Q: This month, two prospectus changes to the All Asset funds were filed with the SEC: 1) the addition of Chris Brightman as a portfolio manager on the funds, and 2) an increase in the All Asset All Authority Fund’s international equity limit from 33⅓% of total assets to 50%. Can you comment on the rationale for both of these changes?
Arnott: As the only named PM on the All Asset suite, I’ve received a disproportionate share of the credit for building and managing this path-breaking diversifier for investors, most of whom tend to be over-reliant on a two-pillar-centric 60/40 stock/bond balanced portfolio. Jason Hsu was heavily involved in the first decade of the All Asset suite, and he continues to serve as Research Affiliates’ Vice Chairman as he builds a new business as the majority owner of Rayliant (formerly our Greater China division, and a business in which Research Affiliates continues to maintain a substantial minority interest). As Jason increasingly focused there, Chris Brightman became more involved in the management of the All Asset funds over the past four years. Adding Chris as co-PM is both an acknowledgment of his already important role in the All Asset suite and a signal that I’m far from alone in managing and enhancing our investment process.
Since becoming our CIO in 2014, Chris has become instrumental in managing the All Asset suite, and helped us to weather a difficult three-year bear market in most “Third Pillar” asset classes (inflation-linked assets, credit and emerging markets). Chris has 33 years of investment management experience in equities, fixed income, currencies and asset allocation, with a dozen years as CIO or CEO of substantial institutional investment management businesses. Since joining us in 2010, he has led teams in research, investment management and portfolio infrastructure, in addition to product development and client reporting. Today Chris partners with me in examining all major decisions relating to the All Asset strategies and coordinates the day-to-day implementation. In my eyes, this official filing is simply a well-earned recognition of Chris’s partnership with me in managing the All Asset suite.
While giving credit where credit is due, I’d be remiss not to highlight the deep roster of research and investment talent at Research Affiliates that also supports the All Asset funds. Currently, we have 42 investment professionals firmwide – including 11 Ph.D.s – all with an average experience of 10 years in quantitative finance. We hope that, in time, their significant contributions are as recognized externally as they are internally. So as a first step, we’re going to do something about that – beginning next month, we’re renaming this monthly commentary “All Asset All Access” to emphasize the breadth and depth of our resources and our commitment to making them all available to you. While Chris and I are the portfolio managers, there’s far more to our team than many people realize. Meet the leaders of our Research and Investment Management team!
Does this change suggest that my role is shrinking? Hardly. It’s a reflection of the growing depth of the Research Affiliates team. I cannot imagine anything that I could enjoy more than shepherding your (and my) investment resources through the storms that undoubtedly lie ahead.
Now, let me explain why we are raising the international equity limit in the All Asset All Authority Fund from 33⅓% to 50%, a change that will go into effect next month. First and foremost from an investment perspective, our models are looking for modest incremental exposure to international equities in the current environment, given meaningful valuation differentials versus the U.S. market (which we describe in more detail below). We think the move will benefit investors.
Does this reflect a structural shift in our investment view toward international equities, or a structural increase in our long-term average equity beta or total volatility? No. In fact, the revised international equity limit for All Authority simply brings it in line with what we can already do in All Asset. Recall that All Asset has a lower total equity limit – 50% versus 66⅔% for All Authority – but doesn’t specify an international equity limit within that 50%. In 2003, when we launched All Authority with a higher total equity limit than All Asset, PIMCO had only had one international equity fund; today we have 12!
Bottom line, our total equity limit remains unchanged; we are simply harmonizing our ability to access international markets at a time when our tactical views are calling for that. Additionally, recall that the All Asset All Authority fund has the flexibility to take a short position in the U.S. equity market equal to 20% of total assets. So if we modestly increase exposure to international equities, we can simultaneously reduce our overall equity beta and volatility profile if we believe that is the best approach. Of course, we will continue to be tactical in dialing up or down these measures in response to the ever-changing environment.
