In Defense of an International Equity Allocation
Most of the retail investing world has now finally received the message that index investing is both easy and good for you—one of the few things I’ve found in this life that can lay claim to exhibiting both of those characteristics at the same time. Much of the beauty and utility of index investing is found in the reality that once you adopt it as an investing world-view, you can sit back, put your feet up, and bask in the glory of efficiently “managing” your investments while essentially doing nothing. In fact, laziness in managing your index fund assets is actually a positive attribute in that it often allows you to avoid the common behavioral patterns that lead to error and long-term under-performance (see, e.g., the so-called “behavior gap“). By enbracing a laziness bias you can avoid all those nasty results—buy high and sell low anyone?—attributable to action bias.
But even as a committed index investor devoted to laziness there is a function that you need to actively engage in: determining portfolio allocation. A multitude of factors can and should enter into this analysis and decision, but the two most critical factors are your time horizon and your tolerance for volatility. (The industry usually speaks in terms of “risk tolerance,” but I think this is a minsnomer. It is my (somewhat historically informed) belief that if you have 20-plus years to invest your money, there is almost zero risk that you will lose principal in equity investments over that period. So the real question is whether you can stomach the ups and downs that are guaranteed to occur on that long road.) Because the vast majority of our portfolio is still long-term money, it is predominately allocated to equities. Check. Done … right? Nope. Even if you are a “total market” indexer like we are, you still have to decide which geographic markets and economies you will buy. For those of us in the US, the question really boils down to how much—if any—you want to allocate to foreign equity markets. For nearly all of the 17 years that we have been equity index investors I have targeted an allocation of 25-30% of our equity portfolio to international equity index funds. I figured now would be a good time to sit down and re-visit that decision and see if I can convince myself all over again that it’s the right decision (or at least the decision that I’m most comfortable with).
The Argument In Support of International Equity Diversification
Let’s go through the base case in support of an allocation to international equities. The factors in support of an allocation include (i) additional portfolio diversification; (ii) exposure to corporations and industries that are not present (or as prevalent) in the US indexes; (iii) exposure to rapidly growing so-called “emerging market” economies; and (iv) theoretically higher risk-adjusted returns over long periods of time.
(i) Additional Diversification. I like diversification and the idea of spreading my bets among investments that don’t always move in lock-step with one another. While this figure varies over time as the various global markets move up and down, the fact remains that roughly 50% of global equity market capitalization is outside the US. This chart from Vanguard shows the movement over time between US and non-US market capitalization as of 12.31.13 (see the full report here):
I have never been comfortable with the idea of ignoring fully one-half of the world’s equity market capitalization just because it was non-US based. And while the US market comprises roughly 50% of global equity market cap, it accounts for only roughly 25% of global GDP.
International equity exposure—if not currency hedged—also brings additional diversification from exposure to the foreign currencies in which the foreign companies generate their earnings. These currencies are translated back to US dollars for the US-based investor that holds an equity index investment denominated in the greenback, adding currency movement to gains and losses from the stock movement of the underlying businesses. Vanguard depicts this additional foreign-currency exposure here:
The upshot here is that currency movement has historically not been correlated to stock price movements, and thus over time currency fluctuation has helped to reduce the correlation between US and non-US equities. I have no idea what the US dollar will do in decades to come, so I like the idea of also being somewhat currency diversified (even if over the long term it is expected that the net result of these currency fluctuations will largely net to zero).
In sum, while I remain confident that the US economy will likely out-perform most non-US economies over the coming years and decades, that doesn’t necessarily mean that US equity markets will out-perform global equity markets. Valuations matter, and the US is mucho expensivo right now.
(ii) Exposure to Specific Corporations and Industries Not Present in US Indexes. This is of course part of the additional diversification factor noted in (i) above, but it’s worth considering from a closer perspective. If your only index investment is the S&P 500 or US total market, you would have no exposure to exceptional global businesses like Toyota, Samsung, Nestle, Royal Dutch Shell, and Siemens, to name a few. What if the next hot Samsung smart-device kills the iPhone? Not likely it seems to me, but I also didn’t see the death of the blackberry coming back in 2006 when I was toting around my hip corporate-gig ball-and-chain on my belt. Check out how BBRY has done since, say, August 2007 (hint: if you had $100 invested at that date you may have enough today to buy a Happy Meal). And more to the point: I don’t want to have to worry about crap like that. It is also the case that while the US share of global market cap has been growing, that larger market cap is becoming increasingly concentrated in a shrinking group of publicly traded US corporations.
