“Purgatory for Pessimists” — International Equities
An Unemotional, Factor-Based Approach
The current Bull Market has been unkind to non-U.S. allocations. At a conference I recently attended, the term TINA1 (there is no alternative) was overheard, more than once, in the context of allocating investor portfolios. Expected returns on fixed income are likely to be low, and identifying top quartile alts managers — let alone gaining access to them — is increasingly difficult.
This conundrum is further exacerbated by U.S. stocks’ dramatic outperformance on the global equity stage. As shown in the table below, International and Emerging Markets have lagged their U.S. counterparts by over 5% annualized for the trailing 10-years.2 Of course, the persistent climb has lifted U.S. valuations in turn and, no matter what metric or timeframe you look at,3 the U.S. is now more expensive than it was a few years back. While value is not necessarily a good timing metric, the opportunity beyond the U.S.— International markets 17% cheaper and Emerging Markets 35% cheaper — is hard to ignore:
Performance & Valuation4 (As of 9/30/17)
Though 10 years can seem like an eternity for investors, it’s really just one full market cycle. As shown below in the chart of rolling 3-year performance,5 the current market environment represents the 5th “round trip” over the past 4½ decades of leadership cycles.
Performance — U.S. vs. Foreign
(Rolling 3-Year Return)
One allocation approach would be to simply buy the respective index ETFs for International and Emerging Markets and call it a day. While market-level valuation metrics can be useful as an initial pass, they tend to obscure the risk & reward within broad averages. Explanations abound as to why non-U.S. stocks have lagged,6 but I tend to think of non-U.S. markets as a “Purgatory for Pessimists” because there is always something of justifiable concern. Because humans are hard-wired to form simple heuristics for efficient decision-making, that line of thinking (in the years following The Great Recession) might’ve gone something like this:
- Crisis in Greece,
- Greece is in Europe,
- European crisis!
- Europe is outside the U.S.,
- Foreign = unsafe!
- Rinse & repeat for
- North Korean hostilities, and
- China’s ascendancy.
A simple indexing approach to these markets manages these “risks” by naively obfuscating them within averages of index returns.
Digging Beyond Market Cap to Identify Opportunity
The return & valuation metrics presented (see “Performance & Valuation” table above) are all based on common cap-weighted indexes that cover hundreds, if not thousands, of stocks across their respective market. They tend to favor larger-cap firms, thus omitting at the worst and underweighting at the least, a large portion of the global opportunity set. As believers that a stock’s characteristic profile predisposes its future performance, we often look at factor spreads for evidence of opportunity to generate excess return.
Factor Spreads: High Decile minus Low Decile
(% from 1991–2016)
As described in prior osamresearch.com articles, the 6 themes that we’ve built collectively define the factor profile of a stock — agnostic on geography, market cap, or style. From these themes, 3 are used specifically to select stocks: Value, Momentum, and Shareholder Yield.7 The 3 remaining themes help eliminate from consideration the stocks that fail on certain quality tests: Earnings Quality, Financial Strength, and Earnings Growth.8
In the “Factor Spreads” table above, we look at the return spread between the highest-ranked and lowest-ranked decile of all 6 factor themes from 1991 through 20169 within 3 different universes: U.S. stocks, International Ordinary Shares, and American Depository Receipts (ADRs).10 While factor investing is more commonly applied within the U.S. market, the table above demonstrates that stock selection based on factors can be as effective, if not more so, outside of the U.S. market.11
The key takeaway from these factor spreads is that tremendous performance differentiation underlies average returns for broad groups of stocks. Wider spreads generally suggest a larger opportunity to harvest excess return when using a disciplined, factor-based approach. In the same way a pure indexing approach neglects the management of seemingly obvious risks, indexing also (naively) turns a blind eye to large, consistent, and persistent structural trends that are more likely to generate alpha.
