Fixed Income Manager Selection: Beware of Biases

Transcription

1 QUANTITATIVE RESEARCH April 216 Fixed Income Manager Selection: Beware of Biases AUTHORS Ravi K. Mattu Managing Director Global Head, Analytics Mukundan Devarajan Executive Vice President Quantitative Research Analyst Steve Sapra Executive Vice President Quantitative Research Analyst Dzmitry Nikalaichyk Vice President Quantitative Research Analyst Asset owners typically use a two-stage process to construct their investment portfolio. The first step is the creation of a policy portfolio. This portfolio makes the tradeoff between expected returns and risk, and involves forecasting expected returns, volatilities and correlations for the various asset classes under consideration. The policy portfolio provides the baseline set of factor risks desired by the investor and serves as the anchor to the final portfolio. After the components of the policy portfolio are determined, the second step involves manager selection, whereby specific managers are hired under the assumption that, in aggregate, the managers portfolios reflect the factor risks of the policy portfolio. While this process makes portfolio construction tractable, it may lead to unintended risk exposures at the overall portfolio level. To the extent managers portfolios imbed structural tilts to certain risk factors, asset allocators may be systematically exposing themselves to risks beyond those expected from their policy portfolio.

2 2 April 216 Quantitative Research In this paper we employ a factor-based approach to analyze the historical returns of 27 core fixed income managers from the Morningstar Direct Database. We find compelling evidence that a majority of core fixed income managers are systematically exposed to certain risk factors, in particular credit risk. This exposure increases both the effective equity beta as well as the left-tail risk of the overall bond portfolio. Additionally we find a more subtle reason for the unintended credit exposure of core fixed income managers. Over the past 25 years there has been a secular increase in the weight and riskiness of credit in standard fixed income benchmarks. This, in conjunction with the systematic exposure of most core bond managers to credit risk, means that investors are likely exposing themselves to higher levels of credit risk than they may realize. EVIDENCE FROM RELATED STUDIES While fixed income managers long credit bias has been documented by others, previous research has focused on aggregate performance across managers (Bosse, Wimmer and Phillips, 213), or anecdotal evidence (see, for example, Friedman, Otieno and Gessinger, 212, and Dopfel, 24). In contrast, we run multi-factor regressions to measure structural tilts for each of the 27 fixed income managers in our database. The simulations-based framework of Brown and Marshall (21) provides a theoretical foundation for our approach, although the authors do not run regressions on actual fund returns. We also control for exposures to smart fixed income factors, by using the same factor set as Worah and Mattu (214) to describe super-secular themes in fixed income investing. The framework we use is similar in spirit to recent research on the exposures of equity fund managers to risk factors (i.e. market or beta, size, style, and momentum) by Bender, Hammond and Mok (214), but applied to sources of fixed income risk premia and fixed income managers. DATA We use monthly after-fee total returns from Morningstar Direct from January 25 to November We apply the following selection criteria to the universe of funds to identify the funds to include in the study: The fund is classified as belonging to the U.S. Fixed Income category as defined by Morningstar The inception date of the fund was before January 1, 25 The analyst-assigned benchmark or primary prospectus benchmark is the Barclays U.S. Aggregate Index The fund is classified as actively managed Our study includes funds no longer in existence as of November 215. These so-called dead funds are used to ensure that our results do not suffer from survivorship bias. Dead funds must have at least two years of data to be included. This filtering results in 154 live and 116 dead funds, for a total of 27. For each fund in the universe, we calculate a time series of returns in excess of the Barclays U.S. Aggregate Index. 2 In aggregate, the funds in our sample represent approximately 6% of the assets-under-management by U.S.-domiciled active bond funds that have a core bond benchmark. Figure 1 shows the volatility and tracking error for the average as well as the 25th, 5th and 75th percentile of our set of managers. Figure 1. Standard deviation of total and active fund and benchmark returns (January 25 to November 215) Average 25th percentile Median 75th percentile Fund return volatility (per year) 4.1% 3.1% 3.5% 4.3% Tracking error volatility (per year) 2.7% 1.3% 2.1% 3.1% Benchmark volatility (per year) 3.3% Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 The risk factors we use to analyze the systematic exposures of