Vanguard makes the case that international equity indexes also comprise a greater exposure to old-world industries that are no longer as prevalent in the US indexes, and thus a focus on the US market exclusively would overweight new-technology industries like bio-tech and IT while under-weighting more old-school industries like electrical equipment, durable goods, and automobiles. To the left and below is a side-by-side look at the different top-10 stock holdings and industry sector concentrations in VXUS (total international) versus VTI (total domestic stock):
(iii) Exposure to Rapidly Growing Emerging Market Economies. Earnings growth drives equity returns and emerging market economies outside the US have the greatest growth potential. Think of places like Brazil, Russia, India, China, South Africa, Mexico, Thailand, Malaysia, etc. This type of explosive growth potential does not exist in the US market outside of individual stock offerings—and I’m not interested in placing concentrated bets on Amazon, Snapface, or Facenotes. One of my favorite stats in support of long-term equity investment is that rolling 10-year annualized total returns (i.e., including dividends) for the US market since 1900 have averaged roughly 10%. This is great data in support of long-term equity investment in the US, but it must also be considered that for much of this period the US was itself an emerging market. That type of growth opportunity likely doesn’t exist in the US any more. But of course, with young and fragile emerging market economies, you also get stories like this (political turmoil in Brazil roils emerging markets recovery).
(iv) Higher Risk-Adjusted Returns. Greater “risk-adjusted returns” are often cited in support of an international equities allocation. There are a number of different ways to calculate a risk-adjusted return, but in concept they are all getting at the idea of return per amount of risk assumed. Because I’m a long-term investor that generally accepts maximum “risk” in the form of potential maximum volatility, I’m not really all that concerned about risk-adjusted return. I just want maximum return over the long-term. Higher risk-adjusted returns with lower absolute returns are of no interest to me, so I largely dismiss this factor cited in support of an international allocation. (I’ll note too that since the diversification benefit on volatility has largely disappeared since around 2008, as discussed below, even if I was concerned with risk-adjusted returns, I’m not sure that this factor is valid support for an international allocation.)
The Argument Against International Equity Allocation
I have recently been re-acquainted with the reality that a number of well-respected (ok, totally globally respected) market luminaries are somewhat opposed to the idea of any allocation to international equities. This list includes Jack Bogle (founder of Vanguard), Warren Buffet (Chairman of Berkshire Hathaway and widely regarded as the best investor alive today), and JL Collins (FI blogger and all-around bad-ass). I don’t like being at odds with this group over what to order for lunch, let alone over a technical investing issue. Their arguments in opposition to an international allocation are largely consistent and straight-forward: an international allocation adds unnecessary complexity to the portfolio while at the same time adding additional expense. Why add the complexity of global geo-political concerns and foreign exchange exposure (at additional cost) when the US market with its large multi-national corporations that do business everywhere can provide the same diversification benefit? Keep it simple they say. Even the Bogleheads can’t agree on this issue. (Footnote here: One thing we can all agree on, and that someone needs to inform Johnny Depp, is that international investing does not include buying private islands in the Bahamas or compounds in the South of France.)
A Wee Bit of Data
Here are some interesting data points that are relevant to the discussion:
–Emerging market economies currently trade at a price/earning ratio of approximately 16 and the total international index is trading at around 20.6x earnings. The US is currently trading at nearly 27x earnings. Data from Vanguard:
Global equity indexes historically trade at p/e levels that are similar to the US indexes, whereas emerging market indexes tend to trade at a slightly lower p/e than the foreign and domestic developed markets. I don’t think that “this time is different,” and we are likely to experience some form of reversion to the mean, which likely portends a reduction in the p/e multiple for the US indexes. At 20.6 the international index is also trading at a somewhat lofty level, impacted significantly by gains over the last 6 months.
—Using data as of March 2016, an allocation of 25% to international equities produced the same 10.2% annualized return as an all-US stock portfolio over the past 46 years, but with lower volatility and less significant drawdowns.
