Liability Driven Investing: Finding Your Match

Institutional Group Driven Investing: Finding Your Match Customization and Active Management are the Keys to Success As the end of 2014 nears, many defined benefit (DB) pension plan sponsors are breathing a sigh of relief as they witness another year of funding level improvements, albeit with a few bumps along the way. With most plans near or at fully funded status, pension committees are implementing or contemplating de-risking strategies such as liability driven investing (LDI). Essentially, LDI creates the framework or mind-set to manage pension risk by ensuring investment decisions are based jointly on assets and liabilities. The goal is to generate the best possible match between assets and liabilities, and ensure the two sides move in the same direction. Regardless of the end game for the DB plan, (whether the plan sponsor is committed to maintaining an open plan, or seeks to exit the DB game through an annuity purchase or closing/freezing the plan) LDI establishes the road map to arrive at that destination with minimal volatility in cash contributions and/or accounting disclosures along the way. The roadmap, also known as the glidepath or journey plan, is in essence an active approach to asset allocation designed to shed risk at key trigger points which are based on funding thresholds. As the plan moves along the glidepath, the portfolio becomes more conservative. Approaching a new year, plan sponsors should be reviewing pension strategies for 2015, particularly in an environment where de-risking opportunities exist with the help of current strong equity returns and low interest rates. While there are many ways to implement an LDI strategy, this paper goes back to the basics in how to optimize the liability match. A skilfully crafted solution begins with understanding the major factors contributing to a mismatch (otherwise known as tracking error) and depends on customization and ongoing active management for success. 2014 SEI 1

4 6 8 1 1 14 16 18 2 2 24 26 28 3 3 34 36 38 4 4 44 46 48 5 5 54 56 58 6 6 Payments ($ Millions) Understanding the Mismatch Closing in on the gap between assets and liabilities begins with the basic understanding of how liabilities are calculated. The exercise begins with projecting the benefit payment stream over a given period of time, and then calculating the present value of the future cash flows by discounting the benefit payment stream against a yield curve of high-quality corporate bonds (Chart 1). Because the discount rate ties the benefit payment stream to the performance of interest rates, the plan s liabilities respond accordingly: when rates go down, liabilities go up and vice versa. Chart 1: Discounted Liabilities Using High-Quality Corporate Bonds 70 60 50 40 30 20 10 0 Benefit Payments Present Value Maturity The sensitivity of an asset to the movement of interest rates is known as the duration. To illustrate, a duration of 10 means that for every 1% change in interest rates, the asset value will change by 10%. In an LDI strategy, duration looks at the benefit payment streams and calculates the sensitivity of each stream to the discount rates. A benefit payment that is further in the future is going to have a higher sensitivity to a change in interest rates. Calculating sensitivity would be fairly simple if there were only one payment stream; however, a pension plan has many payment streams, all with different durations. The effective duration of a pension plan is the combination of all of these benefit payments discounted back to the present. To use this data efficiently, the present value sensitivities are aggregated into key rate duration buckets. The following example illustrates how liabilities and durations are calculated and the effects of a customized solution. In Chart 2, the sample pension plan is 100% funded and uses a 100% core fixed income portfolio LDI implementation. The fixed income portfolio has a duration of 4.4 years and the plan has a duration of 13.4 years. In the early years of this example, the fixed income portfolio provides good coverage and closely matches the liability, but as time progresses a mismatch develops, becoming significant in year seven. As a result, the plan has an effective hedge duration of 32.8% and a funded ratio volatility of 9.4%. This means that the 100% funded plan, with all of its assets invested in a portfolio of core fixed income securities, has an unpredictable and fairly volatile funded ratio of between 82.2% and 112% on an annual basis. Chart 2: 100% Fixed Income LDI Implementation Asset Duration 4.4 Years + Duration 13.4 Years =Effective Duration 32.8% Funded Ratio Volatility 9.4% Funded Ratio (90% Confidence Interval) 82.2-11% Core Fixed Income 2

