Why we need to keep talking about LDI—even in a low-rate environment
It might seem odd to talk about buying long-dated bonds at a time when the 30-year bond yield is getting close to a record low. But it’s essential for pension plans to understand how their assets are performing relative to liabilities, and that understanding lies at the heart of liability-driven investing (LDI).
LDI is investing the portfolio in a way designed to pay pension and benefits to plan members when they retire. Although the largest pension plans in Canada have embraced it, uptake on LDI has been relatively limited, and falling interest rates haven’t helped move the conversation forward.
Traditionally, LDI refers solely to the fixed income portion of the portfolio. But this perspective should be broader and take into account the impact of other asset classes driving returns, such as equities and real estate. An LDI framework also means getting a better understanding of the risk of rising deficits or contributions and deciding whether or not these risks are worth taking.
Here are six aspects of an LDI strategy plan sponsors need to consider.
1 | Hedging Interest Rate Risk As rates go down, so does the level of rates used to discount liabilities, which drives them up. If assets don’t go up by the same amount, there is a mismatch, and the funding position worsens.
The current low level of interest rates is often given as a reason for only hedging a minimal amount of this risk. The rationale is, as rates increase, the liabilities will go down by a larger amount than the assets. In practice, however, a large number of market participants expect yields to rise—a factor already priced into long-term yield. A pension plan would only benefit from staying away from long bonds if long-term yields rise by more than expected.
The other key consideration is the size of the underweight on long bonds relative to liabilities. Most plans have hedged only a small portion of their interest rate risk, and their assets would therefore significantly outperform their liabilities if rates were to go up (everything else being equal), as the liabilities would fall by a larger amount than the assets.
2 | Picking a Benchmark The obvious choice for a benchmark is pension liabilities, as meeting those is the ultimate objective of LDI, but there are complications. First, there isn’t a single way to discount liabilities; different methodologies can be used (accounting, going concern, solvency, economic, etc.) depending on the plan’s objectives.
Second, liabilities do not comprise an investable benchmark; it’s impossible to buy securities that will exactly match a stream of liabilities. For example, the solvency liability is discounted using federal bond yields plus spreads that change every quarter, based on market surveys by the Canadian Institute of Actuaries. This is different from a traditional benchmark, whose performance is simply a weighted average of the underlying constituents. Also, liabilities are estimates that will change over time.
A simple solution is to use a long bond benchmark (such as the FTSE TMX Canada Long Term Index)—a transparent benchmark that is independently calculated and easily accessible. The main drawback, however, is the benchmark was never designed to hedge liabilities; it simply reflects where borrowers are issuing long-term debt. It may be acceptable as an approximation for a plan investing a small portion of its assets in fixed income, but as this amount increases, it’s essential to consider a more accurate approach.
Read: Lessons learned in LDI
3 | Selecting a Manager Choosing an investment manager is an important part of implementing an LDI mandate. An LDI mandate is tailored to the specific pension plan and therefore requires a good understanding between the client and the manager to adapt the solution to the plan’s specific needs. When considering an LDI fund manager, look for the following characteristics.
Stability – An LDI mandate is a long-term investment, and it can be disruptive if there are changes at the manager level. The manager should be committed to LDI and have a large enough asset base in the space to make it viable in the long run.
Experienced investment personnel – Setting up and managing an LDI mandate requires skills outside of traditional fixed income management. A good understanding of the liabilities and the different discounting curves is essential. It also helps if the team is managing a wide range of LDI mandates, as they can gain experience from solutions developed for other clients. A dedicated team is an asset, but it is not a requirement; there are also benefits with an integrated team working closely with other fixed income portfolio managers, particularly if the mandate is more on the active side. A team with a strong depth of skills ensures resilience to staff turnover.
Flexible and robust investment process – The management firm should be able to customize its process to the specific plan’s requirements. It also needs robust portfolio management systems to model, adjust and monitor the performance of the portfolio versus the liabilities.
Innovative risk systems – In Canada, there are currently no comprehensive off-the-shelf risk systems designed to manage a portfolio against liabilities. Managers have generally decided to build their own systems, and the complexity and features vary significantly. If LDI is an area of focus for the manager, it helps in justifying a continuing investment in improving LDI risk systems.
4 | Setting the Glide Path Establishing a glide path based on triggers is a good way to implement a phased approach to reducing interest rate risk over time. It may be advantageous to hold a larger portion of growth assets when the plan is underfunded, since asset returns may help reduce any deficit and finance service accruals. Once the plan is closer to fully funded, the allocation to growth assets should most likely be reduced. A dynamic investment strategy ensures discipline in implementation by setting clear instructions. It also gives the portfolio the ability to take advantage of short-term moves in interest rates: if rates increase for a few days going through a trigger and then fall back down, having a glide path means the plan won’t miss the opportunity. This is important, as many pension plans will likely buy a large number of long bonds when rates go up.
The simplest trigger is based on yield level, where exposure to long bonds is increased by a preset amount if rates reach a set level (e.g., if the 30-year federal yield is 3.5%). A trigger based on funded status has the benefit of taking into account the performance of non-fixed income asset classes and moves the plan toward a goal of being fully funded. Finally, a trigger based on time guarantees the plan implements, at least partly, an LDI solution.
The plan should also ensure it has enough liquidity to implement the required steps if it has investments in less-liquid assets such as real estate, infrastructure or private equity. An interest rate overlay can add flexibility by delaying the need for selling growth assets.
Read: LDI keeps HOOPP positive
5 | Going Active or Passive An LDI strategy can take varying degrees of active risk, but it’s not possible to have a fully passive approach, as liabilities are not fully replicable. The degree of active management will vary depending on how much discretion the fund manager has to outperform liabilities. It makes sense to focus most of the active risk on the corporate bond portion of the portfolio, rather than on federal/provincial bonds, as this is where the widest dispersion of returns exists.
Also, ensure the portfolio is diversified. The Canadian long corporate bond market is concentrated: the two largest sectors (utilities and energy) represent more than 40% of the index, and the top 10 issuers comprise 39% of the index.
6 | Choosing Pooled or Segregated Larger plans have the choice of investing in a segregated account or in a pooled fund, while smaller plans will likely find a pooled fund solution the best fit for cost and diversification reasons.
A segregated account offers flexibility in solution design, including client- specific risk constraints, as well as somewhat better matching of liabilities. A pooled fund solution is more cost-effective and simpler to implement, as it doesn’t require a custodian. For plans wanting to use an overlay strategy, in which a plan hedges a larger portion of the liabilities than the amount it has invested in hedging assets, this also removes the need for lengthy negotiations with banks to sign derivatives agreements.
The two options are not mutually exclusive: a client can invest in a pooled fund initially, when the hedge ratio is low, and then move to a segregated solution as LDI assets increase. An in-kind transfer of securities will limit transaction costs when moving from one option to the other.
Ultimately, it’s important to be clear about your LDI objectives. A plan sponsor may want to minimize accounting volatility or offload the liability to an insurance company, or it might simply want to minimize the cost of running the plan.
Interest rates may be low today, but they won’t stay low forever. If you take the time to review LDI options and think about your course of action, you will be able to react quickly when opportunities arise—as they inevitably will.
Erwan Pirou is Canadian head of fixed income research for Aon Hewitt in Toronto.
Get a PDF of this article.