BEST IDEAS FOR THE LIABILITY HEDGE PORTFOLIO

Transcription

1 BEST IDEAS FOR THE LIABILITY HEDGE PORTFOLIO

2 INTRODUCTION Plan sponsors continue to monitor closely the risk exposure of defined benefit retirement plans. We define risk as funded status volatility. The primary contributors to this risk are variances between the impact of interest rate and credit spread movements on the assets and liabilities, and exposure to equity and other growth asset exposure. Because interest rate and credit spread mismatches (when managing to a credit curve mandate) are generally considered to be uncompensated, there has been a strong push to manage these sources of risk by investing in assets that closely mimic liability characteristics. There are a number of methods to value the liabilities, including US GAAP and IFRS, PPA (with and without stabilization) and economic measures based on a risk free rate (i.e., Treasury yields or swap curve). In this paper we consider hedging a creditbased discount rate that is consistent with US GAAP. This paper highlights our views on how to construct a best-in-class hedge portfolio. This includes: recommended interest rate and credit hedge ratios for a generic plan; changes to the hedge portfolio along a dynamic derisking glide-path for a generic plan; portfolio construction including split of Treasuries and credit bonds and use; of derivatives such as interest rate swaps and Treasury futures; use of leverage in the liability hedging portfolio; choice of bond benchmarks; views on active and passive management of the liability hedging portfolio; changes to construction of the liability hedging portfolio in a low interest rate environment as well as construction reflecting an agnostic view on interest rates; monitoring of the liability hedging portfolio We term assets that mimic the liability characteristics the liability hedging assets. Assets that are intended to provide growth (equity, alternatives, real estate, etc) are referred to as growth assets. 1 1

3 OUR BELIEFS The liability hedging portfolio should be designed to mitigate the funded status volatility resulting from movements in interest rates and credit spreads. Note that how much of the volatility can be controlled through the liability hedging portfolio is dependant on several factors including the duration and size of the liabilities relative to the amount of liability hedging assets, the duration of the liability hedging assets and the funded status. We find it helpful to view the desired level of risk management as a function of the interest rate and credit spread hedge ratios relative to the liabilities, which are also referred to as the dollar duration impact. The primary drivers of movement in US liabilities calculated for IRS minimum funding requirements and US GAAP are changes in interest rates and credit spreads embedded in the yields on high quality corporate bonds. Consequently, our recommended liability hedging instruments are fixed income in nature. In constructing a liability hedging portfolio, an understanding of the characteristics of plan liabilities is critical. Plan benefit features range from simple (e.g., traditional annuity and flat dollar plans) to complex (e.g., cash balance, pension equity and traditional plans with lump sum options). Understanding how the liabilities for the more complex plan designs can be impacted in multiple ways by changes in interest rates and credit spreads is essential in correctly designing a liability hedging portfolio. Actuarial assumptions such as salary scale, mortality and other decrements and changes thereto also determine the term structure of the plan cash flows but cannot easily be hedged, so are not addressed here. We define a fair-value rates environment as one in which a long duration interest swap will be expected to have zero net return. The consequence of this type of environment is that the plan can lengthen the duration of liability hedging assets and remove more of the interest rate risk without any expected cost relative to shorter duration instruments. In a fair-value rates environment, we believe that most plans should remove interest rate risk to the maximum extent possible based on the liability hedging instruments available, and seek to generate alpha returns from more traditional growth asset classes with high expected returns and low correlation with liability movements (i.e. equities and alternatives). 2 2

