LDI: Complex Problems, Practical Solutions

May 16, 2017

Printable Version

Each plan sponsor’s LDI journey is unique and, along the way, may require a number of benchmark changes to adequately hedge liabilities. At each decision point, sponsors must choose between standard or more customized solutions. At IR+M, we strive to present sponsors with a range of options, always mindful of the costs of complexity versus the value it adds.

UNDERSTANDING THE RISKS

  • Key Market Risks: Market risks typically impact corporate pension liabilities by affecting three main factors – duration, credit, and curve.
        • Duration – the impact of a parallel shift in the yield curve that is used to discount pension liabilities.
        • Credit – the impact of changing credit spreads in the yield curve that is used to value pension liabilities.
        • Curve – the impact of non-parallel shifts in the yield curve used to value pension liabilities
  • Ranking the Risks: For those plans that are in the earlier stages of their LDI journey, duration and credit risk are frequently the most important drivers of volatility. We find that, in the current low rate environment, duration risk is roughly twice as impactful as credit spread risk. Plan sponsors should begin to focus on hedging curve risk in the later stages of their LDI journey

  • Working Harder (Early-Stage LDI): Plan sponsors in the early stages of their LDI journey, typically have significant equity exposure and a limited fixed income portfolio serving as a liability hedge. At this stage, sponsors often focus on adding duration or credit exposure. They endeavor to make their fixed income allocation work “harder.”
  • Working Smarter (Late-Stage LDI): As their fixed income portfolio grows, plan sponsors in the later stages often employ more complex solutions in an effort to address the specific curve exposures or unique credit profile of the plan liabilities. Their emphasis is on making their fixed income assets work “smarter.”

MAKING YOUR PORTFOLIO WORK HARDER

  • A Hedging Misconception: Early-stage LDI clients often believe, perhaps incorrectly, that “LDI” requires the hedging assets to closely resemble the liabilities. As a result, some plan sponsors use custom benchmarks to better align their fixed income and liability durations. While this is an important step for late-stage clients, it can be counter-productive for early-stage ones.
  • Thinking in Dollars: The failure to account for underfunding and/or low allocations to hedging assets can be costly. While the full value of a sponsor’s liability is often exposed to rate movements, only those assets that are allocated to the hedging portfolio are affected by rate changes. To truly evaluate duration risk, a plan sponsor may wish to consider the dollar exposure on both the asset and liability sides.
  • The Hedge Ratio: We believe that it is important for plan sponsors to think in terms of hedge ratios, as opposed to duration targets. The hedge ratio is the sensitivity of liabilities to parallel interest rate moves relative to the sensitivity of assets to the same move.

  • Extending Beyond The Liability: If the sponsor’s objective is to hedge additional duration risk without committing additional assets to the hedging portfolio, the sponsor should extend the duration of the hedging assets beyond that of the liabilities. In other words, the sponsor needs to make the hedging assets work harder.

MAINTAINING THE HEDGE

  • Rebalancing Duration – How Often? When the plan is well-funded and has a high allocation to hedging assets, there may be significant value in aligning the duration of those assets with that of the liabilities. Once this has been accomplished, how often should the duration target of the hedging assets relative to that of the liabilities be revisited?
  • The Answer May Lie With Convexity: The frequency of rebalancing is a function of the relative convexity (the rate of change of duration with interest rates) of the assets and liabilities. When rates rise/fall, durations will shorten/extend. Convexity measures the extent of these movements.
  • Fixed Duration Targets Require More Frequent Rebalancing: Pension liabilities tend to have high convexity due to the long tails of their payment distributions. If the hedging assets target a specific duration, the fixed income portfolio will have low convexity, especially if it is repeatedly aligned with its target. Based on this methodology, we believe that if a fixed duration target is used, then the liability duration should be measured often, and the fixed income target adjusted accordingly.
  • Blended Indices Reduce Rebalancing: As an alternative to frequent and often costly rebalancing, plan sponsors may target their liabilities by utilizing a blend of benchmarks. Between rebalancing periods, the convexity of the blended benchmarks should broadly capture the duration change of the liabilities. As a result, annual updates to blends are often sufficient.

CUSTOMIZATION IN CREDIT HEDGING

  • Liability Discount Rates Impact Hedging: Most corporate defined benefit plans calculate their liabilities using discount rates based on high-quality corporate bonds. Specifically, accounting liabilities are typically valued using the yield on AA-rated corporate bonds. As a result, liability changes are highly correlated with movements in high-quality long corporate spreads.
  • An Imperfect Choice: Plan sponsors often seek to maintain a corporate spread exposure that is similar to that of their plan liabilities. This helps insulate funded status volatility from corporate spread movements. The decision of what fixed income benchmark to use as a hedge against corporate spread exposure can be a difficult one. We believe the options of using a concentrated AA universe used to value liabilities, or a broader benchmark that introduces a degree of basis risk, are both imperfect.
  • Credit Ratings Can be Misleading: While weighing the merits of incorporating quality restrictions, plan sponsors may become too focused on ratings. Ratings alone are not indicative of the suitability of a particular issue in a client’s hedging portfolio. As the matrix suggests, correlations between rating buckets are so strong that they may be misleading.
  • Not all BBB’s Are Create Equal: The BBB-rated corporate universe contains numerous high-quality bonds that have strong credit fundamentals. However, as a whole, the universe may exhibit volatility. Using an A or better benchmark may be more appropriate if short-term volatility is a key concern, such as for plans that are approaching a buy-out or pension risk transfer. For those sponsors that have significant volatility due to equity allocations, using a broader benchmark may provide active managers with more tools to protect funded status while adding negligible additional volatility.