3 April 216 Quantitative Research 3 funds in our sample are described in Figure 2. Duration refers to a factor that is exposed to parallel shifts in the yield curve. Credit and mortgage backed securities (MBS) are factors that are exposed to changes in average spreads of corporate bond and fixed rate Agency MBS indices, respectively. Curve (steepener) and rates volatility (sales) require some additional description, since they represent more exotic beta, and are not readily accessible through passive funds. The curve factor accounts for the impact of non-parallel shifts in the yield curve to portfolio returns. We measure curve factor returns as the difference between the returns of a duration-neutral portfolio that is long the Barclays US Treasury 1-3 years maturity index and is short the Barclays US Treasury 1+ years maturity index. A positive exposure to this factor indicates a bias in favor of steeper yield curves in the portfolio, i.e., the manager tends to under-weight the long end of the yield curve (relative to the benchmark) and should outperform when the yield curve steepens. Rates volatility represents a strategy that sells at-the-money 1 month x 1 year swaption straddles (i.e., options with one month to expiry, on the 1-year swap rate) at the closing prices on the 5th, 1th, 15th, 2th and 25th of each month, delta hedged on a daily basis. Returns for this strategy are scaled to match the trailing five-year volatility of the U.S. Treasury index. Positive exposure to rates volatility indicates the manager tends to sell volatility. This may be achieved by either explicitly selling options on interest rates or indirectly by constructing bulleted portfolios or by owning callable bonds, including fixed-rate mortgage-backed securities. Figure 2. Description of risk factor excess returns Factor Duration Credit MBS Curve Rates volatility Unit of excess returns US Treasury Index over cash (1M OIS), per year of duration US Corporate Investment Grade Index, over durationmatched Treasuries, per year of spread duration US MBS Index, over duration-matched Treasuries, per year of spread duration US Treasury 1-3 year Index minus US Treasury 1+ year Index (duration-neutral) over cash, per year of duration Delta-hedged ATM 1 month x 1 year swaption straddle sale Source: Barclays for Duration, Credit, MBS, and Curve, and Credit Suisse for Rates Volatility. While there could be factors not included in this list that are utilized in the space, in our view, these risk factors represent the major factors to which fixed income managers are exposed. The economic justification for selecting these particular risk factors is that they have delivered a risk premium over long periods of time. Figure 3 shows the excess return, correlations of these returns and Sharpe ratio of each factor over the last decade. All these factors had positive Sharpe ratios over the last decade. This is reassuring, since many of these factors are true economic risks and investors should be compensated for bearing these risks. 3 For comparison, we have also included return statistics of the S&P 5, since equity risk is both the dominant risk in investment portfolios and should deservedly serve as a benchmark risk factor in any asset allocation exercise. In Figure 3 we also show the correlation matrix of the risk factors based on monthly return data from January 25 to November 215. Returns from credit, MBS, and rates volatility sales are positively correlated, as they all represent risk-on exposures, co-varying positively with general risk assets in the economy. Over the time period used in this study, duration has behaved as a risk-off, or defensive, factor (negatively correlated with the spread factors). Lastly, curve has a negative correlation with duration over this time period. This implies that higher rates were accompanied by a steeper curve which is the opposite of the longer term behavior of yield curves which tend to steepen when rates are declining. This behavior is largely due to fact that the front end of the yield curve has been pinned down at the zero bound since late 28. FACTOR EXPOSURES OF ACTIVE FUNDS: A UNIVARIATE VIEW To provide intuition on the exposures of funds, we first look at simple correlation analysis. In Figure 4, we show the average correlation of fund excess returns with each risk factor as well as the 25th, 5th and 75th percentiles over the entire time period. With the exception of duration, the average correlation between funds excess returns and a given risk factor is positive in all cases. We find that the excess returns of 86% of the funds we have analyzed exhibit a positive correlation with credit and the average correlation is 42%. Moreover, 25% of funds have a correlation of 73% or more with this factor. These correlations are suggestive of the important role that exposure to the credit risk factor plays in active fixed income management. Positive correlations of fund excess returns with MBS and rates volatility

