Risk management Making the difference in fixed income investing

Risk management Making the difference in fixed income investing Pictet Asset Management September 2014 For professional investors only

Overview Tough times lie ahead for fixed income investors. The ballooning of public debt across the developed world and the implementation of ultra-loose monetary policy in the US, Japan and Europe have forced market participants to abandon conventional ideas about risk and return. Complicating matters further is a weakening of market infrastructure. Tighter oversight of the banking industry has seen many large investment banks cut back on fixed income trading, causing a reduction in the number of market makers and a deterioration of trading conditions in the secondary market. That all this is occurring at a time when the primary bond market is witnessing unprecedented growth is a worrying prospect as it suggests that the asset class may become more vulnerable to bouts of volatility. Bond investing is arguably more risky than it has ever been. It is against this backdrop that the analysis, control and mitigation of risk are taking on a more prominent role in the management of fixed income investments. To achieve long-term investment success at a time when the fixed income market is undergoing significant upheaval, a deeper understanding of how risk manifests itself at the security and portfolio level is critical. Risk management Making the difference in fixed income investing 3

Fixed income risk management at Pictet Asset Management at the heart of the investment process Risk management has always been central to the fixed income investment process at Pictet Asset Management. Our Fixed Income Risk (FI Risk) management professionals are members of a standalone unit that is part of the fixed income investment team; it reports directly to the Global Head of Fixed Income. Each fixed income risk manager is a specialist within a specific fixed income asset class and works closely with portfolio management teams. By structuring the team in this way, risk managers develop a deeper understanding of the investment processes of the strategies under their supervision, which enables them to respond to the changing conditions in fixed income markets. The FI Risk team also includes quantitative analysts, who bring their own distinctive insights to the management of our bond investments. The FI Risk unit s duties span a broad spectrum of risk management activities, from portfolio risk analysis and performance monitoring and attributions to the provision of insightful and timely quantitative research that guides investment positioning. The team s approach to managing risk is guided by three main principles: - Portfolios must be managed according to specific tracking error (TE)/ Volatility and/or Value-at-Risk targets - Risk targets are set according to a client s own tolerance for risk - Diversification of risk is a key objective for portfolio managers The team is responsible for evaluating and managing the risks associated with each investment decision, eliminating unintended risks from portfolios and ensuring investment professionals have a clear and holistic view of their portfolios risk profile. It also seeks to ensure that portfolios risk profiles are in keeping with both investment managers philosophy and the strategy s long-term return objectives. Over several years, the team has developed a set of customisable analytical tools that have been adapted to meet the constantly changing needs of our investment managers and clients. For all these reasons, our fixed income risk professionals play a major role in the construction of our bond portfolios. 4 Pictet Asset Management

A quantitative, investment-focused approach to risk It is the team s quantitative, investment-focused approach and the development and deployment of customised quantitative models that set our fixed income risk team apart from those of our peers. In recent years, there have been many occasions when the quantitative analysis provided by the FI Risk team has validated the investment stance taken, led to changes in a portfolio s positioning or been instrumental in the development of new strategies and risk measurement techniques. Timely market risk monitoring In one instance in the summer of 2013, our risk team detected an unusually abrupt shift in the risk profile of one of our global credit strategies. A deeper analysis revealed that the spike in the portfolio s annualised volatility and Value at Risk (VaR) stemmed from a tactical trade in credit default swap indices and credit default swaptions, contracts which give bond investors protection against the risk of default. The FI Risk team immediately informed portfolio managers of this deviation. As it turned out, the temporary shift in the portfolio s risk profile was intentional but required monitoring nevertheless. The investment and risk teams subsequently worked in tandem over the following days to ensure the risk exposure remained within target levels. FIG. 1 A SPIKE IN PORTFOLIO S SPREAD VOLATILITY PROVIDES A TRIGGER FOR HEIGHTENED RISK MONITORING Annualised Vol (in %) 8 7 6 5 4 3 2 1 0 Total Ex-Ante Vol (%) Interest rate Spread FX Vega 8 7 6 5 4 3 2 1 0-1 -1 01.2013 02.2013 03.2013 04.2013 05.2013 06.2013 Source: Pictet Asset Management, Data taken from analysis of EUR High Yield strategy Risk management Making the difference in fixed income investing 5