4 THE EXTENT OF THE LIABILITY HEDGE There are three broad exposures that create volatility between the market value of the liability and the liability hedging portfolio: parallel shifts in interest rates, changes in credit spreads and changes in the shape (i.e., twists) of the yield curve. While each of these creates funded status volatility, the magnitude of the impact, and the importance of reducing the exposure, declines from left to right. Reduce volatility due to parallel movements in interest rates Reduce volatility due to credit spreads Reduce volatility due to non-parallel movements in rates The contribution of each factor to volatility is approximately 70%, 25% and 5% respectively. Given the relative size of the risk factors, the first objective is to structure the liability hedging portfolio to reduce interest rate volatility due to parallel movements in rates. We measure the amount interest rate exposure considering both the assets and liabilities using the interest rate hedge ratio. We use a calculation called dollar duration to measure the present value movement in the assets and liabilities under a one basis point parallel movement in the yield curve. The interest rate hedge ratio is the percentage of liability movement, in terms of dollar impacts, that is matched by the liability hedging portfolio. In determining the target interest rate hedge ratio, there are a number of factors that must be considered including (but not necessarily limited to): the split between growth and liability hedging assets (which determines the amount of physical liability hedging assets available), the funded status of the plan, the sponsor s attitude to risk and view on rates. There are limits to how much interest rate risk can be hedged solely through an allocation to physical liability hedging assets. Overcoming this requires the use of synthetic hedging instruments such as swaps or Treasury futures to create a leveraged exposure, where a dollar of physical assets provides more than a dollar of liability duration exposure. Some plan sponsors; however, are uncomfortable with the use of leverage to gain this exposure. In this paper, we consider the design of the liability hedging portfolio for both physical-only and unconstrained investors. To address where plan sponsors are invested today and where they are likely to move to, we have found it helpful to describe our thoughts on the composition of the liability hedge portfolio along a generic plan glide-path trajectory. To simplify the discussion, we have identified three summary points along the glide path Early, Middle and Late which are further defined below: Early Middle Late Funding ratio <80% 80-95% 95%+ Allocation to liability hedging assets <50% 50-75% 75%+ 3 3

5 We have calculated the funded status volatility of a generic plan, given varying interest rate hedge ratio targets, for the Early, Middle and Late phases. Our results are shown in the following chart. Funded status volatility (% of liablity) 14% 12% 10% 8% 6% Maximum hedge ratio using physical assets. Higher hedge ratio achieved with unfunded swaps. Diminishing improvement in risk reduction over ~50% hedge ratio 4% Early stage Diminishing improvement in risk reduction over ~70% hedge ratio Middle stage Material improvement in risk 2% reduction up to about 90% hedge ratio Late stage 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Hedge ratio Source: Mercer calculations of an example plan using Mercer s Capital Market Assumptions There are a number of observations and conclusions: All else equal, the risk, as measured by the funded status volatility, decreases along the glide-path due to better funding and less allocation to growth assets. Risk decreases as the interest rate hedge ratio increases. Holding only physical liability hedging assets can constrain the hedge ratio, especially in the case where the plan is in the Early Phase. In all cases, there is very little additional risk reduction beyond the 90% interest rate hedge level. Even though a 100% hedge ratio may be theoretically desired, it may not be worth the incremental costs to achieve. There comes a point where increasing the hedge ratio does not continue to materially decrease the funded status volatility. These points are broadly 50% in the Early Phase, 70% in the Middle Phase and 90% in the Late Phase. Adopting a hedge ratio beyond this point may start to be less efficient, although risk reduction is still possible with higher hedge ratios. The decision to extend the hedge ratio to a point where there is diminishing risk reduction will depend on the view on interest rates, a desire to employ leverage and return requirements. 4 4

6 We summarize our findings for the target hedge ratio along our generic glide-path. Minimum target interest rate hedge ratio Early stage 50% Middle stage 70% Late stage 90% The above analysis considers only the hedge ratio and only addresses the impact on funded status volatility of parallel movements in interest rates. It does not consider movements in credit spreads and/or nonparallel movements in interest rates. We address these now. After dollar duration hedging, the next most important consideration is the management of the credit spread volatility. Based on historical data, this is around one-third of the volatility of the interest rate volatility 1. The most desirable liability hedging spread instrument is a long duration high quality bond; however, we recognize that there is also a correlation between credit spread and equities. Considerations in hedging the credit spread risk are discussed further in the section on Portfolio Structuring. We use the credit spread hedge ratio to measure the level of hedging of the credit spread risk. A final objective is to structure the liability hedging portfolio to remove interest rate volatility that arises from non-parallel movements in the yield curve. Exposure to non-parallel movements, illustrated below, is best expressed through comparing key-rate durations. Parallel move Twist Butterfly We recommend that the liability hedging portfolio be structured to maximize the interest rate hedge ratio as the primary objective, using the various approaches discussed hereafter to attempt to reduce key rate imbalances. There is a trade-off in levering up the hedge ratio by investing in assets with a longer duration than the liabilities and maintaining an adequate duration at each key rate. We are not prescriptive of the quantitative level of any imbalance but note that there are mismatch risks in carrying an asset portfolio that has a significantly higher duration than the liabilities and have seen environments, including the Fed s Operation Twist, which negatively impacted the performance of the liability hedging assets due to the mismatch. 1 Source: Mercer calculations and data from Bloomberg LP. We would also employ Duration times Spread (DTS) to measure the degree of credit spread risk, especially at the Late stage and where a manager may tilt to higher credit spread risk than the benchmark 5 5

