 wrote: Tue Oct 17, 2017 7:13 am
alex_686 wrote: Mon Oct 16, 2017 11:15 am Historical there has been a 1% volatility in inflation adjusted FX rates in developed markets. For equities that does not constitute one of the major risk factors.
Where did you find that 1% volatility in currency?

I found previously in this thread (viewtopic.php?f=10&t=222809&start=50#p3440628) that the standard deviation of the annual percentage change in the dollar index is nearly 9%.
Financial Market History: Reflections on the Past for Investors Today, edited by David Chambers, forward by John Bogle, a bit of a brick, the Kindle version is free.

The 9% analysis is problematic. It is for one currency over a selected time range which includes a very atypical political event. So a certain amount of cherry picking. You want to use a time period of greater than 10 years and you want to weigh your basket by asset weight. And if you noticed I said "inflation adjusted FX". Factor that in to get the real change in FX rates and even that 9% falls.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
Northern Flicker
Posts: 15289
Joined: Fri Apr 10, 2015 12:29 am

Re: Currency risk?

Post by Northern Flicker »

I just don't see any evidence that currency risk is a one way street.
If your liabilities (living expenses) are normally denominated in your home currency, then currency changes in your assets and liabilities move up or down in lockstep when assets are also denominated in home currency.

Currency risk arises when assets and liabilities are denominated differently. In that case, it is a 2-way street that can go either way. Because it can go the wrong way, it is a risk.

Multinational companies normally have some direct currency exposure in their revenue stream.

If your home country has a net trade deficit, then one might argue that one’s liabilities have some indirect exposure to other currencies. In this case, the risk is 2-way and a big drop in home currency would trigger or be synonymous with high inflation.
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Currency risk?

Post by nisiprius »

TD2626 wrote: Tue Oct 17, 2017 7:52 am...Fourth, I question the motives of those believing in a 0% international allocation (or 5% or 10%). While many are simply ignorant, and others cite reasons such as currency risk for their home country bias, for a few, is their bias ultimately... bias, a behavioral error? Are some inherently biased against stocks from other countries because of personal beliefs unrelated to investment theory?...
I think it is very likely. And I think it is also likely some advocating full cap weight are factoring in a personal bias against stocks from the United States. Nothing too surprising here, if you believe that your personal macroeconomics predictions, founded partially on ideology, are actionable information. If you believe in United States exceptionalism you might throw more weight on the U.S. side, if you believe the United States is going to hell in a handbasket you might throw more weight on the international side.

Warren Buffett, who always seems to be uttering ambiguously phrased, almost Delphic statements, has said
For 240 years it’s been a terrible mistake to bet against America, and now is no time to start.
Is that the same thing as a recommendation to be 100% U.S.? Hard to tell. Speaking of ambiguity, his next sentence was "America’s golden goose of commerce and innovation will continue to lay more and larger eggs," which... he couldn't have meant... no... surely not...
Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
aj76er
Posts: 1178
Joined: Tue Dec 01, 2015 10:34 pm
Location: Austin, TX

Re: Currency risk?

Post by aj76er »

Some interesting back-testing of the effects of holding international can now be done on www.portfoliocharts.com

Take a typical allocation (40/20/35/5 of domestic stocks/foreign stocks/bonds/cash) and then look at the safe withdrawal rates across different countries and currencies. The addition of foreign helps boost this in nearly all cases. Of course, there are many factors at play here, but 20%-40% or even cap weighted doesn't seem like a bad bet to make.
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
3funder
Posts: 1814
Joined: Sun Oct 15, 2017 9:35 pm

Re: Currency risk?

Post by 3funder »

In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Global stocks, US bonds, and time.
Northern Flicker
Posts: 15289
Joined: Fri Apr 10, 2015 12:29 am

Re: Currency risk?

Post by Northern Flicker »

3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
3funder
Posts: 1814
Joined: Sun Oct 15, 2017 9:35 pm

Re: Currency risk?

Post by 3funder »

jalbert wrote: Tue Oct 17, 2017 3:33 pm
3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
Perhaps I'll feel differently at their age; never say never!
Global stocks, US bonds, and time.
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

jalbert wrote: Tue Oct 17, 2017 3:33 pm
3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
Does this suggest a declining international allocation as age increases? I'm not suggesting that, but I see how someone could make a case for something like that - having high international when younger (when there's enough time to ride out currency fluctuations and when much of the portfolio is in (domestic) human capital) and having lower international when older. I wonder if there's any rigorous glide path that could be used, though?
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