4 4 April 216 Quantitative Research Figure 3. Risk factor correlations (January 25 November 215) Avg. excess returns (%, per year) Sharpe ratio Duration Curve Rates volatility Correlation Duration Curve Rates vol Credit MBS S&P 5 Credit MBS S&P Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 are consistent with the hypothesis that exposure to these risk factors has also been an important source of outperformance for fixed income managers, but to a lesser degree than credit. The negative correlation of fund excess returns with duration is somewhat surprising at first blush. Although detailed results we present later largely confirm a systematic duration underweight for our set of managers, it is important to note that a negative univariate duration correlation may be caused by the strong negative correlation that duration returns have with the credit risk factor. As we show in the next section, a large fraction of reported alpha can be explained by structural credit exposure and thus, one would expect univariate duration betas to be negative for this set of managers even if the analytical duration exposures were neutral. To demonstrate in an intuitive yet powerful way the extent to which exposure to credit drives fixed income fund excess returns, we plot in Figure 5 the average manager excess return versus the return to a portfolio which is long two years of the credit factor. The correlation between the two return series is.93 and the commonality is visually striking. Investors likely operate under the assumption that core bond managers dynamically shift their credit exposure depending on perceived value in the sector at any given time. The correlation results in Figure 4 and the rolling performance chart in Figure 5, however, are inconsistent with this notion and rather indicate that a structural exposure to credit risk is in fact the dominant active tilt amongst active core bond managers. Interestingly, during the 28 crisis period, the average manager significantly underperformed a passive portfolio of two years of credit spread duration, but then nearly made up all of this underperformance during the 29 market recovery. This anecdotal observation may indicate that managers are exposed to latent risks which are not obvious in the majority of time periods, such as liquidity risk. While we don t pursue this matter here, we believe it is an interesting area for future research. Figure 4. Correlation of fund excess returns and risk factor performance for select correlation percentiles (January 25 November 215) Average 25th percentile Median 75th percentile % of funds with positive correlation Duration -39% -52% -42% -3% 7% Curve 17% 3% 15% 35% 77% Rates volatility 15% -3% 21% 35% 72% Credit 42% 16% 51% 73% 86% MBS 23% 6% 31% 47% 79% Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215

5 April 216 Quantitative Research 5 Figure 5. Average fixed income fund excess returns (over benchmark) vs. two years of passive credit factor exposure (rolling 12-months, January 25 November 215) Percent (%) Manager excess return (equally weighted) Excess return of 2 units credit factor exposure Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 MULTIVARIATE EXPOSURES In this section we run the following multivariate time-series regression for each fund i: rr ii,tt rr bb,tt = αα ii + ββ ii,dddddd. rr DDDDDD,tt + ββ ii,cccccccccccc. rr CCCCCCCCCCCC,tt + ββ ii,mmmmmm. rr MMMMMM,tt + ββ ii,cccccccccc. rr CCCCCCCCCC,tt + ββ ii,vvvvvv. rr VVVVVV,tt + εε ii,tt Where r i,t represents the after-fee monthly total return of fund i at time t, r b,t is the total return of the benchmark, α i is the average return not explained by the five risk factors, and the various β i,j represent regression-based exposures of fund i to the j th factor. The regression is run using monthly data from January 25 to November 215, except for dead funds where we use data up through the last month for which returns are available. In Figure 6a, we present estimates of regression coefficients, t-statistics and R-squares averaged across the funds in our sample. Additionally, we show each variable averaged within each information ratio (IR) quartile, in order to show how factor exposures vary by risk-adjusted performance. We exclude funds Figure 6a. Monthly alpha and active factor exposure estimates of funds, averaged in quartiles of full-sample information ratio (January 25 November 215) Realized IR Alpha (bp/yr) Duration Curve Rates vol Credit MBS R 2 Factor exposures Units - Yrs dur Yrs dur % MV Yrs spd dur Yrs spd dur - Trimmed average* % Bottom quartile % 2nd quartile % 3rd quartile % Top quartile % t-statistics Trimmed average Bottom quartile nd quartile rd quartile Top quartile Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 We report Newey-West (1987) t-statistics, which adjust for autocorrelation in risk factor returns. *Trimmed average is computed after excluding the top and bottom 5% funds by full-sample information ratio.