7 A practical example of the tradeoff between duration matching and curve fitting arises in developing a liability hedging portfolio along a de-risking glidepath. At the start of the glide-path, where there is a low allocation to physical bonds, there is the need to lengthen the liability hedging portfolio duration, and we would anticipate in fair-rates conditions a liability hedging portfolio duration in excess of the liability duration. However, this is done at the expense of curve fitting as it is generally difficult or impossible to, simultaneously, increase duration and also match key rate durations to improve the curve fit with any precision. As the glide-path enters the Late stage, where the liability matching portfolio comprises of mostly bonds, then the asset and liability durations will converge and curve matching along the key rates will become more viable. PORTFOLIO STRUCTURING The liability hedging portfolio should be diversified across credit exposure as well as term structures. This is particularly important in the later stages of a glide-path when the plan becomes fully funded and a significant percentage of the portfolio is in liability hedging assets. We favor the following liability hedging assets for their duration and convexity characteristics: medium and long credit bonds, medium and long US Treasury bonds, STRIPS, interest rate swaps where the plan is a fixed receiver, and Treasury futures. We also see a place for swaptions in expressing a view on interest rates. Early Stage. At this point, the growth portfolio is significant and consequently there is a meaningful amount of equity exposure. Given a positive correlation of equity returns and credit spreads, the preference is to concentrate on managing the interest rate risk within the liability hedging portfolio in preference to the credit spread risk. The liability hedging portfolio that reduces overall funded status volatility will usually have a high weighting to long Treasuries or STRIPS in preference to long credit (see figure below). For plan sponsors comfortable with the use of leverage, this may also involve the use of swaps and/or Treasury futures if valuations are appropriate. Surplus volatility 10.5% 10.0% 9.5% 9.0% All Treasury Minimum risk state has substantial Treasury allocation All long credit Source: Mercer calculations of an example plan using Mercer s Capital Market Assumptions. Equity allocation 60%, 80% funded and liability duration 12 years. 6 6

8 Note that hedging interest rate risk primarily through the use of Treasury securities comes at a cost since the returns generated are not sufficient to keep pace with the growth in the corporate bond based liability. For sponsors who are sensitive to the need to generate more returns, including those influenced by the use of a higher expected return on asset assumption in their US GAAP reporting, a large Treasury allocation can be challenging. It is therefore important to quantify the trade-off in risk reduction with the impact on expected returns. This is illustrated in the following figure. GAAP Expected return 7.3% 7.3% 7.2% 7.2% 7.1% 7.1% 7.0% 7.0% 6.9% 6.9% All long Treasury 9.4% 9.5% 9.6% 9.7% 9.8% 9.9% 10.0% 10.1% 10.2% 10.3% 10.4% Funded status volatility All long credit Note: GAAP expected return is illustrative and will vary from plan to plan Middle Stage. As the plan transitions from growth to liability hedging assets, we expect the composition of the liability hedging portfolio to change. We would expect the relative weighting to change from Treasuries to credit bonds to more closely align with the credit based discount rate (see figure below). This is a function of the amount of credit spread hedging provided by equity type assets declining, and the need for the accuracy of the liability hedge, including credit spread hedging, increasing. We would also expect to be able to generate the desired hedge ratio using physical assets with less duration than in the Early Stage, and the allocation to very long duration bonds including STRIPS can be reduced. Minimum risk state has mixture credit and Treasuries Source: Mercer calculations of an example plan using Mercer s Capital Market Assumptions. Equity allocation 40%, 90% funded and liability duration 12 years. 7 7