TD2626 wrote: Thu Oct 19, 2017 9:09 am
jalbert wrote: Tue Oct 17, 2017 3:33 pm
3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
Does this suggest a declining international allocation as age increases? I'm not suggesting that, but I see how someone could make a case for something like that - having high international when younger (when there's enough time to ride out currency fluctuations and when much of the portfolio is in (domestic) human capital) and having lower international when older. I wonder if there's any rigorous glide path that could be used, though?
The argument is false. Holding stocks - foreign or domestic - is risky. Stock are for the long term. Currency risk is one of those rare bounded mean reverting factors out there. If you hold stocks for the long term then currency risk fades into background noise.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

alex_686 wrote: Thu Oct 19, 2017 9:15 am
TD2626 wrote: Thu Oct 19, 2017 9:09 am
jalbert wrote: Tue Oct 17, 2017 3:33 pm
3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
Does this suggest a declining international allocation as age increases? I'm not suggesting that, but I see how someone could make a case for something like that - having high international when younger (when there's enough time to ride out currency fluctuations and when much of the portfolio is in (domestic) human capital) and having lower international when older. I wonder if there's any rigorous glide path that could be used, though?
The argument is false. Holding stocks - foreign or domestic - is risky. Stock are for the long term. Currency risk is one of those rare bounded mean reverting factors out there. If you hold stocks for the long term then currency risk fades into background noise.
Good point - since the stock allocation is by definition a long-term allocation, then by definition the investor would have the time to ride out currency fluctuations.

Just trying to explain why it seems as though older investors are the ones who seem to underweight international investing in many cases.
User avatar
aj76er
Posts: 1178
Joined: Tue Dec 01, 2015 10:34 pm
Location: Austin, TX

Re: Currency risk?

Post by aj76er »

TD2626 wrote: Thu Oct 19, 2017 9:29 am
alex_686 wrote: Thu Oct 19, 2017 9:15 am
TD2626 wrote: Thu Oct 19, 2017 9:09 am
jalbert wrote: Tue Oct 17, 2017 3:33 pm
3funder wrote: Tue Oct 17, 2017 3:15 pm In the long run, I see the currency risk as small enough for me to consider it a wash. My stock allocation is 60/40 US/International, and it doesn't bother me one bit. Just my two cents.
Those close to or in retirement are much more sensitive to short-term currency movements than those with long investment horizons.
Does this suggest a declining international allocation as age increases? I'm not suggesting that, but I see how someone could make a case for something like that - having high international when younger (when there's enough time to ride out currency fluctuations and when much of the portfolio is in (domestic) human capital) and having lower international when older. I wonder if there's any rigorous glide path that could be used, though?
The argument is false. Holding stocks - foreign or domestic - is risky. Stock are for the long term. Currency risk is one of those rare bounded mean reverting factors out there. If you hold stocks for the long term then currency risk fades into background noise.
Good point - since the stock allocation is by definition a long-term allocation, then by definition the investor would have the time to ride out currency fluctuations.

Just trying to explain why it seems as though older investors are the ones who seem to underweight international investing in many cases.
Some older, retired investors could chime in here; but I do think that the currency fluctuations adds an extra risk to the portfolio. Over long time horizons it should cancel out, but in the short-term it can magnify variation in the portfolio (e.g. stddev). There is a reason why the Vanguard website lists VTSAX as a 4 and VTIAX as a 5 on a risk scale (from 1-5).
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
alex_686
Posts: 13286
Joined: Mon Feb 09, 2015 1:39 pm

Re: Currency risk?

Post by alex_686 »

aj76er wrote: Thu Oct 19, 2017 1:30 pm There is a reason why the Vanguard website lists VTSAX as a 4 and VTIAX as a 5 on a risk scale (from 1-5).
Yeah, but is it currency risk? If so, how important? The VTIAX skews small cap, value, and emerging market - all of these factors carries a higher risk. Is currency the thing that shifts it from a 4 to a 5?
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

alex_686 wrote: Thu Oct 19, 2017 1:44 pm
aj76er wrote: Thu Oct 19, 2017 1:30 pm There is a reason why the Vanguard website lists VTSAX as a 4 and VTIAX as a 5 on a risk scale (from 1-5).
Yeah, but is it currency risk? If so, how important? The VTIAX skews small cap, value, and emerging market - all of these factors carries a higher risk. Is currency the thing that shifts it from a 4 to a 5?
Emerging markets have more political risk than developed markets. But is political risk compensated?

(Some risks are compensated and some are uncompensated - single stock risks and currency risks are uncompensated because it's risk with no expected reward. Stock market risk and interest rate risk is expected (but by no means guaranteed) to be rewarded by higher returns over the long run)
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Currency risk?

Post by lack_ey »

TD2626 wrote: Thu Oct 19, 2017 2:00 pm Emerging markets have more political risk than developed markets. But is political risk compensated?

