2000年的影子

QUARTERLY LETTER SHADES OF 2000 Ben Inker | Head of Asset Allocation 3Q 2019 Most of us have times in our professional lives we prefer not to think too much about. In my case, the period from 1997 to early 2000 was that time. It seemed as if every single EXECUTIVE SUMMARY thing that could go wrong for us from an investment standpoint did. Sure, value stocks The years leading up to the 2000 stock were drastically underperforming, but emerging equity and debt didn’t have to go market bubble were extraordinary and through a crisis simultaneously. Long-Term Capital didn’t have to blow up taking with unprecedented. They caused unique it a whole series of assets that had little or nothing in common with each other besides pain to the portfolios of valuation-driven the fact that anyone with an eye to history would have judged them relatively cheap. investors. The valuation extremes, In my own personal contribution to ensuring nothing would go right, I convinced our though, created the greatest opportunity U.S. quantitative equity team to make a tiny change to our stock momentum model. It set for valuation-driven investors since disqualified any stock from being purchased on momentum grounds if it was trading the Great Depression. While the events at over 25 times fair value on our intrinsic value model. It removed a grand total of of the last decade have not been as one stock from our equity portfolios.1 Unfortunately, that stock was America Online; striking as those of the late 1990s, the in the course of 1999, that omission lowered the return of our U.S. Core strategy by recent cycle has gone on for significantly longer and the pain caused to our something over 1 percentage point. portfolios has begun to approach 1990’s The relative performance of our global equity and multi-asset strategies over the last levels. As the current cycle has ground half of the 1990s was, frankly, unimaginably bad. Exhibit 1 shows the performance of on slowly but surely, the valuation each strategy relative to its respective benchmark from 1994 to 1999. extremes have moved wider, creating an opportunity set for valuation-driven investors that looks as extraordinary as what we saw 20 years ago.

EXHIBIT 1GMO GLOBAL EQUITY ALLOCATION AND MULTI-ASSET PERFORMANCE VS. BENCHMARKS (1994-99) Composite Performance vs. Benchmark 10% 5% 0% -5% Multi-Asset: -21% -10% -15% -20% -25% Global Equity Allocation: -26% -30% 1993 1994 1995 1996 1997 1998 1999 As of 12/31/99 | Source: GMO | All performance is net of fees.

1 For strategies with tracking error in the 3% to 4% range, these numbers were the kind of Our universe at the time was the S&P 500. Had we events that should have happened about every 900 years.2 Those of our clients that stayed applied the rule to the NASDAQ our net likely would have caught more of the Internet bubble’s high flyers. with us were unsurprisingly rather annoyed and quite tired of listening to our argument 2 that the prospective returns for our strategies were actually very good. We were often At least if market returns were normally distributed. told, “You just don’t get it,” and asked, “How can you make up for the lost ground?”

Shades of 2000 | p2 Happily, for those clients that believed enough of what we were saying to stick around, on or about the turn of the millennium things changed. The next decade saw the strategies win by amounts that well rewarded their patience.

EXHIBIT 2GMO GLOBAL EQUITY ALLOCATION AND MULTI-ASSET PERFORMANCE VS. BENCHMARKS (1994-2009) Composite Performance vs. Benchmark 60% 50% Global Equity 40% Allocation: +102% 30% 20% Global Equity 10% Allocation: -26% Multi-Asset: +70% 0% -10% Multi-Asset: -21% -20% -30% 1993 1995 1997 1999 2001 2003 2005 2007 As of 2/28/09 | Source: GMO | All performance is net of fees.

Ever since 2000, we have always assumed that neither the pain from the 1990s, nor the opportunity set that the pain created, would come our way again. And on a day-to-day and month-to-month basis, nothing quite like it has occurred. But there is more than one way for things to get to extremes. The performance of some of our strategies is once again approaching the 1990s-style cumulative pain level, as we can see in Exhibit 3.

EXHIBIT 3GMO GLOBAL EQUITY ALLOCATION AND MULTI-ASSET PERFORMANCE VS. BENCHMARKS (1994-2019) 60% Composite Performance vs. Benchmark Global Equity Global Equity 50% Allocation: +102% Allocation: -21% 40% 30% 20% Global Equity Multi-Asset: -16% 10% Allocation: -26% Multi-Asset: +70% 0% -10% -20% -30% Multi-Asset: -21% 1993 1996 1999 2002 2005 2008 2011 2014 2017 As of 8/31/19 | Source: GMO | All performance is net of fees.

Shades of 2000 | p3 While the magnitude of our underperformance has been 20% to 25% less than it was last time around, its duration has been particularly painful. This period has lasted over 10 years instead of the 5 years of the late 1990s. Unsurprisingly, our clients are once again finding their patience wearing thin.