6 6 April 216 Quantitative Research in the top and bottom 5% IR, so as to eliminate the effect of outliers on our aggregate results. We define the IR as the ratio of average annualized excess returns over the benchmark to the annualized volatility of these excess returns. Our regression model explains roughly half of the variance of the returns of funds in excess of their benchmark. The average regression R-squared is 55% and this number is similar across IR quartiles. We also present the distribution of univariate betas in Appendix 1, and find that average univariate exposures are broadly consistent with our multivariate results. The long exposure to credit that we documented on a univariate basis is borne out in our multivariate analysis as well. On average funds are overweight 1.25 years of credit spread duration and this overweight is statistically significant, with an average t-statistic of 3.6. Importantly, we find that credit spread exposure increases monotonically with realized information ratios, ranging from.65 for the lowest quartile managers to 1.94 for the highest. Our results confirm those of Bosse, Wimmer and Phillips (213) who find that credit exposures have historically been an important driver of outperformance among U.S. bond funds. In Figure 6b we show the scatterplot of realized IR versus credit exposure for both live and dead funds. The upwardsloping line reflects the important role that credit risk has played historically in risk-adjusted performance of core bond managers. Consistent with our univariate analysis in the prior section, the average duration exposure is negative and statistically significant for all but the top quartile funds. Similar to credit exposure, duration exposure increases (although not monotonically) with realized IR. We find the duration underweight is statistically significant for all but the best-performing managers. Similarly, funds in the top IR quartile also have a significant exposure to the curve steepening trade. As shown earlier in Figure 3, positive exposures to the duration and curve factors are diversifiers to an overweight credit position, and have outperformed in the time period used in the analysis. These results point to the plausibility that more balanced portfolio construction has contributed to the relative risk-adjusted outperformance of funds in the top quartile. Figure 6b. Credit factor exposure vs. information ratio Information Ratio Live funds Credit factor exposure (yrs spd dur) Dead funds Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 Live funds are presented in blue, dead funds are presented in red To confirm the stability of our results, we present a robustness analysis in Figure 6c. We show trimmed averages of factor exposure estimates and t-statistics for live and dead funds separately. We find no notable variation in the average factor exposures of these two groups of funds. Not surprisingly, we find that estimates of alpha for dead funds are significantly lower than the live funds. IMPLICATIONS FOR ASSET ALLOCATION AND RISK MANAGEMENT An obvious interpretation of these results is that active returns are often dominated by beta, or structural tilts to known factor premia. For a large proportion of fixed income funds, there is some truth to this assertion. However, certain factor exposures, such as curve and rates volatility, cannot be easily accessed in the market. Therefore a manager s exposure to these factors could arguably be more valuable than exposure to standard factors such as duration. Importantly, it is only with hindsight that we know that these factors delivered superior returns over the last decade. Moreover, skilled fund managers have expanded the search for structural tilts beyond our five factors, to include, for example, emerging market credit and duration, non-agency

7 April 216 Quantitative Research 7 Figure 6c. Robustness analysis of multi-variate results Alpha (bp/yr) Duration Curve Rates Vol Credit MBS R 2 Factor exposures Units - Yrs dur Yrs dur % MV Yrs spd dur Yrs spd dur - Trimmed average* % Live funds % Dead funds % t-statistics Trimmed average Live funds Dead funds Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215 We report Newey-West (1987) t-statistics, which adjust for autocorrelation in risk factor returns. *Trimmed average is computed after excluding the top and bottom 5% funds by full-sample information ratio. MBS, and currencies. 