STERLING’S WORLD REPORT May 2013

THE BERNANKE PROPHECY: IMPLICATIONS FOR GLOBAL EQUITIES While these forecasts embody a wide range of underlying models and assumptions, the basic message is clear—long-term interest rates are expected to rise gradually over the next few years, rising (at least according to these forecasts) to around 3% at the end of 2014. 1 --Ben Bernanke, March 1, 2013 Ben Bernanke gave a remarkable speech earlier this year. As indicated in the “money quote” above, its message was clear: investors should be prepared for U.S. long-term interest rates to rise to around 3% by the end of 2014 and to somewhere in the 4%-to-5% range by 2017. Coming from a Fed Chairman, Bernanke’s message was remarkable in several ways. First, it was surprisingly specific. Second, it was notably provocative since its implied message is that investors in long-term U.S. Treasury bonds should be prepared to lose money, possibly quite a bit of money. And while interest rate forecasts are a dime a dozen, Bernanke’s outlook should be taken more seriously than most since he is in a key position to influence the actual outcome. No one who loses money in U.S. Treasury bonds will be able to say that he didn’t warn them. A key exhibit from Bernanke’s speech is shown in Chart 1, which depicts four forecasts of U.S. long-term interest rates from 2013 to 2017. While based on different methodologies, all of these forecasts illustrate his core message: “In short, we expect that as the economy recovers, long-term rates will rise over time to more normal levels.” Of course, Bernanke was careful to emphasize that there is a considerable range of uncertainty surrounding all such forecasts. Chart 2 depicts some complementary approaches to summarizing the uncertainty surrounding such forecasts. For example, according to a market-based measure of uncertainty derived from “swaptions,” the forecast for U.S. 10-year bond yields by the end of 2014 could be amended to say that bond yields are likely to head to 3% plus or minus 1%, with the plus or minus range reflecting a 70% confidence interval. But this is basically saying that an investor’s base case should be that long-term rates will head to 3% by the end of 2014, with 15% odds respectively that they could end up higher than 4% or lower than 3%. 2 The implied message – that bond investors should be prepared to lose money – can be gleaned from the results of Bloomberg’s handy total return calculator, shown in Chart 3. According to the bond arithmetic, if 10-year Treasury rates rise from the current level of 1.76% to 3%-to-4% by the end of 2014, the total return to investors would range from about minus 3.6% to minus 7.8%. The loss in terms of real purchasing power will be considerably more if inflation continues to run at close to 2% over that time period.

1“Long-term Interest Rates,” Remarks by Ben S. Bernanke at the Annual Monetary/Macroeconomics Conference: The Past and Future of Monetary Policy sponsored by the Federal Reserve Bank of San Francisco, March 1, 2013. 2Although not highlighted in Bernanke’s speech, it is worth noting from Chart 2 that the Fed’s own sensitivity analysis is notably skewed toward higher rates by the end of 2014; i.e., it appears to be a 3% base case with a range from about 4.3 to 2.4%.