Now, just because the pain has been notable doesn’t necessarily mean that the opportunity created is correspondingly impressive. To assess the current opportunity set, we have the advantage that we have been publishing our asset class forecasts since the mid-1990s. This history allows us to compare the opportunity set through time on the forecasts we have used to build our portfolios. Exhibit 4 shows this for the two strategies we’ve been looking at so far.

E XHIBIT 4FORECAST OF GLOBAL EQUIT Y ALLOCATION AND MULTI-ASSE T PORTFOLIOS VS. BENCHMARKS (1996-2009) Forecast of Multi-Asset and Global Equity 5% 4% 3% Allocation Portfolios Relative to Benchmarks 2% Global Equity Allocation “ As the Internet bubble burst and valuation 1% 0% Multi-Asset disparities normalized, 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 our portfolios delivered Source: GMO extraordinary Forecasts are the “Mean Reversion” forecasts published by GMO at the time. outperformance and the superiority of our In the run-up to 2000, the forecasts for both portfolios got better and better relative portfolios based on our to their benchmarks, driven largely by the extraordinary discount of value stocks in own forecasts slowly the late 1990s. As the Internet bubble burst and valuation disparities normalized, our portfolios delivered extraordinary outperformance and the superiority of our waned. portfolios based on our own forecasts slowly waned. In the 2000 case, the poor prior performance set the stage for better times for our portfolios, which was appropriately reflected in the forecasts we were seeing at the time.

What about this time around? Exhibit 5 updates the expected return charts through August 2019.

Shades of 2000 | p4 EXHIBIT 5: FORECAST OF GLOBAL EQUIT Y ALLOCATION AND MULTI-ASSET PORTFOLIOS VS. BENCHMARKS Forecast of Multi-Asset and Global Equity 6% 5% Global Equity Allocation 4% 3% Allocation Portfolios Relative to Benchmarks 2% Multi-Asset 1% 0% 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: GMO Forecasts are the “Mean Reversion” forecasts published by GMO at the time.

For both the global equity portfolio and the multi-asset class portfolio, things appear to be even better than they were in 2000. On the face of it, our portfolios have never had a better opportunity set than they have today, at least relative to traditional portfolios. How can that be, when the value opportunity was better in 2000? Part of the reason is that other asset allocation opportunities beyond the value style are even better than they were then. Table 1 shows the percentile rank of the valuation spread for regional stock valuations and style valuations.

TABLE 1: ASSET ALLOCATION OPPORTUNITIES WITHIN EQUITIES VALUE STYLE WITHIN EAFE vs. U.S. EM vs. U.S. U.S. EAFE EM Dec 1999 68.9% 88.7% 99.4% 99.2% 93.4% Aug 2019 99.9% 93.8% 93.1% 96.5% 94.7% As of 8/31/19 | Source: GMO Valuation spreads are calculated on an average of several capital proxies.

In both the U.S. and EAFE markets, 1999 was a more extreme value opportunity. Not only were both spreads into the 99th percentile, but the only datapoints more extreme were the months immediately surrounding December 1999. But EAFE and emerging are both significantly cheaper relative to the U.S. today than they were back then. And while value is not quite as well-positioned in the developed world, it is slightly better positioned in emerging than it was then.

Shades of 2000 | p5 Now come the caveats. The forecasts I have been using for these calculations are our “Mean Reversion” forecasts, which assume that all valuations are going to revert to long-term averages. In 2000, those forecasts assumed it would take 10 years for valuations to revert to normal, whereas now we use a 7-year reversion period. This means today’s forecasts generally have slightly larger amplitudes than the ones from back then. And beyond the calculation details, many investors made the case in 2000 that long-term averages were not meaningful anymore and the future would be far different from the past. From the standpoint of the world from 2000 to 2010, those investors proved to be wrong, as asset prices generally moved toward long-term averages. The case for believing that today’s world is different does seem stronger to us than it did then. While my guess is that future generations will look on a world with 30% of all investment grade developed market bonds trading at a negative yield3 as evidence that we had collectively lost our minds for a while, I’m significantly less confident in that assertion than we were in saying the S&P 500 was stupidly overvalued in 2000.4 Given that uncertainty, today we have another set of forecasts that we also use in building our portfolios – the “partial mean reversion”5 forecasts – that assume interest rates have permanently moved lower and have driven down other assets’ required returns as well. The forecasts in that scenario are marginally less “ friendly to our portfolios, but not materially so, as we can see in Exhibit 6, which also makes an adjustment for the 10-year reversion period of the 2000 forecasts relative to the 7-year reversion period we use today.

The opportunity set today is far and away the best one we have seen in EXHIBIT 6: FORECAST PORTFOLIO SPREAD TO 20 years...