4 There are also nuanced risk factors within these broad betas that have the potential to deliver superior risk adjusted returns. Examples of such exposures include shortduration credit and the corporate bonds of BB-rated companies. Historically, these factors have outperformed the broad credit index, in part due to compensation for exposure to incremental jump-to-default risk in the case of short-duration credit and clientele effects in the case of BB corporates. Our findings of a structural overweight to credit in active fixed income portfolios have important implications for overall portfolio risk. When a manager takes active risk factor tilts, the result is increased total exposure to the risk factor, and more often than not, greater portfolio risk. In the case of the credit factor, this issue is compounded by the fact that over the last two decades, the credit exposure of the Barclays US Aggregate index has risen significantly. In terms of market value, credit has increased from 17% of the index in June 199 to 3% today. Meanwhile, the average maturity of credit has extended. As a consequence of these two secular trends, the spread duration of the index has gone from.8 in June 199 to 2.1 today. While this secular trend has been apparent for over two decades, Figure 7a shows that the trend has continued through our analysis period which starts in 25. Since 25 average spread duration over this period has been 1.6, which is the spread duration that would roughly be implied by using index returns over that time period to model risk exposure. Also, the average rating has declined by over a notch from approximately A1 (Moody s nomenclature) to A2/A3. This downward trend in ratings is shown in Figure 7b and expressed in terms of both a letter grade and a numerical credit rating score.

8 8 April 216 Quantitative Research Figure 7a. Contribution of credit sector to U.S. aggregate market value and duration ( ) Credit as a % of Barclays U.S. aggregate Credit contribution to U.S. agg duration (yrs) Duration contribution of credit Credit market value weight Source: Barclays. As of 3 November Average duration contribution Figure 7b. Average credit rating score of U.S. credit ( ) Credit index rating number A3 A2 A Aa Source: Barclays POINT. As of 3 November 215 Full credit score numerical mapping provided in the Appendix The inescapable implication of this analysis is that asset owners are likely to get both a higher and riskier exposure to credit than expected based on using the historical average profile of the benchmark. This translates into higher volatility (especially in periods of financial market stress) and a higher beta to equities. In Figure 8 we show risk measures for the Barclays Aggregate Index computed on both an ex post (historical) and ex ante (forward-looking) basis using data over the past 1-plus years. Ex ante measures are computed using the current exposures of the benchmark and therefore reflect the secular increase in the credit exposure of the index. Additionally, ex ante results are presented before and after adjusting for the structural credit overlay of fund managers in our sample. We use 1.25 and 1.94 years of spread duration versus the index reflecting the average and top-quartile structural tilts, respectively. Baseline estimates are computed using monthly data over the period from January 25 through November 215, while stressed period estimates are based on data from January 28 through December 21. After adjusting for the structural credit overweight, the volatility increases from 4.1% to 4.4% for the average active core bond manager and 4.7% for the top quartile manager. In stressed conditions, volatility increases noticeably, from 5.8% to 6.5% and 7.2%, respectively. Similarly, the beta to the S&P 5 measured over the full sample (stressed) period rises from.2 (.6) to.6 (.11) and.9 (.14). The increase in volatility and equity beta estimates under stressed conditions highlights the impact of a credit overweight on the potential left-tail return properties of a portfolio. Overall, our risk analysis is likely to under-estimate the risk impact of structural factor tilts, for at least two reasons. First, as we have shown, most fixed income funds take risk-on structural exposures beyond credit, namely: short duration, and long MBS and rates volatility. Second, several individual funds have much higher credit biases than we report, including high yield exposures. During the crisis, a few leading fixed income funds experienced losses greater than 2% a much more extreme outcome than suggested by our analysis.