3M 1Yr 2Yr 3Yr 5Yr Long 7yr Duration 10Yr 15Yr Corporate 20Yr 25YrBond 30Yr 1Yr 2Yr 3Yr 5Yr 7yr 10Yr 15Yr 20Yr 25Yr 30Yr iability r 2Yr 3Yr 5Yr 7yr 10Yr 15Yr 20Yr 25Yr 30Yr lity Yr 3Yr 5Yr 7yr 10Yr 15Yr 20Yr 25Yr 30Yr Yr 5Yr 7yr 10Yr 15Yr 20Yr 25Yr 30Yr 7yr 10Yr 15Yr 20Yr 25Yr 30Yr uration Corporate Bond A possible solution to address this mismatch is to buy long duration bonds, such as the DEX Long Term Bond Indices. Chart 3 shows how adding long bonds would indeed improve the duration match, as the current liability duration is 13.4 years and the asset duration is 1 years, giving an effective hedge of 93.3%. However, sizeable mismatch continues to appear, especially near the long end. The implementation in this example is based upon public market benchmarks, not the unique benefit payment stream of the pension. If the pension fund buys bonds that are different from the public market index, tracking error occurs, which will either yield alpha or lead to loss, relative to the public market benchmark. The challenge is that the true benchmark is the pension plans distribution of key rate durations, not that of the public market benchmark. Chart 3: 100% LDI Implementation Asset Duration 1 Years + Duration 13.4 Years =Effective Duration 93.3% Funded Ratio Volatility 4.7% Funded Ratio (90% Confidence Interval) 91.1-106% Customization to Minimize Tracking Error A fully customized LDI implementation would fill the remaining gaps. An overlay of STRIPS or interest rate swaps can help to create coverage across the entire curve as a result of their liquidity. In Chart 4, an overlay of STRIPS more closely matches the effective duration, and also reduces funded ratio volatility and the range in the funded ratio confidence interval. Chart 4: Customized Portfolio and STRIPS LDI Implementation In this case a diverse fixed income strategy, coupled with an overlay of STRIPS, still does not guarantee a perfect hedge. In tough markets, the sample pension plan has a 92.8% funded ratio, but only 103.9% in good times. Despite best efforts, liability-matching is an imperfect process. Thus, while a skilfully customized LDI implementation goes a long way to reduce risk and volatility in the pension plan, risk exposure continues to exist due to the construction methodology of the liabilities discussed earlier. This is commonly known as basis risk, or the difference between what is used to build the discount rate and what is used in the LDI implementation. Plan sponsors should also be cognizant that mismatch is not solely dependent upon the liability, but also the correlation and impact of the non-matching portfolio as well. 3

Active Management at the Glidepath Level A successful LDI implementation requires active management at two levels the glidepath and the fixed income portfolio. Discussed earlier, the glidepath is in essence an active approach to asset allocation which establishes a predetermined set of investment policies that gradually become more conservative, where changes are driven by the pension plan s funding position. As the plan meets improving funding thresholds, the portfolio begins to take on a more liability-matched flavour. As triggers are hit and the plan moves along the glidepath, it is critical for plan sponsors to actively assess current market conditions to ensure the pre-determined asset mixes continue to be relevant and appropriate. Glidepath allocations are economically dependent. Where volatility is the new normal in our markets, or where organizational goals or circumstances change, the current glidepath and allocation strategy may no longer meet the plan s current hurdle rate and thus either require more in contributions, or longer periods of outperformance to catch back up. Thus consideration should be given to re-risking should markets and other factors render the plan s pre-determined allocation unable to meet pension goals. Understanding the impact of the non-matching portfolio, as well as the components of the LDI mismatch, allows plan sponsors to decide where along the glide path they should be, and whether the glide path should allow rerisking. As plan sponsors approach a fully funded status, the triggers of an LDI strategy should be managed in an active and customized manner in order to properly match the liability stream for the particular plan. Active Management at the Portfolio Level An active LDI strategy monitors, anticipates and adjusts fixed income investments in order to reduce the risk of tracking error and investment loss due to a downgrade in bonds or a material change in credit spreads. Focusing on the three major drivers of liability mismatch duration, credit exposure and curve risk allows plan sponsors to more fully understand their hedge and the performance versus liability changes. Underperformance can occur when a bond in the discount rate universe is downgraded and removed from consideration. When a bond begins to deteriorate, the yield goes up and the spread starts to widen. Once a bond is downgraded and removed from the liability discount rate index, the removal of the high-yielding bond causes the discount rate to go down, and liabilities to go up. If a downgraded bond was held in an LDI portfolio, the pension plan would be forced to recognize an investment loss. Not only would the pension liabilities increase, but the plan would also see a mark-to-market investment loss. To illustrate, recall the first half of 2012 when discount rates declined in the U.S. During this time, interest rates declined slightly, but the impact on discount rates was far more pronounced due to the large number of downgraded U.S. banks that no longer qualified for inclusion in the discount rate universe. During the period prior to the downgrades, spreads were widening, and in June 2012, JP Morgan Chase, Credit Suisse, Barclays, UBS and Deutsche Bank (which were all AA-rated) were downgraded and consequently removed from pension discount rate indices. The impact of this removal caused overall discount rates to fall, as a large number of higher yielding bonds in the discount rate universe were removed. By the end of June 2012, the Barclays Capital Long Corporate Index yield dropped 27 basis points. This reflects both the Index s composition and the yields for a typical pension plan discounted against the Index during this time period. By contrast, a plan discounted against the Citigroup Pension Index would have gone down by almost 40 basis points during the same time period. This Index saw a larger decrease because it contained bonds issued by the previously mentioned five banks, which were downgraded and removed. An active LDI manager who would be closely tracking the portfolio would have noticed the risks and avoided holding these particular issues. What does this actually equate to in numbers? If a plan sponsor owned the Barclays Capital Long Corporate Index during this timeframe, it generated returns of 5.4%, while liabilities increased by 6.2%. If the pension plan had a duration of 12, this underperformance of 80 basis points would have decreased the funded status by almost 10%. This is why actively managing all aspects of the liability hedge is important. Just matching the duration, while ignoring the curve match or spread risk, will eventually lead to underperformance of the hedge. 4