Harnessing Factors to Build Differentiated International Portfolios
Select the ‘Right’ Stocks, Avoid the ‘Wrong’ Stocks
Underlying the factor spreads for Ordinaries12 and ADRs is a relatively consistent stair-step monotonic trend. As you move from highest-ranked to lowest-ranked stocks on each factor, there is a relatively linear degradation in excess returns. To demonstrate, we take the Ordinary and ADR universes, rank them on our Value theme, and then organize them into decile portfolios from cheapest (leftmost) to most expensive (rightmost). This chart shows the annualized excess return on those decile portfolios versus an equal-weighted universe of stocks that meet certain liquidity and market cap criteria.13 We’ve found similar results as with our other two key stock selection themes: Momentum and Shareholder Yield.
Excess Return (1991–2016) <br>
OSAM Value vs. Equal-Weighted Universe
Cap-weighted indexes fail to acknowledge this empirical evidence. Astute investors would seek to concentrate into the highest-ranked deciles and avoid stocks in the lowest-ranked deciles to enhance portfolio return. One way to do this in practice is to tilt towards cheap stocks, owning a little more of them than the cap-weighted index and a little less of the expensive stocks. This “smart beta” approach works for large institutional investors who are more mindful of Tracking Error and care more about Information Ratios than absolute returns. A second way, which I’ll explore next, is to simply own the highest-ranked stocks, while avoiding the rest of the universe.
Incorporate Multiple Factors for Consistency & Diversification
Value by itself is a powerful factor. However, incorporating Momentum and Shareholder Yield provides benefits to risk-adjusted return and consistency. The table below shows the results of a hypothetical multi-factor ADR portfolio from 1991 through 2016:14
This multi-factor ADR portfolio outperforms the MSCI ACWI ex U.S. by 8.3% annualized (1991–2016) with a dramatically higher Sharpe Ratio. Importantly, the multi-factor portfolio’s performance is consistent.15
Concentrate & Weight by Conviction — not Market Cap
Part and parcel with owning only the highest-ranked names by a given factor is accepting the fact that large portions of the eligible universe will go un-owned. This introduces significant differentiation into a portfolio in terms of Active Share, but also higher Tracking Error.16 Given the information on Value shown above, investors should certainly be comfortable not owning the underperforming deciles and accepting the Tracking Error that comes along with it.
To evaluate the importance of concentrated portfolios, we ran portfolios ranging from 5 to 300 stocks and, as the number of names within the portfolio expanded, we charted the excess return relative to the MSCI ACWI ex U.S. We constructed two versions of the concentrated portfolios to show the deleterious impact of market cap-weighting when introduced.17
Diffusion of Factor Excess Return by Concentration
OSAM Value (ADRs 1991–2016)
In both cases, as the number of names allowed in the portfolio expands the degradation of excess return is irrefutable. Also, notice the disparity in excess return between an equal-weighted versus a cap-weighted portfolio construction process. These are the exact same stocks! Yet, a cap-weighted approach underperforms by hundreds of basis points — even in large 300-stock portfolios.
There are No Free Lunches, Embrace Lower Capacity
Sure, all of this looks great on paper. The challenge is implementing effectively. As a portfolio’s concentration increases, strategy capacity decreases. And, as the tie to market cap weighting is severed, capacity decreases and implementation costs increase. Fortunately for U.S.-domiciled investors, ADRs provide cost effective vehicles whose implementation costs are on par with their U.S. common stock counterparts. Below, we compare the all-in market impact cost to make trades of various sizes for U.S. common stocks, ADRs based on portfolios of 50, 100, 200, and 300 stocks at portfolio sizes from $50 million to $1 billion.
Impact Costs by Concentration & Capacity
(5-years ending 10/31/17)
In all cases, portfolios are more expensive to trade as concentration and capacity increase, but the overall cost to trade U.S. and ADR stocks is remarkably similar (skilled traders can narrow the gap even further). Impact costs increase significantly between both regions’ 50-stock and 100-stock portfolios, reinforcing that there are tangible implementation costs to balance out the benefits of concentration.
While the Bull Market rages on in the U.S., it’s important for investors to remember that all market cycles are mean reverting. The U.S. has been the leader but may not be in the future. In fact, 2017 has seen a reversal of the prior years’ trend. While Emerging and Developed Markets have outperformed, it remains to be seen if this trend continues. Valuations are significantly discounted outside the U.S. market. Factor spreads are equally as wide as, or wider than, historical averages, which suggests that disciplined investors have a significant opportunity to harness the power of factors to create differentiated portfolios.