9 April 216 Quantitative Research 9 Figure 8. Estimates of volatility of the Barclays U.S. aggregate Index Realized (historical) volatility Without structural overlays Ex-ante volatility (as of November 215) Baseline estimates With average structural credit overweight With top quartile structual credit overweight Volatility (%, p.a.) Beta to S&P Stressed period estimates: January 28 - December 21 Volatility (%, p.a.) Beta to S&P Source: PIMCO. As of 3 November 215 Baseline estimates use data over the period January 25 November 215. Stressed period estimates use January 28 December 21 CONCLUSION This research provides evidence that the active returns of core fixed income managers are highly correlated to a set of basic risk factors, and in particular, credit. The highest performing managers in our study have more active duration and credit exposure than their lower performing counterparts. The beta to credit for the best performers is 3 times larger than that of the worst performers. In the full sample, credit exposure is by far the most dominant factor exposure. This analysis suggests that fixed income investors are likely earning a substantial proportion of the excess returns by taking (perhaps) unintended exposure to certain factors that impact the risk of their overall portfolio. An additional finding is the increasing role that credit exposure plays in not just the active portfolio but the benchmark as well. The benchmark weight in Credit has nearly doubled in the past 25 years while at the same time the average maturity of credit bonds has extended and the credit quality has declined. The combination of these effects implies a potentially greater impact of a structural tilt to credit than what we have seen historically.

10 1 April 216 Quantitative Research Appendix A - Distribution of univariate betas In each chart, we overlay a normal distribution with mean zero and standard deviation equal to that of the empirical distribution. These results confirm our simple correlation analysis. Also, they provide an estimate of the magnitude of exposures. For example, a beta of 1.5 can be interpreted as follows: if the factor return is +1%, the impact on the manger s excess return will be +1.5%. Figure A1. Univariate betas across managers (January 25 November 215) Duration Credit Mean: -1.3 Std. Dev.: Mean: 1.5 Std. Dev.: MBS Curve Mean: 2 Std. Dev.: Mean:.4 Std. Dev.: Rates Volatility Mean:.4 Std. Dev.: Source: PIMCO, Morningstar, Barclays and Bloomberg. As of 3 November 215

11 April 216 Quantitative Research 11 Appendix B - Credit rating score map Figure B1: Mapping of credit rating to credit rating score Credit rating Credit rating number Aaa 2 Aa1 3 Aa2 4 Aa3 5 A1 6 A2 7 A3 8 Baa1 9 Baa2 1 Baa3 11 Ba1 12 Ba2 13 Ba3 14 B1 15 B2 16 B3 17 Caa1 18 Caa2 19 Caa3 2 Ca 21 C 22 D 23 Source: Barclays References Bender, Jennifer P., Hammond, Brett, and William Mok Can Alpha be Captured by Risk Premia? The Journal of Portfolio Management, Vol. 4, No. 2: (Winter 214) Bosse, Paul M., Wimmer, Brian R., and Christopher B. Philips Active Bond-Fund Excess Returns: Is it Alpha or Beta?. Vanguard Research, September 213. Brown, David T. and William J. Marshall. 21. Assessing Fixed Income Fund Manager Style Performance from Historical Returns. The Journal of Fixed Income, Vol. 1, No. 4: (March 21) Dopfel, Frederick E. 24. Fixed-Income Style Analysis and Optimal Manager Structure. The Journal of Fixed Income, Vol. 14, No. 2: (September 24) Friedman, Eric, Otieno, Francois, and John Geissinger Measuring Success in Fixed Income. Hewitt EnnisKnupp, Aon plc. (November 212) Moore, James Sorry, Mr. Bogle, But I Respectfully Disagree. Strongly. PIMCO Viewpoint. (October 214). Newey, Whitney K., West, Kenneth D A Simple, Positive Semi-definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix. Econometrica, vol. 55, no. 3 (May1987): Worah, Mihir P., and Ravi K. Mattu The Secret of Active Portfolio Management. PIMCO Viewpoint. (October 212). 1 For two funds in our sample we utilized data from Bloomberg in order to fill in missing months in the early part of their returns series. 2 Throughout this article, when we refer to fund excess return, we mean returns in excess of the benchmark (Barclays US Aggregate Index). It is common practice to refer to these returns as alpha, but in this article, alpha refers to the intercept of the regression of excess returns on five factors. This alpha can be thought of as representing true manager skill after adjusting for structural risk factor exposure. 3 Our results indicate lower estimates of risk premia than those reported in Worah and Mattu (214), who use approximately 2 years of return history, whereas our study uses the past 1 years. 4 For a more complete discussion on active management in fixed income markets, see Moore (214).