Need for an Active Viewpoint of the Markets Presumably, constructing the portfolio of an active LDI strategy seems relatively basic: How many and what type of bonds should be held? And how should the three critical factors (duration, credit exposure, and curve risk) be weighted? The answer depends on the organization s active viewpoint of the market. Does the organization want to own A- rated bonds because it thinks spreads are wide? Or does it want to get out of A s and BBB s because spreads are narrow? There are periods when owning Federal bonds will outperform liabilities because spreads are widening. There are periods when owning long corporate bonds will outperform when spreads are tightening. When spreads are widening or compressing, active LDI management adjusts the composition of the overall LDI strategy. Need for Diversification Diversification is also important in an active LDI strategy. By using multiple investment sources with different philosophies within a specific asset class, a plan sponsor can reduce risk for a given level of excess return (alpha) and achieve maximum diversification. Simply meeting the liability index is not sufficient; the plan needs to outperform to account for the favourable selection bias inherent in the discount rate. When creating a fully hedged strategy, it might be beneficial to add small amounts of high-yield bonds or equity. This can generate outperformance to make up for times when spreads are compressing, when the discount rate is artificially low due to higher yielding bonds being removed, or other construction issues. Measuring Success: Benchmark Against Liabilities Once a plan has implemented a custom LDI strategy with ongoing active management, the next natural question is how is success measured? The true benchmark lies in the valuation of liabilities. The goal of an LDI strategy should be to meet or exceed changes in liability valuations, while correlating asset and liability returns over time. Constant monitoring is necessary to ensure that the assumptions used are both accurate and the best representation of the present liability. At the investment manager level, the traditional benchmark is against performance in the form of public benchmark returns. In the case of fixed income managers, performance is benchmarked against plan liabilities, rather than fund or market-based performance. While not all investment managers agree to be benchmarked in this way, it is a growing trend worth exploring for many plan sponsors. As such, the LDI strategy is not only benchmarked against liabilities at the portfolio level, but at the manager level as well. Conclusion An effective LDI framework begins with a thorough understanding of the pension plan s assets and liabilities, followed by a diversified, risk-managed fixed income implementation. Diversification is particularly important in the fixed income space multiple fixed income strategies, as well as multiple managers, are needed within each strategy. This helps to manage risks at the security level, manager level and fund level. A custom overlay of STRIPS or swaps fill in the gaps where overweight or underweight, in order for plan assets and liabilities to look and act very much alike. To obtain the best match possible, plan sponsors should implement an active strategy that monitors, anticipates and adjusts fixed income strategies in an effort to reduce the risk of a liability mismatch and investment loss. An additional allocation to high-yield bonds and equities can help to close any remaining gap that might exist if the opportunity to avoid a bond downgrade is missed. In short, the most effective LDI strategy needs to be customized to a plan s unique goals and liabilities and managed in an active way with an established point of view. For information on SEI and how the company helps clients implement LDI strategies, visit www.seic.com, or contact Michael Chwalka at mchwalka@seic.com. This information should not be relied upon by the reader as research or investment advice. This information is for educational purposes only. Information provided by SEI Investments Canada Company, a wholly owned subsidiary of SEI Investments Company. 5