A multi-faceted risk approach In response to the changing needs of its clients, Pictet AM has launched a number of innovative fixed income strategies in recent years. Many of these give portfolio managers the discretion to invest across a broad range of fixed income markets and the freedom to express both negative and positive views on securities and asset classes. In order to provide investment managers of these funds with the clearest possible view of the risks in their portfolios, the FI Risk team uses several analytical tools that go beyond measuring a portfolio s TE or VaR. These range from those that measure a portfolio s gross and net exposure to more advanced metrics such as the duration time spread (DTS) and the DTS ratio (see glossary), which are used to gauge the portfolio s credit beta. Standalone tracking error is another measure that is regularly monitored to gauge the risks a portfolio is exposed to. Standalone TE breaks down a fixed income portfolio's risk into its individual components interest rate, inflation, spread etc and looks at each in isolation; in doing so, standalone TE assumes there is no correlation between these factors. It gives managers a clearer idea of the risk on a specific position when another one that diversifies it is closed. This risk measure enables the FI risk team to monitor the diversification within a portfolio from various angles. Portfolios are regularly submitted to stress tests to assess the impact of shocks on their returns. By modelling a number of possible future scenarios for each risk factor, the fixed income risk team can forecast changes in the value of the portfolio s investments and hence its overall gain or loss under different market conditions. FIG. 2 DTS ANALYSIS OF A CREDIT PORTFOLIO, BY SECTOR Basis points 600 500 400 300 200 100 0-100 -200 01.2014 02.2014 03.2014 04.2014 600 500 400 300 200 100-100 -200 Telecommunications Technology Sovereign Government Financial Energy Cons., Non-Cyclicals Cons., Cyclicals Capital Goods Basic Industry Index Other Sectors 05.2014 06.2014 07.2014 0 Source: Pictet Asset Management 6 Pictet Asset Management FIG. 3 DIVERSIFICATION INDICATOR FOR VARIOUS FIXED INCOME PORTFOLIOS Basis points Telecommunications Diversification 600 Technology ratio Sovereign 1.0 500 Pictet-EUR Government Bonds fund Pictet-World Government Bonds fund 1.0 Government 0.8 400 Pictet-EUR Inflation Linked Bonds fund PI CH-Foreign Bonds fund Financial 0.8 Energy 0.6 300 Cons., Non-Cyclicals 0.6 0.4 200 Cons., Cyclicals 0.4 Capital Goods 0.2 100 Basic Industry 0.2 0.00 Index Other Sectors 0.0-100 -200 01.2014 01.2014 02.2014 02.2014 03.2014 03.2014 04.2014 04.2014 05.2014 05.2014 06.2014 06.2014 02.2014 Source: Pictet Asset Management 07.2014

Practical quantitative guidance The FI Risk team has also developed models that allow investment managers to understand the behaviour of credit spreads, aiding portfolio construction. For our high-yield debt investment team, for instance, our quantitative risk specialists performed a detailed regression analysis of speculativegrade bonds risk premia. The analysts found that risk premium or spread was heavily influenced by the volatility of the European stock index, the default rate of high-yield bond issuers and the yield spread between peripheral European and German government bonds. The output from this model has now become part of the macroeconomic, or top-down, analysis conducted by the high-yield bond investment team. An active role in product development The FI Risk team also has a major part to play in the development of new fixed income products, helping formulate risk guidelines that serve the investment needs of both clients and portfolio managers. During the development phase of Pictet AM s new range of short duration fixed income strategies, the FI Risk team performed several analyses aimed at establishing investment parameters for the funds in the product family. For each of these strategies, the FI Risk team was able to provide detailed guidance on the suitability of investments by maturity, credit rating and currency. The analysis also encompassed the selection of each portfolio s reference index, VaR limits and country exposure caps. Risk management Making the difference in fixed income investing 7