12 Past performance is not a guarantee or a reliable indicator of future results. All investments contain risk and may lose value. Investing in the bond market is subject to risks, including market, interest rate, issuer, credit, inflation risk, and liquidity risk. The value of most bonds and bond strategies are impacted by changes in interest rates. Bonds and bond strategies with longer durations tend to be more sensitive and volatile than those with shorter durations; bond prices generally fall as interest rates rise, and the current low interest rate environment increases this risk. Current reductions in bond counterparty capacity may contribute to decreased market liquidity and increased price volatility. Bond investments may be worth more or less than the original cost when redeemed. Corporate debt securities are subject to the risk of the issuer s inability to meet principal and interest payments on the obligation and may also be subject to price volatility due to factors such as interest rate sensitivity, market perception of the creditworthiness of the issuer and general market liquidity. The credit quality of a particular security or group of securities does not ensure the stability or safety of the overall portfolio. This paper contains hypothetical analysis based on a set of assumptions that may or may not develop over time. Results shown may not be attained and should not be construed as the only possibilities that exist. Different weightings in the asset allocation illustration will produce different results. Actual results will vary and are subject to change with market conditions. There is no guarantee that results will be achieved. The analysis reflected in this information is based upon data at time of analysis. Forecasts, estimates, and certain information contained herein are based upon proprietary research and should not be considered as investment advice or a recommendation of any particular security, strategy or investment product. PIMCO routinely reviews, modifies, and adds risk factors to its proprietary models. Due to the dynamic nature of factors affecting markets, there is no guarantee that simulations will capture all relevant risk factors or that the implementation of any resulting solutions will protect against loss. Simulated risk analysis contains inherent limitations and is generally prepared with the benefit of hindsight. Realized losses may be larger than predicted by a given model due to additional factors that cannot be accurately forecasted or incorporated into a model based on historical or assumed data. The correlation of various indexes or securities against one another or against inflation is based upon data over a certain time period. These correlations may vary substantially in the future or over different time periods that can result in greater volatility. This material contains the opinions of the authors but not necessarily those of PIMCO and such opinions are subject to change without notice. This material is distributed for informational purposes only and should not be considered as investment advice or a recommendation of any particular security, strategy or investment product. Information contained herein has been obtained from sources believed to be reliable, but not guaranteed. PIMCO provides services only to qualified institutions and investors. This is not an offer to any person in any jurisdiction where unlawful or unauthorized. Pacific Investment Management Company LLC, 65 Newport Center Drive, Newport Beach, CA 9266 is regulated by the United States Securities and Exchange Commission. PIMCO Europe Ltd (Company No ), PIMCO Europe, Ltd Amsterdam Branch (Company No ), and PIMCO Europe Ltd - Italy (Company No ) are authorised and regulated by the Financial Conduct Authority (25 The North Colonnade, Canary Wharf, London E14 5HS) in the U.K. The Amsterdam and Italy branches are additionally regulated by the AFM and CONSOB in accordance with Article 27 of the Italian Consolidated Financial Act, respectively. PIMCO Europe Ltd services and products are available only to professional clients as defined in the Financial Conduct Authority s Handbook and are not available to individual investors, who should not