Vanguard: Optimizing International Diversification

Christine Benz (intro): Hi, I'm Christine Benz for Morningstar.com.

I recently interviewed Christopher Philips, who is a senior investment analyst in Vanguard's Investment Strategy Group. One of the topics we discussed was the optimal foreign-stock weighting in investor portfolios.

...

Benz: Chris, thank you so much for being here.

Chris Philips: My pleasure.

Benz: You and your team have recently done some interesting work on international weightings in a portfolio. I know investors often grapple with this. There has been a stampede of assets into international over the past decade, but investors aren't sure how much of their portfolios, of their equity portfolios, to devote to foreign stocks.

I think if you are truly agnostic, you might think "Well, the U.S. market is a 40-some percent of the global market cap, maybe I'll start there." Is that a good approach?

Philips: Actually it is, and if you think about investment theory, investment theory would suggest that one would want to be market proportional. So, if the market is hypothetically 50-50, U.S. and non-U.S., and that's a reasonable place to start.

What we do know is that the markets aren't perfectly efficient, and a lot of investors have embedded biases that, for better for worse, would skew them towards one allocation that tends to be the domestic allocation.

Benz: Their home-country bias.

Philips: Exactly their home-country bias. We also know that the U.S. is the most liquid market in the world. We know it's the most diversified market in the world. Just think about our sector allocation, we're much closer to the global portfolio than, say, the German sector allocation or Japanese sector allocation.

Benz: So we've got a really broadly diversified market here by sector.

Philips: Exactly. And we also have the lowest cost. So the transaction costs are much lower in the U.S. than elsewhere.

So all these things in aggregate actually suggest that investors might be OK by having that home bias. That it might not be completely out of the ordinary to have a larger portion of their assets in U.S. investments than in non-U.S.

Now that being said, we wouldn't want them to be 100% in the U.S. simply because there is that opportunity for diversification, and we certainly believe that someone would want to be more diversified rather than less diversified, and the way you can do that is by having some allocation to non-U.S. stocks.

Benz: So you ran a bunch of scenarios where you looked at optimal foreign allocations. Where do you come down in terms of recommendations for most people? Where do they get the best risk-reward profile in terms of apportioning their U.S. versus foreign stock assets?

<TRANSCRIPT>

Philips: That's a challenging question--the reason being that, depending on the time period that you are looking at, that answer can actually change. So if you look at the longest time period that we have, say, from 1970s through the most recent market date, then you actually end up with around 30% in non-U.S. and 70% in U.S. as being that "optimal."

If you look back, let's say, just the last 10 years, then the optimal is actually closer to 10% in non-U.S. and 90% U.S., mostly because of the volatility around the global financial crisis, and international stocks actually got hurt a little bit more than U.S. stocks.

So the time period that you are evaluating does make a huge difference. That's why, when we talk with clients, we actually recommend a range, and we talk about a range of anywhere from 20% to 40% making the most sense for the largest number of investors out there. If you skew more towards risk aversion, being more concerned with the local domicile, so how the U.S. is performing, you want a little bit more control of the costs, the risk exposures, then you might want to skew more towards that 20%. If you want more opportunity for diversification, you might skew a little bit more towards that 40%, but that's a reasonable range.

Benz: Does age figure into this at all, and the fact that you oftentimes get foreign currency gains and losses embedded in your foreign stock holdings?

Philips: I think age is a factor only to the extent that you may become more loss averse as you become older. And that loss aversion as opposed to risk aversion in terms of volatility is a significant behavioral aspect, where if the U.S. dollar significantly appreciates, all of a sudden international investments, because they go opposite the dollar in terms of the currency exposure, they would actually get hurt more than they would if there was no dollar exposure in them. So, that loss aversion can play a role.

If we have an investor who is able to take a step back and ignore the short-term currency fluctuations with the idea that currency is unpredictable, it is a random walk, than we would actually say that the same amount of diversification in non-U.S. stocks make sense throughout one's life cycle.

Benz: You hinted about how, over the past decade, foreign stocks have generally underperformed, U.S. stocks, have been a lot more volatile. One other thing that we've seen is just correlations have gone up a lot, so you're not getting the same diversification bang for your foreign stock holdings that you were maybe a couple of decades ago. What do you make of that trend? Is it something that's here to stay, do you expect?

Philips: We don't know. Our assumption is that, all else equal, correlations going forward will likely be higher than they have been historically. The magnitude, the differences, we can't really pinpoint. Reason being that we do know globalization is increasing and should continue to increase. We know that barriers to investment are coming down, so it is easier for any investor out there to gain access to a diversified mutual fund or ETF in the industry, and get that diversification. We know that globally, embedded home biases are starting to come down. So, for example, a U.K. investor today may not have the same degree of home bias that they might have had 10 or 15 years ago. So, you have this global expansion of capital and awareness of this global portfolio, and certainly that does impact correlations.

What we do also know, however, is that correlations aren't the only factor in diversification, and a great example is some of the recent volatility in the marketplace, where we've seen U.S. drop, say, 5%, but European stocks drop, say, 3%. Now, they are both down, so we would say correlations are close to 1, but if you have some of both, then you're going to be in the middle, and so the magnitude difference of returns is also a critical component that I am not sure a lot of people think about, but it certainly weighs into that idea of broad diversification.

Benz: Right. Last question for you, Chris, on this topic. In some respects do you find that classifying companies as foreign or domestic based on their country of domicile is kind of a vestige of a bygone era? Where you've got Coke and McDonald's, for example, doing more than half of their business overseas, and many foreign companies doing a lot of business in the U.S. Does it make sense for investors to classify their portfolios based on where the companies are actually based?

Philips: Again, that's an interesting question, and there has been a lot of research done on this idea of multinational companies and how they play into an investor's portfolio. What we do know is that still today, where a company is domiciled has the most influence on its returns. So, McDonald's is going to be more highly correlated to the U.S. market than it is going to be to, say, the European market or to the Japanese market, or wherever else it has its revenues come from. The biggest reason is because businesses operate in the currency markets as well. So McDonald's wants to take out currency risk from their portfolio, from their sales, their revenues, and so they'll operate in the futures market to hedge out a lot of that currency exposure. That flows through into the income statement, the balance sheet, and what you get is an earnings stream that can actually be more reflective of the performance of the dollar and of U.S.-based assets. That's true for any country and any company in those countries. So we certainly want investors to be aware of multinationals and where the companies they are investing in reside and how they play out in the portfolio, but we also do think there is a significant amount of value to having foreign domiciled investments as well.

Benz: Okay. Chris, Thank you so much for sharing this research. Very helpful.

Philips: You're very welcome.