9 Late Stage. In the Late stage of the glide-path, we expect the growth portfolio to be relatively small and there is a greater need to manage the credit spread risk and to also focus on the curve risk. Consequently, we expect the liability hedging portfolio to be largely corporate bonds that closely matches the duration of the liabilities and the key rate profile. We believe the most effective way to manage the key rate profile is through the use of derivatives including interest rate swaps. The liquidity in these instruments provides more flexibility than the use of physical bonds alone. Regardless of the instruments used, it can be difficult to manage both interest rate and credit spread hedge ratios in excess of 90% and some compromise may be necessary. It is also likely that a plan in the Late stage will be mature, so liquidity for benefit payments might be required. Consequently, we might expect a holding of Treasuries to provide this liquidity. All long Credit Source: Mercer calculations of an example plan using Mercer s Capital Market Assumptions. Equity allocation 10%, 100% funded and liability duration 12 years. The following figure shows an illustrative liability hedging portfolio that we might expect in a fair-value rates environment. The chart highlights the evolution of the liability hedging portfolio for a plan with duration of 12 years, through the Early, Middle and Late stages of de-risking. In the Early stage, a high allocation to long duration bonds and ultra-long Treasuries significantly improves the hedge ratio, reducing the interest rate risk exposure. As discussed in the Late stage, it becomes more important to hedge credit duration exposure as well as the key rate profile. As a result, Treasury / STRIPS exposure is systematically decreased while credit exposure is increased. At the Late stage, credit duration exposure becomes relatively important and therefore requires most of the liability hedging portfolio to be credit bonds. Note also that particular attention is paid to aligning with the term structure profile of the plan cash flows as accurately as possible. In the Late stage we also put more emphasis on active credit management, and this is addressed further in Section STRIPs Long Government Intermediate bonds Long Corporate Early stage Middle stage Late stage 8 8

10 Early Middle Late Funding ratio <80% 80-95% 95%+ Allocation to hedging assets: < 50% 50-75% 75%+ Minimum target interest rate hedge ratio 50% 70% 90% Target credit spread hedge ratio 60% 60-80% 80%+ In implementing the liability hedging portfolio, the use of standard bond mandates and benchmarks is generally reasonable, particularly in the Early and Middle stages; however, the limitation of standard benchmarks must be understood. This is summarized in the figure below. Long Corporate AA Long Corporate Long Credit Long Gov/Credit 2,158 Market value ($B) High concentration of issuers ,255 19% 15% Concentration of financial sector 8% 9% High concentration of financials 55% Concentration of noncorporates 0% 0% 22% Investment grade emerging sovereign debt Basis risk with the liability High Similarity to liability Low From the figure above, it is apparent that there is a trade-off between the liability matching characteristics of the bond benchmark and the concentration of individual sector and security risk. The most liability-like benchmark has the fewest bonds, and the most diverse bond benchmark potentially has the greatest basis risk to the liabilities. In most cases in the Early stage, we would recommend a broader benchmark to capture the duration characteristics of the liability and provide a Treasury allocation that ensures the plan is not overweight to spread movements from equities and credit. If the plan was additionally lengthening the duration then it seems likely that a long STRIPs mandate would be employed requiring a separate benchmark. 9 9

11 In the Late stage, we would recommend a benchmark that closely mimics the liabilities. This can be accomplished by assigning a custom liability benchmark to one or more managers or by employing managers on standard investable benchmarks (albeit with possible caps etc.) and using a completion manager that would have a mandate related to managing mismatches in the key rate profile. This is further described in the Manager Structuring section below. MANAGER STRUCTURING Active/passive manager decisions should be made in the context of the allocation of active management risk in a risk budgeting exercise; however, we generally support active management of the hedging portfolio to help it keep pace with the liabilities. A particular issue is the impact of credit downgrades, which negatively impact the actual liability hedging portfolio, but do not have an impact on the theoretical liability portfolio used to construct yield curves for valuing pension liabilities. If the intent is for the liability hedging portfolio to keep up with liability returns, then some degree of active management will be needed. The goal is generally to win by not losing i.e., defensive alpha. This is generally a strategy to minimize the exposure of the portfolio to defaults and downgrades of credit. We generally do not support active management based on significant duration positions as this degrades the purpose of the liability hedging portfolio. To the extent a sponsor wishes to take a position in interest rates, this should be reflected in the hedge ratio as described above that informs manager mandates and benchmarks, rather than interest rate positions being taken by the managers themselves. Consequently, active Treasury or swaps management should be considered carefully, based on individual managers and client goals. Generally, we see the most value in active management of the credit quality allocation and issuer selection to minimize exposure to defaults and downgrades. We recommend multiple active credit managers to diversify their credit screening methods. The precise number of managers will depend on the size of the portfolio, complementary nature of the mangers investment process, governance budget and appetite for complexity. Passive management is a viable approach, particularly for Treasury mandates. For mandates that include credit, active management is preferred to manage downgrade risk. Given the above points, one possible approach might be to split a Barclays Long Govt/Credit mandate into an active credit mandate and a passive Treasury mandate. However, complexity, governance budget and costs will need to be factored into this decision. To the extent that a passive bond management provides greater liquidity, plan sponsors with a relatively high benefit pay out ratio could find this beneficial and it should be considered. Some fixed income managers specialize in adding excess returns in various ways, especially in areas such as high yield or emerging market debt. These asset classes can play a role in portfolio construction and may be an out of benchmark active management decision. However, strategic allocations to these areas should be included in the growth asset category rather than in the liability hedging assets. We support limiting the credit quality for clients requiring very tight risk controls and this can be most relevant for Late stage implementation. For example we might seek to cap a manager s BBB holdings in a long credit mandate to limit the basis risk with the liabilities