FINE TUNING THE HEDGE

  • Late Stage LDI Clients Focus On Remaining Risks: Many plan sponsors are able to sufficiently mitigate funded status volatility by using standard benchmarks to hedge the duration and credit components of their liabilities. Plan sponsors that are sensitive to small movements in funded status may seek to hedge the remaining curve risk that is associated with plan liabilities.
  • There is No Perfect Hedge: There is no ideal way to hedge the curve risk that is associated with pension liabilities. Future cashflow projections are derived from underlying actuarial assumptions, which are typically revised on an annual basis. As a result, the accuracy of the resulting curve risk hedge is only as good as the accuracy of the actuarial assumptions.
  • Minimizing Curve Risk: Plan sponsors will often attempt to align the cashflows from the hedging portfolio (i.e. coupon and maturity payments) with the expected payouts from the pension plan. By aligning the cashflows at each maturity, the plan sponsor should theoretically be able to make the plan relatively immune to changes in the shape of the yield curve used to discount the liabilities.
  • The Opportunity Cost of Precision: Cashflow matching is not without costs. The expenses are derived from managing to a custom benchmark, as well as from restricting managers from placing their best ideas in the hedging portfolio. These additional limitations can prevent a manager from adding alpha and maintaining pace with downgrade-immune pension liabilities.
  • Stepping Outside The Cashflow Model: While a cashflow matching approach may specify allocations to each maturity, the reality is that certain points along the yield curve can be extremely correlated. In particular, longer-maturity bonds often exhibit high correlations due to trading conventions. Managers may exploit these correlations in an attempt to expand their opportunity set while still constructing effective hedging portfolios. The chart below displays the significantly higher correlation between long-duration Treasury bonds versus shorter-duration ones. Plan sponsors may find that less precision is required to hedge longer-dated cashflows than shorter-duration payments.
  • Key-Rate Hedging: We believe key-rate hedging, which measures the liability’s curve exposure at select points, is a more cost-effective approach to curve hedging. Managers can construct a hedging portfolio that replicates the key-rate exposures of the plan’s liabilities, thereby reducing the curve-risk of non-parallel movements in the yield curve used to value the liabilities.
  • Choosing the Key-Rate Buckets: It is customary to choose points closer together at the front of the yield curve, where correlations are lower. The points tend to be more dispersed as correlations improve in the long-end of the curve. Market convention is to consider five or six key rates, typically at the 0.5, 2, 5, 10, 20, and 30-year points of the curve. However, as the chart to the right shows, the 20-year and 30-year points tend to be well-aligned.

JUDGING THE VALUE-ADD OF COMPLEXITY

  • The Custom Decision: We have discussed just a few of the decisions that plan sponsors face as they consider the appropriate benchmark for their hedging portfolios. There are many additional points at which sponsors will debate the use of a standard benchmark or a more custom option. While the answer will often be specific to the plan sponsor, the decision may weigh similar questions.
  • Does the custom benchmark materially improve the liability hedge? As discussed earlier, the benefits of customizing at an early stage can often be smaller than expected. In the later stages, custom solutions will often be required to hedge curve risk, but sponsors should still be wary of trying to be overly accurate.
  • What is the cost of customization? Understanding the costs of customization can be tricky, as sponsors need to consider both the direct costs, such as higher fees for complex solutions, and the opportunity costs, such as preventing managers’ best ideas from making it into the portfolio.
  • How do we monitor custom performance? Finally, an additional consideration is the ability of plan sponsors to assess an LDI manager’s performance versus a highly custom benchmark. While the ultimate custom benchmark for any LDI manager is the plan’s liability, there are drivers of liability growth that remain unhedgable, even for the most sophisticated LDI manager. We believe an understanding of the nuances, and thorough attribution of returns, is key to reviewing and refining the benchmark decision.

At IR+M, we believe in providing straightforward practical solutions to the complicated problems of liability hedging. Our experienced LDI team works in tandem with plan sponsors to evaluate their hedging goals, and recommend fixed income strategies for each stage of their LDI journey.

¹Source: Bloomberg Barclays ²Source: IR+M Internal Analytics, Bloomberg Barclays Live and Citi Group Pension Discount Curve ³Source: IR+M Internal Analytics, Bloomberg Barclays Live

The views contained in this report are those of IR+M and are based on information obtained by IR+M from sources that are believed to be reliable. This report is for informational purposes only and is not intended to provide specific advice, recommendations for, or projected returns of any particular IR+M product. No part of this material may be reproduced in any form, or referred to in any other publication, without express written permission from Income Research & Management. The illustrations in this report do not constitute an actuarial opinion and should not be used beyond the scope of this report.