What We Do
For nearly the entirety of our investing history I have targeted an international equity allocation of between 25-30%. Because we have historically been effectively 100% allocated to equities, this translated to 25-30% of our entire portfolio. As I begin to contemplate moving more of our portfolio into fixed income (as we age and look to drawing down our principal) this target will shft to 25-30% of our equity portfolio, which will be less than 100% of our total portfolio. With the exception of some random toys-in-the-attic investment vehicles accumulated over time for different purposes and that comprise less than 2% of our portfolio, our international equity vehicle is VXUS/VGTSX (Vanguard total international in either the ETF or mutual fund form) and our US domestic equity vehicle is VTI/VTSMX (Vanguard total stock market in either the ETF or mutual fund form). Our portfolio allocation looks like this (after you include a few dollars that have begun the migration to bonds):
The Vanguard total international vehicle tracks the FTSE Global All Cap ex-US Index, which is described as a “float-adjusted market-capitalization-weighted index designed to measure equity market performance of companies located in developed and emerging markets, excluding the US.” It includes more than 6,000 stocks of companies located in more than 45 countries. The current portfolio geographic composition looks like this:
One of the many reasons I have preferred the Vanguard total international index over the years (aside from all the typical Vanguard advantages like expenses and avoidance of conflict of interest) is because it holds a healthy dose of emerging market stocks in its basket, which not all international equity index funds do. (At the current allocation of 19.3% of the portfolio this is somewhat inflated as a result of recent performance of emerging market stocks.) I have a rough target of a maximum 10% of our equity portfolio allocated to emerging markets (which the Personal Capital allocator tool also helps me to monitor), and so even if the emerging market allocation were to reach up to 30% of the total international fund I would be comfortable and within my target allocation. I’m currently still below this 10% overall threshold even though we have been placing some concentrated bets on emerging markets for the last year or so because of depressed valuations.
Where Does the 25-30% International Allocation Logic Come From?
The target allocation of 25-30% of the portfolio to international equities was driven by a Vanguard study showing that nearly all of the diversification benefit to be derived from an international equity allocation can be achieved at a 30% allocation, and 85% of the benefit was achieved at a 25% allocation. But the Vanguard study also points out that this benefit has changed over time as markets move, and different levels of international allocation produced different volatility benefits at different points in time. The study also points out that for the 10-year period ended Dec. 31, 2008, any amount of allocation to international equity index funds actually increased portfolio volatility. So if you are considering an international allocation strictly for the reduced volatility benefit, you may need to consider whether this is still a realistic possibility.
Performance Over Our Investing Timeline
Below is a picture of what the total international index has done over the last 17 years as compared to the total US market and S&P 500 indexes:
There are a lot of interesting data points here. The most obvious point at first glance is that from day 1 to today, the US market has trounced the international index by more than double the return (total US market returning more than 80% and international returning 31.94%). But upon closer inspection you see that nearly all of this disparity has taken place in the last 3 years as the US markets took off and international markets moved mostly sideways. The interntiaonl index had its own heyday back in 2007 just before the crash when it reached return levels approaching where the US market is today, and then suffered losses that were not as severe as the US market. The point here is to see that while the US and international indexes do appear to generally move in the same direction, they do not move in lock-step and gains and losses of significantly different magnitudes will be experienced. So I’m quite glad that I stayed the course back in 2006-07 as international markets rocketed and didn’t forego my target US allocation. And we’ll do the same thing now in sticking with our international allocation even though the last 3 years have proved much more profitable for the US side of our portfolio. We’ve spent the last couple of years being greedy with respect to international and emerging market indexes as others were fearful, and I’m optimistically confident that this approach will yield fruit in the long run. In fact the tide may be turning; here’s what the last 6 months look like:
International indexes have nearly doubled US returns over this short period (14.96% v. 7.75%). As a long-term investor this is of course irrelevant noise, but again it evidences the point that these asset classes can experience dramatically different results.