Fixed income risk management inextricably linked to the investment process These positive interventions are possible thanks to Pictet AM s sophisticated risk management system, an integrated RiskMetrics platform. The system uses advanced risk methodologies and automated risk analytics that enable it to monitor and model the extent to which an individual security affects a portfolio s risk-return profile. It can also accurately determine how a portfolio s exposure to different countries, industry sectors and ratings categories can affect its returns and volatility. The FI Risk team has adapted and continously seeks to improve this system so that it meets the needs of our investment processes. The credit model, for instance, was recently enhanced to more accurately capture a portfolio s idiosyncratic risks. Each day, the team runs an average of 30 sets of risk analyses per portfolio that capture changes in a portfolio s risk profile along a number of different dimensions. These analyses are based on two types of daily risk reports the manager summary report and the more detailed risk attribution report. The manager summary report provides a snapshot of the portfolios risk exposure as indicated by their VaR and annualised TE/Volatility, grouped by investment team. Alerts are generated automatically whenever a portfolio s specific level of risk is reached. Through the more detailed risk attribution report, meanwhile, investment managers receive a comprehensive analysis of the sources of risk inherent in their portfolios an in-depth risk decomposition that captures the proportion of a portfolio s VaR or TE/Volatility that is attributable to interest rates, currency, yield spread, volatility and inflation. 1 1 The risk team uses a number of statistical tests to check the validity of the risk model including the Kupiec and Christoffersen tests. 8 Pictet Asset Management

FIG. 4 REGULAR RISK MONITORING BREAKING DOWN VOLATILITY BY RISK TYPE Ex-ante volatility, % 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0-0.5-0.5 01.2014 02.2014 03.2014 04.2014 05.2014 06.2014 07.2014 Vega FX Total Ex-Ante Vol (%) Spread Interest rate 10-Day VaR at 95% (%) Source: Pictet Asset Management; Data taken from analysis of Absolute Return Fixed Income Strategy FIG. 5 DECOMPOSING RISK BY CURRENCY AND RISK TYPE Ex-ante volatility, % 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0-0.5-0.5-1.0-1.0 01.2014 02.2014 03.2014 04.2014 05.2014 06.2014 07.2014 USD-SPREAD USD-IR USD-FX TRY-IR NOK-FX NOK-IR NZD-FX JPY-IR GBP-FX GBP-SPREAD GBP-IR EUR-FX EUR-SPREAD EUR-IR DKK-IR BRL-IR BRL-FX AUD-FX Other factors Source: Pictet Asset Management; Data taken from analysis of Absolute Return Fixed Income Strategy Risk management Making the difference in fixed income investing 9

Ensuring reliability of risk models - backtesting To ensure our VaR risk model is accurate and robust, its results are back-tested regularly. The VaR figures are checked against actual returns. Backtests are carried out for all Pictet AM s flagship funds every quarter. Figure 6 shows the outcome of these back-tests for one of our strategies. In this instance, the VaR model s predicted daily returns for the Pictet-Local Asia Emerging Market Debt fund are compared to the fund s actual returns. Under a VaR with a confidence level of 95 per cent, the proportion of daily returns falling outside the VaR figure should be on average 5 per cent for the model to be considered reliable. The green dots in Figure 6 show those instances when the actual daily return was in line with the estimate generated by the model. The orange dots below the red line indicate those occasions when the realised return was outside the predicted range. In this case, the back-test confirms the model was robust. FIG. 6 ABSOLUTE VAR BACKTESTING FOR AN EMERGING DEBT PORTFOLIO Daily Returns (%) 2.0 1.5 1.0 Abs 1D-VaR-95% Percentage of exceptions: 4.5% Basel Test Kupiec s p-value: 0.562 Christofferson s p-value: 0.096 2.0 1.5 1.0 0.5 0.5 0.0 0.0-0.5-0.5-1.0-1.0-1.5-1.5-2.0-2.0 07.2011 09.2011 11.2011 01.2012 03.2012 05.2012 07.2012 09.2012 11.2012 01.2013 03.2013 05.2013 07.2013 09.2013 11.2013 01.2014 03.2014 Source: Pictet Asset Management 10 Pictet Asset Management