rely on this communication. PIMCO Deutschland GmbH (Company No , Seidlstr a, 8335 Munich, Germany) is authorised and regulated by the German Federal Financial Supervisory Authority (BaFin) (Marie-Curie-Str , 6439 Frankfurt am Main) in Germany in accordance with Section 32 of the German Banking Act (KWG). The services and products provided by PIMCO Deutschland GmbH are available only to professional clients as defined in Section 31a para. 2 German Securities Trading Act (WpHG). They are not available to individual investors, who should not rely on this communication. PIMCO (Schweiz) GmbH (registered in Switzerland, Company No. CH ), Brandschenkestrasse 41, 82 Zurich, Switzerland, Tel: The services and products provided by PIMCO Switzerland GmbH are not available to individual investors, who should not rely on this communication but contact their financial adviser. PIMCO Asia Pte Ltd (51 Orchard Road #9-3, Wheelock Place, Singapore 23888, Registration No K) is regulated by the Monetary Authority of Singapore as a holder of a capital markets services licence and an exempt financial adviser. The asset management services and investment products are not available to persons where provision of such services and products is unauthorised. PIMCO Asia Limited (Suite 221, 22nd Floor, Two International Finance Centre, No. 8 Finance Street, Central, Hong Kong) is licensed by the Securities and Futures Commission for Types 1, 4 and 9 regulated activities under the Securities and Futures Ordinance. The asset management services and investment products are not available to persons where provision of such services and products is unauthorised. PIMCO Australia Pty Ltd ABN , AFSL (PIMCO Australia) offers products and services to both wholesale and retail clients as defined in the Corporations Act 21 (limited to general financial product advice in the case of retail clients). This communication is provided for general information only without taking into account the objectives, financial situation or needs of any particular investors. PIMCO Japan Ltd (Toranomon Towers Office 18F, , Toranomon, Minato-ku, Tokyo, Japan 15-1) Financial Instruments Business Registration Number is Director of Kanto Local Finance Bureau (Financial Instruments Firm) No PIMCO Japan Ltd is a member of Japan Investment Advisers Association and The Investment Trusts Association, Japan. Investment management products and services offered by PIMCO Japan Ltd are offered only to persons within its respective jurisdiction, and are not available to persons where provision of such products or services is unauthorized. Valuations of assets will fluctuate based upon prices of securities and values of derivative transactions in the portfolio, market conditions, interest rates and credit risk, among others. Investments in foreign currency denominated assets will be affected by foreign exchange rates. There is no guarantee that the principal amount of the investment will be preserved, or that a certain return will be realized; the investment could suffer a loss. All profits and losses incur to the investor. The amounts, maximum amounts and calculation methodologies of each type of fee and expense and their total amounts will vary depending on the investment strategy, the status of investment performance, period of management and outstanding balance of assets and thus such fees and expenses cannot be set forth herein. PIMCO Canada Corp. (199 Bay Street, Suite 25, Commerce Court Station, P.O. Box 363, Toronto, ON, M5L 1G2) services and products may only be available in certain provinces or territories of Canada and only through dealers authorized for that purpose. PIMCO Latin America Edifício Internacional Rio Praia do Flamengo, 154 1o andar, Rio de Janeiro RJ Brasil No part of this publication may be reproduced in any form, or referred to in any other publication, without express written permission. PIMCO is a trademark of Allianz Asset Management of America L.P. in the United States and throughout the world. 216, PIMCO. Newport Beach Headquarters 65 Newport Center Drive Newport Beach, CA Amsterdam Hong Kong London Milan Munich New York Rio de Janeiro Singapore Sydney Tokyo Toronto Zurich pimco.com blog.pimco.com CMR