12 The following shows an illustrative example in the Late stage. Growth portfolio 10% Hedge portfolio 90% Completion manager Likely to be passive growth exposure with possible illiquidity in private equity Blend of active credit managers to diversify credit screening methods and exposure. Use of custom benchmark Potential passive credit manager to limit fees Potential small Treasury allocation for tail-risk protection. Recommend passive strategy with no duration bets that could compromise hedge Only one manager required to coordinate, distribute and fill-in the gaps in the liability profile e.g. passive swap strategy 6 Use of completion manager to "fill-in" key rates 120% Short position adopted by completion manager (not shown) Long position adopted by completion manager Key rate hedge ratio z 100% 80% 60% 40% 20% 0% 120% 130% 70% 50% 30% 20% Key rate (years) Physical managers Completion manager 45% 11 11

13 We illustrate some of these ideas with two example clients below, a simple client with a low governance budget and a complex client with significant resources to spend on the management of the liability hedging portfolio. Simple client Complex client Governance budget Low High Knowledge level of LDI Low to medium Medium to high Ability/desire to use synthetic assets e.g. swaps Ability/desire to make tactical changes to LDI portfolio based on rates movements Example Early stage manager roster Example Middle stage manager roster Example Late stage manager roster Low Low Single long duration benchmark with single manager. Passive or active management. Custom blend of benchmarks with manager selection split by benchmark. Possible active credit manager diversification. Performance of liability hedging portfolio measured against liability benchmark. Possible use of single manager (e.g. Mercer) as a completion manager. Alternatively, maintain blend of standard benchmarks and report performance against liability. High High Custom blend of benchmarks with multiple managers. Potential differentiation of active and passive managers (e.g. active credit and passive Treasury). Same as Early stage or transitioning to Late stage. Performance of liability hedge portfolio measured against liability. Potentially retain underlying managers with additional completion manager (e.g. Mercer) to manage to hedge ratios or to liability return. Performance reporting is comprehensive with an element for non-investability

14 HEDGE PORTFOLIO MONITORING We recommend that the liability hedging portfolio be monitored with reference to the change in liabilities and the funded status. It is clearly important to measure the performance of the managers relative to their benchmark, but a liability hedging portfolio needs steering at a strategic level and individual fund performance will not capture sufficient management information. We recommend that the change in assets and liabilities are attributed to changes in interest rates, changes in credit spreads, manager tracking error and yield carry mismatch. We would additionally expect the individual managers to be able to attribute their performance to sector, duration, and issuer. The following example illustrates the attribution of change in funded status and provides valuable information as to the performance of the liability hedging portfolio. $0 Surplus (Deficit) 12/31/2011 Yield Carry Mismatch Yield Credit Spread Change Change Mismatch Mismatch Hedge Portfolio Tracking Error Growth Portfolio Return Contributions / Benefit Service Cost Payments Expenses Other Gain (Loss) Surplus (Deficit) 12/31/2012 ($50) ($100) ($150) ($261) The negative effects of declining interest rates outweighed the positive effects of widening credit spreads in the discount curve during the quarter ($323) ($200) ($250) ($300) ($13) Negative equity returns widened the funding deficit during the quarter ($350) ($191) ($11) ($23) ($0.2) ($1) ($37) ($400) ($119) ($450) ($500) 13 13