(Some risks are compensated and some are uncompensated - single stock risks and currency risks are uncompensated because it's risk with no expected reward. Stock market risk and interest rate risk is expected (but by no means guaranteed) to be rewarded by higher returns over the long run)
Empirically you can't really say based on limited returns history, but to the extent that there's sufficient pricing influence from global asset allocators looking at this and other things, then sure, it should be. If you generally believe in efficient markets then yes, that should influence the discount rate priced in.

Given the high degree of investor home-country bias across countries (some are at a level that seems rational; many seem fairly conclusively not) and some other considerations, I'm not generally that convinced that the market is particularly amazing at long-term assessments and large-scale macro issues. There's also less data on that. There's a lot of data on the market being able to figure out things like Ford vs. GM, and plenty of analysts and traders working on that kind of information, thinking about the pathway by which information gets conveyed into pricing.

So I'm not too confident saying that the market is correctly pricing in political and other kinds of considerations into emerging markets. But these are very well known and priced in somehow to some degree. Looking forward it should be compensated, to use the word. Just maybe not enough? Or perhaps even too much (though the evidence has not been in this direction so far). Hard to say.
User avatar
TD2626
Posts: 637
Joined: Thu Mar 16, 2017 3:40 pm

Re: Currency risk?

Post by TD2626 »

lack_ey wrote: Thu Oct 19, 2017 2:25 pm
TD2626 wrote: Thu Oct 19, 2017 2:00 pm Emerging markets have more political risk than developed markets. But is political risk compensated?

(Some risks are compensated and some are uncompensated - single stock risks and currency risks are uncompensated because it's risk with no expected reward. Stock market risk and interest rate risk is expected (but by no means guaranteed) to be rewarded by higher returns over the long run)
Empirically you can't really say based on limited returns history, but to the extent that there's sufficient pricing influence from global asset allocators looking at this and other things, then sure, it should be. If you generally believe in efficient markets then yes, that should influence the discount rate priced in.

Given the high degree of investor home-country bias across countries (some are at a level that seems rational; many seem fairly conclusively not) and some other considerations, I'm not generally that convinced that the market is particularly amazing at long-term assessments and large-scale macro issues. There's also less data on that. There's a lot of data on the market being able to figure out things like Ford vs. GM, and plenty of analysts and traders working on that kind of information, thinking about the pathway by which information gets conveyed into pricing.

So I'm not too confident saying that the market is correctly pricing in political and other kinds of considerations into emerging markets. But these are very well known and priced in somehow to some degree. Looking forward it should be compensated, to use the word. Just maybe not enough? Or perhaps even too much (though the evidence has not been in this direction so far). Hard to say.
In my own personal feeling, with international there are higher risks due to higher political risks, and due to currency risks. This is why Total International is rated by vanguard as higher-risk than Total US. However, I feel that political risks are partially, if not mostly, compensated - or at least in theory should be, over the very long term and assuming efficient markets. Currency risk, on the other hand, is not expected to be compensated.

One thing to note is that with broad international funds like Vanguard Total International, a lot of the investments are in countries/regions like Canada, Western Europe, Japan, or Australia. These areas are generally stable and markets are highly regulated - investors don't generally have concerns about fraud, confiscation, millitary coups, etc to the same extent that this is a major concern in emerging markets. Political risk is always there, and there are never any guarantees. However, with Total International, my opinion is that currency risk is more of a concern than political risk.
columbia
Posts: 3023
Joined: Tue Aug 27, 2013 5:30 am

Re: Currency risk?

Post by columbia »

I know that there are 50% hedged international ETFs, which could be a reasonable compromise (depending on one's outlook). I don't believe I've ever seen such a product discussed here, however.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Currency risk?

Post by lack_ey »

columbia wrote: Thu Oct 19, 2017 4:30 pm I know that there are 50% hedged international ETFs, which could be a reasonable compromise (depending on one's outlook). I don't believe I've ever seen such a product discussed here, however.
Half hedged is not really that useful, given that you can just use two different funds, one hedged and one unhedged, which is probably cheaper.

A little different would be funds that are variably hedged based on currency trading signals (e.g. carry, momentum, value/PPP equalization), though that has been discussed a little bit.
User avatar
aj76er
Posts: 1178
Joined: Tue Dec 01, 2015 10:34 pm
Location: Austin, TX

Re: Currency risk?

Post by aj76er »

As an interesting side note, currency risk is even at play when holding a US total market fund, like VTSAX, to some degree. This is because of the large amount of outsourcing done many US companies; especially in the tech sector (which is a large part of the US economy).

At my own company (a mid-cap tech sector company), we have sizeable R&D centers in other international developed countries (e.g.France, Germany). Our quarterly revenue can be impacted significantly by the USD exchange rate relative to these other currencies, as we pay developers in their native currency. I don't know the exact percentage of the impact, but it can be significant, as I was told by managers familiar with the balance sheets.
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
Valuethinker
Posts: 48954
Joined: Fri May 11, 2007 11:07 am

Re: Currency risk?