BENCHMARK 7.0% 6.0% GMO 7-Year Forecast Spread of 6.0% 4.6% 4.8% 5.0% 4.4% 4.4% 4.0% 3.5% Portfolio vs. Bencharmk 3.0% 2.0% 1.0% 0.0% According to Bloomberg, as of August 29, 2019. https:// Multi-Asset Global Equity Sep 2000 Aug 2019 Aug 2019 Partial Mean Reversion The strongest reason to have had more confidence in Source: GMO the mispricing in 2000 was that the S&P 500’s valuation Forecasts are the “Mean Reversion” forecasts published by GMO at the time. The “7-year Equivalent implied impossibly optimistic future growth. Today’s bond yields, by contrast, imply extremely dismal future Forecast” for Sep 2000 is calculated by multiplying the forecast spread by 10/7. This is a slight growth. We have a hard time believing that the future will oversimplification of what a true 7-year forecast would have said at the time but should be quite close. be quite that dismal, but such an outcome definitely does not qualify as “impossible,” unfortunately.

We used to refer to the two sets of forecasts as Adjusting the 2000 forecasts for full comparability does make today’s opportunity set “Purgatory” and “Hell.” We stopped using those terms look slightly less extraordinary than the 2000 opportunity set and using the Partial Mean less because we were offending people than because Reversion forecasts affect the multi-asset portfolios at the margin as well. However, the clients found it endlessly confusing that the forecasts in opportunity set today is far and away the best one we have seen in 20 years and it is very “Hell” were higher than the forecasts in “Purgatory.” much in the ballpark of the 2000 event.

Shades of 2000 | p6 Ben Inker And for multi-asset portfolios, this is also not the end of the story. In 2000, we had Mr. Inker is head of not yet launched our Benchmark-Free Allocation Strategy, where we are free to GMO’s Asset Allocation put together a portfolio without worrying about our tracking error to a traditional team and a member portfolio.6 In that strategy we have the advantage of not needing to hold any U.S. of the GMO Board of equities in order to control our tracking error against a global equity benchmark. And Directors. He joined GMO in that portfolio, we can do even better than we can in our other multi-asset portfolios. in 1992 following the completion of his B.A. in The forecast differential for that portfolio today is 5.5% in the mean reversion Economics from Yale University. In his years scenario and 4.6% in partial mean reversion.7 A narrowing of these levels of forecast at GMO, Mr. Inker has served as an analyst for the Quantitative Equity and Asset Allocation spreads would serve as strong tailwinds for our portfolios, enabling us to outperform teams, as a portfolio manager of several equity benchmarks as we did in the early to mid-2000s. and asset allocation portfolios, as co-head of There is, of course, no absolute guarantee that valuations will revert even part way International Quantitative Equities, and as CIO to historical levels. This time may in fact be different. But even in a world in which of Quantitative Developed Equities. He is a CFA today’s valuations persist indefinitely, we believe our portfolios still have a strong charterholder. inbuilt advantage due to their substantially higher income and earnings. In that world, it would take a long time to make up for the relative losses we’ve suffered as valuations moved to extremes, but in that dreary, low-return world, an extra percent Disclaimer The views expressed are the views of Ben or so of returns every year would go a long way. Inker through the period ending October 2019, Today is not 2000, and today’s opportunities are not quite the same as the ones we had and are subject to change at any time based at our disposal 20 years ago. But in our estimation, they are similarly extraordinary on market and other conditions. This is not an and far better than anything we had seen before that event or have seen since. offer or solicitation for the purchase or sale of any security and should not be construed as such. References to specific securities and issuers are for illustrative purposes only and are not intended to be, and should not be interpreted as, recommendations to purchase or sell such securities.

Copyright © 2019 by GMO LLC.

All rights reserved.

Exhibit Disclosures For Exhibits 1-6, Multi-Asset refers to GMO Asset Allocation Strategy. Exhibits 4-6 represent real return forecasts for the underlying asset classes and not for any GMO fund or strategy. The forecast above is based upon the reasonable beliefs of GMO and are not a guarantee of future performance. Forward-looking statements speak only as of the date they are made, and GMO assumes no duty to and does not undertake to update forward‐looking statements. Forward‐looking statements are subject to numerous assumptions, risks, and uncertainties, which change over time. Actual results may differ materially from those anticipated in forward-looking statements.

Performance Disclosure Performance data quoted represents past performance and is not predictive of future performance. Returns are presented after the deduction of a model advisory fee and a model incentive fee if We started talking to clients about the strategy in the fall of 1999, but it wasn’t until the second half of 2001 that applicable. Net returns include transaction costs, commissions and withholding taxes on foreign anyone actually took us up on managing such a portfolio. income and capital gains and include the reinvestment of dividends and other income, as applicable. A 7 GIPS compliant presentation of composite performance is available at www.gmo.com. Actual fees are Based on GMO’s August 2019 asset class forecasts relative to a 60% MSCI ACWI/40% Bloomberg U.S. disclosed in Part 2 of GMO’s Form ADV and are also available in each strategy’s compliant presentation. Aggregate benchmark. Fees paid by accounts within the composite may be higher or lower than the model fees used.