15 HIGH-LEVEL SUMMARY The following table gives broad expected values for important metrics along the glide path in a fair-value rates environment. There will clearly be specific considerations for each plan and its sponsor. Early Middle Late Funding ratio <80% 80-95% 95%+ Allocation to growth portfolio 50% % <20% Minimum target interest rate hedge ratio 50% 70% 90% Credit spread hedge ratio 60% 60-80% 80%+ Surplus volatility 14-10% 10-6% <6% Key rate profile Low priority More important Very important Allocation to Treasuries Higher Medium Lower Allocation to long credit Lower Medium Higher Bond benchmark Single standard or blend of standard (with possible adjustments) Blend of standard (with possible adjustments) Blend plus completion manager for key rates 14 14

16 For further details on the issues discussed in this paper please contact the authors or your usual Mercer consultant. About the LDI (Liability Driven Investing) Committee Mercer s LDI Committee is comprised primarily of field consultants from Mercer s Financial Strategy Group, which includes pension risk management and investment specialists as well as members of the Investment Management practices and Mercer s Bond Research Boutique. The Committee seeks to provide a clientfacing perspective on research, new ideas, intellectual capital, and to serve as a resource for questions regarding this asset class and related consulting processes. ABOUT THE AUTHORS Gordon Fletcher is a Partner and an actuary working in Mercer's Financial Strategy Group in the New York office. He helps clients identify the extent of their pension risks and how best to manage them. He works on all areas of managing pension risk including LDI strategies, annuities, lump sums, derivative strategies and longevity swaps. He led Mercer's development of pension-buyout online auctions and advised the trustees of a plan that is the first known longevity swap of non-retired participants. He is currently chair of Mercer's LDI committee in the US. Gordon has a master's degree and PhD in mathematics and the Institute of Actuaries' Certificate in Derivatives. Prior to joining Mercer he held research positions in the Universities of Cambridge and São Paulo, and has published in the areas of pensions and mathematics. New York gordon.fletcher@mercer.com Ricky Kimalat is a Senior Associate and an actuary working in Mercer s Financial Strategy Group in the Chicago office. He works on all aspects of asset-liability analysis as well as risk transfer solutions, and assists clients on how best to opportunistically manage them. He is part of Mercer s LDI committee in the US. Ricky holds a bachelors degree in Mathematics. He is an Associate of the Society of Actuaries as well as a CFA charter holder. Chicago ricky.kimalat@mercer.com Published March

17 IMPORTANT NOTICE References to Mercer shall be construed to include Mercer LLC and/or its associated companies Mercer LLC. All rights reserved. This contains confidential and proprietary information of Mercer and is intended for the exclusive use of the parties to whom it was provided by Mercer. Its content may not be modified, sold or otherwise provided, in whole or in part, to any other person or entity, without Mercer s prior written permission. The findings, ratings and/or opinions expressed herein are the intellectual property of Mercer and are subject to change without notice. They are not intended to convey any guarantees as to the future performance of the investment products, asset classes or capital markets discussed. Past performance does not guarantee future results. Mercer s ratings do not constitute individualized investment advice. Information contained herein has been obtained from a range of third party sources. While the information is believed to be reliable, Mercer has not sought to verify it independently. As such, Mercer makes no representations or warranties as to the accuracy of the information presented and takes no responsibility or liability (including for indirect, consequential or incidental damages), for any error, omission or inaccuracy in the data supplied by any third party. This does not constain investment advice relating to your particular circumstances. No investment decision should be made based on this information without first obtaining appropriate professional advice and considering your circumstances. Investing involves risk. The value of your investment will fluctuate over time and you may gain or lose money. Certain investments, such as securities issued by small capitalization, foreign and emerging market issuers, real property, and illiquid, leveraged or high-yield funds, carry additional risks that should be considered before making an investment decision. Additional investment risks may apply to your particular investments. Consult the offering statement or prospectus for details on additional specific risks. Investment advisory services provided by Mercer Investment Consulting, Inc. Capital Market Assumptions provided by Mercer Investment Consulting, Inc. For a more complete description of the assumptions and methodologies used by Mercer in developing this information, please refer to the semi-annual Capital Market Outlook report. Please contact your investment consultant, who can provide you with a current version of this report

18 For further information, please contact your local Mercer office or visit our website at: Argentina Australia Austria Belgium Brazil Canada Chile China Colombia Czech Republic Denmark Finland France Germany Hong Kong India Indonesia Ireland Italy Japan Malaysia Mexico Netherlands New Zealand Norway Peru Philippines Poland Portugal Saudi Arabia Singapore South Korea Spain Sweden Switzerland Taiwan Thailand Turkey United Arab Emirates United Kingdom United States Venezuela Copyright 2013 Mercer LLC. All rights reserved.