A note here on holding international funds and tax impact. Years ago I moved all of our interntonal index holdings into tax-advantaged accounts (401(k), IRA, etc.) because they were generally less tax efficient (in fact I did this at the trough of the downturn, and was able to harvest a hefty tax loss in the process). The international index has historically produced a higher dividend yield than the US index—currently VXUS (international) yields 2.69% as compared to 1.86% for VTI (US). And in the days before ETFs you also had to worry about capital gain distributions resulting from stock turnover, and for whatever reason the international index fund typically had higher turnover yielding higher capital gain pass-through taxes (note that this higher turnover feature appears to no longer be the case). Most if not all of this capital gain pass-through problem goes away if you are investing through an ETF as opposed to a mutual fund, but you still have the increased dividend yield tax hit. (Note: international index fund dividends can be classified as “qualified dividends” that will receive preferential capital-gains tax treatment, including potentially a 0% tax rate depending upon your individual tax bracket. So if you are an FI-type that is strategically managing your taxable income, you could eliminate the tax hit from higher dividends by planning to maximize the use of that 0% rate. But of course higher taxable dividend amounts will reduce the availability that you have in the 0%-capital-gain tax bracket for other planning maneuvers.)
But also note that this analysis that prefers tax-advantaged accounts for international vehicles is likely not the same for everyone. Determining the proper placement (taxable v. tax-deferred account) for international funds requires a more in-depth look at your personal tax situation, including current marginal tax rates, proportion of dividends coming through the fund that will be “qualified,” state income tax treatment (do you get a foreign tax credit?), and future plans (will the amounts be taxable at any point in the future? Maybe not if you anticipate a tax-free Roth conversion ladder approach). By holding the international funds in a tax-advantaged account you are effectively giving up the benefit of the foreign tax credit that could be claimed for amounts withheld by foreign tax authorities. If these amounts are subsequently subject to US tax when pulled out of a tax-advantaged account (traditional IRA or 401(k)), you will likely end up paying double tax on those amounts (foreign withholding on the initial dividend payment followed by US income tax on the distribution). But since I plan to never owe any US tax on these amounts because they will all eventually be converted to a Roth IRA before distribution, this concern is mostly eliminated. (And another thought: there does not appear to be a tax-planning opportunity resulting from holding international funds in a taxable account, generating qualified dividends that will be taxed at a 0% rate because you are FI, and then claiming a full foreign tax credit. Amounts taxed at the 0% rate in the US will need to be excluded from your calculation of “foreign source income,” effectively eliminating your ability to claim the foreign tax credit.)
When I began investing in VGTSX many years ago the expenses were significantly higher, albeit not exorbitant since we’re talking about Vanguard here. It was also the case that for a long time the international fund did not have an “Admiral” class that significantly reduced the carrying expenses. All of that is now mostly irrelevant with the advent of ETF funds where you can get the lowest expense ratio option from the first dollar invested. Today VXUS costs 0.11% for international diversification, whereas VTI (recently reduced by 1 bps) only costs 0.04%. So the international portion of the portfolio costs almost 3x the domestic piece. If you are allocating say 27.5% of your equity portfolio to international and VXUS, that means for every $1 million invested you will spend $193 more in fees for the international diversification piece (.07% expense ratio delta x $275,000). I can live with that.
Impact on a 4% Safe Withdrawal Rate
The original Trinity Study that serves as the basis for the 4% safe-withdrawal rate did not include international equities in its modeled portfolios that were run through the Monte Carlo analysis. But subsequent safe-withdrawal rate studies have been done that include international equities in the equity side of the model portfolios, and they yield similar results. Here’s one from Vanguard that calculates a safe-withdrawal rate (with 85% success) of 3.6% for an “aggressive” portfolio (80% stocks/20% bonds) where the equity portion is allocated 70% domestic/30% foreign.
My Re-Affirmed Conclusions
I started this post as a defense of my long-held belief in a significant international equity allocation. After initially reviewing the long-term return data (that I admit to having ignored for a few years now since I’m a lazy index investor) I found myself questioning the approach. But after sitting with it for a while and working through the material I was able to resolutely conclude that—for me at least—this is the right approach. I didn’t chase the money of international returns back in 2006-07, and I’m not going to chase the money of the US market returns during 2014-2017. Even though the correlation between the US and global equities has been on the rise since the mid-1990s, and the diversification benefit on volatility has largely disappeared since 2008, we’re going to continue to own a significant stake in the entire world and be fully diversified. This approach has been valuable over the last few years as international equities have proved to be the only equity asset class where you could put new money to work at depressed (or at least not elevated) valuation levels. Since we have another 40-plus years to let much of this investment ride, I think we’ll stick with it.