Regular risk reviews The manager summary and risk attribution reports also serve as the basis for more detailed risk reviews between risk professionals and their investment counterparts. Every week, members of the risk team and investment professionals meet to discuss the results of the risk analysis of their respective portfolios. The reviews focus on metrics such as the historical evolution of a portfolio s risk profile by VaR and TE/Volatility the concentration of risk and the contributors to risk. Risk budgeting In addition, risk budgeting meetings are held twice a year with investment teams and the head of fixed income. They give the FI risk unit the opportunity to work with the investment teams to: - review the allocation of risk between the different sources of return for each portfolio over a set period - understand the reasons for any deviation from pre-determined risk budget targets. In our view, an effective way to dampen the volatility of returns and keep unintended risks to a minimum is to embrace diversification at every stage of the portfolio construction process. It is important to recognise that each source of return is also a potential source of volatility and risk. This is why the risk team work closely with investment managers to ensure the portfolio s risk budget is efficiently distributed across a portfolio s sources of return (currency, duration, spread, etc.) Performance attribution The risk team complements its risk analysis with the regular monitoring of portfolio performance. Portfolio returns are monitored on a daily basis and unusual deviations are analysed through a proprietary performance attribution system. Detailed performance attribution analysis is also provided to investment managers on a monthly basis. This ex-post review allows investment managers to review their risk allocation. A more comprehensive review is also carried out at least twice a year. In this process, Pictet AM managing partners, the head of fixed income, investment managers, and the fixed income risk team meet to analyse each portfolio s performance and risk exposure. FIG. 7 MONITORING FRAMEWORK Market risk management Performance monitoring and attribution Daily Control of risk limits (TE, VaR) Monitoring of performance to identify and analyse unexpected variations Weekly Monthly Semi-annual Risk reviews by investment teams Control of internal guidelines* Risk budgeting meetings by investment teams Quality review meetings with the top management Performance analysis Performance attribution adapted to each investment process (YTD, QTD, MTD) *Compliance department controls contractual guidelines and regulatory limits on a daily basis Source: Pictet Asset Management Risk management Making the difference in fixed income investing 11

Concluding remarks A risk management approach for a changing investment landscape The financial crisis and the monetary and regulatory shifts it unleashed have turned fixed income investing into a riskier enterprise. At Pictet Asset Management, understanding and managing risk has always been a fundamental part of the fixed income investment process. The strength of our risk management stems from two factors the first is that our fixed income risk management team has developed a quantitative investment-focused risk management approach that is deeply embedded in the investment process. The second is that such analysis is conducted to ensure that investment decisions are subjected to rigorous and impartial scrutiny. In a period likely to be characterised by increased market volatility, this in an approach that benefits our investment managers and, ultimately, leaves investors better placed to meet their long-term investment objectives. TEAM PROFILE The FI Risk unit is a team of eight. The current head took responsibility for building the team in 2000, while most of the team members have been working together for at least six years. Over time, this stability has enabled the team to develop a robust set of tools specifically designed for the analysis of fixed income portfolios, which they have been adapting to respond to the changing conditions in fixed income markets. 12 Pictet Asset Management