STERLING’S WORLD REPORT WHAT GLOBAL EQUITY SECTORS HAVE THE GREATEST SENSITIVITY TO RISING RATES? If investors take the Bernanke prophecy seriously, then they will not only need to re-think their exposure to government bonds, which have been the asset class of choice during the great “de-risking” following the financial shock of 2008. To the extent they are looking to equities as a hedge against inflation and rising yields, they will also need to carefully consider the implications for different equity sectors and types of equities. As a starting point for this kind of analysis, we asked Trilogy’s Risk Group to conduct a simple exercise using Citi Research’s Global Risk Attribute Model (GRAM). The GRAM is updated every month based on regression analysis of weekly global equity and macroeconomic data over the past three years. The model can be used to stress test the likely performance of portfolios relative to a variety of benchmarks based on shocks to key global macroeconomic variables, like 10-year government bond yields, oil prices, non-oil commodity prices, credit spreads, currency rates, and investment style factors like small vs. large-cap, growth vs. value, or overall market sensitivity. The simple exercise we conducted was to create “Rising Rates” and “Falling Rates” portfolios of 100 names each, equally weighted. The Rising Rates portfolio was based on the top 100 names from the MSCI World Index (MSCI World) universe ranked by Citi’s estimated sensitivity to an increase in 10-year bond yields, while the Falling Rates portfolio was based on the bottom 100 names in the same ranking. We also conducted a similar exercise using the MSCI All Country World Index (MSCI ACWI), which differs from MSCI World by also including many securities from emerging markets. The sector composition of the Rising Rates and Falling Rates portfolios is not too surprising, with a few exceptions. As can be seen in Table 1, compared to MSCI World the Rising Rates portfolio has relatively little exposure to global equity sectors like consumer staples, health care, utilities and telecommunications. Not surprisingly, it favors cyclical sectors like financials, materials, industrials, and information technology. As is seen in Table 2, the same sector tilts are evident when MSCI ACWI is employed. Not surprisingly, the Falling Rates portfolios put more emphasis on defensive sectors like health care, consumer staples, and utilities. That said, they also put substantial weights on several cyclical sectors like consumer discretionary and financials and, perhaps most surprisingly, materials. In each case, however, the surprise is lessened by drilling down to the subsector weights for those sectors, which offer stark contrasts. In the financial sector, for example, diversified financials are favored in the Rising Rates portfolios while real estate is heavily favored within the Falling Rates portfolios. Finally, in the materials sector, steel and chemical shares are heavily favored in the Rising Rates portfolios, while the materials components of the Falling Rates portfolios are almost entirely gold mining shares. Presumably in recent years, falling rates have been associated with negative real interest rates and aggressive quantitative easing that boosts the appeal of gold. From that perspective, we suspect that some of the recent sharp declines in the gold price have come from large investors trying to hedge their exposure to rising long-term yields amid stronger U.S. economic data and some talk from Fed officials of a gradual tapering of Quantitative Easing (QE). WHAT GLOBAL EQUITY REGIONS HAVE THE GREATEST SENSITIVITY TO RISING RATES? The regional exposures of the Rising Rates and Falling Rates portfolios should also be of interest to investors as they think seriously about hedging their exposure to a normalization of long-term interest rates. Starting with the developed markets, Table 2 provides a snapshot of the regional exposures of the two equally weighted portfolios and indicates that both portfolios would have relatively low exposure to the U.S market compared to its 53% weight in MSCI World.

STERLING’S WORLD REPORT WHAT GLOBAL EQUITY REGIONS HAVE THE GREATEST SENSITIVITY TO RISING RATES? - continued Not surprisingly, Europe has a much larger weight in the Rising Rates portfolio than in the Falling Rates portfolio. With Europe struggling against the deflationary shocks of austerity and bank deleveraging, a global economic recovery that brought higher long-term yields would presumably be highly beneficial for European financial stocks, especially in countries like Portugal, Italy, and Spain that have been pummeled by the euro area’s sovereign debt crisis. Japan has the largest exposure in both the Rising Rates and Falling Rates portfolios, but with very different sectors emphasized in each. For example, most of the exposure in the Rising Rates portfolio is in the industrials, information technology, and materials sectors. In contrast, in the Falling Rates portfolio, the largest exposures are to consumer discretionary, consumer staples, and (within industrials) to railways which are prized for their real estate holdings.

Despite Canada’s reputation as a resource-based market, it curiously has almost no exposure in the Rising Rates portfolio. The one exception is a Canadian financial stock. On the other hand, and more curiously, Canada represents 21% of the Falling Rates portfolio. Once again, that is because of the appeal of gold stocks in a world of falling interest rates and aggressive QE: almost all of the Canadian stocks represented in the Falling Rates portfolio are stocks of companies involved in gold mining. Turning to Table 3, which is based on the broader MSCI ACWI universe, it is notable that both the Rising Rates and Falling Rates MSCI ACWI portfolios include a disproportionate share of emerging market securities, especially in the Falling Rates portfolios. The share of Emerging Markets (EM) in the Rising Rates portfolio was a healthy 21%, with cyclical securities and cyclically geared countries like South Korea heavily represented. But the share of EM in the Falling Rates portfolios was a whopping 57%, with a heavy representation of EM consumer staples and financials. Our impression from these results is that financial firms in high-inflation nations like Brazil, Turkey, and India have benefitted from lower inflation, interest rates, and capital inflows associated with deflationary pressures and falling rates in the developed markets. IF BERNANKE’S FORECAST IS CORRECT, HOW SHOULD THE RISING RATES PORTFOLIO PERFORM? Since the Rising Rates and Falling Rates portfolios have some extremely pronounced tilts relative to global equity benchmarks, it is worth asking how they might be expected to perform if the Bernanke long-term interest rate outlook is on target. That is what the Citi Research Global Risk Attribute Model is designed to do, subject to the following disclaimer: “Results may vary with each use and over time. IMPORTANT: The projections or other information generated by the Risk Attribute Model regarding the likelihood of various investment outcomes are hypothetical in nature, do not reflect actual investment results and are not guarantees of future results.”3 We would add to this disclaimer the observations that risk models like GRAM are based on statistical correlations that could well prove to be unstable. For example, if China’s growth is about to experience a structural downshift, as many now expect, the implications for commodity markets and the materials sector could be pronounced even as growth normalizes in the developed world. That would suggest that the materials sector would provide less of a hedge against rising interest rates than historical correlations would imply.