Post by Valuethinker »

aj76er wrote: Thu Oct 19, 2017 7:29 pm As an interesting side note, currency risk is even at play when holding a US total market fund, like VTSAX, to some degree. This is because of the large amount of outsourcing done many US companies; especially in the tech sector (which is a large part of the US economy).

At my own company (a mid-cap tech sector company), we have sizeable R&D centers in other international developed countries (e.g.France, Germany). Our quarterly revenue can be impacted significantly by the USD exchange rate relative to these other currencies, as we pay developers in their native currency. I don't know the exact percentage of the impact, but it can be significant, as I was told by managers familiar with the balance sheets.
Usually what you do to hedge your net asset position ( = Assets - Liabilities) is to borrow against your asset value in that other country.

If you have 100m of assets (in EUR at the current exchange rate) in Germany, say, and you borrow 100m EUR against that, movements up and down in the EUR don't change your balance sheet net assets.

You still have transactional exposure though. If you have a cost in one country, and a revenue in a different currency, then movements up in the sale currency increase your profit margin, and down, decrease it.

What exporting companies tend to do is when they have a signed order (say Airbus selling jets to a US airline) then they hedge the revenue amount back to the cost currency. Same thing on the inputs side-- commodities are usually priced in USD so if I am a German manufacturer and I order aluminium, I will hedge the EUR/ USD rate using a forward contract timed to expire/ deliver just when I have to pay for the metal.

However it's much harder to do that for long term strategic things where you don't know the order size and/or timing.

So if I am Airbus and I am planning to sell more into USD countries, I will move some of my cost base into a USD country- -Airbus assembling in Alabama for example. The Japanese car manufacturers setting up transplants in the USA then over time sourcing parts there, too. That was also partly about tariffs on imported cars, which are 5% under WTO rules.

In the case of your company where it is a cost centre in EUR, unless you have significant profits in the Eurozone, earned in EUR, it's hard to hedge. Or rather, you can hedge the balance sheet exposure fairly easily (see above) but the Profit & Loss exposure is harder. You have a fundamental mismatch between where your costs are and where your revenues are-- thus profit swings. And to the extent that you earn profits in EUR, it's hard (but not impossible) to hedge that exposure back into USD.

There's definitely a difference between how the market reacts to a significant change in the US dollar level for an exporter (Boeing) v. someone with big overseas operations (JP Morgan say) where profits vary because of changes in the translation of exchange rates. The impact on the stock price tends to be much less for the latter.

I believe the accepted estimate is that around 40% of revenues of S&P500 companies are earned outside the USA. For the FTSE 100 (top 100 stocks in UK index or about 84% of those by market value) it is 60-70%.
User avatar
aj76er
Posts: 1178
Joined: Tue Dec 01, 2015 10:34 pm
Location: Austin, TX

Re: Currency risk?

Post by aj76er »

alex_686 wrote: Thu Oct 19, 2017 1:44 pm
aj76er wrote: Thu Oct 19, 2017 1:30 pm There is a reason why the Vanguard website lists VTSAX as a 4 and VTIAX as a 5 on a risk scale (from 1-5).
Yeah, but is it currency risk? If so, how important? The VTIAX skews small cap, value, and emerging market - all of these factors carries a higher risk. Is currency the thing that shifts it from a 4 to a 5?
Political risk could be an additional factor, but probably only significant in emerging markets. On the other hand, I think that currency risk exists throughout all of intl (relative to US), and thus it is probably more of a dominant effect (with emerging market currencies perhaps contributing even more variability).

Holding broad market cap-weighted funds should theoretically contain only the following risks:

* Market risk (e.g. Beta)
* Political risk
* Currency risk

Of these three, only intl stocks carry currency risk. And it could be argued that emerging markets have increased political risk.
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle
User avatar
Noobvestor
Posts: 5944
Joined: Mon Aug 23, 2010 1:09 am

Re: Currency risk?

Post by Noobvestor »

nisiprius wrote: Mon Oct 16, 2017 1:39 pm
Valuethinker wrote: Mon Oct 16, 2017 8:48 am...It's probably not a big error for a US investor to have zero weight in global equities...
Now, that is the key point, I think. John C. Bogle has written that "Successful investing involves doing just a few things right and avoiding serious mistakes." Nobody has ever been able to convince me that anything in the entire range of 0% international up to full cap-weighted international is a "serious" mistake.
Japan. Developed nation, big market, the future (TM). Then 30 years of stagnation. Surely the US isn't immune to the possibility.
rnitz wrote: Mon Oct 16, 2017 8:53 pmTo a man on a ship, the horizon is adding volatility.
+1
"In the absence of clarity, diversification is the only logical strategy" -= Larry Swedroe
Post Reply