Regarding the valuation disparity I mentioned earlier: Over the last three years ending 30 September 2016, U.S. equities (proxied by the S&P 500) have delivered better real returns, at an annualized 10.2% per annum, than almost every other broad global asset class. At the same time, non-U.S. developed market equities (proxied by the MSCI EAFE Index) and emerging market equities (proxied by the MSCI Emerging Markets Index) have delivered lackluster annualized returns of 0.0% and −1.2%, respectively, and they remain the source of disappointment, pain and losses for many investors.
As a result, U.S. stocks are now priced at a Shiller P/E ratio (i.e., price relative to 10-year earnings) of over 26x, a top decile valuation that roughly matches the peak levels before the global financial crisis. True, the U.S. is the healthiest economy in the developed world, but it’s hard to make a case that current economic growth or growth prospects warrant a top decile valuation. Developed ex U.S. equities, on the other hand, are now priced at a Shiller P/E under 14x – that’s over 45% cheaper than the U.S. market. Emerging market (EM) equities are even more compelling at a Shiller P/E under 12x, a bottom decile level. (All asset classes are proxied by the indexes listed in the previous paragraph.) Even after this year’s 16.0% return (as of 30 September 2016), EM equities could double in price tomorrow and still be cheaper than the U.S. market!
We don’t expect these valuation discrepancies to normalize immediately, or fully; the U.S. does deserve a modest structural premium. But the ability to collect higher dividend yields and potentially higher dividend growth while we wait for the valuation disparities to normalize is attractive. Research Affiliates’ long-term return forecasts (which Chris will discuss in more detail below) suggest developed ex U.S. and EM equities may deliver annualized real returns of 5.6% and 6.9%, respectively, for the next five to seven years compared with a mere 0.9% annualized real return forecast for U.S. equities.
Q: A key input that informs the allocation decisions within the All Asset strategies is the return forecasts for the underlying asset classes, which are based on Research Affiliates’ rigorous research and analysis. Looking back, how accurate have these return forecasts been, and over what time horizon?
Brightman: Our industry seems perpetually focused on past performance, even in ever-shifting market environments. Performance is typically reported every day down to the 1/100th of a percentage point, accompanied by many analytical tables and graphs. Does this keen attention to past performance correspond to its importance to our investment decisions? No! Simply extrapolating average past performance into the future in no way encompasses the scale and scope of quantitative research that informs our asset class return forecasts.
As our firm’s name suggests, research is a core function within Research Affiliates. We engage in substantial academic and quantitative research in developing our investment strategies, which includes the asset class return and covariance assumptions that drive our asset allocation views. And we make much of that work publicly available. Most notably, on Research Affiliates’ Asset Allocation website, we display our 10-year real return forecasts for the world’s publicly traded asset classes (proxied by widely available market indexes), explain the building blocks of these future returns, provide documentation of our methodology and include a host of tools that can inform portfolio strategies. But before relying on our asset class forecasts, or the forecasts of anyone else, you should ask, “How accurate are they?”
As we show below, our asset class market forecasts historically have demonstrated a significant correlation to subsequent actual results, though the correlations clearly increase over the time horizon. If you wish to know which asset class will provide the highest return next year, then pay little attention to our forecasts. The short-term volatility of capital market prices is mostly unpredictable noise, producing a very wide distribution of returns over a one-year horizon. Over the longer run, however, the noise of price changes tends to dissipate, and the structural drivers of long-term returns – notably income yield and income growth – more reliably forecast future returns.
So we say, but where’s the proof? We can refer you to Nobel laureate Robert Shiller’s book, “Irrational Exuberance,” but we suspect that you will find visual displays of the quantitative information more accessible.