3Citi

Research Quantitative Analysis, Global Risk Attribute Model (GRAM), Version 2.0, March 7, 2013.

STERLING’S WORLD REPORT IF BERNANKE’S FORECAST IS CORRECT, HOW SHOULD THE RISING RATES PORTFOLIO PERFORM? - continued In Table 3, we show sensitivity analysis based on Citi’s model for both the Rising Rates and Falling Rates portfolios, focusing on the predicted “marginal excess returns” of each portfolio versus MSCI World. The marginal excess returns refer to the expected outperformance or underperformance of those portfolios with respect to a one-standard deviation shock to a variety of macroeconomic variables. Not surprisingly, the impact of a one standard deviation rise in 10-year yields is symmetric, with the Rising Rates portfolio predicted to outperform MSCI World by 7.7 percent while the Falling Rates portfolio is predicted to underperform the benchmark by 7.5%. Although these portfolios were designed to be almost a “pure play” on long-term interest rates, the correlations of rising rates with other macroeconomic variables is quite strong. Not surprisingly, the Rising Rates portfolio has a “reflationary bias” and would also outperform if oil prices or other commodity prices rose, if the dollar were to weaken against the euro, or if global markets rise and credit spreads narrow. The Rising Rates portfolio would also benefit if value stocks outperform growth stocks. Naturally, the sensitivities of the Falling Rates portfolio are all in the opposite direction. Note from the Risk Decomposition statistics that more than 90% of the risk of these portfolios is so-called “systematic” risk (mainly macroeconomic) compared to only 7%-to-8% being “unsystematic” (mainly stock specific). From the point of view of Citi’s risk model, Bernanke’s analysis has rather startling implications for the following reason. If U.S. long-term interest rates rise from 1.7% to 3.0% by the end of 2014, and foreign rates follow suit, that would represent a 3.1 standard deviation move from the point of view of the risk model. Thus the Rising Rates portfolio would be expected to outperform MSCI World by about 24% (= 3.1 * 7.7) while the Falling Rates portfolio would underperform by about 23% (= 3.1 * -7.5), implying a performance spread of nearly 50% between the two portfolios. With that type of potential performance differential between different types of equity portfolios if interest rates normalize, we suspect that Wall Street’s structured product teams will be working overtime to create long-short global equity products specifically designed to help institutional investors hedge against a potential sea change in the interest rate environment. We would also note that, within global equity markets, such insurance seems at the moment to be quite attractively priced. As shown in Chart 4, the relative valuation on a price-to-book value basis of the median stock in the Rising Rates portfolio to that of MSCI World is the cheapest it has been in over a decade. In particular, we find that the typical stock in the Rising Rates portfolio currently trades at a price-to-book value of about 1.0 compared to 1.9 for the typical stock in MSCI World. That represents a discount of about 47% compared to a pre-2008 average discount of 18%. Finally, even though Trilogy’s investment process does not begin with a macroeconomic overlay, we would note that our core Global, International and Emerging Markets strategies all currently have a bias, according to Citi GRAM estimates, to outperform if long-term interest rates rise. Our “growth-at-a-reasonable price” (GARP) investment framework has picked up on the fact that cyclical sectors like financials and information technology currently offer a more attractive tradeoff between growth and valuation characteristics than do highly defensive sectors like consumer staples, health care and utilities. Accordingly, we believe that our portfolios are well positioned if Mr. Bernanke is even half right on the future path of long-term interest rates.

William Sterling Chief Investment Officer Trilogy Global Advisors, LP (212) 703-3100 [email protected]

STERLING’S WORLD REPORT Chart 1 Alternative 10-Year Treasury Forecasts (From Bernanke Speech of March 1, 2013)

In his March speech, Fed Chairman Ben Bernanke presented four different forecasting approaches which all point toward significantly higher long-term interest rates over the next few years.