Glossary of risk management metrics The Value-at-Risk (VaR) of a portfolio is an estimate of the maximum capital loss that the portfolio can be expected to experience within a specified holding period, and with a specified confidence level. Instead of expressing the loss in absolute terms, Relative VaR measures that loss relative to the benchmark loss over the period. The FI Risk unit VaR is based on a Monte Carlo method in which 1,000 possible scenarios are simulated. Tracking Error (TE) quantifies the degree to which the performance of a portfolio differs from that of its benchmark. It is defined as the ex-ante standard deviation between portfolio returns and the benchmark returns, based on a parametric methodology, which assumes that portfolio returns are normally distributed. Effective Duration is the approximate change in a bond s price that will result from change in its yield. It is used for both conventional bonds and those with embedded options or redemption features. Spread Duration measures the price sensitivity of fixed income investments to shifts in their yield spread (the difference in yield between the security and benchmark government debt). Effective Convexity measures the effective duration change that results from a change in a bond's yield. DTS (Duration Times Spread) captures the sensitivity of a bond s price to a relative change of its yield spread. It is calculated by multiplying the spread duration of a bond by its credit spread (in basis points). The DTS ratio compares the fund s DTS to that of its reference benchmark. The diversification indicator takes a value between 0 and 1, with 1 being a perfectly diversified portfolio. (1-{Total Risk / sum of stand-alone risk by category}) Basel test: The Basel Committee recommends a recording of daily exceptions of the 99% VaR over the last year. 1 percent of 250, or 2.5 violations are expected on average over the last year. The Basel Committee has decided that up to four exceptions are acceptable, which defines a green light zone. Five or more exceptions and the fund falls into a yellow or red zone Kupiec s p-value: The Kupiec test is a statistic test that verifies if the observed frequency of violations (exceedances of VaR) is close to the expected number of exceptions. The VaR Model fails the Kupiec test if the p-value of the defined statistic is below a specific value. This test is unconditional as it does not take into consideration whether the exceptions are independent or not. Christoffersen s p-value: This test supplements the Kupiec test by testing the independence of exceptions. It formalises the notion that when exceptions are not independent, the probability of an exception tomorrow, given that there has been an exception today, is no longer equal to the confidence level. Risk management Making the difference in fixed income investing 13

14 Pictet Asset Management

Contacts For further information, please visit our websites: www.pictet.com www.pictetfunds.com This material is for distribution to professional investors only. However, it is not intended for distribution to any person or entity who is a citizen or resident of any locality, state, country or other jurisdiction where such distribution, publication, or use would be contrary to law or regulation. Information used in the preparation of this document is based upon sources believed to be reliable, but no representation or warranty is given as to the accuracy or completeness of those sources. Any opinion, estimate or forecast may be changed at any time without prior warning. Investors should read the prospectus or offering memorandum before investing in any Pictet-managed funds. Tax treatment depends on the individual circumstances of each investor and may be subject to change in the future. Past performance is not a guide to future performance. The value of investments and the income from them can fall as well as rise and is not guaranteed. You may not get back the amount originally invested. This document has been issued in Switzerland by Pictet Asset Management SA and in the rest of the world by Pictet Asset Management Limited, which is authorised and regulated by the Financial Conduct Authority, and may not be reproduced or distributed, either in part or in full, without their prior authorisation. For UK investors, the Pictet and Pictet Total Return umbrellas are domiciled in Luxembourg and are recognised collective investment schemes under section 264 of the Financial Services and Markets Act 2000. Swiss Pictet funds are only registered for distribution in Switzerland under the Swiss Fund Act; they are categorised in the United Kingdom as unregulated collective investment schemes. The Pictet Group manages hedge funds, funds of hedge funds and funds of private equity funds which are not registered for public distribution within the European Union and are categorised in the United Kingdom as unregulated collective investment schemes. For Australian investors, Pictet Asset Management Limited (ARBN 121 228 957) is exempt from the requirement to hold an Australian financial services licence, under the Corporations Act 2001. For US investors, shares sold in the United States or to US Persons will only be sold in private placements to accredited investors pursuant to exemptions from SEC registration under the Section 4(2) and Regulation D private placement exemptions under the 1933 Act and qualified clients as defined under the 1940 Act. The Shares of the Pictet funds have not been registered under the 1933 Act and may not, except in transactions which do not violate United States securities laws, be directly or indirectly offered or sold in the United States or to any US Person. The Management Fund Companies of the Pictet Group will not be registered under the 1940 Act. Copyright 2014 Pictet - Issued in September 2014 Risk management Making the difference in fixed income investing 15

www.pictet.com www.pictetfunds.com