One key finding that informs our forecasts is how, for decades, the yield on the Barclays U.S. Aggregate Bond Index has accurately forecast its subsequent 10-year return. (“Accurate” can be a subjective term, but a 92% correlation – see Figure 2 – qualifies as accurate in our view.) Within that framework, today’s historically low yield of 2% for the aggregate bond market predicts low returns over the next decade. That suggests that even if the Federal Reserve merely hits its inflation target of 2%, then the coming 10-year annualized real return for the U.S. bond market would be approximately 0%.
To many, this historical observation that bond yields accurately predict bond returns is intuitive and unsurprising. In our analysis, we note a less well appreciated trend: The same relationship between yield and return exists for equity markets going back nearly a century. Cyclically adjusted earnings yield, defined as 10-year average real earnings per share divided by current price, provides a reasonably accurate (75% correlation) forecast of the subsequent 10-year annualized real return of the U.S. equity market (see Figure 3).
Because our analysis indicates the market’s cyclically adjusted earnings yield tends to revert to its mean, in our models we add a valuation component to equity market returns. When the market’s earnings yield was far above its historical average – for example, the 10%–12% yields in the 1930s–1940s and again in the 1970s–1980s – subsequent annualized real returns were 15%–20%. Conversely, when the earnings yield was far below its historical average – for example, the 2%–3% yields in the late 1920s and late 1990s – subsequent annualized real returns were close to 0%. With today’s earnings yield of 3.8% well below the long-term average of 6%, this suggests a 10-year annualized real return for the U.S. equity market in the 1% range.
Low starting yields mean that mainstream U.S. stocks and bonds are likely to provide future returns over the next decade far below their long-term historical averages. The important next question is, “What can we do with this information?”
While the yields on U.S. stocks and bonds have been depressed by central bank policies, many other asset classes are priced at higher yields. Our estimated excess return for Third Pillar asset classes (inflation-hedging assets, credit and emerging markets – please see note below Figure 4 for specific Third Pillar asset class proxies) over a 60/40 mix of U.S stocks and bonds is much greater than usual. Over the next five to seven years, we forecast an equally weighted portfolio of these diversifying assets classes to provide future excess returns of 2%–3% relative to a traditional 60/40 mix; see Figure 4.
We can apply our methodology for estimated returns over the past twenty years by looking at previous environments in which Third Pillar markets were, in our view, priced to deliver 2%–3% higher returns than a traditional 60/40 stock/bond portfolio, as is the case today. As evidenced in Figure 5, historically the correlation of the original excess return forecast to the subsequent realized return strengthened considerably over longer time horizons – note how the range of the return quartiles narrowed over the three-, five- and 10-year timeframes. In fact, by year three, not a single diamond fell below the 0% horizontal axis, indicating that the forecast for an excess return in Third Pillar markets came to fruition over the medium term. Most importantly, by year 10, the resulting median return differential (yellow diamond) between the Third Pillar and the 60/40 portfolio was almost exactly in line with our original forecast.
When we look though the noise of short-term price volatility, we can make informed forecasts about which asset classes are likely to provide higher and lower returns in the years ahead. Today’s depressed yields for core bonds and U.S. stocks reduce our estimated return for a traditional portfolio to not much above the likely rate of inflation. The resulting near-zero real returns would fail to meet most investors’ objectives. Rather than accept this failure, we aim to tactically increase our allocations to the more cheaply priced, higher-yielding, and higher estimated return collection of Third Pillar asset classes. Investors in the All Asset suite stand to gain additional potential value-added from PIMCO’s active management within the underlying funds.
The All Asset strategies represent a joint effort between PIMCO and Research Affiliates. PIMCO provides the broad range of underlying strategies – spanning global stocks, global bonds, commodities, real estate and liquid alternative strategies – each actively managed to maximize potential alpha. Research Affiliates, an investment advisory firm founded in 2002 by Rob Arnott and a global leader in asset allocation, serves as the sub-advisor responsible for the asset allocation decisions. Research Affiliates uses their deep research focus to develop a series of value-oriented, contrarian models that determine the appropriate mix of underlying PIMCO strategies in seeking All Asset’s return and risk goals.