Chart 2 Measures of Uncertainty around 10-Year Rate Forecasts (From Bernanke Speech of March 1, 2013)

Bernanke’s speech also highlighted the uncertainty surrounding long-term interest rate forecasts, which increases with time. Note that the Fed’s own forecasts (in blue) are skewed to the upside.

STERLING’S WORLD REPORT Chart 3 Sensitivity Analysis for the Total Return to 10-Year U.S. Treasury Bonds through December 31, 2014

Source: Bloomberg If Bernanke’s illustration of a move of U.S. 10-year bond yields to 3%-to-4% by the end of 2014 is correct, it suggests that the total return to bond investors will vary from minus 3.6% to minus 7.8%.

STERLING’S WORLD REPORT Table 1 Sensitivity of MSCI World Sectors and Regions by Exposure to Long-term Interest Rates based on Citi Research’s Global Risk Attribute Model*

*Rising Rates and Falling Rates portfolios are based on the top 100 and bottom 100 names within the MSCI World universe based on sensitivity to rising 10-year yields Source: Trilogy Global Advisors based on Citi Research’s Global Risk Attribute Model (GRAM)

STERLING’S WORLD REPORT Table 2 Sensitivity of MSCI All Country World (ACWI) Sectors and Regions by Exposure to Long-term Interest Rates based on Citi Research’s Global Risk Attribute Model

*Rising Rates and Falling Rates portfolios are based on the top 100 and bottom 100 names within the MSCI All Country World (ACWI) universe based on sensitivity to rising 10-year yields. Source: Trilogy Global Advisors based on Citi Research’s Global Risk Attribute Model (GRAM)

STERLING’S WORLD REPORT Table 3 Global Equity Sensitivity Analysis Using Citi Research’s Global Risk Attribute Model (GRAM)

*Predicted excess return relative to MSCI World if the macro factor were to increase by one standard deviation. **Rising Rates and Falling Rates portfolios are based on the top 100 and bottom 100 securities within the MSCI World Universe ranked by sensitivity to rising 10-year yields. Source: Trilogy Global Advisors based on Citi Research’s Global Risk Attribute Model (GRAM)

STERLING’S WORLD REPORT Chart 4 Relative Price-to-Book Value of Median Stocks in the Rising Rates Portfolio vs. the MSCI World Index

Relative Price-to-Book

1 0.9 0.8 0.7 0.6 0.5

Dec-12

Jun-12

Dec-11

Jun-11

Dec-10

Jun-10

Dec-09

Jun-09

Dec-08

Jun-08

Dec-07

Jun-07

Dec-06

Jun-06

Dec-05

Jun-05

Dec-04

Jun-04

Dec-03

0.4

Source: Trilogy Global Advisors based on Citi Research’s Global Risk Attribute Model (GRAM)

On the basis of relative price-to-book value ratios, the typical stock in the Rising Rates portfolio is at its cheapest level in more than a decade relative to its counterpart in MSCI World.

STERLING’S WORLD REPORT Important Information: The opinions expressed above are those of Trilogy Global Advisors, LP and are subject to change. There is no guarantee that predictions or expectation will come to pass. This material does not constitute investment advice and is not intended as an endorsement of or recommendation to purchase or sell any specific investment or security. Investment involves risk. Investing in the securities of non-U.S. companies involves special risks not typically associated with investing in U.S. companies. Foreign securities tend to be more volatile and less liquid than investments in U.S. securities, and may lose value because of adverse political, social or economic developments overseas or due to changes in the exchange rates between foreign currencies. In addition, foreign investments are subject to settlement practices, and regulatory and financial reporting standards, that differ from those of the U.S. The risks of foreign investing are heightened for securities of companies in emerging market countries. Emerging market countries tend to have economic structures that are less diverse and mature, and political systems that are less stable, than those of developed countries. In addition to all of the risks of investing in foreign developed markets, emerging market securities are susceptible to illiquid trading markets, governmental interference, and restrictions on gaining access to sales proceed. Certain information herein has been provided by independent third parties whom Trilogy believes to be reliable. Although all content is carefully reviewed, it cannot be guaranteed for accuracy or completeness. Any graphs, charts or formulas shown are not and should not be used as the sole basis of making investment decisions. Source: MSCI. The MSCI data is comprised of a custom index calculated by MSCI. MSCI makes no warranties or representations and shall have no liability whatsoever with respect to any MSCI data contained herein. The MSCI data may not be further redistributed or used as a basis for other indices or any securities or financial products. This report has not been produced or approved by MSCI. © 2013 Trilogy Global Advisors, LP. All rights reserved.