World Market Cap Weighting

[harry]

Charles Ellis known for his book Winning the Loser's Game and as a former chair of Yale's Investment Committee claims that long term investors, including US investors, should not overweight there own country's stock.

In a recent article in Barons Ellis made the following point: The best risk-adjusted returns over the long term can be scored by matching the market capitalization weightings of the world's markets, Ellis says. That would mean putting 45% in domestic stocks, 47% in developed foreign markets and 8% in developing foreign markets.

The idea is to have no bets on whether one market or another will be stronger going forward.

Any thoughts?

[Valuethinker]

Charles Ellis known for his book Winning the Loser's Game and as a former chair of Yale's Investment Committee claims that long term investors, including US investors, should not overweight there own country's stock.

In a recent article in Barons Ellis made the following point: The best risk-adjusted returns over the long term can be scored by matching the market capitalization weightings of the world's markets, Ellis says. That would mean putting 45% in domestic stocks, 47% in developed foreign markets and 8% in developing foreign markets.

The idea is to have no bets on whether one market or another will be stronger going forward.

Any thoughts?

Currency vs. equity exposure.

Japan.

The problem with unehdged currency exposure is that you are taking on a volatility which is greater than stock market volatility, and may not work your way.

A hedged global equity fund pays the price of hedging (which is not trivial) but would be superior from a diversification point of view.

The other problem is Japan: 30% of the world stock market in 1990 v. sub 10% now i believe?

You could get caught in that sort of valuation trap.

Also international stock markets are increasingly correlated with the US-- this is empirically observable. So diversification benefits have been decreasing since the 1980s, steadily. In a crash, correlations move to 1.

If your currency is unhedged, the best guess, based on past performance, is that maximum diversification benefits occur between 20 and 30% of total portfolio value.

[harry]

Valuethinker


Please explain unehdged currency exposure.

[stratton]

Valuethinker


Please explain unehdged currency exposure.

Take a look at this 2 year chart on Yahoo of AGG, PFUIX and PFORX

AGG: Total Bond Market proxy.
PFUIX: PIMCO Foreign Bond (Unhedged) Institutional
PFORX: PIMCO Foreign Bond (USD-Hedged) Institutional

Notice how bouncy PFUIX is. This what happens as the USD gets weaker or stronger against other currencies. The stanfard deviation for PFUIX for the last three years was 7.23 vs. 2.86 for hedged. AGG was 2.76 std. dev. The hedging eliminates the volatility.

Paul

[CyberBob]

Over long periods, it's amazing how similar U.S. and international index are. Except for the big bump in the 80's for the EAFE, it looks like they all got you to about the same place.



January 1, 1972 value = 100

December 31, 2007 values:
3960 - S&P 500
4323 - EAFE
4176 - MSCI World

Note: Both the EAFE and MSCI World indexes are developed-markets only.

Bob

[Valuethinker]

Valuethinker


Please explain unehdged currency exposure.

Unhedged currency exposure, your investment moves with the rise and fall of dollar against other currencies.

hedged, it does not.. Hence 'hedged'.

Note that currency moves are generally bigger than market moves, so if you are unhedged, you will be more playing the US dollar against other currencies, than you will be playing the stock market.

This is why unhedged bonds are probably a pretty bad idea: you are really taking a bet on currencies, because the +/- of the currency in any given year will almost certainly drown out the returns from the bonds.

[stratton]

This is why unhedged bonds are probably a pretty bad idea: you are really taking a bet on currencies, because the +/- of the currency in any given year will almost certainly drown out the returns from the bonds.

Which is why Swensen etc. prefer taking unhedged risk in equities. About 25% to 50% of the foreign returns US investors have had over the last several years is from the depreciating USD.

What You Should Know About Currency Risk

For example, the table shows that the return experienced by US investors who invested in the developed and emerging markets of the world excluding the United States should attribute ¼ to ½ of their return to currency effect – the decline of the US Dollar against those other currencies in this case.