For our part, we suggest a strategy that parses the international landscape to eliminate stocks which rank poorly on quality criteria and then concentrate on names with strong Value, Momentum, and Shareholder Yield.18 For those investors who are able maintain a consistent factor profile over time, this type of factor-based investment strategy can be a key driver of performance in their portfolios.
This article will be discussed in the next episode of host Jim O’Shaughnessy’s “What Works on Wall Street” Podcast, joined by this article’s author Ehren Stanhope, CFA.
Sign up at osam.com/podcast and we’ll keep you informed.
- TINA captures the collective sentiment that equities, despite a massive bull run and rising valuations, are one of few viable asset classes to park capital. For more, see investopedia.com/terms/t/tina-there-no-alternative.asp
- To put that return gap in perspective, the current balance of $1M invested in International and Emerging Markets 10 years ago would be worth about half of a similar investment in the U.S. (about $1.1 million less).
- For example, metrics such as earnings, sales, cash flow, book value; timeframes such as cyclically-adjusted, trailing 12-months, normalized.
- For this table, in lieu of our multi-factor Value theme we use an individual factor (Price-to-Earnings) to measure valuation.
- Expanding our window of returns back to 1970 for U.S. and non-U.S. markets reveals a regular cyclical pattern of leadership.
- Explanations range from central bank intervention to economic growth to geopolitical crises.
- We assess our Value theme through the lens of multiple underlying factors — sales, cash flows, and earnings. Our Momentum theme seeks stocks with strong appreciation over the prior 3, 6, and 9 months, while avoiding those that are highly volatile. Shareholder Yield represents a total return of capital to shareholders through dividends & share buybacks.
- Earnings Quality evaluates the use of accruals to boost earnings. Financial Strength assesses a company’s reliance on outside sources of capital to support its balance sheet. Earnings Growth helps to avoid unprofitable firms.
- Prior to 1991, the ADR universe of stocks was insufficiently large to make an accurate comparison.
- ADRs are traded like stocks in the U.S. but provide exposure to foreign companies).
- As an example, the return spread between the cheapest and most expensive U.S. stocks is 17.8% over the 26-year period. For Ordinaries, the Value spread widens to 19.1%, and for ADRs it’s even wider (21.1%). We present ADRs alongside Ordinaries to show that these uniquely-structured securities provide an enticing alternative to the operational complexity and cost structure of a local share portfolio but without degradation in factor returns.
- For Ordinaries, the universe of stocks is those domiciled outside the U.S. with a market cap greater than $200M and average daily volume greater than $250K. ADRs are compared against a similar universe, but with the added constraint that the security must be an ADR. Both universes are equal-weighted and agnostic to benchmark constituency.
- That is, we don’t begin our assessment from the point of view of a cap-weighted benchmark. As shown, investors ought to own more of the cheap stocks and fewer of the expensive ones.
- The multi-factor ADR portfolio is built by eliminating stocks that rank poorly on our Earnings Quality, Financial Strength, and Earnings Growth themes, then by selecting names that fall into the highest-ranking deciles on our Value, Momentum, and Shareholder Yield themes.
- We measure consistency with Base Rates, which are batting averages for how often a strategy beats its bench-mark in rolling periods. In this case, the multi-factor portfolio outperforms the benchmark 95% of the time in rolling 3-year periods.
- Active Share is a simple calculation that compares the holdings of a fund or portfolio with the holdings of the index used as its benchmark. For example, a 60% Active Share indicates that 60% of a portfolio’s allocations are different than the benchmark’s allocations.
- The equal-weighted version simply ranks all stocks in the universe on our Value theme and then equal-weights the cheapest names. The cap-weighted version selects exactly the same names, but weights them in the portfolio proportionate to their market caps.
- See osam.com/iadr for more background on our “International ADR” strategy, which has successfully used this methodology since its Jan-2006 inception.