Paul

[wearethefall]

I live in the UK, so if I followed the world market cap-weighting I would have only 8.5% of my portfolio in domestic equities. This would mean a wopping 91.5% of my portfolio would be subject to currency risk!

I recently read Triumph of the Optimists and the authors make a good case for a large overseas allocation. They say that in the long run purchasing power parity holds, so exchange rates should track relative inflation levels. They also show that currency movements have a small negative correlation with overseas markets, which should offset some of the risk of investing overseas.

[Adrian Nenu]

Currency vs. equity exposure.

Japan.

The problem with unehdged currency exposure is that you are taking on a volatility which is greater than stock market volatility, and may not work your way.

A hedged global equity fund pays the price of hedging (which is not trivial) but would be superior from a diversification point of view.

The other problem is Japan: 30% of the world stock market in 1990 v. sub 10% now i believe?

You could get caught in that sort of valuation trap.

One strategy is to break the equity allocation into Europe, US, Pacific and Emerging Markets indexes. The use some common sense: if one of them, Pacific Index for instance, reaches an obviously ridiculous valuation for no reason as Japan did in the 80's, then decrease exposure. Same for NASDAQ in the late '90s - another obvious speculative bubble which could have been avoided. Stay the course but use some common sense and don't chase performance over the cliff. Or at least rebalance and don't let it ride until your luck runs out.

Adrian
anenu@tampabay.rr.com

[Valuethinker]

I live in the UK, so if I followed the world market cap-weighting I would have only 8.5% of my portfolio in domestic equities. This would mean a wopping 91.5% of my portfolio would be subject to currency risk!

I recently read Triumph of the Optimists and the authors make a good case for a large overseas allocation. They say that in the long run purchasing power parity holds, so exchange rates should track relative inflation levels. They also show that currency movements have a small negative correlation with overseas markets, which should offset some of the risk of investing overseas.

Yes they teach that theory in class, too .

In the very long run, currencies do track relative inflation rates (for traded goods). But I've seen such large swings plus and minus relative purchasing power parity, in my lifetime, that I doubt I'll ever hit the exact point of parity.

For example London houses are clearly overvalued relative to the world, suggesting the pound is way, way overvalued to the world (or at least London housing is). But this never seems to correct.

Similarly on the negative correlation with overseas markets of currency movements, I doubt it is a large enough effect to fully offset.

What you really need to do is work out what percentage of your consumption is based on non-GBP priced goods and services (lots, I would reckon at least 30% if not 50%) and hedge by holding that level in non-sterling assets.

Since this is a fiendishly difficult thing to do (BP is the largest stock in the UK, but it is mostly producing in US dollars).

[jhd]

Unhedged currency exposure, your investment moves with the rise and fall of dollar against other currencies.

hedged, it does not.. Hence 'hedged'.

Note that currency moves are generally bigger than market moves, so if you are unhedged, you will be more playing the US dollar against other currencies, than you will be playing the stock market.

This is why unhedged bonds are probably a pretty bad idea: you are really taking a bet on currencies, because the +/- of the currency in any given year will almost certainly drown out the returns from the bonds.

Most international index funds are unhedged, right? Are there good, low-cost index funds hedged for (say) the US dollar? If not, how should an investor hedge currency risk?

I suppose you could do this directly, by selling an equivalent number of pounds or euros or yen, or through a fund that does this. But that will either take time and expertise, or incur some additional expenses.

[asset_chaos]

I live in the UK, so if I followed the world market cap-weighting I would have only 8.5% of my portfolio in domestic equities. This would mean a wopping 91.5% of my portfolio would be subject to currency risk!

Only if your portfolio is 100% equities. If you have a balanced portfolio, then you can have your bonds all denominated in pounds. A 50/50 stock/bond portfolio would have about 45% exposure to non-pound currency fluctuations. If that still keeps you up at night, the thing to do might be to hedge the currency exposure explicitly as a separate thing, though perhaps hard for an individual investor to do.

If one sense this is a dilemma of all investors in small market countries---and I'm in Australia, an even smaller market. Americans might be able to convince (fool) themselves that 40-odd percent of the world represented by the US domestic stock market is close enough not to bother with the rest of the world. On the other hand, a Finn who puts all his stock money into the Finnish total stock market index sounds like they are well diversified too but has half of their stock portfolio is in Nokia. (Though to be fair, as Finland uses the euro now, a Finn could index all Europe now.)

[tadamsmar]

Charles Ellis known for his book Winning the Loser's Game and as a former chair of Yale's Investment Committee claims that long term investors, including US investors, should not overweight there own country's stock.

In a recent article in Barons Ellis made the following point: The best risk-adjusted returns over the long term can be scored by matching the market capitalization weightings of the world's markets, Ellis says. That would mean putting 45% in domestic stocks, 47% in developed foreign markets and 8% in developing foreign markets.

The idea is to have no bets on whether one market or another will be stronger going forward.

Any thoughts?

Currency vs. equity exposure.

Japan.

The problem with unehdged currency exposure is that you are taking on a volatility which is greater than stock market volatility, and may not work your way.

A hedged global equity fund pays the price of hedging (which is not trivial) but would be superior from a diversification point of view.

The other problem is Japan: 30% of the world stock market in 1990 v. sub 10% now i believe?

You could get caught in that sort of valuation trap.

Also international stock markets are increasingly correlated with the US-- this is empirically observable. So diversification benefits have been decreasing since the 1980s, steadily. In a crash, correlations move to 1.

If your currency is unhedged, the best guess, based on past performance, is that maximum diversification benefits occur between 20 and 30% of total portfolio value.

At one time the market cap of Japan exceeded that of the US.

Later, the market cap of Pfizer exceeded that of Japan.

Of course, Pfizer used Viagra to do that

[Easy Rhino]

Japan.

If your home country turns out to have the next case of Japanitis, then having extensive non-home assets would be pretty darn useful.

[docneil88]

If your currency is unhedged, the best guess, based on past performance, is that maximum diversification benefits occur between 20 and 30% of total portfolio value.

David Swenson, Ph.D. & Chief Investment Officer of the Yale Endowment, appears to confirm what Valuethinker's is saying.

Fortunately, finance theorists conclude that some measure of foreign exchange exposure adds to portfolio diversification. Unless foreign currency positions constitute more than roughly one-quarter of portfolio assets, currency exposure serves to reduce overall portfolio risk. Beyond a quarter of portfolio assets, the currency exposure constitutes a source of unwanted risk. [Source: p.60 of Unconventional Success: A Fundamental Approach to Personal Investment.]

Unfortunately, I don't know which "finance theorists" he is referring to, and I don't know what statistics they used to arrive at the "one quarter of portfolio assets" conclusion. I also wonder whether that proportion is meant to apply just to US investors, or investors from all countries. Best, Neil

[wearethefall]

I live in the UK, so if I followed the world market cap-weighting I would have only 8.5% of my portfolio in domestic equities. This would mean a wopping 91.5% of my portfolio would be subject to currency risk!

Only if your portfolio is 100% equities. If you have a balanced portfolio, then you can have your bonds all denominated in pounds. A 50/50 stock/bond portfolio would have about 45% exposure to non-pound currency fluctuations. If that still keeps you up at night, the thing to do might be to hedge the currency exposure explicitly as a separate thing, though perhaps hard for an individual investor to do.

I am currently at 100% equities. Swenson has a rule that no asset class should make up more than 30% of a portfolio. I'm currently 50% domestic equities resulting in 50% currency exposure - which means I exceed his currency exposure rule too.

I have trouble trading off between the two, but I'm considering slightly increasing my international exposure. Things are a bit complicated by the fact that I earn in $, making my labour income correlated